Abstract
Aim: The methylome associated with miRNA loci was investigated in oral cancer to explore tobacco specific methylation and potential biomarkers for patient survival. Methods: Methylome data was generated from 16 pairs of cancer-normal tissues by reduced representation bisulfite sequencing method. Differentially methylated regions were identified using the DMAP pipeline. In silico validation and Kaplan-Meier survival analyses were performed on The Cancer Genome Atlas data based on our miRNA methylome data. Results: A total of 4310 unique differentially methylated regions, mapping to 144 miRNA loci, were identified. Three distinct groups of miRNAs were differentially methylated in cancer tissues from smokers, chewers and mixed habitues. Hypermethylation of miR-503, miR-200a/b, miR-320b and miR-489 was associated with worse 5-year survival. Conclusion: Differential methylation patterns in miRNA loci are associated with poor survival underscoring their potential as predictive and prognostic biomarkers in oral cancer.