Integrated transcriptomic and methylome analysis reveals retinoic acid pathway activation after decitabine treatment in EBV associated gastric cancer

Epstein-Barr virus associated gastric cancer (EBVaGC) accounts for ~9–10% of gastric cancers worldwide and is defined by a distinctive molecular profile, including extreme hypermethylation of the DNA. Targeting this aberrant methylation may be a potential therapeutic strategy. EBV + gastric cancer cell lines (YCCEL1, SNU719) and EBV - lines (AGS, SNU16, MKN74) were treated with a DNA methyl-transferase inhibitor (DNMT), decitabine (DCB), for three days followed by RNA sequencing to identify EBV-specific responses. DNA methylation profiling by reduced representation bisulfite sequencing (RRBS) was performed in EBV + cell-lines and integrated with expression data to identify epigenetically regulated networks. While DCB induced broad transcriptional changes across all lines, EBV + cells exhibited the strongest transcriptional response, sharing many upregulated genes. Many of these EBV + specific genes were expressed at lower baseline levels in EBV + tumors from TCGA. DCB predominantly reduced methylation at highly methylated intergenic CpGs, with a subset of promoters undergoing significant demethylation. Integrated analysis revealed a strong inverse correlation between promoter demethylation and gene expression, implicating multiple cancer-relevant pathways. Upstream regulator analysis and motif enrichment indicated that regions losing methylation were enriched for retinoic acid receptor α (RARα) binding motifs, suggesting that DCB-mediated demethylation restores RA pathway accessibility and transcriptional activity. Further, inhibiting RAR signaling reduced DCB induced apoptosis. Although DCB can induce both host gene re-expression and viral lytic gene activation in EBV-positive tumors, its impact on RA signaling in EBVaGC has not been studied. Decitabine promotes extensive epigenetic reprogramming in EBVaGC, with preferential effects in CIMP-positive, EBV-infected cell lines.