Abstract 05: Untangling the gene-epigenome networks: Timing of epigenetic regulation of gene expression in acquired cetuximab resistance gene programs

Abstract

Abstract The current study performs time course RNA-seq and DNA methylation profiling to determine the complex interactions between gene expression and epigenetic changes in acquired therapeutic resistance. The genomics era provides widespread characterization of the genomic landscape of tumors and has enabled precision treatment strategies. Currently, epidermal growth factor receptor (EGFR) inhibitors are the only FDA-approved targeted therapy for clinical use in head and neck squamous cell carcinoma (HNSCC). EGFR inhibitors are only effective in a subset of each of these tumors. Moreover, patients with de novo sensitivity to EGFR inhibitors often subsequently acquire resistance and succumb to their tumors. Numerous genetic and epigenetic alterations occur in tumors with acquired resistance. However, their timing and function remain unknown. Therefore, we develop new computational techniques to find gene programs associated with the acquisition of resistance. High-throughput transcriptional profiling enables unprecedented characterization of individual genes during cancer treatment. However, identifying and targeting mechanisms of EGFR resistance from these high-throughput data requires novel systems biology techniques that can discriminate altered cellular signaling pathways in response to cancer treatment. Therefore, we developed a new bioinformatics algorithm Coordinated Gene Activity in Pattern Sets (CoGAPS) to quantify multiplexed regulation and gene reuse in oncogenic signaling. To test this algorithm, we modified the HaCaT keratinocyte cell line model of premalignancy to simulate cancer cells with constitutive overexpression of wild-type EGFR and with activating mutations in HRAS and PIK3CA in a controlled genetic background. We apply CoGAPS to gene expression data from these models. This algorithm quantified relative changes in cellular signaling pathway activity in these data, not accessible to standard gene comparisons. Increases in CoGAPS pathway signatures from HRAS also occurred in gene expression data from the cetuximab resistant HNSCC cell line (1CC8) relative to its parental, sensitive cell line (UMSCC1). Investigation of the mechanisms of acquired resistance has previously been limited by reliance on case-control comparisons between sensitive and resistant cells, such as the UMSCC1 and 1CC8 cell lines. Since activity in cellular signaling pathways evolves during cetuximab resistance, it is essential to extend these case-control paradigms to quantify the dynamics responsible for resistance. Therefore, we developed a novel time course analysis to determine the molecular mechanisms of acquired cetuximab resistance in HNSCC. Specifically, we treated the cetuximab sensitive SCC25 HNSCC cell line over generations with both cetuximab and PBS. This long-term treatment protocol models the progression of acquired therapeutic resistance, including controls for clonal selection unrelated to treatment. Overexpression of the previous HRAS gene program also occurs in gene expression data measured during acquired cetuximab resistance in SCC25. We also measure DNA methylation during this time course to find the driver of this aberrant signaling associated with resistance. We apply CoGAPS analysis to the time course data for both DNA methylation and gene expression. This analysis distinguished early gene expression changes from cetuximab treatment from longer-term epigenetic alterations to gene expression during acquired resistance. Epigenetic regulation of FGFR1 expression emerged as the dominant mechanism of acquired therapeutic resistance in this system. Thus, our integration of time course DNA methylation and gene expression data enables unprecedented inference of the timing of targetable gene-epigenome programs responsible for acquired resistance. Citation Format: Genevieve Stein-O'Brien, Luciane T. Kagohara, Sijia Li, Manjusha Thakar, Ruchira Ranaweera, Michael Considine, Ludmila V. Danilova, Hiroyuki Ozawa, Joseph A. Califano, Daria A. Gaykalova, Michael F. Ochs, Christine H. Chung, Elana J. Fertig. Untangling the gene-epigenome networks: Timing of epigenetic regulation of gene expression in acquired cetuximab resistance gene programs [abstract]. In: Proceedings of the AACR-AHNS Head and Neck Cancer Conference: Optimizing Survival and Quality of Life through Basic, Clinical, and Translational Research; April 23-25, 2017; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2017;23(23_Suppl):Abstract nr 05.