Identification and clinical assessment of novel biomarkers for sepsis-induced liver injury.

Abstract

BACKGROUND AND AIMS: Sepsis is marked by a dysregulated response to infection, causing circulatory dysfunction, organ failure, and high mortality. Liver dysfunction serves as an early indicator of sepsis severity. However, current interventions remain limited. This study aims to use LC-MS/MS to identify differential proteins, and preliminarily validate them as potential biomarkers for sepsis-induced liver injury (SILI). MATERIALS AND METHODS: We established SILI mouse models using Cecal Ligation and Puncture (CLP) and Lipopolysaccharide (LPS) intraperitoneal injection. And then ELISA, Western blot, and H&E staining were performed to analyze liver pathology and function. Through LC-MS/MS, we explored key proteins and their functional and biological characteristics. Next, we evaluated key proteins' performance in the validation cohort using ROC and external validation from external datasets (GSE60088, PRJEB20791, PXD037868, CNP0002611, PXD027485). Finally, CCK-8 and RT-PCR assays were performed on the LPS-induced in vitro SILI model. RESULTS: A total of 113 shared differential proteins were identified in the CLP model and LPS model. GO, KEGG and WikiPathways analysis highlighted their association with inflammation, protein cascade activation, and lipid metabolism. Among these, TRAF-interacting protein with Forkhead-associated domain (TIFA) and S100 calcium-binding protein A9 (S100A9) were selected for their central roles in inflammation and documented involvement in liver diseases. Plasma TIFA and S100A9 levels were significantly elevated in SILI patients compared to healthy control (HC), with consistent results in external datasets. S100A9, but not TIFA, was also higher in SILI than in disease control (DC). ROC analysis showed that both markers distinguished SILI from HCs, and the combination of TIFA, S100A9 and PCT effectively differentiated SILI from DCs, with their expression positively correlated with liver dysfunction. RT-PCR analysis further showed the transcription levels of TIFA and S100A9 upregulated in the cell model. CONCLUSION: Our findings indicate TIFA and S100A9 as potential biomarkers for SILI. And, we establish the combination of TIFA, S100A9 and PCT as a novel diagnostic model with high sensitivity and specificity, providing new insights into the SILI diagnosis.