The Identification of BRCA1 and BRCA2 Mutation Carriers Using Functional Genomic Assays

Thumbnail Image
Michel, Claire S.
BRCA1 , BRCA2 , Hereditary Breast Cancer , Microarrays , Cancer Screening , Mutation Carriers
An estimated 5-10% of breast cancers are hereditary in nature and are due to the presence of a mutation in a breast cancer predisposition gene; approximately half of these cases possess a mutation in BRCA1 or BRCA2. Many BRCA1/BRCA2 mutations result in a truncated protein and hence are unequivocally disease-causing. However another class of mutations, the Variants of Unknown Significance (VUS), are more problematic as the effect of these mutations on protein function is unclear. The inability to classify these mutations as disease causing generates significant problems in risk evaluation, counseling and preventive care. Accordingly we sought to determine whether carriers of either a BRCA1 or BRCA2 mutation could be identified from non-carriers based on the gene expression patterns of non-cancerous cells. EBV-transformed lymphoblastoid cell lines established from BRCA1/BRCA2 mutation carriers and normal individuals were obtained through the NIH Breast Cancer Family Registries. Cell lines were mock-irradiated or treated with ionizing radiation (2 Gy). Following a recovery period of 6 hours total RNA was extracted and whole genome gene expression profiling was carried out. Molecular classifiers comparing the baseline expression profiles and the radiation-dependent expression profiles of BRCA1/BRCA2 mutation carriers to control individuals were created using a Support Vector Machine (SVM) coupled with a recursive feature removal (RFR) algorithm. Our results suggest that cell populations derived from BRCA1/BRCA2 mutation carriers display unique expression phenotypes from those of control individuals in both the basal and radiation-induced cases. In the task of classification using baseline expression, the BRCA1-classifier correctly classified 15/18 test samples using feature selection based on the training set only, while feature selection using the entire dataset (AD) improved classification to 16/18 samples. The BRCA2-baseline classifier correctly classified 13/17 and 14/17 (AD) samples, respectively. In the task of radiation-dependent classification, the BRCA1-IR classifier correctly classified 12/18 and 16/18 (AD) test samples respectively while the BRCA2-IR classifier correctly classified 13/17 and 16/17 (AD) test samples respectively. These results suggest the possibility of development of this assay into a novel hereditary breast cancer screening diagnostic able to accurately identify the presence of BRCA1 or BRCA2 mutations via a functional assay thereby improving patient outcomes.
External DOI