Abstract 3495: The DNA methylation landscape of long noncoding RNAs in human cancer

Abstract

Abstract Integrating multi-dimensional genomic and epigenetic data from the TCGA, ENCODE and CCLE projects, we comprehensively characterized the epigenetic landscape for lncRNA genes in 9837 tumor samples across 33 cancer types. This analysis revealed that lncRNAs can be both epigenetically silenced and activated by DNA methylation alteration in the promoter region, which is in striking contrast to the well-documented hypermethylation of protein-coding gene promoters in tumor. To determine the corresponding transcriptional changes of lncRNAs with epigenetic alterations in tumors, we applied a heuristic strategy to identify the epigenetic silenced and activated lncRNA by integrating the transcriptome of 58,648 lncRNAs from 6475 TCGA RNA-seq files. This analysis characterized a patient-centric matrix with DNA methylation status of 3380 epigenetically regulated lncRNAs across 6475 tumor samples. Some epigenetically-activated lncRNAs exhibited an “on or off” pattern with completely no expression in normal tissues, which potentiated them as promising diagnostic biomarkers. To rule out the interference of tumor associated stromal and immune cells in lncRNA analysis, we investigated the RNA-seq and DNA methylation profiles of 784 cell lines from breast, lung, and bladder cancer. Most of epigenetically regulated lncRNAs identified in tumor samples showed prominently negative correlation between their expression and promoter methylation in cancer cell lines. We then validated the epigenetically activated lncRNAs by performing decitabine treatment on cell lines. The decitabine treatment led to a dosage and time-dependent induction of lncRNA expression and loss of DNA methylation at lncRNA promoter. Further survival analysis indicated that epigenetically activated lncRNAs were generally correlated with poor survival while the epigenetically silenced lncRNAs were generally correlated with good survival. Among the epigenetically silenced lncRNAs are well-documented tumor suppressor lncRNAs such as MEG3. And lncRNA oncogenes, such as MINCR, are epigenetically activated in multiple cancer types. To delineate the relationship between lncRNAs and known cancer pathways, we performed mutual exclusive analysis by combining the lncRNA epigenetic landscape with known protein-coding tumor genes alterations in same tumors. This analysis nominated several lncRNAs as novel members in cancer pathways, including TP53, PI3K, and EGFR cascades. Indeed, functional experiment validation focus on the top epigenetically activated lncRNA (EPIC1) demonstrated that EPIC1 is a potential oncogene. In breast and ovarian cancer cell lines, EPIC1 promotes cancer cell colony formation and increases cell viability by promoting cell cycle pathway. Collectively, our integrative study provides a resource for investigating lncRNAs in cancer and lays the groundwork for the development of novel diagnostics biomarkers and treatments. Citation Format: Da Yang, Zehua Wang, Bo Yang, Min Zhang, Zhiyuan Wu. The DNA methylation landscape of long noncoding RNAs in human cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3495. doi:10.1158/1538-7445.AM2017-3495