Cell Death by Mitotic Catastrophe: a Molecular Definition

Home , Mitosis, Mitotic catastrophe, Prometaphase

Cell death by mitotic catastrophe: a molecular deﬁnition

Maria Castedo1, Jean-Luc Perfettini1, Thomas Roumier1, Karine Andreau1, Rene Medema2 and Guido Kroemer*,1

1CNRS-UMR 8125, Institut Gustave Roussy, Pavillon de Recherche 1, 39 rue Camille-Desmoulins, Villejuif F-94805, France; 2Division of Molecular Biology H8, Netherlands Cancer Institute, Plesmanlaan 121, Amsterdam 1066 CX, The Netherlands

The current literature is devoid of a clearcut definition of mutant strains (Molz et al., 1989; Ayscough et al., mitotic catastrophe, a type of cell death that occurs during 1992). Accordingly, some authors view ‘mitotic mitosis. Here, we propose that mitotic catastrophe results catastrophe’ of mammalian cells as the failure to from a combination of deficient cell-cycle checkpoints (in undergo complete mitosis (which would be more particular the DNA structure checkpoints and the spindle accurately called ‘mitotic failure’) after DNA damage assembly checkpoint) and cellular damage. Failure to (coupled to defective checkpoints), a situation that arrest the cell cycle before or at mitosis triggers an would lead to tetraploidy (after a single cell cycle) attempt of aberrant chromosome segregation, which (Andreassen et al., 2001a; Margottin-Goguet et al., culminates in the activation of the apoptotic default 2003) or endopolyploidy (after several cell cycles) with pathway and cellular demise. Cell death occurring during extensive DNA damage and repair, perhaps followed by the metaphase/anaphase transition is characterized by the the selection of apoptosis-resistance cells that will ulti- activation of caspase-2 (which can be activated in response mately survive after endoreduplication (Ivanov et al., to DNA damage) and/or mitochondrial membrane per- 2003). meabilization with the release of cell death effectors such During recent years, the expression of ‘mitotic as apoptosis-inducing factor and the caspase-9 and-3 catastrophe’ has been widely used to describe a form activator cytochrome c. Although the morphological of death affecting mammalian cells. Unfortunately aspect of apoptosis may be incomplete, these alterations enough, there is no broadly accepted definition of the constitute the biochemical hallmarks of apoptosis. Cells term ‘mitotic catastrophe’. Igor Roninson attempted to that fail to execute an apoptotic program in response to define mitotic catastrophe in morphological terms, mitotic failure are likely to divide asymmetrically in the namely, as a type of cell death resulting from abnormal next round of cell division, with the consequent generation mitosis, which usually ends in the formation of large of aneuploid cells. This implies that disabling of the cells with multiple micronuclei and decondensed chro- apoptotic program may actually favor chromosomal matin (Swanson et al., 1995; Ianzini and Mackey, 1997; instability, through the suppression of mitotic catastrophe. Roninson et al., 2001). However, reports describing Mitotic catastrophe thus may be conceived as a molecular ‘mitotic catastrophe’ frequently show cells with some device that prevents aneuploidization, which may partici- phenotypic characteristic of apoptosis (such as hyper- pate in oncogenesis. Mitotic catastrophe is controlled by condensed chromatin aggregates) (Heald et al., 1993), numerous molecular players, in particular, cell-cycle- which previously often was interpreted as ‘premature specific kinases (such as the cyclin B1-dependent kinase chromatin condensation’ (Sperling and Ra, 1974; Cdk1, polo-like kinases and Aurora kinases), cell-cycle Chakrabarti and Chakrabarti, 1987). Thus, there is no checkpoint proteins, survivin, p53, caspases and members consensus on the distinctive morphological appearance of the Bcl-2 family. of mitotic catastrophe as far as the extent of chromatin Oncogene (2004) 23, 2825–2837. doi:10.1038/sj.onc.1207528 condensation (which is the morphological hallmark of apoptosis) (Kerr et al., 1972; Ferri and Kroemer, 2001a) Keywords: apoptosis; Cdk1; Chk2; caspases; Bcl-2 is concerned. It has been argued that mitotic catastrophe would be fundamentally different from apoptosis (Roninson et al., 2001) because, in some paradigms, manipulations that Introduction prevent apoptosis (such as overexpression of Bcl-2 in etoposide-treated HeLa cells or MDR1 overexpression ‘Mitotic catastrophe’ has been first described in in irradiated tumor cells) can actually enhance the Schizosaccharomyces pombe as a temperature-sensitive frequency of catastrophic mitoses (Lock and Stribins- lethal phenotype, linked to gross abnormalities of kiene, 1996; Ruth and Roninson, 2000). It has also been chromosome segregation, that was observed in some argued that the assessment of micronucleation (which would indicate mitotic catastrophe) can yield a more accurate picture of cell death induction in vivo than *Correspondence: G Kroemer; E-mail: [email protected] the quantitation of apoptosis (Driessens et al., 2003). Mitotic catastrophe M Castedo et al 2826 Moreover, the fact that caspase inhibitors such as Molecules and processes involved in mitotic catastrophe Z-VAD.fmk can fail to prevent the appearance of dying giant multinucleated cells (induced, for instance, Cdk1/cyclin B1 complex: essential for mitosis and mitotic by treatment with spindle poisons) has been used to catastrophe conclude that mitotic catastrophe would be unrelated to apoptosis (Nabha et al., 2002), assuming that apop- The cyclin-dependent kinase 1 (Cdk1), formerly called tosis must be mediated by caspases. However, it is Cdc2 (or p34Cdc2), interacts with its obligate allosteric now well established that apoptosis may occur in a activator, cyclin B1 to form an active heterodimer, the caspase-independent manner (Susin et al., 1999, 2000; ‘mitosis-promoting factor’. Progression from G2 to Joza et al., 2001), meaning that failure to prevent M phase is driven by the activation of the Cdk1/cyclin cell death by caspase inhibitors cannot be used as B1 complex, whose activity must be sustained from an argument to define this type of cell death as prophase to metaphase. Subsequent entry into the nonapoptotic. anaphase critically relies on the sudden destruction of An alternative definition of mitotic catastrophe the Cdk1/cyclin B1 activity by the anaphase-promoting may be considered as near-to-tautological. Mitotic complex (APC) (Nigg, 2001; Smits and Medema, 2001) catastrophe would be a type of cell death occurring (Figure 1). Accordingly, the activity of Cdk1 is regulated during mitosis, as a result of DNA damage or deran- in a complex spatiotemporal pattern at several levels, ged spindle formation coupled to the debilitation of namely, (i) the transcription of cyclin B1 and (to a lower different checkpoint mechanisms that would normally extent) that of Cdk1, (ii) regulatory Cdk1 (de)pho- arrest progression into mitosis and hence suppress sphorylations by multiple kinases and phosphatases, (iii) catastrophic events until repair has been achieved. endogenous Cdk1 inhibitors, (iv) the subcellular dis- The ‘DNA structure checkpoint’ arrests cells at the tribution of cylin B1, and (v) the regulated degradation G2/M transition in response to unreplicated DNA of cyclin B1. or DNA damage, and the ‘spindle assembly check- The transcription of Cdk1 is modulated in conditions point’ prevents anaphase until all chromosomes have of genotoxic stress. For instance, p53 negatively obtained bipolar attachment. The combination of regulates the transcription of Cdk1 (Yun et al., 1999) checkpoint deficiencies and specific types of damage and cyclin B1 (Taylor et al., 1999), while enhancing the would lead to mitotic catastrophe. Thus, the pharma- transcription of three Cdk1 inhibitors (Gadd45, p21 and cological inhibition or genetic suppression of several 14-3-3s) (Taylor and Stark, 2001). The cyclin B1 gene is G2 checkpoint genes such as ATM, ATR, Chk1, transcribed and its mRNA is stabilized from the end of Chk2, polo-like kinase (Plk)1, Plk2, Plk3, Pin1, Mlh1 the S phase. The phosphorylation of Cdk1 on Thr14 and 14-3-3-s can promote DNA-damage-induced mito- (mainly by Myt1 kinase) and Tyr15 (mainly by Wee1 tic catastrophe (Bunz et al., 1998; Chan et al., 1999; kinase) inhibits its activity during the G2 phase of the Roninson et al., 2001; Chen et al., 2003). Damage cell cycle, while dephosphorylation of Cdk1 on Thr14 leading to mitotic catastrophe can be induced, in and Tyr15 by a member of the Cdc25 family of particular, by microtubule-hyperpolymerizing agents phosphatases results in the activation of Cdk1 during (taxanes, elutherobins, epothilones, laulimalide, sarco- early mitosis. The phosphorylation of Thr161 by CAK dictyins, docodermolide) and by microtubule-depoly- (Cdk-activating kinase, a heterodimer of cyclin H and merizing agents (such as the Vinca alkaloids, Cdk7) is strictly required for Cdk1 to be active. The cryptophyscins, halichondrins, estramustine and colchi- Cdk1 inhibitor p21Cip1/Waf1 directly inhibits Cdk1 activity. cine) as well as by DNA damage (Roninson et al., 2001). Cyclin B1 translocates from the cytosol to the nucleus, Since cancer cells are frequently deficient in cell-cycle early during mitosis. The equilibrium between nuclear checkpoints, they may be particularly susceptible to the import and export of cyclin B1 (as well as that of induction of mitotic catastrophe by such drugs. Mitotic Cdc25C, required for the activation of Cdk1) is catastrophe can also be induced by fusion of mitotic influenced by its phosphorylation status (Nigg, 2001; cells with interphase cells in S or G2, as a result of pre- Smits and Medema, 2001), as well as by cytoplasmic- mature induction of mitosis before the completion of S binding proteins such as 14-3-3s (Chan et al., 1999). At or G2 (Castedo et al., 2004a). Overduplication of centro- the end of the metaphase, the APC (also called somes (which leads to multipolar mitosis) (Sato et al., ‘cyclosome’) must destroy cyclin B1 to allow mitosis to 2000) or failure of centrosomes to undergo duplica- proceed (Peters, 2002). The APC has a critical E3 tion (with consequent failure of chromosomes to ubiquitin ligase activity and directs the assembly of segregate) (Cogswell et al., 2000) may also lead to multiubiquitin chains on specific substrates and their mitotic catastrophe. subsequent degradation by the 26 S proteasome (Yu, The present review is aimed to examine the question 2002). The APC selects substrates by using the Cdc20 or which are the major processes dictating mitotic Cdh1 adaptor proteins to recognize a specific sequence catastrophe, in molecular and genetic terms. Special (destruction box) in cyclin B1. attention will be paid to the relationship between The Cdk1/cyclin B1 heterodimer induces mitosis by mitotic catastrophe and apoptosis. Based on the phosphorylating and activating enzymes regulating information presently available in the literature, we will chromatin condensation, nuclear membrane break- thus propose a tentative molecular definition of mitotic down, mitosis-specific microtubule reorganization and catastrophe. the actin cytoskeleton allowing for mitotic rounding up

Oncogene Mitotic catastrophe M Castedo et al 2827

Figure 1 Place of the Cdk1/cyclin B complex in mitosis. To allow cells to progress from the G2 to the M phase, the Cdk1/cyclin B complex is active, meaning that Cdk1 has undergone an activating phosphorylation (on Thr161) by CAK, that the inhibitory phosphorylation (on Thr14 and Tyr15) has been removed by active Cdc25C phosphatase, that Cdk1 associates with cyclin B and that the complex translocates to the nucleus. In the nucleus, the active Cdk1/cyclin B complex then phosphorylates mitotic substrate proteins. During the anaphase, the APC becomes activated and destroys Cdk1, a step that is a conditio sine qua non for entry into the anaphase. The DNA structure checkpoint activates checkpoint kinases (such as Chk1 and Chk2), which phosphorylate Cdc25C on Ser216, thereby causing its inactivation (and failure to activate Cdk1). The activation of the spindle checkpoint delays maturation of the APC (and thus prevents cyclin B degradation) of the cell (Nigg, 2001). The perfect regulation of the manipulations, for instance, in human colorectal ade- spatiotemporal pattern of Cdk1/cyclin B1 activity is nocarcinoma cell lines treated with fluorouracil (Yoshi- pivotal for the normal cell cycle and is subject to kawa et al., 2001), 14-3-3sÀ/À colon cancer cells treated multiple control steps. Thus, the so-called ‘DNA with doxorubicin (Chan et al., 1999), as well as in structure checkpoints’ (which are activated by incom- heterokarya induced by cell fusion (Castedo et al., 2001, plete DNA replication or by DNA damage such as 2002b). The inhibition of Cdk1 by the expression of a strand breaks) will stimulate Wee1/Myt1 as well as the dominant-negative (DN) mutant or chemical inhibitors so-called checkpoint kinases (Chk1, Chk2) to prevent such as roscovitine or olomoucine prevents the apopto- Cdk1 activation and entry into mitosis. Similarly, the sis of heterokarya (Castedo et al., 2001), including the APC can be inhibited during the ‘spindle assembly mitotic catastrophe promoted by Chk2 inhibition checkpoint’, thus preventing the degradation of cyclin (Castedo et al., 2004b). Prolonged inhibition of the B1 and securin (and that of many other substrates) and APC (which results into prolonged Cdk1 activation) can nuclear division (Nigg, 2001; Smits and Medema, 2001; also result in mitotic catastrophe associated with Melo and Toczyski, 2002). centrosome overduplication. Thus, a recent study Aberrant mitotic entry, before the completion of indicates that overexpression of Emi1, a protein that DNA replication, can result in mitotic catastrophe. inhibits the APC in S and G2, or failure to degrade Emi1 Obviously, this requires the activation of Cdk1, and it is (via action of the SCFBTCRP/Slimb ubiquitin ligase) currently assumed that premature entry of active Cdk1/ leads to the stabilization of APC substrates and mitotic cyclin B1 complex into the nucleus suffices to cause deficiencies (Margottin-Goguet et al., 2003). premature chromatin condensation and apoptosis An increase in Cdk1 activity has also been found in (Heald et al., 1993; Fotedar et al., 1995; Jin et al., numerous apoptotic conditions, and inhibition of the 1998; Porter et al., 2003). Increased nuclear cyclin B1 Cdk1/cyclin B1 complex by a DN Cdk1 mutant, has been found in numerous examples of mitotic antisense constructs or chemical inhibitors have sug- catastrophe induced by pharmacological or genetic gested that this increase in activity may be indeed

Oncogene Mitotic catastrophe M Castedo et al 2828 important for cell killing (Fotedar et al., 1995; Castedo not incorporated into the cells that form the proper et al., 2002a). In HL60 myelomonocytic leukemia cells, embryo. Disruption of the Drosophila Chk2 homologue taxol-induced Cdk1 activation is not affected by Bcl-2 (DmChk2) blocks this mitotic response to DNA lesions nor is it inhibited by the pancaspase inhibitor Z- and also prevents the loss of defective nuclei from the VAD.fmk, although Bcl-2 (but not Z-VAD.fmk) does cortex (Takada et al., 2003). This phenotype, induced by prevent the mitochondrial release of cytochrome c (Cyt. knock out of the DmChk2 gene, is partially suppressed c) (Ibrado et al., 1998). This places Cdk1 upstream of by reducing the gene dosage of the Plk polo, the only the Bcl-2-regulated mitochondrial changes that define member of the Plk gene family in the fly (Xu and Du, apoptosis (Kroemer and Reed, 2000; Ferri and Kroe- 2003). mer, 2001b). Cdk1 can induce mitochondrial membrane In response to DNA damage, Chk2 localizes to permeabilization (MMP) by phosphorylating the proa- centrosome and spindle microtubules in Drosophilia poptotic Bcl-2 protein family member Bad on Ser128, (Takada et al., 2003), while it colocalizes and interacts thereby causing a loss of interaction between Bad and with Plk1, in centrosomes and the midbody in human cytosolic 14-3-3 proteins (Konishi et al., 2002). As a cells (Tsvetkov et al., 2003), suggesting that Chk2 result, Bad can translocate to mitochondria, where it couples centrosome function and spindle assembly to antagonizes Bcl-2-like proteins and/or activates Bax-like genomic integrity via Plk1. In mammals, Chk2 probably proteins (Debatin et al., 2002) and causes cell death functions as an effector through phosphorylation of (Konishi et al., 2002; Letai et al., 2002). Alternatively, additional key substrates such as the transcription factor the prolonged activation of Cdk1 may increase, via p53 (Chehab et al., 2000; Sheih et al., 2000), the cell- indirect mechanisms, the activity of p53, which causes cycle-regulating phosphatases Cdc25A (Falck et al., the transcription of proapoptotic proteins of the Bcl-2 2000) and Cdc25C (Peng et al., 1997), Mdm2 (Sheih family such as Bax and Puma, thereby initiating et al., 2000), BRCA1 (Lee et al., 2000) and PML (Yang mitochondrial apoptosis (Castedo et al., 2001, 2004b; et al., 2002), and perhaps Abl1, Bub1R, Bub1, Bub3, Perfettini et al., 2004). However, the detailed mechan- Psk-H1, Smc3, Plk1, Cdc25B, Dcamkl1, Mre11, Pms1 isms through which Cdk1 can facilitate apoptosis and Xrcc9 (Seo et al., 2003). In addition, Chk2 induction are far from clear and it is well possible that phosphorylates the transcription factor E2F1 on serine Cdk1 facilitates apoptosis through more indirect me- 364, thereby activating the transcriptional upregulation chanisms, for instance by initiating mitosis, thereby of the proapoptotic proteins Apaf-1 and p73 (Stevens laying the grounds for the forthcoming mitotic cata- et al., 2003). In accordance, with these pleiotropic strophe. effects, Chk2 can have an apoptosis-sensitizing effect and can arrest the cell cycle at different stages (Figure 2). DNA damage sensors and Chk2: avoiding mitotic Recent data imply Chk2 as a negative regulator of catastrophe mitotic catastrophe in human cells. The inhibition of Chk2 prevents the activation of a DNA structure An intact DNA structure checkpoint is important for checkpoint, which normally arrests the cell cycle of avoiding mitotic catastrophe. DNA damage such as nonsynchronous HeLa heterokarya at the G2/M bor- double-strand breaks or pyrimidine dimers is perceived by sensors (which include RAD1, RAD9, RAD17 and HUS7) and then relayed to kinases such as the related phosphatidylinositol-3-OH-kinase-like kinase, ataxia telangiectasia mutated (ATM) and ataxia telangiectasia and Rad3 related (ATR) that ultimately activate, via phosphorylation, yet another kinase, such as the Chk2, which halts the cell cycle. This cascade is conserved throughout evolution. Rad53p, the yeast (S. cerevisiae) Chk2 homologue, is activated by Mec1p, the yeast ATM/ATR homologue, during the S phase checkpoint that avoids entry into mitosis before DNA replication is completed (Allen et al., 1994; Weinert et al., 1994; Clarke et al., 1999). In yeast, Mec1p and Rad53 repress the accumulation of the APC specificity factor Cdc20p, a protein whose overexpression during the S phase suffices to cause catastrophic mitosis (Clarke et al., 2003). Figure 2 Contribution of Chk2 to the regulation of apoptosis. DNA damage occurring during the G1, S or G2 ‘phases can lead to In Drosophila melanogaster, Chk2 is also required for the activation of Chk2, which then acts in a pleiotropic manner to the activation of a DNA damage response. In syncytial phosphorylate the transcription factor E2F1 (which facilitates the Drosophila embryos, damaged or incompletely repli- induction of at least two proapoptotic proteins, Apaf-1 and p73). cated DNA triggers centrosome disruption in mitosis, In addition, Chk2 inhibits Plk1 (which function as an endogenous repressor of mitotic catastrophe) activates two proapoptotic thus resulting in defective spindle assembly and ana- proteins, namely, PML and p53 (by direct phosphorylation or phase chromosome segregation. The damaged nuclei indirectly by inhibiting the p53 antagonist MDM2) and provokes a accumulate in the cortex of the syncytial embryo and are cell-cycle arrest via the phosphatases CDC25A and CDC25C

Oncogene Mitotic catastrophe M Castedo et al 2829 der. In a system of continuous syncytium formation, may actually be attributed to the inhibition of Chk2 induced by the coculture of two different HeLa cell lines (rather than that of other kinases), and whether the expressing the HIV-1 envelope (Env) or CD4, respec- inhibition of Chk2 will not ultimately increase or tively (Ferri et al., 2000a, b), polyploid cells normally accelerate genomic instability, thereby favoring the arrest at the early prophase of the cell cycle, after development of therapy-resistant tumor cells or even dismantling of the nuclear envelope but before spindle secondary cancers. formation, correlating with a sudden loss of cyclin B1 Mutations in human Chk2 increase genomic instabil- and then undergo p53-dependent apoptosis after a ity and are linked to cancer. Thus, heterozygous latency phase (Castedo et al., 2001, 2004b). Cell fusion mutations in Chk2 (e.g. the protein-truncating muta- is also accompanied by the activating phosphorylation tion, 1100delC in exon 10 or the R145W mutation that of Chk2 on Thr68, which colocalizes with BRCA compromises the stability of the Chk2 protein) are nuclear bodies (Castedo et al., 2004b) and phosphory- found in a subset of patients with the multicancer Li– lated histone H2AX (Figure 3). The suppression of Fraumeni syndrome (Bell et al., 1999). The 1000delC Chk2 by the overexpression of a kinase-dead DN Chk2 mutation has also been found in a fraction of patients mutant or pharmacological inhibition with debromohy- with breast cancer (Meijers-Heijboe et al., 2002; menialdisine over-rides this arrest, maintains high cyclin Vahteristo et al., 2002). Patients with bilateral mammary B1 levels and Cdk1 activities, and causes the cells to carcinomas exhibit a sixfold increase in the incidence of proceed until metaphase. Thus Chk2-inhibited hetero- the Chk2 1000delC mutation as compared to patients karya manifest metaphases with multiple spindles. The with unilateral tumors (Vahteristo et al., 2002). Other cells then exhibit signs of apoptosis including Cyt. c germline Chk2 mutations affecting the catalytic region release, caspase-3 activation and TUNEL-detectable or regulatory domains (Wu et al., 2001b) have also been DNA fragmentation (Castedo et al., 2004a, b). The linked to breast cancer (Meijers-Heijboe et al., 2002) and inhibition of Chk2 can also abolish the cell-cycle arrest other malignancies including osteosarcoma (Miller et al., normally induced by doxorubicin treatment in HCT116 2002). A somatic mutation of Chk2 has been detected in cells, thereby facilitating apoptosis induction (Castedo myelodysplastic syndrome (Hofmann et al., 2001). This et al., 2004a, b). In this cell line, the ablation of the 14-3- indicates that the pharmacological suppression of Chk2 3-s locus also abolishes the G2 arrest and causes mitotic might indeed have adverse tumor-promoting effects. catastrophe linked to the premature entry of cyclin B/ In summary, it appears that the DNA structure Cdk1 into the nucleus (Chan et al., 1999). Of note, 14-3- checkpoints and Chk2 negatively regulate mitotic 3-sÀ/À cells manifest a reduced activation of Chk2. catastrophe, meaning that the inactivation of Chk2 Moreover, the inhibition of Chk2 abolishes the pheno- (and of its activating partners) can sensitize to this type typic difference between 14-3-3-s-sufficient and -defi- of cell death. Although attractive, the possibility to cient cells (Castedo et al., 2004b), suggesting that the inhibit Chk2 (and similar enzymes) for chemosensitiza- phenotype of 14-3-3-sÀ/À cells is actually due to an tion has to be validated by extensive biological and inhibition of Chk2 activation. clinical tests. The inhibition of Chk2 (and Chk1) by 7-hydroxys- taurosporine (UCN-01) also abrogates the g-radiation- Survivin at the interface between mitotic checkpoint induced G2–M checkpoint and thus enhances g-radia- control and apoptosis suppression? tion-induced apoptosis (Yu et al., 2002). Clinical trials involving a combination of UCN-01 and DNA-dama- In the context of the specific topic of this review, ging agents are underway and have yielded encouraging survivin appears to be a particularly fascinating protein, results for patients with melanoma and refractory for at least three reasons. First, survivin is frequently anaplastic large-cell lymphoma (Senderowicz, 2003). overexpressed in cancers, correlating with abbreviated However, it is not clear which of the effects of UCN-01 survival, unfavorable prognosis, resistance to therapy and accelerated rates of recurrences (Altieri, 2003). Second, survivin is a member of the inhibitor of apoptosis protein gene family and has been postulated to suppress caspase activation (Li et al., 1998). Third, survivin demonstrates a particular cell-cycle-dependent pattern of subcellular distribution, is a substrate of Cdk1, and actually contributes to the spindle checkpoint (Lens et al., 2003). The mitotic checkpoint pathway is regulated by a group of evolutionarily conserved genes that include mitotic arrest deficient MAD1, MAD2, MAD3 (or Figure 3 Colocalization of activated Chk2 and DNA lesions in BUBR1), BUB1, BUB3 and MPS1 (Burke, 2000). These nonsynchronized HeLa syncytia. HeLa syncytia were stained by proteins preferentially bind to the kinetochores of immunofluorescence methods for the detection of phosphorylated unattached chromosomes where they are thought to Chk2 (on Thr 68) and phosphorylated histone H2AX (on Ser 139). Note the colocalization of the two phosphorylation events in foci, generate the signal that suppresses anaphase onset within the nucleus. Chromatin was visualized with Hoechst 33324 (Burke, 2000; Shah and Cleveland, 2000). Treatment counterstaining (blue color in inset) of cells with chemicals that perturb microtubule

Oncogene Mitotic catastrophe M Castedo et al 2830 dynamics will lead to sustained activation of the spindle drial pathway of apoptosis. How this inhibitory effect is checkpoint and often triggers apoptosis. In HeLa cells achieved remains obscure. Direct regulation of caspase treated with nocodazole, activation of p38 has been activity by survivin has been proposed (Li et al., 1999), linked to apoptosis induction. p38 activation would be but later disputed (Verdecia et al., 2000). In addition, it required for the translocation of Bax to mitochondria has been reported that survivin interacts with Smac/ (Deacon et al., 2003). In addition, overexpression of the DIABLO (Song et al., 2003) and it has been suggested MAD checkpoint proteins has also been reported to that this interaction would prevent the activation of sustain the activity of the spindle checkpoint and causes downstream caspases, at the postmitochondrial level. a mitotic arrest that is frequently followed by apoptosis However, this would not explain how survivin can (Geley et al., 2001). Moreover, depletion of CMT2 (an regulate MMP and caspase-independent cell death. MAD2-binding protein) (Habu et al., 2002) or hNuf2 Thus, it is presently unclear if and how the mitotic roles (DeLuca et al., 2002) leads to a blockade in the of survivin relate to its proposed role as an inhibitor of prometaphase/anaphase transition with consequent apoptosis, but it remains tempting to speculate that death. survivin play an important role in the coupling of At the beginning of mitosis, survivin associates with abnormal mitotic progression to the onset of mitotic microtubules of the mitotic spindle, forming a complex catastrophe. with INCENP (inner centromere-binding protein) and et al et al the Aurora B kinase (O’Connor ., 2002; Wall ., p53 and the polyploidy checkpoint 2003). The knock out of survivin and that of its binding partner INCENP has a similar phenotype, with dis- Upon spindle damage, cells become transiently arrested rupted microtubule formation, polyploidy and massive at the metaphase–anaphase transition point and then apoptosis, as well as embryonic cell death on day 4.5 escape from the block (a process termed ‘mitotic (Cutts et al., 1999; Uren et al., 2000). Recent studies slippage’) and exit mitosis without proper segregation provide evidence that survivin is required for a sustained of sister chromatids and cytokinesis. Spindle-damaged mitotic arrest in response to the lack of tension at the cells then definitively arrest at a tetraploid G1 state kinetochore (Carvalho et al., 2003; Lens et al., 2003). because they activate a checkpoint. This arrest is Indeed, when survivin is depleted by small interfering mediated through p53, while the transient arrest at the RNA (siRNA), cells manifest a prometaphase delay metaphase–anaphase transition and mitotic slippage are (B3–4h) and then exit mitosis without attaining probably not influenced by p53 (Lanni and Jacks, 1998), chromosome alignment while reforming a single tetra- although some reports have suggested that the absence ploid nucleus. In survivin-depleted cells, BubR1 is of p53 might increase mitotic slipperage, thus increasing prematurely displaced from kinetochores, and the the frequency of tetraploid cells (Tsuiki et al., 2001). prometaphase delay induced by survivin depletion is There is a broad consensus that p53-deficient cells that completely abrogated when Mad2 and BubR1 are are in the tetraploid G1 state are not prevented from re- codepleted (Lens et al., 2003). Survivin depletion also entering the cell cycle to reduplicate their DNA causes a failure of taxol-induced mitotic arrest (Carval- unchecked, leading to polyploidy and subsequent ho et al., 2003; Lens et al., 2003). Clearly, in these latter chromosomal instability. In the presence of p53, studies depletion of survivin led to a prominent defect in tetraploidy (or polyploidy in general) causes the activa- mitosis, without any apparent direct induction of tion of p21 and an irreversible arrest in the cell cycle (or apoptosis. This would suggest that a role for survivin cell death), thereby preventing the propagation of errors as a protection from apoptosis is rather secondary to its of late mitosis and the generation of aneuploidy (Lanni role during mitosis. and Jacks, 1998; Andreassen et al., 2001b). The absence Transgenic expression of survivin has been reported of p21 also relaxes the polyploidy checkpoint and thus to prevent apoptosis induction by mitochondrion- cause hematopoietic MO7e cells to manifest centriole dependent cell death inducers such as UVB light overduplication with polyploidy and multilobular nuclei (Grossman et al., 2001). Survivin antisense oligonucleo- in response to nocodazole (Mantel et al., 1999), tides administered to a human neuroblastoma (MSN) suggesting that p21 is one of the major p53 target genes cell line cause caspase-independent cell death (with in this context. However, it has been reported that a p53 multiple large macronuclei with no apoptotic bodies) mutant that is transcriptionally silent (p53-135Val) can accompanied by the mitochondrionuclear translocation replace wild-type p53 in the inhibition of polyploidy of apoptosis-inducing factor (AIF) (Shankar et al., (Notterman et al., 1998), suggesting that the functional 2001), an important caspase-independent death effector requirements for p53 in diploid cells are different from (Cande et al., 2002; Penninger and Kroemer, 2003). In those mediating the arrest of tetraploid cells. Impor- contrast, the same survivin antisense oligonucleotide can tantly, the overexpression of Bcl-2 allows polyploid p53- induce caspase activation and full-blown chromatin deficient cells, generated by nocodazole treatment, to condensation in the oligodendroglioma (TC620) line survive (Minn et al., 1996). Thus, p53 abrogation and (Shankar et al., 2001). In HL60 cells, survivin antisense apoptosis inhibition can cooperate to induce rapid and induces the release of mitochondrial apoptosis factors progressive polyploidization following mitotic spindle including Cyt. c, AIF and Smac/DIABLO (Carter et al., damage. 2003). Altogether, these data suggest that survivin The absence of functional p53 is permissive for the functions as an endogenous repressor of the mitochon- generation/survival of polyploid cells in a number of in

Oncogene Mitotic catastrophe M Castedo et al 2831 vitro systems; HeLa cells proliferating in vitro without Importantly, the loss of p53 can sensitize cells to any treatment other than antisense p53 (Iotsova and microtubule poisons such as paclitaxel (Taxol) (Haw- Stehelin, 1995), human leukemia cell lines treated with kins et al., 1996; Wahl et al., 1996; Vikhanskaya et al., nocodazole ((Verdoodt et al., 1999), U87MG glioma 1998) due to its capacity of inactivating the polyploidy cells exposed to 1,3-bis(2choroethyl)-1-nitrosourea (Wu checkpoint. After treatment with spindle inhibitors, cells et al., 2001a), Msh2-deficient fibroblasts treated with possessing an intact p53 system undergo a p21-mediated cisplatin (Strathdee et al., 2001), the 32D murine G1 arrest after adaptation and thus are resistant to myeloid cell line manipulated to overexpress c-myc antimicrotubule agents. Cells deficient for p53 fail to (Yin et al., 1999), murine T cells subjected to mitotic undergo such a G1 arrest (since they fail to upregulate arrest (Baek et al., 2003), or mouse embryonic p21) and endoreduplicate their DNA, leading to massive fibroblasts treated with IC261, an inhibitor of casein apoptosis, mainly during the S phase (Minn et al., 1996; kinase-1d and -e (Behrend et al., 2000), or fused with Di Leonardo et al., 1997; Lanni and Jacks, 1998). polyethylene glycol (Castedo et al., 2001), just to give a Recently, it has been found that there is an additional few examples. The overexpression of Aurora-A, Aurora- mechanism through which the absence of p53 can B and Plk causes cell division defects with consequent sensitize to paclitaxel-induced cell death, namely via multinucleation (mostly tetraploidization) and an in- Plk2 (Burns et al., 2003), as discussed in the following crease in centrosome number. This phenotype is section. exacerbated in p53-negative cells, suggesting that the inability to detect hyperploidy favors the emergence of Plks in mitotic catastrophe numerical chromosome aberrations (Meraldi et al., 2002). Another cancer-relevant example is provided by One of the p53-induced genes is Plk2 (also called serum- endopolyploid cells generated by the expression of K- inducible kinase, Snk). Knock down of Plk2 using cyclin (the Kaposi’s sarcoma-associated herpesvirus siRNA sensitizes cells to apoptosis induction upon cyclin D homologue). Such cells manifest multinuclea- treatment with paclitaxel or nocodazole. This cell death tion, polyploidy, centrosome amplification and conse- occurs during mitosis, in cells containing 4N of quent aneuploidy, provided that p53 is inactivated chromosomes and high cyclin B1 levels. It is accom- (Verschuren et al., 2002). When p53 is present, it is panied by caspase-3 activation (Burns et al., 2003), activated at the stage of polyploidy and causes cell death indicating that it involves the activation of the apoptotic (Verschuren et al., 2002). It is possible that the down- machinery. It thus has been suggested that Plk2 would regulation of p53 by overexpressed MDM2 is also be an endogenous inhibitor of mitotic catastrophe permissive for the physiological polyploidization of whose expression is under the control of p53. megakaryocytes (Datta and Long, 2002). Indeed, Plk2 is just one member of a larger family of kinases megakaryocyte differentiation is tightly linked to that also comprises Plk1 and Plk3. The inhibition of endoreduplication cycles with aborted (metaphase- Plk1 with siRNA results in spindle defects, arrest at arrested) mitoses of a transcriptionally active genome prometaphase and cytokinesis defects resulting into (Raslova et al., 2003), and this correlates with a decrease apoptosis (Spankuch-Schmitt et al., 2002; Liu and in p53 protein levels. Erikson, 2003). Similar effects were found when tumor That p53 can actually suppress polyploidy is also cells were treated with a peptide corresponding to the suggested by a series of observations in vivo, in the amino acids 410–429 of Plk1 fused to an Antennapedia mouse. Thus, the inactivation of p53 by SV40 transgenes peptide that renders it plasma membrane permeable. can lead to tetraploidy in the mouse pancreas (Ramel Upon the addition of this peptide (which is likely to et al., 1995) as well as to cytological abnormalities act as an inhibitor of endogenous Plk1), MCF-7 or (anisokaryosis and anisocytosis) in the mouse liver HeLa S3 cells manifest mitotic alterations with multi- (Gillet et al., 2000). This latter effect is prevented by polar spindles and apoptosis (Yuan et al., 2002). The the overexpression of transgenic wild-type p53 (Gillet overexpression of a DN Plk1 has also been reported et al., 2000). p53 knockout mice manifest the accumula- to induce a G2/M arrest, mitotic alterations (but tion of multinuclear cells in the prostate (Colombel et al., with frequent monoastral mitoses) and apoptosis 1995) and in the testis (Rotter et al., 1993). The absence (Cogswell et al., 2000). An antisense approach designed of p53 allows for the persistence of multinuclear cells to reduce the expression of Plk1 has therapeutic effects upon wound healing in vivo (Oberringer et al., 1999). on human tumors xenotransplanted into immunodefi- The link between p53 abrogation and aneuploidy has cient mice, in particular, when combined with antisense also been established in a series of human cancers. Thus, Bcl-2 (Elez et al., 2003), suggesting that it does involve 17p (p53) allelic losses occur in diploid cells of patients apoptosis. with Barrett’s esophagus, before aneuploidy becomes The overexpression of a kinase-defective mutant of manifest (Blount et al., 1994). In head and neck cancer, Plk3 (Plk3K52R, which may well behave like a DN a close spatial correlation between p53 mutation and Plk3) also arrests mammalian cells in the G2/M phase increased chromosome polysomy has been observed and kills them, accompanied by the accumulation of (Shin et al., 2001). Generally, it appears that p53 cyclin B1 and the caspase-cleaved PARP p85 fragment mutation predisposes to chromosomal instability, per- (Wang et al., 2002). These death-inducing effects are haps due to deficient activation of the polyploidy found even in the absence of stress (such as microtubule checkpoint. poisons), indicating that Plk1 and Plk3 (whose down-

Oncogene Mitotic catastrophe M Castedo et al 2832 regulation/inhibition kills in the absence of stress) and underlining the importance of the apoptotic machinery Plk2 (whose downregulation only kills in the presence of (and in particular of the intrinsic, mitochondrial path- antimicrotubule agents) are not functionally equivalent. way) for the execution of mitotic catastrophe. In accord However, it remains an open conundrum whether with this notion, it has also been shown that inhibition aberrant mitosis (facilitated by the depletion of Plks) of Bcl-2 expression by antisense oligonucleotides can simply induced apoptosis through identical mechanisms, actually facilitate and amplify mitotic catastrophe (Elez without any direct impact of these kinases on the et al., 2003). apoptotic machinery. If mitotic catastrophe is characterized by MMP and caspase activation, what are then the molecular links between the upstream events that link abnormal Mitotic catastrophe and the apoptotic machinery mitosis to the activation of the central executioner of apoptosis? It probably depends on the subtype of As discussed above, mitotic catastrophe can be con- cell death. In the case of p53-dependent mitotic sidered as a type of cell death occurring during mitosis catastrophe, solid evidence is available that the tran- or resulting from mitotic failure. At least two subtypes scription of some proapoptotic p53 target genes may of mitotic catastrophe can be distinguished. First, determine the induction of MMP. Thus, Bax and mitotic catastrophe can kill the cell during or close to Puma have been identiﬁed as the p53 target genes that the metaphase, in a p53-independent manner, as this induce apoptosis during the prophase when nonsyn- occurs in Chk2-inhibited syncytia or Plk2-depleted cells. chronized cells are fused among each other (Roumier Second, mitotic catastrophe can occur after failed et al., 2003; Perfettini et al., 2004). The inhibition of mitosis, during the activation of the polyploidy check- these proteins by antisense oligonucleotides or siRNA point, in a partially p53-dependent manner. prevents the prophase-associated death induced by Mitotic catastrophe is accompanied by chromatin fusion of HeLa cells. The inhibition of p53, Puma condensation (Figure 4a), mitochondrial release of or Bax thus stabilizes the DCm, prevents the release proapoptotic proteins (in particular Cyt. c and AIF) of Cyt. c and AIF from mitochondria and suppres- (Figure 4b and c), caspase activation (Figure 4d) and ses caspase-3 activation (Perfettini et al., 2004). This DNA degradation (Figure 4c). This implies that mitotic indicates that caspase-3 activation occurs downstream catastrophe is accompanied by the key molecular events of mitochondria, presumably as a result of Cyt. c release deﬁning apoptosis, namely, capase activation and MMP and the formation of the Cyt. c/Apaf-1/caspase-9 (Green and Kroemer, 1998; Zamzami and Kroemer, apoptosome complex. Conversely, chemical inhibition 2003). The inhibition of caspases can prevent or at least of caspases with Z-VAD.fmk does not affect the retard the metaphase-associated cell death (not shown). signs of MMP (Ferri et al., 2000a), thus corroborating Importantly, in several paradigms, transfection with the notion that caspase activation is occurring down- MMP inhibitors (Zamzami et al., 1998) such as Bcl-2, stream of MMP, in a model of syncytium-induced Bcl-XL or the Cytomegalovirus vMIA gene (also called apoptosis. UL37), or knock out of Bax can prevent mitotic In the case of metaphase-associated death, which is catastrophe. This has been shown for the two subtypes p53 independent, it appears that one particular caspase of mitotic catastrophe mentioned above (Minn et al., is activated upstream of mitochondria. In Chk2- 1996; Ferri et al., 2000a; Castedo et al., 2001, 2004a, b; inhibited syncytia, caspase-2 activation can be detected Roumier et al., 2003; Perfettini et al., 2004), again even when apoptosis is suppressed by overexpression

Figure 4 Mitotic catastrophe demonstrates signs of apoptosis. HeLa cell syncytia generated in the presence of a chemical Chk2 inhibitor, debromohymenialdesine, undergo apoptosis during the metaphase of the cell cycle. During this stage, the manifest chromatin condensation (as determined by staining with Hoechst 33324) with microtubular dysorganization (as determined by immunostaining with an antitubulin antibody) (a), mitochondrial release of Cyt. c (b) and AIF (c), both detected with speciﬁc antibodies, caspase-3 activation, as determined with an antiserum that recognizes the proteolytically mature caspase-3, yet does not react with inactive procaspase-3 (d), as well as DNA fragmentation as detectable with the TUNEL method (e). In each case, a representative metaphase is shown before and after mitotic catastrophe has occurred

Oncogene Mitotic catastrophe M Castedo et al 2833 of vMIA (Castedo et al., 2004a). Similarly, histone conundrum, however. It has been shown in the past that deacetylase inhibitors such as suberoylanilide hydro- procaspase-2 can be activated in the nucleus as a result xamic acid, oxamﬂatin and depsipeptide can trigger the of DNA damage (Lassus et al., 2002; Paroni et al., 2002; accumulation of cells with a 4N DNA content and Read et al., 2002; Robertson et al., 2002). Moreover, it subsequent apoptosis, and this process involves a has been shown that caspase-2 can act on isolated primary caspase-2 activation that cannot be inhibited mitochondria to trigger MMP, but the mitochondrial by Bcl-2 (Peart et al., 2003). Importantly, in the model target protein of caspase-2 is yet to be identiﬁed (Guo of Chk2-inhibited syncytia, caspase-2 depletion by et al., 2002). Irrespective of these details, it appears RNAi prevents MMP. Both the depletion of caspase-2 however that, in the model of Chk2-inhibited syncytia, and the direct inhibition of MMP (by vMIA) inhibit the pancaspase inhibitors (which also act on caspase-2) can proteolytic maturation of caspase-3 and subsequent actually prevent MMP and mitotic cell death (Castedo mitotic death, indicating that the molecular order in this et al., 2004a), thereby preventing a cell death occurring kind of metaphase-associated mitotic catastrophe is during the metaphase and ultimately inducing the caspase-2-MMP-caspase-3 (Castedo et al., 2004a). asymmetric division of polyploid cells in more than How caspase-2 is activated in this system remains a two daughter cells (Figure 5).

Figure 5 Mitotic catastrophe and its suppression. HeLa cells stably transfected with a histone H2B-GFP construct (which allows for the monitoring of chromosome movement) were fused with PEG, followed by ﬂuorescence videomicroscopy (a–c). Multipolar divisions of syncytia are ﬁlmed. In the presence of the Chk2/Chk1 inhibitor debromohymenialdesine only (a and b), syncytia progress from initial nuclear fusion (karyogamy) to a multipolar metaphase (a) and then to mitotic catastrophe (b). When the pancaspase inhibitor Z-VAD.fmk is added into the system, metaphases are followed by full karyokinesis and cytokinesis, resulting into asymmetric division (c). (d) Schematic overview on the cellular events following syncytium formation (cytogamy). Karyogamy leads to apoptosis, in conditions in which Chk2 is activated. If Chk2 is inhibited, cells progress to the metaphase and undergo mitotic catastrophe unless apoptosis is inhibited, in which case the cells divide asymmetrically

Oncogene Mitotic catastrophe M Castedo et al 2834 Mitotic catastrophe, cytogenetic catastrophe and cancer: Masuda and Takahashi, 2002a, b; Sen et al., 2002; Tort a hypothesis et al., 2003). On theoretical grounds, polyploid cells (which possess more than two centrosomes) should The data discussed in this review suggest under which divide in an asymmetric manner, thus generating conditions mitotic catastrophe is induced. After failure aneuploid cells, containing an excess of chromosomes to activate the G2/M checkpoint(s), cells with DNA or lacking entire chromosomes (Brinkley, 2001; Nigg, lesions (or incomplete DNA replication) activate an 2001; D’Assoro et al., 2002; Lingle et al., 2002; Meraldi apoptotic program that leads to the phenotypic mani- et al., 2002). If chromosome segregation starts at a stage festation of mitotic catastrophe, during the metaphase at which the replication of all chromosomes has not yet of the cell cycle. Suppression of the apoptotic program been completed (which normally is prevented by (and of the spindle checkpoint) then may lead to checkpoints) (Cahill et al., 1998), problems of separa- asymmetric cell division (Figure 5d) or mitotic slippage, tion, possibly linked to DNA strand breaks and repair resulting into the generation of tetraploid cells which, would lead, in addition, to partial deletions of chromo- when the polyploidy checkpoint is inactivated, can somes as well as to translocations (Gisselsson, 2003). generate an aneuploid offspring (Figure 6). It is Most of the daughter cells generated as the result of a tempting to speculate that these mechanisms could be stochastic distribution of chromosomes should be important for cancer development. doomed to death in the absence of genes coding for Most tumors develop in an (oligo)clonal and stochas- essential proteins. However, a small minority of tic manner, through a multistep process. As a working abnormal cells could survive such a catastrophic event. hypothesis, it is tempting to assume that one of the The scenario outlined above thus would develop in mechanisms that contribute to oncogenesis may be what several successive steps: (i) the polyploidization resulting we might call a ‘cytogenetic catastrophe’. Aneuploidiza- from illicit cellular fusion or endomitosis; (ii) the tion could result from the asymmetric division of survival of such polyploid cells due to a suppression of polyploid cells, generated from an illicit cell fusion, as the polyploidy checkpoint (which is executed at least in it may occur in vivo (DeWitt and Knight, 2002; Terada part by apoptosis); (iii) the asymmetric division of et al., 2002; Wurmser and Gage, 2002; Vassilopoulos polyploid cells in daughter cells, perhaps resulting from et al., 2003) or from endoreplication/endomitosis (Nitta the suppression of mitotic catastrophe; (iv) the Darwi- et al., 2003). Indeed, polyploidy is frequently observed in nian survival of the daughter cells. The resulting neoplasia and constitutes a negative prognostic factor, ‘cytogenetic catastrophe’ thus is a theoretical recon- while aneuploidy is a near-to-general characteristic of struction of the stochastic and clonal nature of tumor cancer (Slovak et al., 2000; Choma et al., 2001; Jallepalli cells that would develop in the context of simultaneous and Lengaue, 2001; Shih et al., 2001; Lingle et al., 2002; suppression of cell-cycle checkpoints and apoptosis.

Figure 6 Apoptosis resulting from defective or dysregulated mitosis. (a) Normal cell cycle. (b) Normal consequence of DNA damage. DNA damage either results into immediate arrest in the interphase of the cell cycle or can cause a transient entry into the prophase (with reversible chromatin condensation), with a prolonged ‘antephase’. (c) Failure to activate the checkpoints, including the G2/M checkpoint causes entry into mitosis and death during the metaphase or, in the case of mitotic slippage and consequent mitotic failure, generation of tetraploid cells that undergo apoptosis or generate aneuploid cells

Oncogene Mitotic catastrophe M Castedo et al 2835 Abbreviations European Commission (QLG1-CT-1999-00739 and Contract AIF, apoptosis-inducing factor; Cyt. c, cytochrome c. No. QLK3-CT-20002-01956) (to GK).

Acknowledgements This work has been supported by a special grant from LNC, as well as grants from ANRS, FRM, French Ministry of Science,

References

Allen JB, Zhou Z, Sied W, Friedberg EC and Elledge SJ. Chehab NH, Malikzay A, Appel M and Halazonetis TD. (1994). Genes Dev., 8, 2401–2415. (2000). Genes Dev., 14, 278–288. Altieri DC. (2003). Adv. Cancer Res., 88, 31–52. Chen Z, Xiao Z, Chen J, Ng SC, Sowin T, Sham H, Rosenberg Andreassen PR, Lacroix FB, Lohez OD and Margolis RL. S, Fesik S and Zhang H. (2003). Mol. Cancer Ther., 2, (2001a). Cancer Res., 61, 7660–7668. 543–548. Andreassen PR, Lohez OD, Lacroix FB and Margolis RL. Choma D, Daures JP, Quantin X and Pujol JL. (2001). Br. J. (2001b). Mol. Cell. Biol., 12, 1315–1328. Cancer, 85, 14–22. Ayscough K, Hayles J, MacNeill SA and Nurse P. (1992). Mol. Clarke DJ, Segal M, Andrews CA, Rudyak SG, Jensen S, Gen. Genet., 232, 344–350. Smith K and Reed SI. (2003). Nat. Cell Biol., 5, 928–935. Baek KH, Yoo JK, Cho JH, Choi YH, Sung YC, McKeon F Clarke DJ, Segal M, Mondesert G and Reed SI. (1999). Nat. and Lee CW. (2003). J. Leukocyte Biol., 73, 850–861. Cell Biol., 3, 619–627. Behrend L, Milne DM, Stoter M, Deppert W, Campbell Cogswell JP, Brown CE, Bisi JE and Neill SD. (2000). Cell LE, Meek DW and Knippschild U. (2000). Oncogene, 19, Growth Differ., 11, 615–623. 5303–5313. Colombel M, Radvanyi F, Blanche M, Abbou C, Buttyan R, Bell DW, Varley JM, Szydlo TE, Kang DH, Wahrer DC, Donehower LA, Chopin D and Thiery JP. (1995). Oncogene, Shannon KE, Lubratovich M, Verselis SJ, Isselbacher KJ, 10, 1269–1274. Fraumeni JF, Birch JM, Li FP, Garber JE and Haber DA. Cutts SM, Fowler KJ, Kile BT, Hii LL, O’Dowd RA, Hudson (1999). Science, 286, 2528–2531. DF, Saffery R, Kalitsis P, Earle E and Choo KH. (1999). Blount PL, Galipeau PC, Sanchez CA, Neshat K, Levine DS, Hum. Mol. Genet., 8, 1145–1155. Yin J, Suzuki H, Abraham JM, Meltzer SJ and Reid BJ. D’Assoro AB, Lingle WL and Salisbury JL. (2002). Oncogene, (1994). Cancer Res., 54, 2292–2295. 21, 6146–6153. Brinkley BR. (2001). Trends Cell Biol., 11, 18–21. Datta NS and Long MW. (2002). Exp. Hematol., 30, 158–165. Bunz F, Dutriaux A, Lengauer C, Waldman T, Zhou S, Brown Deacon K, Mistry P, Chernoff J, Blank JL and Patel R. (2003). JP, Sedivy JM, Kinzler KW and Vogelstein B. (1998). Mol. Cell. Biol., 14, 2071–2087. Science, 282, 1497–1501. Debatin KM, Poncet D and Kroemer G. (2002). Oncogene, 21, Burke DJ. (2000). Curr. Opin. Genet. Dev., 10, 26–31. 8786–8803. Burns TF, Fei P, Scata KA, Dicker DT and El-Deiry WS. DeLuca JG, Moree B, Hickey JM, Kilmartin JV and Salmon (2003). Mol. Cell. Biol., 23, 5556–5571. ED. (2002). J. Cell Biol., 159, 549–555. Cahill DP, Lengauer C, Yu J, Riggins GJ, Willson JK, DeWitt N and Knight J. (2002). Nature, 416, 354. Markovitz SD, Kinzler KW and Vogelstein B. (1998). Di Leonardo A, Khan SH, Linke SP, Greco V, Seidita G and Nature, 392, 300–303. Wahl GM. (1997). Cancer Res., 57, 1013–1019. Cande C, Cecconi F, Dessen P and Kroemer G. (2002). J. Cell Driessens G, Harsan L, Robaye B, Waroquier D, Browaeys P, Sci., 115, 4727–4734. Giannakopoulos X, Velu T and Bruyns C. (2003). Br. J. Carter BZ, Wang RY, Schober WD, Milella M, Chism D and Cancer, 18, 727–729. Andreeff M. (2003). Cell Cycle, 2, 488–493. Elez R, Piiper A, Kronenberger B, Kock M, Brendel M, Carvalho A, Carmena M, Sambade C, Earnshaw WC and Hermann E, Pliquett U, Neumann E and Zeuzem S. (2003). Wheatley SP. (2003). J. Cell Sci., 116, 2987–2998. Oncogene, 22, 69–80. Castedo M, Ferri KF, Blanco J, Roumier T, Larochette N, Falck J, Maitland N, Syljuasen RG, Bartek J and Lukas J. Barretina J, Amendola A, Nardacci R, Metivier D, Este JA, (2000). Nature, 410, 842–847. Piacentini M and Kroemer G. (2001). J. Exp. Med., 194, Ferri KF, Jacotot E, Blanco J, Este´ JA, Zamzami A, Susin SA, 1097–1110. Brothers G, Reed JC, Penninger JM and Kroemer G. Castedo M, Perfettini J-L, Roumier T and Kroemer G. (2000a). J. Exp. Med., 192, 1081–1092. (2002a). Cell Death Differ., 9, 1287–1293. Ferri KF, Jacotot E, Geuskens M and Kroemer G. (2000b). Castedo M, Perfettini J-L, Roumier T, Valent A, Raslova H, Cell Death Differ., 7, 1137–1139. Yakushijin K, Horne DA, Feunteun J, Lenoir G, Vainch- Ferri KF and Kroemer G. (2001a). Nat. Cell Biol., 2, E63–E64. enker W and Kroemer G. (2004a). Oncogene, (in press). Ferri KF and Kroemer GK. (2001b). Nat. Cell Biol., 3, E255– Castedo M, Perfettini J-L, Roumier T, Yakushijin K, Horne E263. DA, Medema R and Kroemer G. (2004b). Oncogene, Fotedar R, Flatt J, Gupta S, Margolis RL, Fitzgerald P, (in press). Messier H and Fotedar A. (1995). Mol. Cell. Biol., 15, Castedo M, Roumier T, Blanco J, Ferri KF, Barretina J, 932–942. Andreau K, Perfettini J-L, Armendola A, Nardacci R, Geley S, Kramer E, Gieffers C, Gannon J, Peters JM and Hunt LeDuc P, Ingber DE, Este JA, Modjtahedi N, Piacentini M T. (2001). J. Cell Biol., 153, 137–148. and Kroemer G. (2002b). EMBO J., 21, 4070–4080. Gillet R, Grimber G, Bennoun M, Caron de Fromentel C, Chakrabarti A and Chakrabarti S. (1987). Neoplasma, 34, Briand P and Joulin V. (2000). Oncogene, 19, 3498–3507. 557–562. Gisselsson D. (2003). Adv. Cancer Res., 97, 1–29. Chan TA, Hermeking H, Lengauer C, Kinzler KW and Green DR and Kroemer G. (1998). Trends Cell Biol., 8, Vogelstein B. (1999). Nature, 401, 616–620. 267–271.

Oncogene Mitotic catastrophe M Castedo et al 2836 Grossman D, Kim PJ, Blanc-Brude OP, Brash DE, Tognin S, Masuda A and Takahashi T. (2002b). Oncogene, 21, 6884– Marchisio PC and Altieri DC. (2001). J. Clin. Invest., 108, 6897. 991–999. Meijers-Heijboe H, van den Ouweland A, Klijn J, Wasielewski Guo Y, Srinivasula SM, Druilhe A, Fernandes-Alnemri T and M, de Snoo A, Oldenburg R, Hollestelle A, Houben M, Alnemri ES. (2002). J. Biol. Chem., 277, 13430–13437. Crepin E, van Veghel-Plandsoen M, Elstrodt F, van Duijn Habu T, Kim SH, Weinstein J and Matsumoto M. (2002). C, Bartels C, Meijers C, Schutte M, McGuffog L, Thompson EMBO J., 21, 6419–6428. D, Easton D, Sodha N, Seal S, Barfoot R, Mangion J, Hawkins DS, Demers GW and Galloway DA. (1996). Cancer Chang-Claude J, Eccles D, Eeles R, Evans DG, Houlston R, Res., 56, 892–898. Murday V, Narod S, Peretz T, Peto J, Phelan C, Zhang HX, Heald R, McLoughlin M and McKeon F. (1993). Cell, 74, Szabo C, Devilee P, Goldgar D, Futreal PA, Nathanson KL, 463–474. Weber B, Rahman N and Stratton MR. (2002). Nat. Genet., Hofmann WK, Miller CW, Tsukasaki K, Tavor S, Ikezoe T, 31, 55–59. Hoelzer D, Takeuchi S and Koeffler HP. (2001). Leukemia Melo J and Toczyski D. (2002). Curr. Opin. Cell Biol., 14, Res., 25, 333–338. 237–245. Ianzini F and Mackey MA. (1997). Int. J. Radiat. Biol., 72, Meraldi P, Honda R and Nigg EA. (2002). EMBO J., 21, 409–421. 483–492. Ibrado AM, Kim CN and Bhalla K. (1998). Leukemia, 12, Miller CW, Ikezoe T, Krug U, Hofmann WK, Tavor S, 1930–1936. Vegesna V, Tsukasaki K, Takeuchi S and Koeffler HP. Iotsova V and Stehelin D. (1995). Eur. J. Cell Biol., 68, (2002). Genes Chromosomes Cancer, 33, 17–21. 122–132. Minn AJ, Boise LH and Thompson CB. (1996). Genes Dev., Ivanov A, Cragg MS, Erenpreisa J, Emzinsh D, Lukman H 10, 2621–2631. and Illidge TM. (2003). J. Cell Sci., 116, 4095–4106. Molz L, Boohe R, Young P and Beach D. (1989). Genetics, Jallepalli PV and Lengaue C. (2001). Nat. Rev. Cancer, 1, 773–782. 109–117. Nabha SM, Mohammad RM, Dandashi MH, Coupaye- Jin P, Hardy S and Morgan DO. (1998). J. Cell Biol., 141, Gerard B, Aboukameel A, Pettit GR and Al-Katib AM. 875–885. (2002). Clin. Cancer Res., 8, 2735–2741. Joza N, Susin SA, Daugas E, Stanford WL, Cho SK, Li CYJ, Nigg EA. (2001). Nat. Rev. Mol. Cell. Biol., 2, 21–32. Sasaki T, Elia AJ, Cheng H-YM, Ravagnan L, Ferri KF, Nitta M, Tsuiki H, Arima Y, Harada K, Nishizaki T, Sasaki Zamzami N, Wakeham A, Hakem R, Yoshida H, Kong K, Mimori T, Ushio Y and Saya H. (2003). Genes Cells, 7, Y-Y, Zun˜iga-Pflu¨cker JC, Kroemer G and Penninger JM. 151–162. (2001). Nature, 410, 549–554. Notterman D, Young S, Wainger B and Levine AJ. (1998). Kerr JFR, Wyllie AH and Currie AR. (1972). Br. J. Cancer, Oncogene, 17, 2743–2751. 26, 239–257. O’Connor DS, Wall NR, Porter AC and Altieri DC. (2002). Konishi Y, Lehtinen M, Donovan N and Bonni A. (2002). Cancer Cell, 2, 43–54. Mol. Cell, 9, 1005–1016. Oberringer M, Lothschutz D, Jennewein M, Koschnick M, Kroemer G and Reed JC. (2000). Nat. Med., 6, 513–519. Mutschler W and Hanselmann RG. (1999). Mol. Cell. Biol. Lanni JS and Jacks T. (1998). Mol. Cell. Biol., 18, 1055– Res. Commun., 2, 190–196. 1064. Paroni G, Henderson C, Schneider C and Brancolini C. (2002). Lassus P, Opitz-Araya X and Lazebnik Y. (2002). Science, J. Biol. Chem., 277, 15147–15161. 297, 1352–1354. Peart MJ, Tainton KM, Ruefli AA, Dear AE, Sedelies KA, Lee J-S, Collin KM, Brown AL, Lee C-H and Chung JH. O’Reilly LA, Waterhouse NJ, Trapani JA and Johnstone (2000). Nature, 404, 201–204. RW. (2003). Cancer Res., 63, 4460–4471. Lens SM, Wolthuis RM, Klompmaker R, Kauw J, Agami R, Peng C-Y, Graves PR, Thoma RS, Wu Z, Shaw AS and Brummelkamp T, Kops G and Medema RH. (2003). EMBO Piwnica-Worms H. (1997). Science, 277, 1501–1505. J., 22, 2934–2947. Penninger JM and Kroemer G. (2003). Nat. Cell Biol., 5, Letai A, Bassik M, Walensky L, Sorcinelli M, Weiler S and 97–99. Korsmeyer S. (2002). Cancer Cell, 2, 183. Perfettini J-L, Roumier T, Castedo M, Larochette N, Boya P, Li F, Ackermann EJ, Bennett CF, Rothermel AL, Plescia J, Reynal B, Lazar V, Ciccosanti F, Nardacci R, Penninger Tognin S, Villa A, Marchisio PC and Altieri DC. (1999). JM, Piacentini M and Kroemer G. (2004). J. Exp. Med., 199, Nat. Cell Biol., 1, 461–466. 620–640. Li F, Ambrosini G, Chu EY, Plescia J, Tognin S, Marchisio Peters JM. (2002). Mol. Cell, 9, 931–943. PC and Altieri DC. (1998). Nature, 396, 580–584. Porter LA, Cukier IH and Lee JM. (2003). Blood, 101, 1928– Lingle WL, Barrett SL, Negron VC, D’Assoro AB, Boeneman 1933. K, Liu W, Whitehead CM, Reynolds C and Salisbury JL. Ramel S, Sanchez CA, Schimke MK, Neshat K, Cross (2002). Proc. Natl. Acad. Sci. USA, 99, 1978–1983. SM, Raskind WH and Reid BJ. (1995). Pancreas, 11, Liu X and Erikson RL. (2003). Proc. Natl. Acad. Sci. USA, 213–222. 100, 5789–5794. Raslova H, Roy L, Vourch C, Lecouedic JP, Brison O, Lock RB and Stribinskiene L. (1996). Cancer Res., 56, 4006– Metivier D, Feunteun J, Kroemer G, Debili N and 4012. Vainchenker W. (2003). Blood, 101, 541–544. Mantel C, Braun SE, Reid S, Henegariu O, Liu L, Hangoc G Read SH, Baliga BC, Ekert PG, Vaux DL and Kumar S. and Broxmeyer HE. (1999). Blood, 93, 1390–1398. (2002). J. Cell Biol., 159, 739–745. Margottin-Goguet F, Hsu JY, Loktev A, Hsieh H-M, Robertson JD, Enoksson M, Suomela M, Zhivotovsky Reimann JDR and Jackson PK. (2003). Dev. Cell, 4, B and Orrenius S. (2002). J. Biol. Chem., 277, 29803– 813–826. 29809. Masuda A and Takahashi T. (2002a). Oncogene, 21, 6884– Roninson IB, Broude EV and Chang B-D. (2001). Drug 6897. Resistance Updates, 4, 303–313.

Oncogene Mitotic catastrophe M Castedo et al 2837 Rotter V, Schartz D, Almon E, Goldfinger N, Kapon A, Taylor WR and Stark GR. (2001). Oncogene, 20, 1803–1815. Meshorer A, Donehower LA and Levine AJ. (1993). Proc. Terada N, Hamazaki T, Oka M, Hoki M, Mastalerz DM, Natl. Acad. Sci. USA, 90, 9075–9079. Nakano Y, Meyer EM, Morel L, Petersen BE and Scott EW. Roumier T, Castedo M, Perfettini J-L, Andreau K, Metivier (2002). Nature, 416, 542–545. D, Zamzami N and Kroemer G. (2003). Biochem. Pharma- Tort F, Hernandez S, Bea S, Martinez A, Estller M, Herman col., 66. JG, Puig X, Camacho E, Sanchez M, Nayach I, Lopez- Ruth AC and Roninson IB. (2000). Cancer Res., 60, 2576– Guillermo A, Fernandez PL, Colomer D, Hernandez L and 2578. Campo E. (2003). Blood, 100. Sato N, Mizumoto K, Nakamura M and Tanaka M. (2000). Tsuiki H, Nitta M, Tada M, Inagaki M, Ushio Y and Saya H. Exp. Cell Res., 255, 321–326. (2001). Oncogene, 20, 420–429. Sen S, Zhou H, Zhan RD, Yoon DS, Vakar-Lopez F, Ito S, Tsvetkov L, Xu X, Li J and Stern DF. (2003). J. Biol. Chem., Jiang F, Johnston D, Grossman HB, Ruifrok AC, Katz RL, 278, 8468–8475. Brinkley W and Czerniak B. (2002). J. Natl. Cancer Inst., 94, Uren AG, Wong L, Pakusch M, Fowler KJ, Burrows FJ, Vaux 1320–1329. DL and Choo KH. (2000). Curr. Biol., 10, 1319–1328. Senderowicz AM. (2003). Cancer Chemother. Pharmacol., 52, Vahteristo P, Bartkova J, Eerola H, Syrjakoski K, Ojala S, S61–S73. Kilpivaara O, Tamminen A, Kononen J, Aittomaki K, Seo GJ, Kim SE, Lee YM, Lee JW, Lee JR, Hahn MJ and Heikkila P, Holli K, Blomqvist C, Bartek J, Kallioniemi OP Kim ST. (2003). Biochem. Biophys. Res. Commun., 304, 339– and Nevanlinna H. (2002). Am. J. Hum. Genet., 71, 432–438. 343. Vassilopoulos G, Wang PR and Russell DW. (2003). Nature, Shah JV and Cleveland DW. (2000). Cell, 103, 997–1000. 422, 901–904. Shankar SL, Mani S, O’Guin KN, Kandimalla ER, Agrawal S Verdecia MA, Huang H, Dutil E, Kaiser DA, Hunter T and and Shafit-Zagardo B. (2001). J. Neurochem., 79, 426–436. Noel JP. (2000). Nat. Struct. Biol., 7, 602–608. Sheih SY, Ahn J, Tamai K, Taya Y and Prives C. (2000). Verdoodt B, Decordier I, Geleyns K, Cunha M, Cundari E Genes Dev., 14, 289–300. and Kirsch-Volders M. (1999). Mutagenesis, 14, 513–520. Shih I-M, Zhou W, Goodman SN, Lengauer CL, Kinzler KW Verschuren EW, Klefstrom J, Evan GI and Jones N. (2002). and Vogelstein B. (2001). Cancer Res., 61, 818–822. Cancer Cell, 2, 229–241. Shin DM, Charuruks N, Lippman SM, Lee JJ, Ro JY, Hong Vikhanskaya F, Vignati S, Beccaglia P, Ottoboni C, Russo P, WK and Hittelman WN. (2001). Cancer Epidemiol. Biomar- D’Incalci M and Broggini M. (1998). Exp. Cell Res., 241, kers Prev., 10, 603–609. 96–101. Slovak ML, Kopecky KJ, Cassileth PA, Harrington DH, Theil Wahl AF, Donaldson KL, Fairchild C, Lee FY, Foster SA, KS, Mohamed A, Paietta E, Willman CL, Head DR, Rowe Demers GW and Galloway DA. (1996). Nat. Med., 2, 72–79. JM, Forman SJ and Appelbaum FR. (2000). Blood, 96, Wall NR, O’Connor DS, Plescia J, Pommier Y and Altieri DC. 4075–4083. (2003). Cancer Res., 63, 230–235. Smits VA and Medema RH. (2001). Biochim. Biophys. Acta, Wang Q, Xie S, Chen J, Fukasawa K, Naik U, Traganos F, 1519, 1–12. Darzynkiewicz Z, Jhanwar-Uniyal M and Dai W. (2002). Song Z, Yao X and Wu M. (2003). J. Biol. Chem., 278, 23130– Mol. Cell. Biol., 22, 3450–3459. 23140. Weinert TA, Kiser GL and Hartwell LH. (1994). Genes Dev., Spankuch-Schmitt B, Wolf G, Solbach C, Loibl S, Knecht R, 8, 652–665. Stegmuller M, von Minckwitz G, Kaufmann M and Wu GW, Nutt CL, Zlatescu MC, Keeney M, Chin-Yee I and Strebhardt K. (2002). Oncogene, 21, 3162–3171. Cairncross JG. (2001a). Cancer Res., 61, 4155–4159. Sperling K and Ra PN. (1974). Chromosoma, 14, 121–131. Wu X, Webster SR and Chen J. (2001b). J. Biol. Chem., 276, Stevens C, Smith L and La Thangue NB. (2003). Nat. Cell 2971–2974. Biol., 5, 401–409. Wurmser AE and Gage FH. (2002). Nature, 416, 485–487. Strathdee G, Sansom OJ, Sim A, Clarke AR and Brown R. Xu J and Du W. (2003). FEBS Lett., 19, 209–212. (2001). Oncogene, 20, 1923–1927. Yang S, Kuo C, Bisi JE and Kim MK. (2002). Nat. Cell Biol., Susin SA, Daugas E, Ravagnan L, Samejima K, Zamzami N, 4, 865–870. Loeffler M, Costantini P, Ferri KF, Irinopoulou T, Pre´ vost Yin XY, Grove L, Datta NS, Long MW and Prochownik EV. M-C, Brothers G, Mak TW, Penninger J, Earnshaw WC and (1999). Oncogene, 18, 1177–1184. Kroemer G. (2000). J. Exp. Med., 192, 571–579. Yoshikawa R, Kusunoki M, Yanagi H, Noda M, Furuyama Susin SA, Lorenzo HK, Zamzami N, Marzo I, Snow BE, JI, Yamamura T and Hashimoto-Tamaoki T. (2001). Cancer Brothers GM, Mangion J, Jacotot E, Costantini P, Loeffler Res., 61, 1029–1037. M, Larochette N, Goodlett DR, Aebersold R, Siderovski Yu H. (2002). Curr. Opin. Cell Biol., 14, 706–714. DP, Penninger JM and Kroemer G. (1999). Nature, 397, Yu Q, La Rose J, Zhang H, Takemura H, Kohn KW and 441–446. Pommier Y. (2002). Cancer Res., 62, 5743–5748. Swanson PE, Carroll SB, Zhang XF and Mackey MA. (1995). Yuan J, Kramer A, Eckerdt F, Kaufmann M and Strebhardt Am. J. Pathol., 146, 963–971. K. (2002). Cancer Res., 62, 4186–4190. Takada S, Kelkar A and Theurkauf WE. (2003). Cell, 113, Yun J, Chae HD, Choy HE, Chung J, Yoo HS, Han MH and 87–99. Shin DY. (1999). J. Biol. Chem., 274, 29677–29682. Taylor WR, DePrimo SE, Agarwal A, Agarwal ML, Zamzami N, Brenner C, Marzo I, Susin SA and Kroemer G. Schonthal AH, Katula KS and Stark GR. (1999). Mol. (1998). Oncogene, 16, 2265–2282. Cell. Biol., 10, 3607–3622. Zamzami N and Kroemer G. (2003). Curr. Biol., 13, R71–R73.

Oncogene