REVIEWS
Finishing mitosis, one step at a time
Matt Sullivan and David O. Morgan
Abstract | The final stages of mitosis begin in anaphase, when the mitotic spindle segregates
the duplicated chromosomes. Mitotic exit is then completed by disassembly of the spindle
and packaging of chromosomes into daughter nuclei. The successful completion of mitosis
requires that these events occur in a strict order. Two main mechanisms govern progression
through late mitosis: dephosphorylation of cyclin-dependent kinase (Cdk) substrates and
destruction of the substrates of the anaphase-promoting complex (APC). Here, we discuss
the hypothesis that the order of late mitotic events depends, at least in part, on the order in
which different Cdk and APC substrates are dephosphorylated or destroyed, respectively.
Departments of Physiology
and Biochemistry &
Biophysics, University of
California, 600 16th Street,
San Francisco, California
94158‑2517, USA.
Correspondence to D.O.M.
e-mail:
[email protected]
doi:10.1038/nrm2276
Published online 3 October 2007
A skyscraper is built in a strictly ordered series of steps:
the foundation is laid, the frame goes up, and only after
that are the walls added and the furniture moved in. The
same is true for cell division: chromosome duplication
comes first, after which the duplicated chromosomes
are pulled apart and then packaged in individual cells.
In both these cases, the ordering of events depends on
central regulatory systems that initiate each event at
the appropriate time, based in part on predetermined
schedules. These schedules can be adjusted if necessary:
when problems delay the completion of an event (such
as chromosome duplication), subsequent events (such as
mitosis) can be postponed.
Work over the past 20 years has unveiled the basic
molecular workings of the control system that orders
and coordinates the events of the eukaryotic cell-division
cycle1. The central components of this system are the
cyclin-dependent kinases (Cdks), the activities of which
oscillate during the cell cycle in response to changes in
their association with regulatory cyclin subunits. Distinct
cyclin–Cdk complexes form at specific cell-cycle stages
and initiate the events of the S and M phases. Mitotic
cyclin–Cdk complexes drive the particularly striking
events of early mitosis: chromosome condensation and
resolution, nuclear envelope breakdown, and assembly
of the mitotic spindle. Cdks have completed their main
functions by metaphase, when all sister-chromatid pairs
are bi-orientated on the spindle, pulled towards the poles
but held together by sister-chromatid cohesion.
Cdk activity drives cell-cycle progression as far as
metaphase, but progression into anaphase and beyond
depends on another major regulatory component — a
ubiquitin-protein ligase called the anaphase-promoting
complex (APC) or cyclosome, which ubiquitinates several regulatory proteins and thereby targets them to the
proteasome for destruction2,3. APC activity oscillates in
894 | november 2007 | volume 8
response to changes in the association of the APC with
the activating subunits Cdc20 or Cdh1: association of the
APC with Cdc20 in mid-mitosis leads to the initiation of
anaphase, whereas association with Cdh1 in late mitosis
maintains APC activity throughout the subsequent G1.
A key target of the APC is securin, the destruction of
which initiates chromosome segregation by unleashing
the protease separase, which destroys sister-chromatid
cohesion. The APC also ubiquitinates the mitotic cyclins,
the destruction of which inactivates Cdks and allows
phosphatases to dephosphorylate the many Cdk substrates in the cell. Dephosphorylation of Cdk substrates
is required for normal chromosome and spindle movements in anaphase, as well as for the subsequent events of
telophase: spindle disassembly, reformation of nuclei and
decondensation of chromatin.
The final stages of mitosis are therefore governed by
two main regulatory mechanisms: dephosphorylation
of Cdk substrates and ubiquitination of APC substrates
(FIG. 1). In this Review, we describe our present knowledge
of the highly conserved molecular circuitry that underlies these mechanisms and discuss the emerging concept
that the correct ordering of late mitotic events depends,
at least in part, on the order in which Cdk substrates are
dephosphorylated and APC targets are destroyed.
Cyclin oscillations drive mitosis
Cdk activation in mitosis depends on two or three classes
of mitotic cyclins, each defined by its function and the
timing of its expression. In animal cells, the first major
mitotic cyclin to be expressed is cyclin A, which appears
at the onset of S phase, when it is thought to contribute
to the stimulation of DNA synthesis. Cyclin A levels
remain high after S phase, and in early mitosis it drives
the initiation of chromosome condensation and, possibly,
nuclear envelope breakdown4,5. Cyclin A is destroyed
www.nature.com/reviews/molcellbio
© 2007 Nature Publishing Group
REVIEWS
Arbitrary units
Cdk activity
APC substrate
Ub Ub Ub
Cdk substrate
P P P
APC activity
G2
Metaphase
Early
mitosis
Anaphase
G1
Mitotic
exit
Figure 1 | Control of late mitotic events. Progression through mitosis is shown by the
Reviews
| Molecular
Cell Biology
cells along the bottom of the figure (chromosomes inNature
blue and
spindle
microtubules
in
red). The transition from metaphase to anaphase is triggered by an increase in the activity
of the anaphase-promoting complex (APC) (green line; top of figure), a ubiquitin-protein
ligase that promotes the assembly of chains of ubiquitin (Ub) on its substrates, thereby
targeting them for destruction in the proteasome. The main APC targets are securin, the
destruction of which leads to sister-chromatid separation, and cyclins, the destruction of
which results in a drop in Cdk activity (blue line). Cdk inactivation allows cellular
phosphatases to dephosphorylate Cdk substrates during late mitosis. Cdk-substrate
dephosphorylation is required for the events of anaphase and telophase (not shown).
Hydrophobic patch
A short stretch of amino acids
on the surface of some cyclins
near the Cdk active site. It
interacts with RXL motifs on
cyclin-specific Cdk substrates
or inhibitors.
RXL motif
A degenerate sequence motif
on some Cdk substrates and
inhibitors. It interacts with the
hydrophobic patch region of
specific cyclins.
during prometaphase6. Next in line is cyclin B, the levels
of which rise during G2. Cyclin B promotes the completion of chromosome condensation and spindle assembly,
thereby driving cell-cycle progression as far as metaphase.
Cyclin B (together with securin) is destroyed during meta­
phase, significantly later than cyclin A7. Finally, a third
mitotic cyclin, cyclin B3, has been identified in some
animal species, although its role in mitosis is unclear8,9.
On the basis of studies in Drosophila melanogaster
embryos, cyclin B3 is thought to be destroyed in anaphase,
later than cyclin B. The three mitotic cyclins are therefore
destroyed in a clear sequence: cyclin A, cyclin B and then
cyclin B3 (FIG. 2a).
A series of distinct cyclins also governs mitosis in budding yeast, although in this species there is considerable
overlap in the functions of the different cyclins10. The
major S‑phase cyclin, Clb5, exhibits features that are reminiscent of vertebrate cyclin A: its expression increases in
late G1, it helps to stimulate chromosome duplication and
it appears to have some functions during mitosis. Unlike
cyclin A, Clb5 is destroyed just before anaphase, together
with securin. The major mitotic cyclin, Clb2, is crucial
for spindle assembly and progression to metaphase. Some
Clb2 protein is destroyed at the same time as Clb5 and
securin, but most remains stable until after anaphase.
Cyclin destruction can order late mitotic events. What
function is served by destroying different cyclins at different stages of mitosis? An appealing model is that each
class of cyclin is responsible for driving the phosphorylation of a subset of Cdk substrates, and that each subset is
dephosphorylated when its respective cyclin is destroyed.
nature reviews | molecular cell biology
Specific targets of cyclin A–Cdk, for example, would be
expected to be dephosphorylated before anaphase, earlier
than the targets of cyclin B–Cdk. Given that Cdk-substrate
dephosphorylation drives late mitotic events, it follows
that the ordered destruction of different cyclins could help
to order the events of late mitosis.
A key requirement of this model is that different
cyclins must have different intrinsic functional capacities.
Abundant evidence supports this notion10. For example,
when individual cyclins are mutated to render them
resistant to APC-mediated destruction, the resulting
cellular phenotype generally varies depending on which
cyclin has been mutated. A particularly striking example
is found in embryos of D. melanogaster: stabilization
of cyclin A causes an arrest in metaphase, stabilization of
cyclin B causes an arrest in anaphase and abnormal
chromo­some movements, and stabilization of cyclin B3
allows normal anaphase but causes defects in later
events8,11 (FIG. 2b). The effects of stabilized cyclins vary
in different cell types and species, but the clear implication from these types of studies is that different cyclins
are destroyed at different times to promote the correct
sequence of late mitotic events.
A mechanism that probably underlies the specialization of cyclin function is that cyclins restrict the substrate
specificity or the localization of the associated Cdks10.
Biochemical studies with mammalian cyclins indicate that
several Cdk substrates are phosphorylated rapidly in vitro
by cyclin A–Cdk1 but not by cyclin B–Cdk1 (Ref. 12), and
cyclin specificity in these cases depends on a docking site
in cyclin A (called the hydrophobic patch) that interacts
with a small motif (called the RXL (or Cy) motif) that is
found on its targets13. Similarly, several Cdk1 substrates
in budding yeast are much more rapidly phosphorylated
by Clb5–Cdk1 than by Clb2–Cdk1, again as a result of an
interaction between the hydrophobic patch in Clb5 and
a docking motif in specific substrates14. As might be predicted, many of these Clb5-specific substrates are involved
in chromosome duplication, although some have functions in the mitotic spindle. As we discuss below, dephosphorylation of spindle proteins helps to govern spindle
behaviour during anaphase.
Phosphatases order dephosphorylation
The timing of Cdk-substrate dephosphorylation is not
determined simply by the timing of kinase inactivation:
phosphatases must remove phosphates from Cdk targets,
and substrate specificity could have a considerable impact
on the order in which Cdk substrates are dephosphory­
lated. Analysis of phosphatases in late mitosis has lagged
far behind that of the kinases, but progress is beginning
to be made in this area.
Numerous general phosphatases exist in the cell, and
the basal activity of some of these might help to drive Cdksubstrate dephosphorylation when specific cyclin–
Cdk complexes are inactivated. There is also evidence
that robust dephosphorylation is a result of the activation
of specific phosphatases during late mitosis. The classic
example is the budding yeast phosphatase Cdc14, which is
activated transiently during late mitosis and is required for
the dephosphorylation of several Cdk substrates15,16.
volume 8 | november 2007 | 895
© 2007 Nature Publishing Group
REVIEWS
a
Protein level
Cyclin A
Cyclin B
Cyclin B3
G2
b Stable cyclin A
Prophase
Prometaphase
Metaphase
Stable cyclin B
Anaphase
Telophase
G1
Stable cyclin B3
Figure 2 | The order of mitotic cyclin destruction. a | The three major mitotic cyclins of animal cells are destroyed at
different times during mitosis. Cyclin A is degraded soon after nuclear envelope breakdown,
whereas
cyclin B degradation
Nature Reviews
| Molecular
Cell Biology
(and that of securin) begins immediately after the last sister-chromatid pair is bi-orientated on the spindle at the
beginning of metaphase. An additional mitotic cyclin, cyclin B3, has been studied primarily in Drosophila melanogaster
embryos, in which it is destroyed after cyclin B. Key mitotic transitions are indicated by bold dashed lines.
b | Distinct phenotypes result when different mitotic cyclins are stabilized during mitosis. Stabilized forms of each cyclin
were constructed by mutating the sequences that are required for recognition by the anaphase-promoting complex.
Expression of the indicated cyclin was induced during G2 of cycle 14 in D. melanogaster embryos, resulting in mitotic
arrests (DNA is shown in green and microtubules are shown in red). Stable cyclin A caused an arrest during metaphase,
stable cyclin B caused an arrest during anaphase and abnormal chromosome movements, and stable cyclin B3 caused an
arrest during late anaphase. Panel b is reproduced with permission from Ref. 11  (2001) Elsevier.
Stepwise activation of Cdc14. Before anaphase, Cdc14
is found in the nucleolus, where it is held in an in­active
state by its association with the protein Net1 (also
known as Cfi1). Beginning in early anaphase, Cdc14
is activated by dissociation from Net1 and by release
from the nucleolus17,18. Cdc14 then diffuses throughout
the nucleus and out into the cytoplasm to dephosphory­
late targets in both locations. Activation and release of
Cdc14 from the nucleolus depends on two regulatory
mechanisms (FIG. 3). First, Cdc14 is activated by separase,
the protease that also triggers sister-chromatid sep­
aration. Separase initiates Cdc14 activation by a poorly
understood process that depends on several other proteins, together referred to as the Cdc14 early anaphase
release (FEAR) network19,20. Second, complete Cdc14
activation depends on the mitotic exit network (MEN),
a signalling system that includes a small GTPase and a
pair of protein kinases that are activated during late
mitosis21.
The two Cdc14 activation mechanisms seem to
act in sequence to trigger the dephosphorylation of
distinct populations of Cdk substrates and, thus, distinct mitotic events (FIG. 3). The FEAR network initiates
Cdc14 activ­ation and triggers various processes that
are important for successful anaphase, including the
896 | november 2007 | volume 8
segregation of ribosomal DNA22,23, positioning of the
anaphase nucleus24 and stabilization of the anaphase
spindle25. These events are likely to depend on dephosphorylation during early anaphase of a limited subset of
Cdk targets, including the yeast INCENP homologue
Sli15 (Ref. 26) and the spindle regulators Ase1, Ask1
and Fin1 (Refs 25,27,28). The FEAR network alone,
however, is not sufficient for the dephosphorylation
of all Cdc14 targets. In the absence of MEN activity,
FEAR-dependent Cdc14 activity rises in early anaphase
and an apparently normal anaphase occurs, but Cdc14
activity declines prematurely and the mutant cells arrest
at the end of anaphase19. MEN activation is required for
complete Cdc14 activation and for the dephosphorylation of numerous Cdk substrates, including the APC
activator Cdh1, the transcription factor Swi5 and the
Cdk inhibitor Sic1 (all of which help to drive the completion of Cdk inactivation in late mitosis, as described
below). So, the two steps in Cdc14 activation might provide a mechanism for the ordered dephosphorylation
of different Cdk substrates and, therefore, the ordered
execution of distinct events.
Cdc14 homologues have been identified in higher
eukaryotes, but they do not seem to have the same
central function in late mitosis as in budding yeast29.
www.nature.com/reviews/molcellbio
© 2007 Nature Publishing Group
REVIEWS
MEN
Bfa1
Cdc20
Separase
GTP
Tem1
Bub2
Cdc5
APC
Securin
GDP
Tem1
Lte1
Sister-chromatid separation
Separase
Cdc15
FEAR
Cdc55
PP2A
Mob1
Dbf2
P P P
Net1
Cdc14
Net1
Inactive
Clb2
Cdk1
+
Cdc14
Cdc14
Partly active
Fully active
Dephosphorylation of
early anaphase targets
Chromosome movements,
spindle stability,
rDNA segregation
Dephosphorylation of
late anaphase targets
Clb2–Cdk1 inactivation,
completion of mitosis
Figure 3 | Control of Cdc14 activation by separase and the MEN. During metaphase, Cdc14 is held inactive in the
nucleolus by an interaction with its inhibitor Net1 (also known as Cfi1). The initial release
andReviews
activation
of Cdc14
Nature
| Molecular
Cellduring
Biology
early anaphase is the result of an abrupt increase in Net1 phosphorylation, which is carried out by a complex of the mitotic
cyclin Clb2 and the cyclin-dependent kinase, Cdk1 (Clb2–Cdk1)80. Recent studies20 suggest that Net1 phosphorylation is
removed before anaphase by the phosphatase PP2ACdc55. By unknown mechanisms, separase and the Cdc14 early
anaphase release (FEAR) network reduce PP2ACdc55 activity on Net1, which tilts the balance of activities towards Net1
phosphorylation, thereby initiating Cdc14 activation. Complete Cdc14 activation during late mitosis requires the mitotic
exit network (MEN), the central component of which is the GTPase Tem1 (Ref. 21). Tem1 activation might be promoted by
the guanine-nucleotide exchange factor Lte1 and is opposed by the GTPase-activating protein complex Bfa1–Bub2. Tem1
activates a protein kinase, Cdc15, which in turn activates a kinase complex, Dbf2–Mob1, which stimulates Cdc14 activity.
Activation of the MEN depends on the Polo-like kinase Cdc5, which is activated during mitosis and promotes Tem1
activation by phosphorylating and thereby inhibiting Bfa1–Bub2 (Ref. 81). In addition, Cdc14 dephosphorylates and
thereby activates the kinase Cdc15 (Ref. 82). It is not known how MEN activation promotes Cdc14 function. It is unlikely
that the MEN simply increases the level of Cdc14 activity. Instead, it might somehow give Cdc14 access to a broader range
of targets. For example, separase and the FEAR network might promote the release of Cdc14 within the nucleus, whereas
the MEN might allow Cdc14 access to targets in the cytoplasm.
In fission yeast and Caenorhabditis elegans, for example,
the Cdc14 homologue is not required for mitotic exit,
although it is crucial for cytokinesis. Furthermore, in these
and other species Cdc14 might also help to govern early
mitosis, in part through the dephosphorylation of Cdk
regulators. Therefore, much remains to be learned about
the phosphatases that govern the phosphorylation states
of Cdk substrates during late mitosis in animal cells.
Order of Cdk-substrate dephosphorylation
Despite the copious hints that late mitotic events depend
on the ordered dephosphorylation of Cdk targets, we still
know remarkably little about the identities of these targets.
Several potential Cdk substrates have been identified, but
few have been linked to a specific mitotic process and
even fewer have been studied in sufficient detail to assess
the precise timing of their dephosphorylation in relation
to other substrates. We discuss a few of the best-known
examples in this section.
nature reviews | molecular cell biology
Separase is dephosphorylated before anaphase. In vertebrate cells, separase is a well-established Cdk substrate
that is dephosphorylated and thereby activated before
anaphase (FIG. 4). Cdks phosphorylate separase in early
mitosis and thereby inhibit its protease activity, indicating
that separase dephosphorylation (in addition to securin
destruction) must occur before anaphase can begin30.
Inhibition of separase by cyclin A–Cdk may explain the
observation that the expression of stabilized cyclin A
mutants in flies (FIG. 2b) and in some vertebrate cells
blocks the onset of anaphase6,8,11,31. However, separase is
not strictly a cyclin A target that is dephosphorylated only
when cyclin A is degraded. In vertebrates, cyclin B–Cdk1
also phosphorylates separase (and inhibits it by binding to it)32, and a stabilized cyclin B mutant prevents
anaphase in some mammalian cell types7,33,34. In these
cases at least, separase is probably dephosphorylated at
the end of metaphase, after cyclin B–Cdk1 is inactivated.
The phosphatase that acts on separase is not known.
volume 8 | november 2007 | 897
© 2007 Nature Publishing Group
Phosphorylation level
REVIEWS
destruction and separase-dependent Cdc14 activation27,28.
Thus, in the case of these proteins at least, it is clear that
the timing of cyclin destruction helps to determine the
timing of substrate dephosphorylation and function.
Pre-anaphase substrates
(vertebrate separase)
Early-anaphase substrates
(INCENP, Fin1, Ase1/PRC1)
Late-anaphase substrates
(Cdh1, Sic1, spindle proteins)
Prometaphase
Metaphase
Completion of
sister-chromatid
attachment
Anaphase
Sister-chromatid
separation
Telophase
Initiation of
spindle disassembly,
cytokinesis
Figure 4 | The order of Cdk substrate dephosphorylation during mitosis. The
Nature
Reviews | Molecular
Cell
Biology
curves represent the phosphorylation levels of various
cyclin-dependent
kinase
(Cdk)
substrates, which are dephosphorylated at different times as a result of differences in the
timing of cyclin destruction (FIG. 2) and the activation of phosphatases such as Cdc14 in
budding yeast (FIG. 3). Some substrates, of which vertebrate separase is presently the sole
known example, must be dephosphorylated during late metaphase, before anaphase can
begin. Several other substrates (such as INCENP, Fin1 and Ase1 (known as PRC1 in
mammals)) are dephosphorylated at about this time or during early anaphase; these
proteins help to govern chromosome and spindle behaviour during anaphase. A large
number of substrates are dephosphorylated later during anaphase: these include
regulatory proteins such as Cdh1 and Sic1, the dephosphorylation of which completes
Cdk inactivation, and spindle proteins, the dephosphorylation of which leads to spindle
disassembly. Other major telophase events, such as chromosome decondensation,
nuclear envelope formation and the initiation of cytokinesis, also depend on
Cdk-substrate dephosphorylation during late anaphase or thereafter. The metaphaseto-anaphase transition is indicated with a bold dashed line.
Kinetochore
A large protein structure that
assembles on the chromosome
and mediates the attachment
of the chromosome to
microtubules of the mitotic
spindle.
Spindle midzone
The region at the equator of
the mitotic spindle where
interpolar microtubules
overlap. During anaphase, this
region helps to organize
proteins that govern anaphase
spindle behaviour and that
control the initiation of
cytokinesis.
Dephosphorylation of Cdk substrates in early anaphase.
Dephosphorylation of Cdk substrates has been implicated
in anaphase chromosome movement and spindle stability
(FIG. 4). Expression of stabilized cyclin B in some animal cell
types, including the fly embryo, results in abnormal chromo­
some movements and spindle structure35,36. In budding
yeast, anaphase chromosome and spindle behaviours are
defective when sister-chromatid segregation is artificially
induced under conditions of high Cdk activity and low
Cdc14 activity25. One important Cdk substrate in anaphase is INCENP (Sli15 in yeast), the regulatory subunit of
the kinase Aurora B (Ipl1 in yeast). Dephosphorylation
of INCENP in early anaphase triggers its translocation
from kinetochores to the spindle midzone, where it promotes
spindle stability26. The dephosphorylation of several additional yeast Cdk substrates has also been implicated in the
control of anaphase spindle behaviour. These substrates
include: Ask1, a kinetochore protein that contributes to
normal anaphase chromosome movements and spindle
function25; Fin1, which binds and stabilizes the anaphase
spindle27; and Ase1, which helps to organize the spindle
midzone during anaphase28. PRC1, the human homologue
of Ase1, is also a spindle midzone protein that is activated
by dephosphorylation during anaphase37.
The yeast proteins Fin1 and Ase1 are specific targets of
Clb5–Cdk1 (Ref. 14), and their dephosphorylation during
early anaphase is driven by the combined effects of Clb5
898 | november 2007 | volume 8
Dephosphorylation of Cdk substrates in late anaphase.
Three key components of the budding yeast cell-cycle
control system — Cdh1, Sic1 and Swi5 — are Cdk1 substrates that must be dephosphorylated during anaphase
for late mitotic progression and for entry into G1. Cdh1
is an APC activator, and its phosphorylation by Cdks
restrains its activity from late G1 until anaphase, when it
is dephosphorylated and thereby activated38,39. APCCdh1 is
required for the complete destruction of Clb2, and so the
inactivation of Clb2–Cdk1 is delayed until late anaphase
or thereafter. Sic1 has a similar regulatory function to that
of Cdh1: it is a tight-binding inhibitor of all Clb–Cdk
complexes and its phosphorylation between late G1 and
mitosis triggers its destruction40,41. Its dephosphorylation
during late mitosis allows Sic1 to accumulate and contribute to Clb2–Cdk1 inactivation42. Swi5 is a transcription
factor that drives the expression of SIC1 and other genes.
It is inhibited by Cdk1-dependent phosphorylation and is
activated by dephosphorylation during late mitosis42,43.
Cdh1, Sic1 and Swi5 are all thought to be dephosphorylated during late anaphase as a result of Clb2 des­
truction and MEN-dependent Cdc14 activation (FIGS 3,4).
Phosphorylated Cdh1 and Swi5 are located in the cytoplasm and are then imported into the nucleus upon
dephosphorylation43,44; therefore, their activation might
depend on the export of Cdc14 from the nucleus. The full
activation and nuclear export of Cdc14 could depend on the
MEN, which is activated later than the FEAR-dependent
Cdc14 activation of early anaphase. Thus, the local­
ization and stepwise activation of Cdc14 might provide
a mechanism to introduce a delay between the dephosphorylation of nuclear targets (such as those involved
in spindle stabilization, as mentioned above) and that
of cytoplasmic targets (such as Cdh1, Swi5 and proteins
involved in later events, including cytokinesis45).
Other dephosphorylation events in late mitosis. The
identities of Cdk substrates involved in the major events
of late anaphase and telophase are not yet clear. Spindle
assembly probably depends on the phosphorylation of
many microtubule regulators and motor proteins, and
it is reasonable to expect — although it has not been
proven — that dephosphorylation of these same proteins
is required for spindle disassembly during telophase.
Similar mechanisms probably apply to the proteins that
are phosphorylated to drive chromosome condensation
and to the proteins of the nuclear pore and lamina that
trigger nuclear envelope breakdown — the dephosphorylation of these same proteins is likely to be required to
reverse these events.
How the dephosphorylation of such proteins is delayed
until after anaphase also remains unclear. In budding
yeast, in which a subpopulation of the mitotic cyclin Clb2
is stable until the end of anaphase, a likely answer is that
the proteins involved in telophase events are Clb2 targets
that remain phosphorylated until Clb2 has been destroyed.
www.nature.com/reviews/molcellbio
© 2007 Nature Publishing Group
REVIEWS
Box 1 | APC enzymology
The anaphase-promoting complex (APC) is a multisubunit
E3 ubiquitin-protein ligase of the RING-domain family72,73.
Cdc20
Target
or
Cdc27
Like other members of this family, it binds two substrates:
Ub
Cdh1
the protein target (such as securin or cyclin) and an E2
E2
NH2
ubiquitin-conjugating enzyme that is covalently linked to
Cdc16
the C terminus of a 76-residue protein called ubiquitin
(Ub; see figure). The APC stimulates transfer of the
Cdc23
Doc1
Apc11
C terminus of ubiquitin from the E2–ubiquitin conjugate
to a Lys side chain on the target substrate. In subsequent
Apc5
reactions, it can catalyse the transfer of ubiquitin to a Lys
Apc4
Apc2
in another molecule of ubiquitin, resulting in the
formation of polyubiquitin chains74,75. Long chains that are
Apc1
linked at Lys48 of ubiquitin are efficiently recognized by
the 26S proteasome, resulting in target destruction76.
Clusters of short chains might also drive proteasomedependent destruction74.
Multiubiquitination by the APC is processive, so multiple Lys residues can be modified during a single substrate-binding
Nature Reviews | Molecular Cell Biology
event66,77. Processivity probably depends on differences in the interaction affinities of the enzyme with its two substrates.
Thus, the APC might bind a protein target relatively tightly (that is, with a low dissociation rate) while binding an E2 with
relatively low affinity. Therefore, the protein target would remain bound to the APC as a series of E2–ubiquitin conjugates
bind, unload their ubiquitin on the target and dissociate to make way for the next E2–ubiquitin77,78.
The APC contains 12 or 13 subunits2,3. In yeast, nine subunits are essential for normal activity and are thought to interact
as shown in the figure79. The Apc11 subunit contains the RING domain and probably binds the E2–ubiquitin conjugate.
Protein targets are recruited by the activator subunits Cdc20 or Cdh1, which interact with specific APC core subunits.
Substrates may also interact directly with core subunits such as Doc1 (Refs 50,51).
26S proteasome
A large protease complex that
binds polyubiquitinated
proteins and degrades them.
Destruction box
(D-box). A degenerate
sequence motif (RXXLXXXXN)
that is in most APC targets. The
D‑box mediates an interaction
with APC activator subunits
and is required for target
destruction.
KEN box
A degenerate sequence motif
(KENXXXN) that is in some APC
targets. The KEN box mediates
an interaction with the APC
activator Cdh1.
Polo-like kinase‑1
(Plk1). A protein kinase that is
activated during early mitosis
and that helps to promote
certain mitotic events, such as
spindle assembly.
Aurora kinases A and B
Related protein kinases that
are activated during early
mitosis and govern spindle
assembly, chromosome
attachment to kinetochores
and other mitotic processes.
In animal cells, however, the major mitotic cyclin, cyclin B,
is destroyed in parallel with securin during metaphase.
If dephosphorylation of cyclin B targets promotes telophase events, then why are these targets not dephosphoryl­
ated until long after cyclin B is gone? The answer might
lie in the activity and specificity of mitotic phosphatases,
which remain largely unexplored in animal cells.
Order of APC-substrate destruction
The sequence of Cdk-substrate dephosphorylation
depends, in part, on the order in which cyclins are
destroyed. What, then, determines the order in which
cyclins and other APC targets are destroyed? Multiple
mechanisms are involved, but a central issue is the intrinsic ability of different isoforms of the APC to recognize
and ubiquitinate different targets. Here, we discuss some
of the mechanisms by which APC-substrate specificity
can change at different stages of mitosis.
Like other ubiquitin-protein ligases, the APC catalyses
the transfer of ubiquitin from an E2–ubiquitin conjugate to
a Lys side chain on a protein target (BOX 1). Ubiquitin itself
can also be modified, resulting in polyubiquitin chains that
direct the target to the 26S proteasome for destruction. The
APC reaction is processive, such that multiple ubiquitins
are attached before the substrate dissociates.
The efficiency and processivity of the APC reaction are
determined, at least in part, by the affinity of protein-target
binding to the APC. For example, a decrease in the affinity of a substrate for the active site of the APC probably
reduces processivity — and thereby reduces the length of
polyubiquitin chains on the protein and, hence, its rate
of destruction. Currently, there is only a rudi­mentary
understanding of how the APC interacts with its substrates. APC targets contain short amino-acid sequence
nature reviews | molecular cell biology
motifs, called the destruction (D‑) box and KEN box, that are
required for target destruction during mitosis46,47. These
motifs interact directly with the activator subunits Cdc20
or Cdh1 (Refs 48,49), which are then thought to recruit
the bound substrate to the APC for ubiquitination. There
is also evidence that the APC itself contains binding sites
for the D‑box, although these sites are not understood in
any detail50,51.
APC activators order substrate degradation. The two APC
activators, Cdc20 and Cdh1, bind directly to APC substrates and recruit them to the APC core for ubiquitination. The APC is activated first by Cdc20 before anaphase
and then by Cdh1 during late anaphase. Interestingly,
the two activators have different substrate specificities:
Cdc20 binds a limited range of targets (primarily securin
and cyclins), whereas Cdh1 has a broader specificity that
includes Cdc20 targets and various additional proteins
that are not recognized by Cdc20 (including Cdc20 itself,
Polo-like kinase‑1 (Plk1), Aurora kinases A and B and, in
yeast, the spindle proteins Ase1 and Fin1)2,3. Thus, the
ubiquitination of APCCdc20 targets occurs earlier than that
of APCCdh1 targets, thereby providing a clear mechanism of
APC-substrate ordering (FIG. 5).
Differential substrate targeting by APCCdc20 and
APCCdh1 is presumably achieved by differential recognition of APC substrates by each activator. Cdh1-specific
targets often contain a KEN-box motif that is not recognized by Cdc20 (Ref. 47). The main recognition motif, the
D‑box, is found in most APC substrates, although many
of these substrates are recognized only by APCCdh1 and
not by APCCdc20. It seems likely that subtle differences
in D‑box sequences, or sequence context, can influence substrate recognition by different APC isoforms.
volume 8 | november 2007 | 899
© 2007 Nature Publishing Group
Protein level
REVIEWS
SAC
Cdc20
Cdc20
Cdh1
APC
APC
APC
Cyclin A
Nek2A
HOXC10
Prophase
Cyclin B
Securin
Prometaphase
Nuclear
envelope
breakdown
Cdc20
Metaphase
Completion of
sister-chromatid
attachment
Plk1
Anaphase
Sister-chromatid
separation
Aurora A
Aurora B
Telophase
G1
Initiation of
spindle disassembly,
cytokinesis
Figure 5 | Three windows of APC-dependent destruction in human cells. Three different anaphase-promoting
Reviews
| MolecularofCell
Biology
complex (APC) isoforms predominate during different stages of mitosis and promote Nature
the ordered
destruction
three
different groups of proteins. The APC is first activated during early prometaphase by its activator subunit Cdc20. This form
of the APC is initially inhibited by components of the spindle-assembly checkpoint (SAC), which interact with Cdc20 and
block its ability to recruit substrates to the APC. By unknown mechanisms, SAC-inhibited APC retains the ability to
promote the ubiquitination of several prometaphase targets, including cyclin A, the kinase Nek2A and the transcription
factor HOXC10. When the last sister-chromatid pair is attached to the spindle and the SAC is inactivated, fully activated
APCCdc20 then targets an additional group of proteins — including cyclin B and securin — for destruction during
metaphase. Cyclin destruction leads to cyclin-dependent kinase (Cdk) inactivation, which leads to the dephosphorylation
and activation of the second APC activator, Cdh1. APCCdh1 has broader substrate specificity than APCCdc20 and triggers the
destruction of various additional targets. These targets are destroyed at different times during late mitosis, presumably
providing mechanisms that help to order late mitotic events. APCCdh1 targets include Cdc20, Plk1 and the Aurora kinases,
which are destroyed in the order shown in the figure64. There may also be substrate ordering for the substrates of APCCdc20.
In Drosophila embryos, for example, cyclin B and cyclin B3 are APCCdc20 targets that are destroyed at different times, as
shown in FIG. 2. Key mitotic transitions are indicated with bold dashed lines.
Adding to this complexity is the fact that several APC
targets contain destruction sequences that bear little
resemblance to either the KEN box or the D‑box52–56.
Budding yeast display an interesting case of APCsubstrate ordering that is not easily explained. In this
species, APCCdc20 triggers complete destruction of the
cyclin Clb5 before anaphase, but the levels of the major
mitotic cyclin Clb2 are only partly reduced57,58. APCCdh1
activation later during anaphase is required to drive the
destruction of the remaining Clb2. A subpopulation
of Clb2 might be protected from APCCdc20 by extrinsic
factors such as localization59.
Spindle-assembly
checkpoint
A regulatory system that
monitors chromosome
attachment to the mitotic
spindle and delays APC
activation until all
chromosomes are correctly bi-orientated.
Spindle-assembly checkpoint orders substrate degradation.
In budding yeast, the major Cdc20-specific targets (securin
and Clb5) are degraded together before anaphase. In animal cells, however, some Cdc20-specific targets (cyclin A,
the kinase Nek2A and the transcription factor HOXC10)
are degraded during prometaphase6,31,60,61, whereas
others (securin and cyclin B) are not degraded until meta­
phase7. This ordering of APCCdc20 targets in animal cells
depends on a regulatory system called the spindle-assembly checkpoint (SAC), which prevents the destruction of
900 | november 2007 | volume 8
securin and cyclin B until all sister-chromatid pairs are
correctly attached to the spindle62.
APCCdc20 is first activated during prometaphase, which
leads to the ubiquitination of prometaphase targets such
as cyclin A. However, the unattached sister chromatids
that exist in prometaphase cells generate signals that
allow inhibitory SAC components, such as Mad2, to bind
Cdc20 and block its ability to interact with substrates. By
some mysterious mechanism, Mad2 and other SAC proteins prevent the ubiquitination of some APCCdc20 targets
(securin and cyclin B) while allowing the ubiquitination
of others (cyclin A and Nek2A) (FIG. 5).
Prometaphase targets may be recognized by the APC
through a unique mechanism that does not require binding to Cdc20 (Ref. 60). The prometaphase target Nek2A
has a dipeptide sequence (Met-Arg, MR) at its C terminus
that resembles a sequence (Ile-Arg, IR) that is found at the
C termini of the APC activators Cdc20 and Cdh1. The
IR sequence motif is required for association of the activ­
ators with specific subunits of the APC63. The MR motif
of Nek2A appears to interact directly with the APC core
as well, apparently bypassing the need for an activator to
recruit the substrate. However, Cdc20 is still required for
www.nature.com/reviews/molcellbio
© 2007 Nature Publishing Group
REVIEWS
Nek2A ubiquitination, implying that Cdc20 can activate
the APC independently of its substrate-recruiting function, perhaps by inducing some conformational change
in the APC core. Such a mechanism could help to explain
why the SAC does not prevent the ubiquitination of
prometaphase targets such as Nek2A. That is, the SACinhibited form of Cdc20 may not be able to bind substrates
but may still be able to drive an activating conformational
change in the APC, which allows the ubiquitination of
proteins that are recruited by activator-independent
mechanisms. Cyclin A does not have an IR‑like motif
at its C terminus, but may contain some other sequence
motif that allows it to interact with the APC independently of Cdc20. There is evidence that the APC can bind
D‑box sequences independently of activators50,51, and so
it is possible that the D‑box sequences of prometaphase
targets have a particularly high affinity for binding sites
on the APC core.
APCCdh1 targets are degraded at different times. The activ­
ation of APCCdh1 in anaphase does not lead to synchronous destruction of all of its targets. In human cells, for
example, different APCCdh1 substrates disappear at different times: Cdc20 is degraded first, followed by Plk1, then
Aurora A and finally Aurora B64,65 (FIG. 5). The ordering of
APCCdh1 targets probably depends on two general classes
of mechanism.
First, APCCdh1 might have different intrinsic activities
towards different substrates, such that some substrates
are more rapidly and processively modified than others,
leading to their earlier degradation. There are considerable differences in the intrinsic activity of APCCdh1 towards
different substrates. Recent studies, for example, reveal that
Cdc20, Plk1 and Aurora A are multiubiquitinated in vitro
at decreasing rates and processivities that correlate with
the order of their disappearance in the cell66. The ordering of substrate destruction is therefore likely to depend,
in part, on differences in the affinity of substrates for the
APC. In a sea of multiple different substrates, comp­etition
for binding to the APC would initially result in more
interactions with substrates that have the highest affinities. In addition, given that the affinity of APC-substrate
binding helps to determine the processivity of multiubiquitination (BOX 1), high-affinity substrates will be modified with longer polyubiquitin chains. Upon release from
the APC, substrates that are more extensively modified
have a greater chance of retaining a sufficiently long chain
for recognition by the proteasome, even in the face of
abundant de-ubiquitinating activities in the cell.
What factors underlie these differences between
APCCdh1 substrates? One of the key factors is likely to
be the relative affinity of destruction sequences on the
substrate for binding sites on Cdh1 or the APC core.
In addition, the position and density of available Lys
residues on a substrate could affect the pattern of ubiquitination and, therefore, the likelihood of destruction
by the proteasome.
A second mechanism for ordering APCCdh1 substrates
depends on extrinsic factors that modulate the recognition of some substrates, thereby changing the timing of
their degradation. Changes in the phosphorylation state
nature reviews | molecular cell biology
of substrates, for example, could be crucial in some cases.
The phosphorylation of the replication protein Cdc6
by cyclin E–Cdk2 protects it against APC-dependent
ubiquitination, providing a mechanism that allows Cdc6
accumulation and thus enables origin licensing when cells
enter the cycle from quiescence67. Another potentially
important extrinsic factor is the subcellular localization
of the APC and its substrates. The APC is found on the
spindle, kinetochores and centrosomes in mitosis, and
APC substrates at these sites may be more rapidly ubiquitinated and destroyed in the cell. In Drosophila embryos,
for example, the destruction of cyclin B during late mitosis
begins around the centrosomes and spindle68.
We do not yet understand the importance of ordered
APCCdh1-substrate destruction during late mitosis. In
budding yeast, only the destruction of securin and cyclins is essential for mitotic progression69, and Cdh1 is not
required for viability. Destruction of other substrates,
including most Cdh1-specific targets, is not absolutely
essential but might enhance the robustness of chromosome segregation, telophase events and cytokinesis. The
spindle-stabilizing proteins Ase1 and Fin1, for example,
are destroyed at the end of anaphase, presumably contributing to spindle disassembly; however, non-degradable
mutant forms of these proteins have only minor effects
on spindle behaviour27,70. Similarly, the destruction of
Cdh1-specific targets such as the Polo-like kinase Cdc5
of yeast or Plk1 of vertebrates might not be essential but
might make important contributions to the control of the
substrates of these kinases.
The three windows of APC-substrate degradation. In
summary, APC-substrate degradation in animal cells
occurs in a series of three mitotic windows (FIG. 5). First,
during prometaphase, substrates such as cyclin A are
ubiquitinated by APCCdc20 through mechanisms that may
not require activator-dependent substrate recruitment.
Second, the complete attachment of chromosomes to the
spindle turns off the SAC, triggering APCCdc20-dependent
ubiquitination of metaphase targets such as securin and
cyclin B. Third, activation of APCCdh1 during anaphase
allows continued destruction of metaphase targets and
also promotes the ubiquitination of Cdh1-specific targets.
Different Cdh1 targets are ubiquitinated at different times
owing to intrinsic differences in their affinity for the APC
or owing to other factors, such as substrate phosphorylation and localization, that influence their interactions with
the APC.
Future directions
The hypothesis that the destruction of each class of cyclin
leads to dephosphorylation of the specific targets of that
cyclin is appealing but simplistic, and future models must
incorporate the quantitative complexities of kinase–
substrate interactions in the cell. First, each cyclin–Cdk
complex is likely to possess different intrinsic activities
towards its different substrates, primarily as a result of
different affinities for those substrates. Indeed, studies
of large numbers of yeast Cdk substrates in vitro reveal
a remarkably broad range of activities14,71. Second, it is
important to consider the concentrations of substrates
volume 8 | november 2007 | 901
© 2007 Nature Publishing Group
REVIEWS
in the cell, which are probably sufficiently high for most
cyclin–Cdk complexes to be saturated with substrate.
Under these conditions, destruction of a specific cyclin to
1% of its normal level might have more dramatic effects
on some substrates than others: low-affinity substrates
might be dephosphorylated rapidly, whereas phosphorylation of high-affinity substrates could be maintained
until the kinase is completely inactivated. Third, there
is the issue of the stoichiometry of phosphorylation and
its functional impact. What fraction of the population of
a specific substrate must be dephosphorylated to trigger a
specific late mitotic event? This fraction is likely to vary in
different cases: it is conceivable that dephosphorylation
of 1% of some substrate populations is sufficient to initiate certain processes, whereas other substrate populations
may have to be completely dephosphorylated.
Similarly, we cannot hope for a deep understanding of
phosphatase function during late mitosis without more
quantitative analyses of phosphatase–substrate inter­
actions. Cdc14, for example, has a preference in vitro for
Cdk phosphosites (that is, a phosphoserine followed by
a Pro residue), but studies with various Cdk substrates
in vitro reveal a broad range of activities (L. Holt and
D.O.M., unpublished observations), and some Cdk targets
do not appear to be targets of Cdc14. Thus, some ordering of Cdk-substrate dephosphorylation may result from
intrinsic differences in the activity of Cdc14 towards its
various substrates, not to mention the activities of other
phosphatases that have not yet been discovered.
Studies of the importance of APC-substrate degradation in vivo have often focused on the phenotype that
results if a substrate is not degraded during mitosis. Future
work should begin to address the related but distinct question of what happens when the order of APC-substrate
destruction is altered. What is the effect on mitosis, for
example, if cyclin B is degraded during prometaphase
with Nek2A, if cyclin A is degraded late at anaphase
onset, or if Aurora A is degraded before Cdc20? Once
we have a better understanding of the sequence motifs
that determine the timing of substrate destruction, it may
become possible to address these questions by swapping
destruction motifs between substrates that are degraded
at early and late stages.
The importance of the timing of Cdk-substrate
dephosphorylation is not as tractable as that of APCsubstrate destruction because it is not easy to create
1.
2.
3.
4.
5.
6.
7.
Morgan, D. O. The Cell Cycle: Principles of Control
(New Science Press, London, 2007).
Peters, J. M. The anaphase promoting complex/
cyclosome: a machine designed to destroy. Nature Rev.
Mol. Cell Biol. 7, 644–656 (2006).
Thornton, B. R. & Toczyski, D. P. Precise destruction: an
emerging picture of the APC. Genes Dev. 20,
3069–3078 (2006).
Furuno, N., den Elzen, N. & Pines, J. Human cyclin A is
required for mitosis until mid prophase. J. Cell Biol.
147, 295–306 (1999).
Gong, D. et al. Cyclin A2 regulates nuclear-envelope
breakdown and the nuclear accumulation of cyclin B1.
Curr. Biol. 17, 85–91 (2007).
den Elzen, N. & Pines, J. Cyclin A is destroyed in
prometaphase and can delay chromosome alignment
and anaphase. J. Cell Biol. 153, 121–136 (2001).
Hagting, A. et al. Human securin proteolysis is
mutant substrates that cannot be dephosphorylated.
Substitution of phosphorylation sites with acidic residues sometimes provides clues about the consequences
of continued phosphorylation, but additional methods
will be required to allow systematic analyses of many
substrates. Perhaps a better understanding of the phosphatases that govern Cdk substrates will lead to experimental approaches for constructing substrates that are
resistant to dephosphorylation.
Conclusions
The onset of anaphase and the completion of mitosis
are governed by dephosphorylation of Cdk substrates
and ubiquitination of APC substrates. A growing body
of evidence suggests that the substrates of Cdks and the
APC are modified at different times during mitosis, and
the clear implication of this evidence is that the ordering
of Cdk and APC substrates helps to establish the correct
order of late mitotic events. A great deal of further effort
is required, however, to prove that substrate ordering is
important. We need more detailed information about
the exact time at which Cdk substrates are dephosphorylated or APC targets are destroyed. Similarly, a
complete understanding of the regulation of these substrates requires quantitative analyses of the activities and
affinities of the Cdks and the APC for all of their targets.
A complete picture of this regulatory system must also
include the identities and activities of the phosphatases
and de-ubiquitinating enzymes that oppose Cdk and
APC functions in the cell. Only with this information in
hand will it be possible to thoroughly assess the importance of intrinsic enzyme–substrate interactions in the
timing of substrate modifications.
The cell is guided through the steps of mitosis by an
immensely complex regulatory system that is assembled
from multiple Cdk and APC activities. Studies of this
regulatory system have been directed primarily towards
the mechanisms that generate oscillations in the activity of
these enzymes, on the assumption that these oscillations
provide an accurate reflection of changes in the phosphorylation or ubiquitination states of Cdk or APC substrates,
respectively. A complete understanding of substrate modifications and functions, however, will come only from a
more thorough knowledge of the specific interactions
between cell-cycle regulatory enzymes and all of their
major substrates.
controlled by the spindle checkpoint and reveals when
the APC/C switches from activation by Cdc20 to Cdh1.
J. Cell Biol. 157, 1125–1137 (2002).
8. Sigrist, S., Jacobs, J., Stratmann, R. & Lehner, C. F. Exit
from mitosis is regulated by Drosophila fizzy and the
sequential destruction of cyclins A, B, and B3. EMBO J.
14, 4827–4838 (1995).
9. Jacobs, H. W., Knoblich, J. A. & Lehner, C. F. Drosophila
cyclin B3 is required for female fertility and is
dispensable for mitosis like cyclin B. Genes Dev. 12,
3741–3751 (1998).
10. Bloom, J. & Cross, F. R. Multiple levels of cyclin
specificity in cell-cycle control. Nature Rev. Mol. Cell
Biol. 8, 149–160 (2007).
11. Parry, D. H. & O’Farrell, P. H. The schedule of
destruction of three mitotic cyclins can dictate the
timing of events during exit from mitosis. Curr. Biol. 11,
671–683 (2001).
902 | november 2007 | volume 8
12.
13.
14.
15.
16.
Evidence that the timing of destruction of distinct
cyclins by the APC helps to order late mitotic events.
Peeper, D. S. et al. A‑ and B‑type cyclins differentially
modulate substrate specificity of cyclin-CDK complexes.
EMBO J. 12, 1947–1954 (1993).
Brown, N. R., Noble, M. E., Endicott, J. A. &
Johnson, L. N. The structural basis for specificity of
substrate and recruitment peptides for cyclin-dependent
kinases. Nature Cell Biol. 1, 438–443 (1999).
Loog, M. & Morgan, D. O. Cyclin specificity in the
phosphorylation of cyclin-dependent kinase substrates.
Nature 434, 104–108 (2005).
Stegmeier, F. & Amon, A. Closing mitosis: the functions
of the Cdc14 phosphatase and its regulation. Annu.
Rev. Genet. 38, 203–232 (2004).
D’Amours, D. & Amon, A. At the interface between
signaling and executing anaphase — Cdc14 and the
FEAR network. Genes Dev. 18, 2581–2595 (2004).
www.nature.com/reviews/molcellbio
© 2007 Nature Publishing Group
REVIEWS
17. Shou, W. et al. Exit from mitosis is triggered by Tem1dependent release of the protein phosphatase Cdc14
from nucleolar RENT complex. Cell 97, 233–244
(1999).
18. Visintin, R., Hwang, E. S. & Amon, A. Cfi1 prevents
premature exit from mitosis by anchoring Cdc14
phosphatase in the nucleolus. Nature 398, 818–823
(1999).
19. Stegmeier, F., Visintin, R. & Amon, A. Separase, polo
kinase, the kinetochore protein Slk19, and Spo12
function in a network that controls Cdc14 localization
during early anaphase. Cell 108, 207–220 (2002).
20. Queralt, E., Lehane, C., Novak, B. & Uhlmann, F.
Downregulation of PP2ACdc55 phosphatase by separase
initiates mitotic exit in budding yeast.
Cell 125, 719–732 (2006).
Evidence that separase promotes Cdc14 activation
during early anaphase by blocking the actions of
the phosphatase PP2A.
21. Jaspersen, S. L., Charles, J. F., Tinker-Kulberg, R. L. &
Morgan, D. O. A late mitotic regulatory network
controlling cyclin destruction in Saccharomyces
cerevisiae. Mol. Biol. Cell 9, 2803–2817 (1998).
22. D’Amours, D., Stegmeier, F. & Amon, A. Cdc14 and
condensin control the dissolution of cohesinindependent chromosome linkages at repeated DNA.
Cell 117, 455–469 (2004).
23. Sullivan, M., Higuchi, T., Katis, V. L. & Uhlmann, F.
Cdc14 phosphatase induces rDNA condensation and
resolves cohesin-independent cohesion during
budding yeast anaphase. Cell 117, 471–482 (2004).
24. Ross, K. E. & Cohen-Fix, O. A role for the FEAR
pathway in nuclear positioning during anaphase.
Dev. Cell 6, 729–735 (2004).
25. Higuchi, T. & Uhlmann, F. Stabilization of microtubule
dynamics at anaphase onset promotes chromosome
segregation. Nature 433, 171–176 (2005).
Demonstration that Cdc14 activation during early
anaphase promotes normal anaphase spindle and
chromosome behaviours.
26. Pereira, G. & Schiebel, E. Separase regulates INCENPAurora B anaphase spindle function through Cdc14.
Science 302, 2120–2124 (2003).
27. Woodbury, E. L. & Morgan, D. O. Cdk and APC
activities limit the spindle-stabilizing function of
Fin1 to anaphase. Nature Cell Biol. 9, 106–112
(2007).
28. Khmelinskii, A., Lawrence, C., Roostalu, J. & Schiebel, E.
Cdc14-regulated midzone assembly controls
anaphase B. J. Cell Biol. 177, 981–993 (2007).
29. Trautmann, S. & McCollum, D. Cell cycle: new
functions for Cdc14 family phosphatases. Curr. Biol.
12,R733–R735 (2002).
30. Stemmann, O., Zou, H., Gerber, S. A., Gygi, S. P. &
Kirschner, M. W. Dual inhibition of sister chromatid
separation at metaphase. Cell 107, 715–726 (2001).
31. Geley, S. et al. Anaphase-promoting complex/
cyclosome-dependent proteolysis of human cyclin A
starts at the beginning of mitosis and is not subject to
the spindle assembly checkpoint. J. Cell Biol. 153,
137–148 (2001).
32. Gorr, I. H., Boos, D. & Stemmann, O. Mutual inhibition
of separase and Cdk1 by two-step complex formation.
Mol. Cell 19, 135–141 (2005).
33. Chang, D. C., Xu, N. & Luo, K. Q. Degradation of
cyclin B is required for the onset of anaphase in
mammalian cells. J. Biol. Chem. 278, 37865–37873
(2003).
34. Herbert, M. et al. Homologue disjunction in mouse
oocytes requires proteolysis of securin and cyclin B1.
Nature Cell Biol. 5, 1023–1025 (2003).
35. Parry, D. H., Hickson, G. R. & O’Farrell, P. H. Cyclin B
destruction triggers changes in kinetochore behavior
essential for successful anaphase. Curr. Biol. 13,
647–653 (2003).
36. Wheatley, S. P. et al. CDK1 inactivation regulates
anaphase spindle dynamics and cytokinesis in vivo.
J. Cell Biol. 138, 385–393 (1997).
37. Zhu, C., Lau, E., Schwarzenbacher, R., Bossy-Wetzel, E.
& Jiang, W. Spatiotemporal control of spindle midzone
formation by PRC1 in human cells. Proc. Natl Acad.
Sci. USA 103, 6196–6201 (2006).
38. Jaspersen, S. L., Charles, J. F. & Morgan, D. O.
Inhibitory phosphorylation of the APC regulator Hct1
is controlled by the kinase Cdc28 and the
phosphatase Cdc14. Curr. Biol. 9, 227–236
(1999).
39. Zachariae, W., Schwab, M., Nasmyth, K. & Seufert, W.
Control of cyclin ubiquitination by CDK-regulated
binding of Hct1 to the anaphase promoting complex.
Science 282, 1721–1724 (1998).
40. Nash, P. et al. Multisite phosphorylation of a CDK
inhibitor sets a threshold for the onset of DNA
replication. Nature 414, 514–521 (2001).
41. Verma, R. et al. Phosphorylation of Sic1p by G1 Cdk
required for its degradation and entry into S phase.
Science 278, 455–460 (1997).
42. Visintin, R. et al. The phosphatase Cdc14 triggers
mitotic exit by reversal of Cdk-dependent
phosphorylation. Mol. Cell 2, 709–718 (1998).
43. Moll, T., Tebb, G., Surana, U., Robitsch, H. &
Nasmyth, K. The role of phosphorylation and the
CDC28 protein kinase in the cell cycle-regulated
nuclear import of the S. cerevisiae transcription factor
SWI5. Cell 66, 743–758 (1991).
44. Jaquenoud, M., van Drogen, F. & Peter, M. Cell cycledependent nuclear export of Cdh1p may contribute to
the inactivation of APC/CCdh1. EMBO J. 21,
6515–6526 (2002).
45. Bembenek, J. et al. Crm1-mediated nuclear export of
Cdc14 is required for the completion of cytokinesis in
budding yeast. Cell Cycle 4, 961–971 (2005).
46. King, R. W., Glotzer, M. & Kirschner, M. W. Mutagenic
analysis of the destruction signal of mitotic cyclins and
structural characterization of ubiquitinated
intermediates. Mol. Biol. Cell 7, 1343–1357
(1996).
47. Pfleger, C. M. & Kirschner, M. W. The KEN box: an
APC recognition signal distinct from the D box
targeted by Cdh1. Genes Dev. 14, 655–665
(2000).
48. Burton, J. L., Tsakraklides, V. & Solomon, M. J.
Assembly of an APC–Cdh1–substrate complex is
stimulated by engagement of a destruction box.
Mol. Cell 18, 533–542 (2005).
49. Kraft, C., Vodermaier, H. C., Maurer-Stroh, S.,
Eisenhaber, F. & Peters, J. M. The WD40 propeller
domain of Cdh1 functions as a destruction box
receptor for APC/C substrates. Mol. Cell 18, 543–553
(2005).
50. Yamano, H., Gannon, J., Mahbubani, H. & Hunt, T.
Cell cycle-regulated recognition of the destruction box
of cyclin B by the APC/C in Xenopus egg extracts.
Mol. Cell 13, 137–147 (2004).
51. Carroll, C. W., Enquist-Newman, M. & Morgan, D. O.
The APC subunit Doc1 promotes recognition of the
substrate destruction box. Curr. Biol. 15, 11–18
(2005).
52. Hildebrandt, E. R. & Hoyt, M. A. Cell cycle-dependent
degradation of the Saccharomyces cerevisiae spindle
motor Cin8p requires APCCdh1 and a bipartite
destruction sequence. Mol. Biol. Cell 12, 3402–3416
(2001).
53. Castro, A., Vigneron, S., Bernis, C., Labbe, J. C. &
Lorca, T. Xkid is degraded in a D‑box, KEN-box, and
A‑box‑independent pathway. Mol. Cell Biol. 23,
4126–4138 (2003).
54. Araki, M., Wharton, R. P., Tang, Z., Yu, H. & Asano, M.
Degradation of origin recognition complex large
subunit by the anaphase-promoting complex in
Drosophila. EMBO J. 22, 6115–6126 (2003).
55. Littlepage, L. E. & Ruderman, J. V. Identification of a
new APC/C recognition domain, the A box, which is
required for the Cdh1-dependent destruction of the
kinase Aurora‑A during mitotic exit. Genes Dev. 16,
2274–2285 (2002).
56. Sullivan, M. & Morgan, D. O. A novel destruction
sequence targets the meiotic regulator Spo13 for
anaphase-promoting complex-dependent degradation
in anaphase I. J. Biol. Chem. 282, 19710–19715
(2007).
57. Yeong, F. M., Lim, H. H., Padmashree, C. G. &
Surana, U. Exit from mitosis in budding yeast: biphasic
inactivation of the Cdc28–Clb2 mitotic kinase and the
role of Cdc20. Mol. Cell 5, 501–511 (2000).
58. Wäsch, R. & Cross, F. APC-dependent proteolysis of
the mitotic cyclin Clb2 is essential for mitotic exit.
Nature 418, 556–562 (2002).
59. Eluere, R. et al. Compartmentalization of the functions
and regulation of the mitotic cyclin Clb2 in
S. cerevisiae. J. Cell Sci. 120, 702–711 (2007).
60. Hayes, M. J. et al. Early mitotic degradation of Nek2A
depends on Cdc20-independent interaction with the
APC/C. Nature Cell Biol. 8, 607–614 (2006).
The authors suggest that certain APC substrates,
such as Nek2A, can bind to the APC independently
of Cdc20, triggering their destruction before
metaphase.
61. Gabellini, D. et al. Early mitotic degradation of the
homeoprotein HOXC10 is potentially linked to cell
cycle progression. EMBO J. 22, 3715–3724
(2003).
nature reviews | molecular cell biology
62. Musacchio, A. & Salmon, E. D. The spindle-assembly
checkpoint in space and time. Nature Rev. Mol. Cell
Biol. 8, 379–393 (2007).
63. Vodermaier, H. C., Gieffers, C., Maurer-Stroh, S.,
Eisenhaber, F. & Peters, J. M. TPR subunits of the
anaphase-promoting complex mediate binding to the
activator protein CDH1. Curr. Biol. 13, 1459–1468
(2003).
64. Lindon, C. & Pines, J. Ordered proteolysis in
anaphase inactivates Plk1 to contribute to proper
mitotic exit in human cells. J. Cell Biol. 164,
233–241 (2004).
Evidence for ordered proteolysis of APCCdh1
substrates during anaphase.
65. Pines, J. Mitosis: a matter of getting rid of the right
protein at the right time. Trends Cell Biol. 16, 55–63
(2006).
66. Rape, M., Reddy, S. K. & Kirschner, M. W. The
processivity of multiubiquitination by the APC
determines the order of substrate degradation.
Cell 124, 89–103 (2006).
Biochemical evidence that ordered APC-substrate
degradation is determined in part by the affinity of
the APC for its substrates.
67. Mailand, N. & Diffley, J. F. CDKs promote DNA
replication origin licensing in human cells by
protecting Cdc6 from APC/C-dependent proteolysis.
Cell 122, 915–926 (2005).
68. Huang, J. & Raff, J. The disappearance of cyclin B at
the end of mitosis is regulated spatially in Drosophila
cells. EMBO J. 18, 2184–2195 (1999).
69. Thornton, B. R. & Toczyski, D. P. Securin and
B‑cyclin/CDK are the only essential targets of the APC.
Nature Cell Biol. 5, 1090–1094 (2003).
70. Juang, Y.‑L. et al. APC-mediated proteolysis of Ase1
and the morphogenesis of the mitotic spindle. Science
275, 1311–1314 (1997).
71. Ubersax, J. A. et al. Targets of the cyclin-dependent
kinase Cdk1. Nature 425, 859–864 (2003).
72. Kerscher, O., Felberbaum, R. & Hochstrasser, M.
Modification of proteins by ubiquitin and ubiquitin-like
proteins. Annu. Rev. Cell Dev. Biol. 22, 159–180
(2006).
73. Pickart, C. M. & Eddins, M. J. Ubiquitin: structures,
functions, mechanisms. Biochim. Biophys. Acta 1695,
55–72 (2004).
74. Kirkpatrick, D. S. et al. Quantitative analysis of in vitro
ubiquitinated cyclin B1 reveals complex chain
topology. Nature Cell Biol. 8, 700–710 (2006).
75. Rodrigo-Brenni, M. & Morgan, D. O. Sequential E2s
drive polyubiquitin chain assembly on APC targets.
Cell 130, 127–139 (2007).
76. Thrower, J. S., Hoffman, L., Rechsteiner, M. &
Pickart, C. M. Recognition of the polyubiquitin
proteolytic signal. EMBO J. 19, 94–102 (2000).
77. Carroll, C. W. & Morgan, D. O. The Doc1 subunit is a
processivity factor for the anaphase-promoting
complex. Nature Cell Biol. 4, 880–887 (2002).
78. Eletr, Z. M., Huang, D. T., Duda, D. M., Schulman,
B. A. & Kuhlman, B. E2 conjugating enzymes must
disengage from their E1 enzymes before E3dependent ubiquitin and ubiquitin-like transfer.
Nature Struct. Mol. Biol. 12, 933–934 (2005).
79. Thornton, B. R. et al. An architectural map of the
anaphase-promoting complex. Genes Dev. 20,
449–460 (2006).
80. Azzam, R. et al. Phosphorylation by cyclin B–Cdk
underlies release of mitotic exit activator Cdc14
from the nucleolus. Science 305, 516–519
(2004).
Evidence that Cdk-dependent phosphorylation of
Net1 promotes the release of the phosphatase
Cdc14 during anaphase.
81. Hu, F. et al. Regulation of the Bub2/Bfa1 GAP
complex by Cdc5 and cell cycle checkpoints. Cell 107,
655–665 (2001).
82. Jaspersen, S. L. & Morgan, D. O. Cdc14 activates
Cdc15 to promote mitotic exit in budding yeast.
Curr. Biol. 10, 615–618 (2000).
DATABASES
UniProtKB: http://ca.expasy.org/sprot
Ase1 | Cdc14 | Cdc20 | Cdh1 | Cdk1 | Clb2 | Clb5 | Fin1 | Net1 |
securin | separase | Sic1 | Swi5
FURTHER INFORMATION
David O. Morgan’s homepage:
http://www.ucsf.edu/morgan
All links are active in the online pdf
volume 8 | november 2007 | 903
© 2007 Nature Publishing Group