Adapting classic paradigms to analyze alterations of shoal-wide behavior in early-life stages of zebrafish (Danio rerio) - A case study with fluoxetine.

Given the strong increase in prescription of neuroactive pharmaceuticals, neurotoxicity has received growing concern in science and the public. Regulatory requirements stimulated the development of new methods to evaluate the risk of neurotoxic substances for humans and the environment, and, with respect to potential damage to aquatic ecosystems, a variety of behavior-based assays have been proposed for neurotoxicity testing, most of which, however, are restricted to changes in the behavior of individual fish. Since many fish species form shoals under natural conditions, this may cause important aspects of behavior to be overlooked and there is a need for behavior assays integrating individual behavior with behavior of the entire swarm. In order to combine more environmentally realistic sub-chronic exposure scenarios with undistorted social behavior and animal welfare considerations, two behavioral assays are proposed that might be integrated into early-life stage toxicity studies according to OECD TG 210, which are commonly run for a multitude of regulations: To this end, protocols for a novel tank test and a predator response assay were adapted to also record the behavior of free-swimming zebrafish (Danio rerio) juveniles within shoals. Comparisons of the diving response (novel tank) or the shoal's coherence and position relative to the stimulus (predator) with control groups allow conclusions about the anxiety state of the fish, which might well have an impact on survival chances in the wild. As a model substance, the antidepressant fluoxetine ((RS)-N-Methyl-3-phenyl-3-(4-trifluoromethylphenoxy)propylamine) produced adverse effects down to concentrations three orders of magnitude below the EC from acute fish embryo toxicity tests according to OECD TG 236. With the integration of such behavior tests into OECD TG 210, important population-relevant information on potential neurotoxicity can be collected without increasing the number of experimental animals.