Wang Ying, Yang Xianhai, Wang Juying, Cong Yi, Mu Jingli, Jin Fei
Key Laboratory for Ecological Environment in Coastal Areas, State Oceanic Administration, National Marine Environmental Monitoring Center, No. 42 Linghe Street, Dalian 116023, China.
Nanjing Institute of Environmental Sciences, Ministry of Environmental Protection, Jiang-wang-miao Street, Nanjing 210042, China.
J Hazard Mater. 2016 May 5;308:149-56. doi: 10.1016/j.jhazmat.2016.01.006. Epub 2016 Jan 7.
In the present study, quantitative structure-activity relationship (QSAR) techniques based on toxicity mechanism and density functional theory (DFT) descriptors were adopted to develop predictive models for the toxicity of alkylated and parent aromatic hydrocarbons to Vibrio fischeri. The acute toxicity data of 17 aromatic hydrocarbons from both literature and our experimental results were used to construct QSAR models by partial least squares (PLS) analysis. With consideration of the toxicity process, the partition of aromatic hydrocarbons between water phase and lipid phase and their interaction with the target biomolecule, the optimal QSAR model was obtained by introducing aqueous freely dissolved concentration. The high statistical values of R(2) (0.956) and Q(CUM)(2) (0.942) indicated that the model has good goodness-of-fit, robustness and internal predictive power. The average molecular polarizability (α) and several selected thermodynamic parameters reflecting the intermolecular interactions played important roles in the partition of aromatic hydrocarbons between the water phase and biomembrane. Energy of the highest occupied molecular orbital (E(HOMO)) was the most influential descriptor which dominated the toxicity of aromatic hydrocarbons through the electron-transfer reaction with biomolecules. The results demonstrated that the adoption of freely dissolved concentration instead of nominal concentration was a beneficial attempt for toxicity QSAR modeling of hydrophobic organic chemicals.
在本研究中，采用基于毒性机制和密度泛函理论（DFT）描述符的定量构效关系（QSAR）技术，建立烷基化芳烃和母体芳烃对费氏弧菌毒性的预测模型。利用文献中的17种芳烃急性毒性数据以及我们的实验结果，通过偏最小二乘法（PLS）分析构建QSAR模型。考虑到毒性过程、芳烃在水相和脂相之间的分配及其与目标生物分子的相互作用，引入水相自由溶解浓度获得了最优的QSAR模型。较高的R(2)（0.956）和Q(CUM)(2)（0.942）统计值表明该模型具有良好的拟合优度、稳健性和内部预测能力。平均分子极化率（α）和几个反映分子间相互作用的选定热力学参数在芳烃在水相和生物膜之间的分配中起重要作用。最高占据分子轨道能量（E(HOMO)）是最具影响力的描述符，它通过与生物分子的电子转移反应主导芳烃的毒性。结果表明，采用自由溶解浓度而非标称浓度是疏水有机化学品毒性QSAR建模的有益尝试。
