Oncogene (2005) 24, 2844–2859
& 2005 Nature Publishing Group All rights reserved 0950-9232/05 $30.00
www.nature.com/onc
Getting in and out of mitosis with Polo-like kinase-1
Marcel ATM van Vugt1,2 and René H Medema*,1,2
1
Division of Molecular Biology H8, The Netherlands Cancer Institute, 1066 CX Amsterdam, The Netherlands
Research in different species has shown that Polo-like
kinases are essential for successful cell division. In human
cells, Polo-like kinase-1 (Plk1) has been implicated in the
regulation of different processes, including mitotic entry,
spindle formation and cytokinesis. Recently, a range of
new downstream targets of Plk1 has been identified, as
well as a molecular mechanism that explains recruitment
of Plk1 to potential substrate proteins through its polobox domain. On the basis of these reports, we discuss
possible mechanisms by which Polo-like kinases can exert
their multiple functions during mitosis. Polo-like kinases
also function in DNA damage checkpoints. Plk1 has been
shown to be a target of the G2 DNA damage checkpoint,
while Cdc5, the Polo-like kinase in Saccharomyces
cerevisiae, has long been known to be required for
adaptation to persistent DNA damage. Just recently, a
similar requirement for Polo-like kinases during checkpoint adaptation was demonstrated in multicellular
organisms. Moreover, Plk1 was also shown to be required
for checkpoint recovery following checkpoint inactivation,
that is, in cells where the damage is completely repaired.
Thus, Plk1 appears to play a role at multiple points during
a restart of the cell cycle following DNA damage. Based
on these novel observations, we discuss possible consequences of using Plk1 as a target in anticancer
strategies.
Oncogene (2005) 24, 2844–2859. doi:10.1038/sj.onc.1208617
Keywords: spindle; checkpoint; DNA damage; Cyclin;
cytokinesis; APC/cyclosome
Introduction
Accurate distribution of a duplicated genome over two
daughter cells requires condensed chromosomes, a
perfect spindle, and a mechanism that ensures correct
attachment of these chromosomes to the spindle prior to
cell division. Here we discuss the involvement of Pololike kinase-1 (Plk1) in the broad range of processes from
start to completion of mitosis. In the family of Polo-like
kinases, Plk1 is the human homolog of the founding
member Polo in Drosophila. The original identification
*Correspondence: RH Medema; E-mail: [email protected]
2
Current address: Medical Oncology, University Medical Center,
Stratenum 2.103 Universiteitsweg 100, 3584 CG Utrecht,
The Netherlands
of Polo (Drosophila), Cdc5 (S. cerevisiae) and Plo1 (S.
pombe) by mutational analysis, revealed an essential role
for Polo-like kinases (Plk’s) in mitotic progression. In
human cells, four Polo-like kinases have been identified,
based on the presence of one or more polo-boxes (Plk1,
SNK or Plk2, Fnk/Prk or Plk3 and Sak or Plk4, recently
reviewed in Barr et al. (2004). In human cells, Plk1
appears to be responsible for the main portion of Poloand Cdc5-related functions, and therefore this review
mainly covers recent advances made in Plk1-related
research.
Plk’s are serine/threonine kinases that typically exist
of an N-terminal kinase domain and harbor one (Plk4)
or two (Plk1-3) polo box domains. Plk1 is expressed
from G2 onwards and is degraded at the end of mitosis
(Uchiumi et al., 1997; Fang et al., 1998; Lindon and
Pines, 2004). Although Plk1 is already present in G2, its
kinase activity is first seen at the G2/M transition and
reaches peak levels during mitosis (Golsteyn et al.,
1995). Little is known about the upstream regulation of
Plk1. However, by mutational analysis, two putative
phosphorylation sites (Ser128 and Thr201) were identified in the Xenopus homolog Plx1 (Qian et al., 1999).
Subsequently, xPlkk1, an upstream kinase for Plx1, was
cloned from Xenopus and shown to phosphorylate
Thr201 in vitro (Qian et al., 1998b, 1999). Although no
in vivo evidence exists for a human Plk1-kinase, two
xPlkk1-related kinases were shown to phosphorylate
Plk1 in vitro (Ellinger-Ziegelbauer et al., 2000; Walter
et al., 2003). Recently, mass spectrometric analysis of
Plx1 revealed additional phosphorylation sites, including two autophosphorylation sites (Kelm et al., 2002). In
addition to the identified Plkk’s, Cyclin B–Cdk1 was
also suggested to function as an activating kinase for
Plk1. Indeed, Cyclin B–Cdk1 can phosphorylate human
Plk1 and Xenopus Plx1 in vitro, but the identified Cdk1consensus motif in Plx1 (S304) is not conserved in
human Plk1. Thus, it is unclear if and how Cdk1 can
function as a bona fide upstream kinase for Plk1 in vivo
(Hamanaka et al., 1995; Kelm et al., 2002).
More recently, a model was proposed for Plk1
activation, in which the Polo-box domain rather than
an upstream kinase plays a central role (Elia et al.,
2003b). Binding of the polo box domain to a phosphorylated target results in conformational changes that
releases the inhibitory effect of the polo-box domain
towards the kinase domain. In this model, Plk1 is not
directly activated by upstream kinases, but is coupled to
the activity of other kinases, such as Cyclin B–Cdk1. In
doing so, kinases with specific temporal or spatial
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2845
activities, phosphorylate future Plk1 substrates that
subsequently recruit Plk1. In this way, Plk1 can amplify
signaling pathways or rather act as a switch that only
fires when multiple requirements are met. This of course
does not preclude a role for upstream kinases in
regulating Plk1 through additional phosphorylation
events.
In recent years, a range of molecular targets of Plk1
has been identified that revealed the signaling pathways
by which Plk1 can regulate mitotic entry, DNA damage
checkpoint responses, spindle formation and mitotic
exit. In the following sections, we discuss these targets
and their actions and discuss implications for tumorigenesis and cancer therapy.
nuclear, Cyclin B is localized predominantly in the
cytoplasm during interphase (Pines and Hunter, 1991).
In fact, Cyclin B is continuously shuttled in and out of
the nucleus in G2, but because nuclear export dominates
import at this stage, the bulk of the protein is
cytoplasmic (Hagting et al., 1998). To facilitate this
transport, Cyclin B harbors two domains that control
cytoplasmic-nuclear shuttling; a nuclear localization
signal (NLS) that targets Cyclin B–Cdk1 complexes to
Plk1 and mitotic entry
All cell cycle transitions, such as mitotic entry, are
driven by the activity of specific Cyclin-dependent
kinases (Cdk’s), bound to their cognate Cyclins. Cyclin
subunits are expressed in a cell cycle-dependent fashion
that promotes a forward (one-way) direction in cell cycle
progression. The transition from G2 to mitosis strictly
depends on activation of the MPF (M-phase-promoting
factor), which consists of Cdk1 in complex with Cyclin
B (reviewed in Morgan, 1995). In G2, enhanced
promoter activity as well as increased stability of Cyclin
B mRNA result in sufficiently high amounts of Cyclin B
protein (Pines and Hunter, 1989; Maity et al., 1995;
Piaggio et al., 1995). This newly synthesized Cyclin B
can now bind the inactive Cdk1, but does not result in
an active nuclear Cyclin B–Cdk1 complex yet, because
the complex is kept inactive by different signaling
pathways (Figure 1, upper panel).
Firstly, Cdk1 is phosphorylated by Wee1 and Myt1 in
its ATP-binding site on Tyr15 (by Wee1 and Myt1) and
Thr14 (by Myt1) kinases (Atherton-Fessler et al., 1993;
Booher et al., 1997). Phosphorylation on these sites
prevents ATP-binding of Cdk1 and thereby keeps the
Cyclin–Cdk1 complex catalytically inactive. Secondly,
Cyclin B–Cdk1 complexes are regulated by subcellular
localization. Unlike Cyclin A, which seems to be always
Figure 1 Mitotic entry and the G2 DNA damage checkpoint. In
G2, premature mitotic entry is prevented through inhibitory
phosphorylation on Cdk1 as well as nuclear export of Cyclin B
(upper panel). Plk1 is implicated in several processes that
contribute to the activation of Cyclin B–Cdk1 (middle panel).
Plk1 directly phosphorylates Wee1, Myt1, Cyclin B and Cdc25C.
Phosphorylation of Wee1 results in its degradation. In contrast,
phosphorylation of Cyclin B and Cdc25C, promotes Cyclin B–
Cdk1 activation at the centrosome, and likely involves feedbackloop amplification. Whether Plk1 also signals inhibition towards
Myt1 is unclear. In response to DNA damage, the G2 DNA
damage checkpoint prevents mitotic entry (lower panel), Activation
of checkpoint kinases ATM, ATR, Chk1, Chk2 and p38 mediate
nuclear exclusion of Cdc25C and Cdc25B, degradation of Cdc25A,
activation of Wee1 and inhibition of Plk1. As a result, checkpoint
signaling results in sustained Y15-phosphorylation and an inactive
Cdk1-Cyclin B complex
Oncogene
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2846
the nucleus in an atypical Importin-b-dependent fashion
(Moore et al., 1999; Takizawa et al., 1999), and a
cytoplasmic retention signal (CRS), harboring an NES
(Nuclear Export Signal), which mediates Crm1-dependent nuclear export and ensures cytoplasmic localization
of Cyclin B–Cdk1 before mitotic onset (Pines and
Hunter, 1994; Hagting et al., 1998; Yang et al., 1998).
Entry into mitosis requires full Cyclin B–Cdk1
activity and depends on phosphorylation of Thr161
(located in the T-loop of Cdk1) and dephosphorylation
of Tyr15 and Thr14. The CAK complex (consisting of
Cdk7, Cyclin H and MAT1) was shown to be
responsible for phosphorylation of Thr161, resulting in
increased kinase activity of the Cyclin B–Cdk1 complex
(Fesquet et al., 1993; Poon et al., 1993; Solomon et al.,
1993; Fisher and Morgan, 1994). Dephosphorylation of
Tyr15 and Thr14 is brought about by all members of the
Cdc25 phosphatase family (Cdc25A, B, and C) (Sadhu
et al., 1990; Galaktionov and Beach, 1991; Nagata et al.,
1991). In the process of Cdk1-dephosphorylation at the
G2/M transition, Cdc25B is thought to act as a starterphosphatase that initially activates cytoplasmic Cyclin
B–Cdk1 (Gabrielli et al., 1996; Lammer et al., 1998).
Once activated, Cyclin B–Cdk1 effectively promotes its
own full and rapid activation in a dual positive
amplification loop; Cyclin B–Cdk1 activates Cdc25C
by phosphorylation to further promote Cdk1-dephosphorylation (Figure 1, middle panel) (Hoffmann et al.,
1993). In addition, Cyclin B–Cdk1 phosphorylates Wee1
to promote its degradation and in this way decreases
Cdk1-phosphorylation on Tyr15 (Watanabe et al.,
2004). Thus, the amplification loop that controls Cyclin
B–Cdk1 activity ensures mitotic commitment once
mitotic entry has started. Mitotic entry is also characterized by a rapid accumulation of Cyclin B–Cdk1 in
the nucleus. Preventing nuclear import of Cyclin B
blocks mitotic entry, illustrating the importance for
nuclear localization of this complex (Li et al., 1997).
Moreover, expression of an active Cdk1 mutant that is
constitutively localized in the cytoplasm cannot accelerate mitotic entry (Heald et al., 1993), again indicating
that a collaboration between Cdk1 activation as well as
Cyclin retention in the nucleus controls mitotic entry.
This is further supported by the notion that expression
of a constitutively nuclear mutant of Cyclin B1 or a
Cdk1 mutant lacking the inhibitory Thr14/Tyr15
phosphorylation sites partially overrides a DNA damage-induced G2 arrest, while the combined expression
of both mutants results in a full checkpoint override (Jin
et al., 1998).
Polo-like kinases were suggested to regulate Cyclin B–
Cdk1 activity at several levels (Figure 1, middle panel).
The Xenopus Polo-like kinase Plx1 was shown to directly
phosphorylate and activate Cdc25C (Kumagai and
Dunphy, 1996), and actually be part of the MPF
amplification loop (Abrieu et al., 1998; Qian et al.,
1998a). Indeed, depletion of Plx1 prevented Cdc25C
hyperphosphorylation and delayed Cyclin B–Cdk1
activation (Qian et al., 1998a). Recently, elegant work
of Elia et al. (2003a, b) provided a molecular view of the
amplification loop in MPF by demonstrating that the
Oncogene
polo box of Plk1 is a phospho-binding domain that
determines binding to Cdc25C, once Cdc25C is primed
by Cdk-phosphorylation. Human Plk1 was also shown
to phosphorylate Cdc25C and thereby activate as well as
mediate nuclear translocation of Cdc25C (Smits et al.,
2000; Toyoshima-Morimoto et al., 2002). The contribution of this phosphorylation event to Cyclin B–Cdk1
activity, however, is doubtful, since Plk1-depleted cells
enter mitosis with active Cyclin B–Cdk1 but incompletely phosphorylated Cdc25C (van Vugt et al., 2004a).
Therefore, the requirement for Cdc25C during mitotic
entry may differ between species or between different
kinds of cell division cycles.
Plk1 was recently also shown to regulate Cdk1phosphorylation via another mechanism. Through
direct phosphorylation of Wee1, Plk1 was shown to
promote degradation of Wee1 and thereby decrease
Cdk1 phosphorylation (Watanabe et al., 2004). This
study showed that Wee1 is phosphorylated by Cyclin B–
Cdk1 as well as by Plk1. It could very well be that this
dual phosphorylation is kicked off by a Cdk1-phosphorylation that primes Wee1 for phosphorylation by
Plk1 to create a ‘phosho-degron’ that is recognized by
the b-TrCP-SCF (Watanabe et al., 2004). These
observations do suggest that binding of Plk1 to mitotic
regulators that are primed by Cdk-phosphorylation is a
common theme for Plk1-mediated actions (Elia et al.,
2003b). Whether Plk1 regulates another Cdk1 kinase,
Myt1, via a similar mechanism remains to be elucidated.
However, using a Plk1 consensus phosphorylation
motif, the Wee1-family member Myt1 was identified as
an in vitro Plk1 substrate. In addition, also Cyclin B–
Cdk1 is able to phosphorylate Myt1 in vitro, but
whether these phosphorylation events regulate in vivo
inactivation has to be determined (Booher et al., 1997;
Nakajima et al., 2003).
Besides regulating Cyclin B–Cdk1 activity, Plk1 was
also proposed to regulate the subcellular localization of
this complex. Initially, Plk1 was proposed to fulfill this
role through phosphorylation of Ser147, located in the
NES of Cyclin B (Toyoshima-Morimoto et al., 2001).
This phosphorylation was suggested to disrupt the NES
of Cyclin B and result in nuclear retention of Cyclin B–
Cdk1. Subsequently, Plk1 was shown to phosphorylate
Cyclin B just outside the NES at Ser133 (but not Ser147)
in order to activate Cyclin B–Cdk1 at the centrosome
and promote nuclear import in this way (Yuan et al.,
2002a; Jackman et al., 2003). Yet, the fact that deletion
of this phosphorylation site did not block nuclear
import of Cyclin B–Cdk1 indicates that multiple factors
control nuclear localization of the MPF.
It is clear that the activity of Cyclin B–Cdk1 is
precisely controlled by the balance between phosphatase
activity (Cdc25A, B and C) and kinase activity (Wee1
and Myt1). However, the initial trigger that causes
Cyclin B–Cdk1 activation and drives cells past a point of
no return to ensure mitotic commitment remains a big
mystery. It is clear that in an unperturbed cell cycle
Cyclin B–Cdk1 activation is not allowed until DNA
replication is completed, but the underlying mechanism
that explains the lag between the completion of S phase
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2847
and the onset of mitosis, known as G2 phase, is not
resolved. Possibly, silencing of the replication checkpoint needs a certain amount of time. In addition,
activation of Cdc25B, proposed to act as a starter
phosphatase, could play an important role in the timing
of mitotic entry. However, Cdc25B was shown to be
activated by Cyclin A–Cdk2 and Cyclin A–Cdk2
activity is present from S phase onwards (Gabrielli
et al., 1997). Moreover, deletion of the Cdc25B gene
does not block somatic cell division, so additional levels
of regulation must exist (Lincoln et al., 2002). Interestingly, more and more evidence is accumulating that the
(maturing) centrosome could play a key role in the
initial activation of Cyclin B–Cdk1. Immunocytochemically, a fraction of Cdk1 that is bound to Cyclin B was
shown to localize to centrosomes during interphase
(Bailly et al., 1989, 1992). Initially, this localization
pattern was suggested to regulate centrosome functioning. More recently, however, accumulation of Cyclin B–
Cdk1 complexes at centrosomes, appeared to coincide
with the initial activation of this complex (Jackman
et al., 2003). In addition, Chk1 and Chk2, two negative
regulators of Cyclin B–Cdk1, were also shown to be
localized to centrosomes (Tsvetkov et al., 2003; Kramer
et al., 2004). Very recently, Kramer et al. (2004) showed
that Chk1 is present at centrosomes to prevent
premature Cyclin B–Cdk1 activation. Moreover, Chk1
was shown to recruit Cdc25B to centrosomes, and at the
same time keep Cdk1 temporarily inactive. Taken
together, these data show that the centrosome may
function as a platform for timely Cyclin B–Cdk1
activation. Since Plk1 is present at centrosomes and is
responsible for Cyclin B phosphorylation at this site
(Golsteyn et al., 1995; Jackman et al., 2003), could Plk1
be responsible for initial activation of Cyclin B–Cdk1?
Injection of anti-Plk1 antibodies in primary human cells
or depletion of Plx1 from Xenopus extracts does result in
G2 arrest and argues in favor of this idea (Lane and
Nigg, 1996; Qian et al., 1998a, 2001). However, injection
of anti-Plk1 antibodies in transformed human cells,
expression of dominant-negative Plk1 fragments and
RNA interference of Plk1 in a range of human cells did
not result in G2 arrest but rather caused a mitotic arrest
with high Cyclin B–Cdk1 levels (Lane and Nigg, 1996;
Liu and Erikson, 2002; Seong et al., 2002; SpankuchSchmitt et al., 2002; van Vugt et al., 2004b). Likely,
mitotic entry requires multiple factors, and only in some
cell types or situations does Plk1 become rate limiting
for this process.
Plk1 and G2–M checkpoints
To allow cells time to repair damaged DNA, a complex
network called the ‘DNA damage checkpoint’ prevents
Cyclin–Cdk activation and arrests cell cycle progression.
Accordingly, when cells encounter DNA damage in G2,
the ‘G2 DNA damage checkpoint’ inhibits Cyclin B–
Cdk1 activity to prevent mitotic entry (Figure 1, lower
panel) (reviewed in Elledge, 1996). Different types of
DNA damage can activate the G2 DNA damage
checkpoint and induce a G2 arrest. However, it should
be noted that the G2 DNA damage checkpoint is
particularly useful for repair of double-strand breaks
because the duplicated genomic material offers cells the
possibility to use the duplicated chromatids as templates
for the high-fidelity homologous recombination (HR)
repair system (Haber, 2000). This opportunity is limited
in a temporal fashion, since sister chromatids are
separated during mitosis and as a result neighboring
templates for HR will be lost.
DNA damage results in the activation of checkpoint
kinases ATM and ATR, which subsequently signal to a
range of checkpoint mediators, such as p53, Chk1, Chk2
and Histone H2AX as well as components of DNA
repair mechanisms (reviewed in Shiloh, 2003). Subsequently, activation of this checkpoint pathway results in
inactivation of all Cdc25 family members. Conversely,
checkpoint activation enhances Wee1 function, and as a
net result the balance between positive and negative
regulators generates an Y15-phosphorylated, inactive
Cyclin B–Cdk1 complex. Also, Plk1 was shown to be a
direct target of the G2 DNA damage checkpoint
(Figure 1, lower panel) (Smits et al., 2000). In response
to a range of DNA-damaging agents, Plk1 was shown to
be catalytically inactivated. Moreover, this inhibition
was shown to depend on functional ATM or ATR (van
Vugt et al., 2001). The relevance of this inhibition was
illustrated by the fact that expression of constitutively
active Plk1-mutants can override the DNA damageinduced G2 arrest (Smits et al., 2000). In addition, since
Plk1 is inhibited in response to DNA damage, a picture
is emerging in which all proteins involved in Cyclin B–
Cdk1 regulation are targets of the DNA damage
checkpoint, in such a way that mitotic entry is inhibited
at all costs (Figure 1, lower panel).
Whereas Plk1 is inhibited in response to DNA
damage, Plk3 was shown to be activated after treatment
with DNA damaging agents (Xie et al., 2001). Moreover, activation of Plk3 was shown to coincide with cell
cycle arrest and increased interaction and phosphorylation of p53 by Plk3. Thus, Plk1 and Plk3 appear to have
counteracting rather than overlapping roles in mitotic
entry. Indeed, whereas Plk1 was shown to mediate
nuclear translocation of Cdc25C, Plk3 phosphorylates
Cdc25C on Ser216, a site known to be involved in
nuclear exclusion (Ouyang et al., 1999; ToyoshimaMorimoto et al., 2002). The exact mechanism by which
Plk3 regulates the G2/M transition in general and
Cdc25C in particular is unclear however, as a recent
report showed that Cdc25C translocates to the nucleus
after (direct) phosphorylation by Plk3 (Bahassi el et al.,
2004).
Not only DNA damage activates a checkpoint at the
G2–M transition. All kinds of stress feed into checkpoint signaling, and inhibit Cyclin B–Cdk1 activation.
In response to altered microtubule dynamics for
instance, the ubiquitin-ligase Chfr was shown to be
activated and responsible for delayed mitotic entry
(Scolnick and Halazonetis, 2000). Interestingly, activation of this so-called ‘prophase-checkpoint’ was
Oncogene
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2848
suggested to result in ubiquitination-mediated degradation of Plk1 via Chfr (Kang et al., 2002). Through
downregulation of Plk1, the prophase checkpoint would
lead to inhibition of Cdk1 activity. More recent data
confirm that this checkpoint requires protein ubiquitination, but also demonstrate that proteasomal degradation is not involved (Matsusaka and Pines, 2004). In
fact, Matsusaka and Pines (2004) did not observe Plk1
degradation following activation of the prophase
checkpoint. Thus, the exact nature of this checkpoint
and its role in controlling cell cycle progression remains
largely unknown.
Polo-like kinases and cell cycle resumption after
a checkpoint arrest
In addition to being a target of the DNA damage
checkpoint, Polo-like kinases were also shown to
regulate cell cycle progression after a damage-induced
cell cycle arrest. The first evidence for such a role was
provided by experiments done in S. cerevisiae mutants
that are unable to repair double strand breaks. Yeast
cells ultimately escape the DNA damage checkpoint
arrest in a process called ‘adaptation’. Such a mechanism allows damaged cells to eventually divide and
possibly survive. Alternatively, these cells are allowed
to repair their DNA damage in the subsequent G1.
Although the exact mechanism of checkpoint adaptation remains unclear, the inactivation of Rad53 seems to
be a key switch. In agreement with this notion, yeast
strains lacking functional PP2C phosphatases (known to
inactivate Rad53) cannot adapt in response to a doublestrand break (Leroy et al., 2003). Interestingly, strains
carrying a mutant form of Cdc5 (the Plk1–homolog in
budding yeast) were also unable to perform adaptation
(Toczyski et al., 1997).
Since yeast cells do not run the risk of developing
cancer, ‘adaptation’ for these cells can be beneficial to
help population survival. In contrast, a leaky checkpoint
poses a serious threat to the genomic integrity of a cell
and can potentially cause malignancies in multicellular
organisms. Therefore, the ‘adaptation’ process was long
thought to be limited to single-cell organisms. However,
this notion was challenged by Dunphy and co-workers,
who provided evidence for checkpoint adaptation in
response to stalled replication forks in Xenopus extracts
(Yoo et al., 2004). Interestingly, Plx1 was shown to be
required during this checkpoint adaptation. In this
process, Plx1 was shown to displace the checkpoint
mediator Claspin from chromatin through direct phosphorylation (Yoo et al., 2004). Here again, Plx1 was
reported to dock to its substrate (Claspin), once this
substrate had been ‘primed’ by phosphorylation. In this
case, the priming kinase was shown to be the Xenopus
homolog of ATR (Figure 2) (Yoo et al., 2004). This
observation extends the number and nature of priming
kinases that can direct Plk1 to its future substrate and
shows the versatility of this system (for a list see
Figure 3b) (Elia et al., 2003a; Yoo et al., 2004). The Xatr
Oncogene
phosphorylation site S904 is conserved between Xenopus
and humans. However, while S904 is located in an
ATM/ATR consensus site in Xenopus Claspin, it resides
in a Cdk1 consensus motif in human Claspin. So
whether Plk1-Claspin mediated adaptation is conserved
in all multicellular organisms remains to be clarified.
But more importantly, why would multicellular
organisms perform adaptation? One reason could be
that adaptation allows the clearance of cells with
irrepairable damage through mitotic catastrophe. Indeed, evidence is present for a mechanism in which
mitotic entry in the presence of DNA damage can lead
to cell death via mitotic catastrophe (Nitta et al., 2004).
Another reason for adaptation in multicellular organisms could lie in the fact that in G1 other (more
favorable) types of DNA damage repair are available.
For double-strand breaks this seems unlikely since the
duplicated sister chromatids in G2 create optimal
conditions for HR. For stalled replication, however,
alternative repair mechanisms may be present in G1.
To date, evidence for checkpoint adaptation in
mammalian cells is lacking. However, human cells are
expected to restart the cell cycle once DNA damage is
completely repaired. In fact, Plk1 was recently shown to
play a role in this recovery-process through regulation
of Cdk1 phosphorylation (van Vugt et al., 2004a). Both
Plk1 as well as Cdc25B were shown to be required for
efficient mitotic entry, once the G2 DNA damage
checkpoint was silenced. These results indicate a clear
difference between damaged and undamaged cells.
Whereas depletion of Plk1 or Cdc25B did not notably
affect mitotic entry in an unperturbed cell cycle,
damaged cells critically required Plk1 and Cdc25B for
mitotic entry after checkpoint silencing (van Vugt et al.,
2004a) (see also Figure 2). These results indicate that
damaged and undamaged cell cycles are wired fundamentally different.
Checkpoint recovery driven by Plk1 was shown to
involve the degradation of Wee1 triggered through
direct phosphorylation by Plk1 (Figure 2) (van Vugt
et al., 2004a). However, as Wee1 degradation was
suggested to depend on phosphorylation by both Plk1
and Cyclin B–Cdk1, it is not unlikely that small
amounts of Cyclin B–Cdk1 are required to establish its
own amplification loop through Plk1 after a DNA
damage arrest (Watanabe et al., 2004). This would be
consistent with earlier observations made in yeast, where
Cyclin B–Cdk1 activity is required to inactivate
checkpoint signaling and promote re-entry into the cell
cycle through phosphorylation of Crb2 (Esashi and
Yanagida, 1999).
At first glance it seems contradictory that Plk1 is both
a target of the DNA damage checkpoint as well as an
essential component of the restart mechanism after
checkpoint inactivation (Smits et al., 2000; van Vugt
et al., 2001, 2004a). However, Plk1 must be efficiently
inhibited to ensure proper checkpoint functioning
(Smits et al., 2000). Thus, this apparent contradiction
would be resolved if Plk1 were to be inactivated by
checkpoint kinases as long as the damage persists.
Indeed, Plk1 activity was only restored when DNA
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2849
Xenopus
adaptation after
stalled replication
stalled replication
Xatr
Xatr
Claspin
p
p
Claspin
Plx1
p
p
Claspin
p
Xchk1
Xchk1
Cdc25
cyclin B
ion
iat
oc
s
dis
Cdc25
Cdk1
cyclin B
Cdk1
inactive Cyclin B-Cdk1
active Cyclin B-Cdk1
double strand break
recovery from a
double strand break
Mammalian
ATR
ATR
ATM
p
p
Chk1 Chk2
Plk1
p
Wee1
cyclin B
ATM
p
Cdc25C
Cdc25B
Cdc25A
Ub
Ub
Cdk1
Ub
inactive Cyclin B-Cdk1
Plk1
Ub
Ub
Ub
p
Wee1
p
Chk1 Chk2
Cdc25B
?
cyclin B
Cdk1
active Cyclin B-Cdk1
Figure 2 Adaptation versus recovery. In response to stalled replication in Xenopus, Xatr is activated and promotes Xchk1 activation
in a Claspin-dependent fashion (upper-left panel). In response, Xchk1 can prevent activation of Cyclin B–Cdk1 through inhibition of
Cdc25 phosphatases. During adaptation, Plx1 binds and phosphorylates Xatr-phosphorylated Claspin to displace Claspin from
chromatin and silence checkpoint signaling. As a result, Cyclin B–Cdk1 is activated to promote mitotic entry (upper-right panel).
When the G2 DNA damage checkpoint is activated in response to UV or double-strand breaks in mammalian cells, ATM/ATR
kinases are activated and subsequently activate Chk1/Chk2. Chk1, Chk2 or p38 inhibit Cdc25C and Cdc25B, promote degradation of
Cdc25A and activate Wee1 to prevent activation of Cyclin B–Cdk1 (lower-left panel). During recovery after checkpoint silencing, Plk1
and Cdc25B are both required for Cyclin B–Cdk1 activation and subsequent mitotic entry (lower-right panel). Plk1 (possibly in
combination with Cyclin B–Cdk1) is required for Wee1-degradation during this process through direct phosphorylation
damage checkpoint signaling was silenced, and Cyclin
B–Cdk1 activation as well as mitotic entry were only
observed in cells expressing Plk1 (van Vugt et al.,
2004a), consistent with a role for Plk1 as the trigger
kinase during recovery.
As DNA damage is repaired, checkpoint signaling
will cease and (constitutively active) phosphatases can
revert the inhibition enforced by checkpoint kinases so
that Plk1 activity is allowed to rise. The identity of the
phosphatases involved in the reversion of the checkpoint
Oncogene
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2850
Figure 3 Plk1: localization and substrates. (a) During mitosis Plk1 localizes to distinct structures as indicated. Representative images
of paraformaldehyde-fixed U2OS cells show Plk1 (red) and DNA (blue). Images were obtained by confocal microscopy. In G2, Plk1 is
present on centrosomes. In prophase and metaphase, Plk1 localizes to centrosomes and kinetochores. During anaphase and telophase,
Plk1 is present on centrosomes, kinetochores and the central spindle. (b) List of reported Plk1 substrates and their localization. Protein
sequences were analysed for the presence of Cdk1 consensus motifs that could function as Polo-box binding motifs (Ser-phospho-Ser/
Thr-Pro)
signaling cascades is currently unknown, but in yeast
phosphatases were recently identified that are required
for a restart of the cell cycle following DNA damage
(Leroy et al., 2003; Den Elzen and O’Connell, 2004).
The PP2-related phosphatases Ptc2 and Ptc3 regulate
Rad53 and are required for efficient recovery as well as
adaptation in S. cerevisiae (Leroy et al., 2003). The PP1related Dis2 phosphatase reverses Chk1 phosphorylation by Rad3 (the fission yeast homolog of ATR), and
Oncogene
was shown to be required for recovery from a DNA
damage arrest in S. pombe (Den Elzen and O’Connell,
2004). However, the precise role of related phosphatases
in recovery in mammalian cells remains to be identified.
Interestingly, the mammalian PP2d-phosphatase Wip1
was shown to be required for recovery after a UVinduced arrest through dephosphorylation of p38
(Takekawa et al., 2000). Whether such a mechanism
also regulates Chk1/Chk2 activity in mammalian cells
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2851
and whether PP1-related phosphatases are also involved
in checkpoint recovery requires further research. In
addition, it would be interesting to see whether Plk’s
regulate such phosphatase activity during checkpoint
inactivation, especially since yeast cells with mutant
Cdc5 are unable to dephosphorylate Rad53 and Chk1
(Pellicioli et al., 2001).
Plk1 and mitotic spindle formation
Loss-of-function studies on Polo-like kinases in a range
of organisms, from yeast to mammals, showed severe
mitotic spindle defects (Sunkel and Glover, 1988;
Llamazares et al., 1991; Kitada et al., 1993; SpankuchSchmitt et al., 2002; van Vugt et al., 2004b). In
Drosophila for instance, Polo mutants displayed a high
percentage of monopolar spindles, or bipolar spindles
with one of the poles displaying an abnormal morphology (Sunkel and Glover, 1988). Similarly, in human cells
interference with the function of Plk1 results in the
formation of monopolar spindles (Lane and Nigg, 1996;
van Vugt et al., 2004b). This is not due to a block in
centrosome duplication, but a consequence of impaired
centrosome maturation and separation. Multiple aspects
of centrosome function appear to be affected in Plk1deficent cells. For one, Plk1 was reported to regulate
spindle morphology through g-tubulin recruitment to
centrosomes (Lane and Nigg, 1996). Although Plk1 was
shown to directly bind to g-tubulin (Feng et al., 1999),
regulation of tubulin-nucleation at the centrosome
probably involves multiple Plk1-interacting proteins.
Ninein-like protein (Nlp) and OP18/Stathmin, both
shown to be directly interacting with Plk1, are negative
regulators of microtubule nucleation (Budde et al., 2001;
Casenghi et al., 2003). Moreover, both Nlp and
Stathmin are phosphorylated by Plk1 and at least for
Nlp, this results in displacement from centrosomes to
allow tubulin-nucleation (Casenghi et al., 2003). In
contrast, Polo-like kinases also regulate microtubulestabilizing proteins, including Asp and TCTP (do
Carmo Avides et al., 2001; Yarm, 2002). Polo-mediated
phosphorylation of Asp is required for the aster-forming
capacity of centrosomes and the establishment of a
decent spindle (do Carmo Avides et al., 2001). Whether
TCTP-phosphorylation also plays a role in regulation of
spindle formation or functions to stabilize tubulin-fibers
during cytokinesis is still unclear. The fact that removal
of Plk1-phosphorylation sites in TCTP results in multinucleated cells, may argue for a role during cytokinesis
(Yarm, 2002). However, TCTP looses its association
with the mitotic spindle during the meta-to-anaphase
transition and Plk1 is localized to kinetochores up to
metaphase, which suggests a model in which TCTP on
microtubules can be phosphorylated by Plk1 only when
microtubules attach to kinetochores, potentially resulting in stabilization of the attached microtubule (Arnaud
et al., 1998; Gachet et al., 1999; Yarm, 2002). Whether
Plk1 also regulates the human homolog of Drosophila
Asp remains to be established.
The observed requirement for Plk1 to form a bipolar
spindle could well be the most important function of
Plk1 during somatic cell cycles. For one because most of
the observed phenotypes after interference with Plk1
function concern defects in spindle morphology, giving
rise to a mitotic arrest (Lane and Nigg, 1996; Liu and
Erikson, 2002; Seong et al., 2002; Spankuch-Schmitt
et al., 2002; Kraft et al., 2003; van Vugt et al., 2004b).
However, more importantly, strikingly similar mitotic
defects are observed in Plk1-deficient cells when
compared to cells treated with monastrol, a drug that
causes the formation of monopolar spindles by inhibiting centrosome separation (van Vugt et al., 2004b).
Inactivation of the spindle checkpoint in Plk1-deficient
or monastrol-treated cells allows these cells to exit
mitosis, and midbody formation and cleavage furrow
ingression were not affected, despite the lack of a
bipolar spindle in both conditions (van Vugt et al.,
2004b). This strongly suggests that the centrosome
defects are the prime defects that arise after interfering
with Plk1 function, while other phenotypes may arise as
a secondary consequence of the impaired spindle
function. In addition, these observations underline the
limited role for Plk1 during mitotic entry in unperturbed
cell cycles.
Plk1 and APC activation
The anaphase-promoting complex or cyclosome (APC/
C) is a multisubunit ubiquitin ligase that targets a range
of mitotic proteins for proteasomal degradation. The
APC/C is regulated both by phosphorylation and by
binding to ancillary proteins of the ‘WD-repeat’ family:
Cdc20 and Cdh1 (Peters, 2002). From prophase to
metaphase, the APC/C binds Cdc20, while Cdh1 is
phosphorylated by Cyclin B–Cdk1 complexes and
unable to bind the APC/C at these stages of mitosis.
However, when Cyclin B–Cdk1 is degraded in metaphase, inhibitory phosphorylation on Cdh1 is lost
through dephosphorylation, and Cdh1 can now bind
and activate the APC/C. Cdc20-dependent APC/C
ubiquitinates D-box-containing proteins including Btype Cyclins and securin. In contrast, the Cdh1activated APC/C targets proteins for destruction that
contain either a D-Box or a KEN-box, such as Cdc20,
Aurora kinases and Plk1 (Shirayama et al., 1998;
Lindon and Pines, 2004 and reviewed in Peters, 2002).
During the early stages of mitosis, Cdc20 is kept
inactive by the spindle assembly checkpoint. The
inhibitory signal towards Cdc20 emitted by the spindle
assembly checkpoint is essential for proper chromosome
segregation, as it prevents the resolution of the cohesive
linkage between paired sister chromatids before all of
the chromosomes have correctly attached to the mitotic
spindle (reviewed in Yu, 2002). This cohesive linkage is
formed by a multi-subunit ring-structure called cohesin.
Cleavage of cohesin is dependent on a cysteine protease,
separase, that in turn is inhibited through binding to the
scaffold protein securin (Ciosk et al., 1998). At
Oncogene
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2852
metaphase, securin is targeted for ubiquitin-mediated
destruction by Cdc20-dependent APC/C, and separase
can subsequently cleave cohesin, allowing the sister
chromatids to separate (Cohen-Fix et al., 1996; Visintin
et al., 1997). As Cdc20 is inhibited as long as not all
chromosomes have attained bipolar attachment, cohesin
degradation cannot take place before metaphase, so that
unequal distribution of chromatids and subsequent
aneuploidy in daughter cells is prevented (Jallepalli
et al., 2001). It should be noted that degradation of some
Cdc20-APC/C targets is allowed even before the spindle
assembly checkpoint is silenced. For instance, Cyclin A
and Nek2A are degraded as soon as cells enter mitosis,
independently of spindle checkpoint inactivation (den
Elzen and Pines, 2001; Geley et al., 2001; Hames et al.,
2001). However, it remains unclear how the Cdc20APC/C can distinguish between targets that need to be
degraded before and after spindle checkpoint inactivation.
The APC/C is jointly regulated through phosphorylation by Cyclin B–Cdk1, Polo-like kinases as well as
PKA (Kotani et al., 1998; Zachariae and Nasmyth,
1999; Rudner and Murray, 2000; Golan et al., 2002;
Kraft et al., 2003). Whereas Cyclin B–Cdk1 and Plk1
were shown to activate the APC/C, PKA was shown to
decrease APC/C activity (Kotani et al., 1998). Cyclin B–
Cdk1 phosphorylates a range of APC/C subunits, of
which 15 phosphorylation sites were confirmed in vivo
(Kraft et al., 2003). In addition, Cyclin B–Cdk1 activity
promotes Cdc20-binding to the APC/C and increases its
ubiquitin ligase activity (Kotani et al., 1998; Golan
et al., 2002; Kraft et al., 2003). The significance of APC/
C phosphorylation by Cyclin B–Cdk1 was underlined by
mutational analysis; removal of Cyclin B–Cdk1 sites in
subunits of the yeast APC/C prevented Cdc20-binding
and abolished APC/C activity (Rudner and Murray,
2000).
Polo-like kinases were also shown to be regulators (as
well as targets) of the APC/C. In particular, activity of
Polo-like kinases was shown to be required for
degradation of mitotic Cyclins in yeast and Xenopus
cells (Descombes and Nigg, 1998; Shirayama et al.,
1998). A range of APC/C subunits is phosphorylated by
Plk1 in vitro, including APC6 (or Cdc16), APC3 (or
Cdc27), APC1 (or Tsg24), APC5 and Cdc23 (Kotani
et al., 1998; Golan et al., 2002; Kraft et al., 2003). This
phosphorylation of the APC/C by Plk1 was shown to
increase ubiquitin ligase-activity in vitro (Kotani et al.,
1998; Golan et al., 2002). However, determination of in
vivo APC/C phosphorylation only confirmed few of the
sites that were reported to be phosphorylated in vitro
(Kraft et al., 2003), thus making the impact of APC/C
phosphorylation by Plk’s unclear. And even though
depletion of Plx1 from Xenopus extracts prevented
Cyclin B degradation, this may reflect a specific
requirement for Plx1-mediated APC activation during
exit from meiosis II (Descombes and Nigg, 1998). In
addition, when Cdk1-phosphorylation sites were mutated in yeast APC/C subunits, APC/C was still
phosphorylated by the Polo-like kinase Cdc5 in vitro,
but APC/C activity was diminished (Rudner and
Oncogene
Murray, 2000). Although it does not rule out a role
for Plk1 in APC/C regulation, it does show that Cyclin
B–Cdk1 acts as the prime APC/C regulator in vivo. This
idea is supported by depletion studies in which human
cells lacking Plk1 still degraded the APC/C-Cdc20 target
Cyclin A with normal kinetics (Kraft et al., 2003; van
Vugt et al., 2004b). Degradation of another APC/CCdc20 target, Cyclin B, was suggested to be defective
after Plk1-depletion (Liu and Erikson, 2002). However,
this observation was shown to be an indirect result of
spindle checkpoint activation, rather than an APC/C
activation deficiency, as Plk1-depleted cells degraded
APC/C-Cdc20 targets Cyclin B and securin after spindle
checkpoint inactivation (van Vugt et al., 2004b). Therefore, the importance of APC/C phosphorylation by Plk1
continues to be unclear. Although Plk1 does not seem
essential for APC/C activation, the fact that Plk1 does
increase Cdc20-binding and ubiquitin ligase activity of
the APC/C when added simultaneously with Cyclin B–
Cdk1, suggests that Plk1 might function to ensure
continuous APC/C activity or increase its ability to
recognize (a specific subset of) substrates. Also, the fact
that simultaneous addition of Cyclin B–Cdk1 and Plk1
resulted in additional phosphorylation of a subset of
APC subunits (APC1, APC5 and Cdc23) suggests that
priming phosphorylation may also play a role in APC
regulation (Figure 3b) (Kraft et al., 2003).
During interphase, the APC/C must be kept inactive
to allow accumulation of A and B-type Cyclins. The Fbox protein Emi1 (early mitotic inhibitor-1) prevents
premature APC/C activation and, by doing so, guarantees the capacity for mitotic entry (Reimann et al.,
2001a, b). Emi1 binds to Cdc20, and prevents binding of
Cdc20 to APC/C-targets (Reimann et al., 2001b). In
accordance, depletion of Emi1 strongly delays mitotic
entry, because of the inability to accumulate sufficiently
high levels of mitotic Cyclins. Conversely, proper
progression through mitosis requires timely Emi1degradation, in a process that is regulated by SCF-bTrCP (Guardavaccaro et al., 2003; Margottin-Goguet
et al., 2003). Binding of Emi1 to SCF-b-TrCP was
shown to require phosphorylation of Emi1 by Cdk1 and
an additional kinase. Recently, this additional kinase
was shown to be Plk1 and hence a two-step model was
proposed in which Cdk1 primes Emi for Plk1-mediated
phosphorylation and subsequent degradation (Margottin-Goguet et al., 2003; Moshe et al., 2004). Thus, Plk1
also regulates APC/C in an indirect fashion. However, it
should be noted that the contribution of Emi1phosphorylation by Plk1 to APC/C activation is
unclear, since Plk1-depleted cells arrest in mitosis with
degraded Cyclin A (Kraft et al., 2003; van Vugt et al.,
2004b).
It is still also unclear if functional redundancy exists
between Plk1 and other Plk’s in mitotic progression.
Plk2 is expressed predominantly during G1/S and was
recently shown to play a role in centriole duplication
rather than centrosome separation (Warnke et al.,
2004). Like Plk1, Plk3 is expressed mainly in G2/M
but likely plays a role that is limited to the G2/M
transition (Bahassi el et al., 2004). In contrast, Plk4
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2853
(Sak) was shown to play a role in mitotic progression.
Mice embryo’s lacking Sak displayed elevated levels of
anaphase/telophase cells (Hudson et al., 2001). Moreover, these Sak / cells were positive for Cyclin B,
indicative for defective APC/C activation. These findings, in combination with the observed APC/C activation in the absence of Plk1 (Kraft et al., 2003; van Vugt
et al., 2004b), suggest that the Sak may regulate the
human APC/C in a fashion similar to that of S.
cerevisiae Cdc5 and yeast APC/C. This way, the multiple
functions of Cdc5 may be divided over different Plk’s in
human cells. Future experiments will have to address
how specificity of polo-box mediated interaction of the
different family members is determined, and whether the
different kinase domains contribute to additional levels
of specificity when it comes to downstream targets.
Plk1 and condensation/separation of chromosomes
During DNA replication in S phase cohesin is recruited
to chromosome arms and centromere regions, and
maintains sister-chromatid cohesion. When cells enter
mitosis, the chromatin is tightly packed in a process
called ‘condensation’ (reviewed in Hirano, 2000). This
procedure involves recruitment of multisubunit condensin complexes as well as histone phosphorylation
and is tightly coupled with the displacement of cohesin
complexes from the chromosome arms in prophase.
Nonetheless, the linkage between duplicated sisterchromatids is maintained during this stage through
residual amounts of cohesin at centromeric regions
(Waizenegger et al., 2000). In metaphase all residual
cohesin is then removed from chromosomes through
cleavage of its subunit SCC1, to allow chromosome
segregation (Uhlmann et al., 1999, 2000; Hauf et al.,
2001 and reviewed in Nasmyth et al., 2000).
Polo-like kinases have been implicated to regulate
chromosome condensation and separation. The initial
examination of Polo mutants in Drosophila revealed
hypercondensed chromosomes (Llamazares et al., 1991).
These observations matched later studies that described
condensation defects when Plk1-function was interfered
with (Losada et al., 2002; Gimenez-Abian et al., 2004;
van Vugt et al., 2004a). In contrast to condensin
recruitment, for which no direct involvement for Pololike kinases is described, cohesin removal is directly
regulated by Plk’s. Involvement of Plk’s in cohesin
displacement was originally identified in S. cerevisiae. In
contrast to vertebrate cells, which remove chromatinbound cohesin in a two-step fashion, yeast cells remove
cohesin in one step through separase-mediated cleavage.
During cohesin displacement in S. cerevisiae, the cohesin
subunit SCC1 is phosphorylated by the Polo-homolog
Cdc5 (Alexandru et al., 2001). This phosphorylation
event strongly enhances SCC1-cleavage by separase and
is essential for proper chromosome segregation. Recently, phosphorylation of SCC1 by polo-like kinases
was shown to restrict cleavage by separase to chromatinbound cohesin (Hornig and Uhlmann, 2004). Probably,
the colocalization of polo-like kinases and cohesin to
kinetochores is used to confine cohesin cleavage and
conserve cytoplasmic cohesin.
Polo-like kinases also regulate sister-chromatid cohesion in multicellular organisms. Sumara et al. (2002)
showed that SCC1-phosphorylation by Plk’s is conserved in vertebrates. Using depletion experiments,
Xenopus Plx1 was shown to be required for cohesin
displacement from chromatin through phosphorylation
of SCC1. In vertebrates and higher organisms, however,
Plk1-mediated phosphorylation promotes cleavage-independent removal of cohesin, but this removal is
restricted to cohesin that is associated to the chromosome arms. Essentially, SCC1 phosphorylated by Plx1
has a markedly decreased affinity for chromatin. In line
with this, addition of active Plx1 was sufficient to
mediate cohesin-dissociation from chromatin. In addition, human Plk1 was shown to phosphorylate SCC1 in
vitro on similar sites as in Xenopus SCC1 (Sumara et al.,
2002). It is unlikely that Plk1-dependent phosphorylation of SCC1 also regulates cleavage of SCC1 by
separase after spindle checkpoint inactivation, since
cells lacking Plk1 remove SCC1 from chromosomes
normally when the spindle checkpoint is inactivated
(Gimenez-Abian et al., 2004).
Plk1 and cytokinesis
The observed localization of Plk1 during mitosis nicely
fits its proposed role in cytokinesis (Figures 3a and 4).
Similar to chromosome passenger proteins, Plk1 is
present at kinetochores from prophase until metaphase
and translocates to the central spindle and the midzone
in anaphase and telophase, respectively (Golsteyn et al.,
1995). Different molecular interactions were reported to
be responsible for this change in subcellular localization
(Figure 3). Two possible docking proteins for Plk1 on
the central spindle are the Mitotic Kinesin-like protein-1
and -2 (MKlp1/CHO1 and MKlp2). Plk1 was shown to
phosphorylate both MKlp1/CHO1 and MKlp2, which
are located on the central spindle (Lee et al., 1995;
Adams et al., 1998; Neef et al., 2003; Liu et al., 2004).
For Mklp2, phosphorylation by Plk1 was shown to
create a Plk1-binding site (Neef et al., 2003). In line with
these observations, depletion of MKlp1/CHO1 or
MKlp2 or deletion of the Plk1-phosphorylation sites
in MKlp1/Cho1 or MKlp2 disrupts Plk1 localization to
the central spindle (Lee et al., 1995; Adams et al., 1998;
Neef et al., 2003; Liu et al., 2004). Similar observations
were made with the Golgi-associated protein Nir2
(Litvak et al., 2004). Plk1 was shown to depend on
phosphorylated Nir2 for decent localization to the
central spindle. However, whereas MKlp1/CHO1 and
MKlp2 require Plk1-phosphorylation, Nir2 requires
Cdk1-phosphorylation for Plk1-binding. Furthermore,
both Nir2-depletion or removal of the Cdk1-phosphorylation site in Nir2 resulted in failure to properly
localize Plk1 to the midzone (Litvak et al., 2004). In
addition, Plk1 was shown to phosphorylate NudC, a
Oncogene
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2854
Figure 4 Mitotic exit. In anaphase, Plk1 relocalizes from kinetochores to the central spindle, similar to chromosome passenger
proteins. This relocalization depends on functional Nir2, NudC, MKlp1 and MKlp2. In addition, NudC, MKlp1 and MKlp2 are
directly phosphorylated by Plk1 and these phosphorylation events were shown to be required for successful cytokinesis. During
anaphase, Plk1 is degraded through the Cdh1-activated APC/C. This degradation is required for proper cytokinesis, probably because
activity of Plk1 (as was shown for Cyclin B–Cdk1 activity) inhibits some critical steps in cytokinesis. At least for MKlp2,
phosphorylation by Plk1 was shown to decrease its microtubule-stabilizing effect during cytokinesis
dynein-associated nuclear movement protein that plays
a role in cytokinesis (Zhou et al., 2003). Clearly, multiple
proteins are required to properly localize Plk1 to the
central spindle. Such multiple requirements may arise as
a consequence of complex formation between MKlp1,
MKlp2, Nir2 and possibly NudC. Alternatively, Plk1
could be recruited to the central spindle by one class of
proteins, while its sustained association to the central
spindle is mediated by yet another protein.
Cdk1-directed polo-box recruitment might also apply
to other proteins. Indeed, one of the chromosome
passenger proteins, INCENP, harbors a Cdk1 consensus
phosphorylation site that perfectly fits the requirements
for a polo-box binding site and is probably able to bind
Plk1 (Elia et al., 2003a). Moreover, other Plk’s localize
to the central spindle. Ectopic expression of tagged Plk3
and Plk4, for instance, also shows they can localize to
the spindle midzone (Conn et al., 2000; Hudson et al.,
2001). And although phosphorylation mutants of
MKlp1, MKlp2 and Nir2 are presumably unable to
bind the polo-box of Plk1 and clearly show defective
localization of Plk1, it is also conceivable that such
mutants prevent other Plk’s from binding to the central
spindle. Moreover, involvement of multiple Plk family
members in mitotic exit would explain why phosphomutants of MKlp2 result in severe cytokinesis defects,
while Plk1 depletion does not interfere with furrow
ingression (van Vugt et al., 2004b). In addition, the fact
that overexpression of both Plk1 and Plk3 results in
multinucleation suggests cytokinesis is regulated by
Oncogene
multiple Plk’s (Mundt et al., 1997; Conn et al., 2000).
Functional redundancy between Plk1 and Plk3 may also
play a role in inheritance of cellular organelles during
mitosis. Both Plk1 and Plk3 were suggested to regulate
partitioning of Golgi-stacks during cell division (Sutterlin et al., 2001; Ruan et al., 2004). In addition, Plk1 was
shown to interact and phosphorylate the Golgi-associated proteins Grasp65 and Nir2 (Lin et al., 2000;
Litvak et al., 2004) and Plk3 colocalizes to Golgimarkers Giantin. Finally, overexpression of Plk3 results
in premature Golgi-partitioning (Ruan et al., 2004).
Whether Plk1 and Plk3 regulate distinct processes in
Golgi partitioning remains unclear. However, the fact
that Grasp65 is phosphorylated in vitro by Cyclin B–
Cdk1 and the presence of a polo-box-binding site
suggest that also here Cdk1 can direct phosphorylation
of Plk’s (Figure 3) (Lin et al., 2000).
While different molecular targets for Plk1 were
identified that play a role in correct localization of
Plk1 to the central spindle and cytokinesis, the exact
function of Plk1 at the central spindle is still unclear.
Clearly, interference with Plk1-substrates MKlp1,
MKlp2, NudC or Nir2, as well as Plk1 itself results in
defects in the final stages of cytokinesis rather then
disturbing the whole cytokinesis process (Seong et al.,
2002; Neef et al., 2003; Zhou et al., 2003; Litvak et al.,
2004; Matuliene and Kuriyama, 2004; van Vugt
et al., 2004b). In addition, removal of the Plk1
phosphorylation sites in MKlp1/CHO1 or MKlp2 does
not block furrow ingression, but prevents abscission,
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2855
suggesting that Plk1 plays a role in this final step of
cytokinesis (Neef et al., 2003; Liu et al., 2004).
One possible mechanism through which Plk1 activity
is linked to cytokinesis was provided by experiments in
C. elegans. Cytokinesis in C. elegans requires complex
formation between Zen-4 (homolog of the human
MKlp1) and ZYG-4, a Rho GTPase activator
(Jantsch-Plunger et al., 2000; Mishima et al., 2002).
Regulation of Rho, in turn, was shown to be required
for successful cytokinesis (reviewed in Glotzer, 2001).
Recently, also human MKlp1 was shown to form a
complex with a homologous RhoGAP (Mishima et al.,
2002). Regulation of this complex likely involves
phosphorylation by both Cyclin B–Cdk1 and Plk1.
Cyclin B–Cdk1 directly phosphorylates MKlp1 to
inhibit its motor activity and prevent it from binding
to the mitotic spindle (Mishima et al., 2004). Plk1 also
phosphorylates MKlp1 in vitro, and this phosphorylation was suggested to be required for successful
cytokinesis (Liu et al., 2004). Combining this observation with the fact that MKlp1 is required for Plk1
localization makes it tempting to speculate that this is
another situation in which Plk1 phosphorylates target
proteins that are primed by Cdk1-phosphorylation.
However, the conserved Cdk1 phosphorylation sites in
MKlp1 do not match the Polo-box binding consensus
Ser-Ser/Thr-Pro. Moreover, Plk1 was reported to bind
MKlp1 during anaphase, and proper MKlp1 localization during anaphase requires CDC14-mediated dephosphorylation of the Cdk1 sites (Mishima et al., 2004). So
how exactly Plk1 binds MKlp1 is unclear, but it seems
unlikely that Plk1 uses Cdk1-directed phosphorylation
to dock to MKlp1. Thus, whether Plk1 ultimately
regulates Rho GTPase activity after priming Cdk1
phosphorylation, and whether this involves MKlp1,
MKlp2 or both requires additional experimentation.
In contrast, there is also evidence to suggest that it is
not Plk1 activity, but the efficient elimination of Plk1
that is required for proper mitotic exit. Disruption of a
KEN-box motif in Plk1 prevents its degradation by the
APC/C-Cdh1 and clearly delays mitotic exit (Lindon
and Pines, 2004). This was also suggested by the fact
that Plk1 negatively regulates the microtubule-bundling
capacity of Mklp2 (Neef et al., 2003). Plk1 inactivation
can thus contribute to increased microtubule bundling
at the central spindle. Consequently, enhanced microtubule bundling can cause local decreases in microtubule
concentration, which were suggested to precede cytokinesis (Dechant and Glotzer, 2003).
Taken together, it appears important for successful
cytokinesis that (1) Plk1 is localized correctly to the
central spindle, and (2) Plk1 is degraded efficiently. To
investigate whether these two processes are linked,
further research is required, but the fact that APC/C
also localizes to the central spindle argues in favor of a
model in which midzone localization of Plk1 is required
for its degradation (Kurasawa and Todokoro, 1999).
The difficulty in pinpointing a specific role for Plk’s in
cytokinesis has been the inability to analyse cytokinesis
in the absence of Plk’s, without affecting other mitotic
processes. Using a dominant-negative approach, Seong
and co-workers did find abortive cytokinesis, but in this
assay the dominant-negative effect was not powerful
enough to prevent bipolar spindle formation, and hence
not a full Plk1 inactivation may be observed (Seong
et al., 2002). In addition, this dominant-negative
approach using a Plk1 fragment harboring the polobox domain may interfere with the function of other
Polo-like kinases and reveal general functions of Pololike kinases, not Plk1 specifically. RNAi-mediated
depletion of Plk1 prevents these problems, but also this
technique does not allow observations of late mitosis
because of spindle checkpoint activation. When Plk1
depletion was combined with spindle checkpoint inactivation, no 2N daughter cells were produced (van
Vugt et al., 2004b). However, this effect could equally
well be explained by the observed spindle problems,
since spindle checkpoint inactivation in Plk1 proficient
cells with a monopolar spindle pheno-copied this effect
(Canman et al., 2003; van Vugt et al., 2004b). More
importantly, furrow ingression as well as midbodyformation did occur in these Plk1-depleted cells,
indicating that Plk1 is not essential for these aspects of
cytokinesis (van Vugt et al., 2004b). Timed Plk1
inactivation in cells that obtain a bipolar spindle will
be required to get a clean picture of Plk1 functions in
cytokinesis.
Plk1 and cancer
Plk1 is overexpressed in a range of human tumors,
including prostate tumors, ovarian carcinomas, hepatoblastomas and melanomas, and Plk1 overexpression was
shown to coincide with bad prognosis (Kneisel et al.,
2002; Weichert et al., 2004a, b; Yamada et al., 2004).
Therefore, Plk1 is useful as a prognostic marker for
outcome of disease. In addition, overexpression of Plk1
in murine cells drives tumorigenesis underlining the
strong proliferation-stimulating effect of Plk1 (Smith
et al., 1997). Interestingly, whereas most studies
correlate overexpression of Plk1 with cancer, certain
tumor-associated mutations were found that destabilize
Plk1 (Simizu and Osada, 2000). The fact that Plk1
expression as well as its activity is obviously altered in
transformed cells has prompted the use of Plk1inhibition as a possible anticancer therapy. When antiPlk1 antibodies were injected into transformed and
primary human cells, a striking difference was observed.
Whereas transformed cells arrested in mitosis, primary
cells arrested in G2 (Lane and Nigg, 1996). Others
described the use of dominant-negative fragments of
Plk1, delivered through transfection or linkage to
antennapedia-fragments (Seong et al., 2002; Yuan
et al., 2002b). Again, both these techniques indicated a
strong requirement for Plk1 in mitotic progression as
well as proliferation. More recently, RNA interference
was used to block Plk1 function. Using synthetic
dsRNAs as well as short hairpin RNAs, Plk1-depletion
was shown to efficiently block cell growth of cultured
tumor cells (Spankuch-Schmitt et al., 2002; Liu and
Oncogene
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2856
Erikson, 2003; van Vugt et al., 2004b). In addition,
siRNA-producing vectors injected in nude mice significantly reduced growth of human tumor cells in a
xenograft model (Spankuch et al., 2004). The ultimate
challenge, however, remains to investigate tumor growth
of transformed cells, surrounded by normal somatic
cells of the same species that are sensitive to the same
siRNA. This way, possible undesired effects of Plk1
inhibition on normal somatic cell divisions can be
studied in detail.
Spindle poisons are frequently used in the treatment
of human cancer. These drugs cause a spindle checkpoint-mediated arrest, similar to what is observed in
Plk1-depleted cells (van Vugt et al., 2004b). Therefore,
one might reason that inhibition of Plk1 could be
helpful in cancer-therapy. However, although Plk1depletion appears to efficiently inhibit proliferation of
tumor cells, systemically depleting Plk1 would appear to
be a dangerous track, as interference with Plk1 activity
is also expected to result in severe problems in
centrosome functioning in primary cells. This could
potentially cause aberrant mitoses and result in chromosome instability in normal tissues that could contribute
to de novo tumor formation.
A second possible strategy involves antitumor therapy
through the G2/M DNA damage checkpoint. Genotoxic
therapy can eliminate tumor cells via the DNA damage
checkpoint in G2. Since Plk1 activity is required for
checkpoint recovery after a G2 DNA damage arrest,
inhibition of Plk1 in combination with genotoxic
therapies would enforce a more stringent G2 arrest
and prevent recovery of damaged cells (van Vugt et al.,
2004a). However, as discussed before, Plk1 inhibition
may also lead to an aberrant mitosis and subsequent
chromosomal instability in both tumor and normal
tissue. With this in mind, it may be better to try to
develop a strategy to interfere exclusively with the
function of Plk1 at mitotic entry in recovering cells,
rather than using inhibitors that inhibit Plk1 as a whole.
To accomplish this, one could specifically target those
Plk1-substrate interactions that are important for
mitotic entry, leaving other polo box-mediated interactions intact. Obviously, a better understanding of the
polo-box/phosphopeptide interaction will be required to
accomplish this. Alternatively, a simpler approach may
be to inhibit Cdc25B in combination with genotoxic
therapy. Cdc25B is also required for cells to recover
from a DNA damage checkpoint arrest (van Vugt et al.,
2004a). However, contrary to Plk1, interference with the
function of Cdc25B does not appear to have a major
impact on cell cycle progression in unperturbed cells
(van Vugt et al., 2004a). This is further supported by the
finding that mice lacking functional Cdc25B are viable
and healthy (Lincoln et al., 2002). Therefore, inhibition
of Cdc25B appears to be an attractive target for
combinational therapy without running the risk of
provoking tumor onset through genetic instability.
As a third alternative, Plk1 could be used as a target
for adjuvant therapy. Conventional spindle poisons
(such as paclitaxel) may be more effective in killing
tumor cells, when cells are not delayed at the G2/M
Oncogene
transition. Since Plk1 was suggested (albeit debated) to
be a target of the prophase checkpoint that prevents
mitotic entry in response to microtubule stress imposed
by spindle poisons, it could be useful to overcome this
negative effect on Plk1 activity. Similarly, a G2 DNA
damage checkpoint override, causing cells to enter
mitosis with residual DNA damage almost invariably
leads to cell death. Therefore, preserving Plk1 activity in
cells treated with genotoxic drugs may enhance cell
killing through mitotic catastrophe. This might be
accomplished by specific inhibition of the DNA damage
checkpoint pathways that can inactivate Plk1 activity in
response to damaging agents. However, to interfere with
such a specific interaction, further research is warranted
to elucidate the molecular mechanisms by which Plk1 is
inactivated under these conditions.
Discussion and future directions
The answer to the intriguing question as to how one
protein can regulate so many processes lies (at least in
part) in localization. Plk1’s unique localization pattern
is largely mediated by its phospho-binding polo-box and
an increasing number of reports illustrate the binding of
Plk1 polo-boxes to substrates that primed through
phosphorylation by other kinases (Figure 3). Based on
the work of Elia et al. (2003a), Plk1 is expected to dock
to a phosphorylated Cdk1 consensus motif, which
would limit Plk1 actions to late G2, prometaphase and
metaphase cells. Especially the roles for Plk1 in late
mitotic events, such as cytokinesis, cannot easily be
explained by Cdk-mediated priming. However, different
reports have now shown that Polo-like kinases also bind
to substrates that are primed by mitotic non-Cdk
kinases, notably Plk1 itself (Figure 3) (Neef et al.,
2003). In addition, the nonmitotic kinase, ATR, was
also shown to create a Plk1-binding site (Figure 2) (Yoo
et al., 2004). Whether other nonmitotic or mitotic
kinases, such as Aurora kinases, can direct Plk1
localization, or whether phosphorylation-independent
recruitment can also take place is currently unknown.
However, the fact that Plk1 remains centrosomeassociated from G2 until mitotic exit does argue in
favor of this latter idea.
While the majority of research concerning Plk’s and
cell cycle regulation deals with Plk1, also Sak (Plk4)
might be an important mitotic regulator (Hudson et al.,
2001). As observed for Plk1, Sak localizes to mitotic
structures via its polo box domain (Leung et al., 2002).
Moreover, mice lacking Sak are early embryonic lethal,
and display increased numbers of mitotic cells (Hudson
et al., 2001). However, it is not likely that Plk1 and Sak
are active in similar pathways and are functionally
completely redundant, since deletion of Sak or depletion
of Plk1 both result in mitotic defects. Another interesting feature of these Plk’s is the differential organization
of the polo-box domain; whereas Plk1 contains
two polo-boxes, Sak only has one (Leung et al., 2002).
Since both polo-boxes in Plk1 were shown to participate
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2857
in binding to a phosphorylated target-protein, it
will be interesting to see whether monomers or dimers
of Sak molecules function as a phospho-specific binding
motif.
In conclusion, more and more targets of Plk1 are
identified and, more importantly, the molecular modes
of action of Plk1-target interactions are revealed. These
recent results show that the interplay between Plk1 and
other kinases controls recruitment and activation of
Plk1 on certain subcellular structures and explain how
one kinase can regulate so many different processes in a
timely and spatially controlled fashion.
Acknowledgements
We thank members of the Medema lab for fruitful discussions.
Dr H te Riele, Dr S Lens and B van de Weerdt are
acknowledged for critical reading of this manuscript and
useful comments. We thank B van de Weerdt for confocal
microscopy images. MVV is supported by a grant from the
Dutch Cancer Society (NKI 00-2191).
References
Abrieu A, Brassac T, Galas S, Fisher D, Labbe JC and Doree
M. (1998). J. Cell Sci., 111 (Part 12), 1751–1757.
Adams RR, Tavares AA, Salzberg A, Bellen HJ and Glover
DM. (1998). Genes Dev., 12, 1483–1494.
Alexandru G, Uhlmann F, Mechtler K, Poupart MA and
Nasmyth K. (2001). Cell, 105, 459–472.
Arnaud L, Pines J and Nigg EA. (1998). Chromosoma, 107,
424–429.
Atherton-Fessler S, Parker LL, Geahlen RL and PiwnicaWorms H. (1993). Mol. Cell. Biol., 13, 1675–1685.
Bahassi el M, Hennigan RF, Myer DL and Stambrook PJ.
(2004). Oncogene, 23, 2658–2663.
Bailly E, Doree M, Nurse P and Bornens M. (1989). EMBO J.,
8, 3985–3995.
Bailly E, Pines J, Hunter T and Bornens M. (1992). J. Cell Sci.,
101 (Part 3), 529–545.
Barr FA, Sillje HH and Nigg EA. (2004). Nat. Rev. Mol. Cell.
Biol., 5, 429–440.
Booher RN, Holman PS and Fattaey A. (1997). J. Biol. Chem.,
272, 22300–22306.
Budde PP, Kumagai A, Dunphy WG and Heald R. (2001). J.
Cell Biol., 153, 149–158.
Canman JC, Cameron LA, Maddox PS, Straight A, Tirnauer
JS, Mitchison TJ, Fang G, Kapoor TM and Salmon ED.
(2003). Nature, 424, 1074–1078.
Casenghi M, Meraldi P, Weinhart U, Duncan PI, Korner R
and Nigg EA. (2003). Dev. Cell, 5, 113–125.
Ciosk R, Zachariae W, Michaelis C, Shevchenko A, Mann M
and Nasmyth K. (1998). Cell, 93, 1067–1076.
Cohen-Fix O, Peters JM, Kirschner MW and Koshland D.
(1996). Genes Dev., 10, 3081–3093.
Conn CW, Hennigan RF, Dai W, Sanchez Y and Stambrook
PJ. (2000). Cancer Res., 60, 6826–6831.
Dechant R and Glotzer M. (2003). Dev. Cell, 4, 333–344.
den Elzen N and Pines J. (2001). J. Cell Biol., 153, 121–136.
Den Elzen NR and O’Connell MJ. (2004). EMBO J., 23,
908–918.
Descombes P and Nigg EA. (1998). EMBO J., 17, 1328–1335.
do Carmo Avides M, Tavares A and Glover DM. (2001). Nat.
Cell Biol., 3, 421–424.
Elia AE, Cantley LC and Yaffe MB. (2003a). Science, 299,
1228–1231.
Elia AE, Rellos P, Haire LF, Chao JW, Ivins FJ, Hoepker K,
Mohammad D, Cantley LC, Smerdon SJ and Yaffe MB.
(2003b). Cell, 115, 83–95.
Elledge SJ. (1996). Science, 274, 1664–1672.
Ellinger-Ziegelbauer H, Karasuyama H, Yamada E, Tsujikawa K, Todokoro K and Nishida E. (2000). Genes Cells, 5,
491–498.
Esashi F and Yanagida M. (1999). Mol. Cell, 4, 167–174.
Fang G, Yu H and Kirschner MW. (1998). Mol. Cell, 2,
163–171.
Feng Y, Hodge DR, Palmieri G, Chase DL, Longo DL and
Ferris DK. (1999). Biochem. J., 339 (Part 2), 435–442.
Fesquet D, Labbe JC, Derancourt J, Capony JP, Galas S,
Girard F, Lorca T, Shuttleworth J, Doree M and Cavadore
JC. (1993). EMBO J., 12, 3111–3121.
Fisher RP and Morgan DO. (1994). Cell, 78, 713–724.
Gabrielli BG, Clark JM, McCormack AK and Ellem KA.
(1997). J. Biol. Chem., 272, 28607–28614.
Gabrielli BG, De Souza CP, Tonks ID, Clark JM, Hayward
NK and Ellem KA. (1996). J. Cell Sci., 109 (Part 5),
1081–1093.
Gachet Y, Tournier S, Lee M, Lazaris-Karatzas A, Poulton T
and Bommer UA. (1999). J. Cell Sci., 112 (Part 8),
1257–1271.
Galaktionov K and Beach D. (1991). Cell, 67, 1181–1194.
Geley S, Kramer E, Gieffers C, Gannon J, Peters JM and Hunt
T. (2001). J. Cell Biol., 153, 137–148.
Gimenez-Abian JF, Sumara I, Hirota T, Hauf S, Gerlich D,
De La Torre C, Ellenberg J and Peters JM. (2004). Curr.
Biol., 14, 1187–1193.
Glotzer M. (2001). Annu. Rev. Cell Dev. Biol., 17, 351–386.
Golan A, Yudkovsky Y and Hershko A. (2002). J. Biol.
Chem., 277, 15552–15557.
Golsteyn RM, Mundt KE, Fry AM and Nigg EA. (1995). J.
Cell Biol., 129, 1617–1628.
Guardavaccaro D, Kudo Y, Boulaire J, Barchi M, Busino L,
Donzelli M, Margottin-Goguet F, Jackson PK, Yamasaki L
and Pagano M. (2003). Dev. Cell, 4, 799–812.
Haber JE. (2000). Trends Genet., 16, 259–264.
Hagting A, Karlsson C, Clute P, Jackman M and Pines J.
(1998). EMBO J., 17, 4127–4138.
Hamanaka R, Smith MR, O’Connor PM, Maloid S, Mihalic
K, Spivak JL, Longo DL and Ferris DK. (1995). J. Biol.
Chem., 270, 21086–21091.
Hames RS, Wattam SL, Yamano H, Bacchieri R and Fry AM.
(2001). EMBO J., 20, 7117–7127.
Hauf S, Waizenegger IC and Peters JM. (2001). Science, 293,
1320–1323.
Heald R, McLoughlin M and McKeon F. (1993). Cell, 74,
463–474.
Hirano T. (2000). Annu. Rev. Biochem., 69, 115–144.
Hoffmann I, Clarke PR, Marcote MJ, Karsenti E and Draetta
G. (1993). EMBO J., 12, 53–63.
Hornig NC and Uhlmann F. (2004). EMBO J., 23, 3144–3153.
Hudson JW, Kozarova A, Cheung P, Macmillan JC, Swallow
CJ, Cross JC and Dennis JW. (2001). Curr. Biol., 11,
441–446.
Jackman M, Lindon C, Nigg EA and Pines J. (2003). Nat. Cell
Biol., 5, 143–148.
Jallepalli PV, Waizenegger IC, Bunz F, Langer S, Speicher
MR, Peters JM, Kinzler KW, Vogelstein B and Lengauer C.
(2001). Cell, 105, 445–457.
Oncogene
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2858
Jantsch-Plunger V, Gonczy P, Romano A, Schnabel H, Hamill
D, Schnabel R, Hyman AA and Glotzer M. (2000). J. Cell
Biol., 149, 1391–1404.
Jin P, Hardy S and Morgan DO. (1998). J. Cell Biol., 141,
875–885.
Kang D, Chen J, Wong J and Fang G. (2002). J. Cell Biol.,
156, 249–259.
Kelm O, Wind M, Lehmann WD and Nigg EA. (2002). J. Biol.
Chem., 277, 25247–25256.
Kitada K, Johnson AL, Johnston LH and Sugino A. (1993).
Mol. Cell. Biol., 13, 4445–4457.
Kneisel L, Strebhardt K, Bernd A, Wolter M, Binder A and
Kaufmann R. (2002). J. Cutan. Pathol., 29, 354–358.
Kotani S, Tugendreich S, Fujii M, Jorgensen PM, Watanabe
N, Hoog C, Hieter P and Todokoro K. (1998). Mol. Cell, 1,
371–380.
Kraft C, Herzog F, Gieffers C, Mechtler K, Hagting A, Pines J
and Peters JM. (2003). EMBO J., 22, 6598–6609.
Kramer A, Mailand N, Lukas C, Syljuasen RG, Wilkinson CJ,
Nigg EA, Bartek J and Lukas J. (2004). Nat. Cell Biol., 6,
884–891.
Kumagai A and Dunphy WG. (1996). Science, 273,
1377–1380.
Kurasawa Y and Todokoro K. (1999). Oncogene, 18,
5131–5137.
Lammer C, Wagerer S, Saffrich R, Mertens D, Ansorge W and
Hoffmann I. (1998). J. Cell Sci., 111 (Part 16), 2445–2453.
Lane HA and Nigg EA. (1996). J. Cell Biol., 135, 1701–1713.
Lee KS, Yuan YL, Kuriyama R and Erikson RL. (1995). Mol.
Cell. Biol., 15, 7143–7151.
Leroy C, Lee SE, Vaze MB, Ochsenbien F, Guerois R,
Haber JE and Marsolier-Kergoat MC. (2003). Mol. Cell, 11,
827–835.
Leung GC, Hudson JW, Kozarova A, Davidson A, Dennis JW
and Sicheri F. (2002). Nat. Struct. Biol., 9, 719–724.
Li J, Meyer AN and Donoghue DJ. (1997). Proc. Natl. Acad.
Sci. USA, 94, 502–507.
Lin CY, Madsen ML, Yarm FR, Jang YJ, Liu X and
Erikson RL. (2000). Proc. Natl. Acad. Sci. USA, 97,
12589–12594.
Lincoln AJ, Wickramasinghe D, Stein P, Schultz RM, Palko
ME, De Miguel MP, Tessarollo L and Donovan PJ. (2002).
Nat. Genet., 30, 446–449.
Lindon C and Pines J. (2004). J. Cell Biol., 164, 233–241.
Litvak V, Argov R, Dahan N, Ramachandran S, Amarilio R,
Shainskaya A and Lev S. (2004). Mol. Cell, 14, 319–330.
Liu X and Erikson RL. (2002). Proc. Natl. Acad. Sci. USA, 99,
8672–8676.
Liu X and Erikson RL. (2003). Proc. Natl. Acad. Sci. USA,
100, 5789–5794.
Liu X, Zhou T, Kuriyama R and Erikson RL. (2004). J. Cell
Sci., 117, 3233–3246.
Llamazares S, Moreira A, Tavares A, Girdham C, Spruce BA,
Gonzalez C, Karess RE, Glover DM and Sunkel CE. (1991).
Genes Dev., 5, 2153–2165.
Losada A, Hirano M and Hirano T. (2002). Genes Dev., 16,
3004–3016.
Maity A, McKenna WG and Muschel RJ. (1995). EMBO J.,
14, 603–609.
Margottin-Goguet F, Hsu JY, Loktev A, Hsieh HM, Reimann
JD and Jackson PK. (2003). Dev. Cell, 4, 813–826.
Matsusaka T and Pines J. (2004). J. Cell Biol., 166, 507–516.
Matuliene J and Kuriyama R. (2004). Mol. Biol. Cell, 15,
3083–3094.
Mishima M, Kaitna S and Glotzer M. (2002). Dev. Cell, 2,
41–54.
Oncogene
Mishima M, Pavicic V, Gruneberg U, Nigg EA and Glotzer
M. (2004). Nature, 430, 908–913.
Moore JD, Yang J, Truant R and Kornbluth S. (1999). J. Cell
Biol., 144, 213–224.
Morgan DO. (1995). Nature, 374, 131–134.
Moshe Y, Boulaire J, Pagano M and Hershko A. (2004). Proc.
Natl. Acad. Sci. USA, 101, 7937–7942.
Mundt KE, Golsteyn RM, Lane HA and Nigg EA. (1997).
Biochem. Biophys. Res. Commun., 239, 377–385.
Nagata A, Igarashi M, Jinno S, Suto K and Okayama H.
(1991). New Biol., 3, 959–968.
Nakajima H, Toyoshima-Morimoto F, Taniguchi E and
Nishida E. (2003). J. Biol. Chem., 278, 25277–25280.
Nasmyth K, Peters JM and Uhlmann F. (2000). Science, 288,
1379–1385.
Neef R, Preisinger C, Sutcliffe J, Kopajtich R, Nigg EA,
Mayer TU and Barr FA. (2003). J. Cell Biol., 162, 863–875.
Nitta M, Kobayashi O, Honda S, Hirota T, Kuninaka S,
Marumoto T, Ushio Y and Saya H. (2004). Oncogene, 23,
6548–6558.
Ouyang B, Li W, Pan H, Meadows J, Hoffmann I and Dai W.
(1999). Oncogene, 18, 6029–6036.
Pellicioli A, Lee SE, Lucca C, Foiani M and Haber JE. (2001).
Mol. Cell, 7, 293–300.
Peters JM. (2002). Mol. Cell, 9, 931–943.
Piaggio G, Farina A, Perrotti D, Manni I, Fuschi P, Sacchi A
and Gaetano C. (1995). Exp. Cell Res., 216, 396–402.
Pines J and Hunter T. (1989). Cell, 58, 833–846.
Pines J and Hunter T. (1991). J. Cell Biol., 115, 1–17.
Pines J and Hunter T. (1994). EMBO J., 13, 3772–3781.
Poon RY, Yamashita K, Adamczewski JP, Hunt T and
Shuttleworth J. (1993). EMBO J., 12, 3123–3132.
Qian YW, Erikson E, Li C and Maller JL. (1998a). Mol. Cell
Biol., 18, 4262–4271.
Qian YW, Erikson E and Maller JL. (1998b). Science, 282,
1701–1704.
Qian YW, Erikson E and Maller JL. (1999). Mol. Cell Biol.,
19, 8625–8632.
Qian YW, Erikson E, Taieb FE and Maller JL. (2001). Mol.
Biol. Cell, 12, 1791–1799.
Reimann JD, Freed E, Hsu JY, Kramer ER, Peters JM and
Jackson PK. (2001a). Cell, 105, 645–655.
Reimann JD, Gardner BE, Margottin-Goguet F and Jackson
PK. (2001b). Genes Dev., 15, 3278–3285.
Ruan Q, Wang Q, Xie S, Fang Y, Darzynkiewicz Z, Guan K,
Jhanwar-Uniyal M and Dai W. (2004). Exp. Cell Res., 294,
51–59.
Rudner AD and Murray AW. (2000). J. Cell Biol., 149,
1377–1390.
Sadhu K, Reed SI, Richardson H and Russell P. (1990). Proc.
Natl. Acad. Sci. USA, 87, 5139–5143.
Scolnick DM and Halazonetis TD. (2000). Nature, 406,
430–435.
Seong YS, Kamijo K, Lee JS, Fernandez E, Kuriyama R, Miki
T and Lee KS. (2002). J. Biol. Chem., 277, 32282–32293.
Shiloh Y. (2003). Nat. Rev. Cancer, 3, 155–168.
Shirayama M, Zachariae W, Ciosk R and Nasmyth K. (1998).
EMBO J., 17, 1336–1349.
Simizu S and Osada H. (2000). Nat. Cell Biol., 2, 852–854.
Smith MR, Wilson ML, Hamanaka R, Chase D, Kung H,
Longo DL and Ferris DK. (1997). Biochem. Biophys. Res.
Commun., 234, 397–405.
Smits VA, Klompmaker R, Arnaud L, Rijksen G, Nigg EA
and Medema RH. (2000). Nat. Cell Biol., 2, 672–676.
Solomon MJ, Harper JW and Shuttleworth J. (1993). EMBO
J., 12, 3133–3142.
Regulation of mitotic progression by Plk-1
MATM van Vugt and RH Medema
2859
Spankuch B, Matthess Y, Knecht R, Zimmer B, Kaufmann M
and Strebhardt K. (2004). J. Natl. Cancer Inst., 96, 862–872.
Spankuch-Schmitt B, Bereiter-Hahn J, Kaufmann M and
Strebhardt K. (2002). J. Natl. Cancer Inst., 94, 1863–1877.
Sumara I, Vorlaufer E, Stukenberg PT, Kelm O, Redemann N,
Nigg EA and Peters JM. (2002). Mol. Cell, 9, 515–525.
Sunkel CE and Glover DM. (1988). J. Cell Sci., 89 (Part 1),
25–38.
Sutterlin C, Lin CY, Feng Y, Ferris DK, Erikson RL
and Malhotra V. (2001). Proc. Natl. Acad. Sci. USA, 98,
9128–9132.
Takekawa M, Adachi M, Nakahata A, Nakayama I, Itoh F,
Tsukuda H, Taya Y and Imai K. (2000). EMBO J., 19,
6517–6526.
Takizawa CG, Weis K and Morgan DO. (1999). Proc. Natl.
Acad. Sci. USA, 96, 7938–7943.
Toczyski DP, Galgoczy DJ and Hartwell LH. (1997). Cell, 90,
1097–1106.
Toyoshima-Morimoto F, Taniguchi E and Nishida E. (2002).
EMBO Rep., 3, 341–348.
Toyoshima-Morimoto F, Taniguchi E, Shinya N, Iwamatsu A
and Nishida E. (2001). Nature, 410, 215–220.
Tsvetkov L, Xu X, Li J and Stern DF. (2003). J. Biol. Chem.,
278, 8468–8475.
Uchiumi T, Longo DL and Ferris DK. (1997). J. Biol. Chem.,
272, 9166–9174.
Uhlmann F, Lottspeich F and Nasmyth K. (1999). Nature,
400, 37–42.
Uhlmann F, Wernic D, Poupart MA, Koonin EV and
Nasmyth K. (2000). Cell, 103, 375–386.
van Vugt MA, Bras A and Medema RH. (2004a). Mol. Cell,
15, 799–811.
van Vugt MA, Smits VA, Klompmaker R and Medema RH.
(2001). J. Biol. Chem., 276, 41656–41660.
van Vugt MA, van De Weerdt BC, Vader G, Janssen H,
Calafat J, Klompmaker R, Wolthuis RM and Medema RH.
(2004b). J. Biol. Chem., 279, 36841–36854.
Visintin R, Prinz S and Amon A. (1997). Science, 278,
460–463.
Waizenegger IC, Hauf S, Meinke A and Peters JM. (2000).
Cell, 103, 399–410.
Walter SA, Cutler Jr RE, Martinez R, Gishizky M and Hill
RJ. (2003). J. Biol. Chem., 278, 18221–18228.
Warnke S, Kemmler S, Hames RS, Tsai HL, HoffmannRohrer U, Fry AM and Hoffmann I. (2004). Curr. Biol., 14,
1200–1207.
Watanabe N, Arai H, Nishihara Y, Taniguchi M, Hunter
T and Osada H. (2004). Proc. Natl. Acad. Sci. USA, 101,
4419–4424.
Weichert W, Denkert C, Schmidt M, Gekeler V, Wolf G,
Kobel M, Dietel M and Hauptmann S. (2004a). Br. J.
Cancer, 90, 815–821.
Weichert W, Schmidt M, Gekeler V, Denkert C, Stephan C,
Jung K, Loening S, Dietel M and Kristiansen G. (2004b).
Prostate, 60, 240–245.
Xie S, Wu H, Wang Q, Cogswell JP, Husain I, Conn C,
Stambrook P, Jhanwar-Uniyal M and Dai W. (2001).
J. Biol. Chem., 276, 43305–43312.
Yamada SI, Ohira M, Horie H, Ando K, Takayasu H, Suzuki
Y, Sugano S, Hirata T, Goto T, Matsunaga T, Hiyama E,
Hayashi Y, Ando H, Suita S, Kaneko M, Sasaki F,
Hashizume K, Ohnuma N and Nakagawara A. (2004).
Oncogene, 23, 5901–5911.
Yang J, Bardes ES, Moore JD, Brennan J, Powers MA and
Kornbluth S. (1998). Genes Dev., 12, 2131–2143.
Yarm FR. (2002). Mol. Cell. Biol., 22, 6209–6221.
Yoo HY, Kumagai A, Shevchenko A and Dunphy WG.
(2004). Cell, 117, 575–588.
Yu H. (2002). Curr. Opin. Cell Biol., 14, 706–714.
Yuan J, Eckerdt F, Bereiter-Hahn J, Kurunci-Csacsko E,
Kaufmann M and Strebhardt K. (2002a). Oncogene, 21,
8282–8292.
Yuan J, Kramer A, Eckerdt F, Kaufmann M and Strebhardt
K. (2002b). Cancer Res., 62, 4186–4190.
Zachariae W and Nasmyth K. (1999). Genes Dev., 13,
2039–2058.
Zhou T, Aumais JP, Liu X, Yu-Lee LY and Erikson RL.
(2003). Dev. Cell, 5, 127–138.
Oncogene