Inherited defects of the BRCA1 and BRCA2 genes confer a profound predisposition to breast and ovarian cancer in women. Since the identification of these breast cancer susceptibility genes by positional cloning, we have used gene targeting to develop experimental mouse models for defects in the Brca2 gene. Our studies have focused on mice lacking the carboxy terminal domain of the Brca2 protein (exon 27) because this portion of the gene product has been shown to contain nuclear localization signals and it interacts directly with the Rad51 protein which is thought to be critical for maintaining genomic stability in response to DNA damage. Using homologous recombination techniques, we generated mice that carry one (hemizygous ) or two (homozygous) mutant alleles of Brca2 exon 27. These exon 27 null mice are viable although we have observed a subtle but statistically significant deficiency in the expected number of homozygous Brca2 mutant mice. An initial study with these mice has shown that homozygous mutant Brca2 mice on a C57BL/6J background are predisposed to a low incidence of spontaneous gastric carcinoma compared to their heterozygous and wild type littermates. High- and low-level gamma radiation studies (5 & 0.3 Gy at 5 wks of age) with intercrosses between these C57BL/6 Brca2-exon 27 deficient mice and BALB/c-P53-deficient mice are being followed for neoplastic development in mammary gland and other tissues. Preliminary results from this study indicate that lifetime survival is reduced and carcinoma incidence and latency is increased in Brca2 null mice . Irradiation further increased tumor incidence and decreased survival in heterozygous mice. Fluorescent microsatellite marker-assisted breeding techniques (speed congenics) were used to transfer this Brca2 exon 27 mutation onto genetic backgrounds (BALB/c and SWR) that we have previously shown are more susceptible to radiation-induced mammary carcinogenesis. These mice strains are currently being evaluated to assess tumor susceptibility.