Identification of rare gene variants in South African breast cancer families through next generation sequencing

Abstract

Breast cancer (BC) has become the leading cancer amongst women in South Africa. The overall life time risk for developing this disease is one in 12 (National Cancer registry, 2000- 2011). A strong family history (≥3 affected) is an important factor for inherited predisposition to BC that accounts for approximately 10% of cases worldwide. Mutations in several high- and moderate risk breast cancer genes have been associated with familial BC and includes BRCA1, BRCA2, TP53, PALB2, and CHEK2. Individuals that carry germline mutations in BRCA1 and BRCA2 possess an 80% lifetime risk for BC. Mutations in BRCA1 and BRCA2 are responsible for 29% and 25% of familial BC worldwide. In South Africa BRCA1 mutations account for 19% and BRCA2 for 47% of familial breast cancer. Mutations associated with a moderate risk for BC account for ~1% of cases. This data suggests that ~30% of South African BC families are not characterised by pathogenic mutations in known breast/ovarian (BC/OVC) genes. The purpose of the present study was to identify gene variants that may predispose to breast cancer. Next generation sequencing was performed to investigate the germline DNA of highrisk BC/OVC families that have previously tested negative for premature truncating mutations in BRCA1/2, PALB2 and RAD51C. Paired-end whole exome sequencing was performed with nine index cases, selected from six families with a strong background for BC/OVC. This resulted in the discovery of an average of 26 000 coding variants in index cases. Gene prioritisation strategies were incorporated to filter all exome variants and identify high-priority genes for further analysis. After sequence verification, three high-priority genes were selected for further analysis. The three genes coded for; a novel putative tumour suppressor (TCHP) that is pro-apoptotic; the XPF-endonuclease homolog, EME2; and a POLQ like helicase enzyme (HELQ). Prioritised genes were screened in a total of 61 high-risk families and cohorts of patients with BC or OVC without a family history for their disease. Two potentially damaging variants (stop-gain & inframe amino acid deletion) were identified in TCHP, four (frameshift, nonsense & two in-frame deletions) in EME2 and one frameshift mutation in HELQ in high-risk families and cases that were without a family history for BC/OVC. The analyses performed in the last section of this project was aimed at identifying other potential genes of interest by making use of a list of 516 well recognised and putative DNA repair genes. Through this approach, one additional truncating mutation in POLN (p.Q837SfsX7) was highlighted as a potential gene of interest for future investigation. Despite the key roles that the high-priority genes play in their respective processes, the present study could not verify that the potential loss of function variants discovered make an appreciable contribution towards BC/OVC susceptibility in our setting. Further investigation is necessary to validate their involvement in breast/ovarian cancer predisposition.