RNF19A-mediated ubiquitination of BARD1 prevents BRCA1/BARD1-dependent homologous recombination

Abstract

BRCA1-BARD1 heterodimers act in multiple steps during homologous recombination (HR) to ensure the prompt repair of DNA double strand breaks. Dysfunction of the BRCA1 pathway enhances the therapeutic efficiency of poly-(ADP-ribose) polymerase inhibitors (PARPi) in cancers, but the molecular mechanisms underlying this sensitization to PARPi are not fully understood. Here, we show that cancer cell sensitivity to PARPi is promoted by the ring between ring fingers (RBR) protein RNF19A. We demonstrate that RNF19A suppresses HR by ubiquitinating BARD1, which leads to dissociation of BRCA1-BARD1 complex and exposure of a nuclear export sequence in BARD1 that is otherwise masked by BRCA1, resulting in the export of BARD1 to the cytoplasm. We provide evidence that high RNF19A expression in breast cancer compromises HR and increases sensitivity to PARPi. We propose that RNF19A modulates the cancer cell response to PARPi by negatively regulating the BRCA1-BARD1 complex and inhibiting HR-mediated DNA repair.

Fig. 1. RNF19A inhibits HR and increases sensitivity to DNA-damaging agents.

a–b Control and RNF19A knockdown U2OS cells were treated with or without IR (2 Gy), γ-H2AX foci before or 1 h, 8 h, and 24 h after IR was detected by immunofluorescence. Nuclei were visualized with DAPI (blue). Representative images are shown in a. Quantification of focus signals per cell (each dot represents a single cell, n = 100) is shown in b. Error bars represent means ± s.d. of three independent experiments. Scale bars, 10 μM. c–e The sensitivity of control (Ctrl) and RNF19A knockdown U2OS cells to Olaparib c cisplatin d and IR e was assessed by colony formation assay. Error bars are means ± s.d. of three independent experiments. f–g Control and RNF19A knockdown Clz3 cells, which contains a dual reporter for HR-tdTomato and NHEJ-GFP, were treated with doxycycline (Dox) for 48 h to turn on I-Scel expression and to induce DSBs. Cells were harvested and subjected to FACS analysis. Error bars are means ± s.d. of three independent experiments. h–i Control (Ctrl) or RNF19A knockdown U2OS cells were treated with IR (1 Gy, 1 h for MDC1, 53BP1, FK2, BRCA1, BARD1; 1 Gy, 5 h for RAD51 and 1 Gy, 3 h for RPA32), and indicated foci were detected by immunofluorescence. Nuclei were visualized with DAPI (blue). Representative images are shown h. Quantification of focus signals per cell (each dot represents a single cell, n = 100) is shown in i. Error bars represent means ± s.d. of three independent experiments. Scale bars, 10 μM. p values are determined by unpaired two-sided t test in b–g and i. Source data are provided as a Source Data file.

Fig. 2. RNF19A interacts with the BARD1 RING domain through its RING2 region.

a HEK293T cells were transfected with Vec or Flag-RNF19A. Cell lysates were subjected to immunoprecipitation with Flag beads and immunoblotted with the indicated antibodies. b–c HEK293T cell lysates were subjected to immunoprecipitation with control IgG, RNF19A (b), or BARD1 (c) antibodies and immunoblotted with the indicated antibodies. d–e Diagram of RNF19A-WT and mutation constructs (d). R1, deletion of the RING1 domain (aa132–179); R2, deletion of the IBR domain (aa199–264); and R3, deletion of the RING2 domain (aa301–332). HEK293T cells transfected with Vec, WT, or deletion mutants of Flag-RNF19A outlined in (d) were subjected to immunoprecipitation as in a. f–g Diagram of BARD1-WT and mutation constructs (f). B1, deletion of the RING domain (aa1–138); B2, deletion of BRCT and parts of Ankyrin domain (aa424–777); and B3, deletion of the BRCT domain (aa568–777). HEK293T cells transfected with Vec, WT, or deletion mutants of Flag-BARD1 outlined in f together with GFP-RN19A were subjected to immunoprecipitation with Flag beads and immunoblotted with the indicated antibodies. h Interaction between RNF19A and BARD1 was analyzed by in vitro GST pull-down assays using purified GST-RNF19A and His-BARD1 proteins. i Control (Ctrl) and RNF19A knockdown HEK293T cells stably expressing Vec, WT, or R1 Flag-RNF19A were subjected to DR-GFP-based HR assay. j Control (Ctrl) and RNF19A knockdown U2OS cells stably expressing Vec, WT, or R1 Flag-RNF19A were subjected to colony formation assay for assessment of the sensitivity to Olaparib. Error bars represent means ± s.d. of four (i) or three (j) independent experiments. p values are determined by unpaired two-sided t test in i and j. Source data are provided as a Source Data file.

Fig. 3. RNF19A functions in end resection through its catalytic activity and in a BARD1-dependent manner.

a–d RNF19A knockdown U2OS cells stably expressing Vec, WT, or C316A(CA) Flag-RNF19A were treated with IR (2 Gy, 3 h for RPA32 and 2 Gy, 5 h for RAD51). Cells were fixed and immunostained as indicated (a and c). Cyclin A was used as a marker of S/G2 phase. Quantification of focus signals per cell (each dot represents a single cell, n = 100 for group Cyclin A-; n = 150 in b; and n = 110 in d for group Cyclin A+) is shown in b and d. Scale bars, 10 μM. Error bars represent means ± s.d. of three independent experiments. e Schematic of ER-AsiSI system for quantification of DNA resection. Restriction enzyme AsiSI is fused to the estrogen receptor (ER) and can be induced to the nucleus and generate DSBs at sequence-specific sites by 4-OHT. The genomic DNA was extracted and quantification of ssDNA generated by resection was measured by qPCR. The primer pairs for DSBs are across BsrGI restriction sites. f RNF19A knockdown ER-AsiSI U2OS cells were reconstituted with Vec, WT, or C316A(CA) Flag-RNF19A and were pretreated with 300 nM 4-OHT for 4 h to induce DSBs. Genomic DNA was extracted and digested or mock digested with BsrGI overnight. DNA-end resection adjacent to indicated sites was measured by qPCR. g HEK293T cells stably expressing Vec or Flag-RNF19A were transfected with control (Ctrl) or BARD1 siRNAs for 48 h. HR efficiency was determined using the HR reporter assay. h U2OS cells stably expressing Vec or Flag-RNF19A were transfected with control (Ctrl) or BARD1 siRNAs were subjected to colony formation assay for assessment of response to Olaparib. i U2OS ER-AsiSI cells stably expressing Vec or Flag-RNF19A transfected with control (Ctrl) or BARD1 siRNAs for 48 h and were pretreated with 300 nM 4-OHT for 4 h before digest and measurement of DNA resection. Error bars represent means ± s.d. of three (f, h, and i) or four (g) independent experiments. p values are determined by unpaired two-sided t test in b, d and f–i. Source data are provided as a Source Data file.

Fig. 4. RNF19A ubiquitinates BARD1 and restrains its interaction with BRCA1.

a Control or RNF19A knockdown HEK293T cells were transfected with indicated plasmids. Cell lysates were subjected to immunoprecipitation with His beads and immunoblotted with the indicated antibodies. b HEK293T cells were transfected with Vec, WT, or C316A(CA) Flag-RNF19A together with HA-BARD1 and His-Ub. Cell lysates were subjected to immunoprecipitation with His beads and immunoblotted with the indicated antibodies. c In vitro ubiquitination assay was performed by incubating purified GST-RNF19A and His-BARD1 (aa 26–327) proteins in the presence of recombinant E1, UbcH7, ubiquitin (Ub), and ATP buffer at 37 °C for 1 h. Samples were boiled and immunoblotted with indicated antibodies. d Indicated HA-tagged ubiquitin was transfected into HEK293T cells. Cell lysates were boiled, immunoprecipitated with HA beads, and immunoblotted with the indicated antibodies. e Control or RNF19A knockdown HEK293T cells were transfected with indicated plasmids. Cell lysates were boiled and immunoprecipitated with BARD1 antibody and immunoblotted with indicated antibodies. f HEK293T cells were transfected with indicated constructs and treated with or without IR (10 Gy, 1 h). Cell lysates were immunoprecipitated with HA beads and immunoblotted with indicated antibodies. g Control or RNF19A knockdown HEK293T cells were transfected with indicated constructs. Cell lysates were immunoprecipitated with HA beads and immunoblotted with indicated antibodies. Source data are provided as a Source Data file.

Fig. 5. Ubiquitination of BARD1 by RNF19A is essential for HR regulation.

a HEK293T cells transfected with WT or NES deletion (ΔNES) mutant of Flag-BARD1 were lysed and pulled down by Ni-NTA and immunoblotted with indicated antibodies. b Indicated plasmids were transfected into HEK293T cells. Cell lysates were pulled down by Ni-NTA and immunoblotted as indicated. c WT or 2KR Flag-BARD1 were transfected into HEK293T cells. Cell lysates were immunoprecipitated with Flag beads and subjected to immunoblot with indicated antibodies. d Indicated plasmids were transfected into HEK293T cells. After 6 h MG132 treating, cell lysates were immunoprecipitated with Flag beads and subjected to immunoblot with indicated antibodies. e–g U2OS cells stably expressing Vec or RNF19A (exo) using plvx2-CMV-RNF19A (no tag) were further stably overexpressed with WT or 2KR Flag-BARD1. Each group of cells was transfected with BARD1 siRNAs for 48 h and treated with IR (2 Gy, 1 h). BRCA1 and Flag-BARD1 foci were detected by immunofluorescence (e). Quantification of focus signals per cell (n = 100) is shown in f: Flag-BARD1 and g BRCA1. Scale bars, 10 μM. h HEK293T cells stably expressing Vec or RNF19A (exo) were further stably overexpressed with WT or 2KR Flag-BARD1 and transfected with BARD1 siRNAs for 48 h and subjected to DR-GFP based HR assay. i The sensitivity of cells shown in (e) to Olaparib was assessed using colony formation assay. j U2OS ER-AsiSI cells stably expressing Vec or GFP-RNF19A were further stably overexpressed with WT or 2KR Flag-BARD1 and transfected with BARD1 siRNAs for 48 h and treated with 4-OHT for 4 h. Cells were digested and measured of DNA resection. k U2OS cells stably expressing Vec or GFP-RNF19A were further stably overexpressed with WT or 2KR Flag-BARD1 and treated with IR (10 Gy, 1 h). Nuclear (Nuc) and cytoplasmic (Cyto) proteins were extracted respectively. Cell lysates were immunoblotted with indicated antibodies. Error bars represent means ± s.d. of three (f, g, i, and j) or four (h) independent experiments. p values are determined by unpaired two-sided t test in f–j. Source data are provided as a Source Data file.

Fig. 6. The role of RNF19A in response to cancer therapy.

a–b Sensitivity of control and RNF19A overexpression MDA-MB-231 (a) and HCC1806 cells to Olaparib was assessed by colony formation assay. c–e MDA-MB-231 cells stably expressing Vec or Flag-RNF19A were subcutaneously injected into nude mice. Tumor weight (c) and volume (d) were measured as indicated. Tumor images were acquired as shown in e. f–h Kaplan–Meier estimates of overall survival of breast cancer patients with different expression levels of BARD1 and RNF19A. The log-rank test was used to compare the survival curves between groups. i Nuclear (Nuc) cytoplasmic (Cyto) BARD1 staining was evaluated by the German semi-quantitative scoring system according to the staining intensity and the proportion was compared between tumor (n = 140, mini-0.25, max-2, centre-1, lower quartile-0.8, upper quartile-1) and adjacent tissues (n = 46, mini-1, max-2, centre-1, lower quartile-1, upper quartile-1.33). j Proportion of nuclear (Nuc)/cytoplasm (Cyto) staining of BARD1 was compared between RNF19A high (n = 79, mini-0.25, max-1, centre-0.8, lower quartile-0.75, upper quartile-1) and low (n = 61, mini-1, max-2, centre-1, lower quartile-1, upper quartile-1.33) expression subgroups. k Representative images of IHC analysis of RNF19A and BARD1 in the serial sections of tumor tissues. Scale bars, 100 μM. Values are means ± s.d. of three (a and b) or five (c and d) independent experiments. p values are determined by unpaired two-sided t test in a–d and i–j. Source data are provided as a Source Data file.

Fig. 7. Schematic model of BARD1 regulation by RNF19A.

RNF19A interacts with BARD1 and promotes its ubiquitination. The ubiquitination of BARD1 unveils the NES of BARD1 located within its RING domain, resulting in the dissociation of BRCA1/BARD1 complex and the export of BARD1 to the cytoplasm.