Integrating Cascade Mechanistic Insights into Structure–Activity Relationships for Quantitative Mode-of-Action Analysis: A Novel Risk Assessment Framework for OPFRs

Organophosphate flame retardants (OPFRs), as emerging substitutes for halogenated flame retardants, are widely detected in populations and pose global health concerns. However, their structural diversity and expanding numbers challenge conventional experimental risk assessment. Here, we present a quantitative mode-of-action (qMOA) framework integrating structural attributes and cascade mechanistic insights to enable predictive risk evaluation. From 68 registered OPFRs, 59 representative structurally diverse compounds (average pairwise similarity value: 0.51) were selected for model training. The model identified benzene ring count (55.34%) and substituent type (39.14%) as the primary structural determinants of toxicity. This structure–activity relationship was experimentally validated across tiered events in mitochondrial dysfunction─a recognized MOA for environmental pollutants. Leveraging the validated link and experimental data, we developed a qMOA model to calculate mitochondrial dysfunction MOA-specific threshold doses for individual OPFRs. For example, model-derived thresholds for widely detected OPFRs were 27.84 μg/mL for EHDPP (two benzene rings) and 75.76 μg/mL for TEP (three CH 3 substituents, nonbenzene ring). This enables direct comparison with human exposure data, revealing that relatively multibenzene-ring OPFRs pose higher potential health risks. This study advances a mechanism-informed strategy that enhances the reliability of toxicity inference and enables rapid risk quantification for diverse existing OPFRs and emerging analogues.