Large-scale national and international cancer sequencing programs are generating a compendium of tumor- associated genomic alterations to prioritize the most promising therapeutic targets for drug development. These efforts have uncovered a staggering level of genome complexity in cancer. Although much is known about the function and clinical impact of recurrent aberrations in well-known cancer genes, less is known about which and how the more abundant, low-frequency mutations contribute to tumor progression. Effective translation of tumor genomic datasets into cancer therapeutics will require new experimental systems to inform the functional activity of targets in the relevant biological context encompassing inter- and intra-tumoral heterogeneity. To address these needs, we propose a CTD2 Center that will provide the research community high-throughput informatic and experimental approaches to characterize and validate pathogenic ?driver? mutations and fusion genes as well as identify molecular markers that meaningfully predict responses or resistance to anticancer therapies. We will pursue the following Specific Aims:
In Aim 1 we will implement an algorithmic framework for identifying driver mutations with high sensitivity and specificity. We will focus our algorithm development, training and testing efforts on predicting oncogenic, gain-of-function mutation drivers of glioblastoma multiforme (GBM), pancreatic ductal adenocarcinoma (PDAC) and epithelial ovarian cancer (EOC). These computational approaches will be amenable to the analysis of all cancer types. We will next engineer ~1,500 selected mutations and ~400 fusion genes into expression vectors along with cohorts of personalized, patient-defined coding mutations.
In Aim 2 we will enter mutant alleles and fusion genes into GBM, PDAC and EOC context-specific, in vivo functional screens that take into account the importance of genetic context, tumor microenvironment and heterogeneity in the selection of single and combinatorial drivers of tumorigenesis.
In Aim 3 we will determine the consequences of intra-tumoral heterogeneity on tumor sensitivity and resistance to therapeutic agents using DNA-barcoded, human patient-derived xenograft models that recapitulate the heterogeneity of cancer. We will determine the extent to which single targeted agents and their rational combinations alter tumor population dynamics. We will also leverage Aim 1 informatics and functional characterizations in Aim 2 and 4 to characterize ?persistor? populations to identify aberrations associated with drug resistance.
In Aim 4 we will use high-throughput functional proteomics, innovative protein- protein interaction assays and informer drug library screening studies to elucidate underlying mechanisms and therapeutic liabilities engendered by validated drivers. The foundational platform implemented in our CTD2 Center will provide a validated pipeline for the rapid characterization of gain-of-function aberrations that can be industrialized across tumor lineages to guide clinical management of cancer patients.
/ Relevance A central challenge we face in implementing next generation sequencing into cancer patient care involves determining which gene aberrations in a given heterogeneous tumor are 'drivers' that determine tumor behavior versus 'passengers' that result from the inherently unstable nature of cancer genomes. This proposal provides a tightly integrated approach to address the need for algorithms and cancer context-specific, high- throughput screening and biological response platforms. This systems will be used to prioritize the vast genomic data being generated by large-scale tumor sequencing efforts and will thus enable the identification and selection of optimal therapies targeting novel driver aberrations for each patient.
Showing the most recent 10 out of 13 publications