Identification of genes associated with a macrophage polarization-related transcriptional signature as biomarkers for tendinopathy based on bioinformatics analysis

Tendinopathy is a common overuse-induced musculoskeletal disorder characterized by chronic pain, collagen disorganization, and immune cell infiltration. Macrophage polarization plays a critical role in tendon inflammation and tissue repair, but its diagnostic potential in tendinopathy remains largely unexplored. We utilized two publicly available gene expression datasets, GSE26051 and GSE180836, sourced from the GEO database, serving as the training and screening cohorts, respectively. Importantly, differential expression results from GSE180836 contributed to candidate gene discovery, and this cohort was therefore not used as a validation cohort; the only independent validation of the biomarkers was performed by RT-qPCR in a clinical cohort. Genes associated with a macrophage polarization-related transcriptional signature were integrated with weighted gene co-expression network analysis (WGCNA) and differential expression analysis to identify differentially expressed genes associated with a macrophage polarization-related transcriptional signature (MPTSGs). Three machine learning algorithms—LASSO, SVM-RFE, and Boruta—were employed to screen for candidate biomarkers. A diagnostic nomogram was constructed and validated using receiver operating characteristic (ROC) and calibration curves. Additionally, functional enrichment, immune infiltration, Gene Set Enrichment Analysis (GSEA)/Gene Set Variation Analysis (GSVA), and drug-gene interaction analyses were conducted. Three biomarkers (PTCHD1, NEURL1, and ST8SIA5) were identified through integrative bioinformatics analysis and were significantly downregulated in the training cohort (GSE26051). In the independent clinical cohort (n = 30), RT-qPCR analysis confirmed that PTCHD1 and NEURL1 were significantly down-regulated in tendinopathy samples compared to controls (both P < 0.01). ST8SIA5 also exhibited a downward trend, although the difference did not reach statistical significance (P = 0.074). The nomogram model achieved an AUC > 0.7 with good calibration. GSEA revealed their involvement in key signaling pathways, including PI3K-Akt and ECM-receptor interaction. A total of 16 potential therapeutic compounds were identified via drug-gene network analysis. Our study identifies PTCHD1 and NEURL1 as robust biomarkers associated with macrophage polarization in tendinopathy, and provides new insights into its immunopathogenesis and therapeutic targeting. ST8SIA5 is proposed as a potential candidate showing a consistent trend, warranting further investigation in larger cohorts.