by ecancer reporter Clare Sansom

Currently, many of the most important and interesting targets for cancer therapy are protein kinases: enzymes that control cell signalling through catalysing the addition of phosphate groups to proteins.

The first kinase inhibitor to enter the clinic, imatinib, is a potent and specific inhibitor of the BCR-ABL kinase, which is mutated in almost all cases of chronic myeloid leukaemia.

However, the results of targeting other kinases known to be dysregulated in one or more tumour types have been mixed.

The effectiveness of many kinase inhibitors is limited by toxicity that often arises from “off-target” activity against kinases that are needed in normal human physiology. With over 500 kinases in the human genome, optimum activity profiles for therapeutic kinase inhibitors remain hard to predict.

Kevan Shokat from the University of California San Francisco, San Francisco, USA and his co-workers have now demonstrated that it is possible to identify both kinase targets where inhibition contributes to the efficacy of a therapy and so-called “anti-targets” where inhibition contributes to its toxicity.

They used a Drosophila model of multiple endocrine neoplasia type 2 (MEN2), a genetic disorder that is characterized by a greatly increased propensity to develop medullary thyroid carcinoma. Most people with this syndrome have mutations in the receptor tyrosine kinase RET.

Shokut and his colleagues synthesised a panel of compounds that inhibit RET and also a variable number of kinases downstream of RET in the same signal transduction pathways.

Screening these compounds against the Drosophila assay, in which mutation of the fly orthologue of the RET gene, dRET, prevents survival to adulthood unless this kinase is inhibited, led to the selection of one compound, AD57, as a potent inhibitor of the enzyme.

A close analogue of this compound, AD36, showed limited efficacy, whereas another analogue, AD58, was ineffective. Furthermore, fewer flies survived to the pupal stage when treated with AD58, indicating that this compound was toxic.

These three inhibitors were tested against a broad panel of mammalian kinases in vitro, and the results showed that the relatively small chemical changes between the molecules led to significant differences in their activity profiles.

The potent inhibitor AD57 also inhibited BRAF, S6K, mTOR and SRC kinases, whereas the toxic AD58 was active against mTOR and ineffective against the other kinases. AD57 was also shown to inhibit the growth of MEN2-derived human cancer cells in vitro and in a mouse xenograft model. This indicated that phenotype-based screens using a Drosophila model could identify kinase inhibitors that were effective against human tumour cells.

The researchers then investigated whether the toxicity observed with AD58 was caused by its inhibition of mTOR, a kinase that is known to inhibit one of the pathways involving RET – the RAS pathway – in mammals.

Reducing the expression of the Drosophila analogue of mTOR, dTor, reduced the efficacy of AD57 and increased the toxicity of the analogue AD58. This confirmed mTOR as an “anti-target” for MEN2: an inhibitor of RET, BRAF, S6K, and SRC kinases but not mTOR should therefore be an effective drug for this disease.

To test this hypothesis further, they developed a set of AD57 analogues ]that could be expected to selectively inhibit all the kinases above except mTOR. Two compounds, AD80 and AD81, showed strong activity against human RET, BRAF, S6K, and SRC but were much less active than either AD57 or AD58 against mTOR.

They were also more effective than AD57 in the Drosophila phenotypic screen; in particular, 70-90% of flies treated with AD80 developed into phenotypically normal adults. AD80 was also shown to inhibit the growth of several human tumour cell lines in vitro.

Taken together, these results show that a combination of synthetic medicinal chemistry, kinome profiling and a Drosophila-based phenotypic screen in a method that the authors term “systems pharmacology” can lead to the identification of certain kinases as “anti-targets” in a specific tumour and to the identification of lead compounds with optimum potency and minimum toxicity. This promising approach may be applicable in many other cancer types.

Article: Dar, A.C., Das, T.K., Shokat, K.M. and Cagan, R.L. (2012). Chemical genetic discovery of targets and anti-targets for cancer polypharmacology. Nature 486, 80-84. doi: 10.1038/nature11127