Project 1: Dynamic Genomic and Microenvironmental Models of Acquired Chemoresistance

摘要

ABSTRACT Breast and ovarian cancers are heterogeneous diseases, as a typical tumor contains multiple “subclones”, which are defined as evolutionarily related subpopulations of cells with a different complement of somatically acquired DNA mutations and phenotypes. When chemotherapeutic agents are administered to the patient, some of these subclones may gain a selective advantage and develop resistance to the treatment, resulting in cancer relapse and progression. For this reason, it is imperative to identify these subclones and their evolution across treatment; and to understand how the genomic aberrations within these subclones drive resistance to chemotherapy. We will integrate experimental biology and computational models across temporal samples of patient tumors as they develop a resistant state in order to better understand and combat refractory and terminal cancer. To enable the study of tumor heterogeneity evolution in patients, we will utilize a highly unique collection of metastatic tumor cells from breast and ovarian cancer patients before, during, and after treatments, often across multiple courses of chemotherapy, as well as tumors from a clinical trial taken before and after therapy. We use deep sequencing to find genomic aberrations at each of these time points, and develop systems models to identify the subclones and follow phenotypic changes and their functional impacts of subclone evolution in response to chemotherapy. We hypothesize that 1) Dynamical systems models based on the evolution of subclone structure and acquisition of oncogenic phenotypes during treatment can identify key factors in the development of a chemo-resistant state; and 2) We can delay development of a chemo- resistant cancer state by inhibiting development of phenotypes that emerge over time commonly during treatment. We will model resistant cancer cell populations and both extrinsic and immune microenvironmental factors to identify critical features of acquired resistance and apply these models to a clinical trial aimed at blocking transition to a resistant cancer state. While these components can exhibit co-dependencies, by their nature they can also have vulnerabilities based on these interactive features, and if one can inhibit dependent relationships within a population it may be possible to shift the equilibrium of a tumor from a chemoresistant state to a sensitive state. The algorithms and procedures we are developing in this proposal will for a rational basis for real-time patient monitoring and making treatment choices for refractory patients. The outcomes of this research will deliver approaches to block or reverse the transition to a resistant state for advanced stage breast and ovarian cancer patients.