Precision cancer medicine is predicated on the identification of pharmacologically actionable tumor dependencies. Mutated oncogenes represent one such class of dependencies and are a hallmark of targeted therapy. However, most tumor patients (~75%) do not present with actionable alterations. CaST, thus, proposes to investigate a new class of tumor dependencies that is comprised of master regulator proteins (MRs) that control the regulatory logic of tumor cells. These regulatory modules are referred to as tumor checkpoints, and the aberrant activity of tumor checkpoints is both necessary and sufficient to govern the pathophysiological state of tumor cells, i.e. they maintain tumor homeostasis and effect tumor plasticity. MRs are much more conserved across patients than the individual genetic alterations contributing to tumor etiology. Tumor checkpoints, thus, represent the Achilles' heel of cancer. A key goal of this project is to systematically dissect tumor checkpoints across tumor samples from publicly available repositories and generate a comprehensive, functionally annotated inventory of ~400 tumor checkpoint MRs. This Cancer Homeostasis Protein Database (CHoPD) is the focus of a novel reductionist approach to elucidating tumor homeostasis and plasticity as described in three aims (1) Elucidating key tumor checkpoints and associated MRs across all individual tumor samples in TCGA and related resources to develop CHoPD. Annotation of these proteins will include the upstream genetic alterations and signals responsible for their aberrant activity and their downstream genetic programs. Specific tumor signatures will be used to interrogate tumor-specific regulatory models with an integrated collection of analytical, network-based tools. These analyses will avoid the current lineage-based subtype classification in favor of one based on tumor homeostasis. (2) Elucidating the logic and regulatory architecture of tumor checkpoints responsible for tumor homeostasis and plasticity by decoding their autoregulatory activity, modeling their modular structure (i.e. protein-protein and protein-DNA complexes), and establishing the existence of critical control entry points, whose genetic or drug induced modulation may lead to checkpoint collapse. This will be accomplished by combining protein structure data, genetic and epigenetic data, as well as high-resolution time-series generated with perturbational assays, using small molecules and genetic perturbations. (3) Elucidating the interaction of tumor checkpoints in different tumor compartments, such as tumor cell niches, which represent alternative homeostatic states, and the interactions between tumors and their stromal and immune microenvironments. This will be accomplished by isolating individual cells representing individual tumor and microenvironment niches and studying the interactions between their coupled checkpoints using the tools and methodologies developed in Aims 1 and 2.
Showing the most recent 10 out of 27 publications
