Abstract
BACKGROUND: Sepsis has a high morbidity and mortality rate in critically ill patients, and acute respiratory distress syndrome (ARDS) is one of its most common outcomes. However, there is still no effective biomarker to predict the risk and outcome of ARDS induced by sepsis. METHODS: In this research, the GSE32707 dataset was acquired from the Gene Expression Omnibus (GEO) database and used to identify differentially expressed genes (DEGs). The extracellular protein-related differentially expressed genes (EP-DEGs) were filtered using the Human Protein Atlas (HPA) and UniProt databases. Functional and pathway analyses of the EP-DEGs were conducted through Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses. Additionally, hub genes were identified using STRING, Cytoscape, MCODE, and Cytohubba. The expressions of the hub genes were analyzed in both the training set (GSE32707) and the validation set (GSE66890). The diagnostic potential of lymphocyte subsets was evaluated through ROC curve assessment in the clinical cohort. RESULTS: We identified 86 EP-DEGs from DEGs. These EP-DEGs were found to be significantly enriched in leukocyte mediated immunity. We also identified 5 key extracellular protein genes GNLY, GZMK, CST7, PTPRC and CD19. CD19 expressions were increased in both training and validation sets. ROC curves showed that CD19 expression had a higher accuracy in the diagnosis of sepsis-induced ARDS. Lymphocyte subsets analysis of clinical samples revealed that CD19(+)B cells were elevated in sepsis-induced ARDS, with CD19(+)B cell counts demonstrating a higher diagnostic accuracy (AUC = 0.829) for septic-ARDS compared to other lymphocyte subsets. CONCLUSION: In this study, we employed bioinformatics approaches to identify potential biomarkers for sepsis-induced ARDS and further validated these findings using clinical samples. Our results suggest that peripheral CD19(+) B cells could act as a promising biomarker in sepsis-induced ARDS.