Wang Xiaodong, Yin Chunsheng, Wang Liansheng
State Key Laboratory of Pollution Control and Resources Reuse, The School of Environment, Nanjing University, China.
Chemosphere. 2002 Feb;46(7):1045-51. doi: 10.1016/s0045-6535(01)00148-5.
Inhibition of growth of the yeast Saccharomyces cerevisiae (Cmiz, the minimum concentration that produced a clear inhibition zone within 12 h) for 24 nitroaromatic compounds was investigated and a quantitative structure-activity relationship (QSAR) developed based on hydrophobicity expressed as the l-octanol/water partition coefficient in logarithm form, log K(ow), electrophilicity based on the energy of the lowest unoccupied orbital (E(lumo)). All nitrobenzene derivatives exhibited enhanced reactive toxicity than baseline. The toxicities of mono-nitrobenzenes and di-nitrobenzenes were elicited by different mechanisms of toxic action. For mono-nitro-derivatives, both significant log K(ow) based and strong E(lumo)-dependent relationships were observed indicating that their toxicities were affected both by the penetration process and the interaction with target sites of interaction. The toxicities of di-nitrobenzenes were greater than mono-nitrobenzenes and no log K(ow)-dependent but highly significant E(lumo)-based relationship was obtained. This suggests that toxicity of di-nitrobenzenes was highly electrophilic and involved mainly their in vivo electrophilic interaction with biomacromolecules. In an effort to model the elevated toxicity of all nitrobenzenes, a response-surface analysis was performed and this resulted in a highly predictive two-variable QSAR without reference to their exact mechanisms (Cmiz = 0.41 log K(ow) - 0.89 E(lumo) - 0.46, r2 = 0.87, Q2 = 0.86, n = 24).
研究了24种硝基芳香化合物对酿酒酵母生长的抑制作用(Cmiz,即12小时内产生明显抑菌圈的最低浓度),并基于以对数形式表示的疏水性(正辛醇/水分配系数,log K(ow))和基于最低未占轨道能量(E(lumo))的亲电性建立了定量构效关系(QSAR)。所有硝基苯衍生物均表现出比基线更高的反应毒性。单硝基苯和二硝基苯的毒性由不同的毒性作用机制引起。对于单硝基衍生物,观察到了基于log K(ow)的显著关系和强烈的E(lumo)依赖性关系,这表明它们的毒性既受渗透过程影响,也受与相互作用靶位点相互作用的影响。二硝基苯的毒性大于单硝基苯,未获得基于log K(ow)的依赖性关系,但获得了基于E(lumo)的高度显著关系。这表明二硝基苯的毒性具有高度亲电性,主要涉及其在体内与生物大分子的亲电相互作用。为了模拟所有硝基苯的毒性升高情况,进行了响应面分析,结果得到了一个无需参考其确切机制的高度预测性双变量QSAR(Cmiz = 0.41 log K(ow) - 0.89 E(lumo) - 0.46,r2 = 0.87,Q2 = 0.86,n = 24)。