Autophagy is a common metabolic process in eukaryotes that maintains homeostasis in the intracellular environment, and as an important self-protection mechanism, autophagy participates in various metabolic processes in the body and affects the development of diseases[30]. Some studies suggest that autophagy participates in oxidative stress reactions, helping to mitigate and eliminate damage caused by oxidative stress, thereby slowing down the process of cell death[31]. Biomarkers have been considered important tools for cancer diagnosis and targeted treatment. Recently, several biomarkers have been used for the early detection and treatment of RA.
A total of 101 RA-related DEGs were analyzed in this study. To further study autophagy-related functions, 58 ARDEG were identified, and ARDEG enrichment analysis was performed using GO, KEGG, and GSEA. The findings revealed that the GO analysis was mainly enriched in BP (leukocyte migration, calcium ion homeostasis, and hypophosphite differentiation), CC (external side of the plasma membrane), and MF (receiver light activity, chemokine activity, and metalloproteinase activity). KEGG pathway enrichment analysis revealed associations with immune-related pathways, notably the IL-17 signaling pathway. GSEA enrichment analysis further demonstrated enrichment in pathways such as Reactome_Pi3k_Events_ In_Erbb2_ Signaling, Wp_Mapk_Pathway_In_Congenital_Thyroid_Cancer, and Reactome_Chaperone_ Mediated_Autophagy, among others. These findings suggested that immune or autophagy pathways play an effective role in the treatment of RA. Furthermore, we screened five hub genes in the training set using different PPI algorithms, and found their AUC values to be greater than 76.9%; these genes were significantly differentially expressed in different datasets through comprehensive observation, demonstrating that the test performance of the five established models still requires enhancement, despite their strong verification performance. In addition, we identified many corresponding targets from the hub gene-TF, hub gene-RBP, and hub gene-drug interaction groups, which provided a strong basis for follow-up studies of RA and identified potential targeted drugs.
CXCL10 and CXCL9 are chemokines with many biological functions such as strong recruitment of neutrophils, promotion of cytokine secretion, and inhibition of tumor growth[32]. Chemokines selectively recruit and activate monocytes and lymphocytes in synovial tissues, mediate inflammatory reactions, and play a pivotal role in RA pathogenesis [33]. In a mouse model of CIA, CXCL10 was found to be expressed in the synovial fluid of joints, and synovitis in mice improved after intra-abdominal injection of an anti-CXCL10 antibody[34]. RANKL can promote the expression of CXCL10 in osteoclasts. In RA, CXCL10 mediates RANKL expression in CD4 + T cells through CXCR3[35]. In addition, under the stimulation of injury factors, the body can chemotactically express CXCL9, cause effector T lymphocytes to undergo a large number of directional migrations, and participate in the development of RA[36]. Some clinical trials conducted on patients with RA have demonstrated that CXCL9 and CXCL10 are expressed in the synovial fluid and synovial tissues, suggesting their potential involvement in the pathogenesis of RA. In synovial tissues of a mouse model of inflammatory arthritis, it was observed that CXCL10 could induce RA bone destruction by stimulating the expression of NF-kB receptor activating factor ligand and tumor necrosis factor on CD4 + T cells. The degree of bone and joint destruction in mice was significantly reduced after adding an antagonist, suggesting that CXCL10 might be associated with bone erosion[37]. At present, compared with CXCL10, there are few studies on the mechanism of CXCL9's action in bone destruction. With an increasing understanding of bone tissue homeostasis, the role of CXCL9/CXCL10 chemotaxis in RA bone metabolism could be used as a direction for further exploration in the future[38]. Some studies have shown that decreased autophagy levels lead to increased secretion of inflammatory molecules, such as chemokines, which is an important regulatory mechanism for inducing and aggravating diseases[39]. IL-7 activates the JAK/STAT5PI3k/Akt/mTOR signaling pathway, thereby facilitating T cell survival and promoting cell cycle progression. IL-7R/JAK signaling mediated by IL-7 and IL-7R is necessary for the development and homeostasis of normal T-cells[40]. According to an epidemiological analysis, rs6451231 and rs969129 in IL-7R are associated with an elevated risk of RA, indicating that such IL-7R mutations could potentially serve as susceptibility loci for RA in the Han population[41]. However, the mechanism of studying IL7R in RA has not yet been elucidated. In addition, some studies have shown that the degradation of IL7R through autophagy is essential for T cell proliferation [42]. As a serine protease, GZMB is expressed by cytotoxic T lymphocytes and NK cells. Many studies have indicated a significant upregulation of GZMB in RA, and its expression is associated with the occurrence and progression of RA[43]. It is worth mentioning that GZMB has certain functions in tumor immune infiltration such as skin cutaneous melanoma and colorectal cancer[44, 45]. Studies have shown that CD8 cells of the cloned and expanded GZMB subgroup exist in the synovium of RA patients, and cytotoxic CD8 + T cells targeting the citrulline antigen are conducive to the treatment of synovitis and joint tissue destruction in RA[46]. Similarly, CD2, a novel protein, has not been reported in RA; however, some investigations have demonstrated its potential association with immune inflammation[47], which is also involved in RA pathogenesis. An earlier report indicated that the downregulation of CD2 might weaken the anti-tumor T cell response in colorectal, endometrial, and ovarian cancers and even offset the benefits of PD-1 immunotherapy[48]. At present, the autophagy pathways involving GZMB and CD2 have not yet been reported.
In our study, immune cell infiltration in the microarray dataset of pituitary tumors was estimated using the CIBERSORT algorithm[49]. After conducting an analysis, it was observed that in the RA dataset, there were differences in the activation of NK cells, CD4 memory rest T cells, follicular helper T cells, CD8 T cells, and plasma cells. However, this result is inconsistent with the findings reported by Yu et al.[50]. Only the plasma cells were similar, which needs to be confirmed in follow-up studies. Current evidence suggests that these immune cells could be used in treatments for RA, such as T-cell and B-cell depletion therapies. Antigen-presenting cells (APCs) are essential for the activation of Th17 cells, and determine T cell polarization through their cytokine release patterns [51]. These findings provide novel ideas for selective targeting of chronic destructive arthritis and other inflammatory diseases. This could potentially explain the limited sample size of the GEO database and the differences observed between different datasets. In the present study, five hub genes with diverse functions were identified in patients with RA. The theoretical findings presented in our study were derived from retrospective datasets and it is imperative to conduct further investigations with larger clinical samples for validation purposes.