Abstract 411: Risk-dependent enhancer transcription defines oncogenic gene regulatory networks for high-risk neuroblastoma

Abstract

Abstract Neuroblastoma is the most frequent extracranial solid tumor in childhood. High-risk neuroblastoma (HR-NB) with less than 50% five-year event-free survival rates (EFSR) is distinct from non-high-risk neuroblastoma (NHR-NB) with above 90% EFSR. Investigating risk-dependence is therefore essential, and that of noncoding regulation is extra essential because to date only few coding mutations have been found in pediatric tumors. Among noncoding regulators, active enhancer plays the most crucial role to precisely control gene expression. However, given surrogate marks of enhancers, identifying functional enhancer is challenging. We have recently developed a noncoding transcriptome-based computational approach to define functional enhancers, given that some long noncoding RNAs (lncRNAs) transcribed from enhancers are required for enhancer activity. Here, we apply risk-dependence of lncRNA expression to identify oncogenic enhancers and enhancer-associated lncRNAs, determining an oncogenic regulatory network in HR-NB. LncRNAs are often unstably expressed thus have a higher proportion of risk-dependence than coding genes. We re-analyzed the noncoding transcriptome of sequenced primary HR-NB (n=176) versus NHR-NB (n=322) samples (tumors cell content >60%) and identified 8.3k lncRNAs (3.4k up-regulated and 4.9k down-regulated) for the high-risk patients (FC>2, FDR<0.05). Similarly, from the coding transcriptome profiling of the same patients, we identified 694 up-regulated and 925 down-regulated genes. Noncoding transcriptomic risk-group dependence is a robust metric for enhancer activity. Comparing risk-dependent distal lncRNAs with that of adjacent risk-dependent genes there exists a significant correlation in their direction and magnitude (coefficient=0.70, P<2e-16). Furthermore, risk-dependent lncRNAs reside at active enhancers specific to the neural tube (VISTA, 37 out of 257 validated enhancers, empirical P<0.005). Importantly, we observed three strong oncogenic associations for these 37 “risk-dependent and tissue-specific enhancers.” 1) 70 human transcription factors (TFs) showed significant binding potency that indicates enhancer function, exampled by the tumorigenic FOXM1. These TFs significantly over-represent known oncogenes (FET P=0.01, odds=5) and tumor suppressors (P=0.0003, odds=7.5). 2) Risk-dependent target genes that remarkably over-represent the proliferation of neuroblastoma cell lines (P=0.0001, Ingenuity Pathway Analysis) indicate a directly tumorigenic impact by these risk-dependent enhancers. 3) Risk-dependent enhancer lncRNAs reside at neuroblastoma susceptibility loci more often than by chance. In summary, risk-dependence of noncoding transcripts illuminate oncogenic enhancers with tissue-specific activity and neuroblastoma susceptibility, therefore is a general facet of oncogenic enhancer activity. Citation Format: Zhezhen Wang, Ivan P. Moskowitz, John M. Cunningham, Xinan Holly Yang. Risk-dependent enhancer transcription defines oncogenic gene regulatory networks for high-risk neuroblastoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 411.