Lactate-related biomarkers in severe acute pancreatitis: insights from machine learning, molecular dynamics, and experimental validation

Acute pancreatitis (AP) can be mild and self-limited, but it can also become severe acute pancreatitis (SAP) and lead to multi-organ dysfunction. It is still difficult to identify patients who may develop severe disease at an early stage. Lactate was once viewed only as a metabolic byproduct, but recent work shows that it also acts as a signaling molecule that links cellular energy status with immune control. Its role in the progression of AP is still not clear. Transcriptomic analysis showed clear differences in metabolism and immune features between SAP and non-severe AP (NSAP). Among lactate-related genes (LRGs), 24 were expressed at different levels. Three machine-learning methods identified CCNA2 and H2BC5 as possible diagnostic markers. Both genes had higher expression in SAP clinical samples. In vitro experiments showed that adding lactate increased CCNA2 and H2BC5 expression and raised inflammatory cytokine production. When these genes were knocked down, the inflammatory response decreased. This suggests that both genes are involved in lactate-driven inflammatory signaling. Molecular docking and simulation showed that the MDM2 inhibitor AMG-232 binds strongly to CCNA2, and the mTOR inhibitor Torin-1 binds strongly to H2BC5. Functional tests showed that Torin-1 reduced inflammation and oxidative stress in a sodium-taurocholate–induced cell model. This study shows that CCNA2 and H2BC5 are candidate biomarkers for SAP. They are involved in important immune and metabolic pathways. AMG-232 and Torin-1 are possible therapeutic drugs. These findings give a molecular framework for early risk detection in AP and point to new options for targeted treatment.