Abstract 2376: Mouse mammary tumor model relevant to human triple negative breast cancer

Abstract

Abstract Background: Triple negative breast cancer (TNBC)-breast cancer immunophenotypically identified as negative for estrogen receptor (ER), progesterone receptor (PgR), and HER2-has intrinsically aggressive biological features and is hard to target molecularly with existing drugs. TNBC is known to have frequent mutations of p53 and an abrogated Ink4a-Rb pathway. However, how dual dysfunction of p53 and Ink4a-Rb pathways contributes to TNBC-like phenotypes has not been fully elucidated. Thus, we sought to induce mammary tumors from mouse mammary epithelial cells from mice with knockout (KO) of Ink4a and alternative reading frame (Arf), transcripts of which are upstream accelerators of Rb and p53, respectively. Methods: Mammary glands of 7 week-old female Ink4a/Arf-KO mice were enzymatically digested, and single suspended cells were plated in floating culture to produced mammospheres. To induce tumors, retroviruses carrying Hras(G12V) or c-myc, both of which induce sporadic mammary tumors in transgenic mice, were transduced into mammosphere-cells in floating culture. Wild-type (WT) mammospheres were characterized for cellular target associated with carcinogenesis. Mammosphere-cells infected with the viruses were transplanted into mammary fat pads of syngeneic mice. Results: Although some cells in WT mammospheres were positive for keratin (K)18 and/or K14, most did not express these differentiation markers. When WT mammospheres were cultured on collagen-coated dishes with serum-containing medium, each mammosphere-cell exclusively committed to K18 or K14-expressing cells. This indicated that mammosphere-cell had a potential for bilineage differentiation, suggesting a feature of mammary progenitor or stem cells. Ink4a/Arf-KO mammosphere-cells regenerated mammary glands in vivo. Together, these results indicated that mammosphere formation is a novel method for obtaining mammary epithelial cells with bilineage differentiation potential. Transplantation of Ink4a/Arf-KO mammosphere-cells infected with Hras(G12V), but not with c-myc or mock, induced mammary tumors in vivo. Induced tumors were negative for ER, PgR, and HER2 and had some of the pathologic characteristics of human TNBC, such as pushing border and central necrosis. To molecularly understand this carcinogenesis, we compared gene expressions of WT and induced tumor-derived tumorospheres and performed gene set enrichment analysis. Through the permutation analyses of biological pathways, we purified 41 genes with the highest enrichment scores. Most of these genes are relevant to proliferation or mitosis. In the supervised analysis using the 41 genes as a classifier, TNBCs were clearly separated from non-TNBCs with accuracy of up to 90% in 3 different public datasets of human breast cancer. Conclusion: We have established a mouse mammary tumor model relevant to human TNBC, which give us an opportunity to identify the molecular targets for human TNBC. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 2376. doi:10.1158/1538-7445.AM2011-2376