Negative effects of GM-CSF signaling in a murine model of t(8;21)-induced leukemia

Abstract

The t(8;21)(q22;q22) is common in adult acute myeloid leukemia (AML). The RUNX1-ETO fusion protein that is expressed by this translocation is poorly leukemogenic and requires additional mutations for transformation. Loss of sex chromosome (LOS) is frequently observed in t(8;21) AML. In the present study, to evaluate whether LOS cooperates with t(8;21) in leukemogenesis, we first used a retroviral transduction/transplantation model to express RUNX1-ETO in hematopoietic cells from XO mice. The low frequency of leukemia in these mice suggests that the potentially critical gene for suppression of t(8;21) leukemia in humans is not conserved on mouse sex chromosomes. The gene encoding the GM-CSF receptor α subunit (CSF2RA) is located on X and Y chromosomes in humans but on chromosome 19 in mice. GM-CSF promotes myeloid cell survival, proliferation, and differentiation. To determine whether GM-CSF signaling affects RUNX1-ETO leukemogenesis, hematopoietic stem/progenitor cells that lack GM-CSF signaling were used to express RUNX1-ETO and transplanted into lethally irradiated mice, and a high penetrance of AML was observed in recipients. Furthermore, GM-CSF reduced the replating ability of RUNX1-ETO-expressing cells. These results suggest a possible tumor-suppressor role of GM-CSF in RUNX1-ETO leukemia. Loss of the CSF2RA gene may be a critical mutation explaining the high incidence of LOS associated with the t(8;21)(q22;q22) translocation.

Figure 1

Loss of 1 X chromosome in mice does not promote RUNX1-ETO–induced leukemia. (A) Schematic representation of retroviral constructs. MSCV indicates murine stem cell virus; LTR, long-terminal repeat; HA, hemagglutinin tag; and IRES, internal ribosome entry site. (B) Kaplan-Meier survival curve of animals transplanted with MigRE-transduced XO or XY cells. Pooled results from 6 different transplantation experiments are shown.

Figure 2

GM-CSF is not necessary for maintenance of hematopoiesis. (A) Response of wild-type (βc^+/+) and βc^−/− cells to single cytokines in colony assay. (B) Flow cytometric analysis of myeloid progenitor (MP) and LSK of βc^+/+ and βc^−/− mice. Absolute numbers and percentage of total BM are shown. MP, PI⁻/Lin⁻/IL-7Rα⁻/c-Kit⁺/Sca-1⁻; LSK, PI⁻/Lin⁻/IL-7Rα⁻/c-Kit⁺/Sca-1⁺ (n = 3; C) Flow cytometric analysis of common myeloid progenitor (CMP), PI⁻/Lin⁻/IL-7Rα⁻/c-Kit⁺/Sca-1⁻/FcγR^lo/CD34⁺; granulocyte-macrophage progenitor (GMP), PI⁻/Lin⁻/IL-7Rα⁻/c-Kit⁺/Sca-1⁻/FcγR^hi/CD34⁺; and megakaryocyte-erythrocyte progenitor (MEP), PI⁻/Lin⁻/IL-7Rα⁻/c-Kit⁺/Sca-1⁻/FcγR^lo/CD34⁻ in βc^+/+ and βc^−/− animals. (n = 6).

Figure 3

Deficiency of GM-CSF signaling is favorable for the development of RUNX1-ETO–induced leukemia. (A) Kaplan-Meier survival curve of animals transplanted with βc^−/− cells transduced with MigR1 or MigRE. Pooled results from 4 different transplantation experiments are shown. (B) Immunoblotting analysis of spleen cells from βc^−/− MigRE leukemia animals using anti-HA tag (RUNX1-ETO) Ab. MigRE is a positive control. Wild-type (WT) is a negative control. (C) Kaplan-Meier survival curve of animals transplanted with βc^−/− or βc^+/+ cells transduced with MigR1 or MigRE. (D) Kaplan-Meier survival curve of animals transplanted with βc^−/− or βc^+/+ cells transduced with MigR1 or MigRE9a vectors.

Figure 4

AML is observed in animals transplanted with βc^−/− RUNX1-ETO cells. (A) Cytological analysis of a representative leukemic βc^−/− MigRE animal. Peripheral blood (PB) smears, BM, and spleen (SP) cytospin samples stained with Wright-Giemsa are shown (original magnification, 400×). (B) Histopathological analysis of a representative leukemic βc^−/− MigRE animal. H&E-stained sections of the BM (original magnification, 200×) and SP (original magnification, 100×) are shown. (C) Splenomegaly observed in leukemic βc^−/− MigRE animal. (D) Flow cytometric analysis of the BM from a representative leukemic βc^−/− MigRE animal; numbers represent percentage in GFP⁺BM cells. CD3/B220 are lymphocyte markers; CD11b/Gr-1, granulocyte markers; Ter119/CD71, erythroid markers. (E) Flow cytometric analysis of myeloid progenitor (MP) or LSK from a representative leukemic βc^−/− MigRE animal. Respective percentages in GFP⁺BM are shown. (F) Flow cytometric analysis of MP populations from a representative leukemic βc^−/− MigRE animal. CMP indicates common myeloid progenitor; GMP, granulocyte-macrophage progenitor; and MEP, megakaryocyte-erythrocyte progenitor.

Figure 5

The GM-CSF cytokine negatively affects the self-renewal of RUNX1-ETO cells. (A) Representative result of 3 independent replating assays using BM cells transduced with MIP vector (MIP) or MIP-RE (RE), with or without addition of 10 ng/mL of GM-CSF. (B) Differential counts of cytospin samples from colony cells. Representative results of 2 independent experiments are shown. (C) Differential count of CFUs at the first replating. Representative results of 2 independent experiments. CFU-GM indicates granulocyte macrophage CFU; CFU-G, granulocyte CFU; CFU-M, macrophage CFU; CFU-GEMM, granulocyte, erythroid, macrophage, megakaryocyte CFU; and CFU-E, erythroid CFU. (D) Colony-forming efficiency of Kasumi-1 cells infected with MigR1 (MigR1) or MigR1-CSF2RA (CSF2RA) retrovirus. A representative result of average colony numbers of duplicate samples of 2 independent experiments is shown.

Figure 6

Constitutively elevated phosphorylation of Erk1/2 in RUNX1-ETO cells. (A) Representative immunoblot showing lineage-negative BM cells transduced with MIP or RE and drug selected for 48 hours. Cells were serum and growth factor starved for 6 hours. GM-CSF was added at 50 ng/mL. (B) Quantification of band intensities by densitometry of the blots shown in panel A. Intensities of bands from phosphorylated Erk1/2 were normalized to Erk1/2. (C) Representative immunoblot of βc^+/+ and βc^−/− cells transduced with RE. Cells were starved and stimulated as in panel A.