Identification of copper metabolism-related biomarkers and exploration of mechanisms based on osteoarthritis transcriptomics data

Abstract

Abstract Background Studies have demonstrated that copper metabolism related genes (CMRGs) are tightly associated with a high risk of developing osteoarthritis (OA). However, the details of their regulation are not well understood. Hence, this research intends to explore the mechanism of CMRGs in OA and to provide new clues for the treatment of OA. Methods The GSE48556 and GSE63359 datasets were sourced from the Gene Expression Omnibus (GEO) database. The 133 CMRGs were collected from the literature. Differentially expressed genes (DEGs) between case and control cohorts in the GSE48556 dataset were identified through differentially expressed analysis. Moreover, differentially expressed-CMRGs (DE-CMRGs) were gained via overlapping DEGs and CMRGs. Then, we performed gene enrichment analysis for the DE-CMRGs to identify their regulatory functions. The DE-CMRGs with consistent and markedly divergent expression trends in both datasets were considered as biomarkers. Subsequently, we verified the results using real-time reverse transcription-PCR (qRT-PCR) in clinical blood specimen. Receiver Operating Characteristic (ROC) curves were mapped to assess the predictive accuracy. Finally, Gene Set Enrichment Analysis (GSEA), the Gene-Gene Interaction (GGI) network, immune-related function, and drug prediction were executed, then correlations between biomarkers as well as between biomarkers and immune-related pathways or cells were determined. Results Totally, 4,325 DEGs and 32 DE-CMRGs were selected in GSE48556 dataset, and functional enrichment analysis showed that they were involved in ‘response to copper ion’ and ‘copper ion binding’, which were consistent with the path of our research. KEGG, GSEA and GGI outcomes indicated that there were mainly involved in the pathways of ‘olfactort transduction’, ‘iron ion transport’, ‘ferroptosis’, ‘platinum drug resistance’ and so on. Through simultaneous screening of both datasets, four biomarkers (APP, CUTC, TFRC, and HEPH) were discovered. Then, all of area under curves (AUC) values of the ROC curves exhibited strong prediction accuracy. APP, CUTC and TFRC plasma levels were significantly higher in OA patients compared to controls (p < 0.05). However, the HEPH plasma level of OA patients was significantly decreased compared to controls (P < 0.01). According to correlation analysis, HEPH was positively connected with Th1 cells and the CCR immune path, and negatively correlated with APP, Th2 cells, and the check-point immune pathway. There were 35 drugs predicted by 4 biomarkers such as L-methionine (R)-S-oxide, Mercuribenzoic Acid and Copper. The expression levels of APP, CUTC, and TFRC genes in plasma of OA patients were dramatically lowered (P < 0.05) compared to the control, while the expression levels of HEPH genes were significantly elevated (P < 0.01). Conclusion Four biomakers (APP, CUTC, TFRC, and HEPH) were identified as CM biomarkers in OA, which offered a fresh standpoint to probe the connection between CMRGs and OA.