North American researchers have developed a high throughput screening method that relies on zebrafish behavior to identify new neuroactive molecules and scaffolds. 1 The technology could streamline drug discovery by providing initial in vivo data about neuroactive drug candidates earlier and more cheaply than the rodent models used in preclinical development.Most compound screens use in vitro assays that reproduce only a few of the interactions that occur in whole organisms. This lack of complexity is especially challenging to drug discovery in neurology because in vitro assays cannot model a compound's effect on behavior and brain activity. Meanwhile, studies in the standard neurological rodent models can be long and expensive-factors that preclude their use in upstream drug discovery.Zebrafish (Danio rerio) have long been used as models to study development and gene function in vertebrates. But Randall Peterson and colleagues hit upon the idea of using them to model responses to small molecules after discovering the fish exhibited very specific behaviors in response to light stimulus.Peterson is an assistant professor of medicine at Harvard Medical School and Massachusetts General Hospital (MGH) and an associate member of the Broad Institute of MIT and Harvard.David Kokel, a postdoctoral research fellow in Peterson's MGH group, found that zebrafish embryos moved upon exposure to light. "He was probably not the first person ever to notice this," but he was intrigued by the behavior because zebrafish embryos don't have photoreceptors in their eyes, Peterson said.The team found that the embryos exhibited regular and predictable motor responses to a 30-second pulse of light, which the group suspected could be the basis for a behavior-based screen. They developed a 14-part barcode that represented the intensity of embryos' photomotor responses during different time periods across the 30-second pulse.The barcode for embryos exposed only to the light pulse provided a baseline of normal behavior as a control for embryos used in subsequent screens of library compounds.The researchers screened a library of 14,000 small molecules against the embryos' photomotor responses to a 30-second pulse and compared the resulting barcodes with those for controls. They found that the small molecule-induced barcodes fell into many distinct clusters, each corresponding to a particular phenotype in the zebrafish.For known neuroactive compounds, the clusters corresponded to the mode of action of the compounds, including targets and pathways. For example, barcodes for dopamine receptor agonists indicated intense photomotor activity throughout the second half of the pulse, whereas barcodes for adrenergic receptor agonists indicated intense photomotor activity at the beginning and end of the pulse.Given the results for known compounds, the team suggested the barcode clusters could serve as starting points for identifying the targets of new molecules. To test this hypothesis, they selected two compounds not previously known to be neuroactive...