Exploring toxicity mechanisms of single and binary PFAS on Chlorella vulgaris using integrated QSAR modeling and molecular docking

Per- and polyfluoroalkyl substances (PFAS) are persistent organic pollutants whose toxicity mechanisms remain unclear, particularly in mixtures. This study integrated toxicity experiments, quantitative structure–activity relationship (QSAR) modeling, and molecular docking techniques to investigate the toxicity of 20 individual PFAS and their binary mixtures toward Chlorella vulgaris . The 2D QSAR model identified lipophilicity (MolLogP) and the topological and electronic properties of the molecular surface (PEOE_VSA14) as key descriptors of the toxicity of individual PFAS. The 3D QSAR model (CoMFA) further indicated that both the elongated perfluoroalkyl chains and negatively charged head groups contributed to greater toxicity. For binary mixtures, the concentration addition (CA) and independent action (IA) models achieved the same interaction-classification accuracy of 82%, and the interaction type was ratio-dependent. Mixture-specific QSAR descriptors, including averaged topological structural connectivity (ATSC1m) and dipole moment, were associated with mixture toxicity, indicating additional interactions between components. Molecular docking with spinach superoxide dismutase 1 (SOD1) showed a significant correlation between binding affinity and toxicity (r = –0.78, p < 0.001), suggesting SOD1 as a potential molecular target. The most toxic compound exhibited a binding mode characterized by hydrophobic-chain insertion, hydrogen bonds with ARG-79 and PRO-74, and halogen bonds, which shared key features with the structural requirements predicted by 3D-QSAR. These findings provide atomic-level insights into PFAS–SOD1 interactions, support SOD1 as a potential target, and highlight the need for further experimental validation to confirm functional relevance. They also establish a multiscale framework for ecological risk assessment and the design of safer alternatives.