Coupled In Silico Toxicology Models Reveal Equivalent Ecological Risks from BPA and Its Alternatives in Chinese Surface Waters

As bisphenol A (BPA) has gradually become restricted in production scenarios, the ecological risk level of its main replacement chemicals, i.e., bisphenol S (BPS) and bisphenol F (BPF), should be noted. To overcome the limitations of toxicity data, two kinds of in silico toxicology models (quantitative structure–activity relationship (QSAR) and interspecies correlation estimation (ICE) models) were used to predict enough toxicity data for multiple species. The accuracy of the coupled in silico toxicology models was verified by comparing experimental and predicted data results. Reliable predicted no-effect concentrations (PNECs) of 8.04, 35.2, and 34.2 μg/L were derived for BPA, BPS, and BPF, respectively, using species sensitivity distribution (SSD). Accordingly, the ecological risk quotient (RQ) values of BPA, BPS, and BPF for aquatic organisms were assessed in 32 major Chinese surface waters; they ranged from nearly 0 to 1.86, but were <0.1 in most cases, which indicated that the overall ecological risk level of BPA and its alternatives was low. However, in some cases, the ecological risks posed by BPA alternatives have reached equivalent levels to those posed by BPA (e.g., Liuxi River, Taihu Lake, and Pearl River), which requires further attention. This study provides evidence that the application of coupled in silico toxicology models can effectively predict toxicity data for new chemicals, avoiding time-consuming and laborious animal experiments. The main findings of this study can support environmental risk assessment and management for new chemicals that lack toxicity data.