Identification of a core EMT signature that separates basal-like breast cancers into partial- and post-EMT subtypes

Epithelial-mesenchymal transition (EMT) is a cellular plasticity program critical for embryonic development and tissue regeneration, and aberrant EMT is associated with disease including cancer. The high degree of plasticity in the mammary epithelium is reflected in extensive heterogeneity among breast cancers. Here, we have analyzed RNA-sequencing data from three different mammary epithelial cell line-derived EMT models and identified a robust mammary EMT gene expression signature that separates breast cancers into distinct subgroups. Most strikingly, the basal-like breast cancers form two subgroups displaying partial-EMT and post-EMT gene expression patterns. We present evidence that key EMT-associated transcription factors play distinct roles at different stages of EMT in mammary epithelial cells.