PBRM1 acts as a p53 lysine-acetylation reader to suppress renal tumor growth

Abstract

p53 acetylation is indispensable for its transcriptional activity and tumor suppressive function. However, the identity of reader protein(s) for p53 acetylation remains elusive. PBRM1, the second most highly mutated tumor suppressor gene in kidney cancer, encodes PBRM1. Here, we identify PBRM1 as a reader for p53 acetylation on lysine 382 (K382Ac) through its bromodomain 4 (BD4). Notably, mutations on key residues of BD4 disrupt recognition of p53 K382Ac. The mutation in BD4 also reduces p53 binding to promoters of target genes such as CDKN1A (p21). Consequently, the PBRM1 BD4 mutant fails to fully support p53 transcriptional activity and is defective as a tumor suppressor. We also find that expressions of PBRM1 and p21 correlate with each other in human kidney cancer samples. Our findings uncover a tumor suppressive mechanism of PBRM1 in kidney cancer and provide a mechanistic insight into the crosstalk between p53 and SWI/SNF complexes.

Fig. 1. PBRM1 and p53 display a physical association that is enhanced after DNA damage.

a U2OS cells (left) and HEK293T cells (right) were transfected with vector or Flag-PBRM1 and harvested for immunoprecipitation with Flag-M2 beads and elution with 3X Flag peptide. Inputs and eluates were analyzed by immunoblots. b, c H1299 cells were transfected with Flag-PBRM1 and Myc-p53 and treated with etoposide (50 μM, b) or bleomycin (10 μg/ml, c) for the indicated times. Lysates were subjected to immunoprecipitation with Flag-M2 beads. Inputs and eluates were analyzed by immunoblots. d HEK293 cells were treated with vehicle (DMSO) or etoposide (50 μM) for 8 h and harvested for immunoprecipitation with control IgG and p53 antibodies. The bound PBRM1 and p53 were examined by immunoblots. e HCT116 and HCT116 p53−/− cells were treated with DMSO or 50 μM etoposide for 24 h. Lysates were immunoprecipitated with p53 antibody. The bound PBRM1 and p53 were examined by immunoblots. Source data are provided as a Source Data file.

Fig. 2. Acetylation on K382 of p53 promotes PBRM1 binding.

a Schematic depiction of the functional domains of p53 and truncated constructs. FL full-length, TAD transactivation domain, PRD proline-rich domain, DBD DNA-binding domain, TD tetramerization domain, CTD C-terminal domain. b HEK293T cells were transfected with vectors, Flag-PBRM1, Myc-p53 and Myc-p53 truncated constructs. After immunoprecipitation with Flag-M2 beads, the inputs and eluates were analyzed by immunoblots. c Vectors, Flag-PBRM1, Myc-p53 and HA-Tip60, HA-PCAF, HA-CBP or HA-p300 were transfected into HEK293T cells as indicated. Lysates were used for anti-Flag immunoprecipitation and 3X Flag peptide elution. The inputs and eluates were analyzed by immunoblots. d HEK293T cells were transfected with Flag-PBRM1, Myc-p53 and HA-p300. The amount of HA-p300 was titrated as indicated. Lysates were used in immunoprecipitations with Flag-M2 beads, and inputs and eluates were analyzed by immunoblots. e Vectors, Flag-PBRM1, HA-p300 and Myc-p53 (WT) or Myc-p53 6KR mutant were transfected into HEK293T cells as indicated. Lysates were used for anti-Flag immunoprecipitation and 3X Flag peptide elution. Inputs and eluates were examined by immunoblots. 6KR: lysines mutated to arginines on p53 at K370, K372, K373, K381, K382, and K386. f, g Biotinylated p53 peptides with lysine acetylation at the indicated sites were incubated with lysates from H1299 cells (f) or H1299 cells transfected with full-length PBRM1 or the PBRM1 bromodomains (Flag-PBRM1 FL or Flag-PBRM1-BD1–6, respectively, g). The peptides were pulled down with streptavidin beads and the associated proteins were analyzed by immunoblots. 6KAc: lysines acetylated on p53 at K370, K372, K373, K381, K382, and K386. h Vectors, Flag-PBRM1, HA-p300 and Myc-p53 (WT) or Myc-p53 K382R mutant were transfected into HEK293T cells as indicated. Lysates were used for anti-Flag immunoprecipitation and 3X Flag peptide elution. Inputs and eluates were examined by immunoblots. Source data are provided as a Source Data file.

Fig. 3. Bromodomain 4 of PBRM1 is required for recognition of acetylated K382 on p53.

a Schematic depiction of the functional domains of PBRM1 and truncated constructs. WT: wild-type, BAH: bromo-adjacent homology domain, HMG: high-mobility group domain. b, c HEK293T cells were transfected with vectors, Flag-WT PBRM1 and Flag-PBRM1 truncated constructs (b), and each individual or all the PBRM1 bromodomains (BD1–6, c). After immunoprecipitation with Flag-M2 beads, the inputs and eluates were analyzed by immunoblots. d Schematic of key amino acid residues of binding pockets in PBRM1 bromodomains 4 and 5. The critical YN residues are underlined and mutations are red. e–g H1299 PBRM1 KO cells were transfected with Flag-PBRM1 bromodomains 4 and 5 (Flag-BD45) containing mutations (BD4* or BD5*) that abolish acetyl-lysine recognition in each domain (e), Flag-BD45 containing tumor-derived mutations (f) or full-length PBRM1 containing the BD4* mutation (g). Lysates were incubated with biotinylated p53 peptides with lysine acetylation at the indicated sites. The peptides were pulled down with streptavidin beads and the associated proteins were immunoblotted. h Vectors, Myc-p53, HA-p300, and Flag-PBRM1 with or without the BD4* mutation were transfected into HEK293T PBRM1 knockout cells. The lysates were subjected to anti-Flag immunoprecipitation and 3X Flag peptide elution. Inputs and eluates were analyzed in immunoblots. Source data are provided as a Source Data file.

Fig. 4. PBRM1 is required for full p53 transcriptional activity on a subset of its targets.

a p53 was transfected into H1299 parental cells and five PBRM1 knockout clones, and lysates were immunoblotted. b H1299 parental and PBRM1-KO#1 cells were transfected with increasing amounts of p53 and lysates were immunoblotted. c H1299 parental and PBRM1-KO#1 were transfected with vector or p53, and RNA was extracted and reverse-transcribed. After PCR array, significantly altered p53 target genes were further validated by qPCR. The expression of significantly altered genes is shown as mean ± SEM. p-values were calculated using the paired two-tailed Student’s t-test. d–f Increasing amounts of p53 were transfected into H1299 PBRM1-KO#1 with or without Flag-PBRM1 re-expression and lysates were analyzed by immunoblots. Cells with comparable p53 expression were harvested for RT-qPCR (e) or ChIP followed by qPCR (f). Data are shown as mean ± SEM. g H1299 PBRM1-KO#1 cells stably expressing GFP, wild-type or BD4* mutant PBRM1 were transfected with increasing amounts of p53. Lysates were analyzed by immunoblots. Source data are provided as a Source Data file.

Fig. 5. K382Ac recognition by PBRM1 is critical for p53 function in ccRCC cells.

a, b ACHN (a) or Caki-1 (b) control and PBRM1 shRNA knockdown cells were treated with 50 μM etoposide for the indicated time points. Lysates were analyzed by immunoblots. c RCC4 cells with or without p53 KO were transfected with vector or Flag-PBRM1 at the indicated amounts. Lysates were analyzed by immunoblots. d ACHN control and PBRM1 shRNA #1 knockdown cells were treated with 10 μM Nutlin-3a for the indicated time points. Lysates were analyzed by immunoblots. e RCC4 cells stably expressing GFP, wild-type PBRM1, or BD4* mutant PBRM1 were treated with 100 μM etoposide for the indicated time points. Lysates were analyzed by immunoblots. f RCC4 cells stably expressing GFP, wild-type or indicated PBRM1 mutants were treated with 100 μM etoposide for the indicated times, and lysates were examined via immunoblots. g, h RCC4 cells stably expressing GFP, wild-type or BD4* mutant PBRM1 were treated with 100 μM etoposide for 24 h. Cells were harvested for RT-qPCR (g) or ChIP followed by qPCR (h). Data are shown as mean ± SEM. p-values were calculated using the paired two-tailed Student’s t-test. i, j ACHN control and PBRM1 shRNA knockdown cells were treated with vehicle or 50 μM etoposide for cell cycle analysis. Representative data is shown in i. Quantification of cell cycle phase is shown in j. Source data are provided as a Source Data file.

Fig. 6. Mutation of BD4 abolishes the tumor suppressive function of PBRM1.

a, b GFP, wild-type or BD4* mutant PBRM1 were stably expressed in Ren-01 PBRM1 KO cells (combination of three clones) (a) or PBRM1-null SLR24 cells (b). Cells were treated with 50 μM etoposide for the indicated times, and lysates were analyzed via immunoblots with indicated antibodies. The band intensity of indicated protein was measured with Bio-rad Image Lab 4.1, and the relative ratios were calculated over the signal intensity of Vinculin in the corresponding lanes. c–e. Nude mice xenograft analyses were performed with Ren-01 PBRM1 KO cells expressing GFP (left flank) or PBRM1 (right flank). Representative photographs of a mouse (c, top) and tumors (c, bottom). Tumors were excised and weighed, and data are presented as mean ± SEM (d). p-values were calculated using the paired two-tailed Student’s t-test. Tumors were stained for PBRM1 and p21 expression via IHC (e). Scale bar: 200 μm. f–h Nude mice xenograft analyses were performed with Ren-01 PBRM1 KO cells expressing wild-type PBRM1 (left flank) or the BD4* mutant PBRM1 (right flank). Tumors were excised and weighed, and data are presented as mean ± SEM (g). p-values were calculated using the paired two-tailed Student’s t-test. Representative photographs of a mouse (f, top) and tumors (f, bottom). Tumors were stained for PBRM1 and p21 expression via IHC analysis (h). Scale bar: 200 μm. Source data are provided as a Source Data file.

Fig. 7. PBRM1 loss and p21 loss positively correlate in human ccRCC tumors.

a 160 ccRCC tumors (40 tumors per stage) were used to generate a tissue microarray (TMA). Four foci from different regions of each tumor were selected. PBRM1 expression was analyzed and scored in our previous report^, (a, bottom). p21 expression was analyzed and scored in the tumors (a, top) by a blinded pathologist. Scale bar: 100 μm. b p-values of correlations between PBRM1 loss and p21 loss in human ccRCC samples. c A model depicting how PBRM1 may regulate tumor growth in ccRCC through recognition of p53 K382Ac. Source data are provided as a Source Data file.