BRCA genes: conserved regions and the potential effect of missense changes
Half of coding single-nucleotide polymorphisms in the human genome result in missense changes. Many are rare, weakly penetrant, or occur in genes for which functional assays are not available. Comparative analysis of protein sequences from multiple species is a practical method to predict their effects. Protein-truncating mutations in either of the breast cancer susceptibility genes, BRCA1 or BRCA2, confer a 50--80% lifetime risk of breast cancer. While 34% of BRCA mutations reported in the Breast Cancer Information Core are missense changes, only 8% are known to be dissease-associated. For BRCA1 30% of reported missense changes occur in exon 11, but none has been clearly identified as disease-associated. Initially, in a comparison of sequences from 57 eutherian mammal species, Melissa Fleming and I found seven "highly conserved regions" between amino acids 282 and 1103 and identified 38 missense changes as likely to disrupt gene function. In a second analysis, I hypothesized that using 37 ancestral sequences derived from the 57 GenBank sequences and including eight marsupial sequences would allow me to identify regions unique to mammals and refine my predictions of disease-associated missense changes. I identified 13 conserved regions, three of which appear to be unique to mammals, and 21 likely disease-associated missense changes. Recently the Ostrander and Malone Labs screened BRCA1 and BRCA2 from 1,628 women with breast cancer and 674 controls as part of the Women's Contraceptive and Reproductive Experiences (CARE) study. We identified 80 BRCA1 and 127 BRCA2 missense changes. Based on alignments of full-length sequences from five eutherian mammals, I identified a proline to arginine change at BRCA2 amino acid 655 (P655R) that is likely to be disease-associated. P655R occurred in six out of 84 Jewish women with breast cancer, in zero out of 36 controls, and at a population frequency of 2% in 315 Jewish individuals. A 500-Mb, 12-SNP haplotype surrounding P655R appears to be unique to the Jewish population. Current studies use a combination of functional, structural, and evolutionary methods to determine the likelihood of missense changes in BRCA1 being associated with breast cancer, but the status of most missense changes still remains unclear.