Structure–toxicity relationships for four classes of aliphatic electrophiles to Vibrio fischeri

Quantitative structure–activity relationships (QSARs) that relate the toxicity of a xenobiotic to its physico-chemical properties have particular application in the assessment of pollution to marine species. In this study, predictive QSARs have been developed for the logarithm of the 15-min toxicity to the marine bacterium Vibrio fischeri (pT15) using a hydrophobicity term (logarithm of the octanol-water partition coefficient, log P) and a term to describe electrophilicity (the energy of the lowest unoccupied molecular orbital, ELUMO). QSARs were developed initially for individual chemical classes. Statistically robust, hydrophobic-dependent QSARs were obtained for the haloalcohols and halonitriles. However, a lower quality hydrophobic-dependent equation was observed for diones and only a poor quality hydrophobic-dependent model was developed for chloroesters. These QSARs suggests the four chemical classes exhibit different electrophilic mechanisms of toxic action, as well as different mechanisms occurring in the diones and chloroesters considered. Despite this, a combination of all four classes revealed a highly predictive QSAR based on both descriptors (pT15=0.93 (log P)−0.403 (ELUMO)−0.64; n=47, s=0.478, r2=813). This predictive model is applicable regardless of chemical classes and does not require a priori identification of mechanisms of action.