30 October 2026 in Endocrine, Metabolic & Immune Disorders - Drug Targets
Introduction: Acute respiratory distress syndrome (ARDS), frequently triggered by sepsis, is associated with high mortality and severe disruption of the alveolar-capillary barrier driven by immune dysregulation. This study aimed to characterize immune-stromal crosstalk in sepsis-induced ARDS and develop a predictive risk model. Methods: Bulk transcriptomic datasets were used for differential expression analysis, immune deconvolution, and development of a sepsis-to-ARDS risk model. eQTL-based Mendelian randomization evaluated associations between genetically predicted gene expression and sepsis susceptibility. Human and mouse single-cell datasets were used for pseudotime and inferred ligand-receptor communication analyses. Results: Ten core genes, including CD52 and MACF1, were prioritized by integrating ARDSassociated differential expression with genetic associations for sepsis susceptibility. CD52 showed dynamic expression changes in CD8+ T cells. CD52-low states were associated with stronger inferred MIF- and TGF-β-related communication and with transcriptional programs linked to immune recruitment and barrier dysfunction. The four-gene signature showed moderate discrimination in the external-validation cohort (AUC = 0.688). Discussion: These findings identify CD52 as a marker of a dynamic immunoregulatory state in sepsis- induced ARDS and associate CD52-low states with altered inferred T cell-stromal communication. The four-gene signature provides proof-of-concept support for transcriptomic risk stratification but requires further clinical validation. Conclusion: CD52 marks a time-dependent immune state associated with inferred T cell-stromal communication in sepsis-induced ARDS. The four-gene signature provides proof-of-concept support for risk stratification, while functional studies are required to determine whether CD52 directly modifies barrier-related signaling.