AUTHOR=Yan Qin , Fan Wei , He Xinsen , Zheng Shi , Zhong Xiaolin 

TITLE=Transcriptional responses in different mouse models of septic liver injury differ from those in patients with septic liver injury

JOURNAL=Frontiers in Immunology

VOLUME=Volume 16 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2025.1556392

DOI=10.3389/fimmu.2025.1556392

ISSN=1664-3224

ABSTRACT=IntroductionSepsis, particularly sepsis-induced liver injury (SLI), exhibits acute onset and high mortality (up to 80%). While murine models are widely used for mechanistic studies due to limited human sample availability, their accuracy in replicating human SLI pathophysiology remains debated. MethodsHuman SLI transcriptomes were characterized to identify core genes and immune signatures using Venn analysis and immune infiltration profiling. Transcriptomic features of two murine SLI models—cecal ligation and puncture (CLP) and lipopolysaccharide (LPS) challenge—were benchmarked against human SLI to evaluate pathophysiological relevance. Both models were then utilized to validate core gene expression for SLI biomarker identification. ResultsHuman SLI transcriptomics revealed significant enrichment in apoptotic processes, NF-κB regulation, inflammatory responses, protein phosphorylation, and bacterial response. Key pathways included IL-17 signaling, antigen processing, estrogen signaling, and atherosclerosis. Immune infiltration confirmed multifactorial immune cell involvement. Both murine models recapitulated inflammatory and immune responses, with the LPS model mimicking human SLI via chemotaxis, phagocytosis, NOD-like receptor signaling, and leukocyte migration. The CLP model uniquely replicated neutrophil chemotaxis, apoptosis, ER stress, IL-17, and TNF signaling. SOCS3 was validated as a potential SLI biomarker. DiscussionMurine models partially replicate human SLI pathology but exhibit distinct mechanistic emphases. Careful model selection is essential for biomarker discovery (e.g., SOCS3) and pathogenic mechanism exploration, highlighting inherent model limitations.