Integrative transcriptomic analysis links epithelial PCK1 dysregulation to ulcerative colitis and acetyl-l-carnitine-mediated inflammatory modulation.

BACKGROUND: Ulcerative colitis (UC) is a chronic inflammatory disorder characterized by persistent mucosal injury and disturbed immune-metabolic homeostasis. Increasing evidence indicates that metabolic remodeling of intestinal epithelial cells contributes to UC-associated inflammation; however, the key metabolic regulators involved in this process remain poorly characterized. METHODS: Integrative analyses of bulk transcriptomic datasets from the Gene Expression Omnibus (GEO) were performed to identify UC-associated differentially expressed genes. Weighted gene co-expression network analysis (WGCNA), combined with intersection against an adipogenesis-annotated metabolic gene set, was used to identify candidate metabolism-associated DEGs. Machine learning algorithms were applied to develop an exploratory classification model and prioritize hub genes. Single-cell and spatial transcriptomics were employed to delineate cellular localization. Functional validation was performed in lipopolysaccharide (LPS)/tumor necrosis factor-α (TNF-α)-stimulated NCM460 cells and dextran sulfate sodium (DSS)-induced colitis mice, using siRNA-mediated knockdown of phosphoenolpyruvate carboxykinase 1 (PCK1) in vitro, pharmacological intervention with 3-mercaptopicolinic acid (3-MPA) in vivo, and acetyl-l-carnitine (ALC) treatment. RESULTS: Integrative transcriptomic analysis identified 11 metabolism-associated differentially expressed genes enriched in metabolism- and inflammation-related pathways, particularly AMP-activated protein kinase (AMPK) signaling. Machine-learning-based feature prioritization highlighted five candidate genes, with PCK1 showing the highest contribution to model prediction based on mean absolute SHapley Additive exPlanations (SHAP) values. PCK1 was downregulated in UC, with epithelial enrichment in single-cell analysis and spatially heterogeneous expression in inflamed mucosa. In functional assays, siRNA-mediated PCK1 knockdown enhanced inflammatory gene responses in vitro, while systemic 3-MPA treatment was associated with aggravated DSS-induced inflammatory phenotypes in vivo. ALC treatment partially restored PCK1 expression and alleviated inflammatory responses, with AMPK phosphorylation associated with its protective effects. CONCLUSIONS: PCK1 is an epithelial metabolism-associated candidate linked to inflammatory responses in UC. ALC treatment partially restored PCK1 expression and alleviated inflammatory responses, with AMPK phosphorylation associated with its anti-inflammatory effects.