Nature 486, 80–84 (2012)

Credit: KEVAN SHOKAT

Although most kinase inhibitors clinically used to treat cancer were initially developed against one kinase, part of the efficacy and toxicity of many of these drugs has subsequently been attributed to blocking 'off-target' kinases. However, designing kinase inhibitors that maximize efficacy through polypharmacology while minimizing toxicity has proven challenging. To test an in vivo phenotypic screening approach, Dar et al. turned to a previously developed Drosophila model of multiple endocrine neoplasia type 2 (MEN2), a disease caused by activating mutations in RET kinase. The researchers screened a library of small molecules that inhibited RET as well as a variable spectrum of other kinases. The team next compared the phenotypic effects of the top hit, which resulted in 25% of the flies surviving to adulthood, with other library members that had a similar potency against RET but a distinct activity profile against other kinases. These data suggested that inhibiting Raf and Src increased efficacy, whereas inhibiting mTOR led to toxicity. On the basis of this information, the researchers designed a new inhibitor that rescued about 80% of the MEN2 flies and was more effective in a MEN2 mouse model than vandetanib, a kinase inhibitor approved to treat thyroid cancer resulting from RET mutations that cause MEN2. This work highlights the potential of Drosophila cancer models to facilitate the rational optimization of kinase inhibitor polypharmacology.