Gene expression profiling of Met receptor tyrosine kinase-induced mouse mammary tumors

Breast cancer is a heterogeneous disease comprised of distinct biological entities that correlate with diverse clinical outcomes. Gene expression profiling has divided this heterogeneity into luminal, ERBB2+ and basal molecular subtypes. Basal breast cancers are difficult to treat as they lack expression of candidates suitable for targeted therapies and are associated with poor outcome.Elevated protein level of the hepatocyte growth factor receptor, MET, is observed in 20% of human breast cancers and correlates with poor prognosis. However, the role of MET in mammary tumorigenesis is poorly understood. To address this, we generated a murine model that expresses weakly oncogenic mutants of Met (Metmt) in the mammary epithelium under the transcriptional control of the mouse mammary tumor virus promoter. We demonstrate that Metmt induces mammary carcinomas with diverse phenotypes and used gene expression microarrays to elucidate gene expression changes induced by Met. Since mammary tumors contained variable contents of epithelium and stroma, we used laser capture microdissection to procure epithelial cells for microarray analysis. Based on immunohistochemistry and expression profiling, we show that Metmt produces tumors with luminal or basal characteristics. From hierarchical clustering, Metmt-induced basal tumors clustered with murine models that share features of epithelial to mesenchymal transition and human basal breast cancers. Moreover, Metmt basal tumors clustered with human basal breast cancer. The status of MET among the human breast cancer subtypes has not previously been addressed. We demonstrate that MET levels are variable across molecular subtypes but show elevation in the basal subtype and correlates with poor outcome. We used a candidate gene approach derived from microarray data to gain an understanding of signals required for Met-dependent tumorigenesis. We investigated Nck adaptor proteins and demonstrate a role for Nck in cell motility and actin dynamics of Met-dependent breast carcinoma cells and show elevated expression in human basal breast cancers. By generating a unique mouse model in which Met is expressed in mammary epithelia, with the examination of MET levels in human breast cancer, we have established a novel link between MET and basal breast cancer. This work identifies poor outcome basal breast cancers that may benefit from anti-MET therapies