High-grade serous carcinomas arise in the mouse oviduct via defects linked to the human disease.

Recent studies have suggested that the most common and lethal type of 'ovarian' cancer, i.e. high-grade serous carcinoma (HGSC), usually arises from epithelium on the fallopian tube fimbriae, and not from the ovarian surface epithelium. We have developed Ovgp1-iCreERT2 mice in which the Ovgp1 promoter controls expression of tamoxifen-regulated Cre recombinase in oviductal epithelium - the murine equivalent of human fallopian tube epithelium (FTE). We employed Ovgp1-iCreERT2 mice to show that FTE-specific inactivation of several different combinations of tumour suppressor genes that are recurrently mutated in human HGSCs - namely Brca1, Trp53, Rb1, and Nf1 - results in serous tubal intraepithelial carcinomas (STICs) that progress to HGSC or carcinosarcoma, and to widespread metastatic disease in a subset of mice. The cancer phenotype is highly penetrant and more rapid in mice carrying engineered alleles of all four tumour suppressor genes. Brca1, Trp53 and Pten inactivation in the oviduct also results in STICs and HGSCs, and is associated with diffuse epithelial hyperplasia and mucinous metaplasia, which are not observed in mice with intact Pten. Oviductal tumours arise earlier in these mice than in those with Brca1, Trp53, Rb1 and Nf1 inactivation. Tumour initiation and/or progression in mice lacking conditional Pten alleles probably require the acquisition of additional defects, a notion supported by our identification of loss of the wild-type Rb1 allele in the tumours of mice carrying only one floxed Rb1 allele. Collectively, the models closely recapitulate the heterogeneity and histological, genetic and biological features of human HGSC. These models should prove useful for studying the pathobiology and genetics of HGSC in vivo, and for testing new approaches for prevention, early detection, and treatment.