Long-term toxicity assessment of antibiotics against Vibrio fischeri: Test method optimization and mixture toxicity prediction

Abstract

The current luminescent bacteria test for acute toxicity with short contact time was invalid for antibiotics, and the non-uniformed contact times reported in the literature for long-term toxicity assessment led to incomparable results. Herein, a representative long-term toxicity assessment method was established which unified the contact time of antibiotics and Vibrio fischeri within the bioluminescence increasing period (i.e. 10-100% maximum luminescence) of control samples. The effects of excitation and detoxification of antibiotics such as β-lactams were discovered. Half maximal inhibitory concentration (IC₅₀) of toxic antibiotics (0.00069-0.061 mmol/L) obtained by this method was 2-3 orders of magnitude lower than acute test, quantifying the underestimated toxicity. As antibiotics exist in natural water as mixtures, an equivalent concentration addition (ECA) model was built to predict mixture toxicity based on physical mechanism rather than mathematical method, which showed great fitting results (R² = 0.94). Furthermore, interaction among antibiotics was investigated. Antibiotics acting during bacterial breeding period had strong synergistic inhibition (IC₅₀ relative deviation from 0.1 to 0.6) such as macrolides and quinolones. Some antibiotics produced increasing synergistic inhibition during concentration accumulation, such as macrolides. The discharge of antibiotics with severe long-term toxicity and strong synergistic inhibition effect should be seriously restricted.

Keywords: Bioluminescence; Concentration addition models; Ecological risk; Hormesis; Synergistic inhibition.