Disruption of HOX-PBX Interaction; a Potential Therapeutic Target in Leukemia.

Most cases of acute myeloid leukemia (AML) are closely associated with gene rearrangements. Appraisal of these translocations and analysis of mouse models of leukemia has revealed that several members of the homeodomain containing family of transcription factors are implicated in the pathogenesis of leukemia. Overexpression of HOXA9 in murine models leads to the development of AML. This study focuses on the role of a subset of the HOX genes and their potential as a target for therapeutic intervention. We have designed a synthetic peptide, HXP4, that disrupts the interaction between HOX and PBX leading to growth inhibition of leukemic cells. An in vitro HOX-induced AML model of leukemia was utilised to determine the efficacy of HXP4 as a therapeutic agent. Using this immortalised cell line overexpressing HOXA9 (imHOXA9), we tested the efficacy of HXP4 in vitro. Cells were treated with HXP4 for four days and analysed. All results are expressed relative to untreated control cells. Following a 60μM dose of HXP4, no viable cells were detected as determined by trypan blue staining, suggesting that HXP4 was cytotoxic. Following treatment with a lower dose of 6μM HXP4, and re-suspension in drug-free medium for a further 6 days, cell regrowth was observed, suggesting a cytostatic effect. RT-PCR was performed to identify potential downstream targets of HXP4. Qualitative analysis showed other HOX family members to be unaffected by treatment with either HXP4 dose. A more detailed study was performed using quantitative RT-PCR with imHOXA9. Cells were treated with either 60μM HXP4, 3μM etoposide, or a combination of the two agents (H+Et) and harvested after 1, 2, and 4 hours. In general, no significant change in gene expression was observed in other HOX family members. However, HOXA1 was upregulated 3-fold when treated with HXP4, and HOXA2 was downregulated 2-fold in HXP4 and H+Et treated cells. The reasons for this are as yet unclear. HXP4 also downregulated N-RAS 3.5-fold at two hours. However, complete loss of N-RAS expression following H+Et treatment suggests that HXP4 may be more effective in combination with etoposide. CDC25 expression was slightly downregulated in HXP4-treated cells. The normal function of CDC25 is to activate CDC2 kinase in the nucleus, however in the absence of CDC25, CDC2 remains inactive leading to a delay in mitosis, supporting the proposed cytostatic mode of HXP4 action. For reasons as yet unclear, CD34 expression was upregulated 4-fold and 6-fold in HXP4 and H+Et treated cells respectively. These preliminary results suggest that HXP4 is a cytostatic agent at relatively low concentrations, with a reversible antiproliferative effect. Downstream genes regulated by disrupting the HOX-PBX interaction with HXP4 have been identified by RT-PCR, but microarray analysis will provide a more comprehensive screen for target genes. In vivo experiments are currently in progress. In conclusion blocking the interaction between HOX and PBX may represent a therapeutic strategy in leukemia treatment.