Diverse Programmed Cell Death Patterns and Diagnostic Value in Osteoarthritis Analyzed by Integrative Multiomics Data.

INTRODUCTION

Chronic pain and high disability rates caused by osteoarthritis (OA) significantly impact quality of life. Programmed cell death (PCD) plays a crucial role in OA pathogenesis; however, a comprehensive analysis of PCD patterns in OA is lacking, limiting understanding of their potential role.

METHODS

Batch transcriptomic data related to OA were obtained from the GEO database. Differential expression analysis based on 13 PCD patterns was performed to identify differentially expressed genes (DEGs). Unsupervised clustering algorithms were applied to define molecular subtypes associated with OA. The CIBERSORT algorithm was used to analyze the immune microenvironment and evaluate the immunological relevance of each cluster. Hub PCD-related DEGs were identified using a combination of machine learning algorithms, and an OA diagnostic model was constructed based on these hub genes. Single-cell RNA sequencing (scRNA-seq) dataset GSE169454 was used to classify OA chondrocytes into distinct cell clusters. The AddModuleScore function calculated PCD scores, allowing evaluation of hub-gene expression and PCD variation across clusters. Expression levels of ten hub PCD-related DEGs were further validated in OA cells and rat models.

RESULTS

Differential expression analysis identified 61 PCD-related DEGs. Unsupervised clustering revealed two molecular subtypes of OA (cluster 1 and cluster 2), with immune-related pathways significantly enriched in cluster 1, including potential NK cell activation. Ten hub PCD-related DEGs were identified using three machine learning algorithms, leading to a highly effective diagnostic model (AUC = 0.993). Five distinct cell types were identified in OA chondrocytes, with higher PCD scores observed in OA samples and the HomC subgroup. Functional analysis indicated significant enrichment of the PI3K-AKT signaling pathway in high BINP3-expressing preHTCs, associated with extracellular matrix composition. mRNA and protein levels of the ten hub DEGs were confirmed in animal models.

CONCLUSION

This multi-omics analysis of 13 PCD patterns provides a preliminary evaluation of the diagnostic and classification value of PCD-related genes in OA, highlighting potential biomarkers for clinical application.