Construction of a soluble mediator landscape and analysis of its clinical diagnostic potential in idiopathic pulmonary fibrosis.

BACKGROUND

Idiopathic pulmonary fibrosis (IPF) is a progressive and fatal interstitial lung disease. Current diagnosis primarily relies on imaging and histology, lacking sensitive and specific early biomarkers. Although soluble mediators play pivotal roles in IPF pathogenesis, their systematic characterization and clinical utility remain unclear. This study aimed to construct a soluble mediator-based marker atlas for IPF and evaluate its diagnostic potential.

METHODS

Multiple IPF-related transcriptomic datasets (GSE10667, GSE24206, GSE53845) were integrated from the Gene Expression Omnibus database to identify differentially expressed soluble mediator-related genes (SM-DEGs). Key diagnostic biomarkers were selected using multiple machine learning algorithms, including Least Absolute Shrinkage and Selection Operator, Random Forest, Recursive Feature Elimination, and Boruta, combined with protein-protein interaction network analysis. A diagnostic model was constructed and externally validated. The potential biological significance and therapeutic research value of the candidate genes were further explored through molecular subtyping, immune infiltration analysis, ceRNA network construction, and drug-gene association prediction, and the expression of the candidate diagnostic genes was detected by qRT-PCR in an in vitro IPF cell model.

RESULTS

Sixteen SM-DEGs were identified. By integrating multiple machine learning feature selection methods and PPI network analysis, Cadherin 2 and SPP1 (Secreted Phosphoprotein 1) were finally identified as candidate diagnostic-related feature genes. The infiltration levels of memory B cells, plasma cells, and regulatory T cells in Cluster 2 were higher, and these immune cells may participate in the development of IPF. Drug prediction based on the DGIdb and DrugBank databases showed that most candidate drugs have database-annotated drug-gene associations with SPP1. qRT-PCR demonstrated that CDH2 and SPP1 were significantly upregulated in Transforming Growth Factor-β1-induced human lung fibroblasts.

CONCLUSION

This study identified CDH2 and SPP1 as two candidate diagnostic-related genes for IPF, and provided a basis for further research on their clinical diagnostic value and potential biological roles.