Abstract

Purpose Rheumatoid arthritis (RA) is a common autoimmune disease causing significant bone lesions. This study aims to explore RA through efferocytosis, identify hub genes, and construct a risk prediction model for clinical management and targeted therapy. Method RA data was obtained from the GEO database. Differential gene expression analysis was performed using R packages. Hub genes were identified via LASSO regression and validated with ROC curves. Functional analysis included constructing PPI, mRNA-miRNA, and transcription factor (TF) networks. Gene modules were analyzed by WGCNA, and immune cell infiltration was assessed using CIBERSORT. CD300A expression was knocked down in macrophages and assessed by immunofluorescence and flow cytometry. In vivo , CD300A was silenced in a CIA mouse model using AAV lentivirus, followed by joint pathology and TUNEL staining. Result Six hub genes were identified: IL1R1, WASL, AIM2, NLRP3, CD24, and CD300A, with coefficients of -1.968, -0.445, 1.367, 0.077, 0.248, and 0.523, respectively. B cells, macrophages, and T cells were significantly correlated with the risk score. Additionally, 30 TFs and 56 candidate miRNAs were identified. Knockdown of CD300A in macrophages enhanced efferocytosis in vitro , while silencing CD300A in vivo reduced joint scores, improved joint pathology, and decreased apoptosis in CIA mice. Conclusion Efferocytosis plays a key role in RA pathogenesis. IL1R1, WASL, AIM2, NLRP3, CD24, and CD300A are significant hub genes for RA prediction, with CD300A being a promising therapeutic target.