Identification of Long-Range Epigenetic Silencing on Chromosome 15q25 and Its Clinical Implication in Gastric Cancer.

American Journal of Pathology

PubMedID: 25576785

Kang JY, Song SH, Yun J, Jeon M, Cha Y, Lee SH, Kim HP, Jeong EG, Han SW, Cho NY, Kook MC, Kang GH, Kim TY. Identification of Long-Range Epigenetic Silencing on Chromosome 15q25 and Its Clinical Implication in Gastric Cancer. Am J Pathol. 2015;185(3):666-78.
Recent genome-wide epigenomic and transcription profiling studies have demonstrated that epigenetic silencing can encompass multiple neighboring genes, termed as long-range epigenetic silencing (LRES). Herein, we identified a novel LRES region by comparing gene expression of human colon cancer HCT116 cells with their DNA methyltransferase 1 and DNA methyltransferase 3B double-knockout derivative double-knockout cells. Ten consecutive genes spanning 3 Mb of chromosome 15q25 were coordinately silenced, with eight genes showing promoter CpG island hypermethylation and enrichment of repressive histone marks, which were evaluated by bisulfite sequencing analysis and chromatin immunoprecipitation assay. Comparison of primary gastric tumor specimens with normal tissue confirmed that the long-range silencing of this region was tumor specific. Methylation of genes within the LRES region was evaluated in 190 gastric tumor tissues using the MethyLight assay, and their association with clinicopathological features, such as older age, high-grade differentiation, and diffuse or mixed-type histology, was determined. LRES-positive gastric cancer patients (six or more methylated genes) showed lower recurrence and better survival. Our findings emphasize the differential dynamics of DNA methylation and histone modification, indicating the importance of studying the relationship of each epigenetic modification in the context of chromatin domains. Patients with LRES showed lower recurrence and better prognosis, indicating that stratifying patients according to underlying molecular features, such as LRES regions, may better predict recurrence and survival.