You are on page 1of 13

Cell Cycle Control☆

Randy YC Poon, Hong Kong University of Science and Technology, Hong Kong, People’s Republic of China
ã 2015 Elsevier Inc. All rights reserved.

Introduction 1
Cyclin-Dependent Kinases as Master Regulators of the Cell Cycle 2
The foundation 2
Cyclins and CDKs 3
Regulation of CDK activity 4
Cyclins 4
Phosphorylation 5
CDK inhibitors 6
Control of G2–Mitosis 6
Control of Mitosis–G1 7
Control of G1–S 8
Control of S Phase 9
Checkpoints 11
Spindle-Assembly Checkpoint 11
G1 DNA Damage Checkpoint 11
G2 DNA Damage Checkpoint 11
Replication Checkpoint 12
Acknowledgements 12

Glossary Cyclins A family of proteins first identified as proteins that


Anaphase-promoting complex/cyclosome (APC/C) A oscillate in amount in synchrony with the cell cycle.
multiprotein complex that promotes the entry into Maturation-promoting factor Activity that drives M phase
anaphase by targeting proteins for ubiquitin-mediated of the cell cycle and that is composed of cyclin B and CDK1.
degradation. Also known as M phase-promoting factor.
Centromere Part of the chromosome that links sister Restriction point A point in late G1 phase of the cell cycle,
chromatids and for kinetochore to assemble. after passing which the cell is committed to another round
Centrosomes Microtubule organizing centers that play of the cell cycle. Known as START in yeast.
important roles in the establishment of the bipolar mitotic SCF complexes A family of ubiquitin ligase complexes
spindles. compose of SKP1, CUL1, and a F-box protein.
Checkpoints Mechanisms that delay the cell cycle to ensure Ubiquitin-mediated degradation A mechanism by which
that each stage of the cell cycle is completed before the next many proteins in the cell are tagged for selective proteolysis
stage is initiated. by the proteasome; first modified with covalent attachment
Cyclin-dependent kinases A family of serine/threonine of polyubiquitins.
protein kinases that require cyclin binding for activity.

Introduction

The cell cycle is the sequence of events through which a cell duplicates its genome, grows, and divides into two daughter cells. The
ability of a cell to duplicate is one of the most fundamental properties that define life. Conversely, most cancers are in essence
caused by dysregulation of the cell cycle.
The cell cycle is divided into four phases (Figure 1). After cell division, daughter cells undergo a period of growth (G1) when
cellular proteins, RNA, membranes, and other macromolecules are synthesized. G1 is followed by a period of DNA synthesis
(S) and then by another period of growth (G2). G2 is followed by mitosis (M), during which chromosome condensation, nuclear
envelope breakdown, formation of mitotic spindles, attachment of chromosomes to the mitotic spindles, and separation of sister
chromatids occur. The cell cycle is completed with separation of the two daughter cells (cytokinesis). The stages outside M phase are
collectively known as interphase.


Change History: April 2015. Randy Y.C. Poon rewrote the entire article and made all the Figures.

Reference Module in Biomedical Sciences http://dx.doi.org/10.1016/B978-0-12-801238-3.98748-8 1


2 Cell Cycle Control

Figure 1 The cell cycle. The cell cycle is divided into four phases: Gap 1 (G1), DNA synthesis (S), Gap 2 (G2), and mitosis (M). The cell cycle is
completed with separation of the two daughter cells (cytokinesis). Most nondividing mammalian cells exit the cell cycle at G1 into a quiescent state (G0).
The cyclin–CDK complexes involved in different periods of the cell cycle are shown.

The cell cycle can in essence be considered as a DNA cycle. A somatic cell contains one copy of DNA in G1 phase and
progressively increases to two copies during S phase. The DNA content then returns abruptly to one copy at the end of mitosis,
when the DNA is divided equally into two daughter cells. Most nondividing mammalian cells exit the cell cycle at G1 into a
quiescent state (G0). Alternatively, cells can exit the cell cycle during differentiation and senescence.
Since the 1970s, cell cycle research has moved from merely descriptive studies of events such as chromosome condensation and
spindle organization to a fairly comprehensive molecular understanding through a combination of biochemical and genetic
studies. These studies show that the mechanisms that control the cell cycle are conserved throughout evolution from yeast to
human. It has also become apparent that many oncogenes and tumor suppressor genes are intrinsic components of the cell cycle or
can influence its progression.
The cell cycle is driven by an evolutionarily conserved engine composed of a family of protein kinases called cyclin-dependent
kinases (CDKs). Although the orderly events of the cell cycle depend on many factors, the switching on and off of different CDKs to
drive different phases of the cell cycle remains a good approximation. The activities of CDKs are stringently regulated by protein–
protein interaction and phosphorylation. Ubiquitin-mediated proteolysis of CDK regulators and other proteins is also critical for
cell cycle transitions. On top of the cell cycle engine, surveillance mechanisms termed checkpoints monitor various events, halting
the cell cycle to ensure that each stage of the cell cycle is completed before the next stage is initiated. It was an important revelation
that like so many other processes in the cell, the cell cycle is regulated by phosphorylation and proteolysis.
This chapter summarizes the fundamental concepts of the cell cycle. While the basic mechanisms of cell cycle control are
conserved in all eukaryotic cells, here the emphasis is placed on the somatic cell cycle of mammalian cells.

Cyclin-Dependent Kinases as Master Regulators of the Cell Cycle


The foundation
Progress in the past 30 years has unraveled some of the underlying principles of cell cycle control in eukaryotic cells. Classic cell
fusion experiments with mammalian cells and microinjection experiments with amphibian oocytes demonstrated the existence of
an M phase dominant factor. Fusion of cells in G1, S or G2 to cells in M phase induced nuclear envelope breakdown and
chromosome condensation of the interphase nuclei, suggesting M phase is dominant over other cell cycle phases. When the
cytoplasm from frog eggs (naturally arrested in M phase) is microinjected into oocytes (naturally arrested in interphase), the
oocytes were induced to enter M phase. This activity, termed Maturation-Promoting Factor or M phase-Promoting Factor (MPF),
can be found in mitotic cells in all eukaryotes ranging from yeast to human, indicating that it represents a universal M phase
activator.
Cell Cycle Control 3

Subsequent biochemical purification of MPF revealed that it contains two subunits: a serine/threonine protein kinase CDK1
first identified in yeast and a regulatory subunit called cyclin first identified in murine invertebrates.
Cell Division Cycle (CDC) mutants were identified in the yeast Saccharomyces cerevisiae and Schizosaccharomyces pombe as
mutants that arrest cells at a unique stage of the cell cycle when shifted from the permissive temperature to restrictive temperature.
One of these CDC genes, the CDC28/cdc2 (in S. cerevisiae/S. pombe respectively) encodes a related 34 kDa serine/threonine protein
kinase that is essential for cells to progress through mitosis and a point in late G1 called START. START is similar to the restriction
point in higher eukaryotic cells: after passing START, the cell is committed to another round of cell cycle. Homologues of CDC28/
cdc2 have been found in other eukaryotes (now called CDK1) and can functionally complement the yeast CDC28/cdc2 mutations.
The importance of cell cycle control is underlined by the fact that its components have been conserved down to details in different
eukaryotes. Indeed, human CDK1 was isolated by looking for human cDNAs that could complement a S. pombe cdc2 temperature
sensitive mutation.
The other component of MPF, cyclin, was first discovered as a protein that varied in abundance during the embryonic cell cycles
of marine invertebrates. The level of cyclins oscillates in synchrony with the cell cycle, accumulates progressively throughout
interphase and disappears abruptly at the end of mitosis. Similar to MPF, injection of cyclin mRNA could drive Xenopus oocytes into
M phase.
These important early works established that the mechanisms controlling the cell cycle are conserved throughout evolution
from yeast to human. The presence of CDK1 in MPF accounts for the protein kinase activity associated with MPF. Moreover, the cell
cycle oscillation of the cyclin immediately suggests a mechanism for producing high MPF activity in mitosis and low MPF activity in
interphase (Figure 2) (see ‘Control of G2–mitosis’).

Cyclins and CDKs


Cell cycle regulation in multicellular eukaryotes are necessarily more complex than lower eukaryotes. In yeast, one CDK is involved
in the regulation of both G1-S (START) and G2-M transitions. In contrast, mammalian cells contain a family of CDKs that are
regulated by a family of cyclin subunits (Figure 1).
CDK1 form complexes with the mitotic cyclins (cyclin A and cyclin B) and drives cells into mitosis (see ‘Control of G2–mitosis’).
Destruction of cyclin B turns off CDK1 rapidly and allows the cell to exit mitosis (see ‘Control of mitosis–G1’).
The second member of the CDK family, CDK2, is an important regulator for both G1 phase and S phase. CDK2 associates
mainly with cyclin A and cyclin E. The levels of cyclin E–CDK2 and cyclin A–CDK2 peak during G1–S transition and in S phase
respectively. CDK3 is very similar in sequence to CDK2, but its precise function is not known.
CDK4 and CDK6 are partners of D-type cyclins (D1, D2, and D3) that function in G1 before cyclin E–CDK2. A major target of
cyclin D–CDK4/6 is the retinoblastoma gene product pRb. Hyperphosphorylation of pRb by cyclin D–CDK4/6 (and by cyclin
E–CDK2 and cyclin A–CDK2) releases pRb from E2F, allowing E2F to activate transcription of genes important for S phase (see
‘Control of G1–S’).
Several protein kinases are classified as CDKs simply because they share high sequence homology with other CDKs, especially
within the kinase domain. Whether all of them have cyclin partners remains to be elucidated. It should be noted that several cyclins

Figure 2 Regulation of M phase-promoting factor during the cell cycle. Activation of MPF drives G2 cells into mitosis. MPF is composed to cyclin
B and CDK1. While the expression of CDK1 is constant during the cell cycle, cyclin B accumulates slowly during G2 phase and is degraded abruptly at
the end of mitosis (upper panel). Due to inhibitory phosphorylation on CDK1, the activity of MPF is suppressed during G2 phase even though cyclin
B is accumulating. Dephosphorylation of CDK1 allows rapid activation of MPF, driving the cell into mitosis. MPF also activates the ubiquitin ligase
APC/CCDC20 (lower panel). However, the activation of APC/CCDC20 is suppressed by the spindle-assembly checkpoint. Once the checkpoint is satisfied,
APC/CCDC20 activity surges and degrades cyclin B. A consequence of MPF inactivation is the turning off of APC/CCDC20 and, at the same time,
removing the inhibitory signals on APC/CCDH1. The activity of APC/CCDH1 in G1 is important for degrading any remaining cyclin B, CDC20, and
several other mitotic regulators. To allow cyclin B to build up for the next mitosis, APC/CCDH1 is subsequently turned off by the accumulating cyclin
A–CDK complexes and EMI1.
4 Cell Cycle Control

and CDKs have non-cell cycle regulatory functions. CDK7, which is a component of CDK-activating kinase (CAK) (see ‘Regulation
of CDK activity’), is also a component of the basal transcription factor TFIIH and can phosphorylate the C-terminal repeat domain
(CTD) of the large subunit of RNA polymerase II, an event important for transcription initiation. Other cyclins and CDKs (cyclin
C–CDK8 and cyclin T/K–CDK9) also phosphorylate CTD. CDK8 and CDK19 are components of a multisubunit complex called the
Mediator, which plays pivotal roles in transcriptional regulation. Although CDK5 was identified as one of the CDK partners of
cyclin D in human normal fibroblasts, there is no indication that CDK5 has a function in normal cell cycle control. Instead CDK5 is
activated in postmitotic neurons by binding to a protein called p35, which shares no sequence homology to cyclin, and is essential
in neuronal development.

Regulation of CDK activity

The general mechanism of how CDKs are regulated is discussed here. How individual CDKs are specifically regulated during the cell
cycle is discussed in the sections on different cell cycle transitions.

Cyclins
By definition, the kinase activity of CDKs is dependent on binding to a cyclin subunit. All CDKs are related in sequence to the
archetypal member CDK1 (Figure 3). The levels of most CDKs are relatively constant in the cell cycle, but their activities are tightly
regulated due to the fluctuation of the abundance and activities of their cyclin partners and other regulators.
Cyclins are defined as proteins that are related in sequence to the originally isolated mitotic (A- and B-type) cyclins. The region
that shares the highest homology in the cyclin family is a 100 residues region known as the cyclin box. Following the cyclin box, a
region with low sequence conservation with the cyclin box can nevertheless fold into the same three-dimensional structure to form
a second cyclin box fold.
The crystal structure of an archetypal CDK (CDK2) shows that it has an overall bi-lobed structure typical of serine/threonine
protein kinases, but there are several deviations that explain why CDK2 is not active in the absence of cyclin. These include the fact
that the ATP in the catalytic core of monomeric CDK2 is incorrectly aligned for phospho-transfer. A large structural reorganization

Figure 3 The cyclin-dependent kinase family. Members of the human CDK family are involved in both cell cycle and non-cell cycle functions. The cyclin
partners and functions of many CDKs remain incompletely understood.
Cell Cycle Control 5

occurs in CDK2 following cyclin binding, with CDK2 adopting a structure more closely resembling to that of other active protein
kinases and the ATP now becomes correctly aligned for phospho-transfer.
The oscillation of cyclin during the cell cycle is regulated at the level of transcription and protein degradation. For example, the
mRNAs of the mitotic cyclins behave similarly as the proteins, accumulating during G2 phase and diminishing after mitosis.
Transcription is stimulated and repressed by several transcription factors, including B-MYB, E2F, FOXM1, and NF-Y. Elements in
the promoter, including CCRE/CDE and CHR, are in part responsible for the cell cycle oscillation of transcription.
Compare to the mRNAs, the mitotic cyclin proteins disappear more abruptly at the end of mitosis. Their instability is conferred
by a short sequence at the N-terminal region known as the destruction box (D-box). The D-box targets the mitotic cyclins to a
multisubunit ubiquitin ligase called anaphase-promoting complex/cyclosome (APC/C). A targeting subunit of APC/C, CDC20,
participates in targeting D-box proteins for ubiquitination. The ubiquitinated cyclins are then rapidly degraded by a constitutively
active proteasome complex. Importantly, APC/CCDC20 is activated by cyclin B–CDK1 itself, thereby providing a mechanism for
cyclin B to prime for its own destruction.
Other cyclins such as the G1 cyclins do not contain the D-box and their degradation involves another class of ubiquitin ligase
called the SCF complex. Phosphorylation of cyclin E and cyclin D is important for their SCF-dependent ubiquitination. While
cyclin E can be phosphorylated by its partner CDK2, cyclin D is phosphorylated by GSK3b, thereby linking cyclin D turnover to
mitogen signal transduction pathways (see ‘Control of G1–S’).
In addition to synthesis and proteolysis, cyclins are also regulated by subcellular localization. For example, cyclin B1 is present
in the cytoplasm during G2 and translocates into the nucleus at prophase. Nuclear import of cyclin B1 involves binding of cyclin B1
to importin-b. The cytoplasmic localization of cyclin B1 during G2 is mediated by the binding of the nuclear exporting sequence
(NES) in cyclin B1 to the export mediator CRM1. Phosphorylation of serine residues in the NES (by kinases including CDK1 and
PLK1) is important for its nuclear translocation, possibly through the disruption of CRM1–cyclin B1 interaction. This mechanism
allows the localization of cyclin B1–CDK1 to the nucleus when the complexes are active.

Phosphorylation
Given that the kinase activity of CDK depends on binding to a cyclin subunit, one question is how the progressive synthesis of
cyclin elicits an abrupt activation of some CDKs (Figure 2). It is clear that apart from binding to cyclin, other regulators are involved
in controlling the activity of CDK.
Among the CDKs, the regulation of CDK1 is the best characterized. Monomeric CDK1 is inactive and unphosphorylated.
On binding to cyclin, CDK1 is phosphorylated at three sites: Thr14, Tyr15, and Thr161. As Thr161 is located on an activating
T-loop that blocks the catalytic site, its phosphorylation is required for the full activation of CDK1. By contrast, Thr14 and Tyr15 are
located near the catalytic core and their phosphorylation inhibits CDK1 activity. During G2, cyclin B–CDK1 complexes are held in
the Thr161- and Thr14/Tyr15-phosphorylated inactive state until a threshold of the complexes is reached. The Thr14 and Tyr15
residues are then dephosphorylated to facilitate an abrupt activation of MPF.
CDK1Thr161 is phosphorylated by the CDK-activating kinase (CAK) and dephosphorylated by the CDK-interacting phosphatase
KAP. In higher eukaryotes, CAK itself is composed of a cyclin–CDK pair (cyclin H–CDK7) with a third subunit called MAT1. The
activity of CAK is constant in the cell cycle; and KAP only dephosphorylates Thr161 when cyclin is degraded. Hence CDK1Thr161
phosphorylation mainly relies on whether cyclin–CDK complexes are form.
CDK1Thr14 and CDK1Tyr15 can be phosphorylated by the protein kinases MYT1 and WEE1 respectively. WEE1 is a dual-
specificity kinase that phosphorylates Tyr15 (but not Thr14). MYT1, a kinase that normally bound to the endoplasmic reticulum
and Golgi complex, can phosphorylate both the Thr14 and Tyr15, but has a stronger preference for Thr14.
Both Thr14 and Tyr15 can be dephosphorylated by members of the CDC25 phosphatase family (A, B and C). Significantly,
active CDK1 can inactivate WEE1 and activate more CDC25 by directly phosphorylating these proteins or indirectly phosphory-
lating upstream components. Hence, a small amount of active cyclin B–CDK1 can lead to a rapid and complete activation of all
cyclin B–CDK1 by this autocatalytic loop (Figure 4).

Figure 4 Activation of CDK1 by feedback controls. After cyclin B binds CDK1, the complex is kept in an inactive state by WEE1/MYT1-dependent
phosphorylation of CDK1Thr14/Tyr15. Dephosphorylated of CDK1Thr14/Tyr15 by members of the CDC25 phosphatase family during G2–M transition activates
cyclin B–CDK1. Cyclin B–CDK1 catalyzes its own activation with feedback loops that activate the CDC25 and inactivate WEE1/MYT1 at the same time.
Initial activation of CDC25 may be carried out by PLK1, which in turn is activated by Aurora A and Bora.
6 Cell Cycle Control

CDK inhibitors
Two families of CDK1 inhibitors are involved in negatively regulate the activities of CDKs. One family includes p16INK4A, p15INK4B,
p18INK4C, and p19INK4D; the other family includes p21CIP1/WAF1, p27KIP1, and p57KIP2.
The p16 family of inhibitors contains ankyrin repeats and are specific for cyclin D-CDK4/6 complexes. Remarkably, an
alternative reading frame of INK4A encodes a protein called p14ARF that shares no sequence homology to p16INK4A. Although
p14ARF does not interact with CDKs, it can negatively regulate the cell cycle by inhibition of the p53 regulator MDM2.
Compare to the p16 family, members of the p21 family can bind and inhibit a broader spectrum of CDKs, including CDK2,
CDK4 (weaker), and CDK1 (weakest). Furthermore, the mechanism through which the p21 family and the p16 family inhibit
CDKs appears to be different: p21CIP1/WAF1 forms a stable complex with cyclin-CDK whereas p16INK4A dissociates cyclin from CDK.
The putative functions of different CDK inhibitors are best seen by disruption of these genes in mice. For example, mice with
targeted deletion of the INK4A gene that eliminates both p16INK4A and p14ARF are viable, but develop spontaneous tumors at an
early age and are highly sensitive to carcinogenic treatments. Significantly, a very similar phenotype was obtained when only
p14ARF was knocked out without disrupting p16INK4A, suggesting that p14ARF could be as important, if not more so, as p16INK4A as a
tumor suppressor.
One putative function of p15INK4B is for arresting the cell cycle in response to TGFb. TGFb induces the synthesis of p15INK4B,
which inhibits cyclin D-CDK4/6 and displaces the p21CIP1/WAF1/p27KIP1 that normally associates with cyclin D-CDK4/6 to
redistribute to other cyclin-CDK complexes. Mice with disrupted p18INK4C gene are larger in size and frequently develop pituitary
tumors. On the other hand, mice with disrupted p19INK4D gene have no increase in tumor but instead have some defects in
spermatogenesis.
One major function of p21CIP1/WAF1 is the inhibition cell cycle progression following DNA damage (see ‘Checkpoints’). DNA
damage induces p53 which in turn activates the transcription of p21CIP1/WAF1. A role for p21CIP1/WAF1 in various cell cycle arresting
states such as senescence and differentiation that is independent of p53 has also been suggested. In contrast to mice lacking p53
(which develop tumors spontaneously), mice lacking p21CIP1/WAF1 develop normally. Nonetheless, embryonic fibroblasts from
these animals are significantly impaired in their ability to arrest in G1 following DNA damage, indicating that the DNA damage
checkpoint is important, but not sufficient for p53-dependent tumor suppression.
The level of p27KIP1 is elevated in a number of cell resting states such as quiescence, contact inhibition, and anchorage-
dependent arrest. Mice with disrupted p27KIP1 gene are larger in size due to a general increase in cell number. These mice also
frequently develop pituitary tumors. Mice lacking p57KIP2 have altered cell proliferation and differentiation, resembling some
features of the human hereditary disorder Beckwith-Wiedemann syndrome, which is characterized by a variety of overgrowth and
predisposition to cancer.

Control of G2–Mitosis

The key event for mitotic entry is the activation of CDK1. Although the abundance of CDK1 is constant throughout the cell cycle, it
is active only during mitosis due to the binding to a mitotic cyclin subunit and dephosphorylation. Other kinases including Aurora
kinases and PLK1 are also critical for CDK1 activation and centrosome functions.
In mammalian cells, A- and B-type cyclins are synthesized and destroyed around the time of mitosis and are regarded as mitotic
cyclins (cyclin A also functions in S phase). There are two A-type cyclins (A1 and A2) and three B-type cyclins (B1, B2, and B3) in
mammalian cells. While cyclin A2 is present in all proliferating somatic cells, cyclin A1 is only important during spermatogenesis.
Cyclin B1 is the major mitotic cyclin partner of CDK1. Cyclin B2 is co-expressed with cyclin B1 in the majority of dividing cells but
is less abundant. The expression of cyclin B3 is restricted to developing germ cells and in the adult testis.
One of the most crucial characteristics of the mitotic cyclins is their periodicity. Cyclin A starts to accumulate during late G1
phase and continue through S phase and G2 phase. Cyclin B is synthesized and destroyed slightly later than cyclin A. The periodic
expression of cyclin A and cyclin B is predominantly regulated at the levels of transcription and proteolysis (see ‘Regulation of CDK
activity’).
Identification of physiological substrates for cyclin B–CDK1 is challenging, partly because CDKs have a loose substrate
specificity of a Ser/Thr followed by a Pro. An example of CDK1 substrate that nicely illustrates the effects of CDK1 on mitosis is
nuclear lamin. Nuclear lamins form dimers that make up the intermediate filaments; the intermediate filaments in turn make up
the nuclear lamina that lines the inside of the nuclear envelope. Breakdown of nuclear envelope is essential for mitosis progression
and separation of the chromosome into the two daughter cells. Cyclin B–CDK1 phosphorylates a Ser residue in lamins, leading to
the depolymerization of the intermediate filaments and, in turn, nuclear envelope breakdown.
The defining characteristic of CDK1 activation is a system of feedback loops that converts the slow accumulation of cyclin B into
an abrupt activation of CDK1 (Figure 2). After cyclin B–CDK1 complexes are formed, their kinase activity is initially suppressed by
inhibitory phosphorylation of CDK1Thr14/Tyr15 by MYT1 and WEE1 (see ‘Regulation of CDK activity’). At the end of G2, the stockpile
of inactive cyclin B–CDK1 complexes is rapidly activated by members of the CDC25 phosphatase family. There are three isoforms
of CDC25. CDC25B is believed to be the initial activator of cyclin B–CDK1 at the centrosomes. This is followed by the complete
activation of cyclin B–CDK1 by feedback loops involving CDC25A and CDC25C in the nucleus. Cyclin B–CDK1 orchestrates its
own activation with feedback loops that activate the CDC25 and inactivate WEE1/MYT1 at the same time. Phosphorylation of
Cell Cycle Control 7

WEE1 by CDK1 (as well as by PLK1) also creates a phosphodegron for SCFbTrCP-dependent degradation. Thus cyclin B–CDK1 is
essentially a bistable switch system that becomes autocatalytic once a critical portion is activated (Figure 4).
Given that the activation of cyclin B–CDK1 is autocatalytic, how the initial batch of cyclin B–CDK1 is activated becomes a
salient issue. The polo-like kinase PLK1 appears to kick-start the system by activating CDC25 and inactivating WEE1/MYT1. The
translocation of cyclin B into the nucleus at prophase is also promoted by the phosphorylation of the nuclear exporting sequence
(NES) by CDK1 and PLK1 (see ‘Regulation of CDK activity’).
The activation of PLK1 in G2 phase requires phosphorylation by Aurora A, an event that is assisted by a protein called Bora.
Binding of Bora to PLK1 is stimulated by cyclin B–CDK1-dependent phosphorylation, creating yet another positive feedback loop
in the activation of cyclin B–CDK1. PLK1 then phosphorylates Bora and generates a phosphodegron motif that is recognized by the
ubiquitin ligase SCFbTrCP, targeting Bora to degradation. Degradation of Bora is believed to be important for redistributing Aurora
A from a cytoplasmic Bora-containing complex to a TPX2-containing complex at the mitotic spindle. Aurora A’s activity increases
from late G2 and peaks in mitosis. Activation of the Aurora A requires binding to specific cofactors such as TPX2, leading to the
autophosphorylation of a residue in the T-loop. The activation of Aurora A is indispensable for various centrosome functions,
including centrosome separation, maturation, and mitotic spindle formation.
The centrosome is a small organelle consisting of a pair of centrioles and the surrounding pericentriolar material. The
centrosome is located near the nucleus and contains the microtubule organizing center, playing important roles in the establish-
ment of the interphase cytoplasmic microtubule network and bipolar mitotic spindles. Centrosome duplication occurs during
S phase and is coupled to the cell cycle by cyclin A/E–CDK2 (see ‘Control of S phase’). Centrosome maturation occurs during S and
G2, establishing the microtubule nucleation activities of centrosomes required for the formation of mitotic spindles by recruiting
critical pericentriolar material components. Separation of duplicated and matured centrosomes in late G2 is crucial for the
formation of bipolar mitotic spindles.
Another player that regulates centrosome maturation and separation is the multifunctional protein kinase PLK1. In addition to
its involvement in the activation of cyclin B–CDK1, PLK1 also performs myriad functions in mitosis, including centrosome
functions, kinetochore-spindle attachment, chromosome segregation, and cytokinesis. The protein level of PLK1 fluctuates in a
cell cycle-dependent manner similar to cyclin B. The localization of PLK1 is also tightly regulated during the cell cycle: it is localized
to both cytoplasm and nucleus during G2, present at centrosomes and kinetochores during prophase and metaphase, and degraded
shortly after anaphase by APC/CCDH1. However, a fraction of PLK1 remains at the central spindle during anaphase and then
localizes to the midbody during cytokinesis.
Another kinase called Greatwall (also called MASTL in human) also regulates the CDK1 activating feedback control. Greatwall
phosphorylates Arpp19 and a-endosulfine, leading to their inhibition of the phosphatase PP2A-B55d. As PP2A-B55d is a major
phosphatase that dephosphorylates cyclin B–CDK1 substrates, Greatwall activity is important for maintaining the activity of CDK1
during mitosis. For example, Greatwall can maintain the phosphorylation of CDC25, thereby keeping CDK1 active. Greatwall itself
appears to be activated during mitosis by CDK1 in a feedback loop.

Control of Mitosis–G1

The key event in mitotic exit is the onset of anaphase, which is driven by the APC/C-dependent ubiquitination. Degradation of
APC/C substrates such as cyclin B and securin is important for events during mitotic exit, including sister chromatid separation,
spindle disassembly, chromosome decondensation, cytokinesis, and reformation of nuclear envelope. How to assemble the
spindle and keep APC/C inactivate before all the chromosomes are aligned correctly are the main issues that determine mitotic exit.
Bipolar spindle formation and proper attachment of chromosomes are highly regulated to ensure that chromosomes are
segregated equally to the daughter cells. Several kinases including CDK1, PLK1, and NEK2 are targeted to unattached kinetochores,
phosphorylate key kinetochores proteins such as HEC1, and contribute to the stabilization of microtubule-kinetochore
interactions.
The chromosomal passenger complex (CPC, composed of Aurora B, Borealin, INCENP, and Survivin) plays a major role in
spindle-assembly and cytokinesis. CPC localizes to the kinetochores and chromosomes in early mitosis and functions in
microtubule-kinetochore interactions, sister chromatid cohesion, and the spindle-assembly checkpoint. It promotes chromosome
biorientation by correcting mis-attachments until the bioriented chromosome is under tension. In anaphase, CPC is relocated to
the central spindle at anaphase and subsequently at the midbody to promote cytokinesis.
Both A- and B-type cyclins contain a sequence motif known as the D-box that is required for their destruction by APC/C (see
‘Regulation of CDK activity’). APC/C is a large complex containing more than 10 subunits. APC/C-dependent proteolysis depends
on binding to the targeting subunits CDC20 and CDH1. APC/CCDC20 complexes are present only during mitosis. In contrast, the
level of CDH1 remains constant during the cell cycle, but it only associates with APC/C during G1.
Activated cyclin B–CDK1 stimulates the activity of APC/CCDC20 through phosphorylation of several subunits of APC/C and
CDC20. APC/C is also phosphorylated and activated by PLK1. Hence APC/CCDC20 is activated only after mitotic entry. However,
APC/CCDC20 is suppressed by a surveillance mechanism termed the spindle-assembly checkpoint until all the chromosomes have
achieved correct bipolar attachment to the mitotic spindles. The checkpoint is activated by either the presence of unattached
kinetochores or the absence of tension between paired kinetochores. Unattached kinetochores attract the components of the
checkpoint sensors (including BUB1, BUBR1, BUB3, CENP-E, MAD1, MAD2, and MPS1), catalyzing the formation of diffusible
8 Cell Cycle Control

Figure 5 Regulation of mitotic exit by APC/CCDC20. During early mitosis, APC/CCDC20 is activated by cyclin B–CDK1 and other mitotic kinases. However,
its activity is suppressed by the spindle-assembly checkpoint. The checkpoint is activated by either the presence of unattached kinetochores or the
absence of tension between paired kinetochores. Inhibition of APC/CCDC20 is carried out by binding of CDC20 to MAD2. Once all kinetochores are
properly attached, the spindle-assembly checkpoint is silenced to allow APC/CCDC20 activation. APC/CCDC20 then targets several proteins including
cyclin B and securin to ubiquitin-mediated degradation. Proteolysis of securin releases separase, which in turn cleaves cohesin to allow sister
chromatid separation and anaphase.

mitotic checkpoint complexes (MCC, components include MAD2, BUBR1, and BUB3). These checkpoint components act as signal
transducers, resulting in the inhibition APC/CCDC20 through the sequestration of CDC20 by MAD2. Binding to CDC20 requires a
conformational change in MAD2 from a less stable open conformation (known as O-MAD2) to the more stable close conformation
(C-MAD2). Although the mechanism remains incompletely understood, several lines of evidence suggest that C-MAD2 can convert
more C-MAD2 from O-MAD2 in an autocatalytic manner. Once all kinetochores are properly attached, the spindle-assembly
checkpoint is silenced to allow APC/CCDC20 activation and anaphase onset (Figure 5).
In addition to cyclin B, APC/CCDC20 also degrades several substrates including securin and geminin. Degradation of securin is
important for sister chromatid separation during anaphase. After DNA is replicated, sister chromatids are tethered together by
cohesin, a ring-shaped complex consisting of four SMC subunits. During prophase, the cohesin at chromosome arms are removed
after being phosphorylated by PLK1 and Aurora B. In contrast, the cohesin complexes at the centromere are protected until
anaphase onset through the action of a protein called Shugoshin (SGO1). BUB1 phosphorylates histone H2ASer121 and generates a
platform for the centromeric localization of SGO1. Phosphorylation of cohesin by PLK1 at centrosomes is counteracted by the
phosphatase PP2A that binds to SGO1. During anaphase, SGO1 is removed (in part by PLK1 and APC/C-dependent degradation).
Sister chromatid separation is then triggered by the cleavage of the remaining cohesin by the protease separase. Before anaphase,
separase is inhibited by binding to securin. Degradation of securin by APC/CCDC20 releases separase from the complex, which in
turn cleaves cohesin to allow sister chromatid separation (Figure 5).
Degradation of geminin by APC/CCDC20 releases CDT1, a subunit required for the initiation of DNA replication (see ‘Control of
S phase’). Hence by simultaneously destroying the mitotic cyclins, securin, and geminin, APC/CCDC20 coordinates several
important processes during the mitosis–G1 transition and prepares for the next S phase.
In addition to CDC20, APC/C can also associate with a targeting subunit called CDH1. In contrast to APC/CCDC20, APC/CCDH1
is turned off during mitosis because phosphorylation of CDH1 by cyclin B–CDK1 alters the conformation of CDH1 and prevents
its binding to APC/C. Destruction of cyclin B at anaphase therefore relieves the inhibition of APC/CCDH1, allowing it to degrade
CDC20 and take over the task of degrading any remaining or newly synthesized cyclin B during G1 phase. Finally, APC/CCDH1 is
also responsible in destroying several important mitotic regulators including PLK1, CDC25A, Aurora A, and SGO1.
During the late G1, E2F is released from pRb and activates the transcription of cyclin A (see ‘Control of G1–S’). The
re-accumulation of cyclin A–CDK complexes increases the phosphorylation of CDH1 and prohibits its association with
APC/C. APC/CCDH1 and APC/CCDC20 are also turned off by binding to EMI1, which begins to accumulate at late G1 (also
transactivated by E2F). EMI has to be removed later to allow APC/C to function in mitotic exit. This is achieved by PLK1-
dependent phosphorylation, targeting EMI to ubiquitin-mediated degradation by SCFbTrCP.

Control of G1–S

The key event in G1–S regulation is the restriction point (R). After passing the restriction point, a cell is committed to another round
of cell cycle and becomes independent of proliferation stimulants. In contrast, it enters quiescence (G0) if there are insufficient
mitogenic signals to overcome the restriction point. Mechanistically, the restriction point involves phosphorylation of pRb by the
G1 cyclin–CDK complexes.
Transcription of cyclin D increases when quiescent cells are stimulated to enter the cell cycle by growth factors through
integration of multiple signal transduction pathways (such as the RAS-RAF-MEK-ERK pathway). The strong dependence of cyclin
Cell Cycle Control 9

Figure 6 Regulation of G1–S. Transcription of cyclin D increases when quiescent cells are (G0) stimulated to enter the cell cycle by external
growth signals. Hyperphosphorylation of pRb by cyclin D–CDK4/6 releases pRb from E2F, allowing E2F to activate transcription. Among other
genes, E2F activates the transcription of cyclin E and cyclin A, which activates CDK2 and further boosts the phosphorylation of pRb. Increase of
E2F-dependent transcription allows cells to pass through the restriction point (R).

D expression on extracellular mitogenic cues, coupling to the relative short half-live of the protein (see below), allows cyclin D to
act as an effective mitogenic sensor that relay extracellular signals to the cell cycle.
After cyclin D binds to its partner CDK4/CDK6, the complex then phosphorylates the retinoblastoma gene product pRb.
Hypophosphorylated pRb binds E2F, which is a critical transcription factor for genes important for entry into S phase. As pRb also
associates with histone deacetylase (HDAC), this indirectly brings HDAC to the promoters bound by E2F, thereby repressing their
transactivation through chromatin remodeling. Hyperphosphorylation of pRb by cyclin D–CDK4/6 releases pRb from E2F (and
removing HDAC at the same time), allowing E2F to activate transcription. Among other genes, E2F activates the transcription of
cyclin E and cyclin A, which activates CDK2 and further increases the phosphorylation of pRb. The pRb–E2F pathway functions as a
bistable switch to convert graded growth factor stimulations into all-or-none E2F responses (Figure 6).
Cyclin E is degraded by the SCFFBW7 ubiquitin ligase. The phosphodegron recognized by SCFFBW7 is created by CDK2-dependent
autophosphorylation (as well as by GSK3b). Degradation of cyclin D (cyclin D1 is the best understood) involves phosphorylation
by GSK3b. This generates a phosphodegron that is recognized by the ubiquitin ligase SCFFBX4.
Negative regulators of the G1 cyclin–CDK complexes including the p16 and p21 family of CDK inhibitors (see ‘Regulation of
CDK activity’) can modulate the threshold of the restriction point. The levels of some of the CDK inhibitors are regulated during the
cell cycle. For example, p21CIP1/WAF1 and p27KIP1 are degraded by the ubiquitin ligase SCFSKP2 complex. SKP2 itself is degraded by
APC/CCDH1 (see ‘Control of mitosis–G1’). Low level of SKP2 conferred by APC/CCDH1 in G1 promotes the accumulation of p21CIP1/
WAF1
and p27KIP1, thereby inactivating CDK2 and inhibiting S phase entry.

Control of S Phase

The key issues concerning the control of S phase are (1) how DNA replication occurs only in S phase, and (2) how replication is
initiated once and once only per cell cycle. Centrosome duplication also occurs during S phase and is in part coupled with the
mechanisms that control DNA replication.
It is well established that initiation of DNA replication occurs at chromosomal locations known as origins of replication.
S. cerevisiae is the only known eukaryote with a defined initiation sequence. No specific initiation sequences are found in other
eukaryotes, but they are generally either AT-rich or exhibit bent DNA topology. Several proteins, including origin recognition
complex (ORC, which composed of ORC1-6), CDC6, and CDT1 are assembled at the origins of replication during G1 phase. This
allows the loading of double hexamers of the MCM2-7 core helicase, forming the so-called pre-replication complex (pre-RC)
(Figure 7). The formation of pre-RC on origins is called origin licensing.
During G1–S transition, the origins are activated by CDK2 (activated by cyclin A/E) and another kinase called CDC7 (activated
by binding to DBF4). These kinases phosphorylate components of the pre-RC including MCM2-7, thereby triggering the recruit-
ment of two helicase co-activators, CDC45 and GINS. The MCM2-7 helicase is then activated and unwinds the origin. The single-
stranded DNA is subsequently stabilized by binding to replication protein A (RPA). Finally, the unwound DNA allows the
recruitment of DNA polymerases and other components of the DNA synthesis machinery to initiate DNA synthesis.
Once the genome has been replicated, the pre-RC must not form again until the next cell cycle. The accumulation of CDK
activity during late G1, S, and G2 prevents the re-assembly of the pre-RC by multiple mechanisms. CDK-dependent phosphoryla-
tion excludes MCM2-7 from the nucleus, targets CDT1 and CDC6 for degradation, and dissociates ORC from the chromatin.
Furthermore, accumulation of the protein geminin during S and G2 results in the formation of a tight geminin–CDT1 complex,
thereby preventing CDT1 from loading onto the pre-RC.
10 Cell Cycle Control

Figure 7 Initiation of DNA replication. Initiation of DNA replication occurs at origins of replication. During G1 phase, the origins are licensed by binding
to the pre-replication complex. During G1–S transition, MCM2-7 and other components are phosphorylated by cyclin A/E–CDK2 and DBF4–CDC7,
stimulating the recruitment of CDC45 and GINS. This activates the MCM2-7 helicase to unwind the origin. Finally, the unwound DNA allows the
recruitment of DNA polymerases and other components of the DNA synthesis machinery to initiate DNA synthesis. After DNA replication, CDT1 and
CDC6 are targeted for degradation. CDT1 is also inhibited by binding to newly synthesized geminin. These and other mechanisms prevent the re-firing
the replication origins.
Cell Cycle Control 11

Assembly of the pre-RC only occurs during early G1 because destruction of cyclin A and cyclin B during mitosis provides an
environment of low CDK activity. Proteolysis of geminin by APC/C during mitosis also releases CDT1 from the complex to form
the pre-RC. Hence APC/C resets the mechanisms that prevents re-replication during the last cell cycle.
Cyclin A/E–CDK2 also coordinates the initiation of DNA replication with centrosome cycle. Since each daughter cell receives
only one centrosome after cell division, the centrosome must duplicate once before the next mitosis. Centrosome duplication is
initiated, at least in part, by the activity of cyclin A/E–CDK2.

Checkpoints

The intricate mechanisms that regulate the activities of CDKs, APC/C, and other cell cycle regulators ensure that they are turned on
and off at the correct time in the cell cycle. Several surveillance mechanisms called checkpoints are present to ensure that each stage
of the cell cycle is completed before the next stage is initiated. Major checkpoints are also present to monitor DNA damage to ensure
that damaged DNA is repaired before cell cycle progression. In general, checkpoints include a sensor that monitors cell cycle
defects, signal transducers, and an effector that inhibits the cell cycle. Malfunctioning of checkpoints allows the cell cycle to become
insensitive to external signals, spindles assembly, or DNA damage, giving rise to genome instability.

Spindle-Assembly Checkpoint
Chromosomes that are not attached to spindles send signals to block the inactivation of cyclin B–CDK1 and sister chromatid
separation. This spindle-assembly checkpoint is essential to ensure the equal segregation of chromosomes to daughter cells. The
current paradigm states that checkpoint is activated by either the presence of unattached kinetochores or the absence of tension
between paired kinetochores. A number of checkpoint sensor proteins are localized to unattached or improperly attached
kinetochores, catalyzing the formation of a signal transducer called mitotic checkpoint complex (MCC). The final effector of the
checkpoint is C-MAD2, a specific conformation of MAD2 that can bind and inhibit APC/CCDC20, thereby trapping the cell in
mitosis (see ‘Control of mitosis–G1’). Once all kinetochores are properly attached, the spindle-assembly checkpoint is silenced to
allow APC/CCDC20 activation and anaphase onset. Mutations of the components of the spindle-assembly checkpoint have been
implicated to play significant roles in chromosomal instability in various cancers (Figure 5).

G1 DNA Damage Checkpoint


When cellular DNA is damaged, continue progression through the cell cycle would lead to detrimental DNA mutagenesis.
Damaged cells can pause the cell cycle by checkpoint mechanisms to provide time for DNA repair. Alternatively, the damaged
cells can be eliminated by activating apoptosis. The tumor suppressor p53 is critical for both the cell cycle checkpoint and apoptosis
following DNA damage (Figure 8).
In the absence of DNA damage, p53 is suppressed by one of its own transcriptional targets called MDM2 in a negative feedback
loop. MDM2 binds to the N-terminal transactivation domain of p53 and inhibit p53-mediated transcription, shuttles p53 out of
the nucleus, and promotes ubiquitin-mediated degradation of p53. The last effects on p53 is due to the fact that MDM2 is itself a
ubiquitin ligase.
DNA damage activates sensors that facilitate the activation of the PI-3 (phosphoinositide 3-kinase)-related protein kinases ATM
and ATR, which in turn activate the checkpoint kinases CHK1 and CHK2. ATM/ATR, CHK1/CHK2, and several other DNA damage-
activated protein kinases can phosphorylate the N-terminal region of p53 as well as MDM2. Phosphorylation of these sites
abolishes the MDM2–p53 interaction, leading to a rise in p53 level, nuclear localization, and transcriptional activity.
One of the transcriptional targets of p53 is the CDK inhibitor p21CIP1/WAF1 (see ‘Regulation of CDK activity’). The accumulated
p21CIP1/WAF1 then binds and inhibits cyclin A/E–CDK2. This prohibits the phosphorylation of pRb, thereby stopping the cell cycle
in G1 phase (see ‘Control of G1–S’).
Another important control of the p53 pathway comes from the INK4A gene, which encodes both the CDK inhibitor p16INK4A
and a protein called p14ARF (see ‘Regulation of CDK activity’). The negative regulation of p53 by MDM2 is interrupted by p14ARF
because it sequester MDM2 to the nucleolus. As the INK4A gene is generally activated in response to oncogenic stresses, the
expression of p16INK4A and p14ARF reduces cyclin D–CDK4/6 activity and elevates p53 expression, respectively. Both of these events
eventually suppress pRb phosphorylation and arrest the cell cycle in G1 phase.

G2 DNA Damage Checkpoint


Similarly to the G1 DNA damage checkpoint, the G2 DNA damage checkpoint involves the activation of the protein kinases
ATM/ATR followed by CHK1/CHK2. CHK1/CHK2 then activates WEE1 and inactivates all three isoforms of the CDC25 family
(CDC25A, CDC25B, and CDC25C). Together, these mechanisms promotes CDK1Thr14/Tyr15 phosphorylation, leading to the
inactivation of CDK1 and cell cycle arrest in G2 phase (see ‘Control of G2–mitosis’) (Figure 8).
Phosphorylation of CDC25C by CHK1/CHK2 inactivates its phosphatase activity either directly or indirectly through the
creation of a 14-3-3 binding site. Binding of 14-3-3 masks a proximal nuclear localization sequence and anchors CDC25C in the
12 Cell Cycle Control

Figure 8 The DNA damage checkpoints. DNA damage activates sensors that facilitate the activation of ATM and ATR, which in turn activate the
checkpoint kinases CHK1 and CHK2. Together, these kinases phosphorylate p53 and MDM2, disrupting the inhibition of p53 by MDM2. This results in
the stabilization and activation of p53. One of the transcriptional targets of p53, p21CIP1/WAF1, is a CDK inhibitor and inhibits cyclin E–CDK2, leading to a
cell cycle arrest in G1 phase. The G2 DNA damage checkpoint is likewise carried out by the ATM/ATR-CHK1/CHK2 pathway. CHK1/CHK2 phosphorylates
and inactivates CDC25, preventing the dephosphorylation of CDK1 and arresting the cell in G2 phase.

cytoplasm, preventing efficient access of CDC25C to cyclin B–CDK1. The centrosomal CDC25B is also phosphorylated by CHK1,
creating a docking site for 14-3-3 that prevents access of CDK1 to the catalytic site. In contrast to other CDC25 isoforms, CDC25A is
targeted for rapid degradation by CHK1/CHK2 through a ubiquitin-mediated mechanism. CDC25A stability is controlled by
APC/CCDH1 complexes during mitotic exit and early G1 and by SCFbTrCP complexes during interphase. Importantly, the SCFbTrCP-
dependent turnover of CDC25A is enhanced in response to DNA damage through phosphorylation by CHK1. In addition of acting
on CDC25, CHK1/CHK2 also appears to phosphorylate and activate WEE1 by promoting 14-3-3 binding.

Replication Checkpoint
Stalled replication forks mainly activate the ATR-CHK1 pathway. Replication fork progression can be impaired by insufficient
nucleotide supply or lesions and obstacles on the DNA. Several proteins including ATRIP, TopBP1, and Claspin are involved in
recruiting ATR to single-stranded DNA present at stalled replication forks. Specifically, ATR is activated by binding to the single-
strand binding protein RPA-coated DNA. The activated ATR then phosphorylates and activates CHK1. CHK1 subsequently acts on
CDC25 and WEE1, inhibiting the cyclin–CDK involved in both replication (see ‘Control of S phase’) and mitotic entry (see ‘Control
of G2–mitosis’). Hence the replication checkpoint regulates origin firing, replication forks progression, as well as prevents untimely
mitosis. These provide the cell with time to restart or repair the stalled replication forks. It is notable that the ATR-CHK1 pathway is
essential even in the absence of exogenous stresses during unperturbed S phase.

Acknowledgements

Related works in my laboratory are supported in part by the Research Grants Council.
Cell Cycle Control 13

Further Reading
Fung TK and Poon RYC (2005) A roller coaster ride with the mitotic cyclins. Seminars in Cell & Developmental Biology 16: 335–342.
Kastan MB and Bartek J (2004) Cell-cycle checkpoints and cancer. Nature 432: 316–323.
Ma HT and Poon RYC (2011) How protein kinases coordinate mitosis in animal cells. Biochemical Journal 453: 17–31.
Masai H, Matsumoto S, You Z, Yoshizawa-Sugata N, and Osa M (2010) Eukaryotic chromosome DNA replication: Where, when, and how? Annual Review of Biochemistry 79: 89–130.
Morgan DO (2006) The cell cycle: Principles of control. Sunderland, Massachusetts, USA: Sinauer Associates, Inc.
Woo RA and Poon RYC (2003) Cyclin-dependent kinases and S phase control in mammalian cells. Cell Cycle 2: 316–324.

You might also like