Identification and validation of SUMOylation-related key genes for osteoarthritis through integration of single-cell, bulk RNA sequencing and animal model experiments

Background Osteoarthritis (OA) existed as a degenerative arthropathy that involved the cartilage and many of its surrounding tissues. SUMOylation was an important post-translational modification. Some studies had confirmed that SUMOylation was associated with the occurrence of OA. However, there were relatively few related studies, and the explicit action mechanism of SUMOylation-related genes (SRGs) in OA remained unclear. The objective of this research was to detect and verify key genes related to SRGs in OA. Methods Firstly, GSE51588 (GPL13497), GSE55235 (GPL96) and GSE152805 and SRGs were from public databases. Subsequently, key genes were determined by differential expression analysis, PPI network, machine learning, expression analysis and ROC curves. Following this, using key genes, functional enrichment, immune infiltration, and drug prediction were implemented. Meanwhile, the identification of key cells was achieved by conducting single-cell RNA sequencing (scRNA-seq) analysis. Finally, a rat model (three OA and three sham samples) was established, and the expressions of key genes were verified. Results NUP98 and TOP1 were identified as key genes. Later, the two key genes were co-enriched in multiple pathways, like “ribosome.” Significant differences in 14 key immune cells (like monocytes) were observed between the OA and control groups. Furthermore, the drug-key gene network indicated that eight drugs (like luteolin) were associated with two key genes. Moreover, by scRNA-seq analysis, homeostatic chondrocytes (HomCs) were recognized as key cells. During the differentiation of HomCs, the expression level of both NUP98 and TOP1 first increased and then remained stable. Finally, the expression of two key genes at both gene and protein levels were notably elevated in the sham group. Conclusion In this study, two key genes and one key cell type were identified, which provided new directions for progress in OA research.