Search

Meis1 is a homeodomain transcription factor coexpressed with Hoxa9 in most human acute myeloid leukemias (AMLs). In mouse models of leukemia produced by Hoxa9, Meis1 accelerates leukemogenesis. Because Hoxa9 immortalizes myeloid progenitors in the absence of Meis1 expression, the contribution of Meis1 toward leukemia remains unclear. Here, we describe a cultured progenitor model in which Meis1 programs leukemogenicity. Progenitors immortalized by Hoxa9 in culture are myeloid-lineage restricted and only infrequently caused leukemia after more than 250 days. Coexpressed Meis1 programmed rapid AML-initiating character, maintained multipotent progenitor potential, and induced expression of genes associated with short-term hematopoietic stem cells (HSCs), such as FLT3 and CD34, whose expression also characterizes the leukemia-initiating stem cells of human AML. Meis1 leukemogenesis functions required binding to Pbx, binding to DNA, and a conserved function of its C-terminal tail. We hypothesize that Meis1 is required for the homing and survival of leukemic progenitors within their hematopoietic niches, functions mediated by HSC-specific genes such as CD34 and Fms-like tyrosine kinase 3 (FLT3), respectively. This is the first example of a transcription factor oncoprotein (Meis1) that establishes expression of a tyrosine kinase oncoprotein (FLT3), and explains their coexpression in human leukemia. This cultured progenitor model will be useful to define the genetic basis of leukemogenesis involving Hoxa9 and Meis1.

MEIS1 and Hoxa9 are homeobox transcription factors that promote self-renewal in hematopoietic progenitors. MEIS1 does not induce leukemia, but cooperates strongly with Hoxa9 to produce acute myeloid leukemia (AML). Previously, we demonstrated that Hoxa9 blocks differentiation of myeloid progenitors that do not express MEIS1 and do not induce leukemic. Coexpression of MEIS1 causes transcription of genes that segregate with the leukemia-initiating subset of human AML blasts, such as CD34 and FLT3. We designate these genes as leukemic stem cell genes, or LSC genes. MEIS1 promoted LCS gene transcription by a mechanism that requires interaction with PBX and DNA, and that also requires a short MEIS1 C-terminal transactivation domain (CTD). Here we use a dominant transactivating or transrepressing form of MEIS1 to determine whether the activation or repression function of Pbx:MEIS1 complexes is sufficient to cause myeloid leukemia in combination with coexpressed Hoxa9. Surprisingly, fusion of MEIS1 to the Vp16 transactivation domain (but not the engrailed transrepression domain) produced an autonomous oncoprotein that immortalized progenitors and caused myeloid leukemias without the need for coexpressed exogenous or endogenous Hox genes. Like MEIS1, Vp16MEIS1 required binding to Pbx and DNA for immortalization; however, the CTD was not necessary in the context of Vp16MEIS1. This suggests that the CTD participates in target gene activation in AML blasts, a function replaced by Vp16 in its absence. Retroviral expression of Hoxa9 or Hoxa7 induced a further, strong, transcriptional upregulation of LSC genes in Vp16MEIS1 progenitors and elevated their leukemic potential to the level of bona fide AML blasts. These data suggest that transactivation is the essential function of Pbx:MEIS1 complexes in AML, and that HOX proteins cooperate with Pbx:MEIS1 complexes to activate transcription of early progenitor genes whose expression is required for human AML.