Cell Signaling Regulatory Mechanisms Controlling Epithelial-Mesenchymal Transition in Carcinoma
Epithelial-mesenchymal transition (EMT) is a cellular program normally engaged during development and wound healing that is hijacked in many cancers to drive metastasis and resistance to therapy. The clinical implications of EMT in cancer progression have driven efforts to understand the cellular processes controlling EMT induction. Transforming growth factor-beta (TGFβ) and expression of related transcription factors potentiate EMT induction through complex and incompletely understood mechanisms. In this thesis, we investigated specific intracellular signaling pathways controlling maintenance of mesenchymal characteristics and EMT induction in response to growth factors in lung and pancreatic carcinoma cells. In lung carcinoma cells, extracellular signal-regulated kinase-1/2 (ERK1/2) pathway activation, which promotes cell survival and proliferation, was required for complete EMT induction. Furthermore, chronic ERK1/2 inhibition reversed baseline mesenchymal traits while simultaneously augmenting cellular sensitivity to a clinically approved small molecule EGFR inhibitor in cell lines with multiple clinically relevant modes of therapy resistance. In both lung and pancreatic carcinoma cell lines, TGFβ-mediated EMT was enhanced by co-treatment with epidermal growth factor (EGF), as had been noted in other contexts. We demonstrated that the ability of EGF to enhance TGFβ-mediated EMT depended on SH2 domain-containing phosphatase-2 (SHP2) activation through tyrosine phosphorylated adapter binding, which is required for complete ERK1/2 activation. Though SHP2 was not directly engaged and activated by TGFβ, SHP2 was required for TGFβ-mediated effects. Incomplete or transient effects of ERK inhibition and SHP2 depletion motivated subsequent systematic evaluation of cell signaling processes engaged during EMT induction to identify other pathways that control mesenchymal dedifferentiation in pancreatic carcinoma cells. We thus developed a data-driven computational model to predict the relationships between multivariate signaling events and EMT-associated phenotypes in response to combinations of TGFβ, EGF, and hepatocyte growth factor (HGF). Signaling intermediates that co-varied most with mesenchymal traits provided novel potential targets to inhibit EMT phenotype acquisition or restore epithelial traits in carcinoma. Together, this thesis enhanced mechanistic understanding of EMT regulation by SHP2, identified novel strategies to reverse EMT phenotypes in carcinoma cells, and generated a quantitative model to understand mesenchymal dedifferentiation, which can be leveraged in the future to improve clinical outcomes for cancer patients.