Abstract
Estrogen receptor α (ERα) and tumor suppressor protein p53 exert opposing effects on cellular proliferation. As a transcriptional regulator, p53 is capable of activating or repressing various target genes. We have previously reported that ERα binds directly to p53, leading to down-regulation of transcriptional activation by p53. In addition to transcriptional activation, transcriptional repression of a subset of target genes by p53 plays important roles in diverse biological processes, such as apoptosis. Here, we report that ERα inhibits p53-mediated transcriptional repression. Chromatin immunoprecipitation assays reveal that ERα interacts in vivo with p53 bound to promoters of Survivin and multidrug resistance gene 1, both targets for transcriptional repression by p53. ERα binding to p53 leads to inhibition of p53-mediated transcriptional regulation of these genes in human cancer cells. Transcriptional derepression of Survivin by ERα is dependent on the p53-binding site on the Survivin promoter, consistent with our observation that p53 is necessary for ERα to access the promoters. Importantly, mutagenic conversion of this site to an activation element enabled ERα to repress p53-mediated transcriptional activation. Further, RNA interference–mediated knockdown of ERα resulted in reduced Survivin expression and enhanced the propensity of MCF-7 cells to undergo apoptosis in response to staurosporine treatment, an effect that was blocked by exogenous expression of Survivin. These results unravel a novel mechanism by which ERα opposes p53-mediated apoptosis in breast cancer cells. The findings could have translational implications in developing new therapeutic and prevention strategies against breast cancer. [Cancer Res 2007;67(16):7746–55]
Introduction
In response to various extracellular and intracellular signals, p53 mediates cellular processes, such as cell cycle arrest, apoptosis, senescence, and differentiation, depending on the signal and the cellular context (1–3). There is compelling evidence that transcriptional regulation by p53 is required to execute most of its functions, enabling it to be a tumor suppressor, although transcription-independent mechanisms can also contribute to p53 function (4, 5). Most studies have focused on transcriptional activation by p53 because of the strong association observed between target gene activation and tumor suppression. However, increasing evidence suggests that transcriptional repression of various target genes by p53 also plays an important role in tumor suppression (6). For example, Survivin and multidrug resistance gene 1 (MDR1) belong to a group of genes targeted for transcriptional repression by p53 (7–9). The Survivin gene promoter contains a p53-binding site with four 5-bp consensus sequence 5′-PuPuPuC(A/T)-3′ repeated in two pairs, each arranged as inverted repeats (head to head orientation such as ⇒⇐ ⇒⇐) with a spacer of 3 bp between the two 10-bp repeats (7). MDR1 promoter, on the other hand, has an atypical p53-binding site with the consensus sequence repeated in a head to tail (⇒⇒ ⇒⇒) orientation with a 12-bp spacer between the two pairs of repeats (8). Survivin is a member of the inhibitor of apoptosis (IAP) family. It is expressed in fetal tissues and in most types of cancer, but the expression is very low in normal adult tissues (10, 11). Unlike other members of the IAP family, Survivin can inhibit apoptosis and elicit drug/radiation resistance as well as positively regulate progression through the cell cycle. MDR1 gene encodes P-glycoprotein (PGP), an energy-dependent drug efflux pump. An increase in PGP is largely responsible for the emergence of multidrug-resistant cells (12).
Estrogen receptor (ER) α and ERβ are members of the superfamily of nuclear receptors (13, 14). These receptors mediate the effects of the ligand 17β-estradiol (E2) by functioning as transcriptional regulators that access various target gene promoters either by directly binding to specific estrogen response elements within the promoter or indirectly by interacting with other transcriptional regulators bound to the promoter. Further, several cases of ligand-independent activation of ERα mediated by its phosphorylation by various signaling pathways have been reported (15). Besides its function as a transcriptional regulator, ERα can also mediate several nongenomic effects of estrogen, including mobilization of intracellular calcium, production of cyclic AMP, activation of mitogen-activated protein kinase signaling pathway, increased phosphatidylinositol 3-kinase activity leading to the activation of protein kinase B/Akt and endothelial nitric oxide synthase, activation of membrane tyrosine kinase receptors, and phosphorylation of SRC homology–containing domain (16).
Both ERα (13, 17–20) and p53 (21, 22) play a pivotal role in normal mammary development and in breast oncogenesis. Besides mutations in the p53 gene, a variety of molecular abnormalities in signaling pathways upstream and downstream of p53 can alter its function in breast cancer (21, 22). Further, the balance between opposing cellular signaling pathways impinging on p53 affects its ability to function as a tumor suppressor. A body of accumulating evidence suggests the possibility of a cross-talk between pathways mediated by ERα and p53. For example, treatment of mice with placental hormones resulted in the nuclear accumulation and activation of p53 in response to DNA damage (23). Further, in murine models, early exposure to E2 and progesterone to mimic pregnancy was shown to induce and sustain nuclear p53 expression that protects against subsequent carcinogen challenge (24). Lack of such protection in BALB/c p53-null mammary epithelium further strengthened the pivotal role of p53 in hormone-induced protection against carcinogenesis (25). In MDM2-overexpressing breast cancer cells, enhanced ERα function was observed and E2 stabilized wild-type p53 without any increase in p53 gene transcription (26, 27). Further, p53 has been reported to be in a ternary complex with ERα and MDM2 (28). When MCF-7 cells expressing glucocorticoid receptor driven by a constitutive SV40 promoter was treated with E2, p53 and MDM2 accumulated in the cytoplasm (29). An interesting case of combined action of ERα and p53 bound to separate promoter half-site response elements resulting in synergistic stimulation of transcription was reported recently (30). Importantly, our demonstration of direct binding of ERα to p53 both in vitro and in vivo on endogenous p53 target gene promoters resulting in inhibition of transcriptional activation by p53 (31) has raised the possibility of multiple roles for ERα-p53 interaction in normal and abnormal cellular physiology. Here, we have addressed whether ERα can antagonize transcriptional repression by p53. We show that ERα binds to p53 on endogenous Survivin and MDR1 gene promoters, leading to inhibition of p53-mediated transcriptional repression of these genes. Further, alleviating p53-mediated transcriptional repression of Survivin contributes to the ability of ERα to inhibit apoptosis in human breast cancer cells.
Materials and Methods
Expression vectors and reporter constructs. The ERα and p53 expression vectors (31) and Survivin-luciferase reporter plasmids (32) have been described. The upstream and downstream boundaries of Survivin constructs are numbered with respect to the transcriptional initiation site (+1) at the fourth nucleotide (G) of the sequence GGCGCGCCATTA in the promoter (7). The −171/+22 Survivin-luc spacer mutant (−171/+22 SM Survivin-luc) that lacks the TCC spacer between the two consensus p53-binding half-sites was generated using QuikChange Site-Directed Mutagenesis kit (Stratagene), and the mutation was confirmed by sequencing both strands. Survivin expression construct was prepared by PCR-amplifying Survivin cDNA from MCF-7 cells using Survivin-specific primers (sense, 5′-CACGGATCCATGGGTGCCCCGACGTTGC-3′; antisense, 5′-CAGGAATTCTCAATCCATGGCAGCCAGC-3′). The PCR product was subcloned into pCR3.1 vector as a BamHI-EcoRI fragment. MDR1-luc reporter construct in pGL2-basic vector (8) was a gift from Dr. Kathleen W. Scotto (Cancer Institute of New Jersey, New Brunswick, NJ). The Sp1 expression plasmid and the XETL reporter plasmid were provided by Drs. Guntram Suske (Institute for Molecular Biology and Tumor Research, Marburg, Germany) and Michael Garabedian (New York University Cancer Institute, New York, NY), respectively.
Cell lines and culture. Saos2 (human osteosarcoma cells that are null for p53 and express very low levels of ERα) and MCF-7 (human breast cancer cells expressing both wild-type p53 and ERα) were maintained in DMEM supplemented with 10% fetal bovine serum (FBS) or 10% dextran-charcoal–treated FBS and 100 μg/mL penicillin-streptomycin at 37°C under 5% CO2.
Chromatin immunoprecipitation assay. The chromatin immunoprecipitation (ChIP) assay was done as described previously (31). The immunoprecipitated DNA was amplified by PCR followed by electrophoresis on 2% agarose gel and staining with ethidium bromide. For re-ChIP experiments, primary immunocomplexes were eluted in re-ChIP buffer [0.5 mmol/L DTT, 1% Triton X-100, 2 mmol/L EDTA, 150 mmol/L NaCl, 20 mmol/L Tris-HCl (pH 8.1)] and reimmunoprecipitated with ERα antibody. For re-re-ChIP, the eluted immunocomplexes were further immunoprecipitated with SRC-1, SRC-2, HDAC1, HDAC4, or E2F1 antibodies. The primers used for PCR amplification of eluted DNA are listed in the Supplementary Data.
Transfections and luciferase assays. Transient transfections for luciferase assays were done as described previously using either calcium phosphate precipitation (33) or Fugene 6 (Roche Diagnostics; ref. 31). In the case of treatment with estrogen and antiestrogens, cells were washed 24 h after transfection and exposed to fresh medium containing ethanol (vehicle) or 10 nmol/L E2 (Sigma) for further 24 h before harvesting for luciferase assay. Cells were also treated with 5 μmol/L tamoxifen (Sigma) or 50 nmol/L ICI 182,780 (Tochris) separately or in combination with 10 nmol/L E2.
RNA interference–mediated knockdown of endogenous ERα and p53. MCF-7 cells were plated at 2.5 × 105 per well in a six-well plate and, 24 h later, transfected with either nonspecific, ERα (50 nmol/L), or p53 (20 nmol/L) small interfering RNA (siRNA; Dharmacon) using Lipofectamine 2000 (Invitrogen). Cells were harvested 24 or 48 h after transfection and processed for RNA and protein analysis.
Quantitative real-time PCR. Total RNA from MCF-7 cells was extracted using the Absolutely RNA Miniprep kit (Stratagene). Total RNA (500 ng) was reverse transcribed in 20 μL of reaction using the iScript cDNA Synthesis kit (Bio-Rad). Real-time PCR was done using Taqman PCR master mix, probes, and primers (Applied Biosystems) or SYBR Green Supermix (Bio-Rad) as described previously (31).
Antibodies and Western blotting. Antibodies used for ChIP assays and Western blotting were as follows: p53 (DO-1), ERα (HC-20 and D-12), Survivin (FL-142), SRC-1 (M-341), SRC-3 (M-397), lamin A/C (H110), and poly(ADP-ribose) polymerase (PARP; F-2) from Santa Cruz Biotechnology; PARP (9542) from Cell Signaling Technology, Inc.; HDAC1 (2E10) from Millipore; HDAC4 from Active Motif; β-actin (A2066) and tubulin (T-2046) from Sigma; and horseradish peroxidase–conjugated secondary antibodies from Bio-Rad. Western blotting was done as described previously (31).
Subcellular localization. Cytoplasmic and nuclear proteins from MCF-7 cells were prepared by using Nuclear/Cytosol Fractionation kit (BioVision). Proteins were resolved by SDS-PAGE followed by Western blotting.
Cell proliferation assay. MCF-7 cells maintained in DMEM containing 10% dextran-charcoal–treated FBS for 5 days were plated at 1.5 × 105 per well in six-well plates and transfected the next day with nonspecific, p53, or ERα siRNAs and, 8 h after transfection, treated with vehicle or 100 nmol/L E2. Cells were harvested by trypsinization on day 3 (3rd day after seeding or 2nd day after transfection) and day 4. Cells were suspended in PBS containing 0.25% trypan blue and counted using hemocytometer.
Apoptosis assay. Cellular MCF-7 cells were plated in 10-cm dishes at 2.0 × 106 per dish, and 24 h later, cells were transfected with either nonspecific siRNA (50 nmol/L), ERα (50 nmol/L), or p53 (20 nmol/L) siRNAs and, 48 h later, treated with 1 or 5 μmol/L staurosporine (Calbiochem) for 5 or 24 h. Cellular proteins were resolved by 8% SDS-PAGE followed by Western blot analysis for PARP cleavage products.
Statistical analysis. Data are presented as the mean ± SD. Statistical analysis was done using Student's t test, assuming equal variance, and P value was calculated based on two-tailed test. A P value of <0.05 was considered statistically significant.
Results
ERα interacts with p53 on p53-repressible target promoters. We previously reported that ERα interacts with p53 bound to p53-activated genes, such as p21 and PCNA (31). To analyze the effect of ERα on p53-mediated transcriptional repression, we selected Survivin and MDR1 genes that are targets for transcriptional repression by p53. Functional binding sites for p53 have been characterized on both Survivin (7) and MDR1 (8) gene promoters. First, we analyzed whether ERα-p53 interaction occurs on these gene promoters that contain different types of p53-binding sites. ChIP assays in MCF-7 cells using either ERα or p53 antibodies showed that ERα and p53 coexist on the Survivin and MDR1 promoters in vivo at the p53-binding site–containing region but not at the nonspecific upstream promoter regions that do not contain any p53-binding sites (Fig. 1B). To ascertain that ERα interaction with Survivin and MDR1 promoters is dependent on p53, the ChIP assay was done in Saos2 cells transfected with p53 and ERα expression plasmids separately or together. ERα antibody immunoprecipitated p53-binding site–containing regions of both the endogenous Survivin and MDR1 gene promoters when the cells were transfected with p53 and ERα expression plasmids together but not when the latter plasmid alone was transfected (Fig. 1C, compare lanes 8 and 12 in the Survivin panel and lanes 9 and 13 in the MDR1 panel). Further, the ability of the ERα antibody to pull down the p53-binding site–containing promoter region (−211 to +29) of the endogenous Survivin gene in MCF-7 cells was abolished in cells where p53 was knocked down by specific siRNA (Fig. 1D , compare lanes 3 and 5). Transfecting p53-specific siRNA resulted in >90% reduction in p53 mRNA and protein levels (31). These results show that p53 is necessary to recruit ERα to both the Survivin and MDR1 gene proximal promoter regions containing p53-bindng sites.
Interaction of ERα and p53 proteins on Survivin and MDR1 gene promoters. A, schematic diagram of endogenous human Survivin and MDR1 genes showing promoter regions relevant to the current study. Arrow with +1, transcription start site. Horizontal arrows with numbers, primers used for amplifying p53-specific and nonspecific regions in the ChIP assay. B, ERα interacts with p53 bound to endogenous Survivin and MDR1 promoters. ChIP assay in MCF-7 cells was done with antibodies (ab) against p53, ERα, or rabbit (R IgG) or mouse (M IgG) IgG. DNA from the input and immunoprecipitates was analyzed by PCR using primers specific for p53-binding site on Survivin and MDR1 promoters. As a control, primers against a nonspecific (NS) upstream region of both the promoters were also used. C, p53 is necessary for interaction of exogenously expressed ERα with the p53-binding site in the endogenous Survivin and MDR1 promoters. Saos2 cells (4 × 106) were transfected using Fugene 6 with 30 ng p53 or 1.25 μg ERα alone or in combination. Cells transfected with pUC119 served as negative control. Twenty-four hours after transfection, cells were processed for ChIP assay. Antibodies used are noted above each lane. DNA from the input and immunoprecipitates was analyzed by PCR as in (B). NS site, PCR amplicon of nonspecific promoter region. D, endogenous ERα cannot bind to Survivin promoter when p53 is knocked down. MCF-7 cells were transfected with either nonspecific or p53 siRNA. Twenty-four hours later, cells were processed for ChIP analysis as in (B).
Interaction of ERα and p53 proteins on Survivin and MDR1 gene promoters. A, schematic diagram of endogenous human Survivin and MDR1 genes showing promoter regions relevant to the current study. Arrow with +1, transcription start site. Horizontal arrows with numbers, primers used for amplifying p53-specific and nonspecific regions in the ChIP assay. B, ERα interacts with p53 bound to endogenous Survivin and MDR1 promoters. ChIP assay in MCF-7 cells was done with antibodies (ab) against p53, ERα, or rabbit (R IgG) or mouse (M IgG) IgG. DNA from the input and immunoprecipitates was analyzed by PCR using primers specific for p53-binding site on Survivin and MDR1 promoters. As a control, primers against a nonspecific (NS) upstream region of both the promoters were also used. C, p53 is necessary for interaction of exogenously expressed ERα with the p53-binding site in the endogenous Survivin and MDR1 promoters. Saos2 cells (4 × 106) were transfected using Fugene 6 with 30 ng p53 or 1.25 μg ERα alone or in combination. Cells transfected with pUC119 served as negative control. Twenty-four hours after transfection, cells were processed for ChIP assay. Antibodies used are noted above each lane. DNA from the input and immunoprecipitates was analyzed by PCR as in (B). NS site, PCR amplicon of nonspecific promoter region. D, endogenous ERα cannot bind to Survivin promoter when p53 is knocked down. MCF-7 cells were transfected with either nonspecific or p53 siRNA. Twenty-four hours later, cells were processed for ChIP analysis as in (B).
ERα inhibits transcriptional repression of Survivin and MDR1 promoters by p53. Next, we investigated the functional consequences of the ERα-p53 interaction on p53-mediated transcriptional repression of these genes. First, we analyzed whether transcription from a Survivin minimal promoter (32) can respond to p53 and ERα. We chose a reporter with the minimal promoter of Survivin to exclude any upstream regulatory elements that may also contribute to the ERα effect on Survivin expression. Saos2 cells were transfected with the −171/+22 Survivin-luc reporter in the presence of expression plasmids for wild-type p53 together with ERα wt or a transcriptionally defective ERα L539A point mutant (34), the ERα Δ283-595 mutant that is devoid of the AF2 domain, or the empty vector pCR3.1. Cells were grown in the presence or absence of E2. Transcription of −171/+22 Survivin-luc was repressed by cotransfected p53. ERα wt, but not the ERα L539A and the ERα Δ283-595 mutants, could significantly alleviate the p53-mediated transcriptional repression of the Survivin promoter (Fig. 2A). Analysis of the exogenously expressed proteins showed that similar levels of p53 are present in cells cotransfected with either ERα or ERα Δ283-595 (Fig. 2C,, compare lanes 6 and 7 in the p53 panel), suggesting that the effect of ERα on transcriptional regulation by p53 is not due to altered levels of p53 protein in cotransfected cells. ChIP and re-ChIP analyses in transfected cells showed that both p53 and ERα are recruited to the p53-binding site of the −171/+22 Survivin-luc reporter and these proteins coexist in the complex (Fig. 2D), consistent with the observation that they are bound to the p53-binding site of the endogenous Survivin gene in both MCF-7 and Saos2 cells (Fig. 1B and C) and p53 is necessary for ERα recruitment (Fig. 1C and D). Thus, ERα by interacting with p53 bound to the Survivin proximal promoter is able to alleviate transcriptional repression by p53. In these cells, E2 is not necessary for inhibiting p53-mediated repression of transcription from Survivin promoter. A reporter containing a larger region of Survivin promoter (−2,781/+22 Survivin-luc) also showed comparable pattern of expression (Supplementary Fig. S1A). Similar results were obtained in human breast cancer cell lines ZR-75 (ERα wt, p53 wt) and SKBR3 (ERα negative, p53 mutant) after transient transfection with p53 and/or ERα expression plasmids (data not shown).
ERα wt, but not the AF2 domain mutant, alleviates p53-mediated transcriptional repression of Survivin and MDR1. A, 3 × 104 Saos2 cells maintained in DMEM containing 10% dextran-charcoal–treated FBS were transfected using Fugene 6 with 25 ng ERα, ERα L539A, equimolar amount of ERα Δ283-595 expression plasmids, or empty vector (pCR3.1) alone or in combination with 0.6 ng p53 expression plasmid along with 500 ng −171/+22 Survivin-luc reporter. The pLuc vector without any Survivin promoter region was transfected as a negative reporter control. Twenty-four hours after transfection, cells were treated with vehicle or 10 nmol/L E2 for further 24 h before harvesting for luciferase assay. Columns, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. ***, P < 0.001. Top, Survivin-luc reporter. B, ERα relieves the p53-mediated transcriptional repression of MDR1. Saos2 cells were transfected as in (A) with 10 ng ERα, ERα L539A, an equimolar amount of ERα Δ283-595, or empty vector pCR3.1 alone or in combination with 2 ng p53 in the presence of 500 ng −1,202/+118 MDR1-luc reporter. The pGL2-luc vector without any MDR1 promoter region was transfected as a negative reporter control. Twenty-four hours after transfection, cells were treated with vehicle or 10 nmol/L E2 for further 24 h before harvesting for luciferase assay. Columns, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. ***, P < 0.001. Top, MDR1-luc reporter. C, in an experiment parallel to the one shown in (A), similar DNA transfection cocktail was used to transfect another set of Saos2 cells. Proteins (50 μg) from transfected cells were resolved on a 10% SDS-PAGE gel followed by Western blot analysis with antibodies that recognize wild-type ERα (HC-20), mutant ERα Δ283-595 (D-12), and p53 (DO-1). D, p53 and ERα are present in a single transcriptional complex on exogenous Survivin promoter. Saos2 cells (5 × 106) were transfected with 30 ng p53 or 1.25 μg ERα constructs in the presence of 10 μg −171/+22 Survivin-luc plasmid. Twenty-four hours after transfection, cells were processed for ChIP assay using the p53 antibody (DO-1). Part of this primary ChIP was subjected to re-ChIP assay using the ERα antibody (HC-20). DNA from the input and all immunoprecipitates was analyzed by PCR as described under Fig. 1B. A, top, reporter. Horizontal arrows, primers used for PCR in the ChIP assay.
ERα wt, but not the AF2 domain mutant, alleviates p53-mediated transcriptional repression of Survivin and MDR1. A, 3 × 104 Saos2 cells maintained in DMEM containing 10% dextran-charcoal–treated FBS were transfected using Fugene 6 with 25 ng ERα, ERα L539A, equimolar amount of ERα Δ283-595 expression plasmids, or empty vector (pCR3.1) alone or in combination with 0.6 ng p53 expression plasmid along with 500 ng −171/+22 Survivin-luc reporter. The pLuc vector without any Survivin promoter region was transfected as a negative reporter control. Twenty-four hours after transfection, cells were treated with vehicle or 10 nmol/L E2 for further 24 h before harvesting for luciferase assay. Columns, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. ***, P < 0.001. Top, Survivin-luc reporter. B, ERα relieves the p53-mediated transcriptional repression of MDR1. Saos2 cells were transfected as in (A) with 10 ng ERα, ERα L539A, an equimolar amount of ERα Δ283-595, or empty vector pCR3.1 alone or in combination with 2 ng p53 in the presence of 500 ng −1,202/+118 MDR1-luc reporter. The pGL2-luc vector without any MDR1 promoter region was transfected as a negative reporter control. Twenty-four hours after transfection, cells were treated with vehicle or 10 nmol/L E2 for further 24 h before harvesting for luciferase assay. Columns, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. ***, P < 0.001. Top, MDR1-luc reporter. C, in an experiment parallel to the one shown in (A), similar DNA transfection cocktail was used to transfect another set of Saos2 cells. Proteins (50 μg) from transfected cells were resolved on a 10% SDS-PAGE gel followed by Western blot analysis with antibodies that recognize wild-type ERα (HC-20), mutant ERα Δ283-595 (D-12), and p53 (DO-1). D, p53 and ERα are present in a single transcriptional complex on exogenous Survivin promoter. Saos2 cells (5 × 106) were transfected with 30 ng p53 or 1.25 μg ERα constructs in the presence of 10 μg −171/+22 Survivin-luc plasmid. Twenty-four hours after transfection, cells were processed for ChIP assay using the p53 antibody (DO-1). Part of this primary ChIP was subjected to re-ChIP assay using the ERα antibody (HC-20). DNA from the input and all immunoprecipitates was analyzed by PCR as described under Fig. 1B. A, top, reporter. Horizontal arrows, primers used for PCR in the ChIP assay.
The effect of ERα on transcriptional repression of the MDR1 gene by p53 was similar to that observed in the case of the Survivin gene. Transcription of the −1,202/+118 MDR1 luciferase reporter was repressed in Saos2 cells by exogenous p53, and this repression was reversed by ERα wt but not by the mutants ERα L539A and ERα Δ283-595 (Fig. 2B). Further, as in the case of transcription from Survivin promoter, E2 was not necessary for ERα inhibition of p53-mediated repression of transcription from the MDRI promoter, although the inhibition of p53 function was augmented by E2 in the case of transcription from MDR1 promoter.
The p53-binding site in the Survivin promoter is critical for regulation of transcription by ERα. It has been reported that the 3-bp spacer between the two palindromic half-sites in the consensus p53-binding site in the Survivin promoter is critical for p53-mediated repression and deletion of this spacer in the binding site would transform it into a p53 activation element (7). We deleted the trinucleotide (TCC) spacer in the minimal Survivin promoter, −171/+22 Survivin-luc, to generate the spacer mutant −171/+22 SM Survivin-luc (Fig. 3A). ChIP and re-ChIP assays showed that both p53 and ERα are recruited to this mutant p53 site (Fig. 3B) in a manner similar to what was observed with the wild-type p53-binding site (Fig. 2D). When this modified Survivin-luc reporter was transfected into Saos2 cells in the presence of an increasing concentration of p53 expression construct, dose-dependent activation of transcription was observed, confirming that the p53-binding site, a repressor element in its wild-type form, is indeed converted into a p53-dependent activation site (Fig. 3C). To test if ERα could interact with p53 and affect the p53-mediated transcriptional activation on the −171/+22 SM Survivin-luc promoter, Saos2 cells were transfected with this modified reporter in the presence of p53 and ERα wt or ERα Δ283-595 mutant. Interestingly, ERα wt, but not the deletion mutant, was able to inhibit the p53-mediated transcriptional activation (Fig. 3D). This ability of ERα to counteract p53-mediated transcriptional activation of the Survivin gene carrying a mutated p53-binding element is consistent with our earlier observation where ERα effectively reduced transcription of the p21 and PCNA genes that are transcriptional targets for activation by p53 (31). Thus, ERα is able to bind p53 and counteract both its ability to activate and repress transcription of target genes.
ERα interacts with p53 bound to the mutated response element on Survivin promoter and inhibits transcriptional activation by p53. A, schematic representation of the spacer mutant −171/+22 SM Survivin-luc where the 3-bp spacer located between two palindromic half-sites of the consensus p53-binding site was deleted by site-directed mutagenesis. B, p53 and ERα are present in a single transcriptional complex on Survivin promoter with a spacer mutant p53 response element capable of transcription activation. Saos2 cells (5 × 106) were transfected using calcium phosphate precipitation method with 30 ng p53 or 1.25 μg ERα expression plasmids in the presence of 10 μg −171/+22 SM Survivin-luc plasmid. Twenty-four hours after transfection, cells were harvested followed by ChIP assay with p53 antibody (DO-1). Part of this primary ChIP was subjected to re-ChIP assay with ERα antibody (HC-20). DNA from the input and all immunoprecipitates was analyzed by PCR as described under Fig. 1B. C, dose response of transcriptional activation of Survivin promoter with a spacer mutant p53 response element. Saos2 cells (5 × 104) were transfected with increasing concentrations (0.6–10 ng) of p53 expression plasmid along with 500 ng −171/+22 SM Survivin-luc plasmid. Each transfection was done in triplicate, and 48 h after transfection, cells were harvested for luciferase assay. Columns, mean of three determinations per condition; bars, SD. D, ERα blocks transcriptional activation of mutant Survivin promoter by p53. Saos2 cells (5 × 106) were transfected with 25 ng ERα, equimolar quantities of deletion mutant ERα Δ283-595 expression plasmids, or vector (pCR3.1) alone or in combination with 0.6 ng p53 in the presence of 500 ng −171/+22 SM Survivin-luc reporter. Cells were harvested 48 h after transfection and processed for luciferase assay. Columns, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. ***, P < 0.001.
ERα interacts with p53 bound to the mutated response element on Survivin promoter and inhibits transcriptional activation by p53. A, schematic representation of the spacer mutant −171/+22 SM Survivin-luc where the 3-bp spacer located between two palindromic half-sites of the consensus p53-binding site was deleted by site-directed mutagenesis. B, p53 and ERα are present in a single transcriptional complex on Survivin promoter with a spacer mutant p53 response element capable of transcription activation. Saos2 cells (5 × 106) were transfected using calcium phosphate precipitation method with 30 ng p53 or 1.25 μg ERα expression plasmids in the presence of 10 μg −171/+22 SM Survivin-luc plasmid. Twenty-four hours after transfection, cells were harvested followed by ChIP assay with p53 antibody (DO-1). Part of this primary ChIP was subjected to re-ChIP assay with ERα antibody (HC-20). DNA from the input and all immunoprecipitates was analyzed by PCR as described under Fig. 1B. C, dose response of transcriptional activation of Survivin promoter with a spacer mutant p53 response element. Saos2 cells (5 × 104) were transfected with increasing concentrations (0.6–10 ng) of p53 expression plasmid along with 500 ng −171/+22 SM Survivin-luc plasmid. Each transfection was done in triplicate, and 48 h after transfection, cells were harvested for luciferase assay. Columns, mean of three determinations per condition; bars, SD. D, ERα blocks transcriptional activation of mutant Survivin promoter by p53. Saos2 cells (5 × 106) were transfected with 25 ng ERα, equimolar quantities of deletion mutant ERα Δ283-595 expression plasmids, or vector (pCR3.1) alone or in combination with 0.6 ng p53 in the presence of 500 ng −171/+22 SM Survivin-luc reporter. Cells were harvested 48 h after transfection and processed for luciferase assay. Columns, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. ***, P < 0.001.
Antiestrogens alleviate inhibition of p53 function by ERα. We investigated the effect of E2, tamoxifen, and ICI 182,780 (Faslodex/Fulvestrant) on the ability of ERα to counteract p53 function in human breast cancer cells. MCF-7 cells grown in phenol red–free medium containing 10% dextran/charcoal-stripped serum for 2 days were transfected with the −171/+22 Survivin-luc reporter in the presence or absence of expression constructs for ERα wt, ER mutants ERα539A and ERα Δ283-595, or pCR3.1 vector separately or in combination with p53. Twenty-four hours after transfection, the cells were treated with vehicle or E2. ERα wt, but not the point mutant ERα539A or the AF2 domain deletion mutant (ERα Δ283-595), was able to alleviate p53-mediated transcriptional inhibition of the Survivin promoter both in the presence and absence of E2 (Fig. 4A). Consistent with our earlier finding on antagonism of ERα toward transcription of the p53-activated p21 gene, E2 is not necessary for the inhibitory effect of ERα, but it augments the ability of ERα to counteract p53 function leading to increased Survivin transcription. This effect of E2 is consistent with results obtained with global gene expression profiling where Survivin was one of the genes transcriptionally up-regulated in response to E2 treatment of MCF-7 cells (35). To test the effect of antiestrogens on Survivin promoter activity, cells cotransfected with p53 and ERα expression plasmids were treated with tamoxifen or ICI 182,780 in the presence and absence of E2. Both tamoxifen and ICI 182,780 inhibited the ability of ERα to alleviate p53-mediated repression of Survivin transcription (Fig. 4A). These data further support the idea that ERα, by binding to p53, can inhibit transcriptional repression of Survivin by p53.
Inhibition of p53-mediated transcriptional repression of Survivin by ERα is augmented by E2 and reduced by ERα antagonists. A, both tamoxifen (Tam) and ICI 182,780 (ICI) reduce the ability of ERα to alleviate the p53-mediated transcriptional repression of Survivin promoter. MCF-7 cells maintained in DMEM containing 10% dextran-charcoal–treated FBS were transfected using Fugene 6 with 25 ng ERα, ERα L539A, equimolar quantities of deletion mutant ERα Δ283-595 expression plasmids, or vector (pCR3.1) alone or in combination with 0.6 ng p53 in the presence of 500 ng −171/+22 Survivin-luc reporter. The pLuc vector without any Survivin promoter region was transfected as a negative reporter control. Twenty-four hours after transfection, cells were treated with vehicle or 10 nmol/L E2 for further 24 h before harvesting for luciferase assay. Cells were also treated with 5 μmol/L tamoxifen or 50 nmol/L ICI 182,780 separately or in combination with 10 nmol/L E2. Each transfection was done in triplicate. Columns, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. **, P < 0.01; ***, P < 0.001. B, repression of endogenous Survivin expression by p53 is alleviated by ERα. Saos2 cells (1 × 106) in 10-cm dishes were transfected with 2 μg of p53 or ERα plasmid alone or in combination. Cells were harvested for RNA extraction 48 h after transfection. Survivin RNA level was determined by qRT-PCR using SYBR Green method. Relative expression levels were analyzed by ΔΔCT method and normalized to pCR3.1 vector-transfected cells. Columns, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. **, P < 0.01. C, endogenous ERα and p53 have opposite effects on endogenous Survivin expression in MCF-7 cells. Top, MCF-7 cells were transfected with nonspecific siRNA or siRNA specific to ERα (50 nmol/L) or p53 (20 nmol/L). Twenty-four hours after transfection, cells were harvested and proteins were subjected to Western blotting with antibodies against ERα, p53, and β-actin. NT, nontransfected cells. Bottom, total cellular RNA was isolated 24 h after transfection and Survivin RNA was analyzed by qRT-PCR. Columns, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. **, P < 0.01.
Inhibition of p53-mediated transcriptional repression of Survivin by ERα is augmented by E2 and reduced by ERα antagonists. A, both tamoxifen (Tam) and ICI 182,780 (ICI) reduce the ability of ERα to alleviate the p53-mediated transcriptional repression of Survivin promoter. MCF-7 cells maintained in DMEM containing 10% dextran-charcoal–treated FBS were transfected using Fugene 6 with 25 ng ERα, ERα L539A, equimolar quantities of deletion mutant ERα Δ283-595 expression plasmids, or vector (pCR3.1) alone or in combination with 0.6 ng p53 in the presence of 500 ng −171/+22 Survivin-luc reporter. The pLuc vector without any Survivin promoter region was transfected as a negative reporter control. Twenty-four hours after transfection, cells were treated with vehicle or 10 nmol/L E2 for further 24 h before harvesting for luciferase assay. Cells were also treated with 5 μmol/L tamoxifen or 50 nmol/L ICI 182,780 separately or in combination with 10 nmol/L E2. Each transfection was done in triplicate. Columns, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. **, P < 0.01; ***, P < 0.001. B, repression of endogenous Survivin expression by p53 is alleviated by ERα. Saos2 cells (1 × 106) in 10-cm dishes were transfected with 2 μg of p53 or ERα plasmid alone or in combination. Cells were harvested for RNA extraction 48 h after transfection. Survivin RNA level was determined by qRT-PCR using SYBR Green method. Relative expression levels were analyzed by ΔΔCT method and normalized to pCR3.1 vector-transfected cells. Columns, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. **, P < 0.01. C, endogenous ERα and p53 have opposite effects on endogenous Survivin expression in MCF-7 cells. Top, MCF-7 cells were transfected with nonspecific siRNA or siRNA specific to ERα (50 nmol/L) or p53 (20 nmol/L). Twenty-four hours after transfection, cells were harvested and proteins were subjected to Western blotting with antibodies against ERα, p53, and β-actin. NT, nontransfected cells. Bottom, total cellular RNA was isolated 24 h after transfection and Survivin RNA was analyzed by qRT-PCR. Columns, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. **, P < 0.01.
Opposing effects of p53 and ERα on endogenous Survivin expression. To analyze the effect of ERα and p53 on endogenous Survivin RNA levels, Saos2 cells were transfected with ERα and p53 expression plasmids separately or in combination. Twenty-four hours after transfection, cells were harvested and endogenous Survivin RNA was analyzed by quantitative real-time PCR (qRT-PCR). Consistent with the data from reporter luciferase assays, transfection of exogenous p53 led to decreased Survivin RNA level, whereas the repression was alleviated by cotransfected exogenous ERα (Fig. 4B).
As ectopic expression of ERα can inhibit p53-mediated down-regulation of endogenous Survivin RNA in Saos2 cells, we examined the effect of knocking down ERα and p53 separately in MCF-7 cells. Protein analysis by Western blotting showed that both ERα and p53 expression was efficiently knocked down by respective siRNAs (Fig. 4C , top). qRT-PCR assay showed that endogenous Survivin RNA was significantly decreased when ERα was knocked down. On the other hand, knocking down p53 resulted in increased level of Survivin RNA. Thus, data on the effect of p53 and ERα on endogenous Survivin expression in both Saos2 and MCF-7 cells are consistent with the results from our transient luciferase reporter transcription assays.
ERα inhibits p53-mediated transcriptional repression of Survivin by recruiting histone deacetylases and by facilitating enhanced E2F1 binding. Toward analyzing the mechanism underlying the ability of ERα to inhibit transcriptional repression by p53, we examined whether ERα with or without E2 affects subcellular localization of p53. Cytoplasmic and nuclear fractions were prepared from MCF-7 cells grown in the presence or absence of E2 followed by Western blot analysis of proteins. E2 did no affect the p53 protein levels either in the cytoplasm or in the nucleus (Fig. 5A). Further, ChIP assay in MCF-7 cells grown in the presence or absence of E2 also showed that E2 has no effect on recruitment of p53 or ERα to the Survivin promoter (Fig. 5B). These observations are consistent with our data showing similar p53 binding to endogenous Survivin promoter in the presence or absence of ERα both in Saos2 cells (Fig. 1C,, compare lanes 6 and 10) and in MCF-7 cells (Fig. 5C). Thus, ERα, with or without E2, does not affect nuclear localization of p53 and its accessibility to genomic DNA. Then, we tested the possibility that ERα is counteracting p53-mediated transcriptional repression of Survivin by recruiting its transcriptional coactivators SRC-1 and SRC-3 (AIB1). Sequential ChIP assays showed that neither of these coactivators is present along with p53 on the Survivin promoter in the presence or absence of ERα (Fig. 5C), and hence, countering repressor function of p53 by ERα does not seem to depend on increased ERα coactivator presence. Next, we turned our attention to histone deacetylases (HDAC), as p53 function is regulated by acetylation status of its various amino acids (36, 37). HDAC1 can interact with ERα and is also capable of deacetylating p53 (38). Recently, HDAC4 was reported to interact with ERα (39). We tested whether ERα bound to p53 recruits these two HDACs. As shown in Fig. 5D, sequential ChIP analysis revealed that both HDAC1 and HDAC4 are recruited to the Survivin promoter along with p53 and ERα. When ERα is knocked down, neither of these deacetylases is recruited to the promoter. These results indicate that recruitment of HDAC1 and HDAC4, which in turn may lead to deacetylation and functional inactivation of p53, could be one of the mechanisms by which ERα counteracts transcriptional repression by p53.
ERα does not affect subcellular localization and DNA binding of p53 but affects HDAC recruitment and E2F1 binding to the Survivin promoter. A, cytosolic and nuclear fractions of cell lysates from MCF-7 cells grown in the presence of vehicle or E2 were subjected to Western blot analysis using antibodies against p53 and ERα. Tubulin and lamin A/C expression was monitored as protein loading controls and as markers of cytoplasmic and nuclear compartment integrity, respectively. B, MCF-7 cells grown in medium supplemented with 10% dextran-charcoal–treated FBS were seeded at a density of 3 × 106/10 mL. Twenty-four hours after seeding, cells were treated with vehicle or 10 nmol/L E2. Twenty-four hours after treatment, cell lysates were prepared for ChIP assay. PCR products from the ChIP assay were resolved on an agarose gel. Bottom of each lane, antibodies used in the ChIP assay. C, MCF-7 cells were transfected with nonspecific or ERα siRNA (100 nmol/L each), and after 48 h, cell lysates were prepared for sequential ChIP. Samples were serially immunoprecipitated with p53 antibody followed by ERα antibody and followed by SRC-1 or SRC-3 antibodies. PCR products from the sequential ChIP assay were resolved on an agarose gel. D, MCF-7 cells transfected with siRNAs as in (C) were subjected to sequential ChIP assay. Serial immunoprecipitations were in the following order: p53 antibody followed by ERα antibody followed by HDAC1 or HDAC4 or E2F1 antibodies. PCR products of the sequential ChIP assay were resolved on an agarose gel.
ERα does not affect subcellular localization and DNA binding of p53 but affects HDAC recruitment and E2F1 binding to the Survivin promoter. A, cytosolic and nuclear fractions of cell lysates from MCF-7 cells grown in the presence of vehicle or E2 were subjected to Western blot analysis using antibodies against p53 and ERα. Tubulin and lamin A/C expression was monitored as protein loading controls and as markers of cytoplasmic and nuclear compartment integrity, respectively. B, MCF-7 cells grown in medium supplemented with 10% dextran-charcoal–treated FBS were seeded at a density of 3 × 106/10 mL. Twenty-four hours after seeding, cells were treated with vehicle or 10 nmol/L E2. Twenty-four hours after treatment, cell lysates were prepared for ChIP assay. PCR products from the ChIP assay were resolved on an agarose gel. Bottom of each lane, antibodies used in the ChIP assay. C, MCF-7 cells were transfected with nonspecific or ERα siRNA (100 nmol/L each), and after 48 h, cell lysates were prepared for sequential ChIP. Samples were serially immunoprecipitated with p53 antibody followed by ERα antibody and followed by SRC-1 or SRC-3 antibodies. PCR products from the sequential ChIP assay were resolved on an agarose gel. D, MCF-7 cells transfected with siRNAs as in (C) were subjected to sequential ChIP assay. Serial immunoprecipitations were in the following order: p53 antibody followed by ERα antibody followed by HDAC1 or HDAC4 or E2F1 antibodies. PCR products of the sequential ChIP assay were resolved on an agarose gel.
Interestingly, an E2F-binding site overlaps with p53-binding site on the Survivin promoter (7). E2F1 can bind to p53 (40) and, when recruited to the Survivin promoter, activates transcription (41). Therefore, we looked at E2F1 recruitment to the Survivin promoter in the presence and absence of ERα. E2F1 was recruited to the promoter along with p53 and ERα, but when ERα was knocked down, the recruitment was considerably reduced (Fig. 5D). Thus, facilitating E2F1 recruitment to the promoter could be another mechanism by which ERα opposes transcriptional repression by p53.
Knocking down ERα decreases cell proliferation and down-regulates Survivin expression, leading to an increase in staurosporine-induced, p53-mediated apoptosis in breast cancer cells. To address the physiologic consequence of inhibiting p53 function by ERα, we analyzed proliferation of MCF-7 cells transfected with nonspecific siRNA in comparison with those transfected with ERα or p53 siRNA in the presence and absence of E2. Knocking down ERα significantly reduced cell growth of MCF-7 cells compared with control cells. On the contrary, knocking down p53 resulted in increased proliferation (Fig. 6A). As expected, although the proliferation of control MCF-7 cells was higher in the presence of E2, the opposite effect of down-regulating ERα and p53 was evident both in the presence and absence of E2.
Knocking down ERα in breast cancer cells leads to decreased cell proliferation and increased p53-dependent apoptosis. A, MCF-7 cells maintained in DMEM containing 10% dextran-charcoal–treated FBS were transfected with nonspecific, p53, or ERα siRNAs and, 8 h after transfection, treated with vehicle or 100 nmol/L E2. Cells were harvested by trypsinization on the 3rd and 4th day after seeding, suspended in PBS containing 0.25% trypan blue, and counted using hemocytometer. Proliferation curves for cells in the presence and absence of E2. Points, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. *, P < 0.05; **, P < 0.01. B, MCF-7 cells were transfected with nonspecific, p53, or ERα siRNAs and, 48 h later, subjected to staurosporine (STS; 5 μmol/L) treatment for 24 h. Cell lysates were prepared for PARP cleavage analysis by Western blotting. C, MCF-7 cells were transfected with nonspecific or ERα siRNAs. Forty-eight hours after transfection, they were retransfected with Survivin expression plasmid or empty vector. Twenty-four hours after transfection, cells were treated with 1 μmol/L staurosporine for 5 h. Cells were harvested and 50 μg of total cellular protein were analyzed on 8% SDS-PAGE followed by immunoblotting with PARP antibody. Bands representing cleaved and uncleaved PARP were quantified by densitometry. Numbers at the bottom, ratios of cleaved to uncleaved PARP.
Knocking down ERα in breast cancer cells leads to decreased cell proliferation and increased p53-dependent apoptosis. A, MCF-7 cells maintained in DMEM containing 10% dextran-charcoal–treated FBS were transfected with nonspecific, p53, or ERα siRNAs and, 8 h after transfection, treated with vehicle or 100 nmol/L E2. Cells were harvested by trypsinization on the 3rd and 4th day after seeding, suspended in PBS containing 0.25% trypan blue, and counted using hemocytometer. Proliferation curves for cells in the presence and absence of E2. Points, mean of three determinations per condition; bars, SD. Statistical analysis was done using Student's t test. *, P < 0.05; **, P < 0.01. B, MCF-7 cells were transfected with nonspecific, p53, or ERα siRNAs and, 48 h later, subjected to staurosporine (STS; 5 μmol/L) treatment for 24 h. Cell lysates were prepared for PARP cleavage analysis by Western blotting. C, MCF-7 cells were transfected with nonspecific or ERα siRNAs. Forty-eight hours after transfection, they were retransfected with Survivin expression plasmid or empty vector. Twenty-four hours after transfection, cells were treated with 1 μmol/L staurosporine for 5 h. Cells were harvested and 50 μg of total cellular protein were analyzed on 8% SDS-PAGE followed by immunoblotting with PARP antibody. Bands representing cleaved and uncleaved PARP were quantified by densitometry. Numbers at the bottom, ratios of cleaved to uncleaved PARP.
Because p53 has a profound proapoptotic function, and repression of antiapoptotic Survivin by p53 is countered by ERα, we tested whether ERα plays any role in p53-mediated apoptosis. To determine the effect of decreased ERα expression on apoptosis, MCF-7 cells were transfected with siRNA specific to ERα or p53 and, 48 h after transfection, treated with staurosporine, a broad kinase inhibitor and a potent inducer of apoptosis (42). Apoptosis in these cells was analyzed by the PARP cleavage assay. Control cells untreated with staurosporine showed very low apoptosis even when ERα was knocked down. Cells transfected with nonspecific siRNA were relatively resistant to apoptosis whether they were treated with staurosporine or not, reflecting functional suppression of p53 by ERα. On the other hand, the intensity of cleaved PARP (85-kDa fragment) was enhanced in ERα knock down cells treated with staurosporine (Fig. 6B). These data suggest an antiapoptotic role for ERα by opposing the proapoptotic p53 function. To test if Survivin is a mediator of antiapoptotic effect of ERα, we analyzed whether exogenous expression of Survivin in ERα knockdown cells would decrease staurosporine-induced apoptosis. MCF-7 cells were transfected with either Survivin expression plasmids or vector followed by staurosporine treatment. Expression of exogenous Survivin partially compromised the staurosporine-induced PARP cleavage, suggesting that Survivin is one of the mediators of the antiapoptotic function of ERα (Fig. 6C).
Discussion
The p53 gene is mutated in ∼20% of breast cancers (43). Thus, although the majority of breast cancers have wild-type p53, it remains dysfunctional as a tumor suppressor. Various mechanisms for functionally disabling wild-type p53 have been documented (22, 44). We had recently reported a novel mechanism of functional inactivation of p53 bound to target gene promoters, such as p21, in vivo in the chromatin context by ERα directly interacting with p53 (31). Unlike these genes activated by p53, Survivin and MDR1 genes are targets for transcriptional repression by p53 (6). The data presented here show that ERα interacts with p53 bound to both Survivin and MDR1 promoters, leading to derepression of their transcription. p53 is necessary for the recruitment of ERα to these promoters, as neither exogenously expressed ERα in Saos2 cells nor endogenous ERα in p53 knockdown MCF-7 cells could access the promoters. As was shown in the case of inhibiting transcriptional activation by p53 (31), sequences in the AF2 domain of ERα are necessary for countering transcriptional repression by p53. Intriguingly, in the case of the Survivin promoter, p53 and ERα coexist on the p53-binding site irrespective of whether it is in a repressor or an activator (after a 3-bp spacer deletion) configuration and results in inhibition of both transcriptional repression as well as activation by p53. Therefore, it seems that, once bound to p53, ERα can actively counter transcriptional repression or activation functions of p53. It is likely that ERα is recruiting its coactivators and corepressors when p53 is bound to a promoter response element in the repression and activation configuration, respectively. Alternatively, ERα may recruit other repressors of p53 function. Our observation that HDAC1 and HDAC4, but not coactivators SRC-1 and SRC-3 (AIB1), are recruited to the Survivin promoter favors the latter scenario. The HDACs recruited in ERα-dependent manner may deacetylate p53, disabling it as a repressor of Survivin transcription. As deacetylation of chromatin histones is usually associated with repression rather than activation of transcription, HDAC1 and HDAC4 do not seem to be involved in general histone deacetylation of the Survivin promoter. We have also found that E2F1 is efficiently recruited to the proximal region of the Survivin promoter in the presence of ERα, suggesting that E2F1 may also be contributing to suppression of p53 function by ERα. Besides these interactions, binding of ERα to p53 might also alter the conformation of either or both p53 and ERα affecting their functions. These mechanisms need not be mutually exclusive and future studies should delineate the relevant interactions.
Of note, in cells subjected to genomic damage, interaction between p53 and DNA (cytosine-5)-methyltransferase (DNMT1) concomitant with overall hypermethylation of the Survivin promoter that contributed to repression of transcription has been reported recently (45). Intriguingly, both DNMT1 and ERα bind to the COOH-terminal regulatory domain of p53 (31, 45). An attractive possibility that ERα excludes DNMT1 from binding to p53 resulting in alleviation of repression of Survivin transcription remains to be tested.
As in the case of inhibition of p53-mediated transcriptional activation (31), E2, although not necessary, augments the ability of ERα to alleviate p53-mediated repression of Survivin transcription. That tamoxifen and ICI 182,780, known to alter ERα conformation (46), counter transcriptional activation of Survivin in MCF-7 cells suggests the importance of ERα conformation in its ability to inhibit p53-mediated regulation of Survivin transcription. Various potential mechanisms by which antiestrogens could cause cell death have been proposed (47). Our data suggest that countering the inhibitory effect of ERα on p53 function might be one of the mechanisms underlying the proapoptotic effect of antiestrogens. It is possible that this mode of action may contribute to the response of breast cancer to antiestrogen therapy.
Although various studies have addressed the mechanisms underlying the role of ERα in cell cycle progression, the mechanisms underlying its antiapoptotic capability have remained largely unknown (47, 48). The data presented here show that ERα interacts with p53 bound to the Survivin promoter, resulting in rescue of this antiapoptotic gene from the transcriptional repression imposed by p53, leading to down-regulation of stress-induced, p53-mediated apoptosis in MCF-7 breast cancer cells. The blockage of apoptosis that was lifted when ERα was knocked down in these cells was partially reimposed when Survivin was exogenously expressed. Thus, countering p53-mediated transcriptional repression of Survivin is at least one of the important mechanisms underlying the antiapoptotic function of ERα.
Consistent with these data, overexpression of Survivin in cells sensitive to p53-dependent apoptosis has been reported to result in inhibition of apoptosis (7, 9). Further, expression of Survivin has been related to loss of apoptosis in breast carcinomas, and interestingly, Survivin expression was not limited to tumors with mutant p53 (49). That report, along with a recent study that analyzed transcript profiles of primary breast tumors of known p53 status and concluded transcriptional fingerprint of p53 to be a more definitive indicator of p53 function than p53 mutational status (50), is consistent with our observation that wild-type p53 can be functionally inhibited by ERα leading to up-regulation of Survivin and suppression of apoptosis. Studies addressing Survivin expression and apoptosis in breast tumors differing in ER (positive versus negative) and p53 (wild-type versus mutant) status should provide valuable information on the clinical relevance of the relationship between these proteins. Future studies should also uncover the cellular consequences of ERα affecting transcriptional regulation of other p53 target genes. The findings reported here would contribute to a better understanding of how ER signaling affect p53 function, which in turn could be exploited for developing novel cancer preventive and therapeutic strategies.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
Current address for A. Sayeed: California Pacific Medical Center, San Francisco, CA 94107. Current address for S.D. Konduri: M. D. Anderson Cancer Center Orlando, Cancer Research Institute, Orlando, FL 32806.
Acknowledgments
Grant support: National Cancer Institute grants CA079911 (G.M. Das) and CA 016056 (Roswell Park Cancer Institute Cancer Center Support Grant) and Susan G. Komen Breast Cancer Foundation (G.M. Das).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank S. Alexander for help in apoptosis assays; Drs. A. Levine, S. Smith, Kathleen W. Scotto, Guntram Suske, and Michael Garabedian for plasmids; S. Conrad for providing the MCF-7 cells; and Dr. Margot Ip for critically reading the manuscript.