PROJECT SUMMARY/ABSTRACT: Project 1 Phenotypic plasticity can endow isogenic cell populations with multiple co-existing phenotypes. In the context of cancer, two phenotypes that can be of particular consequence to progression of aggressive cancers are `high proliferation-low migration' (“Proliferative” or P-phenotype) and `low proliferation-high migration' (Aggressive or A-phenotype). The so-called `grow or go' model suggests that cells can switch between these relatively stable phenotypic states, with the rates of transition being functions of the genetic and environmental changes. The increasing rate of switching from proliferative to migratory phenotypic states (the P-A switch) can signal progression to the more invasive (and ultimately metastatic) cancer progression. Currently, the signaling and regulatory networks underlying the P-A switch are poorly understood. In this project, we propose to investigate this switch in high quantitative detail using a novel approach allowing us to separate the two phenotypes based on competitive migration of cells in which genes can be silenced on the kinome- or genome- wide scale. This RACE assay will be performed with primary cells and cell lines of particularly invasive cancer types: glioblastoma multiforme and high grade melanoma. Our preliminary data suggest that the silencing of expression of a wide range of kinases can either enhance or suppress melanoma cell migration. Furthermore, faster, more aggressive cells were found to have differential enrichment of a variety of signaling pathways vs. the proliferative, slower cells. The combined analysis of the signaling and genetic perturbations suggested that the aggressive phenotype is a complex state that not only has different proliferation and migration characteristics, but also has different differentiation and metabolic states, in agreement with previous clinical observations. In addition to the expected molecular players, we identified novel potential regulators, and developed a novel, synthetic biology-based method to identify chemical compounds that can target these regulators. These and other findings led us to propose a program of research aimed at delineation, mathematical and computational modeling, and validation of the networks underlying cell-autonomous (i.e., assuming no cell-cell communication) P-A switching. We further hypothesized that the dynamics of the P-A switching can be influenced by the frequency of fluctuations of environmental factors that can suppress the proliferative phenotype, including drugs commonly used in the clinic. The model accounting for the population dynamics underlying this switching behavior will be trained using both cancer cells and, as a model, synthetically modified yeast cells. We anticipate that this project will lead to an increased understanding of invasive tumor spread estimated to account to more than 90% of deaths in human cancers. This project will be complemented by the second project focused on cell-non-autonomous mechanisms promoting or inhibiting cancer cell invasion.