Integrated bioinformatics analysis of molecular signatures and therapeutic targets in early sepsis

This study aimed to identify key molecular signatures and therapeutic targets in early sepsis through integrated bioinformatics analysis.

We analyzed three independent blood transcriptomic datasets (GSE95233, GSE137340, GSE57065) from the Gene Expression Omnibus. Differential expression, functional enrichment, and protein-protein interaction network analyses were performed to identify overlapping differentially expressed genes (DEGs) and hub genes. The Connectivity Map (cMAP) database was queried to predict potential therapeutic compounds. An independent dataset validated the diagnostic performance and cellular correlates of the hub genes.

We identified 330 overlapping differentially expressed genes enriched in immune pathways like T-cell receptor signaling. Ten hub genes central to T-cell function were pinpointed (CD4, CD247, CD3E, CD2, FYN, ZAP70, CD3G, ITK, LAT, CD5). Independent validation confirmed that these hub genes were significantly down-regulated in sepsis patients, with their expression levels strongly correlating with T-cell exhaustion and inversely correlating with myeloid cell infiltration. While these genes demonstrated excellent diagnostic accuracy (AUC: 0.908-0.999), this high predictive performance likely reflects differences in immune cell composition rather than disease-specific molecular signatures. Additionally, cMAP analysis predicted six potential therapeutic agents (e.g., anastrozole, etofenamate), offering a theoretical framework for drug repurposing in sepsis.

This study independently validates the significant down-regulation of T-cell-related genes in early sepsis, reflecting a profound disruption of adaptive immunity. Our findings confirm the robustness of these genes as surrogate markers of early immune imbalance, establishing a critical foundation for future investigations into T-cell-focused immunomodulatory strategies in sepsis.