We analysed 50 probands with a family history of breast and/or ovarian
cancer for germline mutations in the coding region of the BRCA1 candi
date gene, using single-strand conformation polymorphism (SSCP) analys
is on PCR-amplified genomic DNA. A total Of eight putative disease-cau
sing alterations were identified: four of these are frameshifts and tw
o are nonsense mutations. In addition, we found two missense mutations
, one of which changes the final cysteine of the BRCA1 zinc finger mot
if to glycine. These data are consistent with a tumour suppressor mode
l, and support the notion that this candidate gene is in fact BRCA1. T
he heterogeneity of mutations, coupled with the large size of the gene
, indicates that clinical application of BRCA1 mutation testing will b
e technically challenging.