(100g) Analysis of EGFR Flux Through Different Endocytic Pathways in Cancer Cells with Elevated EGFR Expression

In cancer cells characterized by elevated expression of the epidermal growth factor receptor (EGFR), perturbations to the kinetics of ligand-mediated EGFR endocytosis are frequently observed.  Such perturbations can have important consequences for the ability of EGFR to activate downstream signaling pathways and may therefore affect cancer cell growth rates and response to therapeutics.  Given the existence of endogenous mechanisms which both promote and impede EGFR endocytosis in response to EGFR activation, the question arises as to how these processes act in concert to determine rates of EGFR endocytosis in cancers with elevated EGFR expression.  To quantitatively address this question in the context of non-small lung cancer (NSCLC) cells with elevated EGFR expression, we developed a novel computational model to parse the effects of Sprouty2, an antagonist of EGFR endocytosis through the CBL pathway, and MIG6, which promotes non-CBL-mediated EGFR endocytosis, on overall EGFR endocytosis kinetics.  Our computational framework consisted of a set of coupled ordinary differential equations describing the kinetics of EGFR ligand binding, EGFR dimerization, and EGFR endocytosis through pathways involving CBL (which is sequestered by Sprouty2), MIG6, or neither protein.  To determine the unknown rate constants describing the flux of EGFR through the different endocytosis pathways, we created a panel of H1666 and PC9 NSCLC cell lines with knockdown of Sprouty2, MIG6, or both proteins.  Whereas H1666 cells express wild-type EGFR, PC9 cells express a constitutively active EGFR mutant which displays impaired ligand-mediated endocytosis.  In addition to measuring rates of EGF-mediated EGFR endocytosis in these cell lines, we also measured EGFR expression levels and the fraction of endocytosed EGFR which was recycled to the cell surface.  In both cell lines, SPRY2 knockdown promoted EGFR endocytosis and MIG6 knockdown impeded endocytosis.  With SPRY2 knockdown alone or together with MIG6 knockdown, we also observed a marked reduction in EGFR expression in both cell lines.  Based on these data, our model results suggested that the effects of Sprouty2 knockdown on EGFR endocytosis were almost completely attributable to changes in EGFR expression, a prediction which we validated by reconstituting EGFR expression in H1666 and PC9 cells with Sprouty2 knockdown.  In addition, our model results suggested that the MIG6-mediated internalization pathway plays a surprisingly significant role in setting EGFR endocytosis kinetics of wild-type EGFR in H1666 cells.  Not surprisingly, the CBL-mediated pathway was also found to play a significant role in EGFR endocytosis in H1666 cells.  By contrast, the model results suggested that the CBL-mediated pathway plays essentially no role in mutant EGFR endocytosis in PC9 cells and that MIG6 may promote mutant EGFR endocytosis less efficiently than wild-type EGFR endocytosis.  In addition to these specific conclusions regarding wild-type and mutant EGFR endocytosis in NSCLC cells, the results of this work demonstrate the utility of our computational and experimental approach for identifying the most relevant determinants of EGFR endocytosis in different cellular contexts.  This approach may be of general use in identifying optimal ways to regulate EGFR endocytosis kinetics for therapeutic purposes.