Searching for breast cancer susceptibility genes : a candidate gene approach

Mutations in known predisposition genes account for approximately a third of all multiple-case breast cancer families, suggesting that other breast cancer predisposing genes exist. In order to prioritise candidate breast cancer susceptibility genes, we performed an in silico and literature-based analysis on many genes that interact with BRCA1, BRCA2, or ATM, or play an essential role in the DNA repair pathways. We hypothesized that germline mutations in the BRIP1/BACH1, EIF2S6/Int-6 and ATR genes may account for some of the unexplained multiple-case breast cancer families. The families used in this study were ascertained through the Kathleen Cunningham Foundation Consortium for Research into Familial Breast Cancer (kConFab). Denaturing high performance liquid chromatography (DHPLC) analysis of the coding regions of these three genes was conducted in the index cases of a subset of non-BRCA1/BRCA2 breast cancer families. An additional 253 and 336 index cases from non-BRCA1/BRCA2 breast cancer families were screened for two previously reported breast cancer susceptibility alleles in BRIP1/BACH1 (p.P47A and p.M299I) and for putative pathogenic variants found in ATR, respectively. DHPLC analysis of the BRIP1/BACH1 gene identified one protein truncating variant, as well as four missense variants, a silent change and a variant in the 3' untranslated region. The two previously reported BRIP1/BACH1 mutations were not found in the 253 non-BRCA1/BRCA2 index cases screened. Mutation analysis of EIF3S6/Int-6 identified five novel variants, of which only one was located in the coding region, and another was predicted to affect splicing. Five missense variants, five silent changes, a predicted splicing variant and a large number of intronic variants were found in ATR. Apart from weak support for the ATR c.4846 T>G (p.S1616A) variant, analysis of the truncating, missense and splicing variants of BRIP1/BACH1, EIF3S6/Int-6 and ATR in relatives of the index cases, and in a panel of controls, found no evidence of pathogenicity. There is no evidence that highly-penetrant germline mutations in BRIP1/BACH1, EIF3S6/Int-6 and ATR contribute to familial breast cancer susceptibility, although the ATR c.4846 T>G (p.S1616A) variant may represent a low risk susceptibility allele. Large scale association studies will be necessary to determine whether any of the polymorphisms identified in these genes contribute to breast cancer risk