The majority of ovarian cancers are thought to arise from the single layer of cells on the ovarian surface, the ovarian surface epithelium (OSE). Although ovarian epithelial cancers (ovarian adenocarcinomas, OAs) may share a common origin, they do not have a uniform appearance when examined under the microscope. Rather, OAs can be divided into four major types (serous, mucinous, clear cell, and endometrioid) based on pathologists' recognition of specific features of the tumor cells. The current clinical management of OAs does not, for the most part, take histologic type into account. However, abundant evidence suggests that the different types of OA may represent a collection of distinct diseases, each with different clinical, genetic, and biological features. Given the variable features seen in human OAs, attempts to model human ovarian cancer in the mouse should aim to better reflect this important type specificity. Such type-specific models could ultimately aid in the development of type-specific detection, prevention, and treatment strategies for ovarian cancer.

Cancers arise through clonal selection, essentially an evolutionary process that promotes outgrowth of precancerous and cancerous cells carrying gene mutations that confer the most robust proliferative and survival properties upon the cells. The cellular genes affected by mutation in cancer can be divided into two major classes: proto-oncogenes and tumor suppressor genes. Although mutations of proto-oncogenes and tumor suppressor genes are believed to be critical in virtually all cancer types, the specific genetic defects typically vary from one cancer type to another. Not surprisingly then, each histological type of OA shows some characteristic gene mutations. Our group and others have shown that activating mutations of the proto-oncogene encoding the b-catenin protein are frequently observed in ovarian endometrioid adenocarcinomas (OEAs), but not in the other histological types of OA. b-catenin mutation appears to play a critical role in determining which genes are expressed in OEAs. In a number of in vitro and animal model systems, mutant b-catenin has been shown to play an important role in transforming normal cells to cancerous ones. Similarly, inactivation of the PTEN tumor suppressor protein has been shown to occur frequently in OEAs. PTEN inactivation appears to be an early change because it has been identified in endometriosis, a putative precursor of OEA.

Many types of cancer, including those of the colon, endometrium, lung, brain, breast, and prostate, have been modeled successfully in the mouse. Attempts to generate mouse models of ovarian cancer met with little success until the recent identification of methods allowing selective expression of cancer-associated genes in the OSE. These methods have recently been used to generate transgenic mice that develop poorly differentiated OAs reminiscent of serous adenocarcinomas. We hypothesize that transgenic mice selectively expressing an oncogenic b-catenin protein (b-catS33Y) in the murine ovarian surface epithelium (MOSE) will form endometrioid adenocarcinomas with features similar to human OEAs. We propose studies to develop a mouse model of OEA by expressing a mutant form of b-catenin in the mouse OSE. We will also develop the technology to selectively inactivate OEA-associated tumor suppressor proteins, such as Pten, in the OSE. Any tumors that develop in the animals will be examined for their morphologic and molecular resemblance to their human counterparts.

(1) What will be the ultimate applicability of the research? This research has the potential to help all OA patients by increasing our understanding of how different types of OA develop and progress. Such studies may eventually lead to type-specific preventive, diagnostic, and treatment strategies for ovarian cancer patients, although this outcome will likely take several years to achieve.

(2) If the research is too basic for clinical applicability, what are the interim outcomes? Mouse models of many cancer types have provided important insights into how the tumors develop, progress, and respond to treatment. This study will advance ovarian cancer research by developing a model of OEA. The resources developed by this project (mice expressing or lacking specific cancer-associated genes in the OSE) will be made available to other researchers in the field.