Acute myeloid leukemia (AML), the most common form of acute leukemia in adults, accounts for more than 40% of all leukemia deaths. The AML1-ETO (A) fusion gene resulting from t(8;21) is a recurring genetic alteration in human AML, typically associated with a favorable outcome. Internal tandem duplications (ITD) in the juxtamembrane domain of Fms-like tyrosine kinase 3 (Flt3) result in constitutive ligand-independent signaling. Flt3ITD is one of the most frequently identified molecular abnormalities in AML, and when present, portends unfavorable clinical outcomes. AML1/ETO and FLT3ITD can co-exist in human AML, indicating that these lesions cooperate to promote disease development. Neither AML1/ETO nor Flt3ITD alone are sufficient to induce AML in mice. To determine whether coordinated expression of AML1/ETO and Flt3ITD would cause AML, Mx1-Cre x LSL AML1-ETO/+ x FLT3ITD/ITD (MAFF) mice were generated. In this system, an Mx1 promoter-driven Cre recombinase removes a lox-stop-lox (LSL) cassette upstream of a knock-in human AML1/ETO fusion gene in hematopoietic progenitor cells, while FLT3ITD is constitutively expressed. MAFF mice developed AML at 4-5 weeks of age with 100% penetrance. The AML was associated with severe leukocytosis, splenomegaly and dense bone marrow infiltration by immature-appearing blasts, and was universally lethal. Flow cytometry of bone marrow from MAFF mice revealed a striking expansion of cells harboring a hematopoietic stem cell (HSC) and multipotent progenitor (MP) cell phenotype, indicative of a block in normal myeloid differentiation. No lymphoid leukemias were observed. Interestingly, two copies of the FLT3ITD allele were required for AML to develop in this model. Transplantation of as few as 105 bone marrow cells from primary MAFF mice into lethally-irradiated secondary recipients resulted in 100% engraftment and lethality, with a leukemia identical to that which developed in primary MAFF mice, albeit with delayed latency. Mixed bone marrow chimeras were generated using MAFF marrow along with marrow from syngeneic, wild-type C57BL/6 or Rag2-/-γC-/- mice (lacking mature B, T and NK cells). Chimeras generated with MAFF and Rag2-/-γC-/- marrow succumbed more rapidly than those with MAFF and wild-type marrow, indicating that adaptive immunity and/or NK cells provided some level of active immune surveillance against MAFF AML. MAFF mice were crossed with Rosa26-LSLSIY/+ mice (referred to as MAFFS mice) in order to incorporate inducible expression of a tractable peptide antigen (SIYRYYGL; SIY) in leukemia cells. Through the use of mixed bone marrow chimeric approaches, we are now capable of monitoring SIY antigen-specific CD8+ T cell responses in MAFFS AML-bearing animals, including contexts in which the SIY antigen is shared (leukemia-associated antigen) or expressed only by leukemia cells (leukemia-specific antigen). In conclusion, we have generated the first autochthonous murine AML model in which the role of host immune surveillance can be interrogated. We expect that the MAFF AML model will also be useful for studies regarding leukemia biology and treatment.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.