Abstract
Chronic myelogenous leukemia (CML) is characterized by the BCR/ABL fusion tyrosine kinase which mediates its oncogenic effects in part through constitutive activation of the transcription factor STAT5. BCR/ABL can be inhibited by several kinase inhibitors, including imatinib, which leads to apoptosis of the cells. However, the development of resistance to these agents, most commonly mediated through point mutations in the kinase, is an increasing problem. Since STAT5 is a critical mediator of the effects of BCR/ABL, the development of drugs inhibiting this transcription factor holds promise as an independent means to inhibit BCR/ABL-transformed cells. Therefore, we developed a high throughput cell-based screen to identify drugs that specifically inhibit the transcriptional function of STAT proteins. To accelerate the introduction of active agents into clinical trials, we have focused on drugs that are approved for use, or are otherwise known to be safe, in humans. Using this approach, we have identified pimozide, a neuroleptic drug, as an inhibitor of STAT5. Pimozide decreases STAT5 phosphorylation in CML cells but, in contrast to imatinib, it does not inhibit BCR/ABL kinase activity. Furthermore, pimozide decreases expression of STAT5 target genes in these cells. Consistent with the known role of STAT5 in promoting the survival and proliferation of BCR/ABL-transformed cells, pimozide induces apoptosis in the K562 CML cell line. At similar concentrations, pimozide induces little toxicity in peripheral blood mononuclear cells from normal donors. Given the clinical importance of point mutations in BCR/ABL in the development of resistance to tyrosine kinase inhibitors, we evaluated the effects of pimozide in Ba/f3 cells reconstituted with wildtype BCR/ABL or BCR/ABL containing a T315I mutation, which is highly resistant to current kinase inhibitors. Pimozide reduces the viability of both cell types with similar dose response, providing further support to the hypothesis that this drug works by a mechanism distinct from imatinib. Given this, we considered the possibility that the combination of pimozide and imatinib might be more effective than either drug alone. Treatment of KU812 CML cells with low concentrations of both pimozide and imatinib induced greater loss of viability than either drug alone, with characteristics consistent with pharmacological synergy. In addition, combination treatment with pimozide and imatinib induces a greater level of apoptosis in CML cells than either drug alone. In conclusion, we have identified pimozide as a STAT5 inhibitor that is effective at reducing the viability of CML cells, including those resistant to imatinib. This suggests that the identification of inhibitors of STAT transcription factors may be a promising method for developing new anti-cancer therapies. In addition, the use of pimozide alone or in conjunction with kinase inhibitors may be an important new strategy for the treatment of CML or other diseases characterized by constitutive STAT activation.
