Meis and Hox: a mighty pair defeats apoptosis

Comment on Wermuth and Buchberg, page 1222

Cooperative expression of HoxA9 and Meis1 is one of the molecular hallmarks in myeloid leukemogenesis. It is now evident that coexpression of 2 genes provides protection against apoptotic signals.

Strong oncogenic stimuli sometimes induce an apoptotic response instead of neoplastic proliferation in a certain cellular context. Overexpression of Myc promotes proliferation on the one hand but it leads to apoptosis via the ARF-p53 pathway on the other. One of the important mechanisms to maintain homeostasis is by suppressing an uncontrollable proliferation signal, and thus only a few oncogenes act like a double-edged sword.

Meis1 belongs to a 3–amino acid loop extension (TALE) class homeobox gene and is a cofactor for abdominal B-like Hox such as HoxA9. Meis1 and many Hox genes are expressed at the highest levels in Sca-1⁺ Lin^– hematopoietic stem cells (HSCs).¹  In the neoplastic condition, coexpression of Meis1 and HoxA7/A9 is frequently observed in human acute myeloid leukemia (AML) patients and BXH2 murine AML.^2,3  The cooperative leukemogenic activity of Meis1 and HoxA9 has been further confirmed by the fact that Meis1 overexpression significantly accelerates disease onset of HoxA9-induced AML.⁴  Targeted disruption of Meis1 results in embryonic lethality due to severe hemorrhage related to defective angiogenesis and thrombocytopoiesis.⁵  Collectively these findings suggest that Meis1 possesses a growth-promoting effect for multiple different cell types including hematopoietic cells.

In this issue of Blood, Wermuth and Buchberg demonstrate that Meis1 overexpression strongly induces caspase-dependent apoptosis in both hematopoietic and nonhematopoietic cells. Similarly, PBX1, another TALE class homeobox gene, also induces apoptosis. While HoxA9 suppresses Meis1-induced apoptosis, it has no effect on PBX1-induced apoptosis. More interesting, coexpression of Meis1 and HoxA9 protects cells from multiple apoptosis inducers such as etoposide, staurosporine, and valinomycin, suggesting that interaction between Meis1 and HoxA9 may provide a general antiapoptotic effect. By contrast, HoxA9 expression alone does not exhibit the antiapoptotic function against these reagents. Since the apoptotic activity requires the homeodomain, Meis1 may activate its target genes in the caspase signaling pathway by itself or as a complex with Pbx. HoxA9 prevents this function and, moreover, may form a Meis/Hox complex to activate antiapoptotic molecules. Alternatively, balance among Meis, Hox, and Pbx may be important for apoptotic stimuli.FIG1 

Frequent coexpression of Hox and Meis in HSCs as well as leukemic cells thus indicates that Meis1 requires Hox to perform its growth-promoting function, otherwise the cells die quickly. Several models have been proposed to explain how Meis1 modulates the Hox function: (1) Meis1 and abdominal B-like Hox, including HoxA9, forms a heterodimeric complex and binds DNA; (2) Meis1 interacts with the Pbx/Hox complex on DNA and forms a ternary complex that enhances the transcriptional activity of Hox; and (3) Meis1 recruits Pbx from the Pbx/Hox complex, resulting in accumulation of the Hox homodimer. Thus, it remains to be clarified how Meis, Hox, and Pbx regulate their target genes and what kinds of complexes among these homeodomain proteins are responsible for cellular proliferation, differentiation, and apoptosis. Clarification of the molecular process in Meis1-induced apoptosis and the Hox-dependent antiapoptotic effect will lead to understanding of gene regulation by homeodomain proteins. ▪