Cancer genomics approaches elucidate tumor mutational and microenvironmental landscapes and dynamics
Date of Award
Doctor of Philosophy (PhD)
Next generation sequencing, especially of DNA and RNA, has been rapidly integrated into preclinical and clinical cancer research. Challenges in these studies range from sample type and quality to translation of these results to significantly impact our understanding of human disease. These challenges further complicate analyses designed to elucidate biomarkers and mechanisms associated with treatment response. The projects described in this dissertation begin by characterizing the mutational landscape across tumor cohorts, followed by more in depth analyses of their expression landscapes. In Chapter 2, comprehensive sequencing approaches were used to discover the somatic alterations and signaling events leading to spontaneous tumorigenesis in a genetically engineered mouse model (NRL-PRL) for estrogen receptor positive breast cancers, driven by overexpressed prolactin ligand (PRL) in the mammary glands of female mice. We identified recurrent activating mutations in Kras in 79% of NRL-PRL tumors, suggesting that the hyperplasia due to prolonged PRL exposure resulted in a selective signaling bottleneck for Ras pathway activation. In Chapter 3, we assessed the fidelity of patient-derived xenografts (PDX) as in vivo models for their corresponding oral cavity squamous cell carcinoma (OCSCC) tumors. We found that these PDXs highly maintain both the complex genotypes and tumor-intrinsic gene expression patterns of their matched tumors. While both of these models were established to study human disease, we explored how they fundamentally differ from human cancers. In the case of the immune competent NRL-PRL model, we identified modulation of the immune microenvironment, more prominently in the myeloid compartment, that has similarly been observed in KRAS-driven human tumors. Conversely, PDX models were generated in immunodeficient mice, restricting evaluation of the microenvironmental dynamics associated with tumor growth. However, we still showed that PDX models were established from OCSCC tumors spanning the diverse intratumor heterogeneity previously described across the head and neck squamous cell carcinoma (HNSCC) disease landscape. Chapters 3 and 4 explore the therapeutic efficacy of a targeted MEK1/2 inhibitor, trametinib, and an anti-PD1 immune checkpoint blockade agent, pembrolizumab, in HPV-negative HNSCC. Once again, we used genomic and transcriptomic analyses to characterize the mutational landscapes of these patient cohorts. However, these approaches were more specifically designed to relate the mutational and gene expression landscapes to clinical and pathologic responses to treatment. We used algorithmic approaches to identify mutations that confer immune recognition of tumor cells, as well as deconvolute the infiltrating immune microenvironment in bulk tumor RNA samples. These analyses addressed challenges related to tumor purity and the presence of nontumor cell populations by quantifying and characterizing the microenvironmental cell populations and dynamics in primary tumors. Furthermore, we were able to identify biomarkers and correlates of response by comparing matched untreated and post-treatment tumor samples, further elucidating drug mechanisms related to sensitivity or response to treatment. This dissertation describes studies that effectively address challenges in cancer genomics, including tumor sampling, heterogeneity within and across cancer types, and the role of the microenvironment, in order to successfully establish models for studying human cancers as well as identify correlates of therapeutic response.
Chair and Committee
Obi L. Griffith
Joshua Rubin, Malachi Griffith, Ravindra Uppaluri, Robert Schreiber,
Campbell, Katie, "Cancer genomics approaches elucidate tumor mutational and microenvironmental landscapes and dynamics" (2018). Arts & Sciences Electronic Theses and Dissertations. 1694.
Available for download on Thursday, December 15, 2118
Biogeochemistry Commons, Bioinformatics Commons, Biology Commons
Permanent URL: https://doi.org/10.7936/ga47-0648