Abstract: BACKGROUND: It is known that N6-methyladenosine (m6A) plays a role in the pathogenesis of various diseases and studies have suggested its involvement in the pathologic changes of steroid-induced femoral head necrosis (SNFH). However, research on m6A methylation modifications in steroid-induced femoral head necrosis is limited.
OBJECTIVE: Using bioinformatics methods to identify the differential expression of m6A-related genes in steroid-induced femoral head necrosis and to predict miRNAs associated with these genes to further elucidate the role and mechanism of m6A methylation in steroid-induced femoral head necrosis.
METHODS: Differential gene expression between steroid-induced femoral head necrosis and control groups was analyzed using GSE123568 gene expression data and identified using the “limma” package in R. Functional enrichment analysis was performed on the differentially expressed genes. Differential analysis of the related genes was carried out using the “ggstatsplot” package in R. The differential genes were cross-validated using the GSE74089 dataset. An mRNA-miRNA regulatory network was constructed, and co-expression analysis was performed on the module genes followed by enrichment analysis. Differences in immune cell infiltration between steroid-induced femoral head necrosis and control groups were quantified using the ssGSEA method.
RESULTS AND CONCLUSION: Correlation analysis revealed 13 m6A-related genes, and further analysis through the protein-protein interaction network identification and receiver operating characteristic curve analysis showed that YTHDF2 was expected to be a core differential gene as a potential early biomarker. Enrichment analysis indicated that differentially expressed genes were mainly involved in inflammation and immune response and were closely related to osteoclasts. Cross-validation analysis showed that differential gene expression results between the two datasets were consistent. mRNA-miRNA regulatory network analysis revealed that YTHDF2 was negatively correlated with miRNA-27a. Immune infiltration analysis revealed an increase in immune cell infiltration in steroid-induced femoral head necrosis, and YTHDF2 was positively correlated with the infiltration of CD4+T cells. To conclude, m6A-related gene YTHDF2 can serve as a potential biomarker of steroid-induced femoral head necrosis and is valuable for the early clinical diagnosis and treatment of steroid-induced femoral head necrosis. The negative correlation between YTHDF2 and mir-27a and the positive correlation between YTHDF2 and CD4+T cell infiltration provide new insights into the early diagnosis and treatment of steroid-induced femoral head necrosis and shed light on the mechanism of m6A in steroid-induced femoral head necrosis.

Key words: m6A methylation, steroid-induced femoral head necrosis, bioinformatics analysis, differential gene,  , YTHDF2, enrichment analysis