p53 represses RNA polymerase III transcription by targeting TBP and inhibiting promoter occupancy by TFIIIB

Abstract

The tumor suppressor p53 is a transcription factor that controls cellular growth and proliferation. p53 targets include RNA polymerase (pol) III-dependent genes encoding untranslated RNAs such as tRNA and 5S rRNA. These genes are repressed through interaction of p53 with TFIIIB, a TATA-binding protein (TBP)-containing factor. Although many studies have shown that p53 binds to TBP, the significance of this interaction has remained elusive. Here we demonstrate that the TBP-p53 interaction is of functional importance for regulating RNA pol III-transcribed genes. Unlike RNA pol II-dependent promoter repression, overexpressing TBP can reverse inhibition of tRNA gene transcription by p53. p53 does not disrupt the direct interaction between the TFIIIB subunits TBP and Brf1, but prevents the association of Brf1 complexes with TFIIIC2 and RNA pol III. Using chromatin immunoprecipitation assays, we found that TFIIIB occupancy on tRNA genes markedly decreases following p53 induction, whereas binding of TFIIIC2 to these genes is unaffected. Together our results support the idea that p53 represses RNA pol III transcription through direct interactions with TBP, preventing promoter occupancy by TFIIIB.

Fig. 1. p53 selectively represses RNA pol III transcription of a tRNA gene in vivo. (A) Expression of p53 reduces transcription of a tRNA^Arg gene. H1299 cells were transiently transfected with 2 µg of pArg-maxi and incubated either in the presence (–p53) or absence (+p53) of tetracycline. Total RNA was isolated from the cells 48 h later, and RNase protection assays were performed. A resultant autoradiogram is shown. The arrow designates the 83-nucleotide transcript generated from pArg-maxi, and the brackets denote the half-size tRNA molecules generated by the cleavage of endogenous tRNAs that hybridized with the probe. (B) p53 specifically represses transcription from a tRNA gene promoter. H1299 cells were transfected with pArg-maxi (2 µg), total RNA was isolated and RNase protection assays were performed. For cells transfected with either p-4500/+66hTBP-luc (Foulds and Hawley, 1997) (5 µg), or pG13-luc (el-Diery et al., 1993) (5 µg), total protein was isolated from the cells and luciferase assays were performed. Values shown are the mean ± SEM of at least four independent determinations.

Fig. 2. p53-mediated repression of an RNA pol III promoter is alleviated by overexpression of TBP. (A) Increased TBP expression enhances tRNA gene transcription in cells induced to express p53, but not in non-induced cells. H1299 cells were cotransfected with 2 µg of pArg-maxi and either 0.2–1 µg of the TBP expression construct pLTR-E2TBP or 1 µg of the mutant TBP expression construct pLTR-E2TBP-E284R, where indicated. The percent change in tRNA gene activity was calculated relative to the level of tRNA transcription without p53 induction, which was set as 100%. Immunoblot analysis was carried out using lysates derived from cells transiently transfected with either 1 µg of empty vector or pLTR-E2TBP, and with or without p53 induction as indicated. The endogenous TBP and transiently expressed HA-TBP products are indicated. (B) p53-mediated repression of an RNA pol II-dependent promoter is not relieved by overexpression of TBP. H1299 cells were transiently transfected with 5 µg of the β-3-luc plasmid, containing the human β-3 integrin promoter (Villa-Garcia et al., 1994), with or without p53 induction and, where indicated, cells were cotransfected with pLTR-E2TBP. (C) Brf1 overexpression does not alleviate p53-mediated repression of tRNA gene transcription. H1299 cells were cotransfected with pArg-maxi and 0.1, 0.3 or 1 µg of the Brf1 expression construct containing a double HA tag (Sutcliffe et al., 2000) as indicated. Total RNA was extracted and RNase protection assays were performed. For all panels, the results of at least three independent determinations were quantified, and the values shown are the mean ± SEM. Immunoblot analysis was carried out using lysates derived from cells transiently transfected with either 1 µg of empty vector or the Brf1 expression plasmid. The endogenous Brf1 and transiently expressed HA-Brf1 products are indicated.

Fig. 2. p53-mediated repression of an RNA pol III promoter is alleviated by overexpression of TBP. (A) Increased TBP expression enhances tRNA gene transcription in cells induced to express p53, but not in non-induced cells. H1299 cells were cotransfected with 2 µg of pArg-maxi and either 0.2–1 µg of the TBP expression construct pLTR-E2TBP or 1 µg of the mutant TBP expression construct pLTR-E2TBP-E284R, where indicated. The percent change in tRNA gene activity was calculated relative to the level of tRNA transcription without p53 induction, which was set as 100%. Immunoblot analysis was carried out using lysates derived from cells transiently transfected with either 1 µg of empty vector or pLTR-E2TBP, and with or without p53 induction as indicated. The endogenous TBP and transiently expressed HA-TBP products are indicated. (B) p53-mediated repression of an RNA pol II-dependent promoter is not relieved by overexpression of TBP. H1299 cells were transiently transfected with 5 µg of the β-3-luc plasmid, containing the human β-3 integrin promoter (Villa-Garcia et al., 1994), with or without p53 induction and, where indicated, cells were cotransfected with pLTR-E2TBP. (C) Brf1 overexpression does not alleviate p53-mediated repression of tRNA gene transcription. H1299 cells were cotransfected with pArg-maxi and 0.1, 0.3 or 1 µg of the Brf1 expression construct containing a double HA tag (Sutcliffe et al., 2000) as indicated. Total RNA was extracted and RNase protection assays were performed. For all panels, the results of at least three independent determinations were quantified, and the values shown are the mean ± SEM. Immunoblot analysis was carried out using lysates derived from cells transiently transfected with either 1 µg of empty vector or the Brf1 expression plasmid. The endogenous Brf1 and transiently expressed HA-Brf1 products are indicated.

Fig. 3. p53 does not prevent Brf1 from binding to TBP. (A) Brf1 and TBP co-immunoprecipitate both in the presence and absence of p53 induction. Protein lysates were prepared from H1299 cells that were induced for p53 expression for 0, 2 or 6 days, as designated. Immunoprecipitation assays were carried out using anti-TBP or anti-Brf1 antibodies, as indicated, and 400 µg of protein lysate. The lysates were then subjected to SDS–PAGE, and immunoblot analysis using either anti-TBP or anti-Brf1 antibodies was performed. Mock controls designate assays in which no antibodies were used. Input designates cell lysates (40 µg) that were directly subjected to immunoblot analysis. (B) p53 does not prevent Brf1 from binding to TBP. Anti-TBP antibody was used to immunoprecipitate reticulocyte lysate containing in vitro translated TBP and/or in vitro translated Brf1, as indicated. His-p53 wt (0.6 µg) or His-p53 (98–303) (0.6 µg) was added where indicated. Bound material was separated by SDS–PAGE and visualized by autoradiography. Lane 1 shows 10% input of [³⁵S]TBP and [³⁵S]Brf1. Quantitation of three experiments is shown; values represent the mean and standard deviation of the amount of [³⁵S]Brf1 co-immunoprecipitated, relative to input [³⁵S]Brf1. (C) p53 associates with TBP–Brf1 complexes. Immunoprecipitation assays were performed as described in (A) using 50 µg of protein. Immunoblot analysis was conducted using p53 antibodies. Input designates cell lysates (10 µg) that were directly subjected to immunoblot analysis. The p53 band was determined by using recombinant p53 as a standard. The CRM bands represent non-specific cross-reacting material that is immunoprecipitated with IgG. (D) p53 associates with in vitro translated TBP–Brf1 complexes. Reticulocyte lysate containing ³⁵S-labeled in vitro translated TBP and Brf1 was mixed with recombinant wild-type p53 (0.6 µg) and then immunoprecipitated with antibody against p53 or control antibody against an HA tag. Bound material was separated by SDS–PAGE and visualized by autoradiography. Quantitation of three experiments is shown; values represent the mean and standard deviation of the amounts of [³⁵S]Brf1 and [³⁵S]TBP co-immunoprecipitated with p53 or HA control antibodies.

Fig. 3. p53 does not prevent Brf1 from binding to TBP. (A) Brf1 and TBP co-immunoprecipitate both in the presence and absence of p53 induction. Protein lysates were prepared from H1299 cells that were induced for p53 expression for 0, 2 or 6 days, as designated. Immunoprecipitation assays were carried out using anti-TBP or anti-Brf1 antibodies, as indicated, and 400 µg of protein lysate. The lysates were then subjected to SDS–PAGE, and immunoblot analysis using either anti-TBP or anti-Brf1 antibodies was performed. Mock controls designate assays in which no antibodies were used. Input designates cell lysates (40 µg) that were directly subjected to immunoblot analysis. (B) p53 does not prevent Brf1 from binding to TBP. Anti-TBP antibody was used to immunoprecipitate reticulocyte lysate containing in vitro translated TBP and/or in vitro translated Brf1, as indicated. His-p53 wt (0.6 µg) or His-p53 (98–303) (0.6 µg) was added where indicated. Bound material was separated by SDS–PAGE and visualized by autoradiography. Lane 1 shows 10% input of [³⁵S]TBP and [³⁵S]Brf1. Quantitation of three experiments is shown; values represent the mean and standard deviation of the amount of [³⁵S]Brf1 co-immunoprecipitated, relative to input [³⁵S]Brf1. (C) p53 associates with TBP–Brf1 complexes. Immunoprecipitation assays were performed as described in (A) using 50 µg of protein. Immunoblot analysis was conducted using p53 antibodies. Input designates cell lysates (10 µg) that were directly subjected to immunoblot analysis. The p53 band was determined by using recombinant p53 as a standard. The CRM bands represent non-specific cross-reacting material that is immunoprecipitated with IgG. (D) p53 associates with in vitro translated TBP–Brf1 complexes. Reticulocyte lysate containing ³⁵S-labeled in vitro translated TBP and Brf1 was mixed with recombinant wild-type p53 (0.6 µg) and then immunoprecipitated with antibody against p53 or control antibody against an HA tag. Bound material was separated by SDS–PAGE and visualized by autoradiography. Quantitation of three experiments is shown; values represent the mean and standard deviation of the amounts of [³⁵S]Brf1 and [³⁵S]TBP co-immunoprecipitated with p53 or HA control antibodies.

Fig. 3. p53 does not prevent Brf1 from binding to TBP. (A) Brf1 and TBP co-immunoprecipitate both in the presence and absence of p53 induction. Protein lysates were prepared from H1299 cells that were induced for p53 expression for 0, 2 or 6 days, as designated. Immunoprecipitation assays were carried out using anti-TBP or anti-Brf1 antibodies, as indicated, and 400 µg of protein lysate. The lysates were then subjected to SDS–PAGE, and immunoblot analysis using either anti-TBP or anti-Brf1 antibodies was performed. Mock controls designate assays in which no antibodies were used. Input designates cell lysates (40 µg) that were directly subjected to immunoblot analysis. (B) p53 does not prevent Brf1 from binding to TBP. Anti-TBP antibody was used to immunoprecipitate reticulocyte lysate containing in vitro translated TBP and/or in vitro translated Brf1, as indicated. His-p53 wt (0.6 µg) or His-p53 (98–303) (0.6 µg) was added where indicated. Bound material was separated by SDS–PAGE and visualized by autoradiography. Lane 1 shows 10% input of [³⁵S]TBP and [³⁵S]Brf1. Quantitation of three experiments is shown; values represent the mean and standard deviation of the amount of [³⁵S]Brf1 co-immunoprecipitated, relative to input [³⁵S]Brf1. (C) p53 associates with TBP–Brf1 complexes. Immunoprecipitation assays were performed as described in (A) using 50 µg of protein. Immunoblot analysis was conducted using p53 antibodies. Input designates cell lysates (10 µg) that were directly subjected to immunoblot analysis. The p53 band was determined by using recombinant p53 as a standard. The CRM bands represent non-specific cross-reacting material that is immunoprecipitated with IgG. (D) p53 associates with in vitro translated TBP–Brf1 complexes. Reticulocyte lysate containing ³⁵S-labeled in vitro translated TBP and Brf1 was mixed with recombinant wild-type p53 (0.6 µg) and then immunoprecipitated with antibody against p53 or control antibody against an HA tag. Bound material was separated by SDS–PAGE and visualized by autoradiography. Quantitation of three experiments is shown; values represent the mean and standard deviation of the amounts of [³⁵S]Brf1 and [³⁵S]TBP co-immunoprecipitated with p53 or HA control antibodies.

Fig. 4. p53 disrupts binding of Brf1 to TFIIIC2 and RNA pol III. (A) p53 prevents Brf1 from binding to TFIIIC2. Anti-TFIIIC110 antibody was used to immunoprecipitate in vitro translated Brf1 from reticulocyte lysates. Where indicated, His-p53 wt (1 µg) and His-p53 (98–303) (1 µg) were pre-incubated with 100 µg of HeLa nuclear extract. Bound material was subjected to SDS–PAGE and visualized by autoradiography. Input shows 20% [³⁵S]Brf1. Quantitation of three experiments is shown. (B) p53 prevents Brf1 from binding to RNA pol III. Anti-pol III antibody was used to immunoprecipitate reticulocyte lysate containing in vitro translated Brf1 as described in (A).

Fig. 4. p53 disrupts binding of Brf1 to TFIIIC2 and RNA pol III. (A) p53 prevents Brf1 from binding to TFIIIC2. Anti-TFIIIC110 antibody was used to immunoprecipitate in vitro translated Brf1 from reticulocyte lysates. Where indicated, His-p53 wt (1 µg) and His-p53 (98–303) (1 µg) were pre-incubated with 100 µg of HeLa nuclear extract. Bound material was subjected to SDS–PAGE and visualized by autoradiography. Input shows 20% [³⁵S]Brf1. Quantitation of three experiments is shown. (B) p53 prevents Brf1 from binding to RNA pol III. Anti-pol III antibody was used to immunoprecipitate reticulocyte lysate containing in vitro translated Brf1 as described in (A).

Fig. 5. Wild-type p53 can specifically block recruitment of TFIIIB, but not TFIIIC2, to an immobilized tRNA^Leu gene. Immobilized template assays were performed using beads alone or beads carrying a tRNA^Leu gene fragment (McLaren and Goddard, 1986) and 200 µg of protein isolated from CHO cells stably transfected with HA-tagged Brf1. Assays were carried out by mixing protein lysates with no p53, or 2 or 4 µg of wild-type or mutant p53 (where designated), before adding the reaction to the immobilized template. Amounts of bound TBP, Brf1 and TFIIIC110 were determined by immunoblot analysis.

Fig. 6. Association of the TFIIIB subunits TBP, Brf1 and Bdp1 with a chromosomal tRNA gene is reduced specifically by the induction of p53 in H1299 cells. (A) ChIP analysis of TFIIIB and TFIIIC2 subunits at a tRNA^Arg gene. ChIP assays were conducted as described in Material and methods. The ‘input’ designates PCR product derived from chromatin that was not subjected to immunoprecipitation. (B) Quantitation of ChIP assays. The experimental results for 10 independent ChIP experiments for each antibody (an example is shown in panel A) were quantified. For each antibody, the relative percent occupancy of the subunit was calculated by determining the amount of PCR product derived from chromatin of non-induced cells and setting this value to 100% for each independent experiment. Values shown are the mean ± SEM. (C) Increased p53 expression in H1299 cells does not alter the cellular amounts of TFIIIB or TFIIIC2 subunits. Immunoblot analysis was conducted using 100 µg of protein lysates and antibodies directed against each of the proteins designated.

Fig. 7. Induction of endogenous p53 in HeLa cells by genotoxic stress represses tRNA expression and specifically reduces the amount of TBP associated with chromosomal tRNA genes. (A) p53 levels increase following MMS treatment of HeLa cells. Immunoblot analysis of protein extracted from HeLa cells that are either untreated or treated for 2 h with 0.04% MMS. Blots were probed for p53 and actin, as indicated. (B) RT–PCR analysis of tRNA^Leu, tRNA^Tyr and ARPP P0 mRNA expression. HeLa cells were untreated or treated with 0.04% or 0.08% MMS as indicated. (C) ChIP assay showing levels of TBP and p53 bound to tRNA^Leu, tRNA^Tyr, p21 and TFIIIC220 genes as designated. Control indicates chromatin incubated with beads without antibody. HeLa cells were untreated or treated for 2 h with 0.04% MMS as designated. (D) Quantitation of ChIP assay. Three independent sets of ChIP experiments were performed, and the resultant PCR products were quantified by setting the input value to 100%. Values shown are the mean and standard deviation for each PCR product.

Fig. 7. Induction of endogenous p53 in HeLa cells by genotoxic stress represses tRNA expression and specifically reduces the amount of TBP associated with chromosomal tRNA genes. (A) p53 levels increase following MMS treatment of HeLa cells. Immunoblot analysis of protein extracted from HeLa cells that are either untreated or treated for 2 h with 0.04% MMS. Blots were probed for p53 and actin, as indicated. (B) RT–PCR analysis of tRNA^Leu, tRNA^Tyr and ARPP P0 mRNA expression. HeLa cells were untreated or treated with 0.04% or 0.08% MMS as indicated. (C) ChIP assay showing levels of TBP and p53 bound to tRNA^Leu, tRNA^Tyr, p21 and TFIIIC220 genes as designated. Control indicates chromatin incubated with beads without antibody. HeLa cells were untreated or treated for 2 h with 0.04% MMS as designated. (D) Quantitation of ChIP assay. Three independent sets of ChIP experiments were performed, and the resultant PCR products were quantified by setting the input value to 100%. Values shown are the mean and standard deviation for each PCR product.

Fig. 8. TFIIIB occupancy and RNA pol III transcription are elevated specifically in primary MEFs from p53 knockout mice relative to wild-type controls. (A) ChIP assay showing levels of TBP, TFIIIC2 and RNA pol III bound to tRNA^Leu, tRNA^Tyr and TFIIIC220 genes in primary MEFs from matched p53^+/+ and p53^–/– mice, as designated. Control indicates chromatin incubated with beads without antibody. (B) Quantitation of ChIP results. The value for p53^+/+ cells was set at 100%. Values shown are the mean and standard deviation for each PCR product. (C) Increased tRNA gene occupancy of TBP and RNA pol III in p53^–/– MEFs does not reflect changes in the levels of these proteins. Immunoblot analysis of protein extracted from p53^+/+ and p53^–/– MEFs.