PBX3 is an important cofactor of HOXA9 in leukemogenesis

Although PBX proteins are known to increase DNA-binding/transcriptional activity of HOX proteins through their direct binding, the functional importance of their interaction in leukemogenesis is unclear. We recently reported that overexpression of a 4-homeobox-gene signature (ie, PBX3/HOXA7/HOXA9/HOXA11) is an independent predictor of poor survival in patients with cytogenetically abnormal acute myeloid leukemia (CA-AML). Here we show that it is PBX3, but not PBX1 or PBX2, that is consistently coexpressed with HOXA9 in various subtypes of CA-AML, particularly MLL-rearranged AML, and thus appears as a potential pathologic cofactor of HOXA9 in CA-AML. We then show that depletion of endogenous Pbx3 expression by shRNA significantly inhibits MLL-fusion-mediated cell transformation, and coexpressed PBX3 exhibits a significantly synergistic effect with HOXA9 in promoting cell transformation in vitro and leukemogenesis in vivo. Furthermore, as a proof of concept, we show that a small peptide, namely HXR9, which was developed to specifically disrupt the interactions between HOX and PBX proteins, can selectively kill leukemic cells with overexpression of HOXA/PBX3 genes. Collectively, our data suggest that PBX3 is a critical cofactor of HOXA9 in leukemogenesis, and targeting their interaction is a feasible strategy to treat presently therapy resistant CA-AML (eg, MLL-rearranged leukemia) in which HOXA/PBX3 genes are overexpressed.