BRCA 1 and 2--A Genetic Link to Familial Breast and Ovarian Cancer

BRCA1

In 1990, the first breast cancer susceptibility gene, BRCA1, was localized using linkage analysis to chromosome 17q21.^[12] The following year, it was demonstrated that several ovarian cancer families exhibited linkage to the same locus.^[13] In 1994, Miki and associates^[8] isolated the gene using positional cloning methods. They described a large gene consisting of 22 coding exons distributed over approximately 100,000 base pairs of genomic DNA (Fig. 1). The gene encodes a protein of 1863 amino acids. A 126 base-pair RING finger domain was observed, located from exons 2 to 5. Sequence homology of this domain to a group of genes encoding proteins with DNA binding functions suggested that the BRCA1 protein might play a role in regulation of transcription.^[8] Using yeast and mammalian 2-hybrid screening, Wu and colleagues^[14] recently identified a novel protein that interacts in vivo with the N-terminal region of the BRCA1 protein. This novel protein, BRCA1-associated RING domain (BARD1), was shown to contain an N-terminal RING motif, a C-terminal sequence homology with BRCA1, and 3 tandem ankyrin repeats. Ankyrin repeats are 33 amino acid motifs that have been described in several regulatory proteins.^[15] Furthermore, the BARD/BRCA1 interaction was shown to be disrupted by missense mutations in BRCA1, implying that this protein interaction is an essential component of the tumor suppression role of BRCA1.

Figure 1. Diagrammatic representation of 3-exon gene. Exons represent coding sequence, introns non-coding sequences spliced out during transcription. Translation begins at start codon ATG (here in exon 1), and ends at stop codon TAG, TAA, or TGA. Translation occurs in 5' to 3' direction.

Carriers of germline mutations in BRCA1 have a profound predisposition to the development of breast and ovarian cancer. For female carriers, the lifetime risk of breast cancer exceeds 80%, and for ovarian cancer it approaches 60%.^[16,17,18] The cumulative lifetime risk of developing either breast and/or ovarian cancer, therefore, approaches 90%-100%. In comparison, women in the general population have approximately a 10% lifetime risk for breast cancer and a 1%-2% lifetime risk for ovarian cancer. Male carriers have a small increased risk of developing prostate cancer and colon cancer.^[19] Despite the high penetrance of the mutant gene, not all carriers develop cancer. Hormonal, environmental, reproductive, and other genetic factors may influence penetrance. Many studies are currently underway to elucidate influential genetic factors. Recently, the first genetic modification of penetrance has been reported.^[20] Rare alleles of the VNTR polymorphism downstream of the HRAS gene have been associated with increased risk for ovarian cancer in BRCA1 mutation carriers.^[20] The physiologic mechanism of this interaction is unknown.

In excess of 100 discrete germline mutations in BRCA1 have been identified in over 300 carriers studied.^[21] These mutations are distributed throughout the gene, and the majority (approximately 89%) are insertions, deletions, or nonsense mutations.^[21] (A nonsense mutation occurs when a change in base-pair sequence creates a stop codon, leading to the production of a truncated protein product.) Mutations in the 5' end of the gene have been correlated with a larger increase in ovarian cancer risk than mutations in the 3' end.^[22] Two common changes (185delAG and 5382insC in exons 2 and 20, respectively) account for approximately 19% of the mutations.^[21] The 185delAG mutation is seen in approximately 1% of the Ashkenazi Jewish population.^[23]

It was predicted that somatic mutational inactivation of BRCA1 would account for a significant number of cases of sporadic, or nonfamilial, breast and ovarian cancers,^[8] as is the case with other familial cancer predisposition genes. Somatic mutations are frequently seen, for example, in the APC gene, which produces a syndrome of adenomatous polyposis coli in carriers of germline mutations.^[24] Indeed, loss of heterozygosity (LOH) studies in the BRCA1 region have demonstrated loss in 30% to 70% of sporadic breast and ovarian cancers.^[25,26] Despite an extensive search by several groups, however, only 5 somatic BRCA1 mutations have been identified in patients with ovarian cancer, and none have been reported in patients with breast cancer.^{[27,28,29,30,31]}

However, the lack of somatic BRCA1 mutations in sporadic breast cancer does not necessarily exclude a role for the gene in sporadic disease. Thompson and coworkers^[32] demonstrated a 10-fold reduction in BRCA1 mRNA levels in breast tumor specimens relative to normal breast tissue. Furthermore, in vitro studies performed by the group using normal mammary epithelial cells and MCF7 breast cancer cells linked decreased BRCA1 expression, mediated by antisense oligonucleotides, with increased proliferative rate. More recently, Holt and associates^[33] demonstrated growth inhibition of breast and ovarian cancer cell lines following retroviral transfer of wild-type BRCA1 gene. In support of the earlier observation by Gayther and colleagues,^[22] transfection with 3'-end mutants, but not 5'-end mutants, inhibited ovarian cancer cell line growth. Holt and coworkers^[33] also reported inhibition of MCF7 tumors in nude mice transfected with wild-type BRCA1, and an increase in survival in mice with established tumors transfected with the wild-type gene.

Jensen and colleagues^[34] have suggested that BRCA1 serves as a secreted inhibitor of growth. They conducted immunofluorescence studies using antibodies directed against the C- and N-terminals of the BRCA1 protein in breast epithelial cells. The studies demonstrated low levels of nuclear protein but strong membrane localization. In addition to these observations, they noted that the BRCA1 protein exhibits sequence homology to a family of acidic secretory proteins known as granins.^[34] They concluded that the BRCA1 protein functions as a novel secreted inhibitor of growth. The cellular distribution of BRCA1 reported by Jensen and coworkers differed from the pattern reported by Chen and associates,^[35] who identified BRCA1 as a 220-kD nuclear phosphoprotein that exhibits aberrant cytoplasmic localization in breast and ovarian tumor cells. When the sensitivity of the commercially available C-20 antibody used by Jensen and colleagues^[34] was examined, cross-reactivity with the human epidermal growth factor receptor (EGFR) and HER2 was observed,^[36] and no further support for the theory that BRCA1 acts as a secreted growth inhibitor has been reported.

Using mouse polyclonal antibodies, Chen and associates^[35] recently demonstrated that BRCA1 expression and phosphorylation is cell cycle dependent--it is increased in phases S and M--and that BRCA1 binds to cyclin-dependent kinases associated with cyclins D and A. These findings suggest that BRCA1 activity is regulated by cyclin-dependent kinases. This cell cycle dependency has been confirmed by Vaughn and colleagues,^[37] who showed low levels of expression in G0 and early G1 and high levels in late G1 and S phase.

The uncertainties associated with human BRCA1 function accentuate the importance of development and study of murine models for human disease. In 1995, Bennett and associates^[38] localized the mouse BRCA1 gene to chromosome 11. They found that the gene exhibits 75% identity with human sequence at a nucleotide level and 58% identity at the amino acid level. Three groups have now gone on to develop BRCA1-deficient mice.^[39,40,41] Mice carrying a single inactivated copy of the gene appear to remain healthy up to 11 months of age. Mice homozygous for the defective gene die in utero between 4.5 and 13 days, with disruption of mesoderm and of egg cylinder development. Hakem and coworkers^[40] found these mutant embryos to have decreased expression of cyclin E and mdm2, and increased expression of p21. Interestingly, despite the apparent in utero lethality of homozygosity for a BRCA1 mutation in the mouse, Boyd and colleagues^[42] identified a human BRCA1-mutant homozygote, who survived to age 34 years before developing a breast cancer. This finding, and the differences between mouse and human BRCA1 structure, may limit the conclusions relating to human disease that can be drawn from murine BRCA1 studies.

Cite this: BRCA 1 and 2--A Genetic Link to Familial Breast and Ovarian Cancer - Medscape - Feb 24, 1997.