Flt3 is dispensable to the Hoxa9/Meis1 leukemogenic cooperation

HOX proteins are evolutionarily conserved homeodomain transcription factors that are involved in the regulation of hematopoietic differentiation and self-renewal. HOX genes are frequently up-regulated in leukemias and, clinically, those with the highest expression have been associated with a bad prognosis.^1,2  The oncogenic activity of several HOX genes has been demonstrated in murine leukemia models; however, these leukemias develop with long latencies, suggesting that additional oncogenic events cooperate toward leukemogenesis. The HOX cofactor MEIS1 cooperates with HOX genes to accelerate the onset of acute myeloid leukemias (AMLs) in mouse models (reviewed in Abramovich and Humphries³ ). Interestingly, MEIS1 is frequently found up-regulated along with HOX genes in human leukemias and this is associated with particularly high levels of FLT3 mRNA.⁴  FLT3 is a tyrosine kinase receptor that is transiently expressed during hematopoietic stem cell maturation but whose expression is persistently up-regulated in AMLs. FLT3 signaling triggers cellular proliferation, and mutations that constitutively activate FLT3 are found in approximately 30% of AMLs. Activation of FLT3 signaling can also be triggered by the overexpression of wild-type FLT3, possibly through autocrine or paracrine stimulatory loops in leukemic cells secreting FLT3 ligand (FL) or independently of FL.⁵  Recently, Wang et al⁶  proposed a mechanism by which MEIS1 cooperates with HOXA9 to induce leukemias. Using mouse progenitors immortalized by Hoxa9, they showed that Meis1 coexpression induced the expression of Flt3 and the ability to proliferate in response to FL stimulation, suggesting that the acquisition of leukemogenic potential conferred by Meis1 could be mediated by Flt3 up-regulation.⁶  In another recent study, retroviral transduction of FLT3 was shown to cooperate with NUP98-HOX fusion genes in accelerating the onset of leukemias and, to some extent, recapitulate the leukemogenic cooperative effect of Meis1.⁷  To test whether the proleukemic effect of Meis1 was mediated by the up-regulation of Flt3, we studied the transforming potential of Meis1 in combination with Hoxa9 in the context of Flt3-deficient cells.

Murine Hoxa9 was cloned in the MIE vector (MSCV-IRES-eGFP)⁸  or into the MSCVneo vector.⁹ Meis1a was cloned in the MSCVneo vector. The MSCV-Hoxa9-IRES-Meis1 vector and homozygous Flt3^−/− mice were kindly provided by Guy Sauvageau and Ihor Lemischka,¹⁰  respectively. Lin^−/low bone marrow (BM) cells were spinoculated as previously described.¹¹ 

Lethally irradiated (9 Gy) 129/Sv mice received 60 × 10³ to 70 × 10³ infected progenitors and 10⁵ 129/Sv BM cells for radioprotection. Sick mice were analyzed as previously described.¹²  Tumor transplantability was assessed by injecting 10⁵ leukemic BM cells into 2 to 3 C57Bl/6–Rag-2^−/− mice.

To determine whether Flt3 is essential for the cooperative oncogenicity of Meis1 with Hoxa9, we infected BM progenitors from Flt3^−/− or control 129/Sv Flt3^+/+ cells with retroviral vectors coding for Hoxa9 or Hoxa9 and Meis1. Transduced cells were serially replated in a myeloid colony-forming assay to compare their proliferative potential (Figure 1A). While Flt3^−/− or Flt3^+/+ cells infected with either Meis1 alone or an empty (MIE) vector ceased proliferating after 2 rounds of plating (data not shown), cells infected with Hoxa9 sustained high colony-forming ability and generated more than 500 quaternary colonies (per 10⁴ cells seeded) regardless of Flt3 genotype. A drastic synergistic effect of Meis1 with Hoxa9 was observed generating more than 5000 quaternary colonies (per 10⁴ cells) in both Flt3^−/− and Flt3^+/+ cells. This demonstrates that the ability of Meis1 to synergize with Hoxa9 in conferring to cells a very high clonogenic potential (> 50%) does not require Flt3.

To examine the effect of Meis1 on the level of Flt3 protein in the transformed progenitors, we performed Western-blot analyses of pooled primary colonies (Figure 1B). Compared with cells infected with Hoxa9, cells infected with Hoxa9/Meis1 expressed much greater amounts of Flt3, whereas no Flt3 was detected in cells infected with Meis1 alone. Thus, Meis1 acts in synergy with Hoxa9 to up-regulate Flt3.

To analyze the molecular pathways possibly underlying the Hoxa9/Meis1 oncogenic synergy, we studied the protooncogene c-Myb recently shown to be an important downstream target of Hoxa9 and Meis1.¹³  To explore whether c-Myb expression was affected by Flt3, we analyzed its level in pooled primary colonies (Figure S1, available on the Blood website; see the Supplemental Materials link at the top of the online article). Regardless of Flt3 genotype, we found that Hoxa9 cells expressed c-Myb very moderately, whereas a much stronger expression was seen in Hoxa9/Meis1 cells. This further demonstrates that mechanisms underlying leukemic transformation are independent of Flt3.

To analyze whether Flt3 plays a role in Meis1's collaborative induction of leukemias with Hoxa9,¹⁴  progenitors from Flt3^−/− or Flt3^+/+ mice transduced with either Hoxa9 or Hoxa9/Meis1 were transplanted into irradiated histocompatible 129/Sv recipients. All mice that received Flt3^−/− or Flt3^+/+Hoxa9/Meis1-transduced cells rapidly developed aggressive AMLs with a mean latency of 40 days, whereas the latency of AMLs seen in Hoxa9 mice was longer than 150 days (Figure 1C). The absence of Flt3 did not delay the onset of either Hoxa9 or Hoxa9/Meis1 leukemias. The features of the AMLs are summarized in Table 1and in Figure S1 and Table S1. Whereas important interindividual variations were seen, there was no consistent trend between Flt3^−/− and Flt3^+/+ mice with regards to leukemic blast morphology. Interestingly, there was a large range of Flt3 expression among Hoxa9/Meis1 Flt3^+/+ leukemias (Figure 1D; Table S1). However, we could not see any correlation between the frequency of Flt3-expressing cells, ranging from 16% to 81%, and aggressiveness of the disease, further indicating that Flt3 plays little role in the pathogenesis. Moreover, Flt3 genotype did not influence the transplantability of Hoxa9/Meis1 tumors, as injection of Flt3^−/− or Flt3^+/+ leukemic BM cells induced similar AMLs in all secondary recipients (data not shown).

All mice that received Hoxa9-transduced BM cells eventually developed AMLs (Figure 1C). Once established, the Hoxa9 AMLs displayed similar phenotypic characteristics to those induced by Hoxa9/Meis1 and again were not influenced by Flt3 genotype. Interestingly, some Flt3 expression was detected in the leukemic BM of most Flt3^+/+Hoxa9 mice, possibly resulting from somatic mutations that occurred to trigger the onset of AMLs in collaboration with Hoxa9.

Altogether, our results demonstrate that Flt3 is dispensable for the induction of aggressive leukemias by the concomitant expression of Hoxa9 and Meis1. Although we cannot rule out that certain signaling pathways possibly downstream of another fms-related kinase may be activated to compensate for the absence of Flt3 and may mediate Hoxa9/Meis1-induced transformation in the Flt3^−/− mice, our findings nonetheless provide grounds to question the relevance of therapeutically targeting FLT3 to fight myeloid leukemias associated with an up-regulation of HOXA9 and MEIS1.

Contribution: E.M. performed research and analyzed data; S.A. performed research; and C.L. designed research, analyzed data, and wrote the paper.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Catherine Lavau, CNRS UMR 7151, Hôpital Saint-Louis, 1 Ave Claude Vellefaux 75010 Paris, France; e-mail: catherine.lavau@paris7.jussieu.fr.

We are very grateful to Ihor Lemischka for the Flt3^−/− mice, to Guy Sauvageau for the Hoxa9/Meis1 retroviral vector, and to Marika Pla and the Animal Experimentation Department of the Institut Universitaire d'Hematologie (IUH) for mouse husbandry.

This work was supported by grants from Groupement d'Intérêt Scientifique (GIS) maladies rares, Sang pour Cent La Vie, and the Ligue Comité de Paris. S.A. is a recipient of a Ministère de la Recherche fellowship; C.L. is an Institut National de la Santé et de la Recherche Médicale (INSERM) scientist.