Novel approach to predicting hormetic effects of antibiotic mixtures on Vibrio fischeri.

The determination of the hormetic effects of a mixture is quite difficult because of the moderate simulation and the complexity of measurement in low doses. In the present study, two typical models for mixture toxicity prediction, concentration additive (CA) and independent action (IA), were used to predict the hormetic effects of mixtures. The predictive power of those models was validated by the hormetic effects (24-h exposure) of antibiotic's binary mixtures to Vibrio fischeri. The results showed that CA and IA were unable to predict the hormetic dose-response of mixture, especially those of the interactive mixtures. As an alternative, a novel model, which was named as "six-point" and developed based on the quantitative features in the determined dose-response curve and on the Quantitative Structure Activity Relationships (QSARs) approach, was proposed for predicting the hormetic effects of mixtures in low dose. The results indicated that the "six-point" model can accurately predict the mixture hormetic effects in low dose, not only for non-interactive mixtures but also for interactive mixtures. Therefore, the "six-point" model is a powerful tool to predict the mixture hormetic effects at low dose, and may offer an important approach in the environment risk assessment of mixtures.