Abstract
BACKGROUND: Global consumption of artificial sweeteners (ASs) has risen substantially in recent years. However, their relationship with prostate cancer (PCa) remains poorly characterized. This study investigates the AS-PCa association to identify pivotal genes potentially bridging this relationship. METHOD: This study retrieved target genes associated with ASs and PCa from multiple public databases. Protein-protein interaction (PPI) network analysis and visualization were conducted on overlapping genes, followed by the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses to explore the underlying mechanisms. Subsequently, the optimal predictive model was selected from 101 machine-learning algorithm combinations and validated against 2 external datasets. Molecular docking analysis was then performed to examine the interactions between key genes and AS compounds. Finally, in vitro cellular assays were conducted to validate the specific effects of ASs on PCa. RESULTS: We analyzed seven common ASs-aspartame, acesulfame-K, sucralose, NHDC, sodium cyclamate, neotame, and saccharin-identifying 261 overlapping targets associated with PCa. The GO and KEGG enrichment analyses revealed that these targets primarily regulate cell proliferation, inflammation, and cancer cell metabolism. Machine learning algorithm screening identified the Lasso-SuperPC hybrid model as demonstrating optimal predictive performance, with robust validation in two independent external datasets. Subsequent analysis identified two key regulatory genes: CD38 and MMP11. Molecular docking analysis further confirmed potential interactions between AS compounds and the core target MMP11. Finally, in vitro cellular assays demonstrated that NHDC suppresses MMP11 expression in PCa cells and exhibits anti-PCa pharmacological effects. CONCLUSION: By integrating bioinformatics, machine learning, molecular docking, and in vitro cellular assays, this study demonstrates that ASs inhibit PCa progression through multiple molecular targets and signaling pathways. Collectively, these findings provide important insights into the safety assessment of food additives and cancer risk assessment.