ΜΙSΤRG Mice Support Good-Risk AML Engraftment

Introduction

Acute myeloid leukemia (AML) is a hematopoietic stem cell derived disease with still unsatisfactory treatment options. Its broad biological heterogeneity needs to be dissected by deeper mechanistic understanding to translate functional insights into novel therapeutic strategies. AML engraftment in murine models and subsequent in vivo studies have been limited to clinically aggressive AML and cell lines. In fact there is no published xenograft model for good risk AML. The hypothesis of a humanized environment being favorable for AML engraftment motivates the development of humanized mouse models for faithful in vivo analysis including biologically less aggressive, so called good risk AML.

Material & methods

We transplanted sublethally irradiated newborn MISTRG mice (Rongvaux et al. Nat. Biotech. 2014) that express as knock-ins human M-CSF, IL-3, GM-CSF, TPO and human SIRPa as transgene as well as NSG mice as controls i.h. with human AML samples. AML blasts were purified prior to transplantation by immunomagnetical CD3 and CD19 depletion to avoid xenogeneic graft versus host disease (GvHD) and EBV-driven B-cell proliferation/lymphoma, respectively. We selected AML samples with known genetic alterations representing good risk subgroups (AML with isolated NPM1 mutation or t(8;21)). Engraftment was assessed at different time points by peripheral blood analysis (cell counts and immunophenotyping) and bone marrow aspiration. Final analysis was completed by quantitative and histologic bone marrow analysis such as molecular diagnostics.

Results

Robust engraftment of good risk AML could be observed upon transplantation of patient samples into MISTRG mice: 87% of animals (n=15) engrafted (>1% hCD45+/ live cells) with a mean engraftment level of 61%. Time to and extent of engraftment depends on transplanted cell numbers, aggressiveness of disease and reaches its maximum around 16 to 20 weeks post transplantation in good risk AML when disease burden tends to limit survival. Gender of transplanted mice did not appear to influence engraftment. Engraftment was associated with a distinct phenotype including peripheral pancytopenia with leading anemia and splenomegaly with a significant positive correlation between engraftment level and spleen weight (Spearman r=0.6, p=0.03). We were able to show engraftment by immunophenotyping and molecular prove of preexisting genetic alterations. Engraftment levels in NSG mice were similar or inferior at any time (67% engraftment (n=6) with a mean engraftment level of 37% hCD45+/ live cells) and side to side comparison suggests an AML immunophenotype closer to human AML presentation in MISTRG mice. Engraftment did not seem to depend on SIRPa expression while preliminary data points towards a critical role of IL-3/GM-CSF and M-CSF.

Conclusions

Humanised mouse strains with differential human knock-ins provide suitable models to further dissect the disease supporting environment, which can then be translated into the development of direct therapeutic strategies.