Integration of multi-omics and network toxicology suggests potential mechanisms linking artificial sweeteners to chronic kidney disease

Artificial sweeteners (AS) are widely consumed as sugar substitutes. However, their potential involvement in the pathogenesis of chronic kidney disease (CKD) presents a complex toxicological question. This study aims to explore the potential molecular mechanisms and “toxicant-target-disease” interactions that may link AS exposure to renal function decline. We suggested a systems toxicology framework integrating network pharmacology, ensemble machine learning, and multi-dimensional bioinformatics. Nine machine learning algorithms were applied to human CKD transcriptomic datasets (GEO) to screen for diagnostic markers. The findings were further evaluated via immune infiltration profiling, molecular docking simulations, and Mendelian randomization (MR) analysis. Network analysis identified 448 convergent targets bridging AS exposure to CKD, which were enriched in PI3K-AKT signaling and inflammatory cascades. Machine learning prioritized five core toxicological effectors: MAP2K1 , IKBKB , CASP1 , IL1B and TLR2 . These biomarkers exhibited high diagnostic potential and correlated with dysregulated renal immune microenvironments, notably involving monocyte and mast cell infiltration. Molecular docking predicted stable binding affinities between AS components and these targets, while MR analyses provided genetic evidence supporting a potential causal effect of specific target IL1B on CKD risk. Our findings propose a multi-layered mechanism wherein artificial sweeteners may exacerbate renal decline, potentially via inflammasome-mediated and innate immune pathways. This study offers a theoretical foundation for further safety re-evaluation of sugar substitutes and identifies a candidate biomarker panel for CKD risk assessment.