Molecular mechanisms controlling epithelial to mesenchymal transition and cell invasion in development and cancer

The conversion of epithelial cells into migratory and invasive mesenchymal cells has been established as a fundamental step in the metastatic cascade. In cancer cells the induction of an epithelial-to-mesenchymal transition (EMT) awards invasive properties that allow dissemination from the primary tumor, and promotes acquisition of stem-cell-like properties and prosurvival mechanisms that contribute to therapeutic resistance, relapse, and decreased overall survival. Identifying inhibitors that block EMT is of great therapeutic interest to prevent disseminated tumor cells and target dormant cancer stem cells. Efforts to rationally target EMT in the clinic have been impeded by our incomplete understanding of the molecular mechanisms regulating EMT in normal and cancerous cells. Here we develop a zebrafish in vivo EMT reporter for rapid screening of small molecules called Tg(snai1b:GFP), which labels snai1b-expressing cells during development undergoing EMT from the dorsal neural tube to form neural crest (NC). We identified a multi-kinase inhibitor, called TP-0903, that potently blocked EMT, migration, and differentiation of dorsal neural tube progenitors by activating retinoic acid (RA) signaling. We show for the first time that RA directly controls EMT of NC cells in vivo. Our findings designate RA as an attractive therapeutic strategy to antagonize EMT-dependent pathways and stem-cell-like iv properties to eliminate residual cancer stem cells and metastatic disease. TP-0903 was rationally designed to target AXL RTK, a key oncogenic target that promotes EMT and metastasis. In NC derived cancers like melanoma, AXL can mediate invasion and promote resistance to targeted therapies in vitro. We hypothesize AXL promotes malignant melanoma by inducing EMT of cancer cells in vivo. To test this we generated a zebrafish melanoma model overexpressing AXL. We found AXL accelerated melanocyte transformation, tumor onset, growth and invasion. Additionally, we found that AXL did not activate EMT programs in vivo, despite numerous studies implicating AXL as a driver of EMT in vitro. Nevertheless these studies specify AXL as a driver of melanoma and establish an in vivo zebrafish model of AXL dependent melanoma invasion. This preclinical model will help determine the AXL dependent mechanisms driving malignant melanoma and evaluate effective AXL inhibitors to treat melanoma patients.