Up-Regulation of Flt3 Is a Passive Event in Hoxa9/Meis1-Induced Acute Myeloid Leukemia

Background: The HOXA9, MEIS1, and FLT3 genes are frequently up-regulated and co-expressed in human acute myeloid leukemia (AML). Over-expression of Hoxa9 and Meis1 also cooperate strongly to induce aggressive AML with high Flt3 expression in mice, suggesting an important role for Flt3 in Hoxa9/Meis1 leukemogenesis. However, a previous study using an Flt3^-/- knock-out mouse model indicated that FLT3 is dispensable for HOXA9/MEIS1-induced AML (Morgado et al., Blood 2007). However, the lack of Flt3 dependency seen using this germline Flt3^-/- knock-out model may be the result from an adaptation by the leukemia initiating cells to Flt3-deficiency or that Flt3's functions have been compensated for. Several FLT3-inhibitors have been tested in clinical trials with varying efficacy on AML harboring the FLT3-ITD mutation but the potential of using wild-type FLT3 as a drug target in AML remains unclear.

Aim and methods: To further define the role of expression of wild-type FLT3 in HOXA9/MEIS1-induced AML we used the murine AML model based on concurrent Hoxa9 and Meis1 overexpression. We inhibited Flt3-signaling by two means; with the FLT3 inhibitor AC220 and with the use of an Flt3-ligand (Fl^-/-) knock-out mouse model to minimize the risk that the leukemia initiating cells get adapted to Flt3 deficiency beforehand. In addition, we investigated whether Flt3 can substitute for some of Meis1’s functions and accelerate leukemia onset together with Hoxa9 in vivo.

Results: Hematopoietic cells with concurrent over-expression of both Hoxa9 and Meis1 showed a robust upregulation of Flt3 transcripts compared to cells expressing only Hoxa9 (p < 0.001). In addition, the Hoxa9/Meis1 cells could grow in media supplemented with the Flt3-ligand (Fl) as the only growth factor while cells expressing only Hoxa9 could not. Treatment with the FLT3-inhibitor AC220 effectively inhibited this Flt3-dependent proliferative advantage in vitro. However, the aggressive AML developed in mice transplanted with hematopoietic cells over-expressing Hoxa9 and Meis1 could not be ameliorated with AC220 treatment compared to placebo controls.

Mice without expression of the Flt3-ligand (Fl^-/-) transplanted with bone marrow cells expressing Hoxa9 and Meis1 showed a slower engraftment of leukemic cells (p < 0.001), lower white blood cell counts (p < 0.01), and less organ infiltration (p < 0.001) compared to wild-type recipient mice. However, the rapid development of anemia, leukemia and death were not different in Fl^-/- mice compared to wild-type mice. Also, leukemic cells from Fl^-/- mice expressed similar levels of Flt3 compared to leukemic cells from wild-type mice. In addition, overexpression of Flt3 together with Hoxa9 in hematopoietic cells could not accelerate the progression to AML in mice.

Conclusions: We conclude that FLT3-signaling likely contributes to proliferation and engraftment of leukemic cells with high Hoxa9 and Meis1 expression when its ligand is present. However, bone marrow cells over-expressing Hoxa9 and Meis1 upregulate Flt3 expression also in the absence of Flt3-ligand and inhibition of the signaling through Flt3 does not prevent or delay leukemia development induced by Hoxa9 and Meis1. Thus, the Hoxa9- and Meis1-associated upregulation of Flt3 seems to be a passive event with regard to leukemia development, with limited relevance to the AML pathology. Treating AML with high HOXA9 and MEIS1 expression using FLT3 inhibitors will therefore most likely be unsuccessful.