One way that tumour cells inactivate tumour suppressors is through subcellular mislocalization. For example, in some cancer cells, the forkhead (FOXO) family of transcription factors, which normally inhibit cell proliferation, are functional, but are exported from the nucleus to the cytoplasm. Pam Silver's group has developed a screen for compounds that prevent FOXO transport out of the nucleus and lead to cell-cycle arrest.

FOXO localization is regulated by the PI3K–PTEN–AKT signalling pathway. PTEN is mutated in several tumours, leading to loss of its lipid phosphatase activity and constitutive activation of AKT signalling. AKT phosphorylates FOXO transcription factors at several sites, which promotes nuclear export and thereby prevents their transcriptional activity. As studies showed that forcible localization of FOXO1A to the nucleus can reverse tumorigenicity of PTEN-null cells, Silver and colleagues developed a cell-based chemical screen for inhibitors of FOXO1A nuclear export. Using localization of FOXO1A as a visual assay, they screened over 18,000 compounds for their ability to relocalize FOXO1A to the nucleus in PTEN-null cells. 89 compounds were found to cause FOXO1A nuclear retention, 42 of which were selected for further characterization.

Of these 42 compounds, 19 were also able to block export of the human immunodeficiency virus rev protein — another nuclear factor — and were therefore considered to be 'general export inhibitors'. These compounds were found to function by interfering with the activity of CRM1, an export receptor that shuttles proteins out of the nucleus. The other 23 compounds that were examined were only able to block nuclear export of FOXO1A, through several points along the AKT signalling pathway.

One of the compounds was identified as trifluoperazine, which acts as a calmodulin inhibitor, among other things. The authors therefore treated PTEN-null cells with other types of calmodulin inhibitors, and found that they all caused FOXO1A re-localization to the nucleus. This is the first report of a calmodulin-dependent regulatory mechanism of FOXO1A localization, and indicates that calmodulin could interact with the AKT signalling pathway.

So, the compounds that were identified in the screen cause FOXO1A to stay in the nucleus, but how do they affect tumour growth? Thirty of the nuclear-transport inhibitors that were discovered also blocked proliferation of the PTEN-null cancer cells. Additional studies are required, however, before transport inhibitors can be used as cancer therapies. The drug LMB, which blocks CRM1-mediated nuclear export, has already been found to be highly toxic in Phase I clinical trials. So it will be important to learn whether this toxicity is linked to its effects on CRM1 or to its other possible activities. Nonetheless, the authors hope that some of the inhibitors that were identified in this screen might be developed as leads for new anticancer drugs.