Atrazine Induces Alzheimer's Disease-Like Neurotoxicity by Targeting SDHB and Disrupting Synaptic Function: An Integrated Bioinformatic and In Vivo Study.

Atrazine (ATR) is a globally prevalent herbicide linked to increased risks of cognitive impairment and neurodegeneration, but its molecular mechanism remains unclear. This study integrates bioinformatics, machine learning, and experimental validation to elucidate ATR induced Alzheimer's disease (AD)-like pathology. We combined predicted ATR targets with AD brain transcriptomics, identifying 32 shared genes enriched in neurodegeneration and synaptic signaling pathways. We constructed a diagnostic model using 14 machine learning algorithms; the ensemble model achieved an improved AUC of 0.891 in the training set, with external validation AUCs of 0.833, 0.869, and 0.900. SHAP interpretability analysis identified SDHB as the most important key risk factor. Molecular docking validated the stable binding between ATR and core target proteins, with binding energies ranging from -4.9 to -5.9 kcal/mol. Single-cell sequencing and in silico gene knockdown confirmed that SDHB deficiency primarily disrupts neuronal synaptic signaling pathways. In vivo experiments demonstrated that ATR exposure induces anxiety-like behavior and memory deficits, suppressed hippocampal SDHB, and elevated Aβ and p-tau levels. ATR also reduced synaptic markers PSD95 and SYP in a dose dependent manner, confirming SDHB suppression leads to synaptic damage. Based on these findings, we established the first standardized adverse outcome pathway (AOP) framework delineating the cascade of events from ATR-induced SDHB inhibition to cognitive impairment. This study demonstrates that ATR targets SDHB to induce synaptic dysfunction and AD-like neuropathology, offering a new mechanistic basis for environmental risk assessment.