Results Probl Cell Differ (42)
P. Kaldis: Cell Cycle Regulation
DOI 10.1007/b138827/Published online: 23 September 2005
© Springer-Verlag Berlin Heidelberg 2005
Protein Kinases Involved in Mitotic Spindle
Checkpoint Regulation
Ingrid Hoffmann
Cell Cycle Control and Carcinogenesis (F045), German Cancer Research Center (DKFZ),
Im Neuenheimer Feld 242, 69120 Heidelberg, Germany

Abstract A number of checkpoint controls function to preserve the genome by restraining
cell cycle progression until prerequisite events have been properly completed. Chromo-
some attachment to the mitotic spindle is monitored by the spindle assembly checkpoint.
Sister chromatid separation in anaphase is initiated only once all chromosomes have been
attached to both poles of the spindle. Premature separation of sister chromatids leads to
the loss or gain of chromosomes in daughter cells (aneuploidy), a prevalent form of ge-
netic instability of human cancer. The spindle assembly checkpoint ensures that cells with
misaligned chromosomes do not exit mitosis and divide to form aneuploid cells. A num-
ber of protein kinases and checkpoint phosphoproteins are required for the function of
the spindle assembly checkpoint. This review discusses the recent progress in under-
standing the role of protein kinases of the mitotic checkpoint complex in the surveillance
pathway of the checkpoint.
1
Introduction
During cell division, accurate transmission of the genome is essential for
survival. Entry into mitosis is controlled by checkpoints that monitor DNA
damage and replication, whereas exit from mitosis is controlled by check-
points that monitor assembly and position of the mitotic spindle. The mitotic
spindle checkpoint is activated by the lack of microtubule occupancy and ten-
sion at the kinetochores and leads to cell cycle arrest in prometaphase. It
is a tightly conserved signal transduction pathway that prevents sister chro-

matid separation until all chromosomes achieve bipolar attachment to the
mitotic spindle. The presence of even a single misaligned or unattached chro-
mosome is sufficient to activate the checkpoint. In response to defects in
the mitotic apparatus, it blocks the activity of the anaphase-promoting com-
plex or cyclosome (APC/C), a large multisubunit ubiquitin ligase required
for chromosome segregation. After all sister chromatids have achieved bi-
orientation, the APC/C in association with one of its substrate-binding cofac-
tors, Cdc20, tags the anaphase-inhibiting protein securin with polyubiquitin
chains, leading to its degradation by the proteasome (Peters 2002; Harper
et al. 2002). Sister chromatids are held together by cohesin and cleavage of
94 I. Hoffmann
cohesin will result in loss of sister chromatid cohesion and the onset of sis-
ter chromatid separation (Nasmyth 2002). Degradation of securin activates
separase, a protease which cleaves the Scc1 subunit of cohesin.
2
The Spindle Assembly Checkpoint
The molecular components of the spindle assembly checkpoint were identi-
fied initially in Saccharomyces cerevisiae (Gorbsky 2001; Shah and Cleveland
2000). They include Mad1-3 (mitotic arrest deficiency) (Li and Murray 1991),
Bub1-3 (Hoyt et al. 1991; Roberts et al. 1994), and Mps1 (Weiss and Winey
1996) (Table 1). Homologues of these checkpoint proteins were later found
in other organisms, including mammals. Checkpoint proteins accumulate at
unattached kinetochores in prometaphase, but disappear from kinetochores
later in mitosis or meiosis upon microtubule attachment and tension. In higher
eukaryotes the checkpoint control proteins comprise Mad1, Mad2, Bub3 and
the protein kinases Bub1, BubR1 (Mad3 in budding yeast), and Mps1. In add-
ition to these basic checkpoint components, other proteins such as CENP-E
(a member of the kinesin superfamily) (Abrieu et al. 2000; Yao et al. 2000), Rod,
ZW10 (Chan et al. 2000), Aurora B (Biggins and Murray 2001; Kallio et al. 2002;
Ditchfield et al. 2003) and mitogen-activated protein kinase (MAPK) (Shapiro

et al. 1998; Zecevic et al. 1998) play a role in the spindle checkpoint (Table 1).
Table 1 Proteins involved in mitotic spindle checkpoint regulation
Protein Proposed function in the spindle assembly checkpoint
Mad1 Coiled-coil protein, binds to Mad2 and recruits Mad2 to kinetochores
phosphorylated by Mps1 and Bub1 upon checkpoint activation
Mad2 Binds to Mad1, binds and inhibits APC/C
Cdc20
BubR1 (Mad3) Protein kinase, binds to Bub3 and APC/C
Cdc20
, binds to the mitotic
motor protein CENP-E
Bub1 Protein kinase, binds to and recruits Bub3, Mad1 and Mad2
Bub3 Contains WD-40 repeats, binds to Bub1 and BubR1
Mps1 Protein kinase, essential for establishment and maintenance of the
spindle checkpoint
Aurora B Protein kinase, binds to INCENP, other substrates: CENP-A, Rec8,
vimentin, desmin, the kinesin MCAK, histone H3
Aurora C Protein kinase, binds to INCENP and Aurora B
MAPK Protein kinase
Rod Identified in Drosophila, binds to Zw10
Zw10 Identified in Drosophila, binds to Rod
Protein Kinases Involved in Mitotic Spindle Checkpoint Regulation 95
Subcellular localization studies have placed all these checkpoint proteins at
the kinetochores. Ablation or suppression of function of any of these proteins
substantially compromises mitotic checkpoint control (Lew and Burke 2003).
These checkpoint control proteins form a complex intracellular network, the
mitotic checkpoint complex (MCC), to block the action of APC/C
Cdc20
.
Mad2 interacts with other components of the spindle checkpoint and plays

a key role in the signaling pathway of the checkpoint. In interphase, it binds to
Mad1 and is preferentially found on the nuclear periphery (Chen et al. 1999).
This localization is strictly dependent on Mad1 since in a fission yeast strain
lacking Mad1, Mad2 is no longer found on the nuclear periphery (Ikui et al.
2002). Upon the onset of mitosis, Mad2 translocates into the nucleus and is
guided to unattached kinetochores by Mad1 (Chen et al. 1999). Recruiting
Mad2 to kinetochores is the only known function of Mad1 to date. From early
mitosis on, Mad2 is found in a complex with its target, Cdc20. Human Mad2
is modified through phosphorylation on multiple serine residues in vivo in
a cell cycle-dependent manner. Only unphosphorylated Mad2 interacts with
Mad1 or the APC/C in vivo (Wassmann et al. 2003). Injection of anti-Mad2
antibodies drives prophase cells into a premature anaphase and overrides the
arrest induced by microtubule depolymerization, indicating that the check-
point activation in situations with unattached kinetochores requires Mad2
(Gorbsky et al. 1998). A Mad2 mutant containing serine to aspartic acid muta-
tions mimicking the C-terminal phosphorylation events fails to interact with
Mad1 or the APC/C and acts as a dominant-negative antagonist of wild-type
Mad2 (Wassmann et al. 2003). Although yeast strains lacking Mad2 are vi-
able, deletion of Mad2 in mouse causes cell lethality. Mad2-/-mousecellsdo
not arrest in response to spindle damage, show widespread chromosome mis-
segregation, and undergo apoptosis during initiation of gastrulation (Dobles
et al. 2000).
Upon checkpoint activation, both BubR1-Bub3 and Mad2 are capable of
blocking the activity of APC/C through their direct binding to Cdc20 (Yu
2002; Bharadwaj and Yu 2004). Binding of spindle microtubules to kine-
tochores, disrupts the interaction between Mad1 and Mad2 and ultimately
disables the arrest (Fig. 1a).
3
Regulation of the Spindle Checkpoint by Protein Kinases
3.1

Bub1
Bub1 is a protein kinase and an essential checkpoint component that re-
sides at kinetochores during mitosis. Bub1 was first described in a genetic
screen searching for budding yeast mutants that were sensitive to the spin-
96 I. Hoffmann
Fig. 1 Functions of protein kinases in the spindle checkpoint. Attachment of chromo-
somes to the mitotic spindle is monitored by the spindle checkpoint. Sensing mechanisms
may involve Aurora B/Ipl1 and CENP-E. Upon checkpoint activation both BubR1-Bub3
and Mad2 interact with Cdc20 and lead to an inhibition of APC. Inactivation of the
checkpoint occurs upon bipolar attachment to the mitotic spindle. APC is activated and
ubiquitinates securin. Degradation of securin activates a protease called separase which
cleaves cohesin, resulting in a loss of sister chromatid cohesion, leading to the the onset
of anaphase
Protein Kinases Involved in Mitotic Spindle Checkpoint Regulation 97
dle poison benomyl (Hoyt et al. 1991). Bub1 binds to Bub3 throughout the cell
cycle and phosphorylates Bub3 in vitro (Roberts et al. 1994). Overexpression
of a dominant allele of Bub1 in yeast causes a mitotic delay without spin-
dle damage that is dependent on the functions of Bub2, Bub3, and Mad1-3
(Farr and Hoyt 1998). In yeast, the Bub1–Bub3 complex interacts with Mad1
when the spindle checkpoint is activated (Brady and Hardwick 2000). In ver-
tebrates, Bub1 is required for the kinetochore localization of Mad1 and Mad2
(Sharp-Baker and Chen 2001; Johnson et al. 2004; Vigneron et al. 2004). This
activity of Bub1 seems to be independent of its kinase activity since a kinase-
inactive mutant of Bub1 is fully capable of recruiting Mad1 and Mad2 to
the kinetochores in Xenopus egg extracts (Sharp-Baker and Chen 2001). Im-
munodepletion of Bub1 abolishes the spindle checkpoint and the kinetochore
binding of the checkpoint proteins Mad1-3, Bub3, BubR1 and the kineto-
chore motor protein CENP-E (Sharp-Baker and Chen 2001; Johnson et al.
2004). Recently, it was shown that mammalian Bub1 also has a downstream
function in the spindle checkpoint since it directly phosphorylates Cdc20

(Tang et al. 2004). HeLa cells depleted for Bub1 by RNA interference (RNAi)
are defective in checkpoint signaling. Bub1 directly phosphorylates Cdc20
in vitro and in vivo and inhibits the ubiquitin ligase activity of APC/C
Cdc20
catalytically (Tang et al. 2004). Six Ser/Thr residues in Cdc20 were phospho-
rylated by Bub1 in vitro. Ectopic expression of a Cdc20 protein where all six
residues were mutated to alanine is refractory to Bub1-mediated phosphory-
lation and inhibition. Overexpression of this Cdc20 mutant protein impairs
the function of the spindle checkpoint. Bub1 function seems to be regulated
by several upstream kinases. Bub1 becomes hyperphosphorylated and its ki-
nase activity is induced specifically at unattached chromosomes (Chen 2004).
MAPK contributes to this phosphorylation, as inhibiting MAPK or altering
MAPK consensus sites in Bub1 abolishes the phosphorylation and activation
on chromosomes. The activation of Bub1 seems to be important in maintain-
ing the checkpoint towards late prometaphase when the cell contains only
a few kinetochores or a single unattached kinetochore. It has been shown
that the MAPK downstream target Rsk activates cytosolic Bub1 during frog
oocyte maturation (Schwab et al. 2001), but Rsk does not seem to be in-
volved in Bub1 phosphorylation at kinetochores, because immunodepletion
of Rsk does not have an effect on the phosphorylation. Fission yeast Bub1
is phosphorylated during mitosis and the protein is a substrate for Cdc2
(Yamaguchi et al. 2003). Mutation at four putative Cdc2 sites abolishes the
checkpoint function (Yamaguchi et al. 2003). Both Cdc2 and MAPK have
similar consensus phosphorylation sites. In egg extracts, inhibition of MAPK
abolishes Bub1 phosphorylation without an effect on Cdc2 activity, indicat-
ing that Cdc2 is not involved in Bub1 phosphorylation in the frog. Finally,
Bub1 has a noncheckpoint function at the kinetochores and preserves cohe-
sion through the MEI-S332/shugoshin family of proteins (Salic et al. 2004;
Tang et al. 2004).
98 I. Hoffmann

3.2
BubR1
BubR1 was isolated as a Mad3/Bub1-related protein kinase on the basis of
its similarities with the N-terminal domain of the yeast checkpoint protein
Mad3 (Taylor et al. 1998). Thus, BubR1 is thought to be the homologue of
Mad3; in higher eukaryotes BubR1 directly binds to CENP-E (Chan et al.
1998). It is a mitosis-specific kinase and is inactive during interphase (Chan
et al. 1999). Microinjection of antibodies against BubR1 into HeLa cells ab-
rogated mitotic arrest after nocodazole-induced spindle disassembly (Chan
et al. 1998; Chan et al. 1999). In Xenopus egg extracts, immunodepletion of
BubR1 also prevented mitotic arrest in response to spindle damage (Chen
2002). BubR1 accumulates and becomes hyperphosphorylated at unattached
kinetochores. Immunodepletion of BubR1 greatly reduces kinetochore bind-
ing of Bub1, Bub3, Mad1, Mad2, and CENP-E. These defects can be rescued
by wild-type, kinase-dead, or a truncated BubR1 protein that lacks its ki-
nase domain, indicating that the kinase activity of BubR1 is not essential for
the spindle checkpoint in egg extracts (Chen 2002). Whether phosphorylation
of BubR1 leads to an activation of the kinase is not known. BubR1 accu-
mulates to a higher level and becomes hyperphosphorylated at unattached
kinetochores compared with that at metaphase kinetochores. This phospho-
rylation requires Mad1 or its downstream effector, but not Mad2 (Chen 2002).
Expression of a kinase-inactive mutant of BubR1 abolished mitotic arrest
induced by microtubule disassembly (Chan et al. 1999). RNAi-mediated de-
pletion of BubR1 causes severe chromosome misalignment and results in the
loss of kinetochore-microtubule attachment (Chan et al. 1999). Attachment in
these cells can be restored by inhibition of Aurora kinase, which is known to
stabilize kinetochore-microtubule interactions. BubR1 similar to Mad2 also
associates and can phosphorylate Cdc20 in vitro leading to inactivation of
the APC/C (Sudakin et al. 2001; Tang et al. 2001; Fang 2002). Both Mad2 and
BubR1 can indirectly bind to Cdc20 in vitro and either independently or co-

operatively inhibit polyubiquitination of APC/C
Cdc20
substrates. Quantitative
analysis indicates that BubR1 binds to Cdc20 with a higher affinity and is
more potent than Mad2 in inhibiting the activation of APC by Cdc20 (Fang
2002). The two pathways seem to act synergistically since inactivation of ei-
ther Mad2 or BubR1 by microinjection of inhibitors shows that the activity
of both proteins is required for the metaphase delay at 23
◦
C (Shannon et al.
2002). But why does the cell need two different inhibitors of APC in the same
checkpoint pathway? It is possible that binding of Cdc20 to BubR1 recruits
Cdc20 to kinetochores, where BubR1 promotes the formation of the Mad2–
Cdc20 complex, which subsequently diffuses away from kinetochores and
inhibits APC throughout the cell. Alternatively, BubR1 and Mad2 might in-
hibit APC in response to different checkpoint signals. In Xenopus egg extracts,
CENP-E dependent activation of BubR1 kinase activity at kinetochores is ne-