Transcriptional repression by wild-type p53 utilizes histone deacetylases, mediated by interaction with mSin3a

Abstract

There is growing evidence that the p53 tumor suppressor protein not only can function to activate gene transcription but also to repress the expression of specific genes. Although recent studies have implicated the transcriptional repression function of p53 in the pathway of apoptosis, the molecular basis of this activity remains poorly understood. This study takes a first step toward elucidating this mechanism. We report that trichostatin A (TSA), an inhibitor of histone deacetylases (HDACs), abrogates the ability of p53 to repress the transcription of two genes that it negatively regulates, Map4 and stathmin. Consistent with this finding, we report that p53 physically associates in vivo with HDACs. This interaction is not direct but, rather, is mediated by the corepressor mSin3a. Both wild-type p53 and mSin3a, but not mutant p53, can be found bound to the Map4 promoter at times when this promoter preferentially associates with deacetylated histones in vivo. Significantly, inhibition of p53-mediated transcriptional repression with TSA markedly inhibits apoptosis induction by p53. These data offer the first mechanistic insights for p53-mediated transcriptional repression and underscore the importance of this activity for apoptosis induction by this protein.

Figure 1

Transcriptional repression of Map4 and Stathmin by p53 is inhibited by the HDAC inhibitor TSA. (A) Northern analysis of Map4 mRNA levels in Val5 cells at 39°C (lane 1, mutant p53) and following temperature shift to 32°C for 12 hr [wild-type (wt) p53, G₁ arrest] in the absence (lane 2) and presence (lane 3) of 100 nm TSA. The decrease of Map4 mRNA is largely alleviated by incubation with TSA during temperature shift and p53 induction (lane 3). In contrast, neither temperature shift nor TSA affects the levels of the housekeeping genes GAPDH or β-actin. (B) Northern analysis of Stathmin levels in MCF-7 cells treated with 0.5 μg/ml DOX for 0, 4, 8, and 24 hr indicates that repression of Stathmin expression is evident as early as 4 hr, but is greatest after 24 hr of DOX treatment. (C) Northern analysis of Stathmin levels in MCF-7 cells treated with 0.5 μg/ml DOX for 12 hr indicates that DOX treatment leads to 50% repression of Stathmin levels (lane 3); TSA treatment alone does not repress Stathmin levels (lane 2). The 50% repression of Stathmin by DOX treatment is effectivelyabrogated by TSA treatment at 100 nm (lane 4). In contrast, TSA treatment does not affect p53-mediated transactivation of the p53 response gene p21/waf1. (D) The averaged data from three independent Northern analyses, described above, using probes of comparable specific activity and following normalization to the levels of GAPDH. The error bars mark standard deviations. (E) Western analysis of p53 and MDM2 protein levels in MCF-7 cells treated with adriamycin (+ DOX, lane 2) or adriamycin and TSA (+ DOX + TSA, lane 3). These data indicate that TSA does not affect the post-translational stabilization of p53 following DOX treatment nor does it inhibit the induction of MDM2 protein. Equal microgram amounts of protein were loaded in each lane, and the blot was Coomassie-stained following protein detection to verify equal loading.

Figure 2

Wild-type p53 associates with HDAC1 and mSin3a in vivo. (A) IP–Western analysis of MCF-7 cell extract with antisera specific for rabbit IgG (Sigma, negative control), mSin3a (AK11 and K20 antibodies, Santa Cruz Biotechnology) and HDAC1 (C19, Santa Cruz Biotechnology), followed by Western analysis for p53 protein (DO-1 antibody, Calbiochem). Western analysis indicates the consistent presence of p53 in both HDAC1 (lane 4) and mSin3a (lanes 2,3,8) immunocomplexes. (B) IP with polyclonal antisera to p53 (p53 FL1–393, Santa Cruz Biotechnology) reveals the presence of both mSin3a and HDAC1 in a complex with this protein (lane 6). The amount of p53 present in mSin3a complexes increases approximately fivefold following post-translational stabilization of p53 protein mediated by 6 hr treatment in DOX (lane 9). (C) IP of MCF-7 cell lysate from untreated cells (lanes 1,4), cells treated with 4 J/m² UV irradiation (lanes 2,5), and cells treated with ALLN (lanes 3,6). In lanes 1–3, lysate was immunoprecipitated with antisera for mSin3a (AK-11, Santa Cruz Biotechnology) and Western blotted with DO-1 (Calbiochem). Lanes 4–6 represent Western blots of straight lysate from the indicated samples.

Figure 3

mSin3a interacts with both mutant and wild-type conformations of p53 protein. IP–Western analysis of Val5 cell extract (temperature-sensitive p53 protein) at 39°C (mutant p53) and following temperature shift to 32°C for 24 hr (wild-type p53). (A) Both wild-type and mutant forms of p53 protein are present in mSin3a immunocomplexes (top, lanes 3–6), and in HDAC1 immunocomplexes (lanes 7,8). No p53 is immunoprecipitated with irrelevant antisera (rabbit IgG, lanes 1,2). Fifty micrograms of total protein from each sample is loaded as a control (lanes 9,10). (B) (Top) mSin3a is present in p53 immunocomplexes precipitated by mAb 421, which detects both mutant and wild-type p53, (bottom), even following extensive washes in RIPA buffer (see Materials and Methods). Western analysis of the bottom panel was performed using polyclonal antisera raised against full-length p53 (FL-393, Santa Cruz Biotechnology).

Figure 4

Wild-type p53 and mSin3a interact with the Map4 promoter in vivo, resulting in decreased association of this promoter with acetylated histones. (A) ChIPs assay of the Map4 promoter (nucleotides −70 to +350, where +1 represents that start site of transcription) in Val5 cells cultured at 39°C (mutant p53) or 32°C (wild-type p53). IP of formaldehyde–cross-linked lysate was performed using 10 μg of anti-p53 or anti-mSin3a, followed by PCR using oligonucleotides specific for the Map4 promoter. As negative controls, lysis buffer alone was added to an anti-p53 IP (lanes 6,10), or lysate was incubated without antibody (lanes 5,9). The doublet band represents an occasional internal priming event seen with the Map4 reverse primer. For total input chromatin, 1 μl of a 1:300 dilution of DNA was used for the PCR. (B) ChIPs analysis of the Map4 promoter in Val5 cells using antisera specific for acetylated histone H3 (Upstate Biotechnology, Inc.). Following immunoprecipitation of Val5 extract and PCR of eluted chromatin, the agarose gel was blotted and hybridized to the full-length Map4 promoter. As negative controls, no antisera or 5 μg of anti-acetylated histone H3 was used for the IP of extract and binding buffer alone, respectively. (C) Wild-type p53, but not mutant p53 or mSin3a, associates with the mdm2 promoter in vivo. Immunoprecipitated chromatin from the experiment in A was used for PCR of the murine mdm2 promoter, using oligonucleotides that flank the p53-binding sites in this promoter. Prior to PCR for 25 cycles, immunoprecipitated chromatin was diluted 1:1000. For total input chromatin, 1 μl of a 1:300 dilution was used. (D) ChIPs assays were performed as above using extract from H1299 cells (human lung adenocarcinoma, p53 null) transfected with either human wild-type p53 (lanes 5–7) or human p53 containing mutations of amino acids 22 and 23 (lanes 1–3). Ten micrograms of polyclonal antisera to p53 or mSin3a (Santa Cruz Biotechnology), or no antibody, was used for immunoprecipitations, followed by PCR of the Map4 promoter. Western analysis indicated that equivalent levels of wild-type p53 and the 22/23 mutant were expressed in the transfected cells.

Figure 5

p53 interacts with mSin3a in vitro. (A) GST fusion proteins of p53, which include the p53 amino acids listed, were assayed for binding to the in vitro-translated radiolabeled proteins listed above each panel. Input lanes include 50% of the radiolabeled input protein, unless noted otherwise. Arrows mark the major in vitro-translated product. These data indicate that p53 amino acids 1–160 (lane 3) and 311–393 (lane 5) interact with mSin3a in this assay. Negative controls include GST alone (lanes 1,7), and in vitro-translated HDAC1 (lanes 7,8). (B) Binding assays of in vitro-translated mSin3a (100% input in bottom panel) with in vitro-translated synthetic p53 mutants (25% of input in top panel). Although deletion of amino acids 360–393 of p53 does not inhibit interaction with mSin3a (middle panel, lane 2), deletion of amino acids 320–393 (the oligomerization domain, lane 3) abolished binding. Replacement of the p53 oligomerization domain with a modified leucine zipper tetramerization domain from GCN4 (TZ) allows for mSin3a binding (data not shown), and deletion of p53 amino acids 1–40 does not inhibit this binding (Δ1–40;TZ, lane 6). In contrast, deletion of amino acids 1–100 (Δ1–100;TZ) inhibits the ability of p53 to complex with mSin3a (middle panel, lane 8). As negative controls, luciferase showed undetectable binding to either mSin3a or p53, and mSin3a antiserum was unable to immunoprecipitate the p53 mutants incubated alone (bottom panel). (C) Summary of data obtained from in vitro binding assays; the combined data indicate that p53 amino acids 40–160 comprise the amino-terminal binding domain for mSin3a; amino acids 320–360 comprise the carboxy-terminal binding domain. (D) Binding of GST alone or a GST fusion protein encoding full-length p53 (GST–p53) with in vitro-translated radiolabeled full-length mSin3a, or mSin3a deletion mutants containing the paired amphipathic helical (PAH) domains listed above each panel. The mSin3a construct encoding PAH domains 2–3 constitutes the minimal p53 binding region. (E) A GST fusion protein containing mSin3a amino acids 392–475, which comprises the domain in between PAH domains 2 and 3, is sufficient to bind to in vitro-translated p53.

Figure 6

p53-induced apoptosis in the Vm10 cell line is inhibited by TSA. Caspase-3 activity was analyzed in the Vm10 cell extracts indicated; cells were cultured at 39°C (mutant p53) or 32°C [wild-type (wt) p53, apoptosis] in the presence or absence of 100 nm TSA for 24 hr. As positive controls, Vm10 cells were cultured at 39°C in the presence of 100 nm staurosporine (Calbiochem), TNFα (10 ng/ml, RD Systems) plus cycloheximide (CHX, 40 μg/ml), or dilution vehicle alone (DMSO, 39°C). The activity of caspase-3 is measured as the ability of cell extract to catalyze the cleavage of AMC–DEVD and release the AMC fluorochrome. Readings were performed in duplicate; in the case of temperature-shifted cells, the data plotted are the average results of three independent experiments, with standard deviations plotted on the error bars. For staurosporine and TNFα-treated cells, the data plotted are the averages of two independent experiments. As a negative control, an inhibitor specific for caspase-3 (Ac–DEVD–CHO) was also added to the cell extract (light stippled bars).

Figure 7

p53-mSin3a complexes exist for extended periods of time following DNA damage. IP of MCF7 cells treated with 0.5 μg/ml DOX for the indicated time points with antibodies to p53 (rabbit polyclonal to p53, p53, FL393, Santa Cruz Biotechnology), mSin3a (AK11, Santa Cruz Biotechnology), and p300 (N15, Santa Cruz Biotechnology), followed by Western analysis for p53 (DO-1, Calbiochem). The depicted data are representative of six independent experiments and were reproducible with two different antibodies to p300, which detected similar amounts of total p300 protein (see Materials and Methods).