Head and neck squamous cell carcinoma (HNSCC) is a heterogeneous disease of significant mortality and morbidity, with a 5-year survival rate that has shown only modest improvement. This persistent patient mortality and morbidity can be attributed, in part, to an overall lack of novel targeted treatment options. Indeed, recent genome sequencing studies have highlighted the striking heterogeneity of HNSCC tumors and have failed to identify pervasive and broadly actionable driver mutations and/or other genetic alterations. Despite this genetic heterogeneity, HNSCC tumors can be classified into gene-expression defined subtypes, raising the possibility that such distinct gene expression programs and regulatory networks are closely interwoven with an intrinsic heritable and signature epigenetic state. We hypothesized that there exist key oncogenic transcriptional regulators of subtype dependency and that defining the epigenetic landscape of HNSCC tumors will reveal the identity of these crucial factors. This will be impactful since targeting of tumor addiction to key oncogenic Transcription Factors (TFs), although successful in a number of other tumor types, remains largely unexplored in HNSCC due to limited knowledge of the identity of these regulators in this cancer.
To decipher the genomic and epigenomic drivers of HNSCC tumor subtypes, we examined a large cohort of cell lines that represent the Atypical, Basal and Mesenchymal subtypes. By using global H3K27 acetylation levels as a surrogate marker, we first generated a consensus genome-wide regulatory enhancer and Super-Enhancer map. Interestingly, we observed a differential enrichment pattern of H3K27Ac signals that recapitulated the segregation of cell-lines according to subtype based on differential gene expression profiles. We leveraged the differential enhancer map to identify enriched DNA-Binding Motifs to generate a Transcription Factor Combinatorial Regulatory Network, based on the co-occurrence frequency of TFs relative to genomic background. Our analysis highlighted a Cis-Regulatory Module (CRM) Network, specific to each molecular subtype of HNSCC. Furthermore, we identified key TF motifs that are embedded within these CRMs and enriched for each subtype-specific network, which we posit represent the pertinent TF drivers of the distinct gene-expression network found in various HNSCC tumors. Here, as proof of principle, we have examined the potential interaction between TP63 and ETS1, two oncogenic TFs revealed by our epigenomic and genomic analysis to be crucial mediators of HPV-negative HNSCC tumors.
To better understand the TP63-ETS1 driven molecular processes in HPV-negative tumors, we performed transcriptomic analysis of the effects of knockdown of TP63 and ETS1 by RNA-Seq and identified their global targets by ChIP-Seq. Interestingly, our comprehensive analysis showed that TP63 and ETS1 share a significant number of genomic binding sites, exhibit a marked preference for binding to Super-Enhancers and together act in a concerted fashion to drive expression of unique set of pathways and processes. By integrating our findings with meta-analysis of HNSCC patient datasets, we have uncovered a core TP63-ETS1 driven signature network that serves as a regulatory anchor of common facets of cancer biology, such as proliferation and angiogenesis, as well as drug-targetable tumor-specific processes involving cancer stem cells. Our study has harnessed the power of an integrative genomic/epigenomic approach to characterize HNSCC subtypes and heterogeneity and to provide a robust framework for the discovery of novel tumor drivers that can be leveraged for targeted therapeutics.
Citation Format: Christian Gluck, Isha Sethi, Maria Tsompana, Satrajit Sinha. An oncogenic transcriptional network anchored by ETS1, p63 and subtype specific drivers of HNSCC revealed by epigenomic and genomic interrogation [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 07.