DEVELOPMENTAL BIOLOGY 200, 182–197 (1998)
ARTICLE NO. DB988961
Cdk2 Activity Is Dispensable for the Onset of DNA
Replication during the First Mitotic Cycles
of the Sea Urchin Early Embryo
Jean-Luc Moreau, François Marques, Abdelhamid Barakat,
Philippe Schatt, Jean-Claude Lozano, Gérard Peaucellier,
André Picard, and Anne-Marie Genevière1
Laboratoire Arago, URA 2156, F66650, Banyuls-sur Mer, France
Earlier work reported the important role of Cdk2 as a regulator of DNA replication in somatic cells and in Xenopus extracts.
In the present report we analyze in vivo the involvement of Cdk2 in DNA replication during early embryogenesis using the
first mitotic cycles of sea urchin embryos. Unfertilized Sphaerechinus granularis eggs are arrested after the second meiotic
cytokinesis. Fertilization resumes the block and induces DNA replication after a short lag period, making sea urchin early
embryo a good model for studying in vivo the onset of DNA replication. We show that Cdk2 as well as its potential partner
cyclin A are present in the nucleus in G1 and S phase and therefore available for DNA replication. In accordance with data
obtained in Xenopus egg extracts we observed that Cdk2 kinase activity is low and stable during the entire cycle. However,
in contrast with this in vitro system in which Cdk2 activity is required for the onset of DNA replication, the specific
inhibition of Cdk2 kinase by microinjection of the catalytically inactive Cdk2-K33R or the inhibitor p21Cip1 does not
prevent DNA replication. Because olomoucine, DMAP, and emetine treatments did not preclude DNA synthesis, neither
cyclin A/Cdk1 nor cyclin B/Cdk1 kinase activities are necessary to replace the absence of Cdk2 kinase in promoting DNA
replication. These data suggest that during early embryogenesis Cdks activities, in particular Cdk2, are dispensable in vivo
for the initiation step of DNA replication. However, the specific localization of Cdk2 in the nucleus from the beginning of
M phase to the end of S phase suggests its involvement in other mechanisms regulating DNA replication such as inhibition
of DNA re-replication and/or that its regulating role is achieved through a pathway independent of the kinase activity. We
further demonstrate that even after inhibition of Cdk activities, the permeabilization of the nuclear membrane is required
to allow a second round of DNA replication. However, in contrast to Xenopus egg extracts, re-replication can take place in
the absence of DMAP-sensitive kinase. © 1998 Academic Press
Key Words: cell cycle; Cdk2; cyclin A; DNA replication; sea urchin eggs.
INTRODUCTION
Progression through the cell cycle requires the sequential
activation and inactivation of a family of kinases (Cdks)
whose activity depends on their association with cyclins
(reviewed in Nigg, 1995). In budding yeast, an ordered
sequence of expression of cyclins, through their association
with cdc28, the yeast prototype of Cdk, drives the cycle
through S and M phase. The CLN cyclins control the
transition through start, whereas CLBs regulate entry into S
phase (CLB 5/6) and mitosis (CLB 3/4) (reviewed in Reed,
1992; Nasmyth, 1996). In Schizosaccharomyces pombe cig2
1
To whom correspondence should be addressed. Fax: (33) 4 68 88
73 98. E-mail: [email protected].
182
is the major partner of cdc2 in G1 phase, whereas cdc13/
cdc2 complexes bring about the onset of mitosis (reviewed
in Stern and Nurse, 1996).
In contrast to yeast, several catalytic subunits interacting
with cyclins have been identified in addition to cdc2 (Cdk1)
in higher eukaryotes (reviewed in Sherr 1993, 1996; Doree
and Galas 1994; Nigg, 1995). The transient appearance of
these different cyclin/Cdk complexes drives somatic cell
cycle events such as early G1 transition (cyc D/Cdk4 – 6),
DNA replication (cyc E/Cdk2, cyc A/Cdk2), and cell division (cyc A/Cdk1, cyc B/Cdk1).
However, several lines of evidence suggest that the role of
some cyclin/Cdk complexes could be different during the
embryonic and somatic cell cycles. In particular, it has been
established that cyclin A-dependent kinase plays a key role
0012-1606/98 $25.00
Copyright © 1998 by Academic Press
All rights of reproduction in any form reserved.
Cdk2 and DNA Replication in Sea Urchin Early Embryogenesis
in regulating S phase progression in mammalian cells, since
microinjection of anti-cyclin A antibody (Girard et al.,
1991; Pagano et al., 1992) or antisense cyclin A plasmid
(Girard et al., 1991; Zindy et al., 1992) prevents the entry of
cells into S phase. However, cyclin A mutants in Drosophila embryos arrest only at G2 of the 16th cell cycle (Lehner
and O’Farrell, 1989), suggesting that cyclin A is essentially
required for M phase during early embryogenesis. Experiments in Xenopus extracts have given rise to puzzling
results. Although cycloheximide-treated eggs or interphasic
extracts, which are devoided of cyclin A, could support
DNA replication (Harland and Laskey, 1980; Blow and
Laskey, 1988), cyclin A/Cdk complexes efficiently promoted DNA synthesis in Cdk-depleted extracts (Strausfeld
et al., 1996). Nevertheless, evidence is lacking that cyclin
A-dependent kinases do not play any role in DNA replication during early embryogenesis or to propose an explanation for these apparently contradictory results. The potential partner of cyclin A, Cdk2 has been shown to be required
for DNA replication in mammalian cells, since expression
of a dominant-negative mutant of Cdk2 (Van der Heuvel
and Harlow, 1993) or microinjection of anti-Cdk2 antibody
(Pagano et al., 1993) blocked S phase progression. This
activity is also necessary for the onset of DNA replication
in Xenopus egg extracts as DNA replication is blocked by
immunodepletion of Cdk2 (Fang and Newport, 1991) or
cyclin E (Jackson et al., 1995) or by the Cdk2 kinase
inhibitor p21Cip1 (Strausfeld et al., 1994; Jackson et al.,
1995). More recent studies also suggested that the cell
cycle-dependent compartmentalization of active Cdk2 kinase within nuclei participates in negatively regulating
DNA replication during the cell cycle in Xenopus extracts
(Hua et al., 1997). However, an involvement of Cdk2 in the
onset of DNA replication during early embryogenesis, as
well as its potential role in the control of re-replication,
remain undemonstrated in vivo.
To further analyze the requirement for Cdk2 in vivo
during replication in early embryonic cell cycles, we decided to examine the activity, localization, and role of this
kinase and its potential partner, cyclin A, in the first
mitotic cell cycle of sea urchin embryos. These cells are
especially suitable for studying the mechanisms leading to
DNA replication during early embryogenesis in vivo. Unfertilized sea urchin eggs are arrested in G1 of the cell cycle.
After fertilization sperm chromatin decondenses and forms
a male pronucleus which fuses with the female pronucleus,
and S phase occurs in the zygotic nucleus. The fertilized egg
then divides and subsequent cell cycles proceed without
significant increase in mass as rapidly alternating rounds of
highly synchronous S and M phases. We previously showed
that cyclin B/Cdk1 activity, which is repressed during the
first few minutes after fertilization, undergoes partial activation before S phase, at the same time that cyclin B is
present in the nucleus (Geneviere-Garrigues et al., 1995).
This demonstrated in vivo that this level of kinase activity
(about one-third of the level in M phase) does not inhibit
DNA replication and suggested that cyclin B/Cdk1 could
183
participate to the regulation of some events occurring in S
phase. Our present observations strongly suggest that the
cyclin/Cdk2 activities, but also cyclin/Cdk1 activities, are
dispensable for the onset of DNA replication during the sea
urchin early embryogenesis; however, the localization of
Cdk2 and of cyclin A suggest that they could participate in
the inhibition of DNA re-replication or play a role outside
of the kinase activity.
MATERIALS AND METHODS
Materials
All the following drugs were purchased from Sigma and stored as
stock solutions at 220°C. Emetine was prepared as a 1022 M stock
solution in water. Aphidicolin was dissolved in DMSO as a 1003
solution and used at a 20 mg/ml final concentration. DMAP was
disolved in filtered seawater as a 15 mM solution. Olomoucine was
dissolved to 50 mM in DMSO.
Animals and Handling of Gametes
Sea urchins, Sphaerechinus granularis, were collected over the
year in the Mediterranean Sea near Banyuls (France) and kept under
running seawater until use. Shedding of gametes was induced by
injection of 0.1 ml of 0.2 M acetylcholine through the perioral
membrane. Eggs were filtered through a 150 mesh nylon filter to
remove the debris and rinsed once in 5 mm Millipore-filtered
seawater. Batches containing more than 5% germinal vesicle-stage
oocytes were discarded. Dry sperm was diluted 106-fold before use.
A 5% (v/v) egg suspension was fertilized under slow agitation. For
immunofluorescence experiments eggs were fertilized in the presence of 1 mM ATA (3-amino 1,2,4-triazole, Showman and Foerder,
1979) to inhibit hardening of the fertilization membrane. Only
batches with at least 90% fertilized eggs were further used.
cDNA Cloning of Sea Urchin Cdk1 and Cdk2
To isolate Cdk1 and Cdk2, degenerate oligonucleotides were
synthesized derived from two highly conserved regions of Cdk
proteins (Lehner and O’Farell, 1990; Meyerson et al., 1992),
EKIGEGTY for the 59 oligonucleotide and WYRAPE for the 39
oligonucleotide. PCR amplification was carried out using as template a first-strand cDNA prepared from total RNA of unfertilized
eggs with AMV-reverse transcriptase (Promega) and oligo(dT) primers (PCR conditions: 30 cycles of 94°C for 1 min, 50°C for 2 min,
72°C for 2 min, and a final incubation at 72°C for 10 min). PCR
products were purified and subcloned into M13mp19 and the
inserts were sequenced (Sequenase version 2.0 kit, United States
Biochemical). The cDNAs corresponding to the complete coding
sequences of Cdk1 and Cdk2 were obtained by screening S.
granularis libraries constructed from mRNA respectively of twocell-stage embryos (Lozano et al., 1998) and unfertilized eggs
(SG/UF-pADNS library). In the latter case cDNA was synthesized
as recommended by the kit manufacturer (Amersham) using random primers and mRNA selected by two passes over an oligo(dT)
column (Boehringer Mannheim). Double-stranded cDNA was ligated to HindIII–NotI adapters, size selected (.0.5 kb) by fractionation on low-melting point gel agarose and ligated to HindIIIdigested and dephosphorylated pADNS vector. The ligated
products were transformed into Escherichia coli SCS1 cells (Strat-
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
184
Moreau et al.
PCR products homologous to Cdk2 and the partial cDNA clone
encode a modified PSTAIR sequence in which the second Ser of the
motif changes to an Ala. PCR amplifications from first-strand
cDNA obtained from independently prepared mRNA led to the
same IALLKE sequence. Therefore, this IALLKE sequence is not an
artifact but an authentic sequence present in the second isolated
sea urchin Cdk. Similarly, goldfish Cdk1 displays a PSTAVR
sequence instead of the PSTAIR sequence common to Cdk1 in
other species (Kajiura et al., 1993). A phylogenetic tree produced by
the DARWIN program (Gonnet et al., 1992) classified the two
cDNAs as the sea urchin homologue of Cdk1 and Cdk2, respectively (data not shown). Moreover, an alignment of the c-terminal
peptides of Cdk1, Cdk2, and Cdk3 demonstrated a closer relationship between sea urchin Cdk2 and its species homologue than
between Cdk2 and Cdk1 or 3 (Fig. 1B).
The predicted MWs of the proteins encoded by these two cDNAs
are 34.6 kDa for Cdk1 and 34.1 kDa for Cdk2.
Recombinant Cdk2 protein was produced by cloning the coding
sequence in pET-21a (Novagen) and in pGEX-KG to obtain a
GST-fusion protein. The Cdk2 recombinant protein produced in
pET-21a was purified by preparative gel electrophoresis and the
GST-fusion protein was purified on glutathione Sepharose according to the instructions of the manufacturer (Pharmacia) and dialyzed overnight against 100 mM NaCl, 15 mM Tris–HCl (pH 7.4).
Antibodies Characterization
FIG. 1. (A) Alignment of amino acid sequences of Sphaerechinus
granularis Cdk1 and Cdk2. (:) Identical residues, (.) amino acids
with similar biochemical characteristics. (B) Comparison of the
C-terminal peptide sequences of Cdk1, Cdk2, and Cdk3. Identical
residues specifically found in the Cdk2 sequences from different
species are shown by stars. Amino acid sequences indicated by
underline correspond to the peptides used for rabbit immunization.
agene) to generate the SG/UF-pADNS library. Hybridizations were
performed with 32P-labeled probes (Prime-a-gene labeling system,
Promega) derived from the PCR products in HB (53 SSC) at 65°C
for 16 h and the filter was washed to a final stringency of 0.13
SSC– 0.1% SDS. A Marathon cDNA amplification kit (ClontechChenchik et al., 1995) was used in addition to isolate the fulllength Cdk2 ORF from UF egg mRNA, using the conditions
recommended by the manufacturer.
The sequences of full-length cDNAs were obtained on both
strands from restriction fragment subclones using universal or
reverse primers present on vectors or synthesized specific oligonucleotides. The GenBank data were searched for homologous sequences, using the FASTA or BLAST computer programs, and
aligned by using the CLUSTAL multiple sequence alignment
program.
The cDNA and deduced amino acid sequences of Cdk1 and Cdk2
sea urchin homologues are shown in Fig. 1A. The sequence data are
available from EMBL Data Bank under Accession Nos. AJ 225013
(Cdk1) and AJ 224917 (Cdk2). Cdk1, 2, and 3 are characterized by a
16-aa sequence called the PSTAIR motif (EGVPSTAIREISLLKE)
that is exactly conserved in the sea urchin Cdk1. Both the isolated
Rabbit polyclonal antibodies were raised against the highly
specific 11 and 13 amino acid C-terminal peptides of Cdk1 and
Cdk2 proteins respectively (see Fig. 1). The anti-Cdk2 antibody was
affinity purified on GST-Cdk2 coupled to Hitrap-NHS column
(Pharmacia). The selectivity of the antibodies was checked by
immunoblotting experiments on the bacterially produced Cdk2 or
after immunoprecipitation of sea urchin endogenous Cdk1 (Fig. 2).
Note in Fig. 2A (lane 1) that in whole homogenate the anti-Cdk2
antibodies recognize only the 34-kDa band corresponding to endogenous Cdk2.
A rabbit polyclonal anti-sea urchin cyclin A antibody was raised
against the recombinant S. granularis protein; its specificity has
been described previously (Genevière-Garrigues et al., 1995).
Immunoprecipitation, Western Blot,
and Kinase Assay
Aliquots of cell suspension containing 50 ml of cell volume were
briefly centrifuged (1000g, 1 min) and the pellets were homogenized
by sonication in 500 ml of an ice-cold homogenization buffer containing 0.2 M KCl, 60 mM b-glycerophosphate, 15 mM EGTA, 10 mM
MgCl2, pH 7.3, 0.1% Triton X-100, 1 mM benzamidine, 1% soybean
trypsin inhibitor (SBTI) (w/v), 1 mM PMSF. After centrifugation
(10,000g, 10 min) the supernatant was used for kinase assay, immunodetection by Western blot, or affinity purification on p13-suc1
Sepharose beads (Dunphy et al., 1988).
Immunoprecipitations were done by adding 2 mg of anti-Cdk2 or
anti-cyclin A affinity-purified antibody to the supernatant. After 30
min of continuous shaking at 4°C, 20 ml of protein A–Sepharose
beads (Sigma) was added and the suspension was incubated for
another 30 min. The beads were then pelleted, rinsed three times in
150 mM NaCl, 50 mM Tris–HCl pH 7.5, 0.1% Triton X100, and
then once with the same buffer lacking Triton. In the cyclin A
immunoprecipitation reported in Fig. 4C, the cell pellets were
homogenized in RIPA buffer (150 mM Na Cl, 1% NP-40. 0.5%
deoxycholate, 0.1% SDS, 50 mM Tris–HCl, pH 8).
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
Cdk2 and DNA Replication in Sea Urchin Early Embryogenesis
185
For immunoblot analysis, proteins were resolved on 12.5%
SDS–PAGE and transferred to nitrocellulose membranes (Hybond
C, Amersham). Membranes were saturated for 1 h in buffer S: 0.2 M
NaCl, 5 mM MgCl2, 1 mM CaCl2, 25 mM Tris–HCl pH 7.5, 3%
polyvinylpolypyrrolidone, 0.05% Tween 20 and probed with the
appropriate antibody diluted 1/300 for cyclin A and 1/500 for Cdk2
in buffer S. After overnight incubation, membranes were washed in
buffer S and the bound antibodies were detected with mouse
anti-rabbit IgG conjugated to alkaline phosphatase (Sigma).
For kinase assays, 10 ml of phosphorylation mix containing 10
mM MgCl2, 100 mM ATP, 1 mg histone H1, 80 mM Hepes, pH 7.4,
100 mCi/ml [32P]ATP (3000 Ci/mmol) was added to the immobilized complexes. After 10 min at 20°C the reaction was stopped by
addition of 10 ml of 43 Laemli buffer and boiling for 3 min. The
samples were separated by 12.5% SDS–PAGE. After drying of the
gel and autoradiography, the band containing histone H1 was
excised and counted in a liquid scintillation counter.
Immunofluorescence
Eggs were processed as previously described (Picard et al., 1988).
For cyclin A and Cdk2 immunolocalization affinity-purified anticyclin A (1/100) or anti-Cdk2 (1/100) antibody was used as the
primary antibody and a Texas red– goat anti-rabbit (Amersham,
diluted 1/300) as the secondary antibody. DNA staining with 0.1
mg/ml Hoechst 33258 was performed during the last rinse after the
second antibody incubation.
Monitoring DNA Replication
DNA synthesis in fertilized embryos was monitored by two
different procedures. In the first 250 mCi of [3H]methylthymidine
(Amersham, 120 Ci/mmol) was added to a 50-ml egg suspension 15
min before fertilization. Aliquots of 1 ml (triplicates) were taken as
a function of time and processed for measurement of [3H]thymidine
incorporation into DNA as previously described (GeneviereGarrigues et al., 1995). To determine simultaneously the [3H]thymidine uptake, the embryos were washed rapidly in filtered seawater and lysed by adding 1/10 vol 1 N NaOH and then
neutralizing with 1 N HCl. The final pellet was resuspended in BCS
counting scintillant (Amersham) and counted.
FIG. 2. Characterization of anti-Cdk1 and anti-Cdk2 antibodies.
(A) 0.3 mg of protein from a whole extract of 40 min postfertilization eggs (lane 1), 10 ng of bacterially produced Cdk2 (lane 2), and
50 ml of 40 min postfertilization eggs immunoprecipitated with
anti-Cdk2 antibodies (lane 3) were immunoblotted with antibodies
to Cdk2. This shows clearly that anti-Cdk2 antibodies recognize in
vivo a protein of 34 kDa, the same molecular weight as the
bacterially produced protein. Fifty microliters of 40-min postfertilization eggs was immunoprecipitated with antibodies to Cdk1
(lane 4) or to cyclin B (lane 5) and processed for Western blot
analysis with anti-Cdk2 antibodies. This demonstrates that Cdk2
antibodies do not cross-react in Western blot with Cdk1 and that
Cdk1 antibodies do not recognize native Cdk2 protein in immunoprecipitation experiment. This confirms that the anti-Cdk2immunoreactive p34 protein is not associated to cyclin B. (B) The
efficiency of anti-Cdk1 and anti-cyclin B immunoprecipitation was
checked by immunoblotting the same amount of protein as in A,
lanes 4 and 5, with anti-Cdk1 antibodies (lanes 1 and 2). The same
amount of protein was immunoprecipitated with antibodies to
Cdk2 as in A (lane 3) and 10 ng of recombinant Cdk2 (lane 4) was
immunoblotted with anti-Cdk1 antibodies. This demonstrates that
Cdk1 antibodies do not cross-react with Cdk2 in Western blot.
However, Cdk2 antibodies slightly recognize native Cdk1 in immunoprecipitation experiment. To rule out the possibility that the
slight cross-reaction between antibodies to Cdk2 and Cdk1 will
affect the measurement of Cdk2 kinase activity, we immunoprecipitated 40-min postfertilization eggs (50 ml cell volume) with
anti-cdk2 antibodies (IP1) or preimmune igg (PI) and processed
them for H1 kinase activity. A second immunoprecipitation (IP2)
on the supernatant of IP1 demonstrates that about 80% of the
Cdk2-associated kinase activity (IP1 or 2-PI) is retained in the first
immunoprecipitation. this proved that Cdk2 antibodies immunoprecipitated the Cdk2 kinase activity and not a fraction of Cdk1
kinase activity that remained unchanged after immunodepletion
with anti-Cdk2 antibodies (data not shown).
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
186
Moreau et al.
In the second procedure, DNA synthesis was monitored in
individual cells by incorporation of bromodeoxyuridine. BrdU was
added to the egg suspension at a concentration of 0.1 mg/ml; after
incubation for the indicated times, eggs were fixed and processed as
previously described (Picard et al., 1996). Quantitation of BrdU
staining was accomplished with Image Quant computer software
using pictures photographed with Kodak 5042 EJP and scanned
with Canoscan 2700F.
RESULTS
Cdk2- and Cyclin A-Dependent Kinase Activities
during the First Embryonic Cell Cycle
The abundance and activity of Cdk2 and Cdk1 along
the cell cycle were examined in sea urchin embryos with
the polyclonal antibodies raised against the C-terminal
peptides of these two proteins (see Material and Methods). Immunoblots of p13-suc1 affinity-purified proteins
did not reveal any major changes in the Cdk1 (not shown)
and Cdk2 (Fig. 3A) abundance during the first mitotic
cycle. The Cdk1 and Cdk2 kinases were then immunoprecipitated with the respective affinity-purified antibodies and activities were measured with the common
substrate histone H1. Cdk2 kinase activity is low and did
not vary significantly from fertilization to mitosis (Fig.
3B). In contrast, Cdk1 activity increased 10-fold between
10 min postfertilization embryos and mitotic embryos
(Fig. 3C). The Cdk2 kinase activity remained unchanged
during the following cycles at least until blastula (data
not shown), the latest stage in which immunoprecipitated H1 kinase activity was checked.
As cyclin A is a potential partner of Cdk2, we next
examined the behavior of the cyclin A protein in S. granularis eggs using polyclonal antibodies raised against bacterially produced cyclin A. The level of cyclin A in immunoblots of p13-suc1-isolated proteins gradually increased after
fertilization (Fig. 4A), indicative of accumulation of cyclin
A/Cdk complexes. At the same time cyclin A immunoprecipitated from sea urchin embryos displayed an associated
H1 kinase activity which increased along the cell cycle,
rising at the beginning of S phase to reach a maximum at
NEBD (Fig. 4B). Because cyclin A was not clearly detected
in total protein extracts, we could not analyze the presence
of cyclin A in UF sea urchin eggs. To answer this question
we fertilized eggs in the presence of emetine (1024 M),
which inhibits protein synthesis, and monitored the
amount of cyclin A recovered on p13-suc1 beads. Under
these conditions no cyclin A was visualized in immunoblots after fertilization which confirm that all the cyclin
A/Cdk complexes formed after fertilization are generated
from the newly synthesized cyclin A (data not shown).
However, in contrast to previous reports, we noted that a
significant amount of cyclin B associated to Cdk1 is present
in UF eggs (Fig. 4A).
We then examined whether cyclin A interacted with
Cdk1 and/or Cdk2. The anti-cyclin A immunoprecipitates
contained anti-Cdk1- and anti-Cdk2-reactive 34-kDa pro-
FIG. 3. Levels of Cdk2 protein and Cdk2 kinase activity are stable
along the cell cycle. (A) Aliquots of 30 ml cell volume were taken at
given time after fertilization; the embryo extracts were affinitypurified on p13-suc1 Sepharose beads and processed for immunoblottting with anti-Cdk2 antibodies. (B and C) At given intervals
after fertilization aliquots of 50 ml cell volume were taken and H1
histone kinase activity was assayed in anti-Cdk2 (B) or anti-Cdk1
(C) immunoprecipitates. The reported values correspond to the
difference between anti-Cdk and preimmune IgG immunoprecipitations for a 10-min kinase assay and are representative of five
individual experiments. The mean level of Cdk2 activity is reported as dashed line in C.
teins in S (30 min) as in M (70 min) phases (Fig. 4C). To
estimate the relative contribution of Cdk1 and Cdk2 to the
cyclin A-dependent kinase activity during S phase, egg
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
Cdk2 and DNA Replication in Sea Urchin Early Embryogenesis
187
fected, showing that cyclin A-associated kinase activity
was mainly dependent on Cdk1, even if cyclin A/Cdk2
complexes were present at that time and could contribute
to a low proportion of the cyclin A-associated kinase
activity.
Subcellular Localization of Cdk2 and Cyclin A
along the First Mitotic Cell Cycle
FIG. 4. Fertilization induces synthesis of cyclin A that associates
with Cdk1 and Cdk2 proteins. However, the activation of the
cyclin A-dependent H1 kinase activity is mainly associated to
Cdk1. (A) Aliquots of embryos were taken every 10 min after
fertilization, processed for p13-suc1 affinity, and immunoblotted
with antibodies to cyclin B or cyclin A (A). (B) In an equivalent
experiment, cyclin A immunocomplexes (on 50 ml cell volume)
were assayed for H1 histone kinase activity (10 min assay). (C) The
Cdks interacting with cyclin A were analyzed in embryos at the
indicated times postfertilization (min). The proteins immunoprecipitated with anti-cyclin A antibodies were separated by SDS–
PAGE and the blots were probed with anti-Cdk1 and anti-Cdk2
antibodies.
extracts were depleted by immunoprecipitation with antibodies to Cdk1- and cyclin A-dependent H1 kinase activity
were compared in control and immunodepleted extracts
(Table 1). An almost complete depletion of Cdk1 (96%) left
11% of the cyclin A-dependent H1 kinase activity unaf-
The presence of a low level or an invariant activity
measured in whole extract does not exclude a role for the
corresponding kinase since it could be differentially localized during the cell cycle, leading to a transient local
accumulation of activity at a specific time. To investigate
this possibility, the subcellular localization of Cdk2 during
the first mitotic cell cycle was studied by immunofluorescence using affinity-purified antibodies (Fig. 5). A strong
Cdk2 labeling was detected in the male pronucleus in the
first minutes postfertilization. This intense labeling remained mainly associated with the male chromatin, visualized by Hoechst staining, even after fusion of the male
and female pronuclei (not shown). Later, when paternally
derived chromatin was no longer recognizable, which corresponds to the onset of DNA synthesis (30 min), immunoreactivity was visible on the whole chromatin of the zygotic
nucleus. Because there is no significant variation in Cdk2
abundance during the cell cycle, this suggests either a
nuclear translocation of Cdk2 from the cytoplasmic compartment and/or a repartition of Cdk2 from the male
pronucleus to the zygotic nucleus. Before the end of S phase
(40 min), Cdk2 nuclear staining moved to a dot-like pattern
at the nuclear periphery, being excluded from the chromatin. Cdk2 was then rapidly released to the cytoplasm where
it became associated with centrosomal material (60 min).
Cdk2 moved back to the chromatin during prometaphase
(70 min). During the formation and fusion of karyomeres
the still condensed chromosomes showed an intense Cdk2
staining (95 min).
The cyclin A staining pattern while exhibiting some
similarities with that produced by Cdk2 antibodies also
showed some noteworthy differences in the temporal and
spatial distribution (Fig. 6). Cyclin A was not clearly detected before 10 min postfertilization, at which time it was
detected as a nuclear protein colocalizing with chromatin
(data not shown). This nuclear labeling remained intense
during the entire S phase and the beginning of G2 phase (30
to 50 min) contrasting with the Cdk2 labeling that left
chromatin at the end of S phase. Before the beginning of
prophase, cyclin A became associated to the mitotic asters
while a decreasing staining was still observed on chromatin. During metaphase (75 min) and anaphase (80 min),
cyclin A was dispersed in the mitotic apparatus region, a
pattern distinct from Cdk2, exclusively associated to the
chromatin at that time. Cyclin A progressively went back
to chromatin as the blastomere nuclei reformed (95 min).
The localization of Cdk2 and cyclin A in the nucleus
during S phase suggests a role for these proteins in ongoing
DNA replication.
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
188
Moreau et al.
TABLE 1
Cdk1 kinase activity
pmol incorporated in H1
Cyc A-dependent kinase activity
Control
Cdk1 depleted
Control
Cdk1 depleted
3.90 6 0.06
0.16 6 0.03
1.45 6 0.39
0.16 6 0.01
Note. Aliquots of 50 ml cell volume were harvested 30 min after fertilization; the embryo extracts were depleted with antibodies to Cdk1
covalently attached to protein A–Sepharose beads with dimethyl pimelimidate crosslinking (column 2 and 4) or mock-depleted with protein
A–Sepharose. The supernatants were then immunoprecipitated with protein A–Sepharose crosslinked to antibodies to Cdk1 (columns 1 and
2) or cyclin A (columns 3 and 4) and processed for a 10-min H1 kinase assay.
Inhibiting the Cdk Kinase Activity Does Not
Impede DNA Replication
The potential role of Cdk2 in promoting DNA synthesis
during the sea urchin embryonic cell cycles was investigated by microinjection of p21Cip1, an inhibitor of Cdk2
kinase in both mammalian cells and Xenopus eggs (Harper
et al., 1993; Xiong et al., 1993; Gu et al., 1993; Dulic et al.,
1994). We first verify that p21Cip1 at concentrations above
0.4 mM inhibit more than 90% of the Cdk2 kinase activity
measured in vitro from sea urchin embryo extracts (Cdk2 IP
in picomoles of 32P incorporated in H1, control: 0.65 6 0.02;
with p21Cip1 : 0.06 6 0.01).The microinjection of p21Cip1
(0.6 mM) in UF eggs left BrdU incorporation unaffected (Figs.
7A–71) but mitotic events were delayed probably due to the
partial inhibition of cyclin B/Cdk1 kinase at the concentration used (Fig. 7A, compare 3 and 2). The above results
suggest that the Cdk2 kinase activity is not necessary for
the onset of DNA replication. These results were reinforced
by data obtained from microinjection of the catalytically
inactive mutant protein Cdk2-K33R. This kinase-dead mutant has been shown to efficiently compete with endogenous cyclins and to inhibit appearance of normal effects of
the wild-type kinase (Van der Heuvel and Harlow, 1993).
The mutant protein was microinjected into UF eggs or
one-cell embryos during G2 phase and the effects on DNA
replication were examined respectively in the first and
second mitotic cell cycles. In both cases, DNA replication
was apparently unaffected as judged by BrdU incorporation
into chromatin (Fig. 7B). It should be noted that an early
effect of Cdk2-K33R could be observed in various eggs in
which the fusion of pronuclei was inhibited without affecting DNA replication. At high concentration this negative
dominant is known to titrate even the newly synthesized
cyclin B when microinjected into oocytes of another echinoderm species, the starfish Astropecten aranciacus (Picard
et al., 1996). Accordingly, microinjection of Cdk2-K33R
prior to fertilization delayed the nuclear envelope breakdown (NEBD) by about 20 min. Microinjection in G2 (50
min) no longer inhibited NEBD but disturbed mitotic
events, suggesting that Cdk2-K33R also associates with
cyclin B in sea urchin eggs and thus efficiently inhibits
Cdk/cyclin dimer formation.
The above results support the view that Cdk2 kinase
activity is not critical for DNA synthesis initiation. To
reinforce this conclusion the purine analogues olomoucine
and DMAP were used to inhibit Cdk2 kinase activity (Fig.
8). Olomoucine has been described as a potent competitive
inhibitor of Cdk2/cyclin A, Cdk2/cyclin E, and Cdk1/cyclin
B, in various plant and animal models (Vesely et al., 1994;
Abraham et al., 1995). At a concentration of 0.1 mM
olomoucine completely inhibited the Cdk2-dependent H1
kinase activity measured in vitro from sea urchin embryo
extracts and about 80% of the total Cdk-associated H1
kinase activity (Table 2). DMAP is a nonspecific Ser/Thr
kinase inhibitor that blocked the onset of DNA replication
in Xenopus egg extracts (Blow, 1993). The concentration
required to inhibit replication in this system (2 mM) is
slightly higher than the one necessary to abolish MPF
activity, in Xenopus (Blow, 1993) as well as in S. granularis
(0.4 mM, personal data). At this concentration, DMAP
inhibited 100% of the Cdk2 kinase activity and 97% of the
total H1 kinase activities measured in sea urchin embryo
extracts (Table 2); this last inhibition reached 99% at a
concentration of 5 mM. DMAP also inhibited Cdk in vivo
activation as shown in Table 2. It was also verified by
checking the mitotic indexes (NEBD and cleavage) that
DMAP inhibition is completed in less than 5 min in vivo
(data not shown). Embryos were analyzed for replication by
monitoring the proportion of nuclei which incorporated
BrdU in newly synthesized DNA. As judged by this parameter, DNA replication readily occurred in the presence of
olomoucine (0.1– 0.5 mM), although with a delay (Fig. 8A,
1), while in contrast cleavage was either significantly delayed (by 50 min at 0.1 mM) or completely inhibited (0.5
mM) (data not shown). Similarly, DMAP added to batches
of embryos within the first 10 min after fertilization at a
concentration of 2 mM did not inhibit BrdU incorporation
(Fig. 8A, 1) but completely blocked NEBD (data not shown).
The maximal amount of BrdU incorporated during the
first cell cycle was identical in control and DMAP- or
olomoucine-treated embryos as shown by immunofluorescence quantitation (Fig. 8A, 2). To rule out the possibility
that Cdk activities would not be completely inhibited in
vivo, emetine (1024 M), an inhibitor of protein synthesis
which prevents the formation of complexes between Cdks
and newly synthesized cyclin A and B, was added together
with DMAP to eggs 10 min postfertilization. It was verified
in this case (table 2) that the cdk activities remaining after
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
Cdk2 and DNA Replication in Sea Urchin Early Embryogenesis
189
FIG. 5. Evolution of Cdk2 immunofluorescence staining during the first mitotic cell cycle. Ten- to 95-min postfertilization embryos
were labeled with affinity-purified antisera to Cdk2 (A and C) and the DNA was visualized with Hoechst dye (B and D). Pictures were
taken at the indicated time (min) postfertilization. At 10 min male (upper right) and female pronuclei (lower left) were moving to the
center of the fertilized egg; fusion of pronuclei occurred at 20 min. Cdk2 was colocalized with chromatin during S phase (30 min) and
migrated to the nuclear periphery at the end of S phase (40 min). Cdk2 was found in cytoplasm associated to centrosomal material
during G2. At time of prometaphase (70 min), it became colocalized with chromatin again and remained associated with chromatin
during anaphase (85 min), telophase (95 min), and at the beginning of S phase of the second cell cycle (bar, 50 mm).
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
190
Moreau et al.
FIG. 6. Evolution of cyclin A localization during the first mitotic cell cycle. Postfertilization embryos taken at the indicated time (min)
were labeled with affinity-purified antisera to cyclin A (A, C) and the DNA was visualized with Hoechst dye (B, D). Cyclin A remained
colocalized with chromatin all along the S phase (30 and 40 min). It was translocated to the cytoplasm in G2 (50 min), although a significant
staining remained associated to chromatin at that time. Cyclin A was distributed on the mitotic apparatus during metaphase (75 min) and
anaphase (80 min) and became clearly associated to chromatin again in telophase (95 min) and during S phase of the second cell cycle (not
shown) (bar, 50 mm).
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
Cdk2 and DNA Replication in Sea Urchin Early Embryogenesis
191
FIG. 7. Inhibition of Cdk2 by microinjection of the p21Cip1 inhibitor or the catalytically inactive Cdk2-K33R does not inhibit DNA
synthesis. (A) p21Cip1 was microinjected to a concentration of 0.6 mM in unfertilized eggs and BrdU (0.1 mg/ml) was supplied at fertilization.
The incorporated BrdU was visualized in the 60-min postfertilization embryos (1). At 130 min the second cleavage was completed in the
control embryos (2) as visualized by Nomarski interference contrast, meanwhile p21Cip1 microinjected embryos were undergoing first
mitosis (3). (B) Cdk2-K33R was microinjected in unfertilized eggs (2) or in G2 phase embryos (3) to a final concentration of 0.2 mg/ml. BrdU
(0.1 mg/ml) was supplied respectively at fertilization or in G2 and the embryos were fixed 60 min later. The incorporated BrdU was
visualized with fluorescently labeled anti BrdU antibodies and compared with control (1). It was verified in parallel experiments (A and B)
that no BrdU was incorporated in the presence of aphidicolin (20 mg/ml) (not shown). (bar, 50 mm).
inhibition of cyclinA/Cdk and cyclinB/Cdk activation are
even more efficiently inhibited by DMAP (99.3%). this
treatment did not further modify BrdU incorporation (data
not shown). this set of results confirm that Cdk2 complexes
are not essential for early steps of DNA replication. they
also suggest that no single cdk activity is required to replace
Cdk2 in promoting dna replication as discussed later on.
It should be noted that DMAP as well as olomoucine
slow down BrdU incorporation during the first cell cycle, in
contrast to the emetine treatment. They should be several
reasons for a delayed incorporation. First, BrdU uptake
could be affected by DMAP treatment, leading to an in-
crease in the time necessary to reach the threshold level of
detection. This threshold level is the result of BrdU incorporation taking place during priming events as well as first
elongation events. Thus, if a cyclin/Cdk complex, the
activation of which would be independent of protein synthesis, is involved in the precise regulation of DNA replication, then it might be required during one of these two
steps of DNA synthesis. To further investigate whether the
initiation or elongation step of DNA replication is modulated by a DMAP treatment, we monitored the extent of
DNA synthesis by measuring [3H]thymidine incorporation.
Because DMAP dramatically slowed down [3H]thymidine
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
192
Moreau et al.
FIG. 8. (A) Inhibition of Cdks kinase activities by olomoucine and DMAP does not preclude DNA replication. Olomoucine (0.1, 0.3, 0.5
mM) or DMAP (2 mM) was added with BrdU at the time of fertilization. The percentage of nuclei stained with anti-BrdU antibodies was
plotted as a function of time (1). The amount of fluorescence in individual nuclei was quantitated at the plateau of BrdU incorporation in
control, DMAP-treated (0.2 mM), or olomoucine-treated (0.3 mM) eggs (mean values from 10 representative eggs (2) .(B) Inhibiting kinase
activities by DMAP does not inhibit DNA synthesis but slows down the rate of [3H]thymidine DNA incorporation. The batches of eggs
were supplied with [3H]thymidine 20 min before fertilization and DMAP (2 mM) was added 10 min postfertilization. Aliquots of embryos
were taken at the indicated time (min); uptake and DNA incorporation were measured as described under Material and Methods. [3H]
Thymidine incorporation into DNA was corrected for uptake in treated eggs; the reported values were expressed considering an identical
[3H]thymidine concentration in control and treated eggs. At 180 min postfertilization control embryos have already reached the second S
phase (not shown). At that time the amount of [3H]thymidine incorporated in presence of aphidicolin (20 mg/ml) was measured and taken
as reference to evaluate significant incorporation into DNA.
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
193
Cdk2 and DNA Replication in Sea Urchin Early Embryogenesis
TABLE 2
Control
embryos
Drug added
in assay
Suc 1 beads
Cdk2 IP
Olomoucine
(0.1 mM)
DMAP (2mM)
DMAP (5 mM)
Emetine (0.1 mM)
1 DMAP (5 mM)
2
2
1
2
1
2
1
1
24.60 6 0.45
0.65 6 0.02
10.21 6 0.86
nd
4.23 6 1.29
0
7.68 6 0.47
nd
0.62 6 0.12
0
6.24 6 1.24
nd
0.23 6 1.05
nd
0.18 6 0.04
nd
Note. H1 kinase activities (pmol 32P incorporated in H1) were measured on proteins of sea urchin embryos (50 ml cell volume) either after
purification on p13-suc1 Sepharose beads (row 3) or immunoprecipitation with anti-Cdk2 antibodies (row 4). The values reported in row
4 are the differences between Cdk2 IP and preimmune IgG IP. Drugs were applied 10 min postfertilization, embryos were harvested 60 min
later and drugs were added or not (2), as mentioned in row 2, during protein purification and in vitro kinase assay. The reported values in
the absence of drugs in the kinase assay correspond to the in vivo inhibition of Cdk complexes activation, whereas the addition of drugs
to the assay express the cumulative effects of the competitive inhibition of Cdk complexes and the inhibition of their in vivo activation.
uptake in embryos, it was necessary to simultaneously
quantify [3H]thymidine incorporation into DNA and
[3H]thymidine uptake in the embryos. The corrected values
were registered as a function of time (Fig. 8c). DNA incorporation of [3H]thymidine was detected above background
at the same time in control and treated embryos, which
suggests that initiation is not dependent on a DMAPsensitive mechanism. However, completion of DNA replication was delayed, showing that a later event is affected by
this concentration of DMAP.
When added after NEBD, DMAP allowed a second round
of DNA replication (Fig. 9a) while inhibiting the ongoing
mitotic events. Although DMAP affected nuclear division
and the formation of the cleavage furrow, BrdU incorporation was visualized in the nuclear structures that re-formed
around chromatin. This confirmed the results of Néant and
Dubé (1996) in another sea urchin species, Strongylocentrotus droebachiensis, and demonstrated that this second
round of DNA replication is independent of a DMAPsensitive kinase, showing that Cdk activities are also not
essential for DNA replication in this case. The above
results argue against an essential role of the Cdk activities
in promoting the initiation of DNA replication during sea
urchin early embryogenesis.
In S. pombe the inhibition of Cdk activity in cdc 13
deletion mutants (Hayles, 1994) or in some temperature
sensitive mutants of p34cdc2 (Broek et al., 1991) leads to
extra rounds of DNA replication. As shown in Fig. 8B, Cdk
inhibition by DMAP is not sufficient in sea urchin to
promote more than one round of DNA replication. Similarly, DMAP applied in G2 did not allow further DNA
replication (Fig 9C). The above data suggested that other
factors are necessary in addition to the inhibition of Cdk
activities to regenerate the competence for re-replication. In
Xenopus egg extract/Xenopus sperm nuclei system, permeabilization of the nuclear membrane plays an essential role
in allowing a second round of DNA replication. Indeed,
microinjection of CHAPS in the G2-arrested DMAP-treated
embryos induced DNA incorporation of BrdU after a lag
period (Fig. 9D).
DISCUSSION
A number of studies have shown that activation of DNA
replication at S phase is controlled by Cdk2 kinase activity
in higher eukaryotic cells. This function has been mainly
demonstrated in Xenopus egg extracts (Fang and Newport,
1991, 1993; Strausfeld et al., 1994; Chevalier et al., 1995;
FIG. 9. A second round of DNA replication takes place in the
presence of DMAP if the drug is added after NEBD or if the nuclear
membrane of G2-treated embryos is permeabilized by CHAPS. (A)
and (B) DMAP (2 mM) and BrdU (0.1 mg/ml) were added at time of
NEBD (70 min) in the absence (A) or presence (B) of aphidicolin (20
mg/ml) and fixed 50 min later. The incorporated BrdU was visualized with fluorescently labeled anti-BrdU antibodies. (C) and (D)
The 50 min postfertilization embryos were supplied with DMAP (2
mM) and BrdU (0.1 mg/ml). Ten embryos were microinjected with
1% CHAPS and maintained in the continuous presence of DMAP.
BrdU incorporation was analyzed in microinjected (D) and nonmicroinjected (C) embryos were fixed 50 min later (bar, 50 mm).
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
194
Moreau et al.
Jackson et al., 1995; Yan and Newport, 1995). But how
precisely Cdk2 kinase contributes to the activation of
replication at the molecular level is currently unknown.
Because sea urchin early embryogenesis provides a good
opportunity to study, in vivo, the mechanisms regulating
DNA replication, we analyzed Cdk2 localization and activity during the first mitotic cycle of these embryos. Although the protein is present in the nucleus during S phase
and a measurable activity was found in whole cell extracts,
the Cdk2 kinase activity could be inhibited without precluding the onset of DNA replication. We will discuss a
potential role of the kinase activity in later step of DNA
replication as well as a possible function of the protein
independent of its kinase activity.
Cdk2 Is Present in the Nucleus during Replication
For the first time, Cdk2 was visualized in vivo in the
early embryonic cell cycle. In the first minutes after fertilization the Cdk2 antibodies intensely stained the male
pronucleus and displayed a faint and diffuse labeling of the
cytoplasm and the female pronucleus. The nuclear Cdk2
staining became homogenously distributed on the chromatin of the zygotic nucleus at the time when BrdU incorporation become detectable into DNA (about 30 min postfertilization). This pattern remains unchanged and coincident
with cyclin A labeling during most of S phase. The homogenous distribution of cdk2 and cyclin A contrasts with the
punctuated pattern characteristic of S phase in myotubes in
culture reentering cell cycle (Cardoso et al., 1993). However, the punctuated staining, corresponding to the replication foci, is not a universal feature, even in somatic cells,
because Cdk2 presents a diffuse nuclear distribution in
other cell lines (Brénot-Bosc et al., 1995; Baptist et al.,
1996). This absence of foci could also be explained if this
transient event is too rapid to be observed in sea urchin
embryos or if the organization of replication in these
embryos does not involve the aggregation of numerous
origins in replicating domains. Cdk2 and cyclin A acquire a
distinct distribution at the end of replication. At that time
Cdk2 immunoreactivity assembled into bright dots at the
periphery of the nucleus before being released in the cytoplasm when chromatin began to condense while most of
the cyclin A remains distributed on the whole mitotic
apparatus. Because only a subset of cyclin A is associated
with Cdk2, this suggests a distinct localization for the
cyclin A/Cdk1 complexes during G2 and M phases.
Cdk2 reassociates to chromatin in prometaphase but it
becomes again clearly colocalized with cyclin A on chromatin in telophase which corresponds in Arbacia punctulata to the formation of the karyomers (Longo, 1972) in
which [3H]thymidine incorporation has been observed by
electronic microscopy (Ito et al., 1981).
The presence of Cdk2 in the nucleus during the first and
second S phase of the mitotic cycles is consistent with a
potential role of the kinase during S phase.
Cdk2 Activity Is Dispensable for the Onset
of DNA Replication
Due to the experimental limitations, a direct involvement of Cdk2 kinase in DNA replication has been rarely
investigated in vivo. In early embryonic cycles in particular,
in vivo data are missing to support the conclusion drawn
from experiments in Xenopus egg extracts of a Cdk2 activity requirement for the onset of DNA replication (Fang and
Newport, 1991, 1993; Strausfeld, 1994; Chevalier et al.,
1995; Yan and Newport, 1995). In contrast, our results
strongly suggest that during the first sea urchin mitotic cell
cycles the Cdk2 kinase activity is not essential for DNA
replication in vivo. Several lines of evidence support this
view: microinjection of the dominant-negative Cdk2-K33R
or the inhibitor p21Cip1 did not preclude DNA replication;
moreover, inhibiting Cdk2 kinase activity by treatment
with olomoucine or DMAP before the onset of DNA
replication does not impede BrdU or [3H]thymidine incorporation into DNA; furthermore, endoreduplication could
be induced in the presence of Cdk inhibitors (DMAP), by
permeabilization of the nuclear membrane with mild detergent in the G2-arrested embryos; while all the above treatments severely affected mitotic features, showing that MPF
activity was highly reduced.
These observations are reminiscent of recently published
data showing that the inhibition of Cdk2 activity by a
truncated form of Xic1, highly specific for cyclin E/Cdk2
(Su et al., 1995), does not significantly inhibit DNA synthesis in newly fertilized Xenopus eggs (Hartley et al., 1997).
Similarly, a butyrolactone treatment does not lead to a
significant reduction of cell growth in the Gc4-PF cells, an
established cell line derived from a murine fibrosarcoma
(Dobashi et al., 1997). These data strongly suggest that
Cdk2 kinase activity could be dispensable for DNA replication onset in vivo in some rapidly dividing cells such as
the above-mentioned proliferating cells or during the early
embryogenesis.
In Xenopus egg extracts, the S-phase-promoting function
of Cdk2 is well documented. Since Cdk2 or cyclin E
depletion as well as inhibition by p21Cip1 does not affect the
formation of preinitiation centers but blocks the appearance of replication foci (Jackson et al., 1995; Yan and
Newport, 1995), the Cdk2-sensitive events have been proposed to be the DNA unwinding, the priming activity, or
another event at the transition between initiation and
elongation. Nevertheless, the time at which the S-phasepromoting activity of Cdk2 is required remained unknown.
In the model deduced from genetic experiments in Saccharomyces cerevisiae the requirement for clb/Cdk1 activity
occurred close to the initiation events (Schwob et al., 1994;
Dahmann et al., 1995; Piatti et al., 1996). This contrasts
with Xenopus extracts in which the sensitive period respectively to suc1-affinity depletion and p21Cip1 inhibition
terminated before the completion of nuclear assembly
(Blow and Nurse, 1990; Jackson et al., 1995). It is possible
that the absence of nuclear disassembly during mitosis in
yeast could oblige a strict timing pattern of events that
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
Cdk2 and DNA Replication in Sea Urchin Early Embryogenesis
could have evolved in metazoans into a more flexible
mechanism involving, for example, the modification of
spatial distribution. This flexibility of the mechanisms
controlling DNA replication could also give rise to situations, such as that observed in sea urchin early embryos,
where the S-phase-promoting function of Cdk2 kinase activity becomes a nonessential activity, provided that the
ordering of events can be controlled by the assembly of a
nuclear membrane. This hypothesis is further supported by
the strict requirement of membrane permeabilization for
DNA reinitiation that we observed in DMAP-treated embryos.
Of course we can also not exclude the possibility that the
strict Cdk2 requirement of Xenopus extracts would be a
consequence of the artificial release from a metaphasic
arrest by a calcium spike of CSF extracts or metaphasearrested eggs.
The present data demonstrate that Cdk2 is dispensable
for the onset of DNA replication but also that no single Cdk
activity is required to replace Cdk2 in promoting DNA
replication. It was previously shown by adding emetine to
fertilized sea urchin eggs that the newly formed cyclin A or
cyclin B complexes are not involved in DNA replication
(Wagenaar et al., 1983). The persistence of replication in
presence of DMAP added together with emetine suggests
that the cyclin B complexes preexisting in unfertilized eggs
are also not required for the onset of DNA replication.
However, these data do not exclude the involvement of a
Cdk in one step of DNA replication. Thus the delay
observed to reach the completion of DNA replication could
reflect the role of a Cdk in precisely regulating later step of
replication. Such a role has been already proposed for cyclin
A-dependent kinase activity (Jackson et al., 1995). It should
be noted that emetine treatment does not slow down BrdU
(personal data) or [3H]thymidine incorporation (Wagenaar,
1983); thus, only a cyclin/Cdk complex, the formation of
which would be independent of protein synthesis, could be
involved in the precise regulation of the elongation of DNA.
This could be cyclin E/Cdk2 because it has been shown in
Xenopus extract that cyclin E abundance is not dependent
of protein synthesis and that the associated kinase is
constitutively active during cell cycle (Fang and Newport,
1991). However, because high concentrations of DMAP and
to a lesser extent of olomoucine are inhibitors of kinases
other than Cdks (Labhart, 1995; Vesely et al., 1994), another
possibility would be that the kinase affecting the time
course of BrdU or [3H]thymidine incorporation would not
be a Cdk kinase.
Moreover, recent experiments in Xenopus egg extract
implicate Cdk2 activity in the restriction of DNA replication to one round per cell cycle (Hua et al., 1997), Cdk2
controlling the potentiation of chromatin for replication.
The presence of Cdk2 in the nucleus during S phase of the
first sea urchin mitotic cycles could be consistent with
such a regulation. Because we demonstrated that Cdk2 left
the nucleus at the end of S phase and went back rapidly to
the chromatin after NEBD, this would suggest that the
195
release of Cdk2 kinase activity from the nucleus during this
temporal window will allow the later formation of preinitiation complex. However, the accumulation of Cdks
within the nucleus cannot constitute the only mechanism
regulating the potentiation of chromatin for replication,
because Cdk inhibition by DMAP treatment, in G2 S.
granularis embryos or in G2 mammalian cells (Coverley et
al., 1996), did not induce DNA reinitiation. Re-replication
of G2 chromatin in somatic or embryonic cells (Blow and
Laskey, 1988; Leno et al., 1992; present data) is controlled
by a mechanism involving the breakdown of the nuclear
envelope. In the particular case of somatic cell chromatin,
this requirement can be bypassed if the intact nuclei are
isolated from G2 DMAP-treated cells and subsequently
transferred to interphasic Xenopus extracts (Coverley et al.,
1996, 1998). In contrast, nuclear envelope permeabilization
seems to be strictly required during early embryogenesis as
sea urchin DMAP-treated G2 embryos which are able to
support DNA reinitiation, only do it after nuclear membrane permeabilization. It is noteworthy that this reinitiation in G2 nuclei is not dependent on a DMAP-sensitive
kinase as is the case in Xenopus extracts (Blow, 1993). This
suggests that regeneration of DNA replication competence
and capacity to initiate DNA replication would be dependent on (a) cellular factor(s) which access to chromatin and
efficiency to promote initiation would be differentially
controlled among species or according to development.
Could Cdk2 Have a Function Independent
of Its Kinase Activity?
Our data demonstrate that Cdk2 kinase can be inactivated without precluding DNA replication. although the
protein is localized on chromatin from prometaphase to the
end of S phase. Moreover, depletion of Cdk2 from Xenopus
extracts using specific antibodies or p13-suc1 beads (Blow
and Nurse, 1990) led to inhibition of DNA replication.
Thus, it is tempting to think that the Cdk2 protein itself,
rather than its kinase activity, could potentiate DNA replication. In such a case, addition of an inactive kinase
Cdk2-K33R would not modify this function and p21Cip1
would, or not, affect the physical interaction of Cdk2 with
its unknown partners, depending on the conditions or
species. Nevertheless, the release of Cdk2 from the chromatin could remain necessary to the regeneration of DNA
competence to replication explaining the very specific cellcycle-dependent localization we observed.
ACKNOWLEDGMENTS
We thank Dr. B. Ducommun for providing us the pGEX-p21Cip1
plasmid and Dr. T. Millward for critical reading of the manuscript.
This work was supported by grants from the Association pour la
Recherche sur le Cancer and from the Ligue Nationale contre le
cancer. F. Marqués was recipient of a Fellowship from the Ligues
Regionales contre le cancer of Corrèze and Pyrennées orientales.
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
196
Moreau et al.
REFERENCES
Abraham, R. T., Acquarone, M., Andersen, A., Asensi, A., Belle, R.,
Berger, F., Bergounioux, C., Brunn, G., Buquet-Fagot, C., Fagot,
D., Glab, N., Goudeau, H., Goudeau, M., Guerrier, P., Houghton,
P., Hendriks, H., Kloareg, B., Lippai, M., Marie, D., Maro, B.,
Meijer, L., Mester, J., Mulner-Lorillon, O., Poulet, S. A., Schierenberg, E., Schutte, B., Vaulot, D., and Verlhac M. H. (1995).
Cellular effects of olomoucine, an inhibitor of cyclin-dependent
kinases. Biol. Cell 83, 105–120.
Baptist, M., Lamy, F., Gannon, J., Hunt, T., Dumont, J. E., and
Roger, P. P (1996). Expression and subcellular localisation of
Cdk2 and Cdc2 kinases and their common partner cyclin A in
thyroid epithelial cells: Comparison of cyclic AMP-dependent
and -independent cell cycles. J. Cell. Physiol. 166, 256 –273.
Blow, J. J. and Laskey, R. A. (1988). A role for the nuclear envelope
in controlling DNA replication within the cell cycle. Nature 332,
546 –548.
Blow J. J., and Nurse P. (1990). A cdc2-like protein is involved in the
initiation of DNA replication in Xenopus egg extracts. Cell 62,
855– 862.
Blow J. J. (1993). Preventing re-replication of DNA in a single cell
cycle: Evidence for a replication licensing factor. J. Cell Biol. 122,
993–1002.
Broek, D., Bartlett, R., Crawford, K., and Nurse, P. (1991). Involvement of p34cdc2 in establishing the dependency of S-phase on
mitosis. Nature 349, 388 –393.
Brenot-Bosc, F., Gupta, S., Margolis, R. L., and Fotedar, R. (1995).
Changes in the subcellular localization of replication initiation
proteins and cell cycle proteins during G1- to S-phase transition
in mammalian cells. Chromosoma 103, 517–527
Cardoso, M. C., Leonhardt, H., and Nadal-Girard, B. (1993). Reversal of terminal differentiation and control of DNA replication:
Cyclin A and Cdk2 specifically localize at subnuclear sites of
DNA replication. Cell 74, 979 –992
Chevalier, S., Tassan, J. P., Cox, R., Philippe, M., and Ford C. (1995).
Both cdc2 and Cdk2 promote S phase initiation in Xenopus egg
extracts. J. Cell Sci. 108, 1831–1841.
Coverley, D., Hannah, R., Wilkinson, R., and Downes, S. C. (1996).
A protein kinase-dependent block to reinitiation of DNA replication in G2 phase in mammalian cells. Exp. Cell Res. 225,
294 –300.
Coverley D., Wilkinson H. R., Madine M. A., Mills A. D., and
Laskey R. A. (1998). Protein kinase inhibition in G2 causes
mammalian Mcm proteins to reassociate with chromatin and
restores ability to replicate. Exp. Cell Res. 238, 63– 69.
Dahmann, C., Diffley, J. F. X., and Nasmith, K. (1995). S-phase
promoting cyclin-dependent kinases prevent re-replication by
inhibiting the transition of replication origins to a pre-replicative
state. Curr. Biol. 5, 1257–1269.
Dobashi, Y., Chigira, M., Shoji, M., Wakata, Y., Kitagawa, M.,
Tmauchi, H., Akiyama, T., and Kameya, T. (1997). Cell cycle
control with minimal participation of Cdk2 in murine fibroblaste clone cultured in protein free medium. Biochem. Biophys.
Res. Commun. 232, 622– 625.
Dorée, M., and Galas, S. (1994). The cyclin-dependent protein
kinases and the control of cell division. FASEB J. 8, 114 –121.
Dulic, V., Kaufmann, W. K., Wilson, S. J., Tisty, T. D., Lees, E.,
Harper, J. W., Elledge, S., and Reeed, S. I. (1994). p53-dependent
inhibition of cyclin-dependent kinase activities in human fibroblasts during radiation-induced G1 arrest. Cell 76, 1013–1023.
Fang, F., and Newport, J. (1991). Evidence that the G1–S and G2–M
transitions are controlled by different cdc2 proteins in higher
eukaryotes. Cell 66, 731–742.
Fang, F., and Newport, J. (1993). Distinct roles of Cdk2 and cdc2 in
RP-A phosphorylation during the cell cycle. J. Cell Sci. 106,
983–994.
Geneviere-Garrigues, A. M., Barakat, H., Dorée, M., Moreau, J. L.,
and Picard, A. (1995). Active cyclin B-cdc2 kinase does not
inhibit DNA replication and cannot drive prematurely fertilized
sea urchin eggs into mitosis. J. Cell Sci. 108, 2693–2703.
Girard, F., Strausfeld, U., Fernandez, A., and Lamb, N. J. (1991).
Cyclin A is required for the onset of DNA replication in
mammalian fibroblasts. Cell 67, 1169 –1179.
Gonnet, G. M., Cohen, M. A., and Benner, S. A. (1992). Exhaustive
matching of the entire sequence data base. Science 256, 1443–
1445.
Gu, Y., Turck, C. W., and Morgan, D. O. (1993). Inhibition of Cdk2
activity in vivo by an associated 20k regulatory subunit. Nature
366, 707–710.
Harland, R. M., and Laskey, R A. (1980). Regulated DNA replication of DNA microinjected into eggs of Xenopus laevis. Cell 21,
761–771.
Harper, J. W., Adami, G. R., Wei, N., Keyomarsi, K., and Elledge,
S. J. (1993). The p21 Cdk-interacting protein CIP1 is a potent
inhibitor of G1 cyclin-dependent kinases. Cell 75, 806 – 816.
Hartley, R. S., Sible, J. C., Lewellyn, A. L., and Maller, J. L. (1997).
A role for cyclin E/Cdk2 in the timing of the midblastula
transition in Xenopus embryos. Dev. Biol. 188, 312–321.
Hayles, J., Fisher, D., Woollard, A., and Nurse, P. (1994). Temporal
order of S phase and mitosis in fission yeast is determined by the
state of the p34cdc2-mitotic B cyclin complex. Cell 78, 813– 822.
Hua, X. H., Yan, H., and Newport, J. (1997). A role for Cdk2 kinase
in negatively regulating DNA replication during S phase of the
cell cycle. J. Cell Biol. 137, 183–192.
Ito, S., Dan, K., and Goodenough, D. (1981). Ultrastructure and
3
H-thymidine incorporation by chromosome vesicles in sea
urchin embryos. Chromosoma 83, 441– 453.
Jackson, P. K, Chevalier, S., Philippe, M., and Kirschner, M. W.
(1995). Early events in DNA replication require cyclin E and are
blocked by p21CIP1. J. Cell Biol. 130, 755–769.
Kajiura, H., Yamashita, M., Katsu, Y., and Nagahama, Y. (1993).
Isolation and characterisation of goldfish cdc2, a catalytic component of maturation-promoting factor. Dev. Growth Differ. 35,
647– 654.
Labhart, P. (1995). DNA-dependent protein kinase specifically
represses promoter-directed transcription initiation by RNA
polymerase I. Proc. Natl. Acad. Sci. USA 92, 2934 –2938.
Lehner, F. C., and O’Farrell, H. P. (1989). Expression and function of
Drosophila cyclin A during embryonic cell cycle progression.
Cell 56, 957–968.
Lehner, F. C., and O’Farrell, H. P. (1990). Drosophila cdc2 homologs: A functional homolog is coexpressed with a cognate
variant. EMBO J. 9, 3573–3581.
Leno, G. H., Downes, C. S., and Laskey, R. A. (1992). The nuclear
membrane prevents replication of human G2 nuclei but not G1
nuclei in Xenopus egg extract. Cell 69, 151–158.
Longo, F. J. (1972). An ultrastructural analysis of mitosis and
cytokinesis in the zygote of the sea urchin Arbacia punctulata. J.
Morphol. 138, 207–238.
Lozano J. C, Schatt P., Marques F., Peaucellier G., Fort P., Feral J. P.,
Geneviere A. M., and Picard, A. (1998) A presumptive developmental role for a sea urchin cyclin B splice variant. J. Cell Biol.
140, 283–293.
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.
197
Cdk2 and DNA Replication in Sea Urchin Early Embryogenesis
Meyerson, M., Enders, G. , Wu, C., Su, L., Gorka, C., Nelson, C.,
Harlow, E., and Tsai, L. (1992). A family of human cdc2-related
protein kinases. EMBO J. 11, 2909.
Nasmyth, K. (1996). At the heart of the budding yeast cell cycle.
Trends Genet. 12, 405– 412.
Neant, I., and Dubé, F. (1996). DNA replication initiation
by 6-DMAP treatment in maturing oocytes and dividing
embryos from marine invertebrates. Mol. Reprod. Dev. 44,
443– 451.
Nigg, E. A. (1995). Cyclin-dependent protein kinases: key regulators cf the eukaryotic cell cycle. Bioessay 17, 471– 480.
Pagano, M., Pepperkok, R., Verde, F., Ansorge, W., and Draetta, G.
(1992). Cyclin A is required at two points in the human cell
cycle. EMBO J. 11, 961–971.
Pagano, M., Pepperkok, R., Lukas, J., Baldin, V., Ansorge, W.,
Bartek, J., and Draetta, G. (1993). Regulation of the cell cycle by
the Cdk2 protein kinase in cultured human fibroblasts. J. Cell
Biol. 121, 101–111.
Piatti, S., Bohm, T., Cocker, J., Diffley, J. F. X., and Nasmith, K.
(1996). Activation of S-phase promoting CDKs in late G1
define a ‘‘point of no return’’ after which Cdc6 synthesis
cannot promote DNA replication in yeast. Genes Dev. 10,
1516 –1531.
Picard, A., Harricane , M. C., Labbé, J. C., and Dorée, M. (1988).
Germinal vesicle components are not required for the cell
cycle oscillator of the early starfish embryo. Dev. Biol. 128,
121–128.
Picard, A., Galas , S., and Dorée, M. (1996). Newly assembled cyclin
B-cdc2 kinase is required to suppress DNA replication between
meiosis I and meiosis II in starfish oocytes. EMBO J. 15, 3590 –
3598.
Reed, S. I. (1992). The role of p34 kinases in the G1 to S-phase
transition. Annu. Rev. Cell Biol. 8, 529 –561
Sherr, C. J. (1993). Mammalian G1 cyclins. Cell 73, 1059 –1065.
Sherr, C. J. (1996). Cancer cell cycles. Science 274, 1672–1677.
Schwob, E., Bohm, T., Mendenhall, M. D., and Nasmyth, K. (1994).
The B-type cyclin inhibitor p40sic1 controls the G1 to S transition
in Saccharomyces cerevisiae. Cell 79, 233–244.
Showman, R. M., and Foerder, C. A. (1979). Removal of the
fertilization membrane of sea urchin embryos employing aminotriazole. Exp. Cell Res. 2, 253–255.
Stern, B., and Nurse, P. (1996). A quantitative model for the cdc2
control of S phase and mitosis in fission yeast. Trends Genet. 12,
345–350.
Strausfeld. U., Howell. M., Rempel, R., Maller, J. L., Hunt, T., and
Blow, J. (1994). Cip1 blocks the initiation of DNA replication in
Xenopus extracts by inhibition of cyclin-dependent kinases.
Curr. Biol. 4, 876 – 883.
Strausfeld, U. P., Howell. M., Descombes, P., Chevalier, S.,
Rempel, R. E., Adamczewski, J., Maller, J. L., Hunt, T., and
Blow, J. J. (1996). Both cyclin A and cyclin E have S-phase
promoting (SPF) activity in Xenopus egg extracts. J. Cell Sci.
109, 1555–1563.
Su, J-Y., Rempel, R. E., Erikson, E., and Maller, J. L. (1995).
Cloning and characterisation of the Xenopus cyclin-dependent
kinase inhibitor p27-Xic1. Proc. Natl. Acad. Sci. USA 92,
10187–10191.
van der Heuvel., S., and Harlow, E. (1993). Distinct roles for
cyclin-dependent kinases in cell cycle control. Science 262,
2050 –2054.
Vesely, J., Havlicek , L., Strnad, M., Blow, J. J., Donella-Deana, A.,
Pinna, L., Letham, D. S., Kato, J-Y., Detivaud, L., Leclerc, S., and
Meijer, L. (1994). Inhibition of cyclin-dependent kinases by
purine analogues. Eur. J. Biochem. 224, 771–786.
Wagenaar, E. (1983). The timing of synthesis of proteins Required
for mitosis in the cell cycle of the sea urchin embryo. Exp. Cell
Res. 144, 393– 403.
Xiong, Y., Hannon, G. J., Zhang, H., Casso, D., Kobayashi, R., and
Beach, D. (1993). p21 is a universal inhibitor of cyclin kinases.
Nature 366, 701–704.
Yan, H., and Newport, J. (1995). An analysis of the regulation of
DNA synthesis by Cdk2, Cip1, and licensing factor. J. Cell Biol.
129, 1–15.
Zindy, F. Lamas, E., Chenivesse, X., Sobczak, J., Wang, J., Fesquet,
D., Henglein, B., and Bréchot, C. (1992). Cyclin A is required in
S phase in normal epithelial cells. Biochem. Biophys. Res.
Commun. 182, 1144 –115.
Received for publication April 9, 1998
Revised May 5, 1998
Accepted May 5, 1998
Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.