Mice with megakaryocytic-targeted ablation of GATA-1 expression (GATA-1low mice) develop while age a syndrome similar to idiopathic myelofibrosis. These mice present with life-long thrombocytopenia and extensive proliferation of differentiation-impaired megakaryocytes. Although GATA-1 affects erythroid cell production, GATA-1low mice are not anemic because of increased erythropoiesis in the spleen. At 8–10-month, GATA-1low mice develop anemia, osteosclerosis, marrow and spleen fibrosis and neoangiogenesis. From 15-month, extensive extramedullary hematopoiesis in the liver occurs. Myelofibrosis is developed by hemizygous male and heterozygous females alike. In order to evaluate the role of the spleen in the development of extramedullary hematopoiesis, young (6-month) wild-type littermates (WT) and GATA-1low mice (both hemizygous males and heterozygous females) were splenectomized. Three- and 9–12 months later, the myelofibrotic trait expressed by the treated animals was compared with that of un-manipulated controls. Although all the mice survived surgery, a dramatic difference in mortality was observed between hemizygous males, which died of severe anemia within 2-weeks, and heterozygous females (no death recorded). Heterozygous GATA-1low females developed a modest anemia 9–12-months after surgery (Htc, 40.1±2.7% vs. 48.2±5.5 in splenectomized GATA-1low and WT littermates, respectively, p<0.01), comparable to that presented by the age-matched untreated mutants. Platelet counts increased 2-to-3-fold in both WT and GATA-1low mice, but remained significantly lower than normal in the latter. There was an increase in the number of CD16+ NK cells in the blood of splenectomized GATA-1low mice (287±61 vs. 61±19 CD16+ cells/mL in splenectomized GATA-1low and WT littermates, respectively), but no other abnormality in blood cell counts was observed. At 3 months post-splenectomy, the number of bone marrow cells in GATA-1low mice increased up to normal [from 4.0(±1.2)x106 to 8.0(±1.3)x106 cells/femur in untreated and splenectomized mutants, respectively]. Histologic evaluation of bone sections showed no overt change in the extent of marrow fibrosis compared to age-matched untouched mutants, while the extent of trabecular bone was reduced. No fibrosis developed in WT mice at any time point after splenectomy, nor were changes in total marrow cellularity observed. There was no sign of extramedullary hematopoiesis at 3 month post-splenectomy in either WT or GATA-1low mice, while at the later time point, mutant mice showed a prominent involvement of the liver with high numbers of megakaryocytes, isolated or in clusters, and maturing cells of all hematopoietic lineages, that in extreme cases diffusely infiltrated the parenchyma. Extramedullary hematopoiesis was not found in either the lung or the kidney. In conclusions, splenectomy affected the extent but not the timing of the development of extramedullary hematopoiesis in GATA-1low mice suggesting that the number of stem cells present in the marrow of old splenectomized animals was sufficient to colonize the liver. The observation that splenectomy per se does not result in extramedullary hematopoiesis support the hypothesis that it is not fibrosis of the spleen, but a second hit within the stem cell compartment, that is possibly responsible for the development of extramedullary hematopoiesis in this mouse model of myelofibrosis.

Author notes

Corresponding author

Sign in via your Institution