Effect of DNA methylation on gene transcription is associated with the distribution of methylation sites across the genome in osteoarthritis

Genetics and epigenetics are important subjects in the field of osteoarthritis (OA) research. DNA methylation may affect gene transcription, but the specific mechanisms have remained to be fully elucidated. In the present study, the ChAMP methylation analysis package was used to identify differentially methylated genes (DMGs) from the dataset GSE63695 from the Gene Expression Omnibus (GEO) database. The distribution of differentially methylated sites (DMS) and the total array sites across the genome were analyzed by enrichment analysis. Subsequently, two mRNA expression profiling datasets, GSE114007 and GSE113825, were obtained from the GEO database and common differentially expressed genes (DEGs) were identified using the Limma package. Key genes were screened by analyzing the distribution of DMS across the genome consisting of DEGs and DMGs. A total of 1,662 and 1,986 DEGs were identified between OA and normal human cartilage from the GSE113825 and GSE114007 dataset, respectively. A further screening revealed 292 genes with common differences between the two datasets. A total of 574 DMS containing 394 DMGs were observed between OA and normal cartilage. Integrative analysis revealed a corresponding subset of 15 genes. Of these, 6 genes were verified by reverse transcription-quantitative PCR, confirming that the mRNA expression of 5 genes (MAP1B, FNDC1, ANLN, SCNN1A and STC2) in OA cartilage was consistent with the mRNA expression from the analysis of the datasets. Upon treatment with the DNA methylation inhibitor 5-aza-2'-deoxycytidine, the mRNA levels of FNDC1 and SCNN1A were decreased, and no significant alteration in the mRNA levels of MAP1B, ANLN, KCNN4 and STC2 was observed. The incidence of differential methylation varied in subregions of the genome and the effects on transcription were associated with the distribution of DEGs across the genome. The regulation of this appears more complex than initially postulated. Combining the data on epigenetic differences of OA with the genome or transcriptome data for analysis may improve the understanding of the pathophysiological processes of OA. FNDC1 and SCNN1A may potentially be valuable biomarkers for OA.