Megakaryocytes Regulate Hematopoietic Stem Cell Quiescence Via PF4 Secretion

Hematopoietic stem cells (HSCs) adapt to the varying needs for the replenishment of terminally differentiated blood cells by switching between quiescence, self-renewal and differentiation. A tight regulation of this balance is crucial to ensure hematopoietic homeostasis over a lifetime. In the bone marrow (BM), HSCs reside in specialized microenvironments, or niches, that regulate HSC behavior. Although all niche cells identified thus far have a non-hematopoietic origin (e.g., mesenchymal stem cells, osteoprogenitors, endothelial cells and sympathetic nerves), less is known about HSC regulation by its progeny. During confocal imaging studies of whole-mount BM, we have noted an association between HSCs and megakaryocytes (Mk); indeed up to 50% of Lin^-CD41^-CD48^-CD150⁺ HSC were in direct contact with Mk. HSC were located at an average of 15.4 µm-distance to Mk. This distribution was not random (p=6x10^-8) as demonstrated by two-sample Kolmogorov-Smirnov test. To assess the functional role of Mk in HSC regulation in vivo, we bred PF4-cre mice (in which the Cre recombinase gene is under the control of the Mk-specific platelet factor 4 (PF4) promoter) with iDTR mice (where cre-mediated recombination induces expression of the diphtheria toxin (DT) receptor). After 7 days of DT treatment PF4-cre:iDTR mice showed specific BM Mk depletion (5.3-fold reduction; P<0.001), followed by a concomitant reduction in platelets (7.0-fold reduction; P<0.0001). Strikingly, Mk depletion led to a 11.5-fold increase (P<0.001) in the number of Lin^-c-Kit⁺Sca-1⁺CD105⁺CD150⁺ HSC due to a 5.5-fold increase in proliferation (determined by BrdU incorporation; P<0.001). Since loss of quiescence has been frequently associated with reduced HSC potential, we tested whether the increase in HSC proliferation after Mk depletion impaired HSC function by competitive transplantations in limiting dilutions. Extreme limiting dilution analysis (ELDA) revealed 625 HSCs with repopulating capacity per femur in DT-treated PF4-cre:iDTR mice which is 4.8-fold more (P<0.01) than in DT-treated PF4-cre transgenic control mice. Importantly, Mk depletion did not result in aberrant hematopoiesis as reflected by WBC and RBC counts in the peripheral blood and progenitor cell (CFU-E, GMP, Pre-GM, PreMegE, Pre CFU-E ) numbers in the BM that were indistinguishable from WT mice. To ascertain whether the effect was mediated by circulating platelets rather BM Mk, we injected mice with neuraminidase, which selectively depletes platelets. Neuraminidase treatment led to a 13.9-fold reduction in blood platelets (P<0.0001), but neither BM Mk nor HSC were significantly altered. Thus, Mk and not platelets regulate BM HSC quiescence. To investigate the mechanisms through which Mk regulate HSC, we screened for factors that have been shown to maintain cell quiescence in several cell types. Of those, PF4 itself, which is expressed exclusively by Mk, had the greatest reduction in the bone marrow extracellular fluid (BMEF) upon Mk depletion as determined by ELISA analyses. We therefore hypothesized that PF4 contributed to Mk-mediated maintenance of HSC quiescence. We first carried out in vitro culture experiments which revealed that recombinant murine (rm) PF4 restrained HSC proliferation in serum-free culture conditions (3.1 fold reduction; P<0.01). This effect was abrogated when PF4 was neutralized by heparin. Following 7 days of rmPF4 injections, mice exhibited a 1.5 fold reduction in HSC numbers (P<0.0002). We confirmed the reduction of functional HSCs by competitive transplantations. Sixteen weeks after transplantation recipient mice transplanted with BM cells from PF4-treated mice showed a significant reduction in the percentage of donor-derived cells (p<0.01). Next, we analyzed HSC numbers in Pf4^-/- mice or in huPF4-Tg mice (which overexpress total PF4) in comparison to WT controls. Consistent with its role in HSC quiescence, Pf4^-/- showed increased HSC numbers (1.5 fold; P<0.01) and enhanced proliferation (1.6 fold, P<0.05). By contrast, overexpression of PF4 in huPF4-Tg mice led to 1.5-fold (P<0.01) and 2.3-fold (P<0.05) reductions in HSC numbers and proliferation, respectively. These studies demonstrate a feedback loop where an HSC progeny, the megakaryocyte, can directly regulate HSC pool size and proliferation via PF4.