Functional restoration of BRCA2 protein by secondary BRCA2 mutations in BRCA2-mutated ovarian carcinoma

Abstract

Acquired platinum resistance is a serious problem in the treatment of ovarian carcinomas. However, the mechanism of the drug resistance has not been elucidated. Here, we show functional significance of restoration of BRCA2 protein by secondary BRCA2 mutations in acquired drug resistance of BRCA2-mutated ovarian carcinoma. Three ovarian cancer cell lines (PEO1, PEO4, and PEO6) were derived from a BRCA2 mutation [5193C>G (Y1655X)] carrier with ovarian carcinoma with acquired cisplatin resistance and a secondary BRCA2 mutation [5193C>T (Y1655Y)] that canceled the inherited mutation. PEO1 was BRCA2 deficient and sensitive to cisplatin and a poly(ADP-ribose) polymerase inhibitor, AG14361, whereas PEO4 was resistant. PEO4 and PEO6, derived from ascites at the time of relapse with cisplatin resistance, had the secondary mutation and were BRCA2 proficient. In vitro cisplatin/AG14361 selection of PEO1 led to restoration of BRCA2 due to another secondary BRCA2 mutation. BRCA2 depletion sensitized BRCA2-restored PEO1 clones and PEO4 to cisplatin/AG14361. Thus, restoration of BRCA2 due to secondary BRCA2 mutation is involved in acquired drug resistance of BRCA2-mutated ovarian carcinoma.

Figure 1. PEO1 is a BRCA2-deficient ovarian cancer cell line

(Full-length blots are presented in Supplemental Fig. S5A-B) A. BRCA2 western blotting of 25 ovarian cancer cell lines. B. Clinical course of the patient with an ovarian cancer from which PEO1, PEO4 and PEO6 were derived (reconstituted using published information (14, 15)). Timings of collection of samples used to generate the cell lines were also shown. C. Cisplatin sensitivity assessed by crystal violet assay. 2008 and 2008+FANCF are a cisplatin-sensitive and a cisplatin-resistant control, respectively (13). D. DNA sequences of BRCA2. In PEO1, a nonsense mutation (5193C>G, Y1655X) was observed. In PEO4 and PEO6, a silent mutation on the same base (5193C>T, Y1655Y) was observed. In the non-tumor cells of the patient, a heterozygous mutation (5193C>G) was detected. In tumor cells from the ascites both at the second relapse and at the terminal phase, a secondary BRCA2 mutation (5193C>T) was detected.

Figure 2. Restoration of BRCA2 protein by secondary BRCA2 mutation in cisplatin-selected PEO1 clones

(Full-length blots are presented in Supplemental Fig. S5C) A. BRCA2 western blotting of the 11 clones of PEO1 generated by selecting the cells in the presence of cisplatin. Eight PEO1 clones indicated with a single asterisk restored BRCA2 protein expression. A double asterisk indicates a band presumed to be non-specific. B. DNA sequence of BRCA2 in C4-1. In addition to the original mutation (5193C>G), a secondary mutation (5192A>T) was observed. The same mutation was observed in C4-2, C4-5, C4-6, C4-7, C4-10, C4-12 and C4-14 (data not shown). The mutation was confirmed in cDNAs (data not shown). C. Schematic presentation of BRCA2 proteins encoded by transcripts in PEO1 clones, PEO4 and PEO6. The secondary genetic changes (shown as white arrow heads) cancel the nonsense mutation caused by the original mutation (5193C>G, black arrow heads), and the encoded BRCA2 proteins have intact C-terminal regions containing a single strand DNA (ssDNA) binding domain and nuclear localization signals (NLS). The regions that the BRCA2 antibodies (Ab-1 and Ab-2) recognize are depicted. D. Cisplatin/AG14361 sensitivity of the PEO1 clones. Mean values of at least three independent experiments ± SEM are shown. All of the cisplatin-selected PEO1 clones and PEO4 are cisplatin resistant compared to parental PEO1 (p<0.05, LD₅₀ data were compared by unpaired t test). All of the cisplatin-selected PEO1 clones (except for C4-4) and PEO4 were AG14361 resistant compared to parental PEO1 (p<0.05, LD₅₀ data were compared by unpaired t test).

Figure 3. The restored BRCA2 proteins with secondary BRCA2 mutations are functional

A. Quantitation of homologous recombination induced by I-SceI in VC8-DR-GFP cells transiently transfected with wild-type and mutant forms of FLAG-tagged human BRCA2 cDNA. The proportion of GFP-positive cells for each construct relative to vector control is shown (mean ± SEM). Asterisks (*) indicate significant difference with wild-type BRCA2 cDNA-transfected cells (p<0.05, unpaired t test). B. Ionizing radiation (IR)-induced RAD51 foci formation is restored in the BRCA2-restored PEO1 clones. Indicated cells were irradiated (15 Gy) and fixed 12 hours after IR. Cells were immunostained with RAD51 antibody. Representative pictures of immunostained cells after IR are shown, together with quantification of the cells with at least five RAD51 foci before (−, white bars) and 12 hours after IR (+, grey bars). Mean values of at least three independent experiments ± SEM are shown. Asterisks (*) indicate significant difference with irradiated parental PEO1 cells (p<0.05, unpaired t test). Scale bar = 20 μm. Counterstains for the DNA-specific dye 4’, 6-diamidino-2-phenylindole dihydrochloride (DAPI) are also shown.

Figure 4. BRCA2 restoration is critical for acquired drug resistance

(Full-length blots are presented in Supplemental Fig. S5D) A. BRCA2 western blotting of PEO1, PEO4, C4-2, C4-5, and C4-11 cells treated with indicated siRNA. BRCA2 siRNA #1 was used in this experiment. B. Cisplatin sensitivity assessed by crystal violet assay. PEO4, C4-2 and C4-5 were resistant to cisplatin. BRCA2 siRNA (#1)- treated PEO4, C4-2 and C4-5 cells are cisplatin-sensitive compared to control siRNA-treated cells. (p<0.05, LD₅₀ data were compared by unpaired t test). BRCA2 siRNA had no effect on cisplatin sensitivity of PEO1 and C4-11 cells. (mean ± SEM, n=3) C. AG14361 sensitivity assessed by crystal violet assay. PEO4, C4-2 and C4-5 were resistant to AG14361. BRCA2 siRNA (#1)-treated PEO4, and C4-2 cells were AG14361-sensitive compared to control siRNA-treated cells (p<0.05, LD₅₀ data were compared by unpaired t test). BRCA2 siRNA (#1)-treated C4-5 cells tended to be AG14361-sensitive compared to control siRNA-treated cells, but the difference is not statistically significant (p=0.067). BRCA2 siRNA (#1) had no effect on AG14361 sensitivity of PEO1 and C4-11 cells. (mean ± SEM, n=3)