Abstract P1-03-02: Development of a Novel Animal Model of Human Triple-Negative Breast Cancer by Retroviral Oncogene Transduction of Mouse Mammary Stem Cells

Abstract

Abstract Background: Triple-negative breast cancer (TNBC), which is negative for estrogen receptor (ER), progesterone receptor (PgR), and HER2 expression, represents approximately 20% of all newly diagnosed breast cancers. Patients with TNBC do not derive benefits from endocrine therapy and HER2-targeted therapy. Although TNBC is known to frequently involve mutations of the p53 gene, reduced expression of Rb, and activation of extracellular signal-regulated kinase (ERK), the molecular mechanisms of TNBC's etiology, and carcinogenesis animal models relevant to human TNBC, have not been established. In this study, we developed a mouse model of human TNBC. Methods: To induce TNBC in a murine mammary gland, we used 2 critical genetic events, knockout (KO) of Ink4a/Arf and of the activating form of Hras (12V). Ink4a and Arf play tumor suppressive roles upstream of Rb and p53, respectively; thus, KO of this locus is equivalent to abrogation of the Rb and p53 pathways. Meanwhile, Hras is an oncogene upstream of ERK, Akt, and other kinases. Mammary glands from Ink4a/Arf-KO mice were cultured in serum-free floating culture as single cells, which massed together to form mammospheres. Such mammosphere cells have been shown to repopulate mammary duct-alveolar units in vivo and to differentiate to basal or luminal mammary epithelial cells in adherent culture supplemented with serum, indicating that they are mammary stem cells. To these cellular masses, retroviruses carrying Hras(12V) were infected in floating culture. Then, single-digested mammosphere cells, including the retrovirus-infected cells, were inoculated into mammary fat pads of syngeneic mice. Tumors were evaluated for pathological and immunophenotypical findings and for molecular signature by RT-PCR analysis. Results: Induced tumors showed aggressive features, higher proliferation, dissemination to distant organs, invasiveness to adjacent lesions, and lethality. Pathologically, immature mammary ducts were identified as delineated by positive cells for keratin 14, a marker of basal cells. Immunophenotypically, all induced tumors were negative for ER, PgR, and HER2, although phosphorylation of ERK kinase was strongly and frequently positive. Together, these pathological findings and immune and biological phenotypes are quite similar to those of human TNBC. In addition, the tumors were negative for an epithelial marker, E-cadherin, although the mammospheres were actually derived from mammary epithelial cells. Meanwhile, induced-tumor cells expressed embryonic stem cell-like genes—Oct-3/4, Cripto, and Zfp42, but not Nanog—much more than did Ink4a/Arf-KO or wild-type mammary epithelial cells in RT-PCR analysis. Furthermore, in the induced tumors, we observed heterogeneous expression of CD49f, CD24, and CD44, which are candidate markers of mouse mammary cancer stem cells. Conclusion: Induced tumors had the biological and molecular essence of human TNBC, epithelial-to-mesenchymal transition, dedifferentiation, and cancer stem cells. Therefore, our genetically induced mouse mammary tumor model can help reveal the etiology of this disease from the comprehensive standpoints of pathology, biology, cellular context, and molecular signature. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P1-03-02.