Exploration of molecular biomarkers in ankylosing spondylitis transcriptomics

Background: Inflammation of the spine and sacroiliac joints is a hallmark of the chronic, progressive inflammatory illness known as ankylosing spondylitis (AS). The insidious onset and non-specific early symptoms of AS often lead to delays in diagnosis and treatment, which may result in the onset of disability. It is therefore imperative to identify new biomarkers.

Methods: In this study, datasets GSE73754 and GSE25101 were derived from the Gene Expression Omnibus (GEO). Key genes were identified through differential expression analysis and weighted gene co-expression network analysis (WGCNA). A model was then established using LASSO regression, and then it was subjected to the receiver operating characteristic (ROC) curve analysis for evaluation of the diagnostic accuracy of the genes. Subsequently, immune infiltration analysis was conducted to demonstrate the immune infiltration status of the samples and the correlation between key genes and immune infiltration. Finally, the expression levels of key genes in peripheral blood mononuclear cells (PBMCs) and their correlation with clinical indicators were validated via RT-qPCR assay.

Results: Through WGCNA and differential expression analysis, 6 genes were identified. Ultimately, five key genes (ACSL1, SLC40A1, GZMM, TRIB1, XBP1) were determined using LASSO regression. The area under the ROC curve (AUC) for these genes was greater than 0.7, indicating favorable diagnostic performance. Immune infiltration analysis showed that AS was associated with infiltration levels of various immune cells. RT-qPCR validated that the expression of ACSL1, SLC40A1, GZMM, and XBP1 was consistent with the predictive model, whereas TRIB1 expression was contrary to the predictive model. Clinical correlation analysis of key genes revealed that ACSL1 was positively linked to hsCRP levels, GZMM was negatively linked to, hsCRP levels, and neutrophil absolute values, SLC40A1 was positively linked to ESR, hsCRP levels and neutrophil absolute values, and XBP1 was negatively linked to ESR, hsCRP levels, and neutrophil absolute values.

Conclusion: This study identified key genes that may reveal a potential association between AS and ferroptosis, demonstrating high diagnostic value. Furthermore, the expression levels of these genes in peripheral blood mononuclear cells (PBMCs) are strongly correlated with disease activity. These findings not only suggest potential biomarkers for the diagnosis of AS but also offer important references for exploring new therapeutic targets, highlighting their substantial clinical applicability.