One of the problems with approaches to conventional drug discovery is the need for information on an appropriate molecular target, when more often than not the pathways that need to be targeted for a therapeutic outcome are not obvious. In the May issue of Nature Biotechnology, Peterson et al. use a phenotype-based screen of a zebrafish model of aortic deformities to identify pathways and potential drug leads that correct the disease.

The zebrafish mutation gridlock (grl), which affects the gene hey-2, disrupts aortic blood flow in a similar way to aortic coarctation in humans, a common congenital cardiovascular malformation. The authors screened for chemical suppressors of grl by arraying mutant embryos in 96-well plates and exposing them to small molecules from a structurally diverse chemical library. After 48 hours of treatment, two small molecules out of 5,000 tested suppressed the gridlock phenotype, restoring normal blood circulation. The molecules, 4-[2-(4-methoxy-phenylsulphanyl) ethyl] pyridine and 4-[2-(4-nitro-phenylsulphanyl) ethyl] pyridine, were structurally related compounds. The former (also called GS4012) was more potent and was used for all subsequent experiments.

To elucidate the biological mechanism, the authors treated wild-type embryos with GS4012 and measured the expression level of genes known to regulate vasculogenesis and aortic development. Significant and dose-dependent increases in mRNA levels of vascular endothelial growth factor (VEGF) were observed in response to drug treatment. The ability of GS4012 to cause overexpression of VEGF could be the mechanism that rescues the mutant, as grl is thought to act downstream of VEGF signalling. When tested on primary human endothelial cells, the drug caused an increase in tubule formation, compared with controls.

This work demonstrates that phenotype-based chemical screens in vertebrate models can be valuable in identifying drug leads; however, whether the leads will be developed in light of the scant knowledge of the specific target remains to be seen. Drugs have been developed in the absence of a clear target, such as ezetimibe (Zetia; Merck) and metformin (Glucophage; Bristol-Myers Squibb), but in general it is likely that drug leads with known targets will receive higher priority than those without.