ALBERT

Description: Thrombopoietin (TPO) is the primary regulator of megakaryopoiesis and therefore also the most important determinant of the number of platelets in circulation. The regulation of TPO blood concentration is complex, with at least a significant component mediated by removal and degradation of the hormone by the mature circulating platelet mass. In this way, if the production of TPO were fixed, when the number of platelets rises, an increased amount of TPO is removed from the plasma, resulting in an inverse relationship between platelet counts and TPO concentration. However, several studies hint at the existence of additional mechanisms in which TPO production is altered in response to physiological and pathologic conditions. We, and others, have previously reported that TPO mRNA is increased in the bone marrow of mice or humans after either immune- or radiation-mediated thrombocytopenia, although hepatic levels remain unchanged by this manipulation. To further explore the mechanism(s) of this effect, we utilized in vitro marrow stromal cell models to study the effects of blood proteins on TPO production. As an initial hypothesis we determined if platelet-derived proteins in serum might suppress TPO production from both a marrow stromal cell line, OP9, or primary murine marrow stromal cells derived from long-term hematopoietic cultures. As assessed by quantitative RT-PCR we found that TPO mRNA levels increase a mean of 2.8-fold±0.7 (n = 3) twelve hours following the removal of serum from the culture. Similar results were obtained from primary murine marrow stromal cells; TPO mRNA levels rose a mean of 2.9-fold±0.9 (n = 4) within sixteen hours of serum deprivation. In contrast, TPO transcript levels were unaffected by the same manipulation of serum in the hepatocyte cell line HEPA1c1c7. As the removal of serum might have induced a stress response in the cells, we tested whether other cell stressors might mimic this response; we found that neither UV irradiation nor treatment with toxic metals such as nickel, cobalt, or cadmium produced any rise in TPO mRNA in marrow stromal cells. Furthermore, to test whether a cell cycle arrest triggered by serum deprivation might mediate these effects, we treated stromal cells with cell cycle inhibitors, but failed to find any affect on TPO transcript levels in stromal cells. Further biochemical fractionation of serum suggested that one or more distinct proteins is responsible for this effect, demonstrated by the ability of both ammonium sulfate precipitation and ion exchange chromatography to partition the suppressive effects of serum. However, the active agent in serum is not TPO itself, as the addition of 150 ng/ml of the pure hormone did not suppress stromal cell TPO transcript levels. Knowledge of this novel regulatory mechanism should be useful in treating platelet disorders and perhaps also during stem cell transplantation, a setting in which TPO is known to play a vital role.

Description: Thrombopoietin (TPO), the primary regulator of megakaryocyte (MK) and platelet formation, modulates the activity of multiple signal transduction molecules, including those in the Jak/STAT, p42/p44 MAPK, and phosphatidylinositol 3-kinase (PI3K)/Akt pathways. In the previous study, we reported that PI3K and Akt are necessary for TPO-induced cell cycle progression of primary MK progenitors. The absence of PI3K activity results in a block of transition from G1 to S phase in these cells (Geddis AE et al. JBC2001276:34473–34479). However, the molecular events secondary to the activation of PI3K/Akt responsible for MK proliferation remain unclear. In this study we show that FOXO3a and its downstream target p27Kip1 play an important role in TPO-induced proliferation of MK progenitors. TPO induces phosphorylation of Akt and FOXO3a in both UT-7/TPO, a megakaryocytic cell line, and primary murine MKs in a PI3K dependent fashion. Cell cycle progression of UT-7/TPO cells is blocked in G1 phase by inhibition of PI3K. We found that TPO down-modulates p27Kip1 expression at both the mRNA and protein levels in UT-7/TPO cells and primary MKs in a PI3K dependent fashion. UT-7/TPO stably expressing constitutively active Akt or a dominant-negative form of FOXO3a failed to induce p27Kip1 expression after TPO withdrawal. Induced expression of an active form of FOXO3a resulted in increased p27Kip1 expression in this cell line. In an attempt to assess whether FOXO3a has an effect of MK proliferation in vivo, we compared the number of MKs in Foxo3a-deficient mice and in wild type controls. Although peripheral blood cell counts of erythrocytes, neutrophils, monocytes and platelets were normal in the Foxo3a-deficient mice, total nucleated marrow cell count of Foxo3a-deficient mice were 60% increased compared with wild type controls. In addition, the increase of MKs was more profound than that of total nucleated marrow cells; CD41+ MKs from Foxo3a-deficient mice increased 2.1-fold, and mature MKs with 8N and greater ploidy increased 2.5-fold, compared with wild type controls. Taken together with the previous observation that p27Kip1-deficient mice also display increased numbers of MK progenitors, our findings strongly suggest that the effect of TPO on MK proliferation is mediated by PI3K/Akt-induced FOXO3a inactivation and subsequent p27Kip1 down-regulation in vitro and in vivo.

Description: Thrombopoietin (TPO) is essential for normal megakaryopoiesis, and mice and humans lacking the TPO receptor c-Mpl have significantly impaired platelet production. However, in the c-Mpl-null mouse model platelet counts, while reduced to ~10% of normal, are not zero, suggesting that another cytokine is able to support some degree of residual thrombopoiesis. We and others have reported that elimination or severe reduction of stem cell factor, G-CSF, IL-3, IL-6 or IL-11 does not eliminate residual thrombopoiesis. Because megakaryocytes (MKs) and erythrocytes are derived from a common progenitor, we asked if erythropoietin (EPO) can stimulate thrombopoiesis in c-Mpl-null mice. We administered 90 u recombinant EPO or vehicle by subcutaneous injection every 3 days to c-Mpl-null or WT control mice and measured baseline and weekly platelet counts. In three independent experiments, at 2 weeks platelet counts in c-Mpl-null mice receiving EPO were significantly higher that at baseline (5–7 mice per group, average of mean platelet counts 425,000/mm3 vs. 285,000/mm3, p=0.0015). There was a trend towards higher platelet counts in WT mice receiving EPO but this did not reach statistical significance. No difference in platelet counts was observed in mice injected with vehicle. In one experiment c-Mpl-null or WT mice were injected with EPO for 4 weeks and the platelet response in the c-Mpl-null animals was sustained for the duration of the experiment. Western blotting showed that murine MKs express the EPO receptor. To determine if EPO stimulates MK production directly we stimulated WT murine MKs in vitro with either 6 u/ml EPO, 100 ng/ml rhTPO or both and monitored activation of ERK and STAT5 signaling by immunoblotting. Stimulation of MKs with EPO resulted in phosphorylation of ERK and STAT5 (15- and 14-fold above baseline, respectively), compared to TPO (97- and 75-fold above baseline). Stimulation with EPO and TPO together had an additive effect (phospho-ERK increased 121-fold and phospho-STAT5 increased 100-fold). To determine if EPO acts primarily on early or late MKs, we harvested bone marrow from c-Mpl-null mice after 2 weeks of treatment with EPO or vehicle and measured CFU-MK frequency and MK ploidy. Although there was a small increase in the frequency of CFU-MK in mice treated with EPO compared to vehicle, these differences were not significant (n=3, p=0.7), possibly due to the difficulty in assaying CFU-MK in vitro without TPO. In addition, EPO did not significantly enhance MK ploidy in c-Mpl-null mice, although MKs in the 32N and greater peaks were slightly more numerous. Therefore, we conclude that EPO can augment platelet production in the absence of c-Mpl signaling, although it is not yet clear if EPO primarily acts on early or late cells. Additional experiments are underway to determine if ablating EPO receptors in a TPO-null mouse model will eliminate residual thrombopoiesis. These findings may have clinical relevance for treating patients with congenital amegakaryocytic thrombocytopenia and other causes of thrombocytopenia in which c-Mpl signaling is impaired.

Description: Megakaryopoiesis is tightly regulated by a number of hematopoietic growth factors to maintain a physiological level of circulating platelets. Thrombopoietin (TPO) is the primary regulator of megakaryopoiesis, supporting the proliferation and survival of hematopoietic stem cells, driving megakaryocyte differentiation and promoting endomitosis and proplatelet formation. However, recent findings suggest that the chemokines fibroblast growth factor-4 and stromal-derived factor-1 (SDF-1) can partially restore thrombopoiesis in the absence of TPO. These chemokines enhance survival and maturation of megakaryocyte progenitors, as well as platelet release, by promoting progenitor cell movement from the osteoblastic to the vascular niche. However, little is known at present of the molecular mechanisms involved in controlling megakaryocyte motility. Focal adhesion kinase (FAK) is essential for the migration of most cells as it mediates the assembly and disassembly of focal adhesions and we have found it to be highly enriched in megakaryocytes compared to other cells in the bone marrow. Additionally, FAK is activated by SDF-1, a key regulator of chemotaxis, and in this study we found it phosphorylated (activated) by TPO specifically at Y-577 and Y-925. Therefore FAK may be required for megakaryocyte progenitor chemotaxis from the osteoblastic to the vascular niche. In order to determine the potential role of FAK in murine thrombopoiesis we used Cre/loxP technology to conditionally delete fak specifically from megakaryocytes. Mice expressing floxed fak alleles were crossed to mice expressing Cre recombinase under the control of the platelet factor 4 (PF4) promoter; the progeny failed to express detectable levels of FAK in megakaryocytes by western blotting or immunofluorescence, whilst expression of the gene was unaltered in other tissues, including heart, liver, lung and spleen. While the platelet counts of the mutant mice were normal at steady-state, multiple compensatory mechanisms could be operative. In fact, using megakaryocyte colony assays, we observed a 4-fold increase in the number of colony forming unit-megakaryocytes (CFU-MK) in mice in which fak has been specifically deleted. To clarify the function of FAK in mature megakaryocytes, total bone marrow collected from these and control animals was grown in TPO-containing culture medium for 72 hours and mature megakaryocytes were isolated on a BSA-density gradient. No difference in megakaryocyte adhesion to fibrinogen or fibronectin was found between cells isolated from controls and mutant mice. However, chemotaxis assays using transwell-inserts with an SDF-1α gradient showed a statistically significant increase in chemotaxis in fak null megakaryocytes, compared to controls, suggesting that abnormal cell migration could account for the hematopoietic changes noted in the mice. In summary, we have successfully ablated fak specifically from the megakaryocyte lineage in vivo and are currently using this model to determine a role for this protein in megakaryopoiesis. Specific deletion of FAK in these cells enhances CFU-MK formation and promotes chemotaxis of mature megakaryocytes in response to SDF-1α, which has recently been shown to be required for megakaryocyte motility. Further, we have evidence to suggest that FAK activity is, at least in part, regulated by TPO. Therefore, we propose a previously undescribed role for FAK in megakaryopoiesis and megakaryocyte chemotaxis.

Description: Abstract 1737 Myeloproliferative disorders (MPDs) are a heterogeneous group of bone marrow disorders characterized by increases in one or more blood cell lineages. A single, somatic mutation in JAK2 (V617F) is responsible for many of the features of Philadelphia chromosome-negative MPDs (Polycythemia Vera, Essential Thrombocythemia and Primary Myelofibrosis). Clinically, the most common cause of death in these patients is arterial thrombosis; however some patients display a bleeding diathesis. Although the incidence and potential causes of dysfunctional hemostasis in patients with MPDs has been studied extensively, the critical regulating factors are unclear and therefore it has been difficult to develop an effective therapeutic regimen for these complications. As the formation of stable thrombi requires interactions between endothelial cells, platelets and leukocytes, we have recently generated mice that express human JAK2V617F in each of these cell lineages by crossing a JAK2V617F/Flip-Flop (FF1) mouse with mice expressing lineage-specific Cre recombinases. These crosses have generated the following mice; 1) Tie2-Cre/FF1, where JAK2V617F is expressed in all hematopoietic and endothelial cells, 2) Pf4-Cre/FF1, where JAK2V617F expression is limited to platelets, and 3) LysM-Cre/FF1, where JAK2V617F expression is limited to leukocytes. Expression of human JAK2V617F was confirmed in megakaryocytes, platelets, leukocytes and endothelial cells (Tie2-Cre/FF1), megakaryocytes and platelets (Pf4-Cre/FF1) and leukocytes (LysM-Cre/FF1) by conventional and real-time PCR. Of the 3 mouse strains, only Tie2-Cre/FF1 exhibited a MPD phenotype. Platelet counts were significantly increased compared to Tie2-Cre controls (at 3 months, Tie2-Cre: 779 (±61)/ml;Tie2-Cre/FF1: 2943 (± 217)/ml) without significant increases in any other cells types. Tie2-Cre/FF1 mice also exhibit greatly increased number of CFU-MKs and bone marrow derived megakaryocytes. Therefore, Tie2-Cre/FF1 mouse exhibits an ET-like phenotype. Although circulating platelet counts did not increase in Pf4-Cre/FF1 mice, we did observe an increase in the number of CFU-MKs in colony assays. Next we determined the roles of the lineage-restricted JAK2V617F expression on hemostasis in vitro and in vivo. Aggregometry on washed platelets showed no significant difference between any group and their controls in response to PAR4 (100–400mM), ADP (2–20mM) or collagen (1–10mg/ml). Additionally, we were unable to show a significant difference in GPIIbIIIa activation or surface expression of P-selectin in response to the same agonists. Despite no clear platelet abnormalities in any of the 3 mouse lineages, we identified significant hemostatic abnormalities in vivo in Tie2-Cre/FF1 mice. Tail bleeding time was significantly increased in Tie2-Cre/FF1 mice compared to Tie2-Cre controls (Tie2-Cre average, 2min 47secs; Tie2-Cre/FF1, 6mins 37secs) while Tie2-Cre/FF1 mice also exhibited an increased occurrence of re-bleeding compared to Tie2-Cre controls. Additionally, we performed FeCl3 carotid artery occlusion assays to better determine in vivo thrombosis. We found that at 10% FeCl3, Tie2-Cre control mice exhibited complete artery occlusion in approximate 6 min. In contrast, Tie2-Cre/FF1 mice failed to show any sign of arterial occlusion throughout the duration of the experiment (30 min). Given the significant increase in platelet numbers in Tie2-Cre/FF1 mice, we next determined if acquired von Willibrand Disease (VWD) could account for prolonged bleeding and reduced clotting; plasma vW Factor levels by ELISA were normal. In contrast to Tie2-Cre/FF1 mice, neither the PF4-Cre/FF1 or LysM-Cre/FF1 mice exhibit dysfunctional thrombosis. These data provide compelling evidence that expression of JAK2V617F in cells other than just platelets or just leukocytes is necessary to generate the hemostatic abnormalities seen in patients with MPDs. Recent findings show that some patients express endothelial JAK2V617F and patients with ET exhibit increased numbers of circulating endothelial progenitors. Thus, our data is consistent with the hypothesis that expression of JAK2V617F in endothelial cells, in addition to hematopoietic cells results in the bleeding diathesis seen in patients with MPDs. Disclosures: No relevant conflicts of interest to declare.

Description: Blood levels of TPO, the primary regulator of platelet production, are thought to be regulated primarily by cellular degradation following its binding to, and internalization by its cell surface receptor, c-Mpl. This physiology is supported by in vitro studies, in which TPO can be removed from solution by incubation with platelets, and accounts for the very high levels of circulating TPO in individuals with congenital amegakaryocytic thrombocytopenia (CAMT), as in this disorder there are no c-Mpl receptor-bearing cells to bind and remove the hormone. Previous reports have demonstrated that in addition to hematopoietic cells, c-Mpl is expressed on several types of endothelial cells. We hypothesized that the c-Mpl expressed on endothelial cells would contribute to the regulation of circulating TPO levels. As a model to test this hypothesis, we transplanted 10 Mpl knock-out mice and 10 wild-type (WT) controls with WT marrow cells and analyzed platelet counts and plasma TPO levels over the course of 6 months. The resulting two groups of chimeric mice both express c-Mpl on megakaryocytes and platelets, but only the WT recipients express the receptor on endothelial cells. If c-Mpl receptors normally expressed on endothelial cells take up circulating TPO and degrade it, we expected that the Mpl-null mice reconstituted with WT cells would display increased TPO levels and an increased steady state platelet count compared to the WT recipients. As expected, pre-transplant the platelet counts in the Mpl-null mice were low (average value 398x109/L) and TPO levels were high (average value 8040 pg/ml), as compared to WT controls (platelet counts average 883x109/L, TPO level 745 pg/ml). However, following transplant with WT cells, we found that TPO levels in the chimeric mice were virtually identical (954 pg/ml for the Mpl-null recipients vs. 829 pg/ml for the WT recipients) as were the steady state platelet counts up to 6 months following transplantation (891x109 /L vs. 883x109/Lx109 /L, respectively). These results indicate that the c-Mpl receptor expressed on endothelial cells does not contribute significantly to the regulation of circulating TPO levels or to steady state platelet counts. The apparent lack of TPO adsorbtion by the endothelial cells is surprising, as prior published studies indicate that Mpl receptors expressed on human umbilical vein endothelial cells bind TPO and that c-Mpl is functionally active on hepatic endothelial cells; in addition, the composite surface area of the endothelium is large. These results also imply that patients with CAMT that have successfully engrafted with normal hematopoietic stem cells should have normal (not elevated) TPO levels. Similarly, gene replacement strategies designed to restore c-Mpl in CAMT do not need to be expressed on endothelial cells to establish the normal regulation of TPO.

Description: Thrombopoietin (TPO) is a recently cloned cytokine that binds to its receptor, Mpl, and promotes hematopoietic expansion and maturation, primarily of the megakaryocyte lineage. The signaling pathways responsible for these events are thought to involve the Janus family of nonreceptor tyrosine kinases (JAKs) and the signal transducers and activators of transcription (STATs), which are activated by tyrosine phosphorylation. Previous investigators have studied these molecules in engineered and naturally occurring cell lines. To investigate the molecular basis for TPO signal transduction in a more physiologic target, we determined the pattern of JAK and STAT activation in purified, normal murine megakaryocytes. These results are compared with those of established cell lines that only proliferate (Ba/F3-mMPL and DA-1-TPO) or only differentiate (L8057) in response to TPO. From these findings, a model is proposed to explain the physiologic roles of JAK2, TYK2, STAT3, and STAT5 in TPO signaling. Furthermore, previous studies of the physical interaction between Mpl and the JAKs are extended, showing a difference in the association of JAK2 and TYK2 with the TPO receptor. Finally, we show that, in the cell line Ba/F3-mMPL, the closely related proteins STAT5A and STAT5B are both activated by TPO stimulation and are capable of heterodimerization. Together, these results further our understanding of the early stages of megakaryocyte and platelet development.