Tài liệu Báo cáo khoa học: Structural and functional investigations of Ureaplasma parvum UMP kinase – a potential antibacterial drug target ppt

Structural and functional investigations of Ureaplasma
parvum UMP kinase – a potential antibacterial drug target
Louise Egeblad-Welin
1
, Martin Welin
2,
*, Liya Wang
1
and Staffan Eriksson
1
1 Department of Anatomy, Physiology and Biochemistry, Swedish University of Agricultural Sciences, Uppsala Biomedical Centre, Sweden
2 Department of Molecular Biology, Swedish University of Agricultural Sciences, Uppsala Biomedical Centre, Sweden
Ureaplasma parvum belongs to the class Mollicutes,
which have the smallest genomes known in any free-
living organisms, and a very low G + C content [1]. It
is a human pathogen that normally colonizes the
urogenital tract, where it is involved in a variety of
diseases such as urethritis and prostatitis. During
pregnancy, it is an opportunistic pathogen and can
cause spontaneous abortions and premature birth. The
bacteria can be transferred vertically from mother to
child during birth, and give rise to meningitis and
pneumoniae in newborns [2].
U. parvum uridine monophosphate kinase
(UpUMPK) (EC 2.7.4.22), coded by the PyrH gene,
catalyses the reversible phosphorylation of uridine
monophosphate (UMP) using a nucleoside triphosphate
(NTP) as phosphate donor [3,4]. It has been cloned and
the recombinant enzyme characterized [4]. UpUMPK
has high sequence identity to other UMPKs from bacte-
ria and archaea (Fig. 1); the sequence identity to UMP-

Ks from Escherichia coli, Pyrococcus furiosus, Sulfolobus
solfataricus, Haemophilus influenzae and Streptococcus
pneumoniae is 34, 26, 25, 34 and 38%, respectively. The
Keywords
bacterial UMP kinase; Mollicutes;
mycoplasma; subunit interaction;
Ureaplasma parvum
Correspondence
S. Eriksson, Department of Anatomy,
Physiology and Biochemistry, Swedish
University of Agricultural Sciences, Box 575,
Biomedical Center, S-751 23 Uppsala,
Sweden
Fax: +46 18550762
Tel: +46 184714187
Email:
*Present address
Structural Genomics Consortium, Karolinska
Institutet, Stockholm, Sweden
Database
The structure has been submitted to the
Protein Data Bank under the accession
number 2va1
(Received 29 June 2007, revised 19 October
2007, accepted 22 October 2007)
doi:10.1111/j.1742-4658.2007.06157.x
The crystal structure of uridine monophosphate kinase (UMP kinase,
UMPK) from the opportunistic pathogen Ureaplasma parvum was deter-
mined and showed similar three-dimensional fold as other bacterial and
archaeal UMPKs that all belong to the amino acid kinase family. Recom-

binant UpUMPK exhibited Michaelis–Menten kinetics with UMP, with K
m
and V
max
values of 214 ± 4 lm and 262 ± 24 lmolÆmin
)1
Æmg
)1
, respec-
tively, but with ATP as variable substrate the kinetic analysis showed posi-
tive cooperativity, with an n value of 1.5 ± 0.1. The end-product UTP was
a competitive inhibitor against UMP and a noncompetitive inhibitor
towards ATP. Unlike UMPKs from other bacteria, which are activated by
GTP, GTP had no detectable effect on UpUMPK activity. An attempt to
create a GTP-activated enzyme was made using site-directed mutagenesis.
The mutant enzyme F133N (F133 corresponds to the residue in Escherichia
coli that is involved in GTP activation), with F133A as a control, were
expressed, purified and characterized. Both enzymes exhibited negative coo-
perativity with UMP, and GTP had no effect on enzyme activity, demon-
strating that F133 is involved in subunit interactions but apparently not in
GTP activation. The physiological role of UpUMPK in bacterial nucleic
acid synthesis and its potential as target for development of antimicrobial
agents are discussed.
Abbreviations
NDPK, nucleoside diphosphate kinase; UMPK, uridine monophosphate kinase.
FEBS Journal 274 (2007) 6403–6414 ª 2007 The Authors Journal compilation ª 2007 FEBS 6403
primary sequence and crystal structures of E. coli
UMPK and the archaea P. furiosus and S. solfataricus
UMPKs showed that these enzymes belong to the amino
acid kinase family [5–8]. In eukaryotic cells, the corre-

sponding enzyme is CMP-UMPK (EC 2.7.4.14), which
is a member of the nucleoside monophosphate kinase
family [9].
Studies with Mycoplasma genitalium using transpo-
son mutagenesis showed that UMPK is essential for
the survival of the organism [10,11]. The UMPKs
(PyrH genes) of E. coli, H. influenzae and St. pneumo-
niae have also been shown to be essential [12–14].
Therefore, UMPK is a potential drug target for the
development of antimicrobial agents, and it is of great
importance to study the structure and function of these
enzymes.
E. coli, Bacillus subtilis and St. pneumoniae UMPKs
are hexamers and are activated by GTP, inhibited by
UTP, and show Michaelis–Menten kinetics with
UMP. Furthermore, St. pneumoniae UMPK displays
positive cooperativity with ATP [5,14–16], as was also
recently shown with the enzymes from other Gram-
positive bacteria, e.g. B. subtilis and Staphylococcus
aureus [17]. In E. coli, UMPK residues Arg62, Asp77,
Thr138 and Asn140 have been suggested to be
involved in the interaction with GTP, as mutation of
these residues abolished the GTP activation [6,18].
However, S. solfataricus UMPK is not activated by
Bacteria specific loop
Archaea specific
loop
Fig. 1. Sequence alignment of UMPKs. Accession numbers: H. influenzae, P43890; E. coli, P0A7E9; N. meningitidis, P65931; St. pneumo-
niae, Q97R83; Streptococcus pyogenes, P65938; B. subtilis, O31749; U. parvum, Q9PPX6; S. solfataricus, Q97ZE2; P. furiosus, Q8U122.
Secondary elements for U. parvum UMPK are listed above the alignments. Completely conserved residues are colored in red and similar res-

idues in yellow.
UMP kinase from Ureaplasma parvum L. Egeblad-Welin et al.
6404 FEBS Journal 274 (2007) 6403–6414 ª 2007 The Authors Journal compilation ª 2007 FEBS
GTP, which was previously suggested to be specific
to archaeal UMPKs [8].
In this study, recombinant UMPK from U. parvum
was enzymatically characterized, particular with regard
to the substrates UMP and ATP, the inhibitor UTP
and the potential activator GTP. The crystal structure
was determined by X-ray crystallography in complex
with a phosphate ion. A cross-talk region between two
subunits of UpUMPK was identified, which corre-
sponded to the region in E. coli UMPK that contains
the key residues Thr138 and Asn140 that are involved
in GTP activation. Residue Phe133 of UpUMPK (cor-
responding to Asn140 in E. coli UMPK, Fig. 1) was
mutated to either Asn or Ala, and the resulting mutant
enzymes were characterized.
Results
Overall structure
The structure of the UpUMPK was determined by
X-ray crystallography to a resolution of 2.5 A
˚
with a
final R value of 23.3% and R
free
of 28.5% (Table 1).
The enzyme is a hexamer composed of three dimers
that are related by threefold symmetry (Fig. 2A). The
monomer subunit consists of an a ⁄ b-fold with a nine-

stranded twisted b-sheet surrounded by eight a-helices
and one 3
10
helix. (Fig. 2B). The monomers are simi-
lar, and when each subunit (B, C, D, E and F) is
superpositioned on subunit A, the rmsd varies between
0.169 and 0.419 A
˚
. A flexible loop between b5 and b6
(amino acids 166–175) can only be traced in the elec-
tron density for the E subunit, with B factors of
approximately 50 A
˚
2
. In UMPK from P. furiosus, this
loop is responsible for binding the adenine base of an
ATP analogue (Protein Data Bank accession number
Table 1. Data collection and refinement statistics. Values in paren-
theses refer to the data in the highest-resolution shell. ESRF,
European synchrotron radiation facility.
Parameter Value
Space group P2
1
Cell dimensions (A
˚
,°) a, 79.8
b, 96.6
c, 96.3
b, 105.8
Content of asymmetric unit One hexamer

Resolution (A
˚
) 33.4–2.5 (2.64)
Completeness (%) 99.9 (99.9)
R
meas
(%) 11.7 (41.5)
I ⁄ rI 11.4 (3.1)
Redundancy 3.8
Number of observed reflections 182 623
Number of unique reflections 48 558
Beam line ESRF, ID14 eh4
Wavelength (A
˚
) 0.976
Temperature (K) 100
R (%) 23.3
R
free
(%) 28.5
rmsds
Bond length (A
˚
) 0.007
Bond angle (°) 1.04
Mean B value (A
˚
2
) 34.7
α2

A
B
α3
α1
α8
α6
α7
α5
α4
β4
β2
β1
β6
β9
β8
β7
β5
η1

β3

Fig. 2. (A) UpUMPK as a hexamer; a phosphate ion is seen in every
monomer. The dimeric couples A + B, C + D and E + F are colored
green, pink and blue, respectively. (B) The monomer of UpUMPK
in complex with a phosphate ion. The flexible loop that was only
observed in one subunit is colored in orange.
L. Egeblad-Welin et al. UMP kinase from Ureaplasma parvum
FEBS Journal 274 (2007) 6403–6414 ª 2007 The Authors Journal compilation ª 2007 FEBS 6405
2BRI), and, as no ATP ⁄ ATP analogue is bound to the
enzyme, the loop is not held in a tight position. The

forces that hold the hexamer together are (a) hydro-
phobic interactions between the dimeric couples
(A + B, C + D and E + F), (b) a few hydrogen
bonds, and a hydrophobic interaction between A + C,
B + E and D + F, and (c) electrostatic forces in the
central channel of the hexamer between B, C and F,
and A, D and E. The hydrophobic interactions
between A and B (Fig. 3A,B) are formed between the
antiparallel a3-helices from each subunit, primarily by
Leu, Met and Ile. One hydrogen bond could be identi-
fied at each end of the interacting a-helices between
Asn86 and the carbonyl carbon of Leu62. Between A
and C, the a7 from each subunit is connected via
two hydrogen bonds between Thr197 and Glu204
(Fig. 3A,C), and a hydrophobic interaction between
Thr131 and Phe133 (Fig. 3D). The central channel of
the hexamer is made up of two layers of electrostatic
forces on top of each other, with one layer rotated by
60°. The amino acids found in the electrostatic hole in
each layer are Lys102 and Asp104. Lys102 is held in
position by Asp104, and, in the interaction between A,
D and E, a water molecule is hydrogen-bonded to the
three lysines (Fig. 3E). There is no water molecule fix-
ing the three lysines from the B, C and F subunits,
and there is no direct interaction between subunits A
and F, B and D, or C and E.
A phosphate ion was found in the donor site of all
subunits (Fig. 2B), although the protein was crystal-
lized in the presence of 5 mm GTP. The B factors for
the phosphate ions in the subunits varied from 46–

55 A
˚
2
. Structural alignments with UMPK from E. coli
(Protein Data Bank accession numbers 2BND and
2BNF) indicated that the phosphate ion in UpUMPK
was bound in the position corresponding to the
b-phosphate of either UDP or UTP. Combinations of
co-crystallization and soaking were performed in
order to bind either substrate or inhibitor. However,
in all collected data sets (UpUMPK co-crystallized
with UMP, and UTP or with no added ligand), a
phosphate ion was detected in the active site. This
indicates that phosphate ions were bound to the
enzyme during expression and purification, since no
phosphate buffers were used during the preparation
procedures.
As no structural data for UpUMPK with UMP
bound were obtained, the binding of UMP to the
active site was modeled by structural alignment with
UMPK from E. coli with UMP bound [6], which gave
an rmsd of 1.29 A
˚
for 212 Ca-atoms. The binding of
UMP in E. coli creates a tightly closed conformation
with a2 (Fig. 4A). In UpUMPK, this a-helix has a
more open conformation in the absence of UMP
(Fig. 4A). The amino acid residues responsible for
binding of UMP are relatively conserved (Fig. 4B); the
only difference is a Phe133 found in the position corre-

sponding to Asn140 in E. coli. The probable binding
motif for the uracil base is through hydrogen bonds
from N3 to the backbone O of Phe133, and from O4
to the backbone N and the side chain of Thr131, with
the ribose moiety anchored by two hydrogen bonds,
one from 2¢-OH to the side chain of Asp70, and one
from the backbone N of Gly63, and the phosphate
group forming hydrogen bonds from O1 to Arg55,
from O2 to the backbone N and the side chain of
Thr138, and from O3 to the backbone N of Gly50. A
P-loop (GXXXXGKS ⁄ T) that is usually found in
nucleotide binding enzymes is not present in UpUMPK
[19]. UpUMPK contains instead a glycine-rich motif
within amino acids 44–54 that is responsible for bind-
ing of the phosphate ion. These amino acids are rela-
tively conserved among the UMPKs, with an amino
acid sequence motif as follows: V ⁄ IXV ⁄ IXGGGXXXR
(Fig. 1).
Functional characterization
The substrate specificity of purified recombinant
UpUMPK was explored using a coupled spectrophoto-
metric assay [20], and several ribonucleoside mono-
phoshates and deoxyribonucleoside monophosphates
were tested as phosphate acceptors with ATP as phos-
phate donor. The only effective acceptor was found to
be UMP, and the pH optimum of the reaction was
6.8. It was also observed that a stoichiometry between
Mg
2+
and ATP of 2 : 1 gave 1.6-fold higher catalytic

rates compared to a 1 : 1 stoichiometry (data not
shown). Therefore, the Mg
2+
:ATP ratio was kept at
2 : 1 in all further experiments, in analogy with other
UMPK studies [14,17].
Two-substrate kinetics with UMP and ATP
Initial two-substrate kinetic assays were performed
with varying UMP concentrations (100–2000 lm) and
various fixed ATP concentrations (100, 200, 500 and
1000 lm), and the dependency of the velocity on sub-
strate concentration was hyperbolic for UMP as the
varied substrate (Fig. 5A). We then calculated what
should be the true K
m
and V
max
values for UMP,
giving 214 ± 4 lm and 262 ± 24 lmolÆmin
)1
Æmg
)1
,
respectively.
However, with ATP as the variable substrate, the
kinetic curves showed a detectable deviation from
Michaelis–Menten kinetics, especially at low substrate
UMP kinase from Ureaplasma parvum L. Egeblad-Welin et al.
6406 FEBS Journal 274 (2007) 6403–6414 ª 2007 The Authors Journal compilation ª 2007 FEBS
concentrations (Fig. 5B). The best fit was therefore to

the Hill equation, giving an n value of 1.54 ± 0.10,
demonstrating positive cooperativity with ATP. With
1mm UMP, the K
0.5,app
(ATP) was 316 ± 54 lm and
the V
max,app
(ATP) was 172 ± 23 lmolÆmin
)1
Æmg
)1
,
which is similar to the values calculated in the initial
kinetic analysis.
UTP as end-product inhibitor
The nature of UTP inhibition was investigated in
assays with fixed ATP (1 mm) and variable UMP con-
centrations (50–1000 lm). Double-reciprocal plots at
various UTP concentrations demonstrated that UTP
was a competitive inhibitor towards UMP, with a K
i
C
AB
B
A
A
B
E204
E204
T197

T197
A
C
F133
F133
T131
T131
A
C
D
E
A
K102
D104
K102
D104
D104
K102
C
E
D
Fig. 3. (A) Interactions between subunits A, B and C. (B) Hydrophobic interaction between the a4 helices from A and B. (C) Interaction
between a7 helices from A and C. Hydrogen bonds form between Thr197 and Glu204. (D) Hydrophobic interaction between A and C, formed
by Thr131 and Phe133 (also referred to as the cross-talk region). (E) Electrostatic interactions in the central channel of the enzyme between
subunits A, D and E. Asp104 holds Lys102 in position in each subunit. A water molecule is fixed by the three lysines from each subunit.
L. Egeblad-Welin et al. UMP kinase from Ureaplasma parvum
FEBS Journal 274 (2007) 6403–6414 ª 2007 The Authors Journal compilation ª 2007 FEBS 6407
value of 0.7 mm (Fig. 6A). When ATP was the vari-
able (50–1000 lm) substrate at a fixed UMP concentra-
tion (1 mm), the inhibition by UTP affected primarily

the V
max
values, while the K
0.5
⁄ K
m
values at various
UTP concentrations were in the same range. Thus,
UTP inhibition was noncompetitive towards ATP,
with a K
i
value of 1.2 mm (Fig. 6B). When this data
set was fitted to the Hill equation, the n values were
1.4, 0.98 and 1.0 for UTP at 0, 0.5 and 1.0 mm, respec-
tively, which indicates that the positive cooperativity
behavior with ATP is altered by the presence of UTP.
Determination of enzyme-bound orthophosphate
and inhibition of enzyme activity by
orthophosphate
In the UpUMPK structure, a phosphate ion was found
in the active site, and this raised a question concerning
the actual orthophosphate content in the enzyme used
in the functional studies. Therefore, the phosphate
content in the enzyme preparation was determined
using a colorimetric method. A concentrated enzyme
solution was precipitated with 5% perchloric acid at
low temperature to release bound phosphate ions. The
concentration of free phosphate in the supernatant was
3.52 lm, and the total concentration of enzyme was
A

B
F133
N140
D70
G56
R55
G50
T138
T131
T138
T145
R62
G57
G63
D77
Fig. 4. (A) Superposition of UpUMPK (green) on top of E. coli
UMPK (blue) at the active site. (B) Binding of UMP to E. coli UMPK
and the location of amino acid residues in UpUMPK based on the
superposition.
A
B
C
Fig. 5. (A) Activity versus [UMP] at various concentrations of ATP
(l
M). (B) Activity versus [ATP] at various concentrations of UMP
(l
M). (C) Lineweaver–Burk plot of 1 ⁄ v against 1 ⁄ [UMP] at various
concentrations of ATP (l
M).
UMP kinase from Ureaplasma parvum L. Egeblad-Welin et al.

6408 FEBS Journal 274 (2007) 6403–6414 ª 2007 The Authors Journal compilation ª 2007 FEBS
91.12 lm, giving a molar ratio of UpUMPK ⁄ phos-
phate of 25 ⁄ 1.
The effect of orthophosphate on enzyme activity
was examined. It was shown that phosphate inhibited
UpUMPK activity with an IC
50
value of 1 mm
(Fig. 7).
Functional consequences of F133N and F133A
mutations
GTP is an activator for all bacterial UMPKs studied
to date [5,14–17], and it was therefore tested with
UpUMPK. In an assay with 1 mm UMP and ATP, the
addition of 0.5 or 1 mm GTP resulted in no detectable
change in UpUMPK activity.
In order to find an explanation for the lack of GTP
activation, the UpUMPK structure was compared to
that of E. coli UMPK. In E. coli UMPK, residue
Asn140 forms a hydrogen bond to Thr138 of a neigh-
boring subunit (Fig. 8). The backbone of Asn140 and
the side chain of Thr138 also form hydrogen bonds to
the uracil base. Mutations of either Thr138 or Asn140
to Ala abolished GTP activation, indicating that these
residues are involved in GTP activation of the E. coli
UMPK [6]. In UpUMPK, a region between subunits A
and C had a Phe133 in the position corresponding to
Asn140 in the cross-talk region of E. coli UMPK.
Phe133 is not able to form hydrogen bonds due to its
hydrophobic interactions with Thr131 (Fig. 3D).

To mimic E. coli UMPK, Phe133 of UpUMPK was
mutated to Asn or Ala. The mutant enzymes, F133N
and F133A, were expressed, purified and characterized.
With 1 mm UMP and ATP as substrates, the activities
of F133N and F133A were only 50 and 20% of that
of the wild-type enzyme. Similar to the situation with
wild-type UpUMPK, addition of 0.5 and 1 mm GTP
resulted in no detectable activation of F133N or
F133A mutant enzymes.
A
B
Fig. 6. (A) UTP acts as a competitive inhibitor towards UMP.
Double-reciprocal plot of 1 ⁄ v versus 1 ⁄ [UMP] at 0, 0.1 and 0.5 m
M
UTP. (B) UTP acts as a noncompetitive inhibitor towards ATP.
Activity versus [UMP] at 0, 0.5 and 1 m
M UTP.
0
50
100
150
200
0 1000 2000 3000 4000 5000 6000
v (µmol/min/mg)
[Pi] µM
Fig. 7. Inhibition of UpUMPK activity by P
i
. [UMP] and
[ATP] ¼ 1m
M.

A
A
C
N140
T138
N140
T138
Fig. 8. Cross-talk region of E. coli UMPK between subunits A and
C, with UMP bound to the active site. Amino acid residues T138
and N140 are found in the cross-talk region (Protein Data Bank
accession number 2BNE) [6].
L. Egeblad-Welin et al. UMP kinase from Ureaplasma parvum
FEBS Journal 274 (2007) 6403–6414 ª 2007 The Authors Journal compilation ª 2007 FEBS 6409
With variable UMP concentration and fixed ATP
concentration, both the F133N and F133A mutants dis-
played negative cooperativity, and the Hill coefficients
were 0.65 ± 0.05 and 0.85 ± 0.05, respectively (Fig. 9).
At 1 mm ATP, the K
0.5,app
(UMP) and V
max,app
values
for F133N were 1100 ± 150 lm and 107 ± 15 lmolÆ
min
)1
Æmg
)1
, respectively. For F133A, the K
0.5,app
(UMP)

and V
max,app
values were 896 ± 212 lm and 63 ±
2 lmolÆmin
)1
Æmg
)1
, respectively. The K
0.5,app
(UMP)
values for the mutant enzymes were four- to fivefold
higher than that of the wild-type enzyme. The V
max,app
values were also affected; they were twofold lower in
case of F133N, and approximately fourfold lower for
the F133A mutant.
Discussion
In this study, we have investigated UpUMPK and
shown that the structure resembles UMPKs from bac-
teria and archaea belonging to the amino acid kinase
family. A phosphate ion was bound to all subunits in
the enzyme. However, the molar content of orthophos-
phate in the soluble UMPK was only 4%. The concen-
tration of phosphate giving 50% inhibition of enzyme
activity (IC
50
value) was 1 mm, indicating that the
phosphate did not have a very high affinity for the
enzyme. The discrepancy between the structural and
functional results is not easily explained and may be

methodological. At present, we cannot distinguish the
possibilities that the enzyme contains tightly bound
phosphate ions that cannot be released by acid precipi-
tation, or alternatively that only the phosphate-binding
fraction of the enzyme can form crystals.
The K
m
value for UpUMPK with UMP is high
(214 ± 4 lm) compared to the K
m
values for other
UMPKs, e.g. S. solfataricus,14lm; E. coli,43lm (at
pH 7.4); B. subtilis,30lm; St. pneumoniae, 100 lm
[8,14–16]. A possible reason for the high K
m
value for
UpUMPK with UMP could be the presence of a phos-
phate ion in the active site. However, as discussed
above, the kinetic results most likely reflect the proper-
ties of the native fully active UpUMPK enzyme.
Positive cooperativity with ATP was observed when
the assays were performed with ATP as the variable
substrate (n value of 1.5). However, in the inhibition
experiment with UTP, the n values were close to 1.0,
indicating that the presence of UTP abolished the
positive cooperativity observed with ATP alone. At
present, there is no clearcut explanation for this obser-
vation. Nevertheless, the cooperative behavior of
UpUMPK with varied ATP concentrations is less pro-
nounced than that reported with other Gram-positive

bacterial UMPKs (n values of 1.9–2.5 with 1 mm
UMP) [14,17].
The fact that UTP is a competitive inhibitor for UMP
is in agreement with the structural data from E. coli
UMPK, where it has been shown that UTP binds to the
base moiety in the active site [6]. For S. solfataricus, the
same pattern was observed with UTP and UMP, but in
that case UTP is a competitive inhibitor towards ATP
[8]. The K
i
values for UTP versus UMP (0.7 mm) and
ATP (1.2 mm) are high, and may indicate that UTP is
an inefficient inhibitor in vivo.
UMPKs from E. coli, Salmonella typhimurium, H. in-
fluenzae, Neisseria meningitidis, B. subtilis, St. pneumo-
niae, Staphylococcus aureus and Enterococcus faecalis
were all activated by GTP by a factor of 2.5–18.5
[17]. The UMPK from the archaea S. solfataricus
was the first UMPK for which lack of activation by
GTP was shown [8], and in this study we have shown
that UMPK from U. parvum also lacks activation by
GTP.
The mutational study of residue F133 was per-
formed to clarify whether this residue is involved in
GTP activation. F133 was mutated to Asn in an
attempt to create a GTP-activated enzyme, and F133A
was prepared and tested as a control. Neither of the
UpUMPK mutants F133N and F133A were activated
by GTP. Thus, this residue alone is not responsible for
GTP activation. However, an interesting feature was

observed. UpUMPK F133N exhibited negative cooper-
ativity with UMP as a substrate (n value of 0.65), and
the same was true of UpUMPK F133A to a lesser
extent (n value of 0.85). This is the first time that nega-
tive cooperativity has been described with a bacterial
UMPK. The observed negative cooperativity may be
explained by alteration of the geometry of the active
site in the neighboring subunit when UMP binds to
the enzyme, i.e. mutation of residue F133, which is
F133A
F133N
60
50
40
30
V/(u/mg)
20
10
0
0 200 400 600
[UMP]/µ
M
800 1000 1200
Fig. 9. Substrate saturation curves of UpUMPK mutant enzymes:
F133N and F133A with UMP as variable substrate.
UMP kinase from Ureaplasma parvum L. Egeblad-Welin et al.
6410 FEBS Journal 274 (2007) 6403–6414 ª 2007 The Authors Journal compilation ª 2007 FEBS
located in the interface of two subunits, may have
affected the mode of subunit interaction.
Jensen et al. (2007) have compared the sequence and

structure of S. solfataricus UMPK to those of the
known bacterial UMPKs. A loop between a6 and a7
was only present in archaea, and was referred to as the
archaea-specific loop (Fig. 1). Another difference that
was detected was in the loop between a3 and a4,
which was absent in archaea, and is therefore referred
to as the bacteria-specific loop (Fig. 1). Jensen et al.
(2007) suggested that either the amino acid residues
found in the bacteria-specific loop or the lack of a
known nucleotide binding motif GXXGXG [21] in the
N-terminus was responsible for the lack of GTP acti-
vation [8]. A comparison of UMPKs from U. parvum,
E. coli and S. solfataricus, chosen to represent myco-
plasma, bacteria and archaea, showed that they all
shared the same fold. As UpUMPK has essentially the
same fold as E. coli UMPK, it is unlikely that residues
in the bacteria-specific loop are involved in the activa-
tion. However, the N-terminal GXXGXG motif is not
found in UpUMPK, suggesting that this motif may be
involved in GTP activation.
UMPK is involved in both de novo and salvage syn-
thesis of DNA and RNA precursors. The results pre-
sented here suggest that UpUMPK is an enzyme that is
mainly regulated by the UMP and orthophosphate lev-
els, and is not very sensitive to feedback inhibition by
the end-product UTP. Furthermore, there is no evi-
dence for allosteric activation by GTP, although the
overall structure is highly similar to the other bacterial
UMPKs. The relative simplicity of the apparent regula-
tion of this structurally complex enzyme may be due to

its lifestyle, as it grows in the urinary tract where the
salvage of uridine and uracil may serve as a rich source
for UMP biosynthesis. One of the goals of this investi-
gation was to evaluate whether UpUMPK is a promis-
ing new target for development of antibacterial agents.
Bacterial UMPKs have no sequence or structural
homology to the human enzyme CMP–UMPK, which
makes them potential targets for drug development,
but, in the case of UpUMPK, a search for non-nucleo-
side ⁄ nucleotide inhibitors may be more successful.
Experimental procedures
Site-directed mutagenesis
The expression plasmid pET-14b-UpUMPK has been
described previously by Wang [4]. The mutants UpUMPK-
F133N and UpUMPK-F133A were constructed by site-
directed mutagenesis using the plasmid pET-14b containing
cDNA for UMPK. The F133N mutation was created using
the following primers: F133N-fw (5¢-GATTTTTGTGGCT
GGAACAGGA
AACCCATATTTTACAACTGATTCG)
and F133N-rv (5¢-CGAATCAGTTGTAAAATATGG
GT
TTCCTGTTCCAGCCACAAAAAT), with the altered
nucleotides shown in bold and underlined. The F133A
mutation was created using the following primers: F133A-
fw (5¢-GTGGCTGGAACAGGA
GCGCCATATTTTACA
ACTGATTCG) and F133A-rv (5¢-CGAATCAGTTGTAA
AATATGG
CGCTCCTGTTCCAGCCAC). The mutations

were verified by DNA sequencing using the BigDye termi-
nator cycle sequencing kit and the ABI PRISM 310 genetic
analyzer (PE Applied Biosystems, Foster City, CA, USA).
Expression and purification of recombinant
enzymes
Both wild-type UpUMPK and the mutant enzymes were
expressed in E. coli BL21 (DE3) in Luria–Bertani (LB) med-
ium. The enzymes were overexpressed by induction with iso-
propyl-b-d-thiogalactoside (0.16 mm) overnight at 37 °C,
and bacteria were harvested by centrifugation at 4600 · g
for 15 min at 4 °C. The pellet was resuspended in buffer A,
containing 50 mm Tris ⁄ HCl (pH 7.5), 0.2 m KCl, 5 mm
MgCl
2
and 0.2 mm phenylmethylsulfonyl fluoride. The cells
were then disrupted by sonication for 5 min with 5 s pulses,
and thereafter centrifuged for 30 min at 4600 · gat4°C.
Purification was carried out at 4 °C. The supernatant was
applied to a metal affinity column (TALON resin, BD Bio-
sciences Clontech, Palo Alto, CA, USA) using the gravity
flow procedure. The column was washed first with buffer B,
containing 20 mm Tris ⁄ HCl (pH 7.5) and 0.2 m KCl, and
then with buffer B and 20 mm imidazole. The protein was
eluted with buffer B and 250 mm imidazole. The purity of
the protein was analyzed by SDS ⁄ PAGE [22], and the pro-
tein concentration was determined according to the method
described by Bradford [23], with BSA as the standard pro-
tein. For wild-type UpUMPK, approximately 110 mg pure
protein was obtained from a 1 L culture.
Crystallization

The UMPK contained an N-terminal His-tag with the
sequence MGSSHHHHHHSSGLVPRGSHM. Crystals
were grown by vapor diffusion, under conditions of
0.2 m ammonium fluoride and 20% (w ⁄ v) poly(ethylene
glycol) 3350 at 15 °C. The enzyme concentration was
1.8 mgÆmL
)1
, and 5 mm GTP was added to the protein.
The protein and crystallization solution were mixed equally
(2 lL of each) in a hanging drop.
Data collection
The crystals were flash-frozen in liquid nitrogen, using
mother solution with the addition of 15% (v ⁄ v) poly(ethylene
L. Egeblad-Welin et al. UMP kinase from Ureaplasma parvum
FEBS Journal 274 (2007) 6403–6414 ª 2007 The Authors Journal compilation ª 2007 FEBS 6411
glycol) 400 as cryoprotectant, and data were collected at
ID14-eh4 at the European synchotron radiation facility
(ESRF), Grenoble, France. The data were indexed, scaled
and merged using mosflm [24] and scala [25], and the crys-
tals were found to belong to the space group P2
1
with a sol-
vent content of 48%. The content of the asymmetric unit was
six monomers.
Structure determination and refinement
The structure was solved by molecular replacement using
molrep [26], with the monomer of UMPK from H. influen-
zae (Protein Data Bank accession number 2A1F) as the
search model. Simulated annealing was performed in cns
[27], and further refinement was performed in refmac5

[28], during which noncrystallographic symmetry (NCS)
restraints were applied to residues 5–160 and 193–230 with
tight main-chain and medium side-chain restraints. Model
building was carried out in o [29] and coot [30].
In chain A, residues 1 and 169–176 are missing; in
chain B, residues 167–172 are missing; in chain C, residues
1, 2 and 169–174are missing; in chain D, residues 168–174
are missing; in chain F, residues 1 and 168–175 are missing.
All amino acid residues were present in the E chain, and
part of the His-tag was observed in the B chain. A few
amino acid residues are found in disallowed regions in the
Ramachandran plot; these are A4, A165, A167, B3, B4 and
F165, all of which are found at either the beginning or the
end of a chain. The structure has been deposited to Protein
Data Bank under the accession number 2va1. All figures
were created using pymol [31], and sequence alignments
were created using clustalw [32] and espript [33].
Determination of enzyme-bound orthophosphate
The presence of orthophosphate in UpUMPK was determined
by a colorimetric method [34]. Briefly, the buffer used for
UpUMPK preparation was exchanged with water using
a PD-10 column (GE Healthcare, Uppsala, Sweden), and
then the protein concentration was determined. To 2 mL
UpUMPK solution, perchloric acid was added to a final con-
centration of 5%, and the mixture was incubated on ice for
10 min. The mixture was then centrifuged at 16 000 · gfor
15 min at 4 °C to remove precipitated protein. The superna-
tant was neutralized with KOH, and incubated on ice for
15 min. After centrifugation at 16 000 g for 20 min at 4 °C,
the supernatant was used in the colorimetric assay as described

previously [34]. The concentration of orthophosphate was
3.52 lm,andtheUpUMPK concentration was 91.12 lm.
Enzyme assays
The UMPK activity was determined using a coupled spec-
trophotometric assay [20] with a Cary 3 spectrophotometer
(Varian Techtron, Mulgrave, Australia) at 37 °C. The
reaction medium (final volume 1 mL) contained 50 mm
Tris ⁄ HCl pH 6.8, 5 mm dithiothreitol, 0.5 mg mL
)1
BSA,
1mm phosphoenolpyruvate, 0.3 mm NADH and 4 lmolÆ
min
)1
Æmg
)1
ÆmL
)1
of pyruvate kinase and lactate dehydroge-
nase. Nucleoside diphosphate kinase (NDPK) was not
added, as this did not lead to a significant change in the
rates determined, as observed by Fassy et al. [14], and
avoids the complication of potential UTP formation. The
coupling enzymes (pyruvate kinase and lactate dehydroge-
nase) were tested with ADP and UDP, and ADP showed a
rate that was > 20 times that of UDP.
In order to determine the true K
m
for UMP and ATP, a
two-substrate assay was performed at four concentrations
of UMP and ATP (100, 200, 500 and 1000 lm). In the

GTP-activation experiments, the concentrations of UMP
and ATP were kept at 1 mm. In all experiments, MgCl
2
concentration was kept in a stoichiometry of 2 : 1 towards
NTP. The enzyme concentration was 0.5 lg per assay for
the wild-type and UpUMPK-F133N, and 1 l g per assay
for UpUMPK-F133A. The decrease in [NADH] was moni-
tored at 340 nm.
Analysis of kinetic data
Kinetic data were evaluated by nonlinear regression
analysis using either the Michaelis–Menten equation v ¼
V
max
Æ[S] ⁄ (K
m
+ [S]), or the Hill equation v ¼ V
max
Æ[S]
n
⁄
(K
n
0:5
+ [S]
n
), where K
m
is the Michaelis constant, K
0.5
is

the value of the substrate concentration [S] where v ¼
0.5 V
max
, and n is the Hill coefficient. If n ¼ 1, there is no
cooperativity, if n < 1 there is negative cooperativity, and
if n > 1 there is positive cooperativity. One unit corre-
sponds to 1 lmol min
)1
.
The inhibition studies were analyzed using equations for
competitive and noncompetitive inhibitors. For competitive
inhibition, the equation is v ¼ V
max
Æ[S] ⁄ (K
m
(1 + [I] ⁄
K
i
) + [S]), and for noncompetitive inhibition the equation
is v ¼ V
max
Æ[S] ⁄ (K
m
+ [S])(1 + [I] ⁄ K
i
). K
i
for UTP towards
ATP was determined using the secondary plot of slope
versus [UTP].

Acknowledgements
The authors wish to thank Andrea Hinas and Fredrik
So
¨
derbom (Department of Molecular Biology, Swedish
University of Agricultural Science) for help with muta-
genesis, Hans Eklund (Department of Molecular Biol-
ogy, Swedish University of Agricultural Science) for
help with the structure determination, and Mark Harris
(Department of Molecular Biology, Uppsala Univer-
sity) for proof reading. This work was supported by
grants from the Swedish Research Council and the
UMP kinase from Ureaplasma parvum L. Egeblad-Welin et al.
6412 FEBS Journal 274 (2007) 6403–6414 ª 2007 The Authors Journal compilation ª 2007 FEBS
Swedish Research Council for the Environment, Agri-
cultural Sciences, and Spatial Planning (FORMAS).
References
1 Pollack JD (2001) Ureaplasma parvum: an opportunity
for combinatorial genomics. TRENDS Microbiol 9,
169–175.
2 Volgmann T, Ohlinger R & Panzig B (2005) Ureaplasma
parvum – harmless commensal or underestimated enemy
of human reproduction? A review. Arch Gynecol Obstet
373, 133–139.
3 Glass JI, Lefkowitz EJ, Glass JS, Heiner CR, Chen EY
& Cassell GH (2000) The complete sequence of the
mucosal pathogen Ureaplasma parvum. Nature 407,
757–762.
4 Wang L (2007) The role of nucleoside monophosphate
kinases in the synthesis of nucleoside triphosphates.

FEBS J 274, 1983–1990.
5 Serina L, Blondin C, Krin E, Sismeiro O, Danchin A,
Sakamato H, Gilles A-M & Baˆ rzu O (1995) Escherichia
coli UMP-kinase, a member of the aspartokinase family,
is a hexamer regulated by guanine nucleotides and
UTP. Biochemistry 34, 5066–5074.
6 Briozzo P, Evrin C, Meyer P, Assairi L, Joly N, Baˆ rzu
O & Gilles A-M (2005) Structure of Escherichia coli
UMP kinase differs from that of other NMP kinases
and sheds new light on enzyme regulation. J Biol Chem
280, 25533–25540.
7 Marco-Marı
´
n C, Gil-Ortiz F & Rubio V (2005) The
crystal structure of Pyrococcus furiosus UMP kinase
provides insight into catalysis and regulation in micro-
bial nucleotide biosynthesis. J Mol Biol 352, 438–454.
8 Jensen KS, Johansson E & Jensen KF (2007) Structural
and enzymatic investigation of the Sulfolobus solfatari-
cus uridylate kinase shows competitive UTP inhibition
and lack of GTP stimulation. Biochemistry 46, 2745–
2757.
9 Yan H & Tsai MD (1999) Nucleoside monophosphate
kinases: structure, mechanism, and substrate specificity.
Adv Enzymol Relat Areas Mol Biol 73, 103–134.
10 Hutchison CA III, Peterson SN, Gill SR, Cline RT,
White O, Fraser CM, Smith HO & Venter JC (1999)
Global transposon mutagenesis and a minimal myco-
plasma genome. Science 286, 2165–2169.
11 Glass JI, Assad-Garcia N, Alperovich N, Yooseph S,

Lewis MR, Maruf M, Hutchison CA III, Smith HO &
Venter JC (2006) Essential genes of a minimal bacte-
rium. Proc Natl Acad Sci USA 103, 425–430.
12 Yamanaka K, Ogura T, Niki H & Hiraga S (1992)
Identification and characterization of the smbA gene, a
suppressor of the mukB null mutant of Escherichia coli.
J Bacteriol 174, 7517–7526.
13 Akerley BJ, Rubin EJ, Novick VL, Amaya K, Judson
N & Mekalanos JJ (2002) A genome-scale analysis for
identification of genes required for growth or survival of
Haemophilus influenzae.
Proc Natl Acad Sci USA 99,
966–971.
14 Fassy F, Krebs O, Lowinski M, Ferrari P, Winter J,
Collard-Dutilleul V & Salahbey Hocini K (2004) UMP
kinase from Streptococcus pneumoniae: evidence for
co-operative ATP binding and allosteric regulation.
Biochem J 384, 619–627.
15 Serina L, Bucurenci N, Gilles A-M, Surewicz WK,
Fabian H, Mantsch HH, Takahashi M, Petrescu I,
Batelier G & Baˆ rzu O (1996) Structural properties of
UMP-kinase from Escherichia coli by pH and UTP.
Biochemistry 35, 7003–7011.
16 Gagyi C, Bucurenci N, Sıˆ rbu O, Labesse G, Ionescu M,
Ofiteru A, Assairi L, Landais S, Danchin A, Baˆ rzu O &
Gilles A-M (2003) UMP kinase from the gram-positive
bacterium Bacillus subtilis is strongly dependent on
GTP for optimal activity. Eur J Biochem 270, 3196–
3204.
17 Evrin C, Straut M, Slavova-Azmanova N, Bucurenci N,

Onu A, Assairi L, Ionescu M, Palibroda N, Baˆ rzu O &
Giles AM (2007) Regulatory mechanisms differ in UMP
kinases from gram-negative and gram-positive bacteria.
J Biol Chem 282, 7242–7253.
18 Bucurenci N, Serina L, Zaharia C, Landais S, Danchin
A&Baˆ rzu O (1998) Mutational analysis of UMP
kinase from Escherichia coli. J Bacteriol 180, 473–477.
19 Saraste M, Sibbald PR & Wittinghofer A (1990) The
P-loop – a common motif in ATP- and GTP-binding
proteins. Trends Biochem Sci 15, 430–434.
20 Blondin C, Serina L, Wiesmu
¨
ller L, Gilles A-M &
Baˆ rzu O (1994) Improved spectrophotometric assay of
nucleoside monophosphate kinase activity using the
pyruvate kinase ⁄ lactate dehydrogenase coupling system.
Anal Biochem 220, 219–221.
21 Schulz GE (1992) Binding of nucleotides by proteins.
Curr Opin Struct Biol 2, 61–67.
22 Laemmli UK (1970) Cleavage of structural proteins dur-
ing the assembly of the head bacteriophage T4. Nature
227, 680–685.
23 Bradford MM (1976) A rapid and sensitive method for
the quantitation of microgram quantities of protein
utilizing the principle of protein–dye binding. Anal
Biochem 72, 248–254.
24 Leslie AGW (1992) Joint CCP4 + ESF-EAMCB News-
letters on Protein Crystallography, no.26.
25 Collaborative Computational Project Number 4 (1994)
The CCP4 suite: programs for protein crystallography.

Acta Crystallogr D 50, 760–763.
26 Vagin A & Teplyakov A (2000) An approach to multi-
copy search in molecular replacement. Acta Crystallgr
D 56, 1622–1624.
27 Brunger AT, Adams PD, Clore GM, DeLano WL, Gros
P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges
M, Pannu NS, et al. (1998) Crystallography & NMR
L. Egeblad-Welin et al. UMP kinase from Ureaplasma parvum
FEBS Journal 274 (2007) 6403–6414 ª 2007 The Authors Journal compilation ª 2007 FEBS 6413
system: a new software suite for macromolecular struc-
ture determination. Acta Crystallogr D 54, 905–921.
28 Murshudov GN, Vagin AA & Dodson EJ (1997)
Refinement of macromolecular structures by the maxi-
mum-likelihood method. Acta Crystallogr D 53, 240–
255.
29 Jones TA, Zou JY, Cowan SW & Kjeldgaard M (1991)
Improved methods for building protein models in
electron density maps and the location of errors in these
models. Acta Crystallogr A 47, 110–119.
30 Emsley P & Cowtan K (2004) Coot: model building
tools for molecular graphics. Acta Crystallogr D 60,
2126–2132.
31 DeLano WL (2002) The PyMOL User’s Manual.
DeLano Scientific, San Carlos, CA, USA.
32 Chenna R, Sugawara H, Koike T, Lopez R, Gibson T,
Higgins DG & Thompsen JD (2003) Multiple sequence
alignment with the Clustal series of programs. Nucleic
Acids Res 31, 2497–3500.
33 Gouet P, Courcelle E, Stuart DI & Metoz F (1999)
ESPript: multiple sequence alignments in PostScript.

Bioinformatics 15, 305–308.
34 Gonzalez-Romo P, Sanchez-Nieto S & Gavilanes-Ruiz
M (1992) A modified colorimetric method for determi-
nation of orthophosphate in the presence of high ATP
concentrations. Anal Biochem 200, 235–238.
UMP kinase from Ureaplasma parvum L. Egeblad-Welin et al.
6414 FEBS Journal 274 (2007) 6403–6414 ª 2007 The Authors Journal compilation ª 2007 FEBS