Portal de Periódicos da CAPES

Inactivation of p21Sensitizes Cells to Apoptosis via an Increase of Both p14ARF and p53 Levels and an Alteration of the Bax/Bcl-2 Ratio

2002; Elsevier BV; Volume: 277; Issue: 40 Linguagem: Inglês

10.1074/jbc.m204497200

1083-351X

Delphine Javelaud, Françoise Besançon,

Cancer, Lipids, and Metabolism

p21WAF1 appears to be a major determinant of the cell fate in response to anticancer therapy. It was shown previously that HCT116 human colon cancer cells growing in vitro enter a stable arrest upon DNA damage, whereas cells with a defective p21WAF1 response undergo apoptosis. Here we report that the enhanced sensitivity of HCT116/p21−/−cells to chemotherapeutic drug-induced apoptosis correlates with an increased expression of p53 and a modification of their Bax/Bcl-2 ratio in favor of the pro-apoptotic protein Bax. Treatment of HCT116/p21−/− cells with daunomycin resulted in a reduction of the mitochondrial membrane potential and in activation of caspase-9, whereas no such changes were observed in HCT116/p21+/+ cells, providing evidence that p21WAF1 exerts an antagonistic effect on the mitochondrial pathway of apoptosis. Moreover, the role of p53 in activation of this pathway was demonstrated by the fact that inhibition of p53 activity by pifithrin-α reduced the sensitivity of HCT116/p21−/−cells to daunomycin-induced apoptosis and restored a Bax/Bcl-2 ratio similar to that observed in HCT116p21+/+ cells. Enhancement of p53 expression after disruption of p21WAF1resulted from a stabilization of p53, which correlated with an increased expression of the tumor suppressor p14ARF, an inhibitor of the ubiquitin ligase activity of Mdm2. In accordance with the role of p14ARF in p53 stabilization, overexpression of p14ARF in HCT116/p21+/+ cells resulted in a strong increase in p53 activity. Our results identify a novel mechanism for the anti-apoptotic effect of p21WAF1 consisting in maintenance of mitochondrial homeostasis that occurs in consequence of a negative control of p14ARF expression. p21WAF1 appears to be a major determinant of the cell fate in response to anticancer therapy. It was shown previously that HCT116 human colon cancer cells growing in vitro enter a stable arrest upon DNA damage, whereas cells with a defective p21WAF1 response undergo apoptosis. Here we report that the enhanced sensitivity of HCT116/p21−/−cells to chemotherapeutic drug-induced apoptosis correlates with an increased expression of p53 and a modification of their Bax/Bcl-2 ratio in favor of the pro-apoptotic protein Bax. Treatment of HCT116/p21−/− cells with daunomycin resulted in a reduction of the mitochondrial membrane potential and in activation of caspase-9, whereas no such changes were observed in HCT116/p21+/+ cells, providing evidence that p21WAF1 exerts an antagonistic effect on the mitochondrial pathway of apoptosis. Moreover, the role of p53 in activation of this pathway was demonstrated by the fact that inhibition of p53 activity by pifithrin-α reduced the sensitivity of HCT116/p21−/−cells to daunomycin-induced apoptosis and restored a Bax/Bcl-2 ratio similar to that observed in HCT116p21+/+ cells. Enhancement of p53 expression after disruption of p21WAF1resulted from a stabilization of p53, which correlated with an increased expression of the tumor suppressor p14ARF, an inhibitor of the ubiquitin ligase activity of Mdm2. In accordance with the role of p14ARF in p53 stabilization, overexpression of p14ARF in HCT116/p21+/+ cells resulted in a strong increase in p53 activity. Our results identify a novel mechanism for the anti-apoptotic effect of p21WAF1 consisting in maintenance of mitochondrial homeostasis that occurs in consequence of a negative control of p14ARF expression. tumor necrosis factor inhibitors of apoptosis c-Jun N-terminal kinase pifithrin-α tumor necrosis factor-related apoptosis-inducing ligand p21 antisense oligonucleotide Stress stimuli such as cytotoxic drugs stabilize the tumor suppressor p53 (1Oren M. J. Biol. Chem. 1999; 274: 36031-36034Abstract Full Text Full Text PDF PubMed Scopus (489) Google Scholar, 2Vogelstein B. Lane D. Levine A.J. Nature. 2000; 408: 307-310Crossref PubMed Scopus (5821) Google Scholar, 3Woods D.B. Vousden K.H. Exp. Cell Res. 2001; 264: 56-66Crossref PubMed Scopus (291) Google Scholar). Depending on the cellular context, activation of p53 promotes either cell cycle arrest to enable DNA repair or apoptosis to eliminate defective cells (4Sionov R.V. Haupt Y. Oncogene. 1999; 18: 6145-6147Crossref PubMed Scopus (501) Google Scholar). Recent studies have addressed the question of how p53 triggers either one or the other of these two pathways, and the protein p21WAF1 has been implicated as a major determinant of cell fate. p21WAF1 is responsible for the p53-dependent checkpoint that results in G1 arrest after DNA damage (5Harper J.W. Adami G.R. Wei N. Keyomarsi K. Elledge S.J. Cell. 1993; 75: 805-816Abstract Full Text PDF PubMed Scopus (5239) Google Scholar,6Waldman T. Kinzler K.W. Vogelstein B. Cancer Res. 1995; 55: 5187-5190PubMed Google Scholar). Cells deficient in p21WAF1 display a markedly different response to DNA damaging agents compared with cells with an intact p21WAF1-dependent checkpoint. HCT116 human colon cancer cells growing in vitro enter a stable arrest following DNA damage by γ-irradiation or chemotherapeutic drugs, whereas cells with a defective p21WAF1response undergo apoptosis (6Waldman T. Kinzler K.W. Vogelstein B. Cancer Res. 1995; 55: 5187-5190PubMed Google Scholar, 7Waldman T. Lengauer C. Kinzler K.W. Vogelstein B. Nature. 1996; 381: 713-716Crossref PubMed Scopus (719) Google Scholar, 8Bunz F. Hwang P.M. Torrance C. Waldman T. Zhang Y. Dillehay L. Williams J. Lengauer C. Kinzler K.W. Vogelstein B. J. Clin. Invest. 1999; 104: 263-269Crossref PubMed Scopus (928) Google Scholar, 9Mahyar-Roemer M. Roemer K. Oncogene. 2001; 20: 3387-3398Crossref PubMed Scopus (157) Google Scholar). Moreover, whereas irradiation of xenograft tumors established from cells with the intactp21WAF1 gene resulted in no cure, a significant fraction of tumors generated from p21WAF1-deficient cells were completely eliminated (10Waldman T. Zhang Y. Dillehay L., Yu, J. Kinzler K. Vogelstein B. Williams J. Nat. Med. 1997; 3: 1034-1036Crossref PubMed Scopus (361) Google Scholar). Similarly, p21WAF1antisense therapy of wild-type HCT116/p21+/+ cells led to loss of G1 arrest, and increased apoptosis after ionizing radiation exposure both in vitro and in tumor xenografts in nude mice (11Tian H. Wittmack E.K. Jorgensen T.J. Cancer Res. 2000; 60: 679-684PubMed Google Scholar). An anti-apoptotic effect of p21WAF1 on p53-dependent or p53-independent cytotoxic stimulation has also been observed in several other cell types such as melanoma cells (12Gorospe M. Cirielli C. Wang X. Seth P. Capogrossi M.C. Holbrook N.J. Oncogene. 1997; 14: 929-935Crossref PubMed Scopus (292) Google Scholar) or lung cancer cells (13Wang Y. Blandino G. Givol D. Oncogene. 1999; 18: 2643-2649Crossref PubMed Scopus (124) Google Scholar). Also, we demonstrated that induction of p21WAF1 expression participates in the anti-apoptotic effect of the transcription factor NF-κB during treatment of Ewing sarcoma cells with TNF-α1(14Javelaud D. Wietzerbin J. Delattre O. Besançon F. Oncogene. 2000; 19: 61-68Crossref PubMed Scopus (59) Google Scholar). Expression of p21WAF1 was also shown to play a key role during differentiation of several cellular systems by conferring cell survival. Indeed, inhibition of p21WAF1 expression by treatment of differentiating myocytes or neuroblastoma cells with antisense oligonucleotides resulted in extensive cell death (15Poluha W. Poluha D.K. Chang B. Crosbie N.E. Schonhoff C.M. Kilpatrick D.L. Ross A.H. Mol. Cell. Biol. 1996; 16: 1335-1341Crossref PubMed Scopus (225) Google Scholar, 16Wang J. Walsh K. Science. 1996; 273: 359-361Crossref PubMed Scopus (463) Google Scholar, 17Lawlor M.A. Rotwein P. Mol. Cell. Biol. 2000; 20: 8983-8995Crossref PubMed Scopus (166) Google Scholar). Similarly, p21WAF1 was required for cell survival during vitamin D3- or phorbol 12-myristate 13-acetate-induced differentiation of monocytes to macrophages (18Asada M. Yamada T. Ichijo H. Delia D. Miyazono K. Fukumuro K. Mizutani S. EMBO J. 1999; 18: 1223-1234Crossref PubMed Scopus (537) Google Scholar, 19Pennington K.N. Taylor J.A. Bren G.D. Paya C.V. Mol. Cell. Biol. 2001; 21: 1930-1941Crossref PubMed Scopus (70) Google Scholar). The mechanisms by which p21WAF1 inhibits cell cycle progression has been found to involve its binding to G1cyclin/cyclin-dependent kinase complexes and to the proliferating cell nuclear antigen proliferating cell nuclear antigen (20Dotto G.P. Biochim. Biophys. Acta. 2000; 1471: M43-M56PubMed Google Scholar). In contrast, the molecular events underlying the p21WAF1 anti-apoptotic effect have not been fully elucidated. Several steps that are implicated in the apoptotic process have been shown to be inhibited by p21WAF1. These include inhibition of activation of apoptosis signal-regulating kinase 1, c-Jun N-terminal kinase (JNK), and caspase-3 (18Asada M. Yamada T. Ichijo H. Delia D. Miyazono K. Fukumuro K. Mizutani S. EMBO J. 1999; 18: 1223-1234Crossref PubMed Scopus (537) Google Scholar, 21Shim J. Lee H. Park J. Kim H. Choi E.J. Nature. 1996; 81: 804-806Crossref Scopus (240) Google Scholar, 22Suzuki A. Tsutomi Y. Akahane K. Araki T. Miura M. Oncogene. 1998; 17: 931-939Crossref PubMed Scopus (324) Google Scholar). Alternatively, in human colorectal carcinoma cells, cyclin-dependent kinase binding or cyclin-dependent kinase inhibitory activity of p21WAF1 was required to prevent apoptosis (23Lu Y. Yamagishi N. Yagi T. Takebe H. Oncogene. 1998; 16: 705-712Crossref PubMed Scopus (82) Google Scholar). In this study we investigated the mechanisms that are implicated in the enhanced sensitivity to therapeutic agents ofp21WAF1 gene-deleted colocarcinoma cells. We report that abolition of p21WAF1 protein expression induced a post-transcriptional increase of p53 levels that resulted from an enhanced expression of the tumor suppressor p14ARF. The consequence of the increased activity of p53 was a modification of the balance between the anti-apoptotic protein Bcl-2 and the pro-apoptotic protein Bax in favor of Bax. Wild-type HCT116 human colon carcinoma cells (HCT116/p21+/+), which express normal p53 and p21WAF1, and a derivative in which bothp21WAF1 alleles have been deleted through homologous recombination (HCT116/p21−/−) were kindly provided by Dr. B. Vogelstein (Johns Hopkins University, Baltimore, MD) and have been described previously (6Waldman T. Kinzler K.W. Vogelstein B. Cancer Res. 1995; 55: 5187-5190PubMed Google Scholar). The pp53-TA-luc vector and the pRSV-β galactosidase (pRSV-βgal) vector were from Clontech and Promega, respectively. The pcDNA vector encoding p14ARF was a gift from Dr. S. Gazzeri (INSERM EMI 9924, Institut Albert Bonniot, La Tronche, France). The following antibodies were used for Western blot analysis: monoclonal antibodies against p53, Mdm2, c-Abl, cIAP1, and Bcl-xL (Pharmigen), caspase-9 (Upstate Biotechnology), poly(ADP-ribose) polymerase (Alexis Biochemicals), XIAP (Medical and Biological Laboratories Co.), p21WAF1 (Transduction Laboratories), p14ARF (Oncogene), Bcl-2 (Dako), actin (ICN), rabbit serum against phospho-JNK and Bad (New England Biolabs), and JNK1, cIAP2, and Bax (Santa Cruz). Mouse monoclonal antibody anti-CD95, clone B.G27, was purchased from Biomol. Recombinant human tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) or TNF-α were from R&D Systems and Bender Wien, respectively. Pifithrin-α (PFTα) was purchased from Alexis Biochemicals. Daunomycin, adriamycin, and etoposide were from Sigma. Fragmented nuclei were detected by fluorescence microscopy after staining of cells with the DNA binding dye Hoechst 33258 (Sigma) as described previously (24Besançon F. Atfi A. Gespach C. Cayre Y.E. Bourgeade M.F. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 8081-8086Crossref PubMed Scopus (68) Google Scholar). The scoring was done in a blinded fashion and an average of 400 cells were examined for each point. Variation of the mitochondrial membrane potential during apoptosis was studied using 3,3′-dihexyloxacarbocyanine iodide, DiOC6 (3Woods D.B. Vousden K.H. Exp. Cell Res. 2001; 264: 56-66Crossref PubMed Scopus (291) Google Scholar) (Calbiochem), as described previously (14Javelaud D. Wietzerbin J. Delattre O. Besançon F. Oncogene. 2000; 19: 61-68Crossref PubMed Scopus (59) Google Scholar). Total cellular extracts were prepared in 10 mm Tris, pH 7.4, 1% SDS, 1 mm sodium vanadate, treated with Benzon nuclease (Merck) for 5 min at room temperature, and boiled for 3 min. Thirty micrograms of proteins were loaded on a SDS-polyacrylamide gel. After electrophoresis, proteins were electrotransferred onto a nitrocellulose membrane. Membranes were blocked with 5% nonfat milk, 0.1% Tween 20 and then probed with primary antibodies according to the manufacturer's instructions. The primary antibodies were labeled with horseradish peroxidase-conjugated secondary antibodies, which were then visualized using an enhanced chemiluminescence kit obtained from Amersham Biosciences. The experiments were repeated three times and representative autoradiograms are shown. Bands were quantified with NIH Image software after densitometric scanning of the autoradiogram. Cells (1 × 107) were washed twice with serum-free medium and incubated for 10 min at 4 °C with 4 μg of pp53-TA-luc or 4 μg of control pTA-luc plasmids together with 1 μg of pRSV-β-gal in Opti-MEM medium (Invitrogen). Cells were then electroporated at 960 μF, 230 V, and plated in the appropriate medium prior to harvest. Reporter gene activity was determined with the luciferase assay system (Promega). Transfection efficiencies were normalized on the basis of β-galactosidase activity that was measured using the β-galactosidase enzyme assay (Promega). Antisense phosphorothioate oligonucleotide (10 μm) against p21WAF1(p21AS: 5′-TGTCATGCTGGTCTGCCGCC-3′) that is complementary to the 3′ end of the coding region (15Poluha W. Poluha D.K. Chang B. Crosbie N.E. Schonhoff C.M. Kilpatrick D.L. Ross A.H. Mol. Cell. Biol. 1996; 16: 1335-1341Crossref PubMed Scopus (225) Google Scholar) and the control oligonucleotide (Co: 5′-TGGATCCGACATGTCAGA-3′) were transfected into cells using Lipofectin (Invitrogen). Oligonucleotides were added again to the cells 24 h later at a final concentration of 5 μm. Cells were then incubated for 48 or 72 h before Western blot analysis of protein expression. Total cellular RNA was isolated using RNeasy kit (Qiagen). Single strand cDNAs were synthesized using Moloney murine leukemia virus reverse transcriptase (Invitrogen) and random hexamers as primers. The cDNAs were amplified using the Taq DNA polymerase (Invitrogen) in a PerkinElmer Life Sciences 9600 thermocycler for 35 cycles. Each cycle consisted of 1 min denaturation at 94 °C, 1 min annealing at 60 °C, and 1.5 min extension at 72 °C. β2-Microglobulin was used as an internal control. PCR products were separated on 2% agarose gel and visualized by staining with ethidium bromide. In our experimental conditions, amplification of cDNAs was linear up to 38 cycles. Primers used for PCR were as follow: p53 sense (5′-TACCACCATCCACTACAACT-3′), p53 antisense (5′-GAACATCTCGAAGCGCTCA-3′), β2-microglobulin sense (5′-CATCCAGCGTACTCCAAAGA-3′), and β2-microglobulin antisense (5′-GACAAGTCTGAATGCTCCAC-3′). p21WAF1has been reported to exert a protective effect against radiations and various chemical apoptosis inducers (7Waldman T. Lengauer C. Kinzler K.W. Vogelstein B. Nature. 1996; 381: 713-716Crossref PubMed Scopus (719) Google Scholar, 8Bunz F. Hwang P.M. Torrance C. Waldman T. Zhang Y. Dillehay L. Williams J. Lengauer C. Kinzler K.W. Vogelstein B. J. Clin. Invest. 1999; 104: 263-269Crossref PubMed Scopus (928) Google Scholar, 9Mahyar-Roemer M. Roemer K. Oncogene. 2001; 20: 3387-3398Crossref PubMed Scopus (157) Google Scholar, 10Waldman T. Zhang Y. Dillehay L., Yu, J. Kinzler K. Vogelstein B. Williams J. Nat. Med. 1997; 3: 1034-1036Crossref PubMed Scopus (361) Google Scholar, 11Tian H. Wittmack E.K. Jorgensen T.J. Cancer Res. 2000; 60: 679-684PubMed Google Scholar, 12Gorospe M. Cirielli C. Wang X. Seth P. Capogrossi M.C. Holbrook N.J. Oncogene. 1997; 14: 929-935Crossref PubMed Scopus (292) Google Scholar, 13Wang Y. Blandino G. Givol D. Oncogene. 1999; 18: 2643-2649Crossref PubMed Scopus (124) Google Scholar, 14Javelaud D. Wietzerbin J. Delattre O. Besançon F. Oncogene. 2000; 19: 61-68Crossref PubMed Scopus (59) Google Scholar). Here we investigated whether p21WAF1 similarly influences the cellular response to chemotherapeutic drugs and to death receptor ligands. For this purpose, we used wild-type HCT116/p21+/+ human colon carcinoma cells and HCT116/p21−/− cells in which bothp21WAF1 alleles have been deleted by homologous recombination (6Waldman T. Kinzler K.W. Vogelstein B. Cancer Res. 1995; 55: 5187-5190PubMed Google Scholar). Cells from both cell lines were exposed to the chemotherapeutic drugs daunomycin, adriamycin, and etoposide, and to the death receptor ligands TNFα, anti-CD95 (Fas), and TRAIL. The percentage of apoptotic cells was determined by staining with the DNA binding dye Hoechst 33258, which allows visualization of condensed chromatin and fragmented nuclei. As shown in Fig.1, we observed a drastic increase in the chemosensitivity to daunomycin in cells lacking thep21WAF1 gene. Similar results were obtained for adriamycin and etoposide, in accordance with previously reported observations (8Bunz F. Hwang P.M. Torrance C. Waldman T. Zhang Y. Dillehay L. Williams J. Lengauer C. Kinzler K.W. Vogelstein B. J. Clin. Invest. 1999; 104: 263-269Crossref PubMed Scopus (928) Google Scholar, 9Mahyar-Roemer M. Roemer K. Oncogene. 2001; 20: 3387-3398Crossref PubMed Scopus (157) Google Scholar). Whereas HCT116/p21+/+ and HCT116/p21−/− cells were insensitive to the killing effect of TNFα and anti-CD95, both cell lines undergo cell death to almost similar extents following treatment with TRAIL. This demonstrates that in HCT116 cells, p21WAF1 does not extensively interfere with the death pathway triggered by TRAIL, in contrast to its strong inhibitory effect on chemotherapeutic drug-induced cell death. It has been shown that chemotherapeutic agents can induce apoptosis through multiple pathways, mainly by targeting the mitochondrial pathway (25Grad J.M. Cepero E. Boise L.H. Drug Resist. Update. 2001; 4: 85-91Crossref PubMed Scopus (59) Google Scholar). This results in cytochrome c release and activation of the apoptosome complex that then activates downstream caspases. Binding of TRAIL to its cognate receptors DR4 and DR5 can activate a second pathway leading to direct caspase activation independently of mitochondria (26Suliman A. Lam A. Datta R. Srivastava R.K. Oncogene. 2001; 20: 2122-2133Crossref PubMed Scopus (342) Google Scholar). This explains the fact that TRAIL-induced apoptosis has been reported to be independent of the expression of proteins of the Bcl-2 family in various cell types (27Walczak H. Bouchon A. Stahl H. Krammer P.H. Cancer Res. 2000; 60: 3051-3057PubMed Google Scholar). As we observed that p21WAF1exerted an anti-apoptotic effect against chemotherapeutic drugs but not against TRAIL-induced cell death, we analyzed the consequences ofp21WAF1 gene deletion on the expression of various members of the Bcl-2 family. The constitutive level of Bcl-2 was decreased in HCT116/p21−/− as compared with parental cells (Fig. 2 A). Inversely,p21WAF1 deletion resulted in an increase in Bax protein expression, whereas levels of Bcl-xL and Bad were unchanged. As shown in Fig. 2 B (upper), our results were obtained in the linear range of the assay. Quantitative analysis of the intensity of the bands revealed that the apparent Bax/Bcl-2 ratio was 4-fold higher in the absence of p21WAF1(Fig. 2, A and B, lower), thus demonstrating a change in relative amounts of Bax and Bcl-2 in favor of Bax. Treatment of both cell lines with daunomycin did not significantly change this ratio. As shown in Fig. 2 A, the level of expression of other regulators of apoptosis, the IAP proteins (inhibitor of apoptosis protein) cIAP1, cIAP2, and XIAP known to control caspase activity, were similar in HCT116/p21+/+ and HCT116/p21−/− cells. During the mitochondrial pathway of apoptosis, mitochondria suffer specific damage such as loss of the mitochondrial membrane potential, and one of the critical roles of the anti-apoptotic members of the Bcl-2 family is to protect mitochondrial homeostasis (28Gross A. McDonnell J.M. Korsmeyer S.J. Genes Dev. 1999; 13: 1899-1911Crossref PubMed Scopus (3253) Google Scholar, 29Wang X. Genes Dev. 2001; 15: 2922-2933Crossref PubMed Scopus (94) Google Scholar, 30Adams J.M. Cory S. Trends Biochem. Sci. 2001; 1: 61-66Abstract Full Text Full Text PDF Scopus (811) Google Scholar). This prompted us to investigate whether the increase of the Bax/Bcl-2 ratio was accompanied by a reduction of the mitochondrial membrane potential in HCT116/p21−/− treated with daunomycin as compared with HCT116/p21+/+ cells. Mitochondrial membrane potential (ΔΨm) was measured using DiOC6(3), a cyanine dye that accumulates in the mitochondrial matrix under the influence of the ΔΨm. Treatment of HCT116/p21−/− cells with daunomycin resulted in a significant decrease in DiOC6(3) uptake as compared with untreated cells (Fig.2 C). Such a decrease was not observed in HCT116/p21+/+ cells (Fig. 2 C). This last result further establishes the correlation between the Bax/Bcl-2 ratio and the ability of daunomycin to disrupt the mitochondrial membrane integrity. As expected, treatment of HCT116/p21−/− cells with daunomycin resulted in an activation of caspase-9, the apical caspase in the mitochondrial pathway (Fig. 2 D). Activation of caspase-3 correlated with activation of caspase-9 as revealed by the cleavage of poly(ADP-ribose) polymerase, a substrate of caspase-3 (Fig.2 D). Because Bax and Bcl-2 expression have been shown to be, respectively, up- and down-regulated by the tumor suppressor p53 (31Miyashita T. Krajewski S. Krajewska M. Wang H.G. Lin H.K. Liebermann D.A. Hoffman B. Reed J.C. Oncogene. 1994; 9: 1799-1805PubMed Google Scholar), we compared p53 levels in HCT116/p21+/+ and HCT116/p21−/− cells. Deletion of the p21WAF1 gene resulted in a higher expression of p53 (Fig. 3 A). To test whether this enhancement in p53 expression was accompanied by an increase in its transcriptional activity, cells were transiently transfected with a plasmid containing the luciferase reporter gene under the control of p53-binding elements. As shown in Fig.3 B, the luciferase activity measured in HCT116/p21−/− cells was 5-fold higher than that measured in HCT116/p21+/+ cells. To corroborate the inverse correlation between expression of p21WAF1 and p53 proteins and to eliminate the possibility that this reflects clonal selection of HCT116/p21−/−cells expressing high levels of p53, we used specific antisense oligonucleotides (p21AS) to inhibit p21WAF1 expression in wild-type HCT116 cells. Treatment of HCT116/p21+/+ cells with p21AS resulted in an inhibition of p21WAF1 expression, whereas control oligonucleotide had no effect (Fig. 3 C). No change in actin expression was observed, thus confirming the specificity of action of p21AS. Inhibition of p21WAF1expression by p21AS was accompanied by an increase in p53 level (Fig.3 C), and here again control AS were without effect. Thus, these results further demonstrate that inhibition of p21WAF1 expression results in an enhancement of p53 protein levels. PFTα is a small molecule that was isolated for its ability to reversibly block p53-dependent transcriptional activation (32Komarov P.G. Komarova E.A. Kondratov R.V. Christov-Tselkov K. Coon J.S. Chernov M.V. Gudkov A.V. Science. 1999; 285: 1733-1737Crossref PubMed Scopus (1117) Google Scholar). We used this compound to investigate whether the higher p53 activity present in HCT116/p21−/− cells as compared with HCT116/p21+/+ cells participated in higher sensitivity to daunomycin-induced apoptosis. As shown in Fig.4 A, treatment of HCT116/p21−/− cells with 30 μm PFTα resulted in an inhibition of p53 transcriptional activity. This reduction of p53 activity was accompanied by a reduction of the sensitivity of these cells to daunomycin-induced cell death (Fig.4 B). Furthermore, as shown in Fig. 4, C andD, PFTα-induced inhibition of p53 activity resulted in both an increase in Bcl-2 expression and a decrease of Bax expression, thus further establishing the correlation between p53 activity and the Bax/Bcl-2 ratio. p53 expression is tightly regulated through several mechanisms, the most effective one being regulation of protein stability (1Oren M. J. Biol. Chem. 1999; 274: 36031-36034Abstract Full Text Full Text PDF PubMed Scopus (489) Google Scholar, 2Vogelstein B. Lane D. Levine A.J. Nature. 2000; 408: 307-310Crossref PubMed Scopus (5821) Google Scholar, 3Woods D.B. Vousden K.H. Exp. Cell Res. 2001; 264: 56-66Crossref PubMed Scopus (291) Google Scholar). To investigate whether p21WAF1 affects p53 expression at a transcriptional or a post-transcriptional level, constitutive p53 mRNA expression was analyzed in HCT116/p21+/+ and HCT116/p21−/− cells by reverse transcriptase-PCR. As shown in Fig. 5 A, no difference in p53 mRNA expression was observed between the two cell lines, indicating that regulation of p53 expression by p21WAF1 occurs at a post-transcriptional level. Several mechanisms have been shown to be implicated in the regulation of p53 stability. A complex pattern of regulation of p53 stability by the stress kinase JNK has been described. Nonphosphorylated JNK targets p53 ubiquitination and degradation (33Fuchs S.Y. Adler V. Buschmann T. Yin Z., Wu, X. Jones S.N. Ronai Z. Genes Dev. 1998; 12: 2658-2663Crossref PubMed Scopus (281) Google Scholar) whereas stress-activated JNK stabilizes and activates p53 (34Fuchs S.Y. Adler V. Pincus M.R. Ronai Z. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 10541-10546Crossref PubMed Scopus (456) Google Scholar). Because p21WAF1 has been shown to inhibit JNK activation in various cells, we investigated whether absence of p21WAF1 expression in HCT116/p21−/− cells resulted in an enhanced level of phosphorylated JNK that could participate in p53 stabilization. As shown in Fig. 5 B, such a mechanism was not implicated in p53 stabilization that occurs in the absence of p21WAF1 because no difference in JNK expression or activation was observed between HCT116/p21+/+ and HCT116/p21−/− cells. One of the key regulators of p53 expression is the oncogenic protein Mdm2 (35Kubbutat M.H. Jones S.N. Vousden K.H. Nature. 1997; 387: 299-303Crossref PubMed Scopus (2847) Google Scholar). This protein can transport p53 from the nucleus to the cytoplasm where its ubiquitin ligase activity contributes to the proteasome-mediated degradation of p53. Inversely, the nonreceptor tyrosine kinase c-Abl and the tumor suppressor p14ARFup-regulate p53 levels by preventing its Mdm2-driven nuclear export and ubiquitination (36Sionov R.V. Coen S. Goldberg Z. Berger M. Bercovich B. Ben-Neriah Y. Ciechanover A. Haupt Y. Mol. Cell. Biol. 2001; 274: 8371-8374Google Scholar, 37Pomerantz J. Schreiber-Agus N. Liegeois N.J. Silverman A. Alland L. Chin L. Potes J. Chen K. Orlow I. Lee H.W. Cordon-Cardo C. DePinho R.A. Cell. 1998; 92: 713-723Abstract Full Text Full Text PDF PubMed Scopus (1334) Google Scholar, 38Zhang Y. Xiong Y. Yarbrough W.G. Cell. 1998; 92: 725-734Abstract Full Text Full Text PDF PubMed Scopus (1404) Google Scholar). As shown in Fig. 5 B, p53 stabilization in HCT116/p21−/− cells was not correlated with a decrease of Mdm2 or an increase of c-Abl levels. In contrast, disruption of p21WAF1 resulted in a strong increase in the basal level of p14ARF expression (Fig.5 B). Also, inhibition of p21WAF1 expression in wild-type HCT116 cells with specific antisense oligonucleotides resulted in an increase in p14ARF expression without affecting JNK, phosphorylated JNK, Mdm2, or c-Abl levels (Fig.5 C). These results strongly suggest that disruption ofp21WAF1 resulted in an enhancement of p14ARF expression that in turn allows p53 accumulation by antagonizing Mdm2 activity. In accordance with this hypothesis, overexpression of p14ARF in HCT116/p21+/+ cells resulted in a strong increase in p53 activity (Fig. 5 D). Apoptosis is a tightly regulated process essential to maintain tissue homeostasis. Acquisition of a resistance to apoptosis plays a pivotal role in tumorigenesis by disrupting the balance between cell proliferation and cell destruction and by allowing cancer cells to resist radiation and chemotherapy. Identification of mechanisms that can antagonize apoptosis is essential to devise therapeutic strategies aimed at enhancing the efficiency of cancer treatment (39Reed J.C. Nat. Drugs Disc. 2002; 1: 111-121Crossref PubMed Scopus (603) Google Scholar). Recent technological advances have allowed the successful targeting of individual genes in human somatic cells, thus permitting studies of activation or inactivation of a particular gene on the response of cancer cells to therapeutic agents. Here we observed that disruption of p21WAF1 expression in human colon cancer cells, either by gene targeting or by the use of specific antisense oligonucleotides, increased p53 expression. This effect correlated with a sensitization of cells to chemotherapy-induced apoptosis. It also correlated with a change in the Bax/Bcl-2 ratio as a consequence of increased expression of the proapoptotic protein Bax and reduced expression of the anti-apoptotic protein Bcl-2, respectively. The roles of the observed enhancement of p53 levels in the sensitization to daunomycin-induced apoptosis, and in the change in the Bax/Bcl-2 ratio were demonstrated by the fact that inhibition of p53 activity upon treatment of HCT116p21−/− cells with PFTα reversed both effects. The relative ratio of anti- versus pro-apoptotic Bcl-2 family proteins has been shown to dictate the ultimate sensitivity or resistance of cells to various apoptotic stimuli (28Gross A. McDonnell J.M. Korsmeyer S.J. Genes Dev. 1999; 13: 1899-1911Crossref PubMed Scopus (3253) Google Scholar, 29Wang X. Genes Dev. 2001; 15: 2922-2933Crossref PubMed Scopus (94) Google Scholar, 30Adams J.M. Cory S. Trends Biochem. Sci. 2001; 1: 61-66Abstract Full Text Full Text PDF Scopus (811) Google Scholar). Therefore the observed modification of this ratio in favor of Bax very likely participates in the enhanced sensitivity of HCT116/p21−/−cells to chemotherapeutic drugs as compared with their normal counterparts. However, one cannot exclude the possibility that additional mechanisms may be implicated in the pro-apoptotic effect of p53. For example, p53 was also shown to induce expression of the pro-apoptotic proteins FAS/APO1, KILLER/DR5, NOXA, p53AIPI, PIDD, and PUMA (1Oren M. J. Biol. Chem. 1999; 274: 36031-36034Abstract Full Text Full Text PDF PubMed Scopus (489) Google Scholar, 2Vogelstein B. Lane D. Levine A.J. Nature. 2000; 408: 307-310Crossref PubMed Scopus (5821) Google Scholar, 3Woods D.B. Vousden K.H. Exp. Cell Res. 2001; 264: 56-66Crossref PubMed Scopus (291) Google Scholar, 4Sionov R.V. Haupt Y. Oncogene. 1999; 18: 6145-6147Crossref PubMed Scopus (501) Google Scholar) and to reduce expression of the anti-apoptotic protein, survivin (40Hoffman W.H. Biade S. Zilfou J.T. Chen J. Murphy M. J. Biol. Chem. 2002; 277: 3247-3257Abstract Full Text Full Text PDF PubMed Scopus (709) Google Scholar). We observed that disruption of p21WAF1 up-regulated p53 expression at a post-transcriptional level. A key player in the regulation of p53 stability is the Mdm2 protein that promotes the ubiquitination of p53 and its subsequent degradation by the proteasome (1Oren M. J. Biol. Chem. 1999; 274: 36031-36034Abstract Full Text Full Text PDF PubMed Scopus (489) Google Scholar, 2Vogelstein B. Lane D. Levine A.J. Nature. 2000; 408: 307-310Crossref PubMed Scopus (5821) Google Scholar, 3Woods D.B. Vousden K.H. Exp. Cell Res. 2001; 264: 56-66Crossref PubMed Scopus (291) Google Scholar, 35Kubbutat M.H. Jones S.N. Vousden K.H. Nature. 1997; 387: 299-303Crossref PubMed Scopus (2847) Google Scholar). The ability of Mdm2 to promote p53 ubiquitination can be modulated by the binding of other proteins. Among these proteins, p14ARF, the product of an alternative transcript of the INK4a tumor suppressor locus, has been shown to be involved in the response to oncogenic stress by up-regulating p53 expression (37Pomerantz J. Schreiber-Agus N. Liegeois N.J. Silverman A. Alland L. Chin L. Potes J. Chen K. Orlow I. Lee H.W. Cordon-Cardo C. DePinho R.A. Cell. 1998; 92: 713-723Abstract Full Text Full Text PDF PubMed Scopus (1334) Google Scholar, 38Zhang Y. Xiong Y. Yarbrough W.G. Cell. 1998; 92: 725-734Abstract Full Text Full Text PDF PubMed Scopus (1404) Google Scholar). p14ARF is very likely responsible for p53 stabilization that occurs after p21WAF1 disruption because we observed that it is expressed at a higher level in HCT116/p21−/− cells than in HCT116/p21+/+cells. The ability of p14ARF to control p53 expression was demonstrated by the fact that transfection of HCT116/p21+/+cells with an expression vector encoding p14ARF strongly increased p53 activity. Altogether, our results provide evidence that repression of p14ARF that occurs in consequence of the inhibition of p21WAF1 expression is an initial step in the sensitization of HCT116/p21−/− cells as compared with HCT116/p21+/+ cells to daunomycin-induced apoptosis. The mechanism leading to an increased expression of p14ARFsubsequently to p21WAF1 disruption remains to be established. It has been shown that a transcriptionally functional E2F DNA-binding site is located in the p14ARF promoter region (41Bates S. Phillips A.C. Clarke P. Stott F. Peters G. Ludwig R.L. Vousden K.H. Nature. 1998; 395: 124-125Crossref PubMed Scopus (812) Google Scholar). Control of G1/S transition by p21WAF1 occurs through inhibition of cyclin D-cdk4/6 complex activity in late G1 (20Dotto G.P. Biochim. Biophys. Acta. 2000; 1471: M43-M56PubMed Google Scholar). Inhibition of such activity maintains the RB protein in an hypophosphorylated form that sequesters E2F1 and inhibits its transcriptional capacity (42Helin K. Curr. Opin. Genet. Dev. 1998; 8: 28-35Crossref PubMed Scopus (428) Google Scholar). Thus, the mechanism by which inhibition of p21WAF1expression increased the p14ARF level could be the activation of E2F1-dependent transcription. However, although as expected, inactivation of p21WAF1 in HCT116/p21−/− cells resulted in the disappearance of the hypophosphorylated form of RB, it did not correlate with an increased E2F1 transcriptional activity (not shown). One explanation could be a repression of E2F1 by p14ARF, because it was recently reported that p14ARF physically interacts with E2F1 and negatively regulates its activity (43Eymin B. Karayan L. Séité P. Brambilla C. Brambilla E. Larsen C.J. Grazzéri S. Oncogene. 2001; 20: 1033-1041Crossref PubMed Scopus (147) Google Scholar). p14ARF expression was also shown to be regulated by other transcriptional regulators such as c-Myc, β-cathenin, and E1A and by the extent ofp14ARF gene methylation (44Sherr C.J. Nat. Rev. Mol. Cell. Biol. 2001; 2: 731-737Crossref PubMed Scopus (825) Google Scholar, 45Esteller M. Tortola S. Toyota M. Capella G. Peinado M.A. Baylin S.B. Herman J.G. Cancer Res. 2000; 60: 129-133PubMed Google Scholar, 46Damalas A. Kahan S. Shtutman M. Ben-Ze'ev A. Oren M. EMBO J. 2001; 20: 4912-4922Crossref PubMed Scopus (181) Google Scholar). Whether control of p14ARF expression by p21WAF1 implies one of these mechanisms is presently under investigation. Mutations in the p53 gene are the most frequent genetic alterations in cancer. Intensive efforts are presently aimed to offer new approaches to cancer therapy based on restoring mutant p53 function by small molecule drugs (47Foster B.A. Coffey H.A. Morin M.J. Rastinejad F. Science. 1999; 286: 2507-2510Crossref PubMed Scopus (691) Google Scholar, 48Bullock A.N. Fersht A.R. Nat. Rev. Cancer. 2001; 1: 68-76Crossref PubMed Scopus (488) Google Scholar). Our results, which show that in the absence of p21WAF1 p53 participates in the triggering of cell death induced by DNA damaging agents, suggest that during chemotherapy high benefits could result from strategies combining inhibition of p21WAF1 expression and restoration of p53 activity in p53-deficient cancer cells. We are deeply grateful to Dr. B. Vogelstein for providing the HCT116/p21+/+ and HCT116/p21−/− cells. We thank Dr. S. Gazzeri for the p14ARF-encoding vector, Dr. H. Ankel for editing assistance, and C. Silvestri for skilful technical help.