Might as well face it: MLL's addicted to HOX.

Chromosomal rearrangements involving the MLL gene are a characteristic feature of a variety of human leukemias, including acute lymphoid (ALL) and acute myeloid leukemias (AML) as well as secondary leukemias resulting from treatment with topoisomerase inhibitors.1 The MEIS1 and HOXA9 genes have been shown to be major downstream targets of MLL fusion proteins.2 The mammalian HOX genes encode DNA-binding homeobox proteins that have a role in developmental patterns and tissue fate during embryogenesis. MEIS1 is a homeodomain protein that acts as a DNA-binding cofactor of HOX proteins from paralog groups 9 through 13.3 MEIS1 and HOXA genes are normally expressed in hematopoietic progenitor cells and their expression is down-regulated on differentiation.4 Overexpression of MEIS1 and HOXA9 is seen in a variety of leukemic cell lines and primary AML samples, suggesting roles for these genes in leukemia initiation and maintenance.5 In addition, fusion of HOXA9 to NUP98 in some AMLs further supports an important role for HOXA9 in leukemia.6 In this issue of Blood, Faber et al ask if HOXA9 expression is required to maintain the leukemic state.7 The authors investigated the effects of knocking down HOXA9 expression in MLL-rearranged and nonrearranged human leukemia cells. After shRNA-mediated HOXA9 suppression, induction of apoptosis and decreased colony formation were observed in leukemia cell lines. An increase of cells in the G1 phase of the cell cycle and some cellular differentiation was also observed prior to the cells undergoing apoptosis. The apoptotic phenotype was rescued by expression of nontargetable HOXA9. The authors demonstrated that a greater induction of cell death was observed in MLL-rearranged cell lines and primary AML samples compared with MLL germ line cells after shRNA knockdown of HOXA9. In addition, the level of apoptosis positively correlated with levels of HOXA9 expression prior to shRNA knockdown. Gene expression profiling on MLL-rearranged cells after knockdown of HOXA9 expression showed decreased expression of genes, such as MEIS1 and PBX3, previously shown to play a role in leukemogenesis, suggesting these genes are downstream of HOXA9. Finally, MLLrearranged cells were subjected to knockdown of HOXA9 and transplanted into mice. These mice showed a significant decrease in leukemia burden, demonstrating that continued HOXA9 expression is required for survival and proliferation of human MLL-rearranged cells in vivo. The results provided by Faber et al demonstrate that MLL-rearranged cells are addicted to HOXA9 expression to maintain their leukemic state, as shown in the figure. Another major downstream target of MLL fusion proteins is MEIS1.2 Previous work done by Wong et al and work by Kumar et al used similar strategies to demonstrate that maintenance of the leukemic state in MLL-rearranged cells is also dependent on expression of MEIS1.8,9 Knockdown of MEIS1 expression results in decreased proliferation and survival of these cells. Although MEIS1 is important in maintenance of the leukemic state, elevated levels of MEIS1 are not sufficient to induce leukemia.10 In contrast, elevated levels of HOXA9 are sufficient to induce leukemia with a long latency.10 Coexpression of MEIS1 along with HOXA9, however, reduces the latency of HOX-induced leukemia.4 Taken together, these data suggest a role for both MEIS1 and HOXA9 in leukemia initiation and maintenance. The data from the gene expression profiling experiment in Faber et al suggest that MEIS1 is downstream of HOXA9. Is continued HOXA9 expression required to maintain high levels of MEIS1 expression in leukemic cells? Knockdown of HOXA9 expression in MLL-rearranged cells and simultaneous ectopic expression of MEIS1 will further elucidate the complex interactions of these transcription factors in leukemia maintenance. Clarification of this gene partnership will lead to an understanding of gene regulation in MLL-rearranged leukemias. MEIS1 and HOXA9 have been shown to function as a heterodimer and in higher order complexes. Small molecule inhibitors could provide alternate therapeutic strategies by disrupting MEIS1/HOXA9 dimer formation, leading to loss of proliferation and survival in MLLrearranged leukemic cells. Conflict-of-interest disclosure: The authors declare no competing financial interests. ■