Pan-Cancer Landscape of Aberrant DNA Methylation across Human Tumors

• Published in: Cell reports (Cambridge) Vol. 25; no. 4; pp. 1066 - 1080.e8
• Main Authors: Saghafinia, Sadegh, Mina, Marco, Riggi, Nicolo, Hanahan, Douglas, Ciriello, Giovanni
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 23.10.2018; Elsevier
• Subjects: aberrant DNA methylation; cancer epigenetics; DNA methylation instability; pediatric cancer; TCGA pan-cancer cohort; TCGA pan-cancer cohort; DNA methylation instability; cancer epigenetics; aberrant DNA methylation; pediatric cancer
• ISSN: 2211-1247; 2211-1247
• DOI: 10.1016/j.celrep.2018.09.082

The discovery of cancer-associated alterations has primarily focused on genetic variants. Nonetheless, altered epigenomes contribute to deregulate transcription and promote oncogenic pathways. Here, we designed an algorithmic approach (RESET) to identify aberrant DNA methylation and associated cis-transcriptional changes across >6,000 human tumors. Tumors exhibiting mutations of chromatin remodeling factors and Wnt signaling displayed DNA methylation instability, characterized by numerous hyper- and hypo-methylated loci. Most silenced and enhanced genes coalesced in specific pathways including apoptosis, DNA repair, and cell metabolism. Cancer-germline antigens (CG) were frequently epigenomically enhanced and their expression correlated with response to anti-PD-1, but not anti-CTLA4, in skin melanoma. Finally, we demonstrated the potential of our approach to explore DNA methylation changes in pediatric tumors, which frequently lack genetic drivers and exhibit epigenomic modifications. Our results provide a pan-cancer map of aberrant DNA methylation to inform functional and therapeutic studies. [Display omitted] •Mutations in WNT and chromatin modifiers are associated with DNA methylation instability•Concurrent genetic and epigenomic changes converge on the same pathways•Cancer-germline antigen expression correlates with response to anti-PD1•Diffuse anaplasia Wilms tumors exhibit DNA methylation instability Saghafinia et al. present an algorithmic approach to identify cancer-associated DNA methylation changes affecting gene expression. By analyzing >6,000 adult and pediatric tumors, the authors identify numerous DNA methylation changes at gene promoters that coalesce into a few pathways and that were associated with patient prognosis and response to therapy.