Inhibition of p53 Transcriptional Activity by Bcl-2 Requires Its Membrane-anchoring Domain*

We show here that the anti-apoptosis protein Bcl-2 potently inhibits p53-dependent transcriptional activation of various p53-responsive promoters in reporter gene co-transfection assays in human embryonic kidney 293 and MCF7 cells, without affecting nuclear accumulation of p53 protein. In contrast, Bcl-2(Δtransmembrane (TM)), which lacks a hydrophobic membrane-anchoring domain, had no effect on p53 activity. Similarly, in MCF7 cells stably expressing either Bcl-2 or Bcl-2(ΔTM), nuclear levels of p53 protein were up-regulated upon treatment with the DNA-damaging agents doxorubicin and UV radiation, whereas p53-responsive promoter activity and expression of p21CIP1/WAF1 were strongly reduced in MCF7-Bcl-2 cells but not in MCF7-Bcl-2(ΔTM) or control MCF7 cells. The issue of membrane anchoring was further explored by testing the effects of Bcl-2 chimeric proteins that contained heterologous transmembrane domains from the mitochondrial protein ActA or the endoplasmic reticulum protein cytochrome b5. Both Bcl-2(ActA) and Bcl-2(Cytob5) suppressed p53-mediated transactivation of reporter gene plasmids with efficiencies comparable to wild-type Bcl-2. These results suggest that (a) Bcl-2 not only suppresses p53-mediated apoptosis but also interferes with the transcriptional activation of p53 target genes at least in some cell lines, and (b) membrane anchoring is required for this function of Bcl-2. We speculate that membrane-anchored Bcl-2 may sequester an unknown factor necessary for p53 transcriptional activity.

Cell death is a physiological process that plays a major role in the maintenance of tissue homeostasis. Deregulation of the delicate balance of genes that encode proteins that either induce or inhibit cell death, contributes to the development of human neoplasia.
The bcl-2 gene was first discovered because of its involvement in the t(14;18) chromosomal translocations commonly found in lymphomas (1). Bcl-2 expression can also be altered in a variety of other cancers through various mechanisms, including loss of p53 tumor suppressor, which down-regulates Bcl-2 expression in some tissues (reviewed in Refs. 2 and 3). Bcl-2 functions as an inhibitor of cell death, thus contributing to cell expansion by inhibiting physiological cell turnover. Bcl-2 has also been shown to block cell death induced by a number of cytotoxic drugs as well as ␥-irradiation and is therefore believed to contribute to resistance to anticancer therapies (reviewed in Refs. 2 and 3). In vertebrates, several Bcl-2 homologues have been identified, some of which function as inhibitors (Bcl-X L , Mcl-1, and A1), and others as promoters of cell death (Bcl-X S , Bax, Bak, and BAD). These Bcl-2 family members are characterized by their ability to interact and form homo-and heterodimers (4). The relative abundance of antiand proapoptotic Bcl-2 homologues determines the sensitivity of cells to apoptotic signals (reviewed in Refs. 2 and 5), but the mechanisms by which Bcl-2 inhibits cell death remain to be fully elucidated.
It has been proposed that Bcl-2 may inhibit cell death by interfering with the function of proapoptotic Bcl-2 homologues, by repressing the release of cytochrome c from mitochondria, by the sequestration of caspase activators, such as Apaf-1, by interfering with the production of free radicals by cytotoxic agents, or by regulating intracellular calcium homeostasis (for reviews, see Refs. 6 -8). In addition, Bcl-2 has also been shown to bind and possibly modulate the function of other mammalian proteins, including R-Ras (9), Raf-1 (10), nuclear factor-B (11), calcineurin (12), SMN (13), and the p53-binding protein 53BP2 (14). Interestingly, the Bcl-2 protein has a transmembrane C-terminal domain (TM) 1 that targets it to intracellular membranes, most notably the outer membranes of mitochondria and nuclei and the endoplasmic reticulum membranes (15)(16)(17). Membrane anchorage appears to be required for optimal Bcl-2 function but may not be necessary for all of its antiapoptotic functions (18,19).
The tumor suppressor p53 functions primarily as a transcription factor that controls genomic stability. The levels of p53 protein are normally very low but rise dramatically when DNA damage has occurred (20,21). High levels of p53 protein cause cell cycle arrest, mostly in G 1 , or apoptosis depending of the cellular context and the gravity of the damage (22)(23)(24). The p53 gene is often deleted or mutated in cancer cells (25,26). This leads to increased susceptibility to malignant transformation and blunted responses to DNA damaging cytotoxic agents (reviewed in Ref. 27). The functional status of p53 strongly correlates with its ability to bind specific DNA sequences on target genes and to activate transcription. In addition, p53 can repress transcription of some promoters lacking specific p53 binding sites, and it also possess some transcription-independent functions that are probably relevant to DNA-repair responses (28,29). Among the genes that are activated by p53 are the cyclin-dependent kinase (cdk) inhibitor waf-1/p21 (30), cyclin G (31), GADD45 (32), the p53 inhibitor mdm2 (33), and the apoptosis-inducing genes bax (34) and KILLER/DR5 (35).
Interestingly, an inverse relationship has been found between mutated p53 and Bcl-2 expression in some types of carcinomas (36 -39). This suggests that p53 and Bcl-2 may serve as inducer and repressor, respectively, of the same apoptotic pathway. Studies with bcl-2 transgenic mice and p53 knockout mice also support this concept (40). In this regard, it is well known that Bcl-2 is a potent inhibitor of p53-induced apoptosis (41,42). In addition, in some cell systems, it has been demonstrated that Bcl-2 inhibits p53 translocation to the nucleus upon DNA damage (43) or when co-expressed with particular proteins (44). Bcl-2 may also relieve p53-mediated transcriptional repression (45,46).
In this study, we report that high levels of Bcl-2 protein significantly impair p53-dependent induction of the cdk inhibitor p21 CIP1/WAF1 after DNA damage in MCF7 breast cancer cells, without, however, blocking p53 nuclear import. In agreement with these findings, we show that Bcl-2 strongly inhibits p53-dependent transactivation of several promoters, including bax, p21 CIP1/waf-1 , mdm2, cyclin G, and GADD45 in transient co-transfection assays, again without preventing p53 entry into the nucleus. We also demonstrate that these inhibitory effects of Bcl-2 require its anchorage in intracellular membranes. These findings suggest that in some circumstances, Bcl-2 can block p53 function by inhibiting its transcriptional activity. Thus, Bcl-2 may be capable of interfering with p53 at several levels, probably depending on cell context.
Cell Culture-The human breast cancer cell line MCF7, transformed human embryonic kidney 293 cells, 293T, containing the large T antigen, and 293-EBNA cells, carrying the Epstain-Barr virus EBNA-1, were obtained from the American Type Culture Collection (Manassas, VA). Cells were maintained in a humidified atmosphere with 5% CO 2 in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 1 mM glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin (Life Technology, Inc.). In some experiments, MCF7 cells were treated with 0.4 M doxorubicin (Sigma).
Generation of MCF7-Bcl-2 Stable Transfectants-MCF7 breast cancer cells (3 ϫ 10 5 ) were plated in 60-mm dishes and grown overnight. For transfection, plasmid DNA (2 g) was diluted into 0.1 ml of Opti-MEM medium (Life Technologies) and combined with 5 l of Lipo-fectAMINE (Life Technologies) in 0.1 ml of Opti-MEM. After incubation for 20 min, 0.8 ml of Opti-MEM was added, and the mixtures were overlaid onto cells. After 4 h, 1 ml of Opti-MEM containing 20% fetal calf serum was added to the cultures. The next day, medium was replaced with complete Dulbecco's modified Eagle's medium. Transformants were selected by growth for 14 days in complete medium containing 0.5 mg/ml G418 (Life Technologies). Polyclonal populations were grown and assayed for stable transgene expression by immunoblotting.
Transfections and Enzyme Assays-Cells at 60% confluency were transfected by a standard calcium phosphate precipitate method. The total amount of plasmid DNA used was normalized to 2 g/well and 7 g/plate for transfections in 12-well and 6-cm plates, respectively, by the addition of empty plasmid. For reporter gene assays, 0.2 g of the ␤-galactosidase expression plasmid pCMV-␤-galactosidase was cotransfected with the luciferase reporter gene to normalize for variations in transfection efficiency. Cells were exposed to the precipitate for 5 h at 37°C. For MCF7 cells, a glycerol shock was applied. Cells were exposed to 15% glycerol in HBS buffer (25 mM HEPES, pH 7.05, 0.75 mM Na 2 HPO 4 , 140 mM NaCl) for 4 min. The glycerol was removed by washing three times with PBS and replacement with fresh medium. For 293 cells, the medium was replaced without applying a shock. MCF7 stable cell lines were transfected using the lipofection method described above. Cell extracts were prepared 48 h after transfection. For reporter gene experiments, cells lysates were made as described by the manufacturer (Promega) and assayed for luciferase and ␤-galactosidase activity. All transfection experiments were carried out in triplicate, repeated at least three times, and normalized for ␤-galactosidase activity.
Cell Extracts and Subcellular Fractionation-For gene expression experiments, cells were washed two times in PBS and lysed in RIPA buffer containing protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 0.28 TIU/ml aprotinin, 50 g/ml leupeptin, 1 M benzamidine, 0.7 g/ml pepstatin). For protein localization experiments, nuclear and nonnuclear fractions were prepared according to the method of Schreiber et al. (54). Briefly, cells were collected and washed two times with ice-cold PBS. Cell pellets were resuspended in Buffer A (10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 2.5 mM dithiothreitol, protease inhibitors) and left on ice for 15 min prior addition of Nonidet P-40 to a final concentration of 0.6% (v/v). After centrifugation, supernatants were collected, and the nuclear pellets were washed two times in the same buffer. Pellets were finally resuspended in Buffer B (20 mM HEPES, pH 7.9, 400 mM NaCl, 25% glycerol, 0.1 mM EDTA, 0.1 mM EGTA, 2.5 mM dithiothreitol, protease inhibitors) and vigorously shaken for 10 min, and the postnuclear supernatants were cleared by centrifugation. Fractions were normalized for protein content based on the bicinchoninic acid method (Pierce) prior to the SDS-PAGE/immunoblot assay.

Bcl-2 Inhibits p53-dependent Transactivation of Target
Genes-To investigate the influence of Bcl-2 on the transactivation activity of p53, a co-transfection reporter gene assay was carried out. For these experiments, MCF7 breast cancer and human embryonic kidney 293T cells were transiently co-transfected with plasmids encoding wild-type p53 and Bcl-2, together with a reporter plasmid in which the luciferase gene is driven by a fragment of the human bax gene promoter containing its p53-response element and TATA box. As shown in Fig.  1, A and B, expression of wild-type p53 induced luciferase expression in both cell lines. Co-transfection of Bcl-2 strongly inhibited p53 transcriptional activity in a concentration-dependent manner in both cell lines, whereas co-transfection of a control protein (BAG-1L) did not affect p53 transactivation. To preliminarily explore the mechanisms responsible for the ob-served suppression of p53 transcriptional activity by Bcl-2, nuclear and cytoplasmic extracts were prepared from 293T transfected cells and tested for expression of p53 and Bcl-2 by immunoblotting. As shown in Fig. 1C, Bcl-2 had no obvious effects on total p53 protein levels. In addition, Bcl-2 also did not inhibit p53 accumulation in the nucleus, as has been reported in some but not all types of cells (41,(43)(44)(45)(46)(47)(48)(49)(50)(51)(52)(53)(54)(55). Tubulin staining was used to confirm the accuracy of the nuclear extraction procedure and to confirm loading of equivalent amounts of the samples. Similar results were obtained in MCF7 cells (data not shown). Taken together, these data indicate that Bcl-2 not only inhibits downstream events in p53-mediated apoptosis but, at least in certain cell lines, it also blocks p53 transcriptional activity.
Bcl-2 Suppresses p53-mediated Transactivation of Multiple p53-responsive Promoters-To assess whether Bcl-2 affects the ability of p53 to transactivate promoters other than the bax gene promoter, similar experiments were performed using luciferase-expressing reporter constructs driven by the p53-responsive p21 CIP1/WAF1 , mdm2, or cyclin G promoters. As shown in Fig. 2, as with the bax promoter, Bcl-2 inhibited the p53-dependent transcriptional activity of all promoters, without, however, affecting their basal levels of activity. Under these same conditions, Bcl-2 did not impair the transactivation of the p53independent SV40 early region promoter used as a control (Fig.   2). Thus, Bcl-2 specifically inhibits p53-mediated transactivation of target genes, apparently without significantly affecting transactivation by other transcription factors.
Inhibitory Effect of Bcl-2 Depends on Membrane Anchoring-The C-terminal domain of Bcl-2 contains a stretch of hydrophobic amino acids that anchors the protein into cellular membranes (56,57,58). Previously, Bcl-2 has been found in association with mitochondrial membranes, the nuclear envelope and the endoplasmic reticulum (15)(16)(17). To assess the importance of the subcellular localization of Bcl-2 for the observed inhibitory effect on p53 transactivation, we compared p53 transcriptional activity in 293-EBNA cells transiently transfected with p53 and either wild-type Bcl-2 or a Bcl-2 mutant, which lacks the C-terminal TM domain. Interestingly, unlike Bcl-2, co-expression of Bcl-2(⌬TM) with p53 did not interfere with p53-mediated transactivation. In fact, this cytoplasmically distributed Bcl-2 deletion mutant slightly stimulated p53 activity, possibly by competing with endogenous Bcl-2 protein. Immunoblot analysis confirmed that the failure of Bcl-2(⌬TM) to suppress p53 was not attributable to lower levels of expression of the truncated protein compared with wild-type p53 (Fig. 3B).
The inability of the Bcl-2(⌬TM) mutant to suppress p53 transcriptional activity could be due either to a specific requirement for the TM domain of Bcl-2 or because of a need for membrane anchoring. To distinguish between these two possibilities, we examined the effects on p53 of Bcl-2 chimerical proteins in which the normal C-terminal TM domain was substituted with membrane-anchoring domains from proteins that reside in either the mitochondrial outer membrane (ActA) or the endoplasmic reticulum (cytochrome b5 (Cb5)) (48). Similar to wild-type Bcl-2, plasmids encoding either Bcl-2(ActA) or Bcl-2(Cb5) proteins caused a concentration-dependent suppression of p53-mediated gene transcription (Fig. 3A). The mitochondria-targeting Bcl-2(ActA) protein was somewhat more potent that the endoplasmic reticulum-targeting Bcl-2(Cb5) 293T cells were transiently transfected with p53, Bcl-2, and ␤-galactosidase expression plasmids along with bax-luc, cyclinG-luc, wwp-luc, mdm2-luc, or pGL3-promoter reporter genes. Luciferase activity was measured 2 days after transfection. Results are expressed as luciferase relative to ␤-galactosidase enzymatic activity (n ϭ 3).
protein. Comparable levels of expression of the wild-type and mutant proteins were demonstrated by immunoblot analysis (Fig. 3B). Again, p53 protein production and nuclear localization were unaffected (data not shown). These observations indicate that the inhibitory effect of Bcl-2 on p53 transactivation requires its insertion into membranes.
Bcl-2 Inhibits p53-dependent p21/WAF-1 Up-regulation Following DNA Damage-The role of Bcl-2 in p53-mediated transactivation of target genes induced following DNA damage was investigated using MCF7 breast cancer cells that had been stably transfected with plasmids encoding Bcl-2, Bcl-2(⌬TM) or no inserted cDNA (neo-control). Polyclonal cell populations expressing comparable levels of ectopic proteins were chosen for experiments (Fig. 4). First the effects of DNA damage on p53 protein levels were examined. Cells were exposed to 0.4 M doxorubicin, and samples were harvested at different times. Nuclear and cytoplasmic extracts were prepared from each sample and subjected to immunoblot analysis for p53 expression. No p53 protein could be detected before cytotoxic treatment (data not shown). In contrast, as shown in Fig. 4A, p53 nuclear expression was strongly induced 24 h after doxorubicin treatment in MCF7 transfectants expressing either wild-type or mutant Bcl-2 protein, as well as in MCF7 control transfectants expressing only the neomycin resistance gene. Thus, as previously observed in transient transfection experiments (see above), the presence of high levels of Bcl-2 proteins did not interfere with p53 accumulation in the nucleus. Similar results were obtained for cells exposed to UV radiation (data not shown).
Next, to assess whether Bcl-2 also inhibits p53 transcriptional activity in the setting of DNA damage, MCF7 cells that stably expressed Bcl-2 or Bcl-2(⌬TM) were co-transfected with a p53-responsive luciferase reporter gene plasmid (wwp-luc), which contains a 2-kb fragment of the p21/waf-1 promoter. A ␤-galactosidase expression plasmid was co-transfected to normalize for transfection efficiency. One day later, cells were treated with 0.4 M doxorubicin or 10 J/m 2 UV radiation. Cell extracts were prepared 24 h later and assayed for luciferase and ␤-galactosidase enzyme activities. As shown in Fig. 4B, both doxorubicin and UV irradiation resulted in strong induction of the p21/waf-1 promoter in MCF7-Neo control cells. In contrast, transactivation of the p21/waf-1 promoter was re- Cytosolic (Cyto) and nuclear (Nuclei) extracts were prepared 24 h later for analysis by immunoblotting using anti-p53 antibodies. B, cells were transiently cotransfected with a p53responsive WWP-luciferase plasmid containing a 2-kb fragment of the p21/WAF-1 promoter and a ␤-galactosidase expression plasmid used to normalize for transfection efficiency. One day after transfection, cells were treated with 0.4 M doxorubicin or 10 J/m 2 UV radiation. Cell extracts were prepared 24 h later and assayed for luciferase and ␤-galactosidase enzyme activities. Results are expressed as a percentage of the maximal promoter activity measured in neo-control cells (mean Ϯ S.D.; n ϭ 3). C, Bcl-2 and p21 protein levels were assessed using lysates from stably transfected MCF7 cells before and after 24 h doxorubicin treatment. SDS-PAGE/immunoblot analysis was conducted using 35 g of total protein per lane and antibodies specific for Bcl-2 and p21. duced by ϳ70% in cells expressing high levels of Bcl-2. Unlike wild-type Bcl-2, MCF7 cells expressing Bcl-2(⌬TM) did not exhibit defects in their p53-mediated transactivation of the p21/waf-1 gene promoter after cytotoxic treatment (Fig. 4B). Cultures of UV-irradiated MCF7-Bcl-2 cells also became confluent, suggesting that cell proliferation was not inhibited despite treatment with DNA damaging agents. In contrast, cell cycle arrest was observed for Bcl-2(⌬TM) expressing cells (not shown).
To test whether the observed inhibition of the p53-mediated transactivation of a plasmid-borne p21/waf-1 promoter also applies to the endogenous p21/waf-1 gene, we examined the relative levels of p21 protein induction by immunoblot analysis of whole cell lysates prepared from MCF7 cells after exposure to DNA-damaging agents. As shown (Fig. 4C), p21 protein induction following treatment with doxorubicin was completely suppressed in MCF7-Bcl-2 but not in MCF7-Bcl2(⌬TM) cells. Fig. 4C also presents immunoblot analysis of the Bcl-2 and Bcl-2(⌬TM) proteins in stably transfected MCF7 cells, demonstrating that these two proteins were produced at comparable levels, thus excluding differences in their expression as a trivial explanation for the failure of Bcl-2(⌬TM) to suppress p21 protein induction.
Finally, similar observations were made by immunofluorescence microscopy, using antibodies specific for the Bcl-2, p53 or p21/WAF-1 proteins. As shown in Fig. 5A, the wild-type Bcl-2 protein was located in the nuclear envelope and perinuclear membranes, in a patter similar to that reported previously (15)(16)(17). In contrast, Bcl-2(⌬TM) was present diffusely through the cytosol and nucleus of MCF7 cells (Fig. 5B). Doxorubicin treatment resulted in marked up-regulation of p53 protein immunostaining in both MCF7-Bcl-2 and MCF7-Bcl-2(⌬TM) cells. In both lines, most of the p53 immunoreactivity was found in the nuclei of these cells (Fig. 5, C and D). Although p53 protein accumulated in the nuclei of doxorubicin-treated MCF7-Bcl-2 cells, expression of the p21/WAF-1 protein was not induced (Fig. 5E). In contrast, striking induction of nuclear p21/WAF-1 immunostaining was observed in MCF7-Bcl-2(⌬TM) cells (Fig. 5F). Thus, in agreement with the results obtained in transient transfection reporter gene assays, overexpression of Bcl-2 strongly reduced the induction of the p53 target gene p21/waf-1 after DNA damage, and this effect was again strictly dependent on the presence of the C-terminal membrane-anchoring domain of Bcl-2.
It is interesting to note that whereas MCF7-Neo cells died within 2 days after treatment, Bcl-2(⌬TM) and Bcl-2 transfectants survived for about 4 and over 7 days, respectively (data not shown). Thus, despite the inability of Bcl-2(⌬TM) to suppress p53-mediated transcriptional activity, it can apparently interfere with some downstream steps in the p53 pathway for apoptosis.

DISCUSSION
The Bcl-2 oncoprotein is known to arrest p53-mediated apoptosis in many cell types by mechanisms that are still not fully elucidated. It has been demonstrated that Bcl-2 inhibits downstream apoptotic events, such as the release of cytochrome c from the mitochondria and subsequent activation of caspases (59,60). However, Bcl-2 can also interfere with the activation of some transcription factors such as NFAT (12) and NF-B (11), promoting us to explore whether Bcl-2 is also capable of regulating upstream events in the p53 pathway. The data provided here indicate that overexpression of Bcl-2 can suppress p53mediated transactivation of p53 target genes. This effect was observed both in transient co-transfection reporter gene assays in which p53 was expressed from plasmids and in experiments in which endogenous p53 protein was activated by DNA-damaging agents. Thus, in addition to interfering with apoptotic events downstream of p53 target genes, Bcl-2 appears to be capable of suppressing an upstream step required for p53 function as a transcription factor.
We observed that anchoring of Bcl-2 to membranes is a requirement for its inhibitory effect on p53 transcriptional activity. Targeting of Bcl-2 to specific membrane compartments, such as endoplasmic reticulum and mitochondria, did not significantly alter the ability of Bcl-2 to inhibit p53. These observations are in agreement with the idea that at least some of the functions of Bcl-2 can be attributed to its ability to sequester cytosolic proteins at membranes (reviewed in Ref. 6). For instance, Bcl-2 was shown to inhibit NFAT-4 transcription factor signaling through sequestration of calcineurin, a phosphatase required for its dephosphorylation and subsequent nuclear translocation (12). A mutant of Bcl-2 lacking the membrane-anchoring domain still bound calcineurin but did not sequester it at membranes and did not suppress signaling through NFAT-4. Other reports have similarly described the recruitment of cytosolic proteins to internal membranes by overexpression of Bcl-2, including the protein kinase Raf1 (10) and the chaperone regulator BAG-1 (61).
The presence of the C-terminal membrane-anchoring domain has been reported to enhance the anti-apoptotic activity of Bcl-2 in most cellular contexts, but it may not be absolutely required at least for some of its functions (18,19). In this FIG. 5. Bcl-2 but not Bcl-2(⌬TM) inhibits p53-dependent p21/ WAF-1 protein induction following doxorubicin treatment. MCF7-Bcl-2 (A, C, and E) and MCF7-Bcl-2(⌬TM) (B, D, and F) stable transfectants were treated with 0.4 M doxorubicin for 24 h. Cells were then fixed, permeabilized, and immunostained with antibodies specific for Bcl-2 (A and B), p53 (C and D) and p21 (E and F). Antibodies were detected by indirect immunofluorescence microscopy as described under "Experimental Procedures." Immunostaining of MCF7-Neo cells is not reported because of the high mortality of these cells when seeded into polylysine-coated 8-well chambers. regard, our observation that Bcl-2(⌬TM) had no inhibitory effect on p53 transactivation of target genes, but did partially suppress p53-induced apoptosis, indicates that some of the cell death protective functions of Bcl-2 are independent of its membrane insertion. These observations may be relevant to other reports that have demonstrated that targeting of Bcl-2 to different locations results in a preferential protection against distinct apoptotic stimuli (48).
In contrast to previous reports that examined a Myc-transformed erythroleukemia or prostate cancer cell lines (43,44), in MCF7 and 293 cells, Bcl-2 did not inhibit the accumulation of p53 in the nucleus. Similarly, nuclear translocation of p53 is reportedly normal in a variety of other types of cell lines despite overexpression of Bcl-2 (40 -42, 46, 57, 62). We therefore speculate that Bcl-2 reduces the capacity of p53 to transactivate genes by regulating the activity of p53 itself or by affecting transcriptional coactivators required for p53 function. For instance, Bcl-2 might directly inactivate p53 by regulating posttranslational modifications required for its function as a transcription factor. In this regard, p53 has been shown to be phosphorylated at multiple sites by different kinases, such as DNA-activated kinases (63), the protein kinases CKI and CK II (64,65), cyclin-dependent kinases (66,67), stress-activated kinases (68), mitogen-activated protein kinases (69), Raf-1 (70), and PKC (71). Phosphorylation status regulates p53 conformation and ability to transactivate target genes (72). As mentioned above, Bcl-2 interacts with some kinases and phosphatases, regulating their intracellular distribution. It is therefore conceivable that Bcl-2 indirectly regulates p53 function by sequestration of a kinase or phosphatase that phosphorylates or dephosphorylates p53. Alternatively, Bcl-2 might interact with transcriptional coactivators required by p53, inhibiting their entry into the nucleus. Interestingly, Bcl-2 and p53 have been reported to compete for binding to 53BP2 (14), a protein of unknown function that was initially discovered in a two-hybrid based screen for proteins that bind wild-type but not mutant p53 (73).
Recently, it has been demonstrated that p53 requires p300/ CBP-family transcriptional co-activators for its transcriptional function (74). Proteins that disrupt this interaction, such as the E1a oncoprotein, reduce p53-dependent transactivation of the bax and p21/waf-1 promoters. Interestingly, Chawla et al. (75) recently reported that CBP contains a signal-regulated transcription activation domain that is controlled by nuclear calcium and calcium/calmodulin-dependent protein kinase IV. Calcium has also been shown to regulate p53 transcriptional activity under some circumstances (76 -78). Because Bcl-2 has been reported to regulate calcium efflux from the endoplasmic reticulum under some circumstances (79) and prevent nuclear accumulation of calcium (80), it is conceivable that Bcl-2 might inhibit p53 transcriptional activity by hindering Ca 2ϩ -dependent CBP activation.
Whatever the mechanism responsible for the suppression of p53-mediated transcription by Bcl-2, comparison with other reports suggests that the overall importance of this action of Bcl-2 may be dependent on cell context. For instance, in hematopoietic cells, Bcl-2 overexpression has been reported to inhibit apoptosis induced by a temperature sensitive mutant allele of p53 without impairing its induction of G 1 cell cycle arrest (62). Thus, in these cells, Bcl-2 apparently does not abrogate p53-mediated transcription of genes involved in inhibiting cell proliferation such as p21/waf-1. In contrast, Upadhyay et al. (81) reported that Bcl-2 overexpression can interfere with p53-mediated p21/WAF-1 induction in MCF10A breast cancer cells. Similarly, in some cell lines, it has been reported that Bcl-2 can inhibit p53-mediated repression of some promoters under circumstances where p53-mediated transactivation was unaffected (45,46). We therefore speculate that Bcl-2 may have several potential mechanisms for interfering with specific steps of p53 activation or function and that the ultimate effect of overexpressing Bcl-2 may be dictated by the relative levels or differential expression of various p53 cofactors or regulators that Bcl-2 can perturb within cells. In summary, although much remains to be learned about the specific mechanisms involved, the observation that Bcl-2 inhibits p53dependent transactivation of target genes might have important consequences for the response of cancer cells to cytotoxic treatments and thus contribute to the ability of Bcl-2 to promote chemo-and radioresistance.