Integrative multi-omics profiling reveals coordinated immunometabolic reprogramming and host-microbiome interactions in acute pancreatitis.

BACKGROUND

Acute pancreatitis (AP) is a life-threatening inflammatory disorder characterized by diverse etiologies and complex pathophysiological mechanisms involving immune dysregulation, systemic metabolic reprogramming, and gut microbiota disturbances. Although single-omics studies have provided partial insights into AP pathogenesis, comprehensive integrative multi-omics analyses investigating the intricate interactions among immunity, metabolism, and the microbiome in AP remain limited.

METHODS

We conducted an integrative multi-omics analysis of peripheral blood transcriptomics, untargeted plasma metabolomics, and fecal whole-metagenome sequencing in 15 patients with AP and 15 age- and sex-matched healthy controls. Differentially expressed genes (DEGs), metabolites (DEMs), and gut microbial species (DGMs) were identified. Subsequently, functional enrichment analysis, correlation network analysis, and exploratory machine learning approaches were employed to investigate molecular interactions and identify candidate biomarkers.

RESULTS

Transcriptomic profiling identified 4, 776 DEGs, including 409 immune-related genes significantly enriched in the NF-κB, IL-17, and cytokine-cytokine receptor interaction pathways, indicating pronounced inflammatory activation. Metabolomic analysis detected 296 DEMs, with prominent alterations in amino acid and lipid metabolism, mong which 9 metabolites showed potential discriminatory value (AUC > 0.75), with representative metabolites including xanthine, homocarnosine, and tetradecanedioic acid. Metagenomic sequencing revealed significant microbial compositional and functional remodeling, characterized by enrichment of pro-inflammatory taxa such as Escherichia coli and Streptococcus anginosus, alongside depletion of SCFA-producing commensals including Faecalibacterium prausnitzii and Blautia wexlerae. Functional profiling demonstrated disrupted amino acid metabolism, gut-brain signaling, and SCFA synthesis. Multi-omics integration revealed 215 significant correlations between host genes, metabolites, and microbes, highlighting key interaction hubs. An exploratory random forest model identified Lachnospira pectinoschiza, Megamonas funiformis, and SRGN as candidate biomarkers, showing promising classification performance within the current cohort (AUC = 0.951).

CONCLUSIONS

This study provides a systems-level characterization of the immune, metabolic, and microbial alterations in AP. The identified molecular signatures and cross-omics interaction networks offer mechanistic insights into AP pathogenesis and highlight candidate biomarkers that warrant further validation in larger, independent cohorts.