Prognostic significance of 5-methylcytosine-related signature genes as biomarkers in oral cancer

Purpose To evaluate the prognostic value of RNA 5-methylcytosine (m5C)-related biomarkers in oral squamous cell carcinoma (OSCC). Method The TCGA-OSCC cohort (n = 324) was used as the discovery cohort, and GSE41613 (n = 97) served as the independent validation cohort. Gene set variation analysis (GSVA) quantified m5C scores in The Cancer Genome Atlas (TCGA) OSCC cohort. Kaplan–Meier analysis assessed survival, and single-sample gene set enrichment analysis (ssGSEA) evaluated immune infiltration. Weighted gene co-expression network analysis (WGCNA), differential expression, and Cox regression constructed the prognostic signature. Drug sensitivity and Gene Ontology (GO)/Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were conducted. Biomarker expression was validated by reverse transcription quantitative PCR (RT-qPCR), Western blotting, and immunofluorescence. Result Patients with a lower level of m 5 C had a longer lifespan than those with a high level of m 5 C. Different immune cells in the tumour microenvironment were correlated with m5C levels. Integration of m 5 C score-associated WGCNA modules with tumor-versus-normal DEGs identified 217 m 5 C-related genes, from which Cox regression further selected ZNF662 and HMMR to establish the dual-gene prognostic signature. Pharmacological sensitivity analysis showed that the therapeutic effects of the high-risk and low-risk groups were not the same. Functional enrichment showed neuroactive ligand-receptor interaction and calcium-signaling pathways. RT-qPCR, Western blotting and immunofluorescence confirmed decreased ZNF662 and increased HMMR expression in OSCC cell lines, mirroring the low-ZNF662/high-HMMR expression pattern associated with the high-risk group in silico. Conclusion The ZNF662/HMMR signature represents a parsimonious m5C score-associated prognostic model that may complement conventional clinicopathological assessment for OSCC risk stratification and help identify high-risk patients who may require closer surveillance and individualized management, pending prospective validation.