Schultz T W, Lin D T, Arnold L M
College of Veterinary Medicine, University of Tennessee, Knoxville 37901-1071.
Sci Total Environ. 1991 Dec;109-110:569-80. doi: 10.1016/0048-9697(91)90210-6.
The relative toxicities (log IGC50(-1] of a fairly heterogeneous series of 66 anilines that were monosubstituted in the 2-, 3-, and 4-positions have been evaluated in the 48 h static Tetrahymena population growth impairment system. Quantitative structure-activity relationships (QSARs) were examined using the 1-octanol/water partition coefficient (log KOW) and the summation of the Hammett sigma electronic substituent constant (sigma sigma) as orthogonal independent predictors. Four chemicals, 4-decylaniline, 4-dodecylaniline, 4-tritylaniline and 4-aminophenethyl alcohol, did not elicit the measured response at saturation. Five chemicals, 2-aminophenol, 2-phenylenediamine, 4-aminophenol, 4-phenylenediamine and 4-nitroaniline, had an altered HPLC spectrum with time. As previously reported, the parent compound, aniline, as well as the 4-position halogen derivatives have been shown to be aberrantly more toxic than expected. None of these chemicals were included in QSAR development which employed linear regression analysis. Log IGC50(-1) = 0.599(log KOW) - 0.905 (n = 52, r2 = 0.885, s = 0.265, f = 383.65, Pr greater than f = 0.0001) was found to be a good predictor of relative toxicity of these monosubstituted anilines. The addition of sigma sigma as a second predictor did not improve the predictability of this QSAR. Abiotic loss over the duration of the assay varied markedly with derivative; however, chemical persistence was not considered to be a factor with this model. The above equation was strikingly similar to log IGC50(-1) = 0.595(log KOW) - 0.996 (n = 67, r2 = 0.866, s = 0.254, f = 421.64, Pr greater than f = 0.0001), a previously developed QSAR from data on a heterogeneous set of monosubstituted phenols. The data for these two equations were combined and subsequent regression analysis resulted in log IGC50(-1) = 0.588(log KOW) - 0.939 (n = 119, r2 = 0.873, s = 0.261, f = 806.49, Pr greater than f = 0.0001). The authors feel that this represents the log KOW-dependent QSAR in the Tetrahymena population growth assay for the polar narcosis mechanism of toxic action.
在48小时静态四膜虫种群生长抑制系统中,对66种在2、3和4位单取代的相当异类的苯胺的相对毒性(log IGC50(-1))进行了评估。使用1-辛醇/水分配系数(log KOW)和哈米特σ电子取代常数总和(σσ)作为正交独立预测因子,研究了定量构效关系(QSAR)。四种化学品,4-癸基苯胺、4-十二烷基苯胺、4-三苯甲基苯胺和4-氨基苯乙醇,在饱和时未引发测量响应。五种化学品,2-氨基酚、2-苯二胺、4-氨基酚、4-苯二胺和4-硝基苯胺,其HPLC光谱随时间发生了变化。如先前报道,母体化合物苯胺以及4位卤素衍生物的毒性异常高于预期。这些化学品均未纳入采用线性回归分析的QSAR开发中。发现log IGC50(-1) = 0.599(log KOW) - 0.905(n = 52,r2 = 0.885,s = 0.265,f = 383.65,Pr大于f = 0.0001)是这些单取代苯胺相对毒性的良好预测因子。添加σσ作为第二个预测因子并未提高该QSAR的可预测性。在测定期间的非生物损失因衍生物而异;然而,化学持久性在该模型中未被视为一个因素。上述方程与log IGC50(-1) = 0.595(log KOW) - 0.996(n = 67,r2 = 0.866,s = 0.254,f = 421.64,Pr大于f = 0.0001)惊人地相似,后者是先前根据一组异类单取代酚的数据开发的QSAR。将这两个方程的数据合并,随后的回归分析得出log IGC50(-1) = 0.588(log KOW) - 0.939(n = 119,r2 = 0.873,s = 0.261,f = 806.49,Pr大于f = 0.0001)。作者认为,这代表了在四膜虫种群生长试验中针对极性麻醉毒性作用机制的依赖log KOW的QSAR。