Báo cáo khoa học: Mouse RS21-C6 is a mammalian 2¢-deoxycytidine 5¢-triphosphate pyrophosphohydrolase that prefers 5-iodocytosine pdf

Mouse RS21-C6 is a mammalian 2¢-deoxycytidine
5¢-triphosphate pyrophosphohydrolase that prefers
5-iodocytosine
Mari Nonaka, Daisuke Tsuchimoto, Kunihiko Sakumi and Yusaku Nakabeppu
Division of Neurofunctional Genomics, Department of Immunobiology and Neuroscience, Medical Institute of Bioregulation, Kyushu
University, Fukuoka, Japan

Keywords
5-I-dCTP; CpG methylation; dCTPase;
modified nucleotide; nucleotide metabolism
Correspondence
D. Tsuchimoto, Division of Neurofunctional
Genomics, Department of Immunobiology
and Neuroscience, Medical Institute of
Bioregulation, Kyushu University, Maidashi
3-1-1, Higashi-ku, Fukuoka 812-8582, Japan
Fax: +81 92 642 6804
Tel: +81 92 642 6802
E-mail:
(Received 2 September 2008, revised
8 January 2009, accepted 12 January 2009)
doi:10.1111/j.1742-4658.2009.06898.x

Free nucleotides in living cells play important roles in a variety of biological reactions, and often undergo chemical modifications of their base
moieties. As modified nucleotides may have deleterious effects on cells, they
must be eliminated from intracellular nucleotide pools. We have performed
a screen for ITP-binding proteins because ITP is a deaminated product of
ATP, the most abundant nucleotide, and identified RS21-C6 protein, which
bound not only ITP but also ATP. Purified, recombinant RS21-C6 hydrolyzed several canonical nucleoside triphosphates to the corresponding
nucleoside monophosphates. The pyrophosphohydrolase activity of RS21C6 showed a preference for deoxynucleoside triphosphates and cytosine
bases. The kcat ⁄ Km (s)1Ỉm)1) values were 3.11 · 104, 4.49 · 103 and

1.87 · 103 for dCTP, dATP and dTTP, respectively, and RS21-C6 did not
hydrolyze dGTP. Of the base-modified nucleotides analyzed, 5-I-dCTP
showed an eightfold higher kcat ⁄ Km value compared with that of its corresponding unmodified nucleotide, dCTP. RS21-C6 is expressed in both proliferating and non-proliferating cells, and is localized to the cytoplasm.
These results show that RS21-C6 produces dCMP, an upstream precursor
for the de novo synthesis of dTTP, by hydrolyzing canonical dCTP. Moreover, RS21-C6 may also prevent inappropriate DNA methylation, DNA
replication blocking or mutagenesis by hydrolyzing modified dCTP.

In living organisms, nucleotides play various roles, as
signal transmitters, molecular switches, coenzymes or
as carriers of energy, in addition to their important
role as precursors of DNA ⁄ RNA synthesis. For example, ATP is a major carrier of energy, a phosphate
group donor in kinase reactions, and an extracellular
signal transmitter. GTP is a molecular switch in signal
transduction pathways and an initiator complex for
translation. UTP and CTP are utilized to form active

intermediates in the biosynthesis of polysaccharides or
phospholipids, respectively. In such roles, recognition
of nucleotides by specific proteins is very important.
Intracellular nucleotides, however, undergo chemical
modifications caused by endogenous reactive molecules, such as reactive oxygen species, or by exogenous
factors, such as chemicals and ionizing irradiation.
Chemical modification may alter the characteristics of
nucleotides, including their recognition by proteins.

Abbreviations
2-Cl-dATP, 2-chloro-(2¢-deoxy)adenosine 5¢-triphosphate; 2-OH-(d)ATP, 2-hydroxy-(2¢-deoxy)adenosine 5¢-triphosphate; 5-Br-dCTP, 5-bromo-2¢deoxycytidine 5¢-triphosphate; 5-F-dUTP, 5-fluoro-2¢-deoxycytidine 5¢-triphosphate; 5-I-dCTP, 5-iodo-2¢-deoxycytidine 5¢-triphosphate;
5-Me-dCTP, 5-methyl-2¢-deoxycytidine 5¢-triphosphate; 5-OH-dCTP, 5-hydroxy-2¢-deoxycytidine 5¢-triphosphate; 8-oxo-(d)GTP, 8-oxo-(2¢deoxy)guanosine 5¢-triphosphate; DCTD, dCMP deaminase; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; NTP, nucleoside
5¢-triphosphate; RNR, ribonucleotide reductase; TS, thymidylate synthase.

1654

FEBS Journal 276 (2009) 1654–1666 ª 2009 The Authors Journal compilation ª 2009 FEBS


M. Nonaka et al.

Some modified deoxynucleotides are incorporated into
DNA by DNA polymerases and accumulate in newly
synthesized DNA. This may prevent DNA replication
or transcription, resulting in cell death and degenerative diseases in humans [1]. Normal functions of nucleotides, other than DNA synthesis, may also be
adversely affected by modified nucleotides. Cells are
equipped with defense systems against such modified
nucleotides. Some modified nucleotides in intracellular
nucleotide pools are hydrolyzed by specific enzymes
[1,2]. Of these enzymes, dUTPase and MTH1 are the
best studied in human cells. The former hydrolyzes
deoxyuridine triphosphate to prevent its incorporation
into DNA. The latter hydrolyzes oxidized purine
nucleoside triphosphates, including 8-oxo-(deoxy)guanosine triphosphate [8-oxo-(d)GTP] and 2-hydroxy(deoxy)adenosine triphosphate [2-OH-(d)ATP], to the
corresponding (deoxy)nucleoside monophosphates and
pyrophosphates to avoid their incorporation into
DNA or RNA [3]. The spontaneous mutation rate in
MTH1-null mouse embryonic stem cells was twofold
higher than that in wild-type cells. Further, MTH1null mice showed more frequent tumorigenesis in the
liver compared to wild-type mice [4].
In addition to oxidization, deamination of the amino
group is another major chemical modification that
occurs in purine nucleotides. Deamination of the
amino group at C6 of adenine or C2 of guanine generates hypoxanthine or xanthine, respectively. Thus,

(d)ITP and (d)XTP are generated from (d)ATP and
(d)GTP, respectively. Incorporation of these modified
nucleotides into DNA during DNA replication or into
RNA during transcription results in gene mutations or
the synthesis of abnormal proteins because hypoxanthine and xanthine can mis-pair with cytosine or
thymine, respectively. Recently, mammalian inosine
triphosphate pyrophosphohydrolases (ITPases) have
been reported to hydrolyze deaminated purine nucleoside triphosphates to the corresponding nucleoside
monophosphates and pyrophosphates [5,6]. ITPasenull mice, in which accumulation of ITP was observed,
showed abnormal development and died within
14 days after birth (M. Behmanesh, K. Sakumi,
S. Toyokuni, S. Oka, Y. Ohnishi, D. Tsuchimoto &
Y. Nakabeppu, unpublished results). The ITP that
accumulates in these mice may have deleterious effects
on cell functions, for example via DNA ⁄ RNA synthesis, or, because of its structural similarity to ATP, by
interaction with ATP-related proteins.
In the present study, we prepared ITP-agarose and
purified ITP-binding proteins to identify additional
ITP-hydrolyzing enzymes or target proteins whose
function can be inhibited by ITP. As a result, we iden-

A mammalian dCTPase that prefers 5-iodocytosine

tified RS21-C6, which was previously reported to be a
thymocyte development-related molecule [7], as well as
ITPase, as ITP-binding proteins. Because RS21-C6
contains a typical MazG domain conserved in the bacterial NTP pyrophosphatase MazG, it has been
described as a member of the all-a NTP pyrophosphohydrolase superfamily with all-a helix structures [8]. A
preliminary structure of RS21-C6 without substrate
has been initially determined [9]. Recently, it was

shown that RS21-C6 hydrolyzes 5-methyl-dCTP, and
the crystal structure of truncated RS21-C6 complexed
with 5-methyl-dCTP indicated that tetramer formation
is required for substrate binding [10]. We examined the
NTP pyrophosphohydrolase activity of purified recombinant RS21-C6 protein towards various nucleotides,
and found that it hydrolyzes some deoxynucleotides,
particularly dCTP, but not dITP or ITP. Furthermore,
we found that iodination at C5 of cytosine significantly
increases the kcat ⁄ Km value of RS21-C6.

Results
Preparation of ITP-agarose
We prepared ITP-agarose from ATP-agarose as
described in Experimental procedures. Analysis of
bases excised from agarose beads revealed that most
adenine bases on the agarose were converted to hypoxanthine after deamination (Fig. 1). We also confirmed
that most free ATP was converted to ITP after the
same treatment (data not shown), demonstrating that
the nucleotides on the treated agarose were ITP. Quantification of released bases indicated that the amounts
of nucleotide on ATP- and ITP-agarose were 17.3 and
2.6 nmol per 25 lL bed volume, respectively.
ITP-binding proteins
ITP-binding proteins were purified from mouse thymocyte extract by a pulldown method using ITP-agarose.
Proteins were then fractionated by SDS–PAGE
(Fig. 2A). After staining the acrylamide gel, we chose
ten ITP-specific bands, and, from these, identified 11
proteins by LC-MS ⁄ MS analysis (Table 1). We
detected ten peptides of ITPase and four peptides of
RS21-C6 in bands 4 and 5, respectively. Specific
binding of ITPase to ITP-agarose was confirmed by

western blot analysis of pulldown samples using antiITPase serum [5] (Fig. 2B, upper panel). To analyze the
interaction of RS21-C6 with ITP in detail, RS21-C6
cDNA and recombinant RS21-C6 protein were prepared as described in Experimental procedures. We
performed pulldown experiments, using ITP-agarose

FEBS Journal 276 (2009) 1654–1666 ª 2009 The Authors Journal compilation ª 2009 FEBS

1655


A mammalian dCTPase that prefers 5-iodocytosine

248.5 nm (AU)

0.06

M. Nonaka et al.

Table 1. ITP-binding proteins identified by LC-MS ⁄ MS analysis.

Before deamination

Band
no.

0.04

1

0.02

0

248.5 nm (AU)

0.06

Acetyl coenzyme A acetyltransferase
1 precursor
Glyceraldehyde-3-phosphate
dehydrogenase
Glutathione S-transferase
Inosine triphosphatase
Similar to ISOC2 protein
RS21-C6
Phospholipid hydroperoxide glutatione
peroxidase
Ribosomal protein L30
Divalent cation tolerant protein
CUTA isoform 1
Unidentified
Isochorismatase domain-containing 1
Es protein 1
Unidentified

2

0.00
4

8

12

16

20

3
4

After deamination

5

0.04
0.02

6

*
0.00
0

4

8
12
Retention time (min)

16

20

7
8
9
10

114
84.7

gi|21450129
gi|50233866
gi|2781337
gi|31982664
gi|20818892
gi|13435502
gi|2522259
gi|6677783
gi|62198210

gi|31541909
gi|20070420

from cell extracts of Escherichia coli BL21-CodonPlus
(DE3)-RIL that had been transformed with pET3a:
RS21-C6 and induced for RS21-C6 expression. Recombinant RS21-C6 protein in bacterial cell extracts also
bound to both ITP- and ATP-agarose (Fig. 2C).
D

Agarose

ATP-agarose

(kDa)
25. 7

(kDa)

ITP-agarose

B

ITP-agaros e

A

ATP-agaros e

Fig. 1. Preparation of ITP-agarose. ATP-agarose were incubated in
1 M HCl before (upper graph) or after (lower graph) deamination,
and the bases released were analyzed by HPLC. The peaks indicated by an open arrowhead and a closed arrowhead were compared with peaks of standard samples, and were identified as
adenine base and hypoxanthine base, respectively. The peak indicated by an asterisk was also observed in a sample released from
agarose without any nucleotide, suggesting that it was derived
from the carrier agarose (data not shown).

NCBInr
accession no.

Protein description

(kDa)
250
150
100
75
ITPase

17. 4
25.7

50
37

47.3
1

7

(kDa)
25
20

8
9
10

15

Recombinant

RS21-C6

15

Agarose

4
5
6

20

C

ITP-agarose

3
25.7

17.4

25

2
ATP-agarose

31.6

RS21-C6

17.4

Recombinant
RS21-C6

Fig. 2. Purification of ITP-binding proteins. (A) Pulldown of ITP-specific proteins from mouse thymocyte extract. Proteins were pulled down
using ATP- or ITP-agarose and then separated by SDS–PAGE. The lanes were loaded with samples pulled down from 5.0 · 107 cells. The
gel was subjected to silver staining. Arrowheads indicate the numbered ITP-specific bands that were recovered and subjected to a mass
spectrometry. (B) Western blot of pulled down samples with anti-ITPase serum and anti-RS21-C6 Ig. Samples pulled down from 2.5 · 107
cells were loaded on the gel as described in (A). (C) Pulldown of recombinant RS21-C6. Recombinant RS21-C6 protein was expressed in
E. coli. Binding proteins were pulled down using agarose beads, separated by SDS–PAGE and stained by silver staining. (D) Purification of
recombinant RS21-C6 protein. Recombinant RS21-C6 protein, expressed in E. coli, was purified as described in Experimental procedures.
The purified protein (100 ng) was separated by SDS–PAGE and stained by silver staining.

1656

FEBS Journal 276 (2009) 1654–1666 ª 2009 The Authors Journal compilation ª 2009 FEBS


M. Nonaka et al.

A mammalian dCTPase that prefers 5-iodocytosine

Expression of endogenous RS21-C6

A

pcDNA
:RS21-C6

pcDNA

BALB/3T3

NIH/3T3

WEHI231

J774A.1

3BW5147.3

Thymocyte

(kDa)
20

(kDa)
114
84.7

RS21-C6

15

47.3

GAPDH

31.3

25.7

C

W N

Mt Cyt
RS21-C6

17.4

RS21-C6
Lamin B
HSP60
GAPDH

GAPDH

20

15

10

5

reb
rum
reb
ellu

m
Ey
Sp
e
ina
l co
rd
He
art
Kid
ne
y
Liv
er
Es Lung
op
ha
gu
Ly
s
mp
hn
od
e
Bo Sple
en
ne
ma
rro
w

Th
ym
us
Te
stis
Ute
r
Sto us
ma
Sa
liva
ch
ry
gla
nd
Sk
Ov
ele
ary
tal
mu
Th
scl
yro
e
id
gla
nd

0

Ce

Relative expression

D

Ce

Fig. 3. Expression of endogenous RS21-C6.
(A) Western blot analysis of RS21-C6 protein in various mouse cell lines. A whole-cell
extract from 1.0 · 105 cells of each cell line
was loaded in each lane. Proteins were separated by SDS–PAGE, and then transferred
to a poly(vinylidene difluoride) membrane.
Signals for RS21-C6 protein and of GAPDH
were detected using anti-RS21-C6 and antiGAPDH Ig, respectively. (B) Western blot
analysis of endogenous or recombinant
RS21-C6 proteins in A20 cells. A20 cells
were transfected with plasmids expressing
recombinant RS21-C6 or with control vectors by electroporation. The cells were incubated for 24 h, and whole-cell extracts from
1.0 · 105 cells were loaded on each lane.
(C) Intracellular localization of RS21-C6 protein. Aliquots (20 lg protein) from whole-cell
extract (W) or each cell fraction were loaded
into each lane. Lamin B, HSP60 or GAPDH
were detected as nuclear (N), mitochondrial
(Mt) or cytoplasmic (Cyt) markers, respectively. (D) Real-time quantitative PCR analysis of RS21-C6 expression in various mouse
tissues. The mRNA expression levels of
RS21-C6 were normalized to those of 18S
rRNA. Error bars represent SD (n = 3). The
expression level of RS21-C6 in spleen was

arbitrarily set as 1.0, and the expression
levels in the other tissues are expressed
relative to that in spleen.

A20

B

pIRES2-EGFP
:RS21-C6

Western blot analysis using anti-RS21-C6 Ig and
whole-cell extracts from several mouse cell lines produced an intense band corresponding to a polypeptide
with a molecular mass of about 19 kDa in each lane
(Fig. 3A), although additional, non-specific bands were
detected in some lanes. We then transfected the mouse
B cell lymphoma line A20 individually with two plasmids to express non-tagged recombinant RS21-C6, and
independently transfected vector controls without
inserts. After incubation for 24 h, whole-cell extracts
were prepared and subjected to western blot analysis
using anti-RS21-C6. We detected a band with a very
intense signal that corresponded to a size of approximately 19 kDa in each of the samples overexpressing
RS21-C6. We detected a band with identical mobility
but a weak signal in each of the vector control samples, indicating that anti-RS21-C6 specifically reacts

pIRES2-EGFP

TrxA-RS21-C6
protein
was

expressed
from
pET32a(+):RS21-C6 and was used as an antigen to
prepare anti-RS21-C6 rabbit serum. Western blot analysis, using affinity-purified anti-RS21-C6 Ig, showed
that endogenous RS21-C6 also binds to both ITP- and
ATP-agarose (Fig. 2B, lower panel). Glyceraldehyde3-phosphate dehydrogenase (GAPDH), detected in
band 2, bound to the negative control deaminated
agarose, as well as to ITP-agarose, as shown by western
blot analysis with anti-GAPDH Ig, suggesting that the
binding of GAPDH was not nucleotide-specific (data
not shown).
The non-tagged recombinant RS21-C6 protein was
purified to a nearly homogeneous state by SDS–
PAGE to analyze its enzyme activity (Fig. 2D). Its
molecular mass was estimated, based on its SDS–
PAGE mobility, as approximately 19 kDa, which is
almost identical to the calculated molecular weight of
18 783.

FEBS Journal 276 (2009) 1654–1666 ª 2009 The Authors Journal compilation ª 2009 FEBS

1657


A mammalian dCTPase that prefers 5-iodocytosine

A

0.08

M. Nonaka et al.

with both recombinant RS21-C6 and mouse RS21-C6
that is endogenous to A20 cells (Fig. 3B). We then
analyzed the intracellular localization of the RS21-C6
protein. Nuclear, mitochondrial and cytosolic fractions
were prepared from mouse liver. Western blot analysis
of each fraction revealed that RS21-C6 is exclusively
located in the cytosol (Fig. 3C). Finally, we examined
the expression levels of RS21-C6 mRNA by real-time
quantitative PCR, and found that RS21-C6 is ubiquitously expressed and that expression was highest in the
liver and heart, and to a lesser extent the salivary
gland (Fig. 3D).

dCTP + buffer

0.04

272 nm (AU)

0.00
0.08

dCTP + RS21-C6

0.04
0.00
0.08

dCMP standard

0.04

Nucleoside triphosphate pyrophosphohydrolase
activity of RS21-C6 protein

0.00
0

80
60
40
20
0

10

20
30
40
Temperature (°C)

50

60

PIPES-Na
Tris-HCl
AMPD-HCl

80
60
40
20
0

6

7

8
pH

9

10

Mg2+
Mn2+

100

Product (%)

D

0

100

Product (%)

C

50

100

Product (%)

B

40
10
20
30
Retention time (min)

80
60
40
20
0

0

20

40
60

80
Metal ion concentration (mM)

100

Product (%)

E 100
80
60
40

NaCl
KCl

20
0

1658

In a preliminary analysis, using canonical nucleotides,
purified RS21-C6 protein showed strong pyrophosphohydrolase activity on dCTP, producing dCMP
(Fig. 4A, middle panel). We also analyzed the reaction
product using BIOMOL GREEN reagent (Enzo Biochem, Inc., New York, NY, USA) and detected no free
phosphate, indicating that RS21-C6 hydrolyzes dCTP
to dCMP and pyrophosphate (data not shown). Next,
we analyzed the optimal conditions for the pyrophosphohydrolase activity of RS21-C6 using dCTP as a
substrate. RS21-C6 showed a temperature-dependent
increase of activity up to 60 °C (Fig. 4B). RS21-C6
demonstrated strongest activity at pH 9.5, the highest

pH analyzed here (Fig. 4C). The divalent metal cation
requirements of RS21-C6 were tested using MgCl2 and
MnCl2. No activity was detected in reactions without
added metals, and maximum activity was measured in
reactions containing 100 mm MgCl2. At 100 mm,
MnCl2 did not support full activity (Fig. 4D). RS21-C6
showed the same activity with various concentrations
of KCl between 0 and 1000 mm (Fig. 4E). NaCl moderately reduced RS21-C6 activity. Based on these
results and in view of physiological conditions, we

0

200

400
600
800
Salt concentration (mM)

1000

Fig. 4. dCTP pyrophosphohydrolase activity of RS21-C6 protein. (A)
Hydrolysis of dCTP to dCMP by RS21-C6. Substrate dCTP (300 lM)
was incubated for 20 min in reaction buffer supplemented with
50 nM of purified RS21-C6 protein. Reaction products were analyzed by HPLC (middle panel), and were compared with substrate
dCTP incubated in the reaction buffer without RS21-C6 (upper
panel), and with standard dCMP prepared in the reaction buffer
(lower panel). The dependency of RS21-C6 activity on temperature
(B), buffer pH (C), divalent metal cations (D) and salts (E) was analyzed. The amounts of dCMP produced were measured and are
shown as percentages of the highest value. Each point and error

bar indicates the mean and standard deviation of three reactions.

FEBS Journal 276 (2009) 1654–1666 ª 2009 The Authors Journal compilation ª 2009 FEBS


M. Nonaka et al.

A mammalian dCTPase that prefers 5-iodocytosine

Table 2. Kinetic parameters for nucleoside triphosphate pyrophosphohydrolase activity of RS21-C6 protein.
Km (lM)

ATP
GTP
CTP
UTP
ITP
2-OH-ATP
dATP
dGTP
dCTP
dTTP
dUTP
dITP
2-OH-dATP
2-Cl-dATP
8-oxo-dGTP
5-OH-dCTP
5-Me-dCTP
5-Br-dCTP

5-I-dCTP
5-Cl-dCTP
5-F-dUTP
dCDP
dADP

No activity
No activity
529
No activity
No activity
No activity
118
No activity
44.1
407
484
No activity
169
107
No activity
164
48.5
21.7
3.9
50.0
304
No activity
No activity

kcat (s)1)

kcat ⁄ Km (s)1ỈM)1)

0.31

156

0.53

4490

1.37
0.76
0.76

31 100
1870
1570

1.39
1.57

8220
14 700

0.87
1.33
1.49
0.94

1.50
0.58

5300
27 400
68 700
241 000
30 000
1900

performed further analysis of RS21-C6 activity under
the conditions described in Experimental procedures.
The substrate concentration was set at 10, 30, 100 or
1000 lm, except for that of 5-I-dCTP, which was set at

A

2 µM
3.8 µM
7.8 µM
15.6 µM
31.3 µM

[Product] (µM)

10

5

0

1

2
3
Time (min)

4

5

0

5

10
Time (min)

15

20

C

3

rate)–1 (s·µM–1)

2

50

(Reaction

Reaction rate (µM·min–1)

Fig. 5. Kinetic analysis of the reaction of
RS21-C6 with 5-I-dCTP. The concentration
of 5-I-dCTP substrate was set at 2, 3.9, 7.8,
15.6, 31.3, 62.5, 125, 250, 500 or 1000 lM.
(A) Enzyme reaction curves. (B) Michaelis–
Menten plot using the results at substrate
concentrations from 0 to 62.5 lM, showing
the saturation pattern. (C) A Lineweaver–
Burk plot prepared using all the results.

30

0

0

B

62.5 µM
125 µM
250 µM
500 µM
1000 µM

60
[Product] (µM)

Substrate

2, 3.9, 7.8, 15.6, 31.3, 62.5, 125, 250, 500 or 1000 lm,
because the Km value for 5-I-dCTP is < 10 lm. The
Km and kcat values at 50 nm RS21-C6 for various
nucleotides were determined from Lineweaver–Burk
plots, and are shown in Table 2. Except for CTP,
RS21-C6 did not hydrolyze the analyzed ribonucleotides, even ITP. For the analyzed deoxynucleotides, the kcat ⁄ Km values for dCTP, dATP and
dTTP were 3.11 · 104, 4.49 · 103 and 1.87 · 103, and
RS21-C6 did not hydrolyze dGTP and dITP.
Thus, RS21-C6 shows a preference for cytosine base
and deoxyribose. Among the base-modified nucleotides, 5-I-dCTP had an eightfold higher kcat ⁄ Km value
compared to its corresponding unmodified deoxynucleotide, dCTP. We analyzed the hydrolysis of
dCTP by RS21-C6 in the presence of ITP, and found
that ITP did not inhibit it even at 500 lm (data not
shown).
Enzyme reaction curves, a Michaelis–Menten plot
and a Lineweaver–Burk plot for 5-I-dCTP are shown
in Fig. 5. The oligomeric structure of RS21-C6,
together with the conformational flexibility of its active
sites, suggests cooperativity between the sites and allosteric regulation. However, we obtained a typical
Michaelis–Menten-type saturation curve rather than
the sigmoid curve typical of non-Michaelis–Mententype reactions. This result indicates that the RS21-C6
tetramer does not show any cooporative binding of
substrates to the multiple active sites under the present
conditions.

1

25

0
0

10

20 30 40 50
[Substrate] (àM)

FEBS Journal 276 (2009) 16541666 ê 2009 The Authors Journal compilation ê 2009 FEBS

60

70

0.2

0
0.2
[Substrate]1 (àM1)

0.4

1659


A mammalian dCTPase that prefers 5-iodocytosine

M. Nonaka et al.

Discussion
We have identified ITPase, a known ITP-hydrolyzing
enzyme, as an ITP-binding protein, using ITP-agarose.
This indicates that our screening technology is a valid
approach for identification of novel proteins that
bind various modified nucleotides. RS21-C6, another
protein identified in this study, was an ITP-binding
protein which can also bind ATP. Because of its
homology in amino acid sequence to known nucleoside
triphosphate (NTP) pyrophosphohydrolases, including
bacterial MazG proteins [8], we analyzed its catalytic
activity for hydrolyzing canonical nucleotides. We
showed that RS21-C6 is a deoxynucleoside triphosphate pyrophosphohydrolase that prefers dCTP. In
mammalian cells, no other protein has been reported
as a pyrimidine dNTP-specific pyrophosphohydrolase,
although Orf135 and iMazG, a novel bacterial MazG
protein, have been reported as bacterial dCTPases
[11,12]. The likely biological unit of RS21-C6 is a tetramer [10], suggesting that the multivalency of the
RS21-C6 tetramer may stabilize its interaction with
ITP immobilized on agarose beads. This may explain
why RS21-C6, which is not an ITP-specific protein,
has been identified as an ITP-binding protein.
It has been shown that the substrate-binding pocket
of RS21-C6 comprises several residues, including
His38, Trp47, Trp73, Tyr102, Glu63, Glu66, Glu95
and Asn98, that the nitrogenous base and deoxyribose
of 5-methyl-dCTP are located in a hydrophobic cavity,

and that the phosphate groups interact with the four
electronegative amino acid residues [10]. Among the
known all-a-NTP pyrophosphohydrolases, we found
that residue Asn125 in RS21-C6 is conserved in various
dUTPases
[Campylobacter jejuni
dUTPase
(CjdUTPase), Leishmania major dUTPase, Trypanosoma cruzi dUTPase (TcdUTPase)], dCTPases (enterobacteria phage T2 dCTPase, enterobacteria phage T4
dCTPase, bacteriophage RB15 dCTPase), and iMazG
[8,12]. Moroz et al. and Harkiolaki et al. analyzed the
crystal structures of CjdUTPase [13] and TcdUTPase
[14], respectively, with their substrates. They showed
that Asn179 of CjdUTPase and Asn201 of TcdUTPase, the residues corresponding to Asn125 of RS21C6, bind to the 2¢-deoxyribose moiety of substrates.
This residue is not conserved in either the HisE family,
which has phosphoribosyl-ATP pyrophosphatase activity (E. coli HisIE, Corynebacterium glutamicam HisE,
Pyrococcus furiosus HisIE, Saccharomyces cerevisiae
HIS4, Arabidopsis thaliana HisIE), or the MazG family
(SSO12199, E. coli MazG, Thermotoga maritime
MazG, Bacillus subtilis YABN, Streptomyces cacaoi
YBL1). Because enzymes in the HisE and MazG fami1660

lies but not those of the dUTPase, dCTPase, iMazG
and RS21-C6 families hydrolyze ribonucleotides, we
suggest that the Asn125 residue in RS21-C6 may be
involved in the preference for deoxyribose sugar.
Wu et al. [10] did not mention an interaction between
Asn125 and the substrate 5-methyl-dCTP in their analysis of the crystal structure of RS21-C6 with the substrate. They used the core fragment of RS21-C6
(RSCUT: residues 21–126) in which the Asn125 residue is located very close to the C-terminal end. Therefore, Asn125 might not be appropriately located in the
truncated molecule.
Recognition of the cytosine base by RS21-C6

appears to be supported by the His38 residue in
helix 1, which forms a hydrogen bond with the O2 of
the cytosine base [10]. This corresponds to the glutamine residues in the first helix of dUTPases (Gln14 of
CjdUTPase or Gln22 of TcdUTPase), which form a
hydrogen bond with the O2 of a uracil base [8]. It has
been shown that His58 of CjdUTPase [13] and Trp61
of TcdUTPase [14] form a hydrogen bond with O4 of
the uracil base, and this appears to be involved in their
discrimination of uracil from cytosine. These residues
are not conserved in RS21-C6, supporting its lower
affinity for dUTP in comparison to dCTP as revealed
in the present study.
RS21-C6 has essentially similar affinities towards
both dCTP and 5-methyl-dCTP, indicating that there
may not be a specific residue that recognizes the
methyl group at the C5 position. However, halogenation at the C5 position, particularly iodination, significantly increased its affinity to RS21-C6. It is possible
that one of the residues that form the substrate-binding pocket may recognize the iodine at the C5 position. An analysis of the crystal structures of RS21-C6
complexed with 5-I-dCTP will help to delineate the
structural basis for its substrate recognition.
What is the biological role of dCTPase in mammalian cells? RS21-C6 hydrolyzes dCTP and produces
dCMP. In mammalian cells, deamination of dCMP by
dCMP deaminase is the most important pathway for
dUMP production. dUMP is then converted to dTMP
by thymidylate synthase [15–18], and dTMP is converted to dTTP by a two-step phosphorylation reaction. The activity of dCMP deaminase is positively
regulated by dCTP, and negatively by dTTP [19].
Hence, the conversion of dCTP to dTTP appears to be
a key pathway for regulating the ratio of dCTP ⁄ dTTP
in the nucleotide pool. Based on pulse-chase experiments using [5-3H]cytidine, Bianchi et al. [20] estimated
that dCTP was incorporated into DNA at a rate of
16 pmolỈmin)1 in rapidly growing mouse 3T6 cells,

and that dTMP was formed from free dCMP at a rate

FEBS Journal 276 (2009) 1654–1666 ª 2009 The Authors Journal compilation ª 2009 FEBS


M. Nonaka et al.

of 10 pmolỈmin)1. A defect in dCMP deaminase in
hamster fibroblasts was reported to cause an imbalanced dCTP ⁄ dTTP ratio, and to mildly affect the
fidelity of DNA replication [21]. dCTP is not an
effector molecule that allosterically controls ribonucleotide reductase (RNR). Therefore, cells must have
other mechanisms to regulate the cytosolic dCTP
concentration. In this regard, dCTP pyrophosphohydrolase plays an important role in avoiding accumulation of excess dCTP and supplying sufficient levels
of dCMP, an upstream precursor of de novo synthesis
of dTTP.
Expression of RS21-C6 mRNA was detected in all
tissues examined in this study, and was particularly high
in the liver and heart. These data suggest that RS21-C6
plays a role in both proliferating and non-proliferating
cells. Additionally, we found that RS21-C6 was localized to the cytosol. In mammalian cells, two types of
RNR, R1 ⁄ R2 RNR and R1 ⁄ p53R2 RNR, regulate the
synthesis of dNTPs for DNA replication [22,23]. In
proliferating cells, the R1 ⁄ R2 RNR complex, which
consists of R1 and R2 subunits, is localized to the cytoplasm and supplies deoxynucleotides for nuclear DNA
synthesis [22]. Even in non-proliferating cells, dNTPs
are necessary for mitochondria DNA replication. These
cells express the p53R2 subunit instead of the R2
subunit in the cytoplasm. Mitochondrial DNA depletion caused by mutations in the RRM2B gene, which
encodes the p53R2 subunit, demonstrates that the
R1 ⁄ p53R2 RNR complex plays a critical role in dNTP

supply for mitochondrial DNA replication [24]. Similarly to the RS21-C6 gene, expression of the RRM2B
gene is found in many tissues, in contrast to the R2
subunit, which is undetectable in the heart, brain and
muscle [25]. dTMP and thymidine synthesized in the
cytoplasm are imported into the mitochondria, phosphorylated by mitochondrial enzymes and used for
mitochondrial DNA replication [26]. On the other
hand, it has been reported that dCTP transport activity
exists in human mitochondria [27]. These reports raise
the possibility that the cytosolic concentration of dCTP
and dTMP influences the balance of mitochondrial
deoxypyrimidine nucleotide pools (Fig. 6A).
Several members of the NTP pyrophosphohydrolase
family have also been shown to eliminate non-canonical nucleotides from the intracellular NTP pool.
Here, we have shown that RS21-C6 has the highest
kcat ⁄ Km value for the modified deoxynucleotide 5-IdCTP of the various nucleotides examined, including
its corresponding canonical deoxynucleotide, dCTP.
Our data indicate that 5-I-dCTP or its analogs might
be true substrates of RS21-C6. It is unlikely that the
intracellular level of ITP prevents RS21-C6 activity,

A mammalian dCTPase that prefers 5-iodocytosine

A

B

Fig. 6. Models of two biological roles of RS21-C6 protein. (A)
RS21-C6 may supply dCMP as an upstream precursor of de novo
synthesis of dTTP. CDP is reduced to dCDP by ribonucleotide
reductase (RNR). dCTP is synthesized by phosphorylation of dCDP.

Excess dCTP is hydrolyzed to dCMP by RS21-C6. dCMP is
converted to dTMP by dCMP deaminase (DCTD) and thymidylate
synthase (TS). dTMP is converted to dTTP by two steps of
phosphorylation. RS21-C6 plays a role in regulation of the
dCTP ⁄ dTTP ratio in dNTP pools for nuclear and mitochondrial DNA
synthesis. (B) RS21-C6 may hydrolyze 5-I-dCTP or its structurally
related nucleotides to prevent inappropriate CpG methylation.

because 500 lm ITP did not inhibit hydrolysis of
dCTP by RS21-C6 in our preliminary experiment.
5-I-dCTP is a derivative molecule of dCTP, in which
C5 of the cytosine base is iodinated. Halogenation of
cytosine at C5, including chlorination and bromination, has been shown to occur under physiological conditions. Such halogenation occurs in the presence of
either myeloperoxidase or eosinophil peroxidase, which
are produced by phagocytic cells [28,29]. In these reactions, hypohalous acids, inter-halogen and haloamines
are candidate intermediate molecules that can diffuse
across plasma membranes into the cytoplasm and
halogenate cytosine in cells. Kawai et al. [30] have previously detected halogenated cytosine in DNA in
inflamed tissues. Their results indicate in vivo halogenation of cytosine, but do not provide direct evidence
for intra-cellular halogenation. Valinluck et al.
reported that 5-iodocytosine, 5-bromocytosine and
5-chlorocytosine, at a CpG site of DNA, can mimic
5-methylcytosine and induce inappropriate DNA methyltransferase 1-dependent methylation within the CpG
sequence [31,32]. The induction effect of 5-iodocytosine
was the greatest among the 5-halogenated cytosines,
and was even better than that of 5-methylcytosine.
5-halogenated cytosine, at a CpG site, may enhance
the binding of methyl-CpG-binding protein 2 [33].
These reports suggest that the 5-halogenated dCTP
generated in chronic inflamed tissues might be incorporated into promoter regions of important genes, such

as tumor-suppressor genes, and induce their silencing
through inappropriate CpG methylation of the
promoter regions or by binding of methyl-CpG-binding protein 2, resulting in tumorigenesis. RS21-C6 may

FEBS Journal 276 (2009) 1654–1666 ª 2009 The Authors Journal compilation ª 2009 FEBS

1661


A mammalian dCTPase that prefers 5-iodocytosine

M. Nonaka et al.

prevent such deleterious effects by hydrolyzing modified deoxynucleotides including 5-I-dCTP or its structurally related molecules (Fig. 6B).
5-Methyl-dCTP is also a potential inducer of inappropriate CpG methylation by DNA methyltransferase 1 when it is incorporated in a CpG site. Although
5-methyl-dCMP is a poor substrate for mammalian
nucleoside monophosphate kinases [34,35], 5-methyldCMP has been shown to be incorporated into the
DNA of Chinese hamster ovary cells with low dCMP
deaminase activity [36]. Using preliminary comparative modeling, Moroz et al. [8] found that 5-methyldCTP is a potential substrate candidate of RS21-C6,
and 5-methyl-dCTP hydrolyzing activity of RS21-C6
was recently reported by Wu et al. [10]. In the present
study, we show that the kcat ⁄ Km value of RS21-C6
for 5-methyl-dCTP is almost the same as that for
dCTP. More detailed comparisons of the enzyme
kinetics for dCTP and 5-methyl-dCTP are necessary
to establish the physiological role of RS21-C6 for
5-methyl-dCTP.
XTP3TPA (gi|13129100) is a human homolog of
RS21-C6. Our study demonstrated the substrate preference of RS21-C6 for deoxynucleotides, cytosine bases
and iodination at C5 of cytosine. These data suggest

that a defect of human XTP3TPA might cause a
nuclear DNA replication block or mitochondrial DNA
depletion, as a result of an imbalanced dCTP ⁄ dTTP
ratio. Moreover, XTP3TPA might be involved in
tumorigenesis in chronically inflamed tissues, as a
result of accumulation of modified deoxynucleotides.

Experimental procedures
Synthetic oligonucleotides
The synthetic oligonucleotides listed below, used as PCR
primers, were purchased from Genenet Co. Ltd (Fukuoka,
Japan), Sigma-Aldrich Japan (Tokyo, Japan) and Takara
Bio Inc. (Ohtsu, Japan): 5¢Nde-mMAZG, 5¢-ATACATATG
TCCACGGCTGGTGACGGTGAGCG-3¢; 5¢Nco-mMAZG,
5¢-ATACCATGGCCTCCACGGCTGGTGACGGTGAGC3¢; 3¢mMAZG-BamHI, 5¢-ATAGGATCCTTATGTGGAAG
CCTGGTCTCTC-3¢; RS21-C6 forward, 5¢-GCGAGCTGGC
AGAACTCTTC-3¢; RS21-C6 reverse, 5¢-TTTGGTGGCCA
TGCTTGA-3¢; 18S rRNA forward, 5¢-AGGATGTGAAGG
ATGGGAAG-3¢; 18S rRNA reverse, 5¢-ACGAAGGCCCC
AAAAGTG-3¢.

Nucleotides
The nucleotides used as substrates for RS21-C6 were purchased from Sigma-Aldrich (St Louis, MO, USA), TriLink

1662

Biotechnologies Inc. (San Diego, CA, USA) or Jena Bioscience (Jena, Germany).

Preparation of mouse thymocyte extract
Five-week-old C57BL ⁄ 6J male mice (Clea Japan, Tokyo,

Japan) were dissected under pentobarbital anesthesia
(75 mgỈkg)1, intraperitoneally), and killed by blood drainage from abdominal vessels. The thymus was removed
and ground between glass slides to prepare thymocyte suspensions. Thymocytes (6 · 108) were suspended in 3 mL of
lysis buffer {25 mm 2-[4-(2-Hydroxyethyl)-1-piperazinyl]
ethanesulfonate-Na pH 7.2, 150 mm NaCl, 60 mm MgCl2,
0.05% Nonidet P-40 (Nacalai tesque, Kyoto, Japan), 1 mm
dithiothreitol, 1% protease inhibitor cocktail (Nacalai
Tesque)}, and were disrupted by sonication. Cell lysates
were then centrifuged at 100 000 g for 30 min. The supernatant was collected as the thymocyte extract. Handling
and killing of all animals used in this study were in accordance with the national prescribed guidelines, and ethical
approval for the studies was granted by the Animal Experiment Committee of Kyushu University (Fukuoka, Japan).

Preparation of ITP-agarose
ATP-agarose (adenosine 5¢-triphosphate agarose, SigmaAldrich; 25 lL bed volume) or 25 lL agarose carrier matrix
were washed for 2 min twice in 1 mL 3 m sodium acetate
buffer (pH 3.2), and then suspended in 150 lL of deamination buffer [100 mm sodium nitrite (NaNO2), 500 mm
sodium thiocyanate (NaSCN), 3 m sodium acetate (NaCH3COO) pH 3.2], and incubated at 37 °C for 60 min. Then
each agarose aliquot was washed for 2 min twice in 1 mL
of water and used as ITP-agarose or as deaminated agarose, respectively. To confirm the nucleotide immobilized
on each agarose, the base moiety was excised by incubation
in 1 m HCl at 100 °C for 1 h, and analyzed by HPLC after
neutralization and filtration.

Purification and identification of ITP-binding
proteins
ITP-agarose was resuspended in 450 lL thymocyte extract
in a microtube, and mixed by vertical rotation for
30 min. Agarose that had been subjected to deamination
was used as the negative control. Each agarose sample
was then washed for 1 min three times in 1 mL of lysis

buffer without protease inhibitor. All procedures were performed at 4 °C. Each agarose was then resuspended in
40 lL of 2· SDS sampling buffer (Sigma-Aldrich), and
incubated at 95 °C for 5 min. The supernatant was collected after centrifugation at 140 g for 5 s at room
temperature. Proteins in each sample were separated by
SDS–PAGE and analyzed by LC-MS ⁄ MS as described

FEBS Journal 276 (2009) 1654–1666 ª 2009 The Authors Journal compilation ª 2009 FEBS


M. Nonaka et al.

previously [37]. Collision-induced dissociation spectra were
acquired and compared with those in the International
Protein Index (IPI version 3.16; European Bioinformatics
Institute Hinxton, UK) using the MASCOT search engine
(Matrix Science, Boston, MA, USA). The high-scoring
peptide sequences (MASCOT score > 45) assigned by
MASCOT were manually confirmed by comparison with
the corresponding collision-induced dissociation spectra.
Finally we selected as candidate proteins those proteins for
which multiple peptides were identified in this analysis.

Isolation of RS21-C6 cDNA
RS21-C6 cDNA fragments, RS21-C6(NdeI/BamHI) and
RS21-C6(NcoI/BamHI) were amplified by PCR from a
mouse fibroblast cell line, NIH/3T3, prepared as described
previously [38],
using primer sets 5¢Nde-mMAZG/
3¢mMAZG-BamHI
and

5¢Nco-mMAZG/3¢mMAZGBamHI. Amplified fragments were subcloned into pT7Blue-2
T-vector (Novagen, Madison, WI, USA) to generate
the plasmids pT7Blue2T:RS21-C6(NdeI ⁄ BamHI) and
pT7Blue2T:RS21-C6(NcoI ⁄ BamHI), respectively.

Construction of expression plasmids
Plasmids pET3a:RS21-C6, pET32a(+):RS21-C6, pcDNA3.1hyg(+):RS21-C6 and pIRES2-EGFP:RS21-C6 were
prepared by inserting DNA fragments containing the
ORF of RS21-C6 cDNA into the NdeI ⁄ BamHI site of
pET3a (Novagen), the NcoI ⁄ BamHI site of pET32a(+)
(Novagen) or the XhoI ⁄ BamHI site of pcDNA3.1hyg(+)
(Invitrogen, Carlsbad, CA, USA) or into the XhoI/BamHI
site of pIRES2-EGFP (Clontech Laboratories Inc., Mountain View, CA, USA), respectively.

Expression and purification of recombinant
RS21-C6 protein
Expression of recombinant RS21-C6, without any tag
sequence, was induced in E. coli BL21-CodonPlus(DE3)RIL cells (Stratagene, La Jolla, CA, USA) transformed
with pET3a:RS21-C6, as described previously [39]. Cells
were suspended in buffer A (50 mm Tris ⁄ HCl pH 8.0,
100 mm NaCl, 5 mm EDTA, 5 mm 2-mercaptoethanol,
1 mm phenylmethanesulfonyl fluoride, 1 lgỈmL)1 pepstatin A, 1 lgỈmL)1 chymostatin, 1 lgỈmL)1 leupeptin), disrupted by sonication, and clarified by centrifugation at
20 000 g for 30 min at 4 °C. Proteins in the supernatants
were precipitated using ammonium sulfate (40–50% saturation), and re-dissolved in buffer B (50 mm Tris ⁄ HCl
pH 8.0, 100 mm NaCl, 5% glycerol, 5 mm MgCl2, 5 mm
2-mercaptoehanol). Dissolved samples were dialyzed three
times against 1 L of buffer B and loaded onto HiTrap-Q
HP anion exchange columns (GE Healthcare, Chalfont

A mammalian dCTPase that prefers 5-iodocytosine

St Giles, UK) equilibrated with buffer C (50 mm Tris ⁄ HCl
pH 8.0, 50 mm NaCl, 5% glycerol, 5 mm MgCl2, 5 mm
2-mercaptoehanol). Binding proteins were eluted using a
linear gradient of NaCl (50–1000 mm). Fractions containing
RS21-C6 protein were applied onto Superdex 75 HR10 ⁄ 30
size exclusion columns (Sigma-Aldrich) equilibrated with
buffer B. Fractions containing RS21-C6 were then loaded
onto a MonoQ HR5 ⁄ 5 anion exchange column (GE Healthcare) equilibrated with buffer C, and eluted using a linear
gradient of NaCl (50–1000 mm). Fractions containing
RS21-C6 were loaded sequentially onto HiTrap-S HP and
HiTrap heparin columns (GE Healthcare), and flow-through
fractions were collected. RS21-C6 protein was concentrated
using HiTrap-Q columns, dialyzed against buffer D (50 mm
Tris ⁄ HCl pH 8.0, 100 mm NaCl, 50% glycerol, 5 mm
MgCl2, 1 mm dithiothreitol), and stored at )30 °C as purified RS21-C6 protein.

Nucleotide-hydrolyzing assay with RS21-C6
protein
Substrate nucleotides were incubated in 18 lL of reaction
buffer [50 mm Tris ⁄ HCl pH 8.0, 100 mm KCl, 5 mm
MgCl2, 100 lgỈmL)1 BSA (New England Biolabs Inc.,
Ipswich, MA, USA), 1 mm dithiothreitol] at 30 °C for
10 min. Then, 2 lL of 500 nm RS21-C6 protein, in reaction buffer, was added to the reaction and further incubated at 30 °C for 0–30 min. Sample solutions were
mixed with 10 lL of ice-cold 50 mm EDTA to stop the
reactions, clarified by centrifugation at 9000 g for 10 min
at 4 °C, and separated on SunFire C18 5 lm 4.6 ·
250 mm columns (Waters, Milford, MA, USA) or TSK
gel DEAE-2SW columns (Tohso, Tokyo, Japan) using an
Alliance photodiode array HPLC system (Waters), at a

flow rate of 1 mLỈmin)1 with HPLC buffer 1 (0.1 m
potassium phosphate pH 6.0, 5% methanol) or HPLC
buffer 2 (75 mm sodium phosphate pH 6.0, 20% acetonitrile, 0.4 mm EDTA). The amounts of nucleotide were
quantified by UV absorption.

Anti-RS21-C6 Ig
An antigen, TrxA-RS21-C6 protein, was expressed in
E. coli BL21-CodonPlus (DE3)-RIL cells transformed with
pET32a(+):RS21-C6, and purified by metal affinity chromatography with TALON beads (Clontech). Preparation of
rabbit anti-TrxA-RS21-C6 serum and affinity purification
of anti-RS21-C6 Ig were performed as described previously
[39,40].

Western blot
Western blot analysis using antibodies against anti-RS21C6, Lamin B (Santa Cruz Biotechnology Inc., Santa Cruz,

FEBS Journal 276 (2009) 1654–1666 ª 2009 The Authors Journal compilation ª 2009 FEBS

1663


A mammalian dCTPase that prefers 5-iodocytosine

M. Nonaka et al.

CA, USA), HSP60 (LK-1; StressGen Biotechnologies
Corp., Victoria, Canada), GAPDH (Chemicon International Inc., Temwcula, CA, USA) or using anti-ITPase
serum was performed as described previously [5,39].

Cell culture

Mouse embryonic fibroblast cells were prepared as
described previously [41]. Mouse embryonic fibroblast cells,
NIH ⁄ 3T3 cells and BALB ⁄ 3T3 cells were grown in DMEM
(Invitrogen) supplemented with 10% fetal bovine serum,
100 unitsỈmL)1 penicillin and 100 lgỈmL)1 streptomycin.
A20, BW5147.3, WEHI231 and J774A.1 cells were obtained
from the American Type Culture Collection (ATCC,
Manassas, VA, USA) and maintained in RPMI-1640 (Invitrogen) supplemented with 10% fetal bovine serum,
100 unitsỈmL)1 penicillin, 100 lgỈmL)1 streptomycin and
50 lm 2-mercaptoethanol. Transfection of plasmid DNA
into A20 cells was performed using an MP-100 microporator (Digital Bio Technology, Seoul, Korea).

Fractionation of mouse liver cells
After fasting overnight, a six-month-old C57BL ⁄ 6J female
mouse (Clea Japan) was dissected under pentobarbital anesthesia (75 mgỈkg)1, intraperitoneally). After draining blood
by cutting abdominal vessels, the liver was removed and
rinsed with 0.25 m sucrose at 4 °C. The following procedures were all performed at 4 °C. The liver was homogenized in a Teflon Potter–Elvehjem homogenizer, and
fractionated by centrifugation at 700 g for 10 min. The pellet was rinsed with 0.25 m sucrose, and resuspended in
0.25 m sucrose as a nuclear fraction. The supernatant (postnuclear supernatant) was centrifuged at 10 000 g for
10 min. The pellet was rinsed with 0.25 m sucrose and suspended in 0.25 m sucrose. The resulting solution was used
as the mitochondrial fraction. The supernatant obtained
after centrifugation at 10 000 g was re-centrifuged at
110 000 g for 60 min. The resulting supernatant was used
as the cytosolic fraction.

Real-time quantitative PCR
A 10-month-old C57BL ⁄ 6J male mouse (Clea Japan) was
dissected under pentobarbital anesthesia (75 mgỈkg)1, intraperitoneally). After transcardiac perfusion with saline,
tissues were removed. Total RNA samples from mouse
tissues except the thyroid gland were prepared using an

Isogen kit (Nippon Gene, Tokyo, Japan) according to the
manufacturer’s instructions. Thyroid gland total RNA was
purchased from Clontech. Total RNAs were treated using
RNase-free DNase I (Boehringer Mannheim, Mannheim,
Germany) at 37 °C for 60 min. cDNA was synthesized
from 2 lg total RNA using a high-capacity cDNA reverse
transcription kit (Applied Biosystems, Foster City, CA,

1664

USA) using random primers in a total volume of 20 lL.
Real-time quantitative PCR was performed to measure the
levels of RS21-C6 mRNA and 18S rRNA using an ABI
Prism 7000 sequence detection system (Applied Biosystems)
with 10 ng cDNA, 50 nm primers and Power SYBR Green
PCR Master Mix (Applied Biosystems) in a total volume of
25 lL. The PCR reaction was performed as follows: a
single cycle of 50 °C for 2 min, a single cycle of 95 °C for
10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C
for 1 min. The primers were designed using primer express
software (Applied Biosystems) and their sequences are
described above. The primers spanned intron and exon
junctions. The specificity of the PCR products was established by dissociating curve analysis, and by running the
products on a 2% agarose gel to verify their size. The
RS21-C6 levels are expressed relative to the 18S rRNA
levels. Serially diluted cDNA was used to obtain a standard
curve for each transcript.

Acknowledgements
This work was supported by a Grant-in-Aid for

Scientific Research (B) from the Japan Society for the
Promotion of Science (grant number 19390114). We
thank Drs Masaki Matsumoto, Mizuki Ohno and Eiko
Ohta (Medical Institute of Bioregulation, Kyushu University) for helpful discussions, and Mizuho Oda,
Emiko Fujimoto and Masumi Ohtsu (Laboratory for
Technical Support of Medical Institute of Bioregulation, Kyushu University) and Kazumi Asakawa (Medical Institute of Bioregulation, Kyushu University) for
technical assistance.

References
1 Nakabeppu Y, Tsuchimoto D, Furuichi M & Sakumi
K (2004) The defense mechanisms in mammalian cells
against oxidative damage in nucleic acids and their
involvement in the suppression of mutagenesis and cell
death. Free Radic Res 38, 423–429.
2 Galperin MY, Moroz OV, Wilson KS & Murzin AG
(2006) House cleaning, a part of good housekeeping.
Mol Microbiol 59, 5–19.
3 Nakabeppu Y (2001) Molecular genetics and structural
biology of human MutT homolog, MTH1. Mutat Res
477, 59–70.
4 Tsuzuki T, Egashira A & Kura S (2001) Analysis of
MTH1 gene function in mice with targeted mutagenesis.
Mutat Res 477, 71–78.
5 Behmanesh M, Sakumi K, Tsuchimoto D, Torisu K,
Ohnishi-Honda Y, Rancourt DE & Nakabeppu Y
(2005) Characterization of the structure and expression
of mouse Itpa gene and its related sequences in the
mouse genome. DNA Res 12, 39–51.

FEBS Journal 276 (2009) 1654–1666 ª 2009 The Authors Journal compilation ª 2009 FEBS

M. Nonaka et al.

6 Lin S, McLennan AG, Ying K, Wang Z, Gu S, Jin H,
Wu C, Liu W, Yuan Y, Tang R et al. (2001) Cloning,
expression, and characterization of a human inosine
triphosphate pyrophosphatase encoded by the itpa
gene. J Biol Chem 276, 18695–18701.
7 Li Y, Jiang DD, Jin C, Liu YF & Chen WF (2004)
[Prokaryotic expression, purification and preparation of
polyclonal antibody and immunohistochemistry analysis
of RS21-C6 molecule]. Beijing Da Xue Xue Bao 36,
268–271 (in Chinese).
8 Moroz OV, Murzin AG, Makarova KS, Koonin EV,
Wilson KS & Galperin MY (2005) Dimeric dUTPases,
HisE, and MazG belong to a new superfamily of
all-alpha NTP pyrophosphohydrolases with potential
‘house-cleaning’ functions. J Mol Biol 347,
243–255.
9 Wesenberg GE, Philips GN Jr, McCoy JG, Bitto E,
Bingman CA & Allard STM, Center for Eukaryotic
Structural Genomics (2005) X-ray structure of protein
from Mus musculus MM.29898. PDB ID: 2A3Q (http://
www.wwpdb.org/), doi:10.2210/pdb2a3q/pdb
10 Wu B, Liu Y, Zhao Q, Liao S, Zhang J, Bartlam M,
Chen W & Rao Z (2007) Crystal structure of RS21-C6,
involved in nucleoside triphosphate pyrophosphohydrolysis. J Mol Biol 367, 1405–1412.
11 O’Handley SF, Dunn CA & Bessman MJ (2001) Orf135
from Escherichia coli is a Nudix hydrolase specific for

CTP, dCTP, and 5-methyl-dCTP. J Biol Chem 276,
5421–5426.
12 Robinson A, Guilfoyle AP, Harrop SJ, Boucher Y,
Stokes HW, Curmi PM & Mabbutt BC (2007) A putative house-cleaning enzyme encoded within an integron
˚
array: 1.8 A crystal structure defines a new MazG subtype. Mol Microbiol 66, 610–621.
13 Moroz OV, Harkiolaki M, Galperin MY, Vagin AA,
Gonzalez-Pacanowska D & Wilson KS (2004) The
crystal structure of a complex of Campylobacter jejuni
dUTPase with substrate analogue sheds light on the
mechanism and suggests the ‘basic module’ for dimeric
d(C ⁄ U)TPases. J Mol Biol 342, 1583–1597.
14 Harkiolaki M, Dodson EJ, Bernier-Villamor V, Turkenburg JP, Gonzalez-Pacanowska D & Wilson KS (2004)
The crystal structure of Trypanosoma cruzi dUTPase
reveals a novel dUTP ⁄ dUDP binding fold. Structure 12,
41–53.
15 Jackson RC (1978) The regulation of thymidylate biosynthesis in Novikoff hepatoma cells and the effects of
amethopterin, 5-fluorodeoxyuridine, and 3-deazauridine.
J Biol Chem 253, 7440–7446.
16 Maley GF & Maley F (1963) Nucleotide interconversions. X. Deoxyribo- and ribonucleoside 5¢-phosphate
synthesis via a phosphotransferase reaction in chick
embryo extracts. Arch Biochem Biophys 101, 342–349.
17 Nicander B & Reichard P (1985) Evidence for the
involvement of substrate cycles in the regulation of

A mammalian dCTPase that prefers 5-iodocytosine

18

19

20

21

22

23

24

25

26

27

28

29

deoxyribonucleoside triphosphate pools in 3T6 cells.
J Biol Chem 260, 9216–9222.
Pontarin G, Ferraro P, Hakansson P, Thelander L,
Reichard P & Bianchi V (2007) p53R2-dependent ribonucleotide reduction provides deoxyribonucleotides in
quiescent human fibroblasts in the absence of induced
DNA damage. J Biol Chem 282, 16820–16828.
Ellims PH, Kao AY & Chabner BA (1983) Kinetic
behaviour and allosteric regulation of human deoxycytidylate deaminase derived from leukemic cells. Mol Cell
Biochem 57, 185–190.

Bianchi V, Pontis E & Reichard P (1986) Interrelations
between substrate cycles and de novo synthesis of pyrimidine deoxyribonucleoside triphosphates in 3T6 cells.
Proc Natl Acad Sci USA 83, 986–990.
Dare E, Zhang LH, Jenssen D & Bianchi V (1995)
Molecular analysis of mutations in the hprt gene of V79
hamster fibroblasts: effects of imbalances in the dCTP,
dGTP and dTTP pools. J Mol Biol 252, 514–521.
Tanaka H, Arakawa H, Yamaguchi T, Shiraishi K,
Fukuda S, Matsui K, Takei Y & Nakamura Y (2000)
A ribonucleotide reductase gene involved in a p53dependent cell-cycle checkpoint for DNA damage.
Nature 404, 42–49.
Engstrom Y & Rozell B (1988) Immunocytochemical
evidence for the cytoplasmic localization and differential
expression during the cell cycle of the M1 and M2
subunits of mammalian ribonucleotide reductase.
EMBO J 7, 1615–1620.
Bourdon A, Minai L, Serre V, Jais JP, Sarzi E, Aubert
S, Chretien D, de Lonlay P, Paquis-Flucklinger V,
Arakawa H et al. (2007) Mutation of RRM2B, encoding p53-controlled ribonucleotide reductase (p53R2),
causes severe mitochondrial DNA depletion. Nat Genet
39, 776–780.
Zhou B, Liu X, Mo X, Xue L, Darwish D, Qiu W, Shih
J, Hwu EB, Luh F & Yen Y (2003) The human ribonucleotide reductase subunit hRRM2 complements p53R2
in response to UV-induced DNA repair in cells with
mutant p53. Cancer Res 63, 6583–6594.
Rampazzo C, Fabris S, Franzolin E, Crovatto K, Frangini M & Bianchi V (2007) Mitochondrial thymidine
kinase and the enzymatic network regulating thymidine
triphosphate pools in cultured human cells. J Biol Chem
282, 34758–34769.
Bridges EG, Jiang Z & Cheng YC (1999) Characterization of a dCTP transport activity reconstituted from

human mitochondria. J Biol Chem 274, 4620–4625.
Henderson JP, Byun J, Williams MV, McCormick ML,
Parks WC, Ridnour LA & Heinecke JW (2001) Bromination of deoxycytidine by eosinophil peroxidase: a mechanism for mutagenesis by oxidative damage of nucleotide
precursors. Proc Natl Acad Sci USA 98, 1631–1636.
Henderson JP, Byun J, Williams MV, Mueller DM,
McCormick ML & Heinecke JW (2001) Production of

FEBS Journal 276 (2009) 1654–1666 ª 2009 The Authors Journal compilation ª 2009 FEBS

1665


A mammalian dCTPase that prefers 5-iodocytosine

30

31

32

33

34

M. Nonaka et al.

brominating intermediates by myeloperoxidase. A transhalogenation pathway for generating mutagenic
nucleobases during inflammation. J Biol Chem 276,
7867–7875.
Kawai Y, Morinaga H, Kondo H, Miyoshi N,

Nakamura Y, Uchida K & Osawa T (2004) Endogenous formation of novel halogenated 2¢-deoxycytidine. Hypohalous acid-mediated DNA modification
at the site of inflammation. J Biol Chem 279, 51241–
51249.
Valinluck V & Sowers LC (2007) Endogenous cytosine
damage products alter the site selectivity of human
DNA maintenance methyltransferase DNMT1. Cancer
Res 67, 946–950.
Valinluck V & Sowers LC (2007) Inflammation-mediated cytosine damage: a mechanistic link between
inflammation and the epigenetic alterations in human
cancers. Cancer Res 67, 5583–5586.
Valinluck V, Liu P, Kang JI Jr, Burdzy A & Sowers
LC (2005) 5-halogenated pyrimidine lesions within a
CpG sequence context mimic 5-methylcytosine by
enhancing the binding of the methyl-CpG-binding
domain of methyl-CpG-binding protein 2 (MeCP2).
Nucleic Acids Res 33, 3057–3064.
Vilpo JA & Vilpo LM (1991) Biochemical mechanisms
by which reutilization of DNA 5-methylcytosine is prevented in human cells. Mutat Res 256, 29–35.

1666

35 Vilpo JA & Vilpo LM (1993) Nucleoside monophosphate kinase may be the key enzyme preventing salvage
of DNA 5-methylcytosine. Mutat Res 286, 217–220.
36 Holliday R & Ho T (1998) Evidence for gene silencing
by endogenous DNA methylation. Proc Natl Acad Sci
USA 95, 8727–8732.
37 Matsumoto M, Hatakeyama S, Oyamada K, Oda Y,
Nishimura T & Nakayama KI (2005) Large-scale analysis of the human ubiquitin-related proteome. Proteomics
5, 4145–4151.
38 Torisu K, Tsuchimoto D, Ohnishi Y & Nakabeppu Y

(2005) Hematopoietic tissue-specific expression of
mouse Neil3. J Biochem (Tokyo) 138, 763–772.
39 Tsuchimoto D, Sakai Y, Sakumi K, Nishioka K, Sasaki
M, Fujiwara T & Nakabeppu Y (2001) Human APE2
protein is mostly localized in the nuclei and to some
extent in the mitochondria, while nuclear APE2 is
partly associated with proliferating cell nuclear antigen.
Nucleic Acids Res 29, 2349–2360.
40 Nakabeppu Y & Nathans D (1991) A naturally occurring truncated form of FosB that inhibits Fos ⁄ Jun transcriptional activity. Cell 64, 751–759.
41 Yoshimura D, Sakumi K, Ohno M, Sakai Y, Furuichi
M, Iwai S & Nakabeppu Y (2003) An oxidized purine
nucleoside triphosphatase, MTH1, suppresses cell death
caused by oxidative stress. J Biol Chem 278, 37965–
37973.

FEBS Journal 276 (2009) 1654–1666 ª 2009 The Authors Journal compilation ª 2009 FEBS