Identification and Immune Characteristics of Cuproptosis-Related Genes in Osteoarthritis

Abstract

Background: Knee osteoarthritis (OA) is the most common osteoarthritic condition and a major contributor to disability. Increasing evidence indicates the pathological changes in the synovial membrane occur earlier than OA. Cuproptosis is a novel kind of copper-induced cell death. However, the role of cuproptosis-related genes (CRGs) in OA is still unknown. Our aim was to explore the role of CRGs in the expression and immunomodulation of osteoarthritis disease. Methods: The gene expression profile-related datasets including 27 normal samples and 27 OA samples were retrieved from the Gene Expression Omnibus (GEO) database. Cluster specific differentially expressed genes were identified using the weighted gene co-expression network analysis (WGCNA) algorithm. Based on least absolute shrinkage (LASSO), support vector machine–recursive feature elimination (SVM-RFE) and RandomForest (RF) algorithm, we constructed a CRG-based predictive model of osteoarthritis. Next, we used the nomogram, calibration curves to test the prediction power of our predictive model, used quantitative real-time PCR (polymerase chain reaction) (qRT-PCR) to verify differences of key genes expression in human synovial tissues. The functional enrichment of CRGs in OA was then evaluated using the gene set enrichment analysis (GSEA) approach. Finally, we constructed targeting drugs and ceRNA network based on online database. Results: We constructed a predictive model consisted of six CRGs (FDX1, CDKN2A, GSS, ATP7B, NDUFA1 and NDUFB1) by the LASSO, SVM-RFE and RF algorithm. The nomogram, calibration curves and qRT-PCR manifested that the predictive model had satisfactory performance. Analysis of immune infiltration revealed immune heterogeneity between OA and healthy individuals. Further immune correlation analysis showed that four genes (FDX1, CDKN2A, GSS and ATP7B) was closely associated with the immune status of OA. Single-cell RNA sequencing analysis showed that the cells type of the six CRGs aggregation was mainly related to immunity. Finally, we identified CDKN2A, ATP7B and five miRNAs (has-miR-30b-3p, has-miR-3163, has-miR-570-3p, has-miR-548x-3p, and has-miR-576-5p) as potential therapeutic targets. Conclusions: We systematically evaluated the immunological traits between individuals with normal tissues and OA tissues and the CRGs that were closely associated with immunity in OA. In addition, we constructed a CRG-based predictive model to disclose the patients of OA with higher immune scores. These findings may offer additional evidence for the function of cuproptosis in the immune regulation of OA and indicate that CRGs are potential diagnostic biomarkers of OA.