Association of MAD2 expression with mitotic checkpoint competence in hepatoma cells

Chromosomal instability (CIN) refers to high rates of chromosomal gains and losses and is a major cause of genomic instability of cells. It is thought that CIN caused by loss of mitotic checkpoint contributes to carcinogenesis. In this study, we evaluated the competence of mitotic checkpoint in hepatoma cells and investigated the cause of mitotic checkpoint defects. We found that 6 (54.5%) of the 11 hepatoma cell lines were defective in mitotic checkpoint control as monitored by mitotic indices and flow-cytometric analysis after treatment with microtubule toxins. Interestingly, all 6 hepatoma cell lines with defective mitotic checkpoint showed significant underexpression of mitotic arrest deficient 2 (MAD2), a key mitotic checkpoint protein. The level of MAD2 underexpression was significantly associated with defective mitotic checkpoint response (p<0.001). In addition, no mutations were found in the coding sequences of MAD2 in all 11 hepatoma cell lines. Our findings suggest that MAD2 deficiency may cause a mitotic checkpoint defect in hepatoma cells.     

NEK2A interacts with MAD1 and possibly functions as a novel integrator of the spindle checkpoint signaling

Chromosome segregation in mitosis is orchestrated by protein kinase signaling cascades. A biochemical cascade named spindle checkpoint ensures the spatial and temporal order of chromosome segregation during mitosis. Here we report that spindle checkpoint protein MAD1 interacts with NEK2A, a human orthologue of the Aspergillus nidulans NIMA kinase. MAD1 interacts with NEK2A in vitro and in vivo via a leucine zipper-containing domain located at the C terminus of MAD1. Like MAD1, NEK2A is localized to HeLa cell kinetochore of mitotic cells. Elimination of NEK2A by small interfering RNA does not arrest cells in mitosis but causes aberrant premature chromosome segregation. NEK2A is required for MAD2 but not MAD1, BUB1, and HEC1 to associate with kinetochores. These NEK2A-eliminated or -suppressed cells display a chromosome bridge phenotype with sister chromatid inter-connected. Moreover, loss of NEK2A impairs mitotic checkpoint signaling in response to spindle damage by nocodazole, which affected mitotic escape and led to generation of cells with multiple nuclei. Our data demonstrate that NEK2A is a kinetochore-associated protein kinase essential for faithful chromosome segregation. We hypothesize that NEK2A links MAD2 molecular dynamics to spindle checkpoint signaling.     

Aneuploid colon cancer cells have a robust spindle checkpoint

Colon cancer cells frequently display minisatellite instability (MIN) or chromosome instability (CIN). While MIN is caused by mismatch repair defects, the lesions responsible for CIN are unknown. The observation that CIN cells fail to undergo mitotic arrest following spindle damage suggested that mutations in spindle checkpoint genes may account for CIN. However, here we show that CIN cells do undergo mitotic arrest in response to spindle damage. Although the maximum mitotic index achieved by CIN lines is diminished relative to MIN lines, CIN cells clearly have a robust spindle checkpoint. Consistently, mutations in spindle checkpoint genes are rare in human tumours. In contrast, the adenomatous polyposis coli (APC) gene is frequently mutated in CIN cells. Significantly, we show here that expression of an APC mutant in MIN cells reduces the mitotic index following spindle damage to a level observed in CIN cells, suggesting that APC dysfunction may contribute to CIN.     

Integrin-linked kinase regulates senescence in an Rb-dependent manner in cancer cell lines

Anti-integrin-linked kinase (ILK) therapies result in aberrant mitosis including altered mitotic spindle organization, centrosome declustering and mitotic arrest. In contrast to cells that expressed the retinoblastoma tumor suppressor protein Rb, we have shown that in retinoblastoma cell lines that do not express Rb, anti-ILK therapies induced aberrant mitosis that led to the accumulation of temporarily viable multinucleated cells. The present work was undertaken to: 1) determine the ultimate fate of cells that had survived anti-ILK therapies and 2) determine whether or not Rb expression altered the outcome of these cells. Our data indicate that ILK, a chemotherapy drug target is expressed in both well-differentiated, Rb-negative and relatively undifferentiated, Rb-positive retinoblastoma tissue. We show that small molecule targeting of ILK in Rb-positive and Rb-deficient cancer cells results in increased centrosomal declustering, aberrant mitotic spindle formation and multinucleation. However, anti-ILK therapies in vitro have different outcomes in retinoblastoma and glioblastoma cell lines that depend on Rb expression. TUNEL labeling and propidium iodide FACS analysis indicate that Rb-positive cells exposed to anti-ILK therapies are more susceptible to apoptosis and senescence than their Rb-deficient counterparts wherein aberrant mitosis induced by anti-ILK therapies exhibit mitotic arrest instead. These studies are the first to show a role for ILK in chemotherapy-induced senescence in Rb-positive cancer lines. Taken together these results indicate that the oncosuppressive outcomes for anti-ILK therapies in vitro, depend on the expression of the tumor suppressor Rb, a known G₁ checkpoint and senescence regulator.     

Checkpoint genes required to delay cell division in response to nocodazole respond to impaired kinetochore function in the yeast Saccharomyces cerevisiae.

Inhibition of mitosis by antimitotic drugs is thought to occur by destruction of microtubules, causing cells to arrest through the action of one or more mitotic checkpoints. We have patterned experiments in the yeast Saccharomyces cerevisiae after recent studies in mammalian cells that demonstrate the effectiveness of antimitotic drugs at concentrations that maintain spindle structure. We show that low concentrations of nocodazole delay cell division under the control of the previously identified mitotic checkpoint genes BUB1, BUB3, MAD1, and MAD2 and independently of BUB2. The same genes mediate the cell cycle delay induced in ctf13 mutants, limited for an essential kinetochore component. Our data suggest that a low concentration of nocodazole induces a cell cycle delay through checkpoint control that is sensitive to impaired kinetochore function. The BUB2 gene may be part of a separate checkpoint that responds to abnormal spindle structure.     

p53-Independent Apoptosis and p53-Dependent Block of DNA Rereplication Following Mitotic Spindle Inhibition in Human Cells

We have studied the response of human transformed cells to mitotic spindle inhibition. Two paired cell lines, K562 and its parvovirus-resistant KS derivative clone, respectively nonexpressing and expressing p53, were continuously exposed to nocodazole. Apoptotic cells were observed in both lines, indicating that mitotic spindle impairment induced p53-independent apoptosis. After a transient mitotic delay, both cell lines exited mitosis, as revealed by flow-cytometric determination of MPM2 antigen and cyclin B1 expression, coupled to cytogenetic analysis of sister centromere separation. Both cell lines exited mitosis without chromatid segregation. K562 p53-deficient cells further resumed DNA synthesis, giving rise to cells with a DNA content above 4C, and reentered a polyploid cycle. In contrast, KS cells underwent a subsequent G1 arrest in the tetraploid state. Thus, G1 arrest in tetraploid cells requires p53 function in the rereplication checkpoint which prevents the G1/S transition following aberrant mitosis; in contrast, p53 expression is dispensable for triggering the apoptotic response in the absence of mitotic spindle.     

Differential kinetochore requirements for establishment and maintenance of the spindle checkpoint are dependent on the mechanism of checkpoint activation in Saccharomyces cerevisiae

The spindle checkpoint in the yeast Saccharomyces cerevisiae is an intracellular signal transduction pathway comprised of two branches that inhibit two different mitotic transitions in cells treated with benzimidazole drugs such as nocodazole. The kinetochore is an integral component of the MAD2 branch of the spindle checkpoint pathway. Current models propose that the kinetochore is required for both the establishment and maintenance of the spindle checkpoint but a role for the kinetochore in the maintenance of spindle checkpoint in yeast has never been directly tested. We used a temperature sensitive ndc10-1 mutant to inactivate kinetochores before and after arresting cells in mitosis to determine the role of kinetochores in the establishment and maintenance of the spindle checkpoint. We show that both establishment and maintenance requires kinetochore function in response to spindle damage induced by benzimidazole drugs. Excess expression of the Mps1 protein kinase causes wild type cells and ndc10-1 cells to arrest in mitosis. Unlike the spindle checkpoint arrest activated by benzimidazoles, this arrest can be maintained independently of kinetochores. The arrest induced by excess Mps1p is independent of BUB2. Therefore, mitotic arrest induced by excess Mps1p expression is due to the action of the MAD2 branch of the spindle checkpoint pathway and excess Mps1p acts downstream of the kinetochore.     

Differential Kinetochore Requirements for Establishment and Maintenance of the Spindle Checkpoint Are Dependent on the Mechanism of Checkpoint Activation in Saccharomyces cerevisiae

The spindle checkpoint in the yeast Saccharomyces cerevisiae is an intracellular signal transduction pathway comprised of two branches that inhibit two different mitotic transitions in cells treated with benzimidazole drugs such as nocodazole. The kinetochore is an integral component of the MAD2 branch of the spindle checkpoint pathway. Current models propose that the kinetochore is required for both the establishment and maintenance of the spindle checkpoint but a role for the kinetochore in the maintenance of spindle checkpoint in yeast has never been directly tested. We used a temperature sensitive ndc10-1 mutant to inactivate kinetochores before and after arresting cells in mitosis to determine the role of kinetochores in the establishment and maintenance of the spindle checkpoint. We show that both establishment and maintenance requires kinetochore function in response to spindle damage induced by benzimidazole drugs. Excess expression of the Mps1 protein kinase causes wild type cells and ndc10-1 cells to arrest in mitosis. Unlike the spindle checkpoint arrest activated by benzimidazoles, this arrest can be maintained independently of kinetochores. The arrest induced by excess Mps1p is independent of BUB2. Therefore, mitotic arrest induced by excess Mps1p expression is due to the action of the MAD2 branch of the spindle checkpoint pathway and excess Mps1p acts downstream of the kinetochore.     

Human CENP-I specifies localization of CENP-F,MAD1 and MAD2 to kinetochores and is essential for mitosis

The kinetochore, a macromolecular complex located at the centromere of chromosomes, provides essential functions for accurate chromosome segregation. Kinetochores contain checkpoint proteins that monitor attachments between the kinetochore and microtubules to ensure that cells do not exit mitosis in the presence of unaligned chromosomes. Here we report that human CENP-I, a constitutive protein of the kinetochore that shares limited similarity with Mis6 of Schizosaccharomyces pombe, is required for the localization of CENP-F and the checkpoint proteins MAD1 and MAD2 to kinetochores. Depletion of CENP-I from kinetochores causes the cell cycle to delay in G2. Although monopolar chromosomes in CENP-I-depleted cells fail to establish bipolar connections, the cells are unable to arrest in mitosis. These cells are transiently delayed in mitosis in a MAD2-dependent manner, even though their kinetochores are depleted of MAD2. The delay is extended considerably when the number of unattached kinetochores is increased. This suggests that no single unattached kinetochore in CENP-I-depleted cells can arrest mitosis. The collective output from many unattached kinetochores is required to reach a threshold signal of 'wait for anaphase' to sustain a prolonged mitotic arrest.     

Spindle checkpoint proteins and chromosome-microtubule attachment in budding yeast

Accurate chromosome segregation depends on precise regulation of mitosis by the spindle checkpoint. This checkpoint monitors the status of kinetochore-microtubule attachment and delays the metaphase to anaphase transition until all kinetochores have formed stable bipolar connections to the mitotic spindle. Components of the spindle checkpoint include the mitotic arrest defective (MAD) genes MAD1-3, and the budding uninhibited by benzimidazole (BUB) genes BUB1 and BUB3. In animal cells, all known spindle checkpoint proteins are recruited to kinetochores during normal mitoses. In contrast, we show that whereas Saccharomyces cerevisiae Bub1p and Bub3p are bound to kinetochores early in mitosis as part of the normal cell cycle, Mad1p and Mad2p are kinetochore bound only in the presence of spindle damage or kinetochore lesions that interfere with chromosome-microtubule attachment. Moreover, although Mad1p and Mad2p perform essential mitotic functions during every division cycle in mammalian cells, they are required in budding yeast only when mitosis goes awry. We propose that differences in the behavior of spindle checkpoint proteins in animal cells and budding yeast result primarily from evolutionary divergence in spindle assembly pathways.     

Cell cycle checkpoints: Arresting progress in mitosis

Cell cycle arrest in M phase can be induced by the failure of a single chromosome to attach properly to the mitotic spindle. The same cell cycle checkpoint mediates M phase arrest when cells are treated with drugs that either disrupt or hyperstabilize spindle microtubules. Study of yeast mutants that fail to arrest in the presence of microtubule disruptors identified a set of genes important in this checkpoint pathway. Two recent papers report the cloning of human and Xenopus homologues of one of these yeast genes, called MAD2 (for mitotic arrest deficient-2)^(1,2). Introduction of antibodies to the MAD2 protein into living mammalian cells or Xenopus egg extracts abrogates the M phase arrest induced by microtubule inhibitors. This and other recent developments suggest a model for the M phase checkpoint in which unattached kinetochores inhibit the ubiquitination of proteins whose proteolysis is necessary for chromatid separation and exit from mitosis.     

BUBR1 phosphorylation is regulated during mitotic checkpoint activation.

Eukaryotic cells have evolved a mechanism that delays the progression of mitosis until condensed chromosomes are properly positioned on the mitotic spindle. To understand the molecular basis of such monitoring mechanism in human cells, we have been studying genes that regulate the mitotic checkpoint. Our early studies have led to the cloning of a full-length cDNA encoding MAD3-like protein (also termed BUBR1/MAD3/SSK1). Dot blot analyses show that BUBR1 mRNA is expressed in tissues with a high mitotic index but not in differentiated tissues. Western blot analyses show that in asynchronous cells, BUBR1 protein primarily exhibits a molecular mass of 120 kDa, and its expression is detected in most cell lines examined. In addition, BUBR1 is present during various stages of the cell cycle. As cells enter later S and G2, BUBR1 levels are increased significantly. Nocodazole-arrested mitotic cells obtained by mechanical shake-off contain BUBR1 antigen with a slower mobility on denaturing SDS gels. Phosphatase treatment restores the slowly migrating band to the interphase state, indicating that the slow mobility of the BUBR1 antigen is attributable to phosphorylation. Furthermore, purified recombinant His6-BUBR1 is capable of autophosphorylation. Our studies indicate that BUBR1 phosphorylation status is regulated during spindle disruption. Considering its strong homology to BUB1 protein kinase, BUBR1 may also play an important role in mitotic checkpoint control by phosphorylation of a critical cellular component(s) of the mitotic checkpoint pathway.     

Requirement of a functional spindle checkpoint for arsenite-induced apoptosis

To understand the potential influence of spindle checkpoint function in response to arsenic trioxide (ATO)-induced apoptosis observed in cancer cell lines, we examined the correlation between activation of the spindle checkpoint and susceptibility to ATO-induced apoptosis in 10 cancer cell lines lacking functional p53. The ability to functionally activate the spindle checkpoint in each cancer cell line was assessed by the induction of mitotic arrest after Taxol treatment. Bromodeoxyuridine (BrdU) pulse-chase analysis of Taxol-treated cell lines with low mitotic arrest showed that they were not arrested at mitosis but divided abnormally, confirming that spindle checkpoint activation was impaired in these cell lines. Our results demonstrate that apoptosis was significantly induced by ATO in cancer cell lines with functional activation of the spindle checkpoint and substantial induction of mitotic arrest. Cell lines with negligible mitotic arrest exhibited little ATO-induced apoptosis. However, no such correlation was observed following treatment of cells with camptothecin, a topoisomerase I inhibitor. Furthermore, attenuation of the spindle checkpoint function by small interfering RNA-mediated silencing of BubR1 and Mad2 in cancer cells that were susceptible to ATO-induced mitotic arrest and apoptosis greatly reduced the induction of mitotic arrest and apoptosis by ATO and increased the formation of micronuclei or multinuclei in survived cells. The marked correlation between ATO-induced mitotic arrest and apoptosis indicates that the induction of apoptosis by ATO was highly dependent on the functional activation of the spindle checkpoint in cancer cells lacking normal p53 function.     

Mad1p, a phosphoprotein component of the spindle assembly checkpoint in budding yeast

The spindle assembly checkpoint prevents cells from initiating anaphase until the spindle has been fully assembled. We previously isolated mitotic arrest deficient (mad) mutants that inactivate this checkpoint and thus increase the sensitivity of cells to benomyl, a drug that interferes with mitotic spindle assembly by depolymerizing microtubules. We have cloned the MAD1 gene and show that when it is disrupted yeast cells have the same phenotype as the previously isolated mad1 mutants: they fail to delay the metaphase to anaphase transition in response to microtubule depolymerization. MAD1 is predicted to encode a 90-kD coiled-coil protein. Anti-Mad1p antibodies give a novel punctate nuclear staining pattern and cell fractionation reveals that the bulk of Mad1p is soluble. Mad1p becomes hyperphosphorylated when wild-type cells are arrested in mitosis by benomyl treatment, or by placing a cold sensitive tubulin mutant at the restrictive temperature. This modification does not occur in G1- arrested cells treated with benomyl or in cells arrested in mitosis by defects in the mitotic cyclin proteolysis machinery, suggesting that Mad1p hyperphosphorylation is a step in the activation of the spindle assembly checkpoint. Analysis of Mad1p phosphorylation in other spindle assembly checkpoint mutants reveals that this response to microtubule- disrupting agents is defective in some (mad2, bub1, and bub3) but not all (mad3, bub2) mutant strains. We discuss the possible functions of Mad1p at this cell cycle checkpoint.     

Bone morphogenetic protein (BMP) signaling regulates mitotic checkpoint protein levels in human breast cancer cells

Aberrant expression of mitotic checkpoint genes compromises mitotic checkpoint, leads to chromosome instability and tumorigenesis. However, the cell signals that control mitotic checkpoint gene expression have not been reported so far. In the present study we show that, in human breast cancer cells, chemical inhibition of Bone morphogenetic proteins (BMPs), but not Transforming Growth Factor-β (TGF-β), abrogates the mitotic arrest induced by nocodazole. Protein expression analysis reveals that inhibition of BMP signaling dramatically down regulates protein levels of mitotic checkpoint components BUB3, Hec1, TTK and MAD2, but inhibition of TGF-β has relatively minor effect on the expression of these proteins. Activation of BMP signaling specifically up regulates BUB3, and activation of Activin A signaling globally down regulates these proteins level. Furthermore, overexpressing MAD2, TTK, BUB3 or Hec1 significantly rescues the mitotic arrest defect caused by BMP inhibition. Our results demonstrated for the first time that TGF-β family cytokines are cellular signals regulating mitotic checkpoint and perturbations in intrinsic BMP signaling could lead to suppression of mitotic checkpoint signaling by downregulating key checkpoint proteins. The results suggest a possible mechanism by which dysregulation of TGF-β signaling causes mitotic checkpoint defects and drives tumorigenesis. The finding also provides a potential and more specific strategy for cancer prevention by targeting BMP and mitotic checkpoint connection.     

Characterization of regions in hsMAD1 needed for binding hsMAD2. A polymorphic change in an hsMAD1 leucine zipper affects MAD1-MAD2 interaction and spindle checkpoint function

In eukaryotes, the mitotic spindle assembly checkpoint provides a monitor for the fidelity of chromosomal segregation. In this context, the mitotic arrest deficiency protein 2 (MAD2) censors chromosomal mis-segregation by monitoring microtubule attachment/tension, a role that requires its attachment to kinetochores. Studies in yeast have shown that binding of MAD1 to MAD2 is important for the checkpoint function of the latter. The interactions between human MAD1 (hsMAD1) and human MAD2 (hsMAD2) have, however, remained poorly characterized. Here we report that two leucine zipper domains (amino acids 501-522 and 557-571) in hsMAD1 are required for its contact with hsMAD2. Interestingly, in several cancer cell lines, we noted the frequent presence of a coding single nucleotide Arg to His polymorphism at codon 558 located within the second leucine zipper of hsMAD1. We found that hsMAD1H558 is less proficient than hsMAD1R558 in binding hsMAD2 and in enforcing mitotic arrest. We also document a first example of loss-of-heterozygosity for a spindle checkpoint gene (at the hsMAD1 558 locus) in a human breast cancer. Based on our findings, it is possible that hsMAD1H558 could be an at-risk polymorphism that contributes to attenuated spindle checkpoint function in human cells.     

PRP4 is a spindle assembly checkpoint protein required for MPS1, MAD1, and MAD2 localization to the kinetochores

The spindle checkpoint delays anaphase onset until every chromosome kinetochore has been efficiently captured by the mitotic spindle microtubules. In this study, we report that the human pre–messenger RNA processing 4 (PRP4) protein kinase associates with kinetochores during mitosis. PRP4 depletion by RNA interference induces mitotic acceleration. Moreover, we frequently observe lagging chromatids during anaphase leading to aneuploidy. PRP4-depleted cells do not arrest in mitosis after nocodazole treatment, indicating a spindle assembly checkpoint (SAC) failure. Thus, we find that PRP4 is necessary for recruitment or maintenance of the checkpoint proteins MPS1, MAD1, and MAD2 at the kinetochores. Our data clearly identify PRP4 as a previously unrecognized kinetochore component that is necessary to establish a functional SAC.     

The Saccharomyces cerevisiae spindle pole body duplication gene MPS1 is part of a mitotic checkpoint

M-phase checkpoints inhibit cell division when mitotic spindle function is perturbed. Here we show that the Saccharomyces cerevisiae MPS1 gene product, an essential protein kinase required for spindle pole body (SPB) duplication (Winey et al., 1991; Lauze et al., 1995), is also required for M-phase check-point function. In cdc31-2 and mps2-1 mutants, conditional failure of SPB duplication results in cell cycle arrest with high p34CDC28 kinase activity that depends on the presence of the wild-type MAD1 checkpoint gene, consistent with checkpoint arrest of mitosis. In contrast, mps1 mutant cells fail to duplicate their SPBs and do not arrest division at 37 degrees C, exhibiting a normal cycle of p34CDC28 kinase activity despite the presence of a monopolar spindle. Double mutant cdc31-2, mps1-1 cells also fail to arrest mitosis at 37 degrees C, despite having SPB structures similar to cdc31-2 single mutants as determined by EM analysis. Arrest of mitosis upon microtubule depolymerization by nocodazole is also conditionally absent in mps1 strains. This is observed in mps1 cells synchronized in S phase with hydroxyurea before exposure to nocodazole, indicating that failure of checkpoint function in mps1 cells is independent of SPB duplication failure. In contrast, hydroxyurea arrest and a number of other cdc mutant arrest phenotypes are unaffected by mps1 alleles. We propose that the essential MPS1 protein kinase functions both in SPB duplication and in a mitotic checkpoint monitoring spindle integrity.     

八氯腺苷诱导SH-SY5Y和SK-N-SH细胞G2/M期阻滞的分子机制不同

为探索八氯腺苷的抗肿瘤作用机制,以神经母细胞瘤SH-SY5Y和SK-N-SH细胞为对象,采用四唑盐比色实验（MTT法）证明,八氯腺苷具有明显的抑制肿瘤细胞增殖的作用,这种抑制作用呈剂量-时间依赖性.流式细胞分析显示,10 μmol/L八氯腺苷作用48 h后可导致靶细胞生长停滞于G ^₂/M期;SH-SY5Y细胞发生明显细胞凋亡,但SK-N-SH细胞却未见凋亡.Hoechst 33342染色显示,SK-N-SH细胞发生了核分裂异常.蛋白质免疫印迹分析证明,10 μmol/L 八氯腺苷处理SH SY5Y 48～72 h后,G^₂检验点调节蛋白ATM、Chk1、Cdc25C和Cdc2磷酸化形式明显上调,同时伴有caspase-3的激活,提示SH-SY5Y细胞发生了G^₂检验点通路和细胞凋亡途径的激活.与SH-SY5Y细胞不同,在SK-N-SH细胞中,八氯腺苷处理24～96 h时,磷酸化ATM、磷酸化Chk1/Chk2、磷酸化Cdc25C以及磷酸化Cdc2的水平呈现逐渐降低的趋势.结果提示,SK-N-SH细胞在八氯腺苷处理后发生了G^₂检验点失败.蛋白质免疫印迹分析还显示,八氯腺苷可诱导p53在SH-SY5Y细胞的表达,但却不能影响SK—N-SH细胞的p53组成性表达水平.p21在SK-N-SH的组成性表达随八氯腺苷处理时间延长而逐渐减少,但在处理前后的SH-SY5Y细胞均未检测到p21蛋白的表达.上述实验结果提示,八氯腺苷抑制两种细胞增殖的机制不同：在SH-SY5Y细胞,八氯腺苷可激活ATM-Chk-Cdc25C-Cdc2/cyclin途径和凋亡通路,使细胞发生G_²/M期阻滞和细胞凋亡;在SK-N-SH细胞,八氯腺苷诱导G^₂检验点失败,导致细胞阻滞在有丝分裂期,并发生有丝分裂异常.2种不同的细胞命运可能还与p53和p21表达不同有关.