Abstract
Signaling mediated by the epidermal growth factor receptor (EGFR) plays a ubiquitously important role in regulating proliferation, survival, and migration of mammalian cells, and its dysregulation leads to cancer and other diseases. In cancer, EGFR is often mutated or overexpressed and is an important therapeutic target. Although EGFR-targeted therapies have shown remarkable efficacy in certain cancer backgrounds, their success is highly context-dependent and frequently hampered by resistance. These limitations underscore an incomplete understanding of how EGFR signaling is regulated across different cancer settings. This dissertation provides new mechanistic insights into EGFR signaling, particularly in cancer contexts that have previously been overlooked, through development of novel computational mechanistic models. In our first study, we investigated signaling consequences of low expression of the RAF kinases that connect receptor tyrosine kinases (RTKs) to the extracellular signal-regulated kinase (ERK) pathway. RAFs are the least abundantly expressed RTK-ERK pathway proteins in many cancer cells, with only a few hundred copies per cell localized and active at the plasma membrane. However, the consequences of limited RAF expression on pathway dynamics remain unclear. We trained a mechanistic computational model of the EGFR-ERK pathway and comprehensively characterized it using multivariate parameter sensitivity analyses and model Sloppiness analyses. Continuum and stochastic model predictions revealed that low RAF abundance suppresses EGFR-mediated ERK activation, limits the effects of oncogenic RAS mutants, and creates stochastic RAF dynamics that can propagate downstream. In our second study, we examined whether variability in endocytic EGFR trafficking contributes to cell-to-cell heterogeneity in ERK activation—a hallmark of many cancer backgrounds associated with drug resistance. While ligand-mediated endocytosis separates EGFR from membrane-bound signaling partners, EGFR can continue signaling from endosomes via reaction-diffusion processes that link endosomal EGFR-driven complexes to membrane-bound signaling molecules. However, existing models neglect natural variation in endosome number, size, and rate of movement that could impact the strength of these reaction-diffusion mechanisms. We developed a novel moving-boundary model of EGFR trafficking into moving endosomes and examined the effects of heterogeneous endosomal properties on ERK activation. The analyses of the coupled partial and ordinary differential-based model identified the number of EGFR-containing endosomes as an important determinant of heterogeneity in ERK activation and are supported by high-content image analyses of mammalian cells. Altogether, our findings provide fundamental insights into the role of protein abundances and spatiotemporal regulation in EGFR-ERK signaling that help identify key pharmacological targets for efficient antagonism of receptor-mediated signaling in cancer, suggest strategies for patient stratification, and provide a new framework for computational prediction and experimental validation of the context-dependent function of EGFR.