Regulatory roles of 13 types of RNA modifications in osteoarthritis: based on bulk and single-cell RNA analysis.

Abstract

Osteoarthritis (OA) is a prevalent degenerative joint disease characterized by pain, joint deformity, and disability, emerged as a primary cause of disability in the elderly. This study aims to investigate the regulatory roles of 13 RNA modification patterns in OA through integrated multi-omics analysis. Genes associated with 13 types of RNA modification patterns were selected for analysis. Pathway analysis was conducted using single-sample gene set enrichment analysis (ssGSEA), and differentially expressed genes (DEGs) were identified. DEGs were further screened using random forest (RF) and support vector machine-recursive feature elimination (SVM-RFE) to develop a diagnostic model. Internal validation of the model was performed via the bootstrap algorithm. The correlation between key genes and specific immune cell types was assessed through immune infiltration analysis. Consensus clustering and weighted gene co-expression network analysis (WGCNA) were utilized to identify key subtypes and regulatory modules. Single-cell analysis was employed to further investigate the expression of modification-related genes in different chondrocyte populations. Cell-cell communication patterns among chondrocytes were characterized using CellChat. Monocle was used to construct chondrocyte developmental trajectories, and SCENIC was applied to detect cell type-specific regulatory networks. Expression of genes associated with the 13 RNA modification patterns was analyzed in OA. ssGSEA revealed significant downregulation of N1-methyladenosine (m1A), N6-methyladenosine (m6A), and uridylation, and significant upregulation of N7-methylguanosine (m7G) and pseudouridine in OA. By integrating RF and SVM-RFE, YTHDC1, RBBP6, LSM1, EIF3D, METTL3, ELP2, and IGF2BP2 were identified as key diagnostic genes. Among them, YTHDC1, METTL3, and IGF2BP2 are associated with m6A, while LSM1 and EIF3D are associated with m7G. Internal validation using the bootstrap algorithm demonstrated the model's stable sensitivity and specificity. Further immune infiltration analysis showed that YTHDC1, METTL3, and IGF2BP2 (m6A-related) exhibit immunosuppressive effects in the OA immune microenvironment, whereas LSM1 (m7G-related) promotes immune cell activation and inflammatory responses. Consensus clustering identified two distinct subtypes; subtype 1, characterized by high LSM1 expression, exhibited higher inflammatory levels and OA positivity rates. Single-cell analysis revealed downregulation of m6A in RegC (regulator chondrocytes) and dynamic changes in YTHDC1, METTL3, and IGF2BP2 with inflammation progression. LSM1 was highly expressed in EC (effector chondrocytes) and LSM1 + ECs communicated with HomC through pathways such as FGF-FGFR. SCENIC analysis showed that YTHDC1, METTL3, and IGF2BP2 were mostly positively correlated with transcription factors, while LSM1 was mostly negatively correlated. SOX5, MEF2A, JUNB, JUND, and CEBPD were identified as specific transcription factors for RegC in OA. This study comprehensively analyzed the regulatory roles of 13 RNA modification patterns in OA and identified key modifications and regulatory genes. YTHDC1, METTL3, and IGF2BP2 (m6A-related) exhibit immunosuppressive effects in OA, while LSM1 (m7G-related) promotes inflammation, contributing to the precise diagnosis and targeted therapy for OA.

Supplementary information: The online version contains supplementary material available at 10.1007/s13205-025-04448-6.