Báo cáo khoa học: Site-specific phosphorylation of MCM4 during the cell cycle in mammalian cells pot

Site-specific phosphorylation of MCM4 during the cell
cycle in mammalian cells
Yuki Komamura-Kohno
1
, Kumiko Karasawa-Shimizu
1
, Takako Saitoh
1
, Michio Sato
1
,
Fumio Hanaoka
2
, Shoji Tanaka
1
and Yukio Ishimi
1,3
1 Mitsubishi Kagaku Institute of Life Sciences, Tokyo, Japan
2 Graduate School of Frontier Biosciences, Osaka University, Japan
3 Faculty of Science, Ibaraki University, Mito, Japan
MCM2–7 proteins are essential for eukaryotic DNA
replication and are the most likely candidates for the
replicative DNA helicase responsible for unwinding
DNA at the replication forks [1–3]. Consistent with
their primary amino acid sequences, a subcomplex of
MCM4 ⁄ 6 ⁄ 7 functions as DNA helicase in vitro [4]. It
has been suggested that MCM2, -3 and -5 play a regu-
latory role in the function of MCM4 ⁄ 6 ⁄ 7 DNA heli-
case, because addition of MCM2 or MCM3 ⁄ 5to
MCM4 ⁄ 6 ⁄ 7 complex resulted in inhibition of the
MCM4 ⁄ 6 ⁄ 7 DNA helicase [5,6]. Thus MCM2–7 com-

plex, a major MCM complex on chromatin during the
G
1
phase, has to be activated to show DNA helicase
activity. It is possible that several proteins, including
CDC7 kinase and CDC45 are involved in this activa-
tion. Evidence suggests that MCM2–7 proteins may
have additional functions during the cell cycle [3].
Cyclin-dependent kinases (CDK), which play a critical
Keywords
CDK; cell cycle; DNA replication; MCM
proteins; phosphorylation
Correspondence
Y. Ishimi, Faculty of Science, Ibaraki
University, 2-1-1 Bunkyo, Mito 310-8512,
Ibaraki, Japan
Fax: +81 29 228 8439
Tel: +81 29 228 8439
E-mail:
(Received 5 September 2005, revised 6
January 2006, accepted 18 January 2006)
doi:10.1111/j.1742-4658.2006.05146.x
MCM4, a subunit of a putative replicative helicase, is phosphorylated dur-
ing the cell cycle, at least in part by cyclin-dependent kinases (CDK), which
play a central role in the regulation of DNA replication. However, detailed
characterization of the phosphorylation of MCM4 remains to be per-
formed. We examined the phosphorylation of human MCM4 at Ser3,
Thr7, Thr19, Ser32, Ser54, Ser88 and Thr110 using anti-phosphoMCM4
sera. Western blot analysis of HeLa cells indicated that phosphorylation of
MCM4 at these seven sites can be classified into two groups: (a) phos-

phorylation that is greatly enhanced in the G
2
and M phases (Thr7, Thr19,
Ser32, Ser54, Ser88 and Thr110), and (b) phosphorylation that is firmly
detected during interphase (Ser3). We present data indicating that phos-
phorylation at Thr7, Thr19, Ser32, Ser88 and Thr110 in the M phase
requires CDK1, using a temperature-sensitive mutant of mouse CDK1,
and phosphorylation at sites 3 and 32 during interphase requires CDK2,
using a dominant-negative mutant of human CDK2. Based on these results
and those from in vitro phosphorylation of MCM4 with CDK2 ⁄ cyclin A,
we discuss the kinases responsible for MCM4 phosphorylation. Phosphor-
ylated MCM4 detected using anti-phospho sera exhibited different affinities
for chromatin. Studies on the nuclear localization of chromatin-bound
MCM4 phosphorylated at sites 3 and 32 suggested that they are not gener-
ally colocalized with replicating DNA. Unexpectedly, MCM4 phosphoryl-
ated at site 32 was enriched in the nucleolus through the cell cycle. These
results suggest that phosphorylation of MCM4 has several distinct and
site-specific roles in the function of MCM during the mammalian cell cycle.
Abbreviations
CDK, cyclin-dependent kinases.
1224 FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS
role in the G
1
to S and G
2
to M transitions in the
cell cycle are also required to prevent re-replication of
DNA in a single cell cycle. Inactivation of CDK1 leads
to re-replication of DNA in eukaryotic cells including
human cells [7]. Targets of the kinase in the regulation

of DNA replication include ORC2, Cdc6 and Mcm
proteins in Saccharomyces cerevisiae [8], and disregula-
tion of these proteins leads to limited over-replication
of the genome. In higher eukaryotic cells, MCM4
is phosphorylated extensively in the M phase, and
CDK1 ⁄ cyclin B appears to be responsible for the
phosphorylation [9–11]. It has been proposed that
phosphorylation of MCM4 in the M phase may be
required to prevent binding of the MCM complex to
chromatin in Xenopus [6]. Partly consistent with this
notion, it has been shown that CDK2 ⁄ cyclin A phos-
phorylates MCM4 to prevent binding of MCM com-
plex to chromatin [12]. In contrast, a recent finding
suggests that an intermediate level of phosphorylation
of MCM4 is required for chromatin binding during
interphase [11]. We showed that the MCM4 ⁄ 6 ⁄ 7 com-
plex purified from HeLa cells in the G
2
and M phase
shows a lower level of DNA helicase activity compared
with complex purified from logarithmically growing
cells [13]. Thus, phosphorylation of MCM4, together
with the presence of geminin, which inhibits the ability
of CDT1 to load MCM proteins onto chromatin, may
help prevent re-replication of DNA in the G
2
and M
phase. During interphase, chromatin-bound MCM4 is
partially phosphorylated and its level is higher than
that of MCM4 that is not bound to chromatin [10,11].

The characterization and functional significance of
MCM4 phosphorylation during interphase remains to
be clarified.
We report that in vitro phosphorylation of human
MCM4 ⁄ 6 ⁄ 7 complex with CDK2 ⁄ cyclin A leads to
inactivation of the DNA helicase activity of the
complex [14]. We identified six Ser or Thr residues (3,
7, 19, 32, 53, 109) in the N-terminal region of mouse
MCM4 as the sites required for phosphorylation with
CDK2 ⁄ cyclin A and CDK1⁄ cyclin B in vitro [13]. Con-
version of these six Ser or Thr residues to Ala resulted
in the MCM4 ⁄ 6 ⁄ 7 complex showing resistance to inhi-
bition with CDK2 ⁄ cyclin A, indicating that phos-
phorylation at these six sites is responsible for the
inactivation of MCM4 ⁄ 6 ⁄ 7 DNA helicase. We charac-
terized the phosphorylation of MCM4 during the cell
cycle in human and mouse cells using antiphospho sera
against these sites. We show that phosphorylation at
sites Thr7, Thr19, Ser32, Ser87 and Thr109 requires
CDK1 in the M phase of mouse FM3A cells, and
phosphorylation at sites 3 and 32 requires CDK2 dur-
ing interphase in human HeLa cells. Changes in the
phosphorylation level during the cell cycle and the
nuclear localization of phosphorylated MCM4 suggest
that MCM4 phosphorylated at these sites has several
distinct and site-specific roles in MCM functions.
Results
Characterization of antiphosphoMCM4 sera
We identified six SP or TP sites (Ser3, Thr7, Thr19,
Ser32, Ser53 and Thr109) in the N-terminal region of

mouse MCM4 as being required for the phosphoryla-
tion of MCM4 with CDK2 ⁄ cyclin A in vitro [13]. All
six sites are conserved between mouse and human
MCM4, although the numbers of sites 53 and 109 in
mouse MCM4 was changed to 54 and 110 in human
MCM4 (Fig. 1). We prepared antiphosphoMCM4 sera
against these sites of human MCM4 in addition to
those against Ser88. Figure 2 shows the specificity of
the antibodies as measured by ELISA. The data indi-
cate that all six antiphosphoMCM4 sera (P-3, -7, -19,
Fig. 1. Amino acid alignment of human and mouse MCM4 in the N-terminal region. Amino acid sequences in the N-terminal region of human
and mouse MCM4 in which SP and TP sites are clustered are aligned. These CDK sites are indicated by bold and italicized letters. Among
these sites, those that are required for phosphorylation with CDK2 ⁄ cyclin A (13) in addition to site 88 are indicated by their amino acid
numbers.
Y. Komamura-Kohno et al. MCM4 phosphorylation in mammalian cells
FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS 1225
-32, -54 and -110) bound almost specifically to their
corresponding phosphopeptides. The binding depend-
ency of the antibodies on phosphorylation is examined
in Fig. 3A. Human MCM4 ⁄ 6 ⁄ 7 complexes were incu-
bated in the presence or absence of purified CDK2 ⁄
cyclin A in vitro and MCM4 proteins were then
analyzed by western blotting using six phosphoantibod-
ies. In addition to wild-type MCM4 ⁄ 6 ⁄ 7 complex, a
mutant complex in which six Ser or Thr residues (3, 7,
19, 32, 54, 100) in MCM4 were converted to Ala was
also incubated with CDK2 ⁄ cyclin A. All the antibodies
reacted to MCM4 in the wild-type complex but not in
the mutant complex. The signals from the wild-type
complex were detected with P-3, -32 and -54 antibodies

even in the absence of CDK2 ⁄ cyclin A, indicating that
MCM4 is phosphorylated at these sites during pre-
paration from insect cells. Incubation of wild-type
MCM4 ⁄ 6 ⁄ 7 complex with CDK2 ⁄ cyclin A enhanced
the signals detected with the antibodies (P-32 and -54)
or induced the signals with the antibodies (P-7, -19 and
-110). However, the kinase barely stimulated the signal
with P-3 antibodies under these conditions. The signal
detected with P-3 antibodies in the absence of Cdk2 ⁄
cyclin A disappeared after incubation of the complex
with k phosphatase (Fig. 3B). These results indicate that
all the signals detected with the six phosphoantibodies
are dependent on the phosphorylation of MCM4.
Binding of these antibodies to human MCM4 in log-
arithmically growing cells and cells synchronized at the
G
2
and M phase was examined (Fig. 4). Because the
MCM proteins, including MCM4, are almost exclu-
sively detached from chromatin in the G
2
and M
phase, they were recovered in a Triton-soluble (S) frac-
tion, which was detected using anti-MCM4 sera. At
this stage, MCM4 was extensively phosphorylated, as
revealed by the retarded mobility of the protein in
SDS gel, compared with the mobility of protein pre-
pared from logarithmically growing HeLa cells. All
seven antiphosphoMCM4 sera recognized the retarded
MCM4 prepared from cells in the M phase, indicating

that these sites are indeed phosphorylated in the M
phase in HeLa cells. We classified the mode of phos-
phorylation into two groups: (a) phosphorylation is
greatly enhanced in phases G
2
and M (Thr7, Thr19,
Ser32, Ser54, Ser88 and Thr110), and (b) phosphoryla-
tion is firmly detected at interphase (Ser3). Phosphoryl-
ated MCM4 in cells at interphase was weakly detected
by the P-32, -54 and -88 antibodies.
Phosphorylation of MCM4 during the HeLa cell
cycle
Changes in the levels of MCM4 phosphorylation at
sites 3 and 32 during the HeLa cell cycle were analyzed
(Fig. 5). Logarithmically growing HeLa cells, pulse-
labeled with BrdU, were stained with antiphospho-
MCM4 sera [P-3 (A) and P-32 (B)] and anti-BrdU sera.
We quantified the fluorescence intensity in the nucleus
and cytoplasm separately. In the M phase, we meas-
ured fluorescence in regions surrounding total con-
densed chromosomes and showed it to be the same as
in the nucleus. Phosphorylation at site 3 increased in
the nucleus during phases G
1
and S, and was detected
in the cytoplasm during the M phase. Phosphorylation
at site 32 increased gradually in the nucleus during
phases S and G
2
, and was detected maximally in the

cytoplasm in the M phase. The timing of phosphoryla-
tion and dephosphorylation of MCM4 during phases
G
2
and M was compared among the sites (Ser3, Thr7,
Ser32, Ser54 and Thr110) (Fig. 6). Figure 6A shows a
typical example of the staining pattern seen using con-
forcal microscopy. The data suggest that all the phos-
phorylated MCM4 is not bound with mitotic
chromosomes. The level of staining during phases G
2
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
P-3Ab
P-7Ab
P-19Ab
P-32Ab
P-54Ab
P-110Ab
A
450

P-Ser3 P-Thr7 P-Thr19
P-Ser54
P-Ser32
P-Thr110
phosphopeptides
Fig. 2. Specificity of binding of phospho-
MCM4 antibodies. The binding specificity of
six antiphosphoMCM4 sera (P-3, -7, -19,
-32, -54 and -110) to phosphopeptides was
examined by ELISA. The six phosphoanti-
bodies were incubated with six correspond-
ing phosphopeptides (P-Ser3, -Thr7, -Thr19,
-Ser32, -Ser54 and -Thr110) and binding was
detected by absorbance at 450 nm.
MCM4 phosphorylation in mammalian cells Y. Komamura-Kohno et al.
1226 FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS
and M was quantified and compared among the five
sites (3, 7, 32, 54, 110) (Fig. 6B). Phosphorylation at
sites 3 and 110 was maximal in the G
2
phase, in con-
trast, phosphorylation at the other sites (7, 32, 54) was
maximal in the M phase. Phosphorylation of MCM4 at
sites 7 and 32 decreased during anaphase. Changes in
phosphorylation at sites 7 and 32 during the M phase
appear to parallel changes in CDK1 ⁄ cyclin B activity.
These results indicate that the CDK sites in MCM4 are
differently phosphorylated and dephosphorylated dur-
ing the HeLa cell cycle in a site-specific manner.
Cyclin-dependent protein kinase is involved in

the phosphorylation of MCM4
To determine which kinase is involved in the phos-
phorylation of MCM4 in phases G
2
and M, we used
mouse FT210 cells in which CDK1 activity is tempera-
ture sensitive [15]. After synchronizing the cells at the
G
1
⁄ S boundary, they were allowed to progress to
phases S and G
2
. At permissive temperatures, cells
accumulated in the M phase in the presence of noco-
dazole (mitotic index: 30%). At nonpermissive temper-
atures, cells are arrested in the G
2
phase, this is caused
by inactivation of CDK1 which is crucial for entry
into the M phase. We compared the phosphorylation
level of MCM4 between these two cells (Fig. 7A). Only
Triton-soluble fractions were examined for MCM4
phosphorylation. We confirmed that all the antiphos-
phoMCM4 sera (P-3, -7, -19, -32, -54, -88 and -110)
recognized mouse MCM4 that had been prepared from
baculovirus-infected insect cells and then phosphorylat-
ed with CDK2 ⁄ cyclin A in vitro (data not shown).
Extensively phosphorylated MCM4, which showed
MCM4 P7P-3 P-19
P-32 P-54 P-110

A
B
P-3
MCM4
1
234
56
1
2
3
1
2
3456 1
2
34 56
1
2
345
6
1
2
3456
1
2
3456
1
2
34 5 6
1
2

3
-
-
wild
6A
CDK2
-
phosphatase
Fig. 3. In vitro phosphorylation of MCM4
with CDK2 ⁄ cyclin A. (A) Wild-type human
MCM4 ⁄ 6 ⁄ 7 complex (wild) (lanes 1–3) and
a mutant complex (6A) (lanes 4–6) in which
six Ser or Thr residues (3, 7, 19, 32, 54 and
110) of MCM4 had been converted to Ala
were incubated in the presence or absence
of increasing amounts of CDK2 ⁄ cyclin A.
Proteins were analyzed by
Western blot
using anti-phospho and anti-MCM4 sera as
indicated. (B) Wild-type MCM4 ⁄ 6 ⁄ 7 com-
plex was incubated in the presence or
absence of increasing amounts of k phos-
phatase under recommended conditions
(New England Biolabs). The proteins were
analyzed by western blot using anti-P-3 and
anti-MCM4 sera. Arrows on the right-hand
side of the gel indicate the 95 kDa position.
Y. Komamura-Kohno et al. MCM4 phosphorylation in mammalian cells
FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS 1227
retarded mobility, was detected in extracts from cells

cultured at a permissive temperature but not in cells
cultured at a nonpermissive temperature, which was
detected by anti-MCM4 sera. The P-7 and P-19 phos-
phoantibodies recognized MCM4 with retarded mobil-
ity as well as MCM4 at a nonretarded position in the
extracts prepared from cells cultured at a permissive
temperature. In contrast, only MCM4 at the nonre-
tarded position was detected using these two antibod-
ies in extracts from cells cultured at a nonpermissive
temperature. The phosphoantibodies (P-32, -88 and
-110) recognized almost exclusively MCM4 with retar-
ded mobility in the cells cultured at a permissive tem-
perature. Weak bands recognized with P-88 and P-110
were detected near the nonretarded position but no
bands were recognized with P-32 antibodies in cells
cultured at a nonpermissive temperature. Overall, the
intensity of the bands detected with these antibodies
(P-7, -19, -32, -88, and -110) decreased in cells cultured
at a nonpermissive temperature, because the intensity
ratio (39 ⁄ 33) was calculated as 0.2–0.58. In contrast,
bands detected with the P-3 and P-54 antibodies were
almost unchanged between cells cultured at a permis-
sive temperature and those cultured at a nonpermissive
temperature, because the intensity ratio (39 ⁄ 33) was
calculated as 0.91 and 1.1. These results suggest that
CDK1 is involved in the phosphorylation of mouse
MCM4 at five sites (Thr7, Thr19, Ser32, Ser87 and
Thr109) but not phosphorylation at the other two
(Ser3 and Ser53) in the M phase. Involvement of
CDK1 for MCM4 phosphorylation at sites 32, 87 and

109 in the M phase is almost absolute. However,
involvement at sites 7 and 19 may be partial, because
signals detected at the nonretarded position were not
decreased at the nonpermissive temperature.
To address the question of whether CDK2 is res-
ponsible for the phosphorylation of MCM4 at CDK
sites during interphase, a dominant-negative mutant of
CDK2 [16] was expressed in HeLa cells, and the effect
of its expression on the phosphorylation of MCM4 was
examined (Fig. 7B). The level of MCM4 phosphory-
lation at sites 3 and 32 was compared between cells that
express the mutant CDK2 and those that do not. Phos-
phorylation of MCM4 at these two sites was signifi-
cantly decreased in cells expressing mutant CDK2, as
shown in Fig. 7B. For quantification, we separated the
immunofluorescence intensity from each cell into two
(strong and weak). In cells that do not express CDK2,
strong signals detected with P-3 antibodies were
observed in 30% (100 ⁄ 328) of cells, and in those that
do express the CDK2, strong signals were observed in
1.6% (2 ⁄ 129) of cells. For P-32 antibodies, strong sig-
nals were detected in 42% (100⁄ 238) of cells that do
not express CDK2, and 15% (14 ⁄ 92) of cells that do
express CDK2. Thus, CDK2 is almost exclusively
involved in phosphorylation at site 3 and is partly
involved in phosphorylation at site 32 during inter-
phase in HeLa cells. We also examined the effect of the
expression of mutant CDK2 on phosphorylation of
MCM4 at site 54 (data not shown). The results
MCM4

P-3
P-32
P-54
P-7
P-19
P-110
SPSPSP
SPSP
S
P
log
G2
M
M
G2
log
P-88
Fig. 4. Detection of phosphorylated MCM4
using antiphospho sera by western blot ana-
lyses. (A) Logarithmically growing HeLa cells
were incubated with nocodazole. Cells
detached from the bottle by shaking were
collected and named mitotic (M) cells.
Residual cells were collected and named G
2
cells. These cells and logarithmically grow-
ing cells were separated into Triton-soluble
(S) and Triton-insoluble (P) fractions. After
electrophoresis, proteins in these fractions
were analyzed by using anti-MCM4 or an-

tiphosphoMCM4 (P-3, -7, -19, -32, -54, -88
and -110) sera as indicated. Arrows on the
right-hand side of gel indicate the 95 kDa
position.
MCM4 phosphorylation in mammalian cells Y. Komamura-Kohno et al.
1228 FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS
indicated that phosphorylation at site 54 was almost
entirely resistant to the expression of mutant CDK2,
suggesting that CDK is not involved in the phosphory-
lation.
Phosphorylated MCM4 on chromatin
HeLa cells were separated into Triton-soluble and Tri-
ton-insoluble fractions and the insoluble fraction was
further separated into DNaseI-soluble and DNaseI-
insoluble fractions. Proteins in these fractions were
analyzed using antiphosphoMCM4 sera (Fig. 8). West-
ern blotting analysis using anti-MCM4 sera showed
the distribution of total MCM4 proteins in these frac-
tions under these conditions. MCM4 phosphorylated
at sites 3 and 54 was preferentially detected in the
Triton-soluble fraction. In contrast, MCM4 phos-
phorylated at sites 7 and 32 was mainly detected in the
chromatin-bound fractions. Although the phosphoanti-
body against site 7 recognizes mainly MCM4 in the M
phase (Fig. 4), it can detect MCM4 during interphase
to a lesser extent. These results suggest that MCM4
phosphorylated at different sites shows different affin-
ity for chromatin. To study the relationship between
chromatin-bound phosphorylated MCM4 and DNA
synthesis, logarithmically growing HeLa cells, pulse-

labeled with BrdU, were treated with Triton and then
stained with antiphosphoMCM4 sera (P-3 and P-32)
(Fig. 9A,B). BrdU-negative cells were differentiated
into G
1
and G
2
cells using nuclear mass, and BrdU-
positive cells were differentiated into the three phases
of early S (eS), middle S (mS) and late S (lS) from the
pattern of nuclear staining with BrdU. MCM4 phos-
phorylated at sites 3 and 32 was not largely colocalized
A
B
Fig. 5. Changes in MCM4 phosphorylation during the HeLa cell cycle. Logarithmically growing HeLa cells that had been pulse-labeled with
BrdU were incubated with antiphosphoMCM4 sera [P-3 (A) and P-32 (B)] and anti-BrdU sera, and were observed using a CCD camera. In
each cell, the fluorescence intensities of secondary antibodies were measured. Using image cytometry, intensities in the cytoplasm and
nucleus were individually quantified. In mitotic cells, the intensity in a region containing whole chromosomes was quantified and shown to
be that in nucleus. From the intensity of DAPI fluorescence and the reactivity to anti-BrdU sera, the cell-cycle stage was determined in each
cell.
Y. Komamura-Kohno et al. MCM4 phosphorylation in mammalian cells
FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS 1229
A
B
G1
Intensity of
fluorescence
G2 Prometaphase
Anaphase
Prophase Metaphase Telophase

Fig. 6. Immunostaining of mitotic cells with
phosphoantibodies. (A) Logarithmically
growing HeLa cells were stained with
antiphosphoMCM4 sera (P-3,-7, -32, -54 and
-110) and propidium iodide (10 l
M), and
observed using a confocal laser scanning
microscope. Cells in the M phases (pro-
phase, prometaphase, metaphase, anaphase
and telophase) were collected in addition to
those in the G
1
and G
2
phases. A typical
example of these cells is shown. Binding of
antiphosphoMCM4 sera and PI is shown in
green and red, respectively. (B) The fluores-
cence intensities of antiphosphoMCM4 sera
in cells in phases G
1
,G
2
and M were meas-
ured, and their averages (and standard devi-
ation) are shown as relative values.
MCM4 phosphorylation in mammalian cells Y. Komamura-Kohno et al.
1230 FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS
to BrdU-incorporated DNA in the three periods of the
S phase. Similar results were obtained with anti-

MCM4 sera (data not shown). They are in agreement
with previous findings [17–19]. The fluorescence inten-
sity generated by these antibodies was quantified
(Fig. 9C). Phosphorylation of MCM4 at sites 3 and 32
on chromatin greatly increased in the S phase com-
pared with the G
1
phase. MCM4 phosphorylation on
chromatin at these sites began to decrease during the
late S phase and was greatly reduced in the G
2
phase.
MCM4
33 39
P-7 P-19 P-32 P-110
33 39
92K
33 39 33 39 33 39 33 39
P-88
33°C
0 16hr 28hr
CDK1 inactivation(G2 arrest)
CDK1 activation(M entry)
Aphidicolin Aphidicolin removal
Nocodazole addition
(G1/S)
Recovery of cells
(M)
33°C
39°C

phospho-
MCM4
92K
P-3
33 39
P-54
33 39
0.91
1.1
0.40
0.58 0.31
0.20
0.42
1.0
DN-CDK2(HA)
merged
P-3
P-32
DN-CDK2(HA) merged
A
B
Fig. 7. CDKs are mainly responsible for the
phosphorylation of MCM4. (A) The experi-
mental design is presented at the top.
Mouse FT210 cells were synchronized at
the G
1
⁄ S boundary by incubating the cells
with aphidicolin for 16 h. After removal of
the drug, cells were cultured for 12 h in the

presence of nocodazole at permissive
(33 °C) or nonpermissive (39 °C) tempera-
tures. Cells were lyzed and Triton-soluble
fraction was examined for the presence of
phosphorylated MCM4 using western blot
analyses, which is shown at the bottom.
Antibodies used and temperature for culture
are indicated at the top. In the bottom, the
ratio (39 ⁄ 33 °C) of the intensity of the
signals is shown. (B) A dominant-negative
mutant of human CDK2 was expressed as
fusion proteins with HA in HeLa cells. The
effect of the expressed CDK on the
phosphorylation of MCM4 was examined by
costaining with anti-HA and antipho-
sphoMCM4 sera (P-3 or P-32). Each of the
single stainings and their merged image are
presented. Arrows indicate HeLa cells
expressing HA–CDK2 proteins.
Y. Komamura-Kohno et al. MCM4 phosphorylation in mammalian cells
FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS 1231
These changes were essentially similar to those with
anti-MCM4 sera. The finding that the amounts of
phosphorylated MCM4 (Ser3 and Ser32) in the chro-
matin-bound form decrease during phases S and G
2
is
in contrast to results shown in Fig. 5, in which the
phosphorylated MCM4 in a cell increases during these
periods, indicating that phosphorylated MCM4 is

detached from chromatin as the cell cycle progresses.
Unexpectedly, it has been shown that chromatin-
bound MCM4 phosphorylated at site 32 was concen-
trated in the nucleus during the cell cycle (Fig. 9B).
Similar results were also observed to a lesser extent
with anti-MCM4 sera, but not with anti-MCM3 sera
(data not shown). This finding on P-32 antibodies may
be consistent with the notion that the fluorescence
intensity detected by the antibodies in the G
2
phase
was slightly higher than detected by other antibodies
(Fig. 9C). Nuclear localization of MCM4 phosphoryl-
ated at site 32 was examined by costaining Triton-trea-
ted HeLa cells with the P-32 antibodies and antibodies
to C23 nucleolar protein (Fig. 10A). The data suggest
that these two proteins are colocalized, indicating that
MCM4 phosphorylated at site 32 is enriched in a
nucleolar region. Localization of MCM4 phosphory-
lated at site 32 in HeLa cells was also immunochemi-
cally examined using electron microscopy (Fig. 10B).
Signals with P-32 antibodies were detected in the entire
nucleus but were clustered in several regions including
the nucleolus. In the nucleolus, signals were detected
near densely stained structures. However, as enrich-
ment of the signals in the nucleolus is not obvious in
this system it may indicate that the immnunoreactions
are not saturated under these conditions. From these
results, it is suggested that MCM4 phosphorylated at
different CDK sites shows a unique affinity for chro-

matin.
MCM4
P-32P-7P-3 P-54
S1 S3
P’
S1
S3 P’
S1
S3 P’
S1 S3
P’
S1 S3 P’
S1 S3 P’
105
85
105
85
kDa
kDa
histone
A
C
B
Fig. 8. Detection of phosphorylated MCM4 in chromatin-bound fractions. Logarithmically growing HeLa cells were separated into Triton-sol-
uble (S) and Triton-insoluble fractions. The insoluble fraction was further separated into DNaseI-soluble (S3) and DNaseI-insoluble (P¢)
fractions. Proteins in these fractions were stained with Coomassie Brilliant Blue (A). They were examined by western blot analysis using
anti-MCM4 (B) or antiphosphoMCM4 sera (P-3, -7, -32 and -54) (C).
MCM4 phosphorylation in mammalian cells Y. Komamura-Kohno et al.
1232 FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS
Discussion

We showed that seven SP and TP sites in the N-terminal
region of MCM4 are uniquely phosphorylated during
the cell cycle. CDK1 is required for phosphorylation at
five sites (Thr7, Thr19, Ser32, Ser87 and Thr109) during
the M phase in mouse FM3A cells, and CDK2 is
required for phosphorylation at least at two sites (Ser3
and Ser32) during interphase in HeLa cells, suggesting
that CDK is involved in phosphorylation at these sites.
Changes in the phosphorylation level during the cell
cycle and the different affinities for chromatin suggest
that phosphorylation of MCM4 plays several distinct
roles in MCM function in mammalian cells. The finding
that phosphorylated MCM4 is not largely colocalized
to replicated DNA may be consistent with the notion
that the phosphorylation of MCM4 at CDK sites has a
negative role in MCM function.
All MCM2–7 members have an essential role in
the initiation and elongation of DNA replication [20],
possibly as a replicative DNA helicase [21,22]. It is
possible that MCM4 ⁄ 6 ⁄ 7 DNA helicase [4,23,24] is
generated from the MCM2–7 complex as the function
of the MCM complex. We have reported that the
MCM4 ⁄ 6 ⁄ 7 DNA helicase activity is inhibited by
phosphorylation of MCM4 with CDK2 ⁄ cyclin A at
the six SP or TP sites [13]. Our data indicate that
phosphorylation at these sites is not equivalent in
terms of cell cycle changes, localization in the nuclei or
the role of CDK. Phosphorylation at sites 7 and 32
Fig. 9. Immunostaining of chromatin-bound MCM4. (A) Logarithmically growing HeLa cells that had been pulse-labeled with BrdU were trea-
ted with Triton. They were stained with anti-BrdU sera and antiphosphoMCM4 sera (P-3 and 32). Cells at G

1
, early S (eS) middle S (mS), late
S (lS) and G
2
phases were collected. BrdU-negative cells were differentiated into G
1
and G
2
cells from total area of TOTO-stained nucleus.
Cells at the S phase are differentiated using their staining pattern with anti-BrdU sera. Fluorescent signals detected by antiphosphoMCM4
sera are shown in red and BrdU staining is shown in green, respectively, and these signals are presented individually and combined. (B) The
intensity of fluorescence detected with antibodies (P-3, -32 and MCM4) was quantified in each cell at G
1
, eS, mS, lS and G
2
. Averages of
the intensity in each cell are presented (with standard deviations) and they are shown as relative to values from the G
1
phase.
Y. Komamura-Kohno et al. MCM4 phosphorylation in mammalian cells
FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS 1233
begins to increase in the G
2
phase and to decrease dur-
ing anaphase, the kinetics of which seems to parallel
the change in CDK1 ⁄ cyclin B activity. These results
are consistent with the notion that CDK1 ⁄ cyclin B is
required for the phosphorylation of mouse MCM4 at
these sites. Because changes in the level of phosphory-
lation at other sites during phases G

2
and M differ
from those at sites 7 and 32, other factors may be
involved in the phosphorylation of MCM4. During
interphase, phosphorylation at sites 3 and 32 was
sensitive to the expression of a dominant-negative
mutant of CDK2 in HeLa cells. However, in vitro
studies (Fig. 3) showed that CDK2 ⁄ cyclin A barely
phosphorylates MCM4 at site 3, suggesting that
CDK2 ⁄ cyclin A is indirectly involved in phosphory-
lation. Phosphorylation at site 54 was relatively resist-
ant to the expression of mutant CDK2 during
interphase, and occurs in phases G
2
and M even in the
absence of CDK1 activity, suggesting that kinase(s)
other than CDK may be involved in phosphorylation
at site 54 during the cell cycle. Consistently, it appears
that CDK2 ⁄ cyclin A does not efficiently phosphorylate
site 54 in vitro. One unique feature of the nuclear local-
ization is that MCM4 phosphorylated at site 32 was
enriched in the nucleolus. Using electron microscopy,
the signals detected with P-32 antibodies were near
densely stained structures that probably correspond to
dense fibrillar components in which transcription by
RNA polymerase I occurs [25]. It is possible that
MCM4 phosphorylated at site 32 has a unique role in
the function of the nucleolus, including ribosomal
RNA transcription. It has been reported that MCM
proteins rebind to chromatin at late telophase and

DNA replication licensing is completed at the G
1
phase [26,27]. Unexpectedly, the level of chromatin-
bound MCM4 protein was greatly reduced in the G
1
phase compared with the S phase (Fig. 9C). Because
the extraction conditions used are relatively stringent,
it is possible that a large part of the MCM proteins on
licensed chromatin in the G
1
phase was detached from
chromatin. Phosphorylation of MCM4 and other fac-
tors may be involved in stronger binding of MCM4
protein to chromatin in the S phase compared with the
G
1
phase. This assumption seems to be consistent with
the finding that MCM4 phosphorylated at sites 7 and
32 is enriched in the chromatin fraction (Fig. 8).
In HeLa cells, MCM4 phosphorylation at site 110
was detected not only in the M phase, but also in the G
2
phase (Figs 4 and 6B). These results appear to be incon-
sistent with those shown in Fig. 7A, in which signals
detected with P-110 antibodies decreased in G
2
-arrested
FT210 cells cultured at a nonpermissive temperature.
This discrepancy remains to be resolved but may be
explained by the difference between normal G

2
and
arrested G
2
. For example, the balance of phosphoryla-
tion and dephosphorylation at the site might differ
between the two G
2
conditions. Another apparent
inconsistency concerns the level of phosphorylation at
site 3 in the M phase. The level of phosphorylation at
the site is relatively high in the M phase in Fig. 5A, but
relatively low in Fig. 6B. Immunofluorescence with P-3
antibodies was detected using a CCD camera in Fig. 5A
but was detected using a confocal laser scanning micro-
C
Intensity of
fluorescence
Fig. 9. (Continued).
MCM4 phosphorylation in mammalian cells Y. Komamura-Kohno et al.
1234 FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS
scope in Fig. 6. Because quantification of total fluores-
cence in a cell using a CCD camera is more accurate, the
data in Fig. 5A are more reliable for understanding any
change in phosphorylation at site 3 during the cell cycle.
With regards to the role of MCM4 phosphorylation on
MCM function, it has been suggested that a high level
of phosphorylation in phases G
2
and M might inhibit

binding of the MCM complex to chromatin in G
2
and
M [9]. To address this possibility, we expressed mutant
MCM4 in which six amino acids (Ser or Thr) were con-
verted to Ala or Glu to examine their binding to chro-
matin (data not shown). These two mutant MCM4s
were recovered in the chromatin-bound fraction to a
similar extent to wild-type MCM4, suggesting that
phosphorylation of MCM4 may not directly affect its
chromatin binding. It is also possible that glutamic acid
substitution may not mimic phosphorylation, however,
further analysis is required to address this.
Our findings of cell-cycle changes in phosphoryla-
tion and a unique affinity for chromatin suggest that
phosphorylation of MCM4 has different roles in
MCM function. They present a starting point from
which to explore the molecular mechanisms underly-
ing these phenomena and their functional signific-
ances. Ten SP and TP sites were present in the
N-terminal region of Drosophila MCM4 and three
(3, 54, 110) of the six human CDK sites were con-
served in Drosophila. Essentially, no sequence conser-
vation was detected between human and yeast, but
nine and four SP or TP sites are clustered in this
region of MCM4 in Schizosaccharomyces pombe and
S. cerevisiae, respectively. These findings would be
consistent with the notion that the interplay between
CDK and MCM4 plays an important function in
eukaryotic cells.

Experimental procedures
Cell culture and antibodies
HeLa cells were cultured in Dulbecco’s modified Eagle’s
medium supplemented with 10% bovine serum. FT210 cells
were cultured in RPMI-1640 with 10% fetal bovine serum.
Logarithmically growing HeLa cells were incubated with
A
B
C23 mergedP-32
Fig. 10. Nuclear localization of phosphorylated MCM4. (A) HeLa cells treated with Triton were costained with anti-P-32 and anti-C23 nucleo-
lar protein sera. The images of each staining are shown both singly and combined. (B) HeLa cells were immunostained with P-32 antibodies
and the signals, enhanced using silver, were detected by electron microscopy. Images showing an entire nucleus (left) and the nucleolus
(right) of the same section are presented.
Y. Komamura-Kohno et al. MCM4 phosphorylation in mammalian cells
FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS 1235
50 ngÆmL
)1
nocodazole for 20 h. Cells at mitosis were recov-
ered after shaking a culture bottle. Those that remained
attached to the bottle were recovered after treatment with
trypsin and were named G
2
cells. Anti-MCM4 and anti-
(phospho-human MCM4) (P-7, -19, -32, -54 and -110) rabbit
sera were prepared as reported previously [28] and those
against sites 3 and 88 (P-3 and P-88) were also prepared
reported previously [29]. Anti-C23 mouse sera (MS-3) and
anti-HA mouse sera (Y-11) were obtained from Santa Cruz
Biotechnology (Santa Cruz, CA).
ELISA

Each of six peptides containing phospho-residue at the sites
(Ser3, Thr7, Thr19, Ser32, Ser54, Thr110) was suspended at
5 lgÆmL
)1
in 0.05 m sodium cabonate buffer, pH 9.6. They
were added to 96-well plate (100 lLÆwell
)1
) and the plate
was incubated at 4 °C overnight. After washing the wells
with NaCl ⁄ P
i
containing 0.2% Tween-20, six antiphos-
phoMCM4 sera (200–500 lgÆmL
)1
) diluted 3000-fold with
NaCl ⁄ P
i
containing 0.05% Tween-20 were added to the
wells (100 lLÆwell
)1
) and the plate was incubated at 37 °C
for 1.5 h. After washing the wells, peroxidase-conjugated
anti-(rabbit IgG) sera (Bio-Rad Laboratories, Hercules,
CA) diluted 10 000-fold were added to the wells and the
plate was incubated at 37 °C for 1.5 h. After washing,
tetramethylbenzidine liquid substrate (Sigma, St Louis,
MO) was added to each well (100 lLÆwell
)1
). The reaction
was carried out for 10–20 min at room temperature and

stopped by adding 100 lLof1m sulfuric acid. Absorbance
was measured at 450 nm.
Preparation of cell fractions and western blotting
HeLa cells were lysed at 2 · 10
6
cells per 100 lL in modified
CSK buffer (10 mm Pipes, pH 6.8, 100 mm NaCl, 1 mm
MgCl
2
and 1 mm EGTA) containing 0.1% Triton X-100,
1mm ATP, proteinase inhibitors (Pharmingen) and phos-
phatase inhibitors (10 mm sodium b-glycerophosphate,
5mm sodium pyrophosphate, 1 mm sodium orthovanadate
and 50 mm sodium fluoride) (solution A) and placed on ice
for 15 min. The cell suspension was centrifuged (2000 g for
5 min in a microcentrifuge), and its supernatant was saved
(S1). Recovered precipitate was suspended in solution A,
and the supernatant after centrifugation was saved (S2).
Fractions S1 and S2 were combined and used as an S frac-
tion in some cases. The precipitate was suspended in a vol-
ume of solution A to yield 4 · 10
6
cells per 100 lL (P) and
the suspension was briefly sonicated in the presence of a
loading buffer for SDS ⁄ PAGE. When DNA in the precipi-
tate was digested with DNaseI, cells were lyzed as described,
except that the phosphatase inhibitors were replaced with
phosphatase inhibitor cocktails I and II (Sigma). The preci-
pitated fraction was suspended and then incubated with
DNaseI (Takara, Japan) at 200 lg Æ mL

)1
at 30 °C for
15 min, and then soluble (S3) and insoluble (P¢) fractions
were recovered after centrifugation. The proteins in these
fractions were electrophoresed through 10% acrylamide gels
containing SDS and then transferred to membranes (Immo-
bilon, Millipore Corp., Bedford, MA). Approximately 30 lg
of total proteins in the S fraction was loaded on to the gels.
After membranes had been incubated with a blocking solu-
tion (Blockace, Dai-nippon Pharmaceuticals, Japan) for 1 h
at room temperature, they were incubated at 4 °C overnight
with primary antibodies in the blocking solution (for anti-
MCM4 and anti-HA sera) or 5% bovine serum albumin in
Tris-buffered saline (TBS; 50 mm Tris ⁄ HCl, pH 7.5 0.15 m
NaCl) plus 0.1% Triton X-100 (for antiphosphoMCM4
sera). After washing with TBS plus 0.1% Triton X-100,
membranes were incubated with peroxidase-conjugated sec-
ondary anti-rabbit sera (Bio-Rad) in the blocking solution.
The immunoreacted proteins were detected by Cool Saver
AE-6935 (Atto) using a chemiluminescent detection system
(SuperSignal West Pico or Femto Maximum Sensitivity
Substrate, Pierce, Rockford, IL).
In vitro phosphorylation
Human MCM4 ⁄ 6 ⁄ 7 complex, in which six histidines were
attached to MCM4 at the N-terminus, was purified from
baculovirus-infected High cells using Ni-NTA chromatogra-
phy and then by glycerol gradient centrifugation [30]. A
mutant MCM4 ⁄ 6 ⁄ 7 complex in which six Ser or Thr resi-
dues (3, 7, 19, 32, 54, 110) in the N-terminal region of
MCM4 had been converted to Ala using the QuickChange

site-directed mutagenesis kit (Stratagene, La Jolla, CA) was
also prepared. These complexes were incubated with puri-
fied CDK2 ⁄ cyclin A complex as reported [14]. Proteins
were analyzed as described above.
Immunostaining
HeLa cells were cultured on eight-well chambers (Falcon
Becton Dickinson Labware, Franklin Lakes, NJ) or cover-
slips. They were pulse-labeled with BrdU (50 lm) for
15 min (Fig. 9). After washing with NaCl ⁄ P
i
, cells were
fixed by incubation with 4% paraformaldehyde in NaCl ⁄ P
i
for 5 min at room temperature. To extract soluble proteins,
cells were immersed in buffer containing Triton X-100 used
for cell fractionation and then incubated at 37 °C for
15 min in the same buffer before fixation (Figs 9 and 10).
Cells were washed with NaCl ⁄ P
i
and then permeabilized
and blocked by incubation with 0.1% Triton X-100, 0.02%
SDS and 2% nonfat dried milk in NaCl ⁄ P
i
for 1 h at
37 °C. Incubation of the cells with antiphosphoMCM4,
anti-C23 or anti-HA sera (2.5 lgÆmL
)1
) was performed by
incubation overnight at 4 °C in the above blocking solu-
tion. Cells were washed with the same solution and then

incubated with Cy3-conjugated anti-rabbit or -mouse sera
(Jackson Immuno-Research, West Grove, PA) and
MCM4 phosphorylation in mammalian cells Y. Komamura-Kohno et al.
1236 FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS
FITC-conjugated anti-(mouse IgG) or anti-(rabbit IgG)
sera (Cappel, Durham, NC) for 1.5 h at 37 °C in the block-
ing solution. Washed cells were stained with 2 lgÆmL
)1
DAPI for 15 min at room temperature. After washing with
NaCl ⁄ P
i
, cells were mounted in 90% glycerol and 10%
NaCl ⁄ P
i
solution containing 1,4-diazabicyclo[2,2,2]-octane
(DABCO, Sigma) (2.3%) and observed using fluorescence
microscopy (AX-80, Olympus, Tokyo, Japan). In the
experiment shown in Fig. 9, cells that had been incubated
with Cy3-conjugated anti-rabbit sera were refixed, treated
with 4 m HCl for 30 min at room temperature and incuba-
ted with rat anti-BrdU sera (clone BU1 ⁄ 75; Harlan Sera
Laboratory, Bicester, UK) followed by the incubation with
FITC-conjugated anti-(rat IgG) sera (Cappel).
Cells labeled with BrdU (50 lm) for 15 min were fixed
and permeabilized as reported for the experiments shown in
Fig. 5. To quantify nuclear DNA content, cells were stained
with DAPI solution (5.7 mm DAPI, 1 ⁄ 10 concentrated
McIlvaine buffer, pH 7.0, 0.15 m NaCl, 0.004 m KCl) for
30 min, rinsed in McIlvaine buffer and mounted with
McIlvaine buffer ⁄ glycerol mixture (1 : 1 v ⁄ v). To detect

incorporated BrdU, cells were pretreated with the following
reactions of a DNA nicking with 0.5 n HCl for 5 min at
room temperature and a mild digestion with ExoIII nucle-
ase (0.5 UÆmL
)1
, Toyobo, Osaka, Japan) for 90 min at
37 °C. Linearity of DAPI-stained DNA content of a nuc-
leus was preserved throughout such treatments; i.e. the
DNA content of G
2
, M cells are almost double that of G
1
cells and the DNA content of S cells ranges between that
of G
1
cells and G
2
or M cells (data not shown). BrdU was
detected by the incubation with anti-(BrdU rat IgG) sera
(Sera Laboratory) and then with anti-(rat IgG) sera conju-
gated with FITC (Cappel). Phosphorylated MCM4 proteins
were detected by incubating with antiphosphoMCM4 sera
and then with anti-(rabbit IgG) conjugated with Cy3 (Jack-
son Immuno-Research).
Microfluorometry
An improved method for multiparametric microfluorometry
[31,32] was used to measure dual parameters on an identical
cell for the experiments in Fig. 5. Cells were selected under
phase-contrast illumination, and each was brought to the
center of the field of the microfluorometer. First, the intensity

of DAPI fluorescence of a nucleus was measured using a
UG 306-380 excitation filter and a LP 410 barrier filter
(Zeiss, Jena, Germany). The position of the cell in reference
to the x- and y-axes of the scanning stage was recorded using
a microcomputer (PC-9801UX, NEC, Tokyo, Japan). Sec-
ond, the fluorescence intensity of the immunocytochemically
stained phosphoMCM4 protein was measured using a
BP 450-490 excitation filter and a LP 520 barrier filter
(Zeiss). Data were processed using two microcomputers (PC-
9801 UX, NEC; Macintosh Quadra950, Apple Computers)
with software, some of which was designed for this study.
Image cytometry
To independently quantify nuclear and cytoplasmic phos-
phoMCM4 protein content, an imaging method was devel-
oped for the experiments shown in Fig. 5. In addition, the
cell-cycle phases of individual cells were identified based on
the dual parameters of DNA content and BrdU incorpor-
ation. Images of triple-stained cells on a slide were succes-
sively captured using a cooled CCD camera (SenSys,
NIPPONROPER, Tokyo, Japan) equipped to an AX-80
fluorescence microscope (Olympus), controlled by a soft-
ware (IpLab, Scananalytic, Rockville, MD) installed on a
microcomputer (Power Mac G3, Apple Computers). DAPI
image was measured using a BP360-370 excitation filter and
a BA420-460 barrier filter (Carl Zeiss Japan, Tokyo,
Japan). The FITC image was measured using a UG BP470-
490 excitation filter and a BA515-550 barrier filter. Cy3
image was measured using a BP520-550 excitation filter and
a BA580 barrier filter. By using personally developed
scripts of iplab software, images were processed to correct

shading effects and gray value fluctuation. Individual cell
shapes were obtained from the images stained with anti-
phosphoMCM4 sera following the interactive segmentation
procedure. The nuclear image was extracted from DAPI-
stained image. A particle analysis program was carried on
the nuclear and cell image, respectively. To make a corres-
pondence between data from a cell and that of a nucleus,
the nucleus measurement number was renumbered accord-
ing to that of the cell.
Confocal laser scanning microscope
Confocal laser scanning microscope observations were per-
formed using a MRC 1024 (Bio-Rad) mounted on an Axio-
plan microscope (Zeiss). Two stains (FITC and Cy3) were
excited simultaneously at wavelengths of 488 and 568 nm
emitted from a Kr ⁄ Ar ion laser followed by detection at
522 and 605 nm, respectively. Three stains (FITC ⁄ BrdU,
Cy3 ⁄ MCM and TOTO-3 ⁄ DNA) were excited simulta-
neously at wavelengths of 488, 552 and 642 nm emitted
from a Kr ⁄ Ar ion laser followed by detection at 522, 570
and 660 nm.
Electron microscope observation
HeLa cells treated with Triton were fixed with 4% parafor-
maldehyde and 0.1% glutaraldehyde in 0.1 m sodium phos-
phate buffer (pH 7.2) for 1 h in an ice bath and the
suspension was mixed with an equal volume of 1.5% low-
melting-point agarose (Sigma type VII). Hardened agarose
was incubated with NaCl ⁄ P
i
[-] (Mitsubishi Kagaku Iatron
Inc., Tokyo, Japan) containing 15% sucrose and then 25%

sucrose overnight. The agarose was dropped in a Tissue-Tek
OCT compound (Sakura Finetechnical Co. Ltd, Japan) and
frozen using dry ice. Frozen sections were obtained by taking
Y. Komamura-Kohno et al. MCM4 phosphorylation in mammalian cells
FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS 1237
10 lm slices of the compounded agarose using a Leica
CM3050 Cryomicrotome. Sections were incubated with 20%
Blockace in NaCl ⁄ P
i
. They were incubated with P-32 anti-
bodies ( 2 lgÆmL
)1
) diluted with 5% Blockace (Dai-nippon
Pharmaceuticals, Osaka, Japan) in NaCl ⁄ P
i
for 48 h at 4 °C
and then incubated for 48 h at 4 °C with anti-(rabbit IgG)
sera linked to 1 nm gold (Ultra-small; Aurion, Netherland)
that had been diluted 40-fold with 5% Blockace in NaCl ⁄ P
i
.
After washing with NaCl ⁄ P
i
for 3 h, the sections
were incubated with 2% glutaraldehyde in phosphate-
buffered saline for 30 min and then washed with phosphate-
buffered saline. The sections were incubated with 50 mm
Hepes–NaOH, pH 5.8 for 45 min and then with distilled
water for 30 min. They were incubated with an HQ silver kit
(Nanoprobe, Gibson Research, CA) for 8–10 min at room

temperature in the dark and then washed with distilled water.
Sections were re-fixed with 0.5% OsO
4
in distilled water for
20 min. After dehydration, the sections were embedded in
Epoxy resin (Epon 812, TAAB, Aldermaston, England).
Thin sections were obtained by cutting on a LeicaUltracut
UCT ultramicrotome and collected on grids. Sections were
contrasted by exposure with 4% uranyl acetate and then
observed using a JEOL JEM-1230 transmission microscope.
Acknowledgements
We thank Dr Taku Chibazakura for his useful sugges-
tions. This study was supported in port by a grant-
in-aid for scientific research from the Ministry of
Education, Science, Sports and Culture of Japan.
References
1 Tye BK (1999) MCM proteins in DNA replication.
Annu Rev Biochem 68, 649–686.
2 Bell SP & Dutta A (2002) DNA replication in eukaryo-
tic cells. Annu Rev Biochem 71, 333–374.
3 Forsburg SL (2004) Eukaryotic MCM proteins: beyond
replication initiation. Microbiol Mol Biol Rev 68, 109–131.
4 Ishimi Y (1997) A DNA helicase activity is associated
with an MCM4–6, and )7 protein complex. J Biol
Chem 272, 24508–24513.
5 Ishimi Y, Komamura Y, You Z & Kimura H (1998)
Biochemical function of mouse minichromosome main-
tenance 2 protein. J Biol Chem 273, 8369–8375.
6 Sato M, Gotow T, You Z, Komamura-Kohno Y,
Uchiyama Y, Yabuta N, Nojima H & Ishimi Y (2000)

Electron microscopic observation and single-stranded
DNA binding activity of the Mcm4,6,7 complex. J Mol
Biol 300, 421–431.
7 Itzhaki JE, Gilbert CS & Porter AC (1997) Construc-
tion by gene targeting in human cells of a ‘conditional’
CDC2 mutant that rereplicates its DNA. Nat Genet 15,
258–265.
8 Nguyen VQ, Co C & Li JJ (2001) Cyclin-dependent
kinases prevent DNA re-replication through multiple
mechanisms. Nature 411, 1068–1073.
9 Hendrickson M, Madine M, Dalton S & Gautier J
(1996) Phosphorylation of MCM4 by cdc2 protein
kinase inhibits the activity of the minichromo-
some maintenance complex. Proc Natl Acad Sci USA
93, 12223–12228.
10 Fujita M, Yamada C, Tsurumi T, Hanaoka F, Mats-
uzawa K & Inagaki M (1998) Cell cycle- and chromatin
binding state-dependent phosphorylation of human
MCM heterohexameric complexes. A role for Cdc2
kinase. J Biol Chem 273, 17095–17101.
11 Pereverzeva I, Whitmire E, Khan B & Coue
´
M (2000)
Distinct phosphoisoforms of the Xenopus Mcm4 protein
regulate the function of the Mcm complex. Mol Cell
Biol 20, 3667–3676.
12 Findeisen M, El-Denary M, Kapitza T, Graf R &
Strausfeld U (1999) Cyclin A-dependent kinase activity
affects chromatin binding of ORC, Cdc6, and MCM in
egg extracts of Xenopus laevis. Eur J Biochem 264, 415–

426.
13 Ishimi Y & Komamura-Kohno Y (2001) Phosphoryla-
tion of Mcm4 at specific sites by cyclin-dependent
kinase leads to loss of Mcm4,6,7 helicase activity. J Biol
Chem 276, 34428–34433.
14 Ishimi Y, Komamura-Kohno Y, You Z, Omori A &
Kitagawa M (2000) Inhibition of Mcm4,6,7 helicase
activity by phosphorylation with cyclinA ⁄ Cdk2. J Biol
Chem 275, 16235–16241.
15 Mineo C, Murakami Y, Ishimi Y, Hanaoka F &
Yamada M (1986) Isolation and analysis of a mamma-
lian temperature-sensitive mutant defective in G2 func-
tions. Exp Cell Res 167, 53–62.
16 van den Heuvel S & Harlow E (1993) Distinct roles for
cyclin-dependent kinases in cell cycle control. Science
262, 2050–2054.
17 Todorov IT, Attaran A & Kearsey SE (1995) BM28, a
human member of the MCM2–3)5 family, is displaced
from chromatin during DNA replication. J Cell Biol
129, 1433–1445.
18 Madine MA, Khoo C-Y, Mills AD, Musahl C & Las-
key RA (1995) The nuclear envelope prevents reinitia-
tion of replication by regulating the binding of MCM3
to chromatin in Xenopus egg extracts. Curr Biol 5,
1270–1279.
19 Krude T, Musahl C, Laskey RA & Knippers R (1996)
Human replication proteins hCdc21, hCdc46 and
P1Mcm3 bind chromatin uniformly before S-phase and
are displaced locally during DNA replication. J Cell Sci
109, 309–318.

20 Labib K, Tercero JA & Diffley JF (2000) Uninterrupted
MCM2-7 function required for DNA replication fork
progression. Science 288, 1643–1647.
MCM4 phosphorylation in mammalian cells Y. Komamura-Kohno et al.
1238 FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS
21 Lee J-K & Hurwitz J (2001) Processive DNA helicase
activity of the minichromosome maintenance proteins 4,
6, and 7 complex requires forked DNA structures. Proc
Natl Acad Sci USA 98, 54–59.
22 You Z, Ishimi Y, Mizuno T, Sugasawa K, Hanaoka F
& Masai H (2003) Thymine-rich single-stranded DNA
activates Mcm4 ⁄ 6 ⁄ 7 helicase on Y-fork and bubble-like
substrates. EMBO J 22, 6148–6160.
23 Lee J-K & Hurwitz J (2000) Isolation and characteriza-
tion of various complexes of the minichromosome main-
tenance protein of Schizosaccharomyces pombe. J Biol
Chem 275, 18871–18878.
24 Kaplan DL, Davey MJ & O’Donnell M (2003)
Mcm4,6,7 uses a ‘pump in ring’ mechanism to unwind
DNA by steric exclusion and actively translocate along
a duplex. J Biol Chem 278, 49171–49182.
25 Lam YW, Trinkle-Mulcahy L & Lamond AI (2005) The
nucleolous. J Cell Biol 118, 1335–1337.
26 Dimitrova DS, Prokhorova TA, Blow JJ, Todorov IT &
Gilbert DM (2002) Mammalian nuclei become licensed
for DNA replication during late telophase. J Cell Sci
115, 51–59.
27 Prasanth SG, Mendez J, Prasanth KV & Stillman B
(2004) Dynamics of pre-replication complex proteins
during the cell division cycle. Phil Trans Soc Lond B

Biol Sci 359, 7–16.
28 Ishimi Y, Komamura-Kohno Y, Yamada K & Naka-
nishi M (2003) Identification of MCM4 as a target of
the DNA replication block checkpoint system. J Biol
Chem 278, 24644–24650.
29 Ishimi Y, Komamura-Kohno Y, Karasawa-Shimizu K
& Yamada K (2004) Levels of MCM4 phosphorylation
and DNA synthesis in DNA replication block check-
point control. J Struc Biol 146, 234–241.
30 You Z, Komamura K & Ishimi Y (1999) Biochemical
analysis of the intrinsic Mcm4–Mcm6–Mcm7 DNA heli-
case activity. Mol Cell Biol 19, 8003–8015.
31 Tanaka S (1990) Methods of successive multiparametric
cytochemistry and microfluorometry on identical cells
with special reference to cell cycle phases in a chick
embryo. Exp Cell Res 186, 6–14.
32 Tanaka S, Ueda T, Nakajima K & Higashinakagawa T
(1996) Replication patterns of repetitive DNA sequences
on the W chromosome are altered during development
of the chick embryo. Exp Cell Res 223, 233–241.
Supplementary material
The following supplementary material is available
online:
Fig. S1. Staining with P-3 antibodies of HeLa cells
pulse-labeled with BrdU. Logarithmically growing
HeLa cells pulse-labeled with BrdU were stained with
P-3 antibodies. Staining with P-3 antibodies, anti-BrdU
sera and DAPI, and a cell image to determine area are
shown. These are original data for quantification in
Fig. 5A.

This material is available from ckwell-
synergy.com
Y. Komamura-Kohno et al. MCM4 phosphorylation in mammalian cells
FEBS Journal 273 (2006) 1224–1239 ª 2006 The Authors Journal compilation ª 2006 FEBS 1239