Identification Of The Tumor Transition States Occurring During Emt And Their Implications For Metastasis

Epithelial to Mesenchymal Transition (EMT), rather than being a binary switch, has been hypothesized to represent a spectrum of intermediate tumor states, but direct evidence supporting this concept in vivo in primary tumors is still lacking. In addition, the number of distinct states and the mechanisms regulating the transition between them, their plasticity, and their implication in tumor progression and metastasis are hitherto unknown. Here, we used a genetic mouse model of skin squamous cell carcinoma mediated by oncogenic KRasG12D and p53 loss of function that undergoes spontaneous EMT to identify different subpopulations of epithelial (Epcam+) and mesenchymal (Epcam-) tumor cells. To this end, we screened a large panel (u003e250) of cell surface markers, uncovering 15 markers that were heterogeneously expressed during EMT. Using multicolour FACS analysis we identified up to 6 different tumor subpopulations representing the different EMT transitional states that occurred spontaneously during tumorigenesis in vivo. Cytospin using K14 and Vimentin as well as qRT-PCR with well-characterized markers of epithelial and mesenchymal states confirmed that these different populations represent EMT transitional states, from very early hybrid epithelial/mesenchymal states to completely mesenchymal states. Transplantation of limiting dilution of these different populations into immunodeficient mice showed that they exhibit similar tumor propagating cell frequency. However, phenotypic analysis of the secondary tumors revealed that these populations present different plasticity with the populations characterized by the early steps of EMT giving rise preferentially to tumors with epithelial features whereas the tumor cell populations corresponding to the late stages of EMT preferentially giving rise to tumors with mesenchymal features, identifying the point of no return during EMT. To determine the molecular mechanism controlling these cell state transitions, we assessed their transcriptional and epigenetic landscapes using RNA-seq and ATAC-seq. We uncovered transcription factors (TFs) and signaling pathways specifically enriched in the different tumor subpopulations. Additional functional experiments using CRISPR-Cas9 deletion of the newly identified TFs, as well as the ligands identified the key molecular features that stimulate the different transition states during EMT. Finally, we found that the distinct populations of tumor cells exhibit great differences in metastatic frequency and identified the tumor cell population with highest metastatic potential, opening new avenues for therapy. Altogether our findings identified for the first time the different transitional states associated with EMT in primary tumors in vivo and uncovered the molecular mechanisms associated with these transition states, as well as their importance for metastasis. Citation Format: Ievgenia Pastushenko, Panagiota Sotiropoulou, Cedric Blanpain. Identification of the tumor transition states occurring during EMT and their implications for metastasis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr LB-146. doi:10.1158/1538-7445.AM2017-LB-146