Integrated transcriptomic and metabolomic analyses identify key metabolites and genes in sepsis-associated acute lung injury

Abstract

Introduction Sepsis-associated acute lung injury (SALI) is a life-threatening pulmonary complication of sepsis with high mortality. This study aimed to prioritize serum metabolites and genes potentially involved in SALI by integrating multi-omics datasets. Methods A rat SALI model treated with Xuanbai Chengqi Decoction (XCD) was subjected to untargeted serum metabolomics. Bulk and single-cell RNA-seq datasets were retrieved from GEO, and metabolite-related GWAS statistics were obtained from IEU OpenGWAS. scPagwas mapped metabolite genetic signals to cell populations; hdWGCNA identified cell signature genes. Candidate genes were filtered by overlapping bulk DEGs, ten machine-learning models and expression validation. Functional enrichment, immune infiltration, cell -cell communication, pseudotime and scTenifoldKnk virtual knockdown analyses were performed. Results Oleoylcarnitine showed the strongest genetic association with T cells. Cross-platform integration prioritized GIMAP4 and RCBTB2 as hub genes, with WBP2 retained for further validation. Their expression correlated with altered immune infiltration, especially eosinophils, and the three-gene profiles shared 62 biological processes. SALI exhibited increased number and strength of cell interactions compared with isolated sepsis. Five T-cell subsets were identified, and virtual knockdown of hub genes modified regulation of eight DEGs. Discussion Oleoylcarnitine-associated metabolic changes are correlated with T-cell dysregulation mediated by GIMAP4, RCBTB2, and WBP2 in SALI. These findings are descriptive, and further experiments are required to establish causal relationships.