Targeting the molecular chaperone Hsp90 is emerging as a promising anticancer strategy, as highlighted in two recent papers in Chemistry and Biology on a class of Hsp90 inhibitors. In contrast to other 'targeted' anticancer agents, which are typically aimed at a single cancer-associated molecular alteration, these Hsp90 inhibitors have multiple antitumour-specific effects, meaning that they might be active in a considerably broader range of cancers.

Molecularly targeted therapies have now been the focus of anticancer drug development programmes for several years, and have proved highly successful in a few cases in which the cancer is dependent on the targeted oncoprotein, as, for example, is the case with imatinib (Gleevec; Novartis). However, most tumours are characterized by multiple molecular alterations, which are specific to the tumour. So, targeting just one alteration might be insufficient, and is only likely to show significant success for the subset of tumours in which the targeted protein has a key role.

How might these problems be circumvented? One strategy that has been proposed is to target the machinery that allows cancer cells to function despite the abnormalities that they carry. Hsp90 is a key component of this machinery, as it is required for the stability and functioning of a number of mutated, chimeric and overexpressed signalling proteins that promote the growth and/or survival of cancer cells. Furthermore, it seems that differences between the activation state of Hsp90 in tumours and normal cells allow Hsp90 inhibitors to show good selectivity for tumour cells.

However, Hsp90 inhibitors reported so far — which are primarily natural products or natural product derivatives — have not yet lived up to the promise of their target, largely owing to problems with bioavailability, solubility and toxicity. The structural complexity of the natural products makes the search for related compounds that are more drug-like very challenging, and so researchers have been looking for new families of Hsp90 inhibitors without these limitations.

In the first of the two papers, Chiosis and colleagues present PU24FCl, which is a representative of the first class of designed Hsp90 inhibitors. This compound inhibited the growth of wide range of cancer cell lines in vitro, and these effects correlated with its effects on Hsp90 client proteins thought to be involved in cancer cell growth, such as the receptor tyrosine kinase ERBB2 (HER2/neu). And in mice bearing breast-cancer xenografts, PU24FCl selectively accumulated in tumour tissue, and again showed a dose-dependent effect on Hsp90 client proteins. Finally, in a 30-day study of antitumour efficacy in this xenograft model, treatment with PU24FCl resulted in a 72% reduction in tumour burden compared with a control group.

The second paper, by Wright et al., reports crystal structures of several other Hsp90 inhibitors in the same class as PU24FCl. These structures allowed the authors to develop a detailed explanation for the observed affinity of members of this class for Hsp90, and should prove valuable in the design of further compounds to fully realize the potential of Hsp90 inhibition as an anticancer strategy.