The Mll-AF9 Knock-In Fusion Gene Results in Reprogramming and Enhanced Self-Renewal of Hematopoietic Progenitors, Especially in the HSC and CLP Populations.

The relative importance of the major hematopoietic progenitor/stem cell populations in MLL fusion gene leukemias remains largely undefined. Several studies have demonstrated that retrovirally transformed populations will produce leukemia in mice. However, because the levels of gene expression are known to alter cellular effects and MLL fusion gene expression in previous studies is generally supraphysiologic, uncontrolled and variable, comparison of individual population is difficult. In this study, we used knock-in mice with the fusion gene inserted in the genomic DNA by homologous recombination as an alternative model, permitting study of cells in which the fusion gene is under the control of the endogenous promoter and thus expressed at physiological levels in all progenitor/stem cells. Such knock-in cells also have the advantage that there is haploinsufficiency of the Mll gene expression which may be important in cellular dysfunction. Mll-AF9 knock-in mice develop myeloproliferation as early as 6 days of age and develop myeloproliferative disease-like (MPD-like) myeloid leukemia after six months latency. We postulated that Mll-AF9 induced deregulation would result in progenitor cells with quantifiably different biologic properties. In this study, we evaluated marrow progenitor/stem cells from young mice before leukemia development. Wild type or Mll-AF9 marrow cells from 3–6 young mice (8 weeks old) were sorted into HSC, CLP, CMP and GMP populations using standard protocols and grown in methylcellulose medium containing IL-3, IL-6, SCF, and GM-CSF. After 7 days’ culture, 3.6 fold more myeloid colonies were found in GMP and 2.5 fold more in CMP groups from Mll-AF9 compared to WT mice. After the third plating generation (21 days in culture), a very different pattern emerged. As expected, only rare colonies were found from WT mice whereas Mll-AF9 HSC and CLP groups resulted in significantly more myeloid colonies (60.1±5.5 and 67.5±10.5, respectively) than the CMP and GMP groups (27.2±3.8 and 20.9±4.1, respectively). Dense compact (Type I) colonies composed of immature myeloid cell were frequent in the Mll-AF9 HSC and CLP groups. Two-dimensional Sca-1 and c-kit plots showed that Mll-AF9 resulted in increases from a mean of 6.4% to 12.9% in HSCs and 12.4 to 27.3% in CLPs. Remarkably, Mll-AF9 CLP day 7 cultures had notably high number of CD11b+Gr1+ cells (mean of 57%). Mll-AF9 CLPs cultured under myeloid conditions also showed mixed lineage characteristics: expression of myeloid genes (GM-CSF receptor, c-fms, and lysozyme) and early B lymphoid genes (IL-7 receptor, TdT but not PAX5) and D-J rearrangement. In summary, our results show that Mll-AF9, when expressed at physiologic levels, reprograms progenitor/stem cells from young mice months before development of leukemia. CLPs are remarkable because in contrast to lymphoid restricted wild type CLPs, Mll-AF9 CLPs are mixed lineage in phenotype. Mll-AF9 resulted in increased self-renewal in all four HSC, CLP, CMP, and GMP populations compared with WT populations. However, importantly, in Mll-AF9 mice, self-renewal was most enhanced in cells with more intrinsic self-renewal potential and/or more evidence of reprogramming (HSCs, CLPs) than in committed myeloid progenitors. A predicted consequence of these differences is that cells with the highest Mll-AF9 induced self-renewal (HSCs, CLPs) will be more efficient than transformed CMPs and GMPs in inducing leukemia.