Partial Downregulation of MAD1 Causes Spindle Checkpoint Inactivation and Aneuploidy, but Does Not Confer Resistance Towards Taxol

Partial downregulation of MAD1 causes spindle checkpoint inactivation and aneuploidy, but does not confer resistance towards taxol

Anne Kienitz1,2, Celia Vogel1,2, Ivonne Morales1, Rolf Mu¨ ller1 and Holger Bastians*,1

1Institute for Molecular Biology and Tumor Research (IMT), Philipps University Marburg, Emil-Mannkopff-Strasse 2, D-35037 Marburg, Germany

The mitotic spindle assembly checkpoint ensures proper promoting complex/cyclosome (APC/C) (Li et al., chromosome segregation during mitosis by inhibiting the 1997; Fang et al., 1998; Gorbsky et al., 1998). This onset of anaphase until all kinetochores are attached to large ubiquitin ligase complex is required for the the mitotic spindle and tension across the kinetochores is ubiquitination and destruction of securin and cyclin B, generated. Here, we report that the stable partial down- prerequisites for anaphase onset and mitotic exit, regulation of the spindle checkpoint gene MAD1, which is respectively (for a review see Peters, 2002). The spindle observed in human cancer, leads to a functional inactiva- checkpoint is also activated in response to various tion of the spindle checkpoint resulting in gross aneu- spindle poisons like nocodazole, a drug that depoly- ploidy. Interestingly, although Mad1is thought to act as merizes microtubules, and thus, prevents the attachment a kinetochore based activator of Mad2 during checkpoint of microtubules to the kinetochores. On the other hand, activation, we show that normal levels of Mad2, but not anticancer drugs, like taxanes (e.g. paclitaxel/taxol), of Mad1, are required for preventing premature sister inhibit the dynamic instability of the spindle and allow chromatid separation and for maintaining the timing of an microtubule attachment, but prevent the generation of undisturbed mitosis, suggesting a Mad1independent tension across kinetochores. Treatment of cells with function of Mad2 that operates independent of its these drugs inhibits chromosome alignment and results checkpoint function. Most significantly, a partial repres- in a mitotic arrest before anaphase, which is followed by sion of either MAD1 or MAD2 confers resistance to the induction of apoptosis (Jordan and Wilson, 2004). nocodazole, a drug that inhibits microtubule attachment. To date, the molecular pathways of the spindle In contrast, sensitivity to clinically relevant drugs like checkpoint are not well understood, but it is established taxol or monastrol that inhibit the generation of tension that several evolutionary conserved proteins including across kinetochores is not modulated by partial down- Cenp-E, BubR1, Bub1, Bub3, Mad1, Mad2 and Mps1 regulation of MAD1, suggesting a functional bifurcation are required for spindle checkpoint function (for a of spindle checkpoint dependent apoptotic pathways. review see Yu, 2002; Bharadwaj and Yu, 2004; Taylor Oncogene (2005) 24, 4301–4310. doi:10.1038/sj.onc.1208589 et al., 2004). These and most likely additional yet Published online 14 March 2005 unknown proteins are sequentially recruited to unattached or tension-lacking kinetochores leading to the Keywords: cell cycle; mitosis; mad2; apoptosis; genomic generation of a diffusible APC/C inhibitor (Shah and instability Cleveland, 2000; Musacchio and Hardwick, 2002; Vigneron et al., 2004). Mad2, BubR1, and also a mitotic checkpoint complex (MCC) consisting of Mad2, BubR1, Bub3 and Cdc20 can function as potent Introduction inhibitors for APC/C activity, but the mechanism of the MCC formation is still unclear (Fang et al., 1998; The faithful segregation of chromosomes during mitosis Sudakin et al., 2001; Tang et al., 2001; Fang, 2002). requires that all chromosomes are correctly attached to According to a current model, an unattached kineto- the mitotic spindle apparatus and properly aligned at chore serves as an activating platform for the diffusible the metaphase plate before anaphase onset is allowed. APC/C inhibitor. Consistently, Mad1 is required for During the early phases of mitosis when chromosomes recruitment of Mad2 to kinetochores and complex are not yet aligned, the mitotic spindle assembly formation of Mad1 and Mad2 at the kinetochore is checkpoint is activated (Campbell and Gorbsky, 1995; important for checkpoint activation, suggesting that Li and Nicklas, 1995; Rieder et al., 1995) and prevents Mad1 might function as an activator of Mad2 at the the onset of anaphase by inhibiting the anaphase kinetochore (Jin et al., 1998; Chen et al., 1999; Campbell et al., 2001; Sironi et al., 2001; Luo et al., 2002; Martin- Lluesma et al., 2002). *Correspondence: H Bastians; E-mail: [email protected] 2These authors contributed equally to this work Interestingly, Mad2 and BubR1 appear to be also Received 11 August 2004; revised 27 January 2005; accepted 4 February involved in regulating the normal timing of mitosis. 2005; published online 14 March 2005 Inhibition of Mad2 or BubR1, but none of the other MAD1 downregulation causes aneuploidy A Kienitz et al 4302 Mad and Bub proteins, causes an acceleration of mitosis A function of Mad1 together with Mad2 for the (Gorbsky et al., 1998; Canman et al., 2002; Meraldi spindle checkpoint has been demonstrated in cells et al., 2004). Moreover, the inhibition of their kineto- depleted of Mad1 by transient siRNA transfections chore localization does not shorten the time from (Luo et al., 2002; Martin-Lluesma et al., 2002; Meraldi prophase to anaphase indicating that Mad2 and BubR1 et al., 2004). However, severe depletion or homozygous might be part of a kinetochore independent mitotic disruption of spindle checkpoint genes results in cell timer mechanism (Meraldi et al., 2004). death indicating that spindle checkpoint genes are The majority of human cancer cells are aneuploid, essential for viability (Dobles et al., 2000; Kalitsis which can be caused by gaining or loosing whole et al., 2000; Kops et al., 2004; Michel et al., 2004). chromosomes during defective chromosome segregation Therefore, previous studies could not address the long- events (Duesberg et al., 2000; Sieber et al., 2003). This term physiological consequences of Mad1 inactivation. phenotype is called chromosomal instability (CIN) and Moreover, in human cancer, only partial downregula- although its molecular basis is still unknown, it is now tion of MAD1 or MAD2 is observed and it is important apparent that malfunction of the spindle checkpoint can to resolve whether a partial downregulation of MAD1 directly contribute to CIN (Lengauer et al., 1998; Michel or MAD2 is functionally redundant. Therefore, we et al., 2001; Masuda and Takahashi, 2002; Babu et al., investigated the phenotypes associated with partial 2003; Rajagopalan and Lengauer, 2004). Indeed, defects downregulation of MAD1 or MAD2. Our results in the spindle checkpoint are frequently found in various indicate a differential requirement of Mad1 and Mad2 types of human cancer including lung, breast, ovarian, for spindle checkpoint activation, mitotic timing and colon and hepatocellular tumors and these defects are induction of apoptosis in response to chemotherapeutic associated with CIN and aneuploidy (Cahill et al., 1998; drugs in human cancer cells. Takahashi et al., 1999; Masuda and Takahashi, 2002; Shichiri et al., 2002). While spindle checkpoint defects are frequent, spindle checkpoint genes appear to be Results rarely altered in human tumors (Cahill et al., 1998; Nomoto et al., 1999; Hernando et al., 2001). Instead, Generation of human colon carcinoma cells exhibiting downregulated expression of spindle checkpoint genes reduced levels of Mad1 might contribute to a deactivated spindle checkpoint in cancer cells. In fact, the genes of the MCC, MAD2, We wished to generate karyotypically stable human cells BUB3 and BUBR1, have been shown to be haplo- exhibiting partial downregulation of MAD1. For this, insufficient (Michel et al., 2001; Babu et al., 2003; Baker we stably expressed an siRNA targeting MAD1 in et al., 2004). Heterozygous deletion of MAD2 in human HCT116 cells, which are well characterized for their colon carcinoma cells and heterozygous MAD2, BUB3 intact cell cycle checkpoints and their relatively stable or BUBR1 disruptions in mice result in partially karyotype. In addition, HCT116 cells have been used downregulated checkpoint protein levels, an impaired previously for heterozygous gene deletion of MAD2 spindle checkpoint and aneuploidy (Michel et al., 2001; (HCT116-MAD2 þ /À) resulting in cells with approxi- Babu et al., 2003; Baker et al., 2004). Most significantly, mately 20% reduction of Mad2 protein level (Michel MAD2 þ /À, BUB3 þ /À and BUBR1 þ /À mice are prone to et al., 2001). This enabled us to investigate and to tumor development suggesting that aneuploidy can compare the long-term cellular consequences of MAD1 contribute directly to tumorigenesis (Michel et al., and MAD2 downregulation. Upon stable transfection 2001; Babu et al., 2003; Dai et al., 2004). To date, it is we selected two independent clones, termed knockdown unclear if a moderate downregulation of other spindle clone 2-2 (MAD1-kd-2-2) and knockdown clone 2-31 checkpoint genes directly contributes to aneuploidy and (MAD1-kd-2-31) for the experiments throughout this tumorigenesis; however, a reduced expression of MAD2 study. These clones showed about 35and 50%reduction and BUBR1, but also of BUB1 and MAD1, has been of endogenous Mad1 protein, respectively, when com- found in different human cancers (Li and Benezra, 1996; pared to HCT116-wild-type cells (HCT116-wt) and to Shichiri et al., 2002; Wang et al., 2002). puromycin resistant control cells (HCT116-pSUPER) as Surprisingly, spindle checkpoint defects have recently shown on Western blots using specific anti-Mad1 been associated with resistance to spindle damaging antibodies (Figure 1a). When grown asynchronously in anticancer drugs, for example, paclitaxel/taxol (Kasai culture, HCT-MAD1-kd cells did not show significant et al., 2002; Anand et al., 2003; Masuda et al., 2003; changes in cell cycle distribution, proliferation rate or Sudo et al., 2004) and with the resistance to topoiso- apoptosis indicating that a repression of MAD1 is merase poisons, for example, etoposide, adriamycin/ neither affecting the cell cycle timing per se nor causing doxorubicin (Vogel et al., 2004b). These findings indicate lethality (Figure 1b). that the spindle checkpoint might be directly involved in the induction of apoptosis upon mitotic damage and Partial downregulation of MAD1 or MAD2 result imply that tumors harboring a compromised spindle in a defective spindle checkpoint checkpoint are resistant to such chemotherapeutic treatments. However, the molecular mechanisms of the To investigate whether a partial downregulation of crosstalk between the spindle checkpoint and apoptotic MAD1 would account for a defective spindle assembly pathways remain to be determined. checkpoint function, we determined the mitotic arrest in

Oncogene MAD1 downregulation causes aneuploidy A Kienitz et al 4303 DNA content as demonstrated by FACS analyses (Figure 2f). Furthermore, the premature exit from mitosis in spindle checkpoint deﬁcient cells resulted in a similar high rate of endoreduplication in MAD1-kd and HCT-MAD2 þ /À cells (data not shown; Vogel et al., 2004a). Taken together, partial downregulation of either MAD1 or MAD2 results in a very similar functional inactivation of the spindle checkpoint.

Reduced levels of MAD2 but not of MAD1 induce a high rate of premature sister chromatid segregation Spindle checkpoint inhibition is expected to result in premature sister chromatid separation. To test whether partial downregulation of MAD1 or MAD2 results in premature sister chromatid segregation, we performed Figure 1 Generation of HCT116 cells stably expressing siRNA metaphase chromosome spread analyses (Figure 3a) and targeting MAD1.(a) Reduced protein levels of Mad1 in MAD1 determined the proportion of cells showing disjoined knockdown cells. Lysates from two independent knockdown clones kd kd sister chromatids (Figure 3b). Upon a 2-h nocodazole (Mad1 (2-2) and Mad1 (2-31)), control cells (HCT116 þ pSU- þ /À PER) and parental cells (HCT116-wt) were subjected to Western treatment, HCT-MAD2 cells displayed a high rate blot analysis using anti-Mad1 and anti-actin antibodies. Mad1 of premature anaphases (11–15%) demonstrating, in antibody signals were quantified in four independent experiments, agreement with previous findings (Michel et al., 2001), normalized for actin signals and the Mad1 protein expression level that normal levels of Mad2 are required to prevent gross was calculated in percent of HCT116-wild-type expression. A representative experiment is shown. (b) FACS analyses of premature sister chromatid separation. Upon prolonged asynchronously growing MAD1 knockdown and control cell lines nocodazole treatment (6 h), the proportion was even increased to 16–21%. In contrast, both MAD1-kd cell lines showed much lower proportions of premature anaphases after 2 and 6 h treatment (2.1–3.3% and response to nocodazole, which prevents microtubule 3–5%, respectively) corresponding to just about 20% of attachment to the kinetochores, and thus, chromosome the rate observed in HCT-MAD2 þ /À cells (Figure 3b). alignment (Figure 2a). Both MAD1-kd cell lines (2-2 Our findings suggest the premature sister chromatid and 2-31) showed a mitotic index with maximum values separation might be restrained by a Mad1 independent of 30–40% similar to that observed in HCT-MAD2 þ /À function of Mad2. cells. In contrast, a maximum of approximately 70% of the wild type and control cells accumulated in mitosis. À/À Normal levels of Mad2 but not of Mad1 are required Further, isogenic p53 deficient cells (HCT-p53 ) for normal mitotic timing arrested like wild-type cells indicating no requirement of p53 for normal spindle checkpoint function. Very Previous studies have indicated that the spindle check- similar results were observed after treatment with taxol point regulates the timing of a normal mitosis in the or monastrol, drugs that inhibit the spindle dynamics absence of damage (Taylor and McKeon, 1997; and the formation of a bipolar spindle, respectively, Gorbsky et al., 1998; Shannon et al., 2002; Meraldi allowing microtubule attachment, but not the genera- et al., 2004). Moreover, a kinetochore independent tion of tension (Figure 2b and c). Our results clearly function of Mad2 and BubR1, but not of other Mad or demonstrate that partial downregulation of either Bub proteins, appears to be required for the mitotic MAD1 or MAD2 results in a very similar functional timer function (Meraldi et al., 2004). Therefore, we inactivation of the spindle checkpoint in response to investigated whether partial downregulation of MAD1 lack of microtubule attachment and upon lack of or MAD2, which is observed in human cancer, might tension across kinetochores. disturb the normal timing of mitosis. We treated asyn- Consistently, the defect in spindle checkpoint activa- chronously growing HCT-MAD1-kd, HCT-MAD2 þ /À tion in HCT-MAD1-kd and HCT-MAD2 þ /À cells was and control cells with the proteasome inhibitor ALLN accompanied by a premature exit from mitosis in the for up to 3 h to block mitotic progression beyond presence of nocodazole. This was demonstrated by the metaphase in the absence of spindle damage and premature degradation of cyclin B (Figure 2d) and by determined the proportion of metaphase figures (Fig- generation of multinuclei around groups of chromo- ure 4). All cell lines exhibit similar growth rates somes (Figure 2e). Quantification of multinuclei and similar rates of mitotic entry without the addition revealed that after 18 h of spindle damage 60–65% of ALLN (DMSO control). As expected, ALLN of both, MAD1-kd cells and HCT-MAD2 þ /À cells, treatment increased the proportion of metaphases in had exited without cytokinesis whereas only 3–8% of control cells to 35–38%, indicating that the mitotic the control cells had formed multinuclei after 18 h progression is indeed blocked by proteasome inhibition. of nocodazole treatment (Figure 2e). Under these In contrast, in HCT-MAD2 þ /À cells metaphases were conditions, the multinuclei contained a tetraploid increased to 62–64% indicating an acceleration of

Oncogene MAD1 downregulation causes aneuploidy A Kienitz et al 4304 mitosis. Signiﬁcantly, however, accumulation in meta- is not accelerated upon MAD1 repression. Our results phase of both MAD1-kd clones was not increased support previous ﬁndings showing that Mad2 and compared to control cells indicating that prometaphase BubR1, but no other Mad or Bub proteins, are required

Oncogene MAD1 downregulation causes aneuploidy A Kienitz et al 4305 for the mitotic timer function (Meraldi et al., 2004), and show that partial repression of Mad2, but not of Mad1 as observed in human tumor cells, is expected to disturb mitotic timing.

Partial repression of MAD1 or MAD2 result in aneuploidy Since partial downregulation of MAD1 neither causes gross premature sister chromatid separation nor pertur- bation of normal mitotic timing, we asked whether partial downregulation of MAD1 in tumor cells would account for the generation of aneuploidy. We performed metaphase chromosome spreads of HCT-MAD1-kd cells, HCT-MAD2 þ /À cells and control cells and determined the number of individual chromosomes (Figure 5). The control HCT116 cell lines showed a relatively stable karyotype (76–78% normal karyotype with 45chromosomes). Significantly, HCT- MAD1-kd Figure 4 Reduced levels of MAD2, but not of MAD1, accelerate normal mitotic timing. Asynchronously grown HCT-MAD2 þ /À, cells were found to be highly aneuploid (41 and 53% HCT-MAD1 knockdown and control cells were treated with the proteasome inhibitor ALLN for 3 h to block mitotic progression beyond metaphase or with DMSO as a control. Cells were fixed, DNA was stained with Hoechst dye and mitotic figures were counted. The proportion of metaphases was calculated and the graph shows mean values and standard deviations from three independent experiments. At least 1800 mitotic figures were counted per cell line

Figure 3 Different rates of premature sister chromatid separation upon partial downregulation of MAD1 and MAD2.(a) Example of Figure 5 Partial downregulation of MAD1 or MAD2 result in premature sister chromatid separation in HCT-MAD2 þ /À and gross aneuploidy. HCT116 control cells, HCT-MAD2 þ /À cells and HCT-MAD1-kd (clone 2-31) cells in comparison to intact two independent HCT-MAD1 knockdown cell lines (2-2 and 2-31) chromatid cohesion in HCT116 control cells after treatment with were used for chromosome spread analyses and individual nocodazole for 2 h. (b) Quantiﬁcation of mitotic ﬁgures showing chromosomes were counted. The distribution of chromosome premature sister chromatid separation after treatment with numbers is shown and the percentage of aneuploidy was calculated nocodazole for 2 and 6 h. Mean values and standard deviations (45chromosomes represent the relative euploid karyotype of from four independent experiments are shown. At least 1600 HCT116 cells). At least 150 metaphase spreads were counted for metaphase spreads were counted per cell line and time point each cell line

Figure 2 Partial downregulation of MAD1 or MAD2 results in a defective spindle checkpoint. (a–c) The mitotic arrest in response to the lack of microtubule attachment or upon the lack of tension is impaired in HCT-MAD1 knockdown and HCT-MAD2 þ /À cells. Cells were treated with (a)150nM nocodazole, (b) 100 nM taxol or (c)70mM monastrol for up to 48 h and the mitotic index was determined. The graph shows mean values from at least three independent experiments. (d) Partial downregulation of MAD1 or MAD2 results in premature cyclin B degradation. The indicated cell lines were treated with nocodazole for 14 h and cyclin B protein levels were detected on Western blots. (e) The premature exit from mitosis in MAD2 þ /À and MAD1 knockdown cells is accompanied by the formation of multinuclei. MAD1 knockdown cells were treated with nocodazole for 18 h and nuclei were stained with Hoechst dye (left panel). HCT116 derivative cells were treated with nocodazole for 18 h and the proportion of mitotic, interphase and multinucleated cells were determined. At least 500 cells were counted for each cell line. (f) Multinucleated cells contain a tetraploid DNA content. Cells were treated as in (e) and the DNA content was determined in FACS analyses

Oncogene MAD1 downregulation causes aneuploidy A Kienitz et al 4306 aneuploidy). In agreement with previous findings Therefore, we investigated the induction of apoptosis (Michel et al., 2001), we found HCT-MAD2 þ /À cells to in HCT-wt, HCT-p53À/À,HCT-MAD2 þ /À and HCT- be similarly aneuploid (55% aneuploidy). Thus, the MAD1-kd cells in response to a treatment with drugs prevention of aneuploidy requires normal levels of both, that either inhibit microtubule attachment (nocodazole; Mad1 and Mad2. Figure 6, left panels) or that prevent the generation of tension across kinetochores (taxol and monastrol; Figure 6, middle and right panels). The induction of Tumor cells exhibiting reduced levels of MAD2, but not apoptosis was monitored by the detection of the of MAD1 are resistant to taxol treatment apoptosis specific laddering of chromosomal DNA Spindle damaging agents are clinically highly relevant (Figure 6a), by the detection of the apoptosis specific anticancer drugs (Jordan and Wilson, 2004). Interest- cleavage product of poly-(ADP-ribose)-polymerase ingly, defects in the spindle checkpoint have been (PARP; Figure 6b) and by quantifying the activity of recently associated with a resistance to these chemother- caspase-3 in whole cells (Figure 6c). All three indepen- apeutic drugs (Kasai et al., 2002; Anand et al., 2003; dent assays consistently show that both spindle Masuda et al., 2003; Sudo et al., 2004). However, it is checkpoint compromised cell lines efficiently escape largely unknown how apoptosis is induced in response from apoptosis upon treatment with nocodazole, that is, to these treatments and what the parameters are that in response to the lack of microtubule attachment determine sensitivity or resistance to spindle poisons. (Figure 6, left panels). Significantly and in contrast to

Figure 6 Partial downregulation of MAD2, but not of MAD1, causes resistance to taxol. (a) HCT116 and derivative cells were treated with 150 nM nocodazole, 100 nM taxol or 70 mM monastrol for 48 h and fragmented chromosomal DNA was isolated and resolved on agarose gels. (b) Cells were treated as in (a) and the uncleaved and the apoptosis-speciﬁc cleaved forms of PARP were detected on Western blots. The star indicates a crossreacting band. (c) Cells were treated with nocodazole, taxol or monastrol for 40 h and the apoptotic activity of caspase-3 was quantitatively determined using a ﬂuorogenic peptide substrate. The graphs show mean values and standard deviations from three independent experiments

Oncogene MAD1 downregulation causes aneuploidy A Kienitz et al 4307 MAD2 þ /À cells, spindle checkpoint compromised cells somatic cells (this study) is sufficient to cause spindle with reduced Mad1 levels are sensitive to a treatment checkpoint inactivation and aneuploidy indicating that with drugs that inhibit the generation of tension, that is normal expression of the spindle checkpoint genes is taxol or monastrol (Figure 6, middle and right panels) essential for mediating mitotic arrest in response to indicating a Mad1 independent induction of apoptosis. spindle damage and for maintaining chromosomal Further, we found that HCT116-p53À/À escape from stability. It is important to note that severe down- apoptosis after nocodazole, taxol or monastrol treat- regulation, disruption or homozygous mutational in- ments indicating a general requirement of p53 for the activation of spindle checkpoint genes appears to be induction of apoptosis in response to spindle damage. incompatible with survival (Dobles et al., 2000; Kalitsis Thus, our findings clearly demonstrate that the status of et al., 2000; Kops et al., 2004; Michel et al., 2004), which the spindle checkpoint per se does not determine the might explain why only partial downregulation of sensitivity towards spindle targeting chemotherapeutic spindle checkpoint genes is observed in human cancer. drugs, but rather the status of individual spindle Interestingly, in MAD2 þ /À, BUB3 þ /À and BUBR1 þ /À checkpoint components. In fact, our results suggest a mice, the spindle checkpoint inactivation is also Mad1 independent function of Mad2 to be required for associated with enhanced tumor development (Michel the taxol induced apoptosis and a functional bifurcation et al., 2001; Babu et al., 2003; Dai et al., 2004), for the induction of apoptosis depending on the nature suggesting that aneuploidy can directly contribute to of spindle damage. tumorigenesis. However, the mechanisms of cancer specific downregulation of spindle checkpoint gene expression are currently unknown. Interestingly, in Discussion human colorectal carcinomas of advanced clinical state, a significant silencing of mRNA expression of BUB1 Our data presented here establish the following main and BUBR1 was observed, suggesting that epigenetic points. (i) A stable partial downregulation of either mechanisms like promotor methylation might account MAD1 or MAD2, similar to the situation observed in for reduced gene expression in cancer cells (Shichiri human cancer cells, is sufficient to cause a functional et al., 2002). Nevertheless, the expression level of spindle inactivation of the mitotic spindle assembly checkpoint checkpoint genes is of great diagnostic value to predict in response to the lack of microtubule attachment or spindle checkpoint activity and CIN. upon the lack of tension across the kinetochores. Most importantly, this phenotype is associated with the Different phenotypes associated with the downregulation generation of gross aneuploidy. (ii) Normal expression of MAD1 or MAD2 of MAD2, but not of MAD1, is required to prevent premature sister chromatid separation during check- A number of previous studies have suggested that Mad1 point activation. (iii) Partial repression of MAD2, but acts as an activator of Mad2 at the checkpoint activated not of MAD1, causes an acceleration of normal mitotic kinetochore (Sironi et al., 2001; Luo et al., 2002; Martin- timing. (iv) Partial downregulation of either MAD2 Lluesma et al., 2002). However, an elegant recent study or MAD1 confers resistance to spindle-damaging has demonstrated a requirement of Mad2 and BubR1, agents that disrupt the microtubule–kinetochore attach- but not of other checkpoint proteins for the normal ment. In contrast, reduced levels of Mad2, but not of timing of mitosis (Meraldi et al., 2004). Further, this Mad1, lead to resistance to chemotherapeutic drugs work proposed a cytosolic and kinetochore independent that inhibit the generation of tension, for example, function of Mad2 and BubR1 during an undisturbed paclitaxel/taxol. mitosis. Our data strongly support the idea of additional kinetochore and Mad1 independent functions of Mad2, Partial downregulation of spindle checkpoint in which might be separable from its checkpoint function. human cancer Since partial downregulation of MAD1 neither accel- erates mitosis nor significantly elevates the proportion of Obvious candidate genes that might be involved in premature sister chromatid separation when compared causing CIN in cancer are the spindle checkpoint genes, to cells with reduced levels of Mad2, we suggest a Mad1 but these appear to be rarely inactivated by mutation in independent function of Mad2 not only for normal human cancer (Cahill et al., 1998; Nomoto et al., 1999; mitotic timing but also for preventing the premature Takahashi et al., 1999; Hernando et al., 2001; Masuda onset of anaphase. According to this model, a cytosolic and Takahashi, 2002). However, partial downregulation pool of Mad2 and BubR1 might be involved in of spindle checkpoint genes including MAD1, MAD2, preventing the APC/C-mediated proteolysis of securin, BUB1 and BUBR1 has been observed in cancer cells (Li which is required for sister chromatid separation and Benezra, 1996; Shichiri et al., 2002; Wang et al., (Nasmyth et al., 2000). This idea is further supported 2002). In fact, experimental partial repression of MAD2 by similar observations that were made in BUB3 þ /À leads to spindle checkpoint inactivation (Wang et al., MEFs, where reduced expression of BUB3 leads to an 2002) and causes aneuploidy (Michel et al., 2001; and impairment of the spindle checkpoint in response to this study). Similarly, reduced expression of BUBR1 or spindle damage but not to an induction of premature BUB3 in heterozygous mice (Babu et al., 2003; Baker sister chromatid segregation (Babu et al., 2003). Inter- et al., 2004; Dai et al., 2004), or of MAD1 in human estingly, the model of separable functions of Mad1 and

Oncogene MAD1 downregulation causes aneuploidy A Kienitz et al 4308 Mad2 might also explain the concomitant downregu- important determinant to predict a resistance towards lation of MAD1 and MAD2 observed in ovarian taxol or related chemotherapeutic drugs. carcinoma cells (Wang et al., 2002).

Crosstalk of the mitotic spindle checkpoint and apoptotic Materials and methods pathways in response to chemotherapeutic treatment Cell culture Chemotherapeutic drugs that target the mitotic spindle, HCT116-wild type, HCT116-p53À/À and HCT116-MAD2 þ /À including taxanes and vinca alkaloids, are frequently cells (Michel et al., 2001) were kindly provided by Bert used for treatment of lymphomas, ovarian, breast and Vogelstein (Baltimore, USA), Loren Michel and Robert lung tumors (Jordan and Wilson, 2004). However, it is Benezra (New York, USA), respectively. All HCT116 deriva- still unclear how these drugs induce apoptosis and what tive cells were grown in RPMI 1640 medium supplemented the molecular mechanisms of the resistance are. First with 10% fetal calf serum, 1% glutamine and 100 mg/ml reports have indicated that rapid drug export via ABC streptomycin and 100 U/ml penicillin (all from Gibco, Nether- 1 transporters, the status of bcl-2 or enhanced micro- lands) at 37 C under 95% water and 5% CO2 atmosphere. tubule dynamics can confer resistance (Jordan and Where indicated cells were treated with 150 nM nocodazole Wilson, 2004). In addition, most recent reports have (Sigma, Germany), 100 nM taxol (Sigma) or 70 mM monastrol (Sigma). provided a strong correlation between the status of the spindle checkpoint and the susceptibility to spindle Generation of stable HCT116 MAD1 knockdown cell lines damaging agents (Kasai et al., 2002; Anand et al., 2003; Masuda et al., 2003; Sudo et al., 2004). We have now A double-stranded oligonucleotide corresponding to nt992–1014 established that the induction of apoptosis induced by of the human MAD1 cDNA (50-AAGACCTTTCCAGATT 0 drugs that abrogate the microtubule attachment (e.g. CGTGGTT-3 ) was cloned into the BglII–HindIII site of nocodazole) requires normal levels of both, Mad1 and pSuper (Brummelkamp et al., 2002). HCT116 cells were cotransfected with the pSuper construct and pBabe-puro using MAD1 Mad2. In contrast, partial downregulation of Superfect (Qiagen, Germany) following the manufacturer´ s does not confer resistance to agents that inhibit the instructions. At 2 days post-transfection stable clones were microtubule dynamics and the generation of tension selected in medium containing 5 mg/ml puromycin (Sigma). In (e.g. taxol and monastrol), although it is clearly total, 31 stable clones were tested for activation of the spindle sufficient to cause spindle checkpoint inactivation assembly checkpoint by measuring the mitotic index 16 h after suggesting an uncoupling of mitotic arrest and induction the addition of nocodazole or taxol. Two clones, termed 2-2 of apoptosis. Our results indicate that a Mad1 and and 2-31 were selected and used for further experiments possibly kinetochore independent function of Mad2 throughout this study. might also be involved in the induction of apoptosis in response to the lack of tension. However, the induction Western blotting of apoptosis in response to the lack of attachment Western blotting was performed as described (Vogel et al., requires Mad1 and Mad2. These findings suggest a 2004a) using the following antibodies: monoclonal anti-Mad1 functional bifurcation of apoptotic pathways depending (a generous gift from Dr TJ Yen, Philadephia, USA), anti- on the nature of spindle damage. Significantly, a actin (Sigma), anti-cyclin B (H-433, Santa Cruz, USA), anti- functional bifurcation of the spindle checkpoint is also PARP (BD Pharmingen, USA). All secondary antibodies were observed for the mitotic arrest. The chromosomal from Jackson ImmunoResearch (USA). For quantification of passenger proteins including Aurora B and survivin Mad1 protein we performed Western blots using anti-Mad1 are required for a sustained arrest in response to the lack antibodies and quantified the Mad1 bands using the Image Gauge software. Signal data were normalized for actin bands of tension, but not upon nocodazole treatment (Ditch- and a mean value was calculated from four independent field et al., 2003; Hauf et al., 2003; Lens et al., 2003). experiments. Interestingly, survivin has been shown to regulate directly the induction of apoptosis (Li et al., 1999), Flow cytometry which might reflect a possible link between the spindle checkpoint and the activation of apoptotic pathways. The determination of the DNA content and of the mitotic It will be important to elucidate what the molecular index was performed as described previously (Vogel et al., mechanisms are that mediate the induction of apoptosis 2004a). under these different conditions. Since it is now recognized that the commonly used Microscopy chemotherapeutic drugs like taxanes and vinca alkaloids Cells were fixed on polylysine-treated coverslips with 3% mediate their death-inducing effects through the sup- p-formaldehyde/PBS and DNA was stained with 1 mg/ml pression of the microtubule dynamics rather than by Hoechst to visualize interphase chromatin, mitotic condensed depolymerization of the microtubules (Jordan and chromosomes, multinuclei or apoptotic nuclei. Photos were Wilson, 2004), it seems likely that these drugs employ taken using a digital CCD Camera (Hamamatsu, Japan). the Mad2 dependent, but Mad1 independent apoptotic pathway. Therefore, the status of Mad2 rather than of Determination of mitotic timing Mad1, and not the functionality of the spindle Asynchronously growing cells were grown on coverslips and checkpoint in cancer cells per se, appears to be the treated with 200 mM ALLN (Calbiochem, Germany) for up to

Oncogene MAD1 downregulation causes aneuploidy A Kienitz et al 4309 3 h to block mitotic progression beyond metaphase. Cells were described using anti-PARP antibodies (BD Pharmingen) and ﬁxed and DNA was stained as described above. anti-actin antibodies (Sigma) were used as a control.

Determination of apoptosis Chromosome spread analysis Cells were treated with 70 mM monastrol, 100 nM taxol or Chromosome spreading was performed essentially as described 150 nM nocodazole for up to 48 h. Apoptosis was determined (Hauf et al., 2003). Briefly, cells were treated with 150 nM using the following methods. Apoptotic activation of caspase 3 nocodazole for 1, 2 or 6 h. Cells were then harvested by mitotic was measured using Ac-DEVD-AMC as a fluorogenic caspase shake off and hypotonically swollen in 40% RPMI medium/ 3 substrate (Gurtu et al., 1997). Briefly, whole cells were 60% H2O for 10 min at 371C. Cells were fixed in Carnoy´ s harvested and incubated in a 96-well dish at 371C in a tissue fixative solution (75% methanol/25% acetic acid) with several culture incubator in the presence of 2.5 mM Ac-DEVD-AMC changes of the fixative. Cells were dropped onto cooled glass (BD PharMingen). Cleaved substrate was determined as slides and dried at RT. Chromosomes were stained in 5% triplicates using a Victor2 1420 multilabel counter (Wallac Giemsa (Sigma) for 10 min, rinsed with water, air dried and Oy, Finland) in 30-min intervals for up to 180 min. mounted. To determine apoptotic DNA laddering chromosomal DNA was isolated essentially as described (Herrmann et al., 1994) Acknowledgements with minor modifications. Briefly, attached and floating cells We thank Tim Yen for generously providing Mad1 antibodies, were harvested and extracted twice for 10 min with lysis buffer Bert Vogelstein, Loren Michel and Robert Benezra for (50 mM Tris-HCl pH7.5, 20 mM EDTA, 1% NP40, 1% SDS). providing HCT116 and derivative cell lines and Rene Bernards The pooled supernatants were digested with 1 mg/ml RNase A for the pSuper vector. We are grateful to Irmgard Hofmann for 2 h at 371C and with 1 mg/ml proteinase K for 4 h at 561C. for help with the generation of the knockdown cell lines and to The chromosomal DNA was precipitated and separated on a Heike Krebber for helpful discussions and critically reading 2% agarose gel and visualized with ethidium bromide. the manuscript. This work was supported by grants from the The apoptotic cleavage product of poly-(ADP-Ribose)- Deutsche Forschungsgemeinschaft, the Deutsche Krebshilfe Polymerase (PARP) was determined on Western blots as and the PE Kempkes Stiftung.

References

Anand S, Penrhyn-Lowe S and Venkitaraman AR. (2003). Gurtu V, Kain SR and Zhang G. (1997). Anal. Biochem., 251, Cancer Cell, 3, 51–62. 98–102. Babu JR, Jeganathan KB, Baker DJ, Wu X, Kang-Decker N Hauf S, Cole RW, LaTerra S, Zimmer C, Schnapp G, Walter and van Deursen JM. (2003). J. Cell Biol., 160, 341–353. R, Heckel A, van Meel J, Rieder CL and Peters JM. (2003). Baker DJ, Jeganathan KB, Cameron JD, Thompson M, Juneja J. Cell Biol., 161, 281–294. S, Kopecka A, Kumar R, Jenkins RB, de Groen PC, Roche P Hernando E, Orlow I, Liberal V, Nohales G, Benezra R and and van Deursen JM. (2004). Nat. Genet., 36, 744–749. Cordon-Cardo C. (2001). Int. J. Cancer, 95, 223–227. Bharadwaj R and Yu H. (2004). Oncogene, 23, 2016–2027. Herrmann M, Lorenz HM, Voll R, Grunke M, Woith W and Brummelkamp TR, Bernards R and Agami R. (2002). Science, Kalden JR. (1994). Nucleic Acids Res., 22, 5506–5507. 296, 550–553. Jin DY, Spencer F and Jeang KT. (1998). Cell, 93, 81–91. Cahill DP, Lengauer C, Yu J, Riggins GJ, Willson JK, Jordan MA and Wilson L. (2004). Nat. Rev. Cancer, 4, Markowitz SD, Kinzler KW and Vogelstein B. (1998). 253–265. Nature, 392, 300–303. Kalitsis P, Earle E, Fowler KJ and Choo KH. (2000). Genes Campbell M and Gorbsky G. (1995). J. Cell Biol., 129, Dev., 14, 2277–2282. 1195–1204. Kasai T, Iwanaga Y, Iha H and Jeang KT. (2002). J. Biol. Campbell MS, Chan GK and Yen TJ. (2001). J. Cell Sci., 114, Chem., 277, 5187–5193. 953–963. Kops GJ, Foltz DR and Cleveland DW. (2004). Proc. Natl. Canman JC, Salmon ED and Fang G. (2002). Cell Motil. Acad. Sci. USA, 101, 8699–8704. Cytoskeleton, 52, 61–65. Lengauer C, Kinzler KW and Vogelstein B. (1998). Nature, Chen RH, Brady DM, Smith D, Murray AW and Hardwick 396, 643–649. KG. (1999). Mol. Biol. Cell, 10, 2607–2618. Lens SM, Wolthuis RM, Klompmaker R, Kauw J, Agami R, Dai W, Wang Q, Liu T, Swamy M, Fang Y, Xie S, Mahmood Brummelkamp T, Kops G and Medema RH. (2003). EMBO R, Yang YM, Xu M and Rao CV. (2004). Cancer Res., 64, J., 22, 2934–2947. 440–445. Li F, Ackermann EJ, Bennett CF, Rothermel AL, Plescia J, Ditchﬁeld C, Johnson VL, Tighe A, Ellston R, Haworth C, Tognin S, Villa A, Marchisio PC and Altieri DC. (1999). Johnson T, Mortlock A, Keen N and Taylor SS. (2003). Nat. Cell Biol., 1, 461–466. J. Cell Biol., 161, 267–280. Li X and Nicklas RB. (1995). Nature, 373, 630–632. Dobles M, Liberal V, Scott ML, Benezra R and Sorger PK. Li Y and Benezra R. (1996). Science, 274, 246–248. (2000). Cell, 101, 635–645. Li Y, Gorbea C, Mahaffey D, Rechsteiner M and Benezra R. Duesberg P, Li R, Rasnick D, Rausch C, Willer A, Kraemer (1997). Proc. Natl. Acad. Sci. USA, 94, 12431–12436. A, Yerganian G and Hehlmann R. (2000). Cancer Genet. Luo X, Tang Z, Rizo J and Yu H. (2002). Mol. Cell, 9, Cytogenet., 119, 83–93. 59–71. Fang G. (2002). Mol. Biol. Cell, 13, 755–766. Martin-Lluesma S, Stucke VM and Nigg EA. (2002). Science, Fang G, Yu H and Kirschner MW. (1998). Genes Dev., 12, 297, 2267–2270. 1871–1883. Masuda A, Maeno K, Nakagawa T, Saito H and Takahashi T. Gorbsky GJ, Chen RW and Murray AW. (1998). J. Cell Biol., (2003). Am. J. Pathol., 163, 1109–1116. 141, 1193–1205. Masuda A and Takahashi T. (2002). Oncogene, 21, 6884–6897.

Oncogene MAD1 downregulation causes aneuploidy A Kienitz et al 4310 Meraldi P, Draviam VM and Sorger PK. (2004). Dev. Cell, 7, Sieber OM, Heinimann K and Tomlinson IP. (2003). Nat. Rev. 45–60. Cancer, 3, 701–708. Michel L, Diaz-Rodriguez E, Narayan G, Hernando E, Murty Sironi L, Melixetian M, Faretta M, Prosperini E, Helin K and VV and Benezra R. (2004). Proc. Natl. Acad. Sci. USA, 101, Musacchio A. (2001). EMBO J., 20, 6371–6382. 4459–4464. Sudakin V, Chan GK and Yen TJ. (2001). J. Cell Biol., 154, Michel LS, Liberal V, Chatterjee A, Kirchwegger R, Pasche B, 925–936. Gerald W, Dobles M, Sorger PK, Murty VV and Benezra R. Sudo T, Nitta M, Saya H and Ueno NT. (2004). Cancer Res., (2001). Nature, 409, 355–359. 64, 2502–2508. Musacchio A and Hardwick KG. (2002). Nat. Rev. Mol. Cell Takahashi T, Haruki N, Nomoto S, Masuda A, Saji S, Biol., 3, 731–741. Osada H and Takahashi T. (1999). Oncogene, 18, 4295–4300. Nasmyth K, Peters JM and Uhlmann F. (2000). Science, 288, Tang Z, Bharadwaj R, Li B and Yu H. (2001). Dev. Cell, 1, 1379–1385. 227–237. Nomoto S, Haruki N, Takahashi T, Masuda A, Koshikawa T, Taylor S and McKeon F. (1997). Cell, 89, 727–735. Fujii Y and Osada H. (1999). Oncogene, 18, 7180–7183. Taylor SS, Scott MI and Holland AJ. (2004). Chromosome Peters JM. (2002). Mol. Cell, 9, 931–943. Res., 12, 599–616. Rajagopalan H and Lengauer C. (2004). Nature, 432, Vigneron S, Prieto S, Bernis C, Labbe JC, Castro A and 338–341. Lorca T. (2004). Mol. Biol. Cell, 15, 4584–4596. Rieder C, Cole R, Khodjakov A and Sluder G. (1995). J. Cell Vogel C, Kienitz A, Hofmann I, Muller R and Bastians H. Biol., 130, 941–948. (2004a). Oncogene, 23, 6845–6853. Shah JV and Cleveland DW. (2000). Cell, 103, 997–1000. Vogel C, Kienitz A, Muller R and Bastians H. (2004b). J. Biol. Shannon KB, Canman JC and Salmon ED. (2002). Mol. Biol. Chem., 280, 4025–4028. Cell, 13, 3706–3719. Wang X, Jin DY, Ng RW, Feng H, Wong YC, Cheung AL Shichiri M, Yoshinaga K, Hisatomi H, Sugihara K and and Tsao SW. (2002). Cancer Res., 62, 1662–1668. Hirata Y. (2002). Cancer Res., 62, 13–17. Yu H. (2002). Curr. Opin. Cell Biol., 14, 706–714.

Oncogene