Several Homeobox transcription factors (Hox) have been shown to be highly involved in the complex regulation of hematopoiesis, controlling proliferation, differentiation and self-renewal capacity of hematopoietic stem cells (HSC). One of these genes, HOXA10, has been shown to be expressed in primitive hematopoietic cells and meyolid progenitors, but is downregulated as the cells mature. The expression of HOXA10 is found in almost all types of human acute myeloid leukemia (AML) and overexpression of HOXA10 in mice by a retroviral vector induces proliferation of myeloid progenitors, that eventual leads to AML. In order to further study the role of HOXA10 in hematopoiesis and to ask whether the effect of HOXA10 is dose dependent, we generated an inducible system based on the tetracycline transactivator system, by mating our previously published transgenic HOXA10 mouse model with the Rosa26rtTA mouse. The Rosa locus is transcriptionally active in many organs including all hematopoietic tissues. Here we show that we can induce the expression of HOXA10 in the bone marrow of inducible mice in vivo by administrating doxycycline and no leakiness was detected in uninduced bone marrow. Double transgenic mice were born at normal ratios and hematopoiesis was normal prior to induction. When the gene was systemically induced by administration of doxycycline (2 mg/ml) in the drinking water the mice died within one to two weeks. Analysis showed small hemorrhages (intestine, skin), low hemoglobin level (HOXA10: 82±21 g/L, wt: 130±4 g/L), low RBC (HOXA10: 5.9±1.7x1012/L, wt: 8.67±0.4x1012/L) and the platelet count was severely reduced (A10: 33±26x109/L, wt: 845±121x109/L, data from three experiments n=6, p<0.002). In order to further study the effect of HoxA10 in hematopoietic cells we transplanted inducible HOXA10 bone marrow cells to lethally irradiated wild type recipients. The expression of HOXA10 was induced as above and analysis revealed severe anemia 3–4 weeks after induction with reduced hemoglobin level (HOXA10: 71±18 g/L wt: 124±6 g/L), and low RBC count (HOXA10: 4.8±1.3x1012/L, wt: 8.1±0.7x1012/L, data from three experiments, n=8, p<0.001). When the mice became terminally ill, they were sacrificed, displaying splenomegaly with immature erythrocytes at different stages of development indicating a block in erythroid maturation. In a separate experiment a lower dose (1.5x106) of inducible HOXA10 BM was transplanted. Most of these mice survived but displayed a reduction in hemoglobin levels at three weeks post BMT. Linage analysis showed that high HOXA10 expression caused a block in CD3+ T-cells (HOXA10: 2.88±1.59%, uninduced: 17±1%, wt: 19±1% p<0.0001) and an increase in the proportion of Gr1/Mac1+ positive cells (HOXA10: 45±5%, uninduced: 18±4%, wt: 18±5% p<0.0001). In summary, we present a new model system for studying the effect of a transcription factor in vivo and demonstrate that high levels of HOXA10 cause a block in erythroid development and a severe anemia. Furthermore, high levels of HOXA10 cause reduced numbers of T-cells and increased myelopoiesis, while B cell development is unaffected.

Author notes

Corresponding author