Under normal conditions, humans maintain a blood content of 150-400 x 109 platelets per liter, whereas mice can reach 1000 x 109 platelets per liter. Thrombocytopenia occurs when the level of platelets becomes too low, a situation that increases the risk of spontaneous bleeding and hemorrhage. Although the number of platelets is continuously very high, they are produced by a rare cell population the megakaryocytes (MKs), which in turn are produced by megakaryocyte-erythrocyte precursors (MEPs) in the bone marrow. The first report of a potential role for the zinc finger transcription factor Gfi1b in megakarypoiesis and thrombopoiesis showed that the full knock-out of this gene in mice leads to a severe impairment of both erythropoiesis and megakaryopoiesis, translating into a severe thrombocytopenia and a lethality by day e15.5 (Saleque et al Genes Dev 2002). Because of this developmental arrest at mid-gestation, the function of Gfi1b in adult differentiated hematopoietic cells could not be analyzed. We have thus generated conditionally deficient mice carrying floxed Gfi1b alleles, to study the role of Gfi1b in adult hematopoiesis. To ablate Gfi1b expression, we crossed Gfi1b flox/flox mice with animals carrying two different Cre transgenes: a ROSA-Cre-ERT2 transgene that quickly deactivates Gfi1b in all cells upon treatment with tamoxifen, allowing to almost instantly measure effects on already differentiated MKs and ii) a PF4-Cre transgene that constitutively expresses the Cre specifically in MKs, excluding any adverse effects due to other cell types.
With these mice, we could show that Gfi1b ablation leads to a strong proliferation and expansion of both MEPs, MK precursors and MKs, which was surprisingly associated with an almost complete loss of platelets (∼99.9% reduction compared to controls in PF4-Cre, Gfi1bflox/flox mice). Most striking was the strong increase in the number of MKs, even from an early stage of differentiation, when Gfi1b was deleted. It was not clear, however, if this expansion of MKs was caused directly by the loss of Gfi1b in early progenitors, or if it was a consequence of the severe thrombocytopenia that could stimulate megakaryopoiesis through a feedback loop. An in vivo kinetic study of platelet loss and MK proliferation in the ROSA-Cre-ERT2, Gfi1bflox/flox mice revealed that the expansion of MKs started as early as 3 days after administration of tamoxifen (over 4-fold increase compared to age matched controls). On the other hand, platelet counts started only to decrease noticeably 4 days after tamoxifen administration (about 400 x 109/L vs 1200 x109/L), although reticulated platelets started to decrease as early as day 2 after tamoxifen injection and almost completely disappeared by day 4 after tamoxifen treatment. This suggests an arrest in platelet release in the absence of Gfi1b. After 4 days of tamoxifen treatment, platelet counts decreased quickly to reach a minimum around day 7-8 (60 x 109/L vs 1350 x 109/L in controls). By this time, the number of MKs literally exploded to reach levels up to 15 fold higher than in wild type controls. These results indicate that the number of MKs increases prior to the decrease in circulating platelet when Gfi1b is deleted. However, the results may also suggest that a feedback loop could contribute to this phenomenon by boosting MK expansion upon platelet loss. A platelet lifespan analysis on the rare remaining platelets in PF4-Cre, Gfi1bflox/flox mice revealed that the low platelet level was not due to accelerated platelet clearance, confirming that the platelet loss was the result of an arrest in platelet release. Gfi1b deficient MKs are still polyploidy, but are significantly smaller and have a different nuclear cytoplasmic ratio than their wt counterparts. In addition, in contrast to normal wt MKs, Gfi1b deficient MKs were unable to properly spread or migrate on fibronectin or fibrinogen surfaces, showed lower F-actin content and an increased expression of the platelet glycoprotein IIb of IIb/IIIa complex (CD41/CD61) on their surface. These data suggest that Gfi1b controls the signaling of the platelet specific integrins to restrict MK proliferation, control MK size and their ability to produce platelets.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.