Tumor suppressor protein p53 recruits human Sin3B/HDAC1 complex for down-regulation of its target promoters in response to genotoxic stress

Abstract

Master regulator protein p53, popularly known as the "guardian of genome" is the hub for regulation of diverse cellular pathways. Depending on the cell type and severity of DNA damage, p53 protein mediates cell cycle arrest or apoptosis, besides activating DNA repair, which is apparently achieved by regulation of its target genes, as well as direct interaction with other proteins. p53 is known to repress target genes via multiple mechanisms one of which is via recruitment of chromatin remodelling Sin3/HDAC1/2 complex. Sin3 proteins (Sin3A and Sin3B) regulate gene expression at the chromatin-level by serving as an anchor onto which the core Sin3/HDAC complex is assembled. The Sin3/HDAC co-repressor complex can be recruited by a large number of DNA-binding transcription factors. Sin3A has been closely linked to p53 while Sin3B is considered to be a close associate of E2Fs. The theme of this study was to establish the role of Sin3B in p53-mediated gene repression. We demonstrate a direct protein-protein interaction between human p53 and Sin3B (hSin3B). Amino acids 1-399 of hSin3B protein are involved in its interaction with N-terminal region (amino acids 1-108) of p53. Genotoxic stress induced by Adriamycin treatment increases the levels of hSin3B that is recruited to the promoters of p53-target genes (HSPA8, MAD1 and CRYZ). More importantly recruitment of hSin3B and repression of the three p53-target promoters upon Adriamycin treatment were observed only in p53(+/+) cell lines. Additionally an increased tri-methylation of the H3K9 residue at the promoters of HSPA8 and CRYZ was also observed following Adriamycin treatment. The present study highlights for the first time the essential role of Sin3B as an important associate of p53 in mediating the cellular responses to stress and in the transcriptional repression of genes encoding for heat shock proteins or proteins involved in regulation of cell cycle and apoptosis.

Figure 1. Phosphorylated hSin3B associates with hp53 in vivo.

(A & B) Cell lysates from KB, HEK293 (A) and HCT116 cell lines (B) were immunoprecipitated (IP) with antibody specific for p53 followed by immunoblot analysis (IB) with antibodies specific for hSin3B (sc-13145 for KB and HCT116 cell lines; sc-55516 for HEK293 cells), phosphorylated serine (anti-pSer), phosphorylated tyrosine (anti-pTyr), phosphorylated threonine (anti-pThr) as indicated above each lane. Western analysis indicates the co-immunoprecipitation of phosphorylated hSin3B with p53 in KB, HEK293 and HCT116 cell extracts. (C) IP-Western analysis in p53-null cell line (Saos2) shows that hSin3B was detectable only in the input lane but not in the immune complex obtained from antibody against p53 or in the mock immunoprecipitates. (D) Reciprocal IP-Western analysis in KB cell extract using the ImmunoCruz™ IP/WB Optima E System (Santa Cruz) as described in the methods section reveals the presence of p53 in a complex with hSin3B. In all the experiments input corresponds to 10% of the total cell lysate used for each immunoprecipitation.

Figure 2. Yeast two Hybrid analysis for the interaction of hSin3B with hp53.

(A) & (B) Yeast AH109 cells were co-transformed with plasmids indicated below the plates for each sector. Successful co-transformations were confirmed by growth on SD LT plates (Drop-out medium lacking Leucine and tryptophan). The protein-protein interactions were checked by growing the co-transformants on selective SD QDO-Xgal medium (Quadruple drop-out medium lacking leucine, tryptophan, adenine and histidine and containing X-gal). Positive interaction was observed only between pGBKT7-Sin3B_1–399 and pGADT7-hp53 (Figure 2A, sector A) as well as pGBKT7-Sin3B_193–468 and pGADT7-hp53 (Figure 2B, Sector A). (C) Schematic representation of the various truncated forms of hSin3B used in the yeast two hybrid assays. Each truncated Sin3B construct was co-tranformed with hp53 in AH109 cells and interaction was checked by observing growth on selective medium (SD QDO-Xgal). A plus sign (+) indicates positive interaction and negative sign (−) indicates no interaction. (D) β-galactosidase assays were performed to quantify two-hybrid interactions. A 9.9±1.813 fold increase in the relative β-galactosidase units was observed for hp53/Sin3B_1–399 interactions while a 1.9±0.107 fold increase was observed for hp53/Sin3B_193–468 interaction. All values are plotted with ±SEM calculated for three independent experiments. (E) Yeast AH109 cells were co-transformed with plasmids indicated below the plates for each sector. Positive interaction was observed between pGBKT7-Sin3B_1–399 and pGADT7-hp53_1–108 (sector B) as indicated by growth on selective medium (SD QDO-Xgal).

Figure 3. Up-regulation of hSin3B in response to Adriamycin is p53-dependent.

(A) KB and HCT116 cells were treated with 1.0 µg/ml Adriamycin for 16 hours followed by propidium iodide staining and cell cycle analysis. Adriamycin treatment induced a predominant G2 cell cycle arrest in KB cells and S/G2 arrest in HCT116 cells. (B) Total RNA was isolated and cDNA was synthesized from KB and HCT116 cells with or without Adriamycin treatment. Semi-quantitative PCR results indicated increased levels of p53 and hSin3B mRNA levels in Adriamycin treated cells. (C) Upper panel shows the results of immuno-fluorescence assays using flow cytometry. Lower panel is a plot of the above results comparing the mean fluorescence intensity for p53 and hSin3B in the untreated and Adriamycin treated cells. A significant increase in p53 (P = 0.0049 in KB and P = 0.0036 in HCT116 cells) and hSin3B proteins (P = 0.0234 in KB and P = 0.0365 in HCT116 cells) was observed following Adriamycin treatment. The values have been plotted with ±SEM calculated from three (n = 3) independent experiments. (D) Western analysis of cell lysates of control and Adriamycin treated KB cells showed an increase in the hSin3B and p53 protein levels upon treatment with 1.0 and 2.0 µg/ml Adriamycin. (E) IP-Western analysis of KB cell extract after treatment with 1.0 µg/ml Adriamycin indicates the co-immunoprecipitations of hSin3B with p53 both before and after Adriamycin treatment. (F) Results of semi-quantitative PCR (upper panel) and immuno-fluorescence assays using flow cytometry (lower panel) showed no significant change in the expression levels of either hSin3B transcript or protein in p53-null cells viz. (i) Saos2 (ii) H1299 and (iii) Hep3B cells following treatment with 1.0 µg/ml Adriamycin. In all the immuno-fluorescence experiments using flow cytometer (C & F) pink histograms represent cells not treated with Adriamycin and Blue histogram represent Adriamycin treated cells. Black and green histograms represent the autofluorescence and isotype controls respectively. For all the RT-PCR experiments 18S rRNA was used as endogenous control and for western blotting, expression of β-actin was used as loading control. Representative results of three independent experimental sets are shown. In panel B and F ** indicates primer dimers.

Figure 4. Human p53 and Sin3B/HDAC1 complex associates in vivo with HSPA8, MAD1 and CRYZ promoters.

(A) Schematic representation of p53 response element and the amplified promoter region of HSPA8, MAD1 and CRYZ genes. The arrows indicate the position of the respective Forward and Reverse primers used in the ChIP Assays. (B)–(D) ChIP assays in KB (B), HCT116 (C) and p53-null cells (D). Equal amounts of cross-linked chromatin were pre-cleared and incubated with anti-p53 (sc-6243), anti-Sin3B (sc-768X) or anti-HDAC1 (sc-8410) as indicated above each lane. Following DNA precipitation samples were analyzed by PCR using primers specific for HSPA8, MAD1, CRYZ promoters. For negative PCR control, template was replaced with PCR-grade water. ** indicates primer dimers or non-specific amplification. Input corresponds to 10% of the total chromatin used for each immunoprecipitation. Representative figure of four independent experiments.

Figure 5. HSPA8, MAD1 and CRYZ promoters are repressed upon treatment with Adriamycin in p53^+/+ cells.

(A & B) Total RNA was isolated and cDNA was synthesized from p53^+/+ cell lines: KB & HCT116 with or without Adriamycin treatment. Semi-quantitative PCR results (A) indicated an Adriamycin treatment induced transcriptional repression of HSPA8, MAD1 and CRYZ promoters and transcriptional activation of p21. Quantitative RT-PCR (B) re-confirmed the repression of the three genes in both KB and HCT116 cell lines. The values have been plotted with ±SEM calculated from three (n = 3) independent experiments. (C) cDNA was synthesized from total RNA isolated from p53^−/− cell lines: Hep3B and Saos2 cells with or without Adriamycin treatment followed by semi-quantitative PCR. No significant change in transcript levels of HSPA8, MAD1 and CRYZ was observed in Hep3B and Saos2 cells. (D) cDNA was synthesized from total RNA isolated from two non-small cell lung carcinoma cell lines, A549 (p53^+/+) and H1299 (p53^−/−). No significant change in the expression was observed for the three genes in H1299 cells. In contrast a strong repression of HSPA8, MAD1 and CRYZ was observed in A549 cells. For all expression studies 18S rRNA was used as endogenous control. Representative figures of three independent experiments are shown. ** indicates primer dimers.

Figure 6. H3 Lysine 9 residue at the HSPA8 and CRYZ promoters is hyper-methylated upon Adriamycin treatment.

KB cells with or without Adriamycin treatment (1 µg/ml) for 16 hours were harvested. Equal amounts of cross-linked chromatin were pre-cleared and incubated with anti-H3K9Me3 antibody. Following DNA precipitation samples were analyzed by PCR using primers specific for HSPA8, MAD1, CRYZ promoters. For negative PCR control, template was replaced with PCR-grade water. Input corresponds to 10% of the total chromatin used for each immunoprecipitaion. ** indicates primer dimers. Arrows indicate the desired amplicon.