MAD2-Induced Sensitization to Vincristine Is Associated with Mitotic Arrest and Raf/Bcl-2 Phosphorylation in Nasopharyngeal Carcinoma Cells

MAD2-induced sensitization to vincristine is associated with mitotic arrest and Raf/Bcl-2 phosphorylation in nasopharyngeal carcinoma cells

Xianghong Wang1, Dong-Yan Jin2,3, Hing Lok Wong1, Huichen Feng1, Yong-Chuan Wong1 and Sai Wah Tsao1,*

1Department of Anatomy 21 Sassoon Road, 1st Floor, Laboratory Block, Faculty of Medicine Building, Hong Kong; 2Department of Biochemistry 21 Sassoon Road, 1st Floor, Laboratory Block, Faculty of Medicine Building, Hong Kong

Mitotic arrest deficient 2 (MAD2)is thought to be a key microtubules, and is thought to be crucial for generating component of the mitotic checkpoint, which ensures the ‘wait’ signal to prevent the onset of anaphase in the accurate chromosome segregation. Reduced expression presence of microtubule disruption (Li and Benezra, of MAD2 protein is associated with mitotic checkpoint 1996; Amon, 1999; Wassmann and Benezra, 2001). abrogation and chromosomal instability in certain types of Inactivation of MAD2 either by microinjection of anti- human cancers. To explore the possibility of developing a MAD2 antibodies or knockout of the MAD2 gene led to novel strategy for the treatment of cancer based on defective mitotic checkpoint and chromosomal instabil- selective killing of mitotic checkpoint-defective or ity (Kallio et al., 1998; Dobles et al., 2000), indicating -competent cells, here we have investigated the effect of the essentiality of MAD2 for mitotic checkpoint control. MAD2 expression on cellular sensitivity to checkpoint- Recently, we and others have reported that decreased targeting anticancer drugs. We reintroduced MAD2 expression of MAD2 protein is associated with a protein in a mitotic checkpoint-defective nasopharyngeal defective mitotic checkpoint in human cancers (Kallio carcinoma cell line, CNE2, using an inducible expression et al., 1998; Wang et al., 2000, 2002)and that exogenous vector. We found that overexpression of MAD2 led to an MAD2 expression leads to the restoration of checkpoint increased sensitivity to vincristine, which was accompa- (Wang et al., 2002). However, the possible association nied by increased mitotic index and G2/M cell cycle between MAD2 expression and cellular sensitivity to arrest. In addition, increased phosphorylation of Raf, microtubule-targeting agents such as vincristine and MEK1/2 and Bcl-2 was observed in MAD2-overexpres- taxol, which remain front-line chemotherapeutic agents sing cells in response to vincristine. Furthermore, inhibi- in cancer treatment, has not been studied. Since mitotic tion of phosphorylation of MEK1/2 by its inhibitor checkpoint defect is common in human cancers (Cahill PD098059 led to reduced sensitivity to vincristine, which et al., 1998; Lengauer et al., 1998; Takahashi et al., 1999; was associated with decreased Bcl-2 phosphorylation. Our Kasai et al., 2002)and the cytotoxicity of most data suggest a role for MAD2 in the sensitization of microtubule-inhibiting agents used in the treatment of cancer cells to certain mitotic checkpoint-targeting antic- human cancer acts through induction of mitotic arrest ancer drugs. and apoptosis (Sherwood et al., 1994; Jordan et al., Oncogene (2003) 22, 109–116. doi:10.1038/sj.onc.1206069 1996), we hypothesized that there might be a difference in drug sensitivity between the mitotic checkpoint- Keywords: MAD2; Bcl-2; vincristine; Raf; Nasophar- defective and -competent human cancer cells. In yngeal carcinoma addition, as MAD2 expression is crucial for mitotic checkpoint control as previously reported (Kallio et al., 1998; Dobles et al., 2000; Wang et al., 2000, 2002), it is possible that MAD2 may play a role in cellular The proper segregation of chromosomes in anaphase is sensitivity to microtubule-disrupting agents. Therefore, essential for the integrity of the genome. The mitotic the expression of MAD2 may prove to be useful in checkpoint, also known as a spindle assembly check- selecting an efficient treatment for human cancer. point, ensures that cell division does not occur until all We have previously demonstrated that the nasophar- duplicated chromosomes align and attach to the spindle. yngeal carcinoma (NPC)cell line CNE2 shows relatively One of the key components of the mitotic checkpoint is low levels of MAD2 expression and a defective mitotic mitotic arrest deficient 2 (MAD2). MAD2 is found on checkpoint compared to other NPC cell lines and the the kinetochores, which link to both chromosomes and mitotic competent HeLa cells (Wang et al., 2000). In this study, we introduced exogenous MAD2 in CNE2 cells *Correspondence: S.W. Tsao, Department of Anatomy, University of by cotransfecting a MAD2 expression plasmid driven by Hong Kong, 21 Sassoon Road, 1st Floor, Laboratory, Faculty of Medicine Building, Hong Kong; E-mail: [email protected] an ecdysone-inducible promoter (pIND(SPl)–MAD2) 3D-Y Jin is a Leukemia and Lymphoma Society Scholar. and the plasmid expressing the heterodimeric ecdysone Received 20 May 2002; revised 20 September 2002; accepted 24 receptor (pVgRXR)(Wang et al., 2002). Three stable September 2002 transfectant clones were isolated (CNE2–MAD2–C1 to MAD2-induced sensitization to vincristine Wang et al 110 C3)by selecting in culture medium containing both However, there was no significant difference in IC50 Zeocin (100 mg/ml)and G418 (800 mg/ml). Exogenous values to taxol among the three cell lines MAD2 expression was induced by exposure to 5 mm (IC50 ¼ 0.3770.03 ng/ml; IC50 ¼ 0.2870.07 ng/ml; CNE1 7 915 ponasterone A. As shown in Figure 1a, ponasterone A IC50CNE3 ¼ 0.34 0.03 ng/ml)( P>0.05). These results treatment (+P)resulted in a two to eight-fold increase further support the findings on CNE2 cells that in MAD2 levels in CNE2 cells compared to the increased MAD2 is associated with cellular sensitivity untreated controls (ÀP). There was no significant to vincristine but not taxol in NPC cells. change in population doubling (PD)time in the cells It is generally accepted that microtubule-inhibiting cultured with [PD ¼ 22.471.3 h (Cl); PD ¼ 21.972.1 h agents such as nocodazole and vincristine can induce (C2); PD ¼ 22.671.1 h (C3)] and without apoptosis following a cell cycle arrest at the G2/M phase [PD ¼ 23.571.9 h (Cl); PD ¼ 20.972.1 h (C2); (Sherwood et al., 1994; Jordan et al., 1996; Minn et al., PD ¼ 23.872.3 h (C3)] ponasterone A, indicating that 1996; Wahl et al., 1996). To investigate whether MAD2- ponasterone treatment alone did not affect growth rate induced sensitization to vincristine is mechanistically in these transfectants. linked to the induction of G2/M arrest and apoptosis, The parental cell line CNE2 and three stable we evaluated the apoptosis rate and the mitotic index of transfectant clones were then tested for their sensitivity vincristine-treated cells in the presence and absence of to two microtubule-inhibiting anticancer agents, vincris- ponasterone A. When we stained the vincristine-treated tine (Figure 1b)and taxol (Figure 1c),using colony- cells with 40 6-diamidino-2-phenylindole (DAPI), there forming assay in the presence (+P)or absence ( ÀP)of were no significant alterations in either apoptosis rate or ponasterone A. Both vincristine and taxol inhibit cell mitotic index in the control CNE2 cells cultured in the proliferation by interfering with microtubule dynamics. presence or absence of ponasterone A (Figure 2), However, it is noteworthy that they have opposite indicating that ponasterone A treatment alone did not effects on microtubule polymerization. Whereas vincris- affect vincristine-induced apoptosis and mitotic index. tine binds to tubulin thereby inhibiting the formation of However, in the transfectants, the percentage of microtubules, taxol binds to polymeric microtubules apoptotic cells was shown to be much higher in the and prevents their disassembly. Thus, vincristine and cells cultured in ponasterone A (solid columns)than in taxol represent two contrasting examples of microtubule the cells cultured without ponasterone A (open columns) inhibitors. We noted that the colony-forming ability of (Figure 2a). In addition, the percentage of apoptotic CNE2 and all three MAD2-transfected clones decreased cells was associated with an increased percentage of cells progressively upon treatment with increasing doses of arrested at mitosis (Figure 2b). Moreover, the expres- vincristine, if compared to the untreated controls, either sion of MAD2 correlated with the hypersensitivity to in the presence or absence of ponasterone A (Figure 1b). vincristine reflected by the decreased colony-forming Thus, these cells are sensitive to vincristine regardless of ability (Figure 1b), and with the induction of apoptosis the MAD2 status. In addition, there was no difference in and mitotic arrest (Figure 2). For instance, the CNE2– cellular sensitivity to vincristine or taxol in the parental MAD2–C3 cells showing the highest MAD2 levels CNE2 cells cultured in the presence or absence of displayed the highest sensitivity to vincristine. However, ponasterone A, indicating that ponasterone A treatment in the cells treated with taxol, there was no significant alone did not have any effect on cellular sensitivity to difference in the percentage of either apoptotic or both drugs (Figure 1b and c). However, in the CNE2– mitotic cells between the cells cultured in the presence MAD2 cells, with the addition of ponasterone A and absence of ponasterone A (data not shown). One (solidcolumns), a significant reduction of colony-form- argument would be that these cells failed to accumulate ing ability was observed in the cells treated with in mitosis simply because the taxol concentrations used vincristine than in the cells cultured without ponaster- in this study were too low. However, as shown in one A (open columns)( Po0.05)(Figure 1b).These Figure 1, the cytotoxicity of taxol (10–90% colony- results indicate that the exogenously introduced MAD2 forming inhibition)was similar to or higher than that of conferred an increased sensitivity to vincristine in CNE2 vincristine (5–50% colony-forming inhibition). There- cells. In contrast, after treatment with taxol (Figure 1c), fore, the failure to observe mitotic arrest in taxol-treated there was only a slight decrease in colony-forming cells was unlikely due to the inadequate concentration of ability in the cells cultured in ponasterone A (+P), and taxol in this study. the differential sensitivity was not statistically significant To investigate the mechanisms involved in the compared to the cells cultured in the absence of MAD2-induced hypersensitivity to vincristine, we as- ponasterone A (ÀP, P>0.05). sessed cell cycle profiles of vincristine-treated CNE2 cells To further confirm these findings, we studied three by the use of flow cytometry. As shown in Figure 3, in additional NPC cell lines that have been shown to have the cells expressing exogenous MAD2 (+P), a clear G2/ relatively high (CNE1 and 915)or low (CNE3)levels of M arrest was observed in the CNE2–MAD2–C3 cells MAD2 protein (Wang et al., 2000). As shown in beginning at 24-h postexposure time. A decreased Figure 1d, we found that CNE1 and 915 cells were percentage of G2/M cells was found at 72 h after much more sensitive to vincristine with IC50 (Inhibition exposure to vincristine (3 ng/ml), which was accompa- concentration of 50% colony-forming ability)values of nied by an increase in the percentage of sub-Gl phase 1.3170.40 and 0.7570.34 ng/ml, respectively, compared cells. In contrast, in the cells expressing low levels of to the CNE3 cells (IC50 ¼ 2.6370.76 ng/ml)( Po0.05). MAD2 (ÀP), there was no evidence of G2/M arrest and

Oncogene MAD2-induced sensitization to vincristine Wang et al 111

Figure 1 Effect of MAD2 expression on cellular sensitivity to vincristine and taxol. A 1.5 kb fragment containing the full-length human MAD2 cDNA was cloned into the EcoRI and Xbal sites of an expression vector pIND(SP1)driven by the ecdysone-inducible promoter (Invitrogen, CA, USA)(Wang et al., 2002). The resulting construct pIND(SPl)–MAD2 (conferring neomycin resistance) was then cotransfected with pVgRXR vector (conferring Zeocin resistance), which permits the formation of a functional heterodimeric ecdysone receptor in mammalian cells. Using the calcium phosphate method, the stable pIND(SPl)–MAD2 and pVgRXR transfectants were generated by selecting in both neomycin (800 mg/ml)and Zeocin (100 mg/ml). To induce MAD2 expression in the transfectants, 5 mm of ponasterone A (Invitrogen, CA, USA)was added. ( a)Introduction of exogenous MAD2 expression in CNE2–MAD2 cells by exposure to ponasterone A (+P, 5 mm)for 24 h. A total of 30 mg of protein was analyzed in the cells cultured in the presence (+P)and absence (ÀP)of ponasterone A by Western blotting. Expression of actin was tested as an internal control for equal loading. Note that increased MAD2 expression is shown in the transfected cells cultured in ponasterone A. (b–d)Colony-forming ability after continuous exposure to vincristine (b)and taxol (c)in CNE2 cells and additional NPC cell lines CNE1, 915 and CNE3 ( d). In total, 700 cells were plated in six-well plates, resulting in approximately 100 colonies per well after 2 weeks culture in the presence (+P)or absence ( ÀP)of ponasterone A (5 mm). Vincristine or taxol was added 24 h after plating the cells, and colonies that consisted of more than 50 cells were scored and compared to the solvent-treated controls after 12 days. Two wells were used for each concentration and each experiment was repeated at least three times. Results represent the means and standard error of the means

Oncogene MAD2-induced sensitization to vincristine Wang et al 112

Figure 2 Induction of apoptosis (a)and mitotic arrest ( b)by vincristine in CNE2 cells. The cells were treated with vincristine (3 ng/ml) at indicated time points in the presence (+P)or absence ( ÀP)of ponasterone A and ﬁxed. The percentage of apoptotic (presence of nuclear fragmentation)and mitotic (presence of chromosome condensation)cells was determined by staining the cells with DAPI (0.05 mg/ml). Note that increased mitotic index and apoptosis rate are found in the cells expressing exogenous MAD2 protein after treated with vincristine compared to the cells with low levels of MAD2

the percentage of sub-G1 cells was much lower. Similar disrupted MAD2-dependent checkpoint have an in- results were also found in CNE2–MAD2–Cl and C2 creased sensitivity to microtubule-disrupting agents such cells (data not shown). These results indicate that as taxol and nocodazole (Zhang and Lees, 2001)and expression of exogenous MAD2 protein induced G2/ that low levels of MAD2 expression correlated with an M arrest in response to vincristine treatment. One increased sensitivity to microtubule-disrupting agents interpretation is that MAD2 plays a role in vincristine- (Li and Benezra, 1996). However, the apoptotic studies induced mitotic arrest, which may have triggered in these reports were performed at a relatively short apoptosis. This hypothesis was also supported by the exposure time (18 h)(Zhang and Lees, 2001).In the flow cytometric data that a G2/M peak was clearly seen present study, we did not observe a difference in in the MAD2-overexpressing cells at an earlier time apoptosis rate between the cells expressing low and high point of 24–48 h but was less evident at later time points levels of MAD2 until 48 h after drug treatment (72–96 h), which was accompanied by the increased sub- (Figure 2a). Our apoptosis studies were also confirmed G1 cells (Figure 3). In addition, the percentage of sub- by colony-forming assay, which measures a cell’s ability G1 cells was increased with extended exposure time to divide for at least five times after drug treatment. It is (from 5.6% at 24 h to 43.4% at 96 h). These results agree to be determined whether cell-type-specific effects may generally with the DAPI staining data that the explain the differential responses to microtubule-disrup- percentage of apoptotic cells was increased with tion. In this regard, our findings are generally consistent prolonged exposure to vincristine (Figure 2a). It is with a recent report on human adult T-cell leukemia in understood that cells may undergo apoptosis subsequent which a loss of mitotic checkpoint conferred cellular to the mitotic block (Minn et al., 1996). resistance to microtubule-disrupting agents such as In this study, there was no evidence of mitotic arrest vincristine (Kasai et al., 2002). in response to microtubule-disruption upon treatment A role for the activation of Raf, MAPK and p38 with vincristine in the mitotic checkpoint-defective kinases in the mitotic checkpoint has been well described CNE2 cells expressing insufficient amounts of MAD2 (Minshull et al., 1994; Takenaka et al., 1997, 1998; (Figures 2 and 3, ÀP). In contrast, an increased Wang et al., 1997). Induction of Raf/MAPK activation percentage of G2/M cells was observed in the MAD2- by microtubule-inhibiting agents has also been reported overexpressing cells (Figures 2 and 3, +P). These results (Hayne et al., 2000). With this in mind, we questioned if further support the notion that MAD2 acts as a the Raf/MAPK pathway played a role in MAD2- molecular switch for mitotic checkpoint control in induced sensitization to vincristine. The activation of human cancers (Pennisi, 1998). Previously, it has been Raf and MEKl/2 in CNE2 and CNE2–MAD2–C3 cells reported that human lung cancer cells (A549)with a was examined by Western blot analysis using antibodies

Oncogene MAD2-induced sensitization to vincristine Wang et al 113 we also noted that while early induction of MAD2 (C3, +P, 1–4 h)was associated with increased phospho-Raf (C3, +P, 2–4 h), there was no correlation between MAD2 accumulation and Raf phosphorylation at later time points (Figure 4a and b, C3, +P, 6–8 h). In contrast, phosphorylation of MEKl/2 was sustained even after the reduction of Raf phosphorylation (Figure 4b, C3, +P), compared to cells cultured in the absence of ponasterone A. At the moment, we do not fully understand why and how Raf phosphorylation is downregulated in these cells. One possibility is that Raf phosphorylation might trigger the MAD2-independent negative feedback mechanism to prevent uncontrolled activation of the pathway. Probably the ubiquitously expressed cellular inhibitors of Raf kinase such as Raf kinase inhibitor protein (RKIP)may play a role in this regulation (Yeung et al., 1999). Since Raf phosphorylation is suggested to be responsible for inducing phosphorylation and thus inactivation of Bcl-2 and Bcl-xl, which leads to apoptosis in certain cell types (Blagosklonny et al., 1997; Huang et al., 1999; Fan et al., 2000; Stadheim et al., 2001), we next studied the expression of Bcl-2 and Bcl-xl in CNE2–MAD2–C3 cells. We found that in the presence of ponasterone A (+P), there was an increase in the phosphorylated Bcl-2 protein (Figure 4, C3, upper band), beginning at 4 h after exposure to vincristine, and its levels were further increased at later time points. In contrast, in the cells cultured in the absence of ponasterone (ÀP), the appearance of phosphorylated Bcl-2 was delayed (beginning at 6 h postexposure)and the relative amounts were much lower than in the cells expressing high levels of MAD2 (+P). However, there was no evidence of phosphorylated Bcl-xl either in the presence or absence of ponasterone A, indicating that Bcl-xl may not be involved in the vincristine-induced apoptosis in these cells. Currently, we do not understand which kinase is Figure 3 Effect of MAD2 expression on vincristine-induced cell responsible for the phosphorylation of Bcl-2 in CNE2– cycle perturbation in CNE2–MAD2–C3 cells. Cell cycle distribu- MAD2–C3 cells treated with vincristine. We noted that tion was analyzed using a published protocol (Wang et al, 1998, ASK1/JNK1 has been suggested to be important in 1999). Briefly, the cells (5 Â 105)were fixed in 70% ethanol and stained with propidium iodide (50 mg/ml)and RNAse (1 mg/ml)for taxol-induced Bcl-2 phosphorylation and its inactivation 30 min at 371C. They were then analyzed on an EPICS profile in certain cell types (Yamamoto et al., 1999). We are in analyzer (Coulter, Miami, FL, USA). Results represent three the process of characterizing whether the ASK1/JNK1 independent experiments pathway is involved in the Bcl-2 phosphorylation in NPC cells. We did not observe the degradation of Bcl-2 protein after treatment with vincristine (Figure 4c). Apparently, against phosphorylated Raf (Ser259)and MEKl/2 Bcl-2 is phosphorylated and dephosphorylated on (Ser217/221). As shown in Figure 4b (+P), hyperpho- different sites by a variety of kinases and phosphatases sphorylated Raf and MEKl/2 were found at 2 h after (Srivastava et al., 1998; Dimmeler et al., 1999). It exposure to vincristine in the presence of ponasterone A, remains controversial as to whether phosphorylation of which corresponds to increased expression of MAD2 Bcl-2 leads to proteosome-dependent degradation (Hal- (Figure 4a, +P)in CNE2–MAD2–C3 cells but not in dar et al., 1995; Brichese et al., 2002; Huang and the control CNE2. However, the levels of unpho- Cidlowski, 2002). Thus, further investigations are sphorylated Raf and MEK1/2 remained unchanged required to elucidate which sides of Bcl-2 are phos- (Figure 4b). In contrast, there was no evidence of Raf or phorylated and exactly how Bcl-2 phosphorylation MEKl/2 hyperphosphorylation in the absence of po- contributes to vincristine-induced apoptosis in CNE2– nasterone A (ÀP), which corresponds to the insufficient MAD2 cells. expression of MAD2 (Figure 4a, ÀP). These results Since it was reported that taxol induced Raf/Bcl-2 suggest that the exogenous expression of MAD2 phosphorylation and apoptosis in certain cell types correlates with the activation of Raf/MEK. However, (Blagosklonny et al., 1997; Yamamoto et al., 1999), we

Oncogene MAD2-induced sensitization to vincristine Wang et al 114

Figure 4 Effect of MAD2 expression on vincristine-induced Raf, MEKl/2, Bcl-2 and Bcl-xl phosphorylation in CNE2–MAD2–C3 cells. Approximately 30 mg protein was separated on a 15% SDS-polyacrylamide gel, transferred to nitrocellulose and incubated with antibodies against MAD2 (monoclonal, 1 : 500; Jin et al., 1998), phospho-Raf (Ser259) (polyclonal, 1 : 500; Cell Signaling Technology, USA), Raf (polyclonal, 1 : 200; Santa Cruz, USA), phospho-MEKl/2 (Ser217/221) (polyclonal, 1 : 500; Cell Signaling Technology, USA), MEKl/2 (polyclonal, 1 : 500; Cell Signaling Technology, USA), Bcl-2 (polyclonal, 1 : 300; Santa Cruz, USA), Bcl-xl (polyclonal, 1 : 300; Santa Cruz, USA), or actin (polyclonal, 1 : 200; Roche, Germany). (a–d)Western blot analysis was performed on CNE2 and CNE2–MAD2–C3 cells treated with vincristine (3 ng/ml)( a–c)or taxol (1.5 ng/ml)( d)at indicated time points in the presence (+P)or absence (ÀP)of ponasterone A. Note that increased Raf, MEK1/2 and Bcl-2 phosphorylation is found in the cells expressing exogenous MAD2 protein (+P)after exposure to vincristine. There is no increase in Raf/Bcl-2 phosphorylation after exposure to taxol. E: Effect of MEK1/2 inhibitor PD098059 on MAD2-induced sensitization to vincristine in CNE2-MAD2-C3 cells. (1)Cells were treated with vincristine (3 ng/ml)and PD098059 for 24 h in the presence (+P)or absence ( ÀP)of ponasterone A. (2)Effect of PD098059 on colony-forming ability after continuous exposure to vincristine. PD098059 was added together with vincristine for 12 days in the presence or absence of ponasterone A. Two wells were used for each concentration and each experiment was repeated at least three times. Results represent the means and standard error of the means

Oncogene MAD2-induced sensitization to vincristine Wang et al 115 next studied the effect of taxol on Raf and Bcl-2 One possibility is that MAD2 acts as an essential expression in CNE2 cells. As shown in Figure 4d, we did component of a signaling pathway that leads to Raf and not observe any changes in the levels of phosphorylated MAPK activation. Vincristine and other microtubule- Raf or Bcl-2 proteins after exposure to taxol (1.5 ng/ml), disrupting agents may target the mitotic checkpoint and regardless of MAD2 status. These results indicate that activate MAD2, thereby inducing Raf and MAPK taxol-induced cell death in CNE2 cells may be indepen- activation. In addition, in the present study, the dent of Raf/Bcl-2 phosphorylation as previously activation of the MAPK pathway was also accompanied reported in human ovarian cancer cells (Wang TH by the increased phosphorylated Bcl-2 in the MAD2- et al., 1999). overexpressing cells compared to the cells with To exclude the possibility that ponasterone A insufficient MAD2 (Figure 4c). Inactivation of MEK1/ treatment alone might affect the phosphorylation of 2 by inhibiting its phosphorylation resulted in the Raf, MEKl/2 and Bcl-2, we also studied the parental decreased Bcl-2 phosphorylation and increased colony- CNE2 cells cultured in identical conditions to forming ability in MAD2-expressing cells (Figure 4e). CNE2–MAD2–C3 as a negative control. As shown in These results implicate that the MAD2-induced hyper- Figure 4 (CNE2), the expression levels of all three sensitivity to vincristine may be through phosphoryla- proteins were similar in CNE2 cells (7P)compared to tion of Bcl-2 as a result of Raf/MAPK activation in the CNE2–MAD2–C3 cultured in the absence of CNE2 cells. ponasterone A, suggesting that ponasterone A Another interesting observation in the present study is treatment had no effect on the phosphorylation of these that exogenous MAD2 expression did not significantly proteins. affect the sensitivity of CNE2 cells to taxol. It is To further support the hypothesis that Raf/Bcl-2 understood that taxol and vincristine are two contrast- phosphorylation is essential for MAD2- induced ing examples of microtubule-interfering agents. Vincris- sensitization to vincristine, we treated CNE2–MAD2– tine binds to tubulin subunits and inhibits microtubule C3 cells with a MEK1/2 inhibitor, PD098059, which has polymerization, thus disrupting spindle dynamics and been shown to inactivate specifically MEK1/2 signaling leading to mitotic block. In sharp contrast, taxol by reducing its phosphorylation (Huang et al., 1999). enhances tubulin polymerization by inhibiting micro- We found that after exposure to PD098059 tubule disassembly, which then prevents the breakdown (50 and 75 mm)for 24 h, the vincristine-treated CNE2– of microtubules that is also required for progression of MAD2–C3 cells showed decreased MEK1/2 phosphor- the cell division cycle. The function of MAD2, however, ylation in the presence of ponasterone A, which was is to sense microtubule attachment and once an accompanied by a dose-dependent decrease in Bcl-2 unattached kinetochore is detected, MAD2 generates phosphorylation (Figure 4e1, upper band). In addition, the ‘wait’ signal to stop progression to mitosis (Zhang the colony-forming ability of these cells was enhanced and Lees, 2001). In the cells treated with vincristine, by PD098059 treatment (75 mm)against vincristine unattached kinetochores are present, while in the cells (Figure 4e2, +P+PD098059), compared to the treated with taxol, most kinetochores are attached to the untreated cells (+P). However, we did not observe microtubules. Therefore, it is not surprising that MAD2 significant changes in either phosphorylation of MEKl/2 governs the cellular sensitivity to vincristine but not to or the colony-forming ability in these cells cultured in taxol. Presumably, vincristine and taxol may target the absence of ponasterone A (Figure 4e2, different steps of the same mitotic checkpoint. ÀP+PD098059; ÀP). These results suggest that In addition, the fact that we did not find in- MAD2 expression is associated with Raf/MAPK creased phosphorylation of Raf/Bcl-2 in CNE2– activation and that Bcl-2 phosphorylation is essential MAD2–C3 cells treated with taxol (Figure 4d)further for MAD2-induced sensitization to vincristine. The fact support this hypothesis. Nevertheless, the functions of that Bcl-2 phosphorylation occurred subsequently to the MAD2 are still largely unknown, neither is it clear activation of Raf/MAPK indicates that activation of exactly how mammalian cells escape mitotic arrest Raf/MAPK may be responsible for the inactivation of induced by microtubule disruption. Thus, further Bcl-2, as previously suggested (Blagosklonny et al., studies are required to elucidate the physiological 1997). roles of MAD2 and the integration of its checkpoint Activation of Raf kinase through phosphorylation, function with the programs of cellular proliferation and mainly at serine residues, has been demonstrated apoptosis. during mitosis (Laird et al., 1995; Lovric and Moelling, In summary, in this study we report for the first time 1996). Recently, it has been suggested that Raf that exogenous MAD2 expression resulted in an activation may be responsible for the cytotoxicity of increased sensitivity to a microtubule-disrupting antic- certain microtubule-2 disrupting agents including vin- ancer agent, vincristine, in NPC cells. Inactivation of cristine, since activated Raf kinase is essential for Bcl-2 by phosphorylation, possibly through the activa- phosphorylation of Bcl-2 leading to apoptosis tion of the Raf/MAPK pathway, may be responsible for (Blagosklonny et al., 1997; Huang et al., 1999; Fan the sensitization effect in these cells. Our results et al., 2000). In the present study, we found increased implicate a novel mechanistic link between MAD2 and phosphorylation of Raf and MEK1/2 in the vincristine- cellular sensitivity to certain microtubule-interrupting treated MAD2-overexpressing CNE2 cells (Figure 4b). agents.

Oncogene MAD2-induced sensitization to vincristine Wang et al 116 References

Amon A. (1999). Curr. Opin. Genet. Dev., 9, 69–75. Pennisi E. (1998). Science, 279, 477–478. Blagosklonny MV Giannakakou P el Deiry WS Kingston DG Sherwood SW, Sheridan JP, Schimke RT. (1994). Exp. Cell Higgs PI Neckers L Fojo T. (1997). Cancer Res., 57, 130–135. Res., 215, 373–379. Brichese L Barboule N Heliez C Valette A. (2002). Exp. Cell. Srivastava RK, Srivastava AR, Korsmeyer SJ, Nesterova M, Res., 278, 101–111. Cho-Chung YS, Longo DL. (1998). Mol Cell. Bio., 18, 3509– Cahill DP, Lengauer C, Yu J, Riggins GJ, Willson JK, 3617. Markowitz SD, Kinzler KW, Vogelstein B. (1998). Nature, Stadheim TA, Xiao H, Eastman A. (2001). Cancer Res., 61, 392, 300. 1533–1540. Dimmeler S, Breitschopf K, Haendeler J, Zeiher AM. (1999). J Takahashi T, Haruki N, Nomoto S, Masuda A, Saji S, Osada Exp Med, 189, 1815–1818 H, Takahashi T. (1999). Oncogene, 18, 4295. Dobles M, Liberal V, Scott ML, Benezra R, Sorger PK. Takenaka K, Gotoh Y, Nishida E. (1997). J. Cell Biol., 136, (2000). Cell, 101, 635–645. 1091–1097. Fan M, Du L, Stone AA, Gilbert KM, Chambers TC. (2000). Takenaka K, Moriguchi T, Nishida E. (1998). Science, 280, Cancer Res., 60, 6403–6407. 599–602. Haldar S, Jena N, Croce CM. (1995). Proc. Natt. Acad. Sci. Wahl AF, Donaldson KL, Fairchild C, Lee FY, Foster SA, USA, 92, 4507–4611. Demers GW, Galloway DA. (1996). Nat.Med., 2, 72–79. Hayne C, Tzivion G, Luo Z. (2000). J. Biol. Chem., 275, Wang TH, Popp DM, Wang HS, Saitoh M, Mural JG, Henley 31876–31882. DC, Ichijo H, Wimalasena JJ. (1999). Biol Chem., 274, 8208– Huang ST, Cidlowski JA. (2002). FASEB J., 16, 825–832. 8216. Huang Y, Sheikh MS, Fomace AJ, Holbrook NJ. (1999). Wang X, Jin DY, Ng RW, Feng H, Wong YC, Cheung AL, Oncogene, 18, 3431–3439. Tsao SW. (2002). Cancer Res., 62, 1662–1668. Jin DY, Spencer F, Jeang KT. (1998). Cell, 93, 81–91. Wang X, Jin DY, Wong YC, Cheung AL, Chun AC, Lo AK, Jordan MA, Wendell K, Gardiner S, Deny WB, Copp H, Liu Y, Tsao SW. (2000). Carcinogenesis, 21, 2293–2297. Wilson L. (1996). Cancer Res., 56, 816–825. Wang X, Liu Y, Chow LS, Wong SC, Tsao SW, Kwong DL, Kallio M, Weinstein J, Daum JR, Burke DJ, Gorbsky, GJ. Wang J, Sham JS, Nicholls JM. (1999). Int. J..Oncol., 15, (1998). J. Cell Biol., 141, 1393–1406. 1097–1102. Kasai T, Iwanaga Y, Iha H, Jeang KT. (2002). J. Biol. Chem., Wang X, Wong SC, Pan J, Tsao SW, Fung KH, Kwong DL, 277, 5187–5193. Sham JS, Nicholls JM. (1998). Cancer Res., 58, 5019–5022. Laird AD, Taylor SJ, Oberst M, Shalloway D. (1995). J. Biol. Wang X, Zhai Y, Ferrell Jr JE (1997). J. Cell Biol., 137, 433–443. Chem., 270, 26742–26745. Wassmann K, Benezra R. (2001). Curr. Opin. Genet. Dev., 11, Lengauer C, Kinzler KW, Vogelstein B. (1998). Nature, 396, 83–90. 643–649. Yamamoto K, Ichijo H, Korsmeyer SJ. (1999). Mol. Cell Biol., Li Y, Benezra R. (1996). Science, 274, 246–248. 19, 8469–8478 Lovric J, Moelling K. (1996). Oncogene, 12, 1109–1116. Yeung K, Seitz T Li S, Janosch P, McFerran B, Kaiser C, Fee Minn AJ, Boise LH, Thompson CB. (1996). Genes Dev., 10, F, Katsanakis KD, Rose DW, Mischak H, SedivyJM, Kolch 2621–2631. W (1999). Nature, 401, 173–177. Minshull J, Sun H, Tonks NK, Murray AW. (1994). Cell, 79, Zhang Y, Lees E. (2001). Mol. Cell. Biol., 21, 5190–5199. 475–486.

Oncogene