Network toxicology screening of plastic-associated chemicals reveals candidates with structural mimicry to Alzheimer’s disease-relevant molecules and predicted neurotoxicity

Alzheimer's disease (AD) pathogenesis involves environmental factors, with plastic-associated chemicals emerging as potential neurotoxicants. This study employed a computational toxicology framework to screen plastic-associated chemicals for potential interference with AD-related pathways. From a compiled library of 12,751 plastic-associated chemicals, 2,315 compounds with high blood-brain barrier (BBB) permeability were identified. Using Morgan fingerprints and Tanimoto similarity (threshold ≥ 0.6), we pinpointed chemicals structurally analogous to known AD-active molecules, including 3,5-Dimethoxystilbene (similarity 0.731 to a resveratrol derivative) and Isoeugenol (similarity 0.710 to isorhapontigenin). Structural similarity alone does not imply toxicity; therefore, we further evaluated these candidates using target prediction, molecular docking, and in silico toxicity profiling. Target prediction and molecular docking indicated that several compounds may interact with core AD-related targets (e.g. AChE, MAOB, APP). Notably, ProTox-3.0 profiling revealed that both 3,5-Dimethoxystilbene and Isoeugenol exhibit a potential dual profile of concern: high predicted BBB penetrance (probability > 0.7) and predicted potential for neurotoxic effects (probability 0.54-0.68), although these moderate probability scores warrant cautious interpretation. High-affinity binding was validated for 3,5-Dimethoxystilbene with AChE (-9.1 kcal/mol), comparable to known AChE inhibitors in our docking validation. This study prioritizes plastic-associated chemicals with a 'brain-penetrating and potentially neurotoxic' profile based on integrated computational evidence, providing a data-driven foundation for experimental validation.