Endothelial Cells and Bleeding in Myeloproliferative Neoplasms

The Philadelphia chromosome-negative myeloproliferative neoplasms (MPNs) polycythemia vera (PV), essential thrombocythemia (ET), and primary myelofibrosis are clonal hematopoietic stem cell disorders. Mutations in the Janus kinase 2 (JAK2) gene, especially the JAK2V617F mutation, are common in these disorders, leading to hyperactive kinase activity and overproduction of myeloid progenitor cells. Patients with ET or PV have a higher incidence of both thrombosis and bleeding. Considerable effort has been devoted to understanding the hemostatic abnormalities in these patients. Increases in intrinsic prohemostatic functions of platelets and neutrophils have been implicated as causal factors for thrombosis. JAK2 V617F+ expressing endothelial cells (ECs) have been reported in a subpopulation of MPN patients, and EC expression was coupled with an increased risk of thrombosis Additionally, activation of ECs leading to increased production of P-selectin and E-selectin and decreased production of nitric oxide has been identified in some patients. Conversely, the bleeding diathesis in some patients has been attributed to acquired von Willebrand syndrome, which is due in part to thrombocytosis.

To determine which cell types contribute to hemostatic disorders associated with MPNs, Dr. Leah Etheridge and colleagues in the laboratory of Ian Hitchcock in the Department of Medicine at Stony Brook University used the FF1 transgenic mouse model to express human JAK2 V617F in specific cell lineages. The FF1 transgenic mouse was developed in the laboratory of Radek Skoda at University Hospital in Basal, Switzerland.¹ It contains an inactive gene encoding JAK2 V617F combined with antiparallel loxP sites. In the presence of Cre recombinase the region of the JAK2 V617Fgene flanked by the loxP sites is inverted, activating the gene. Etheridge et al. crossed FF1 mice with either Pf4-Cre or Tie2-Cre mice, which express platelet-specific or hematopoietic/EC-specific Cre recombinase, respectively. Thus, the authors were able to produce mice that express JAK2 V617F exclusively in platelets or in hematopoietic/EC cells. Tie2-Cre/FF1 mice developed a MPN phenotype characterized by increased megakaryopoiesis, thrombocytosis, splenomegaly, neutrophilia, and osteopetrosis. In contrast, Pf4-Cre/FF1 mice did not develop an MPN phenotype, which the authors attributed to the relatively late stage of activation during hematopoietic development of Cre recombinase under control of the PF4 promoter. 

The authors used a ferric chloride–induced carotid artery occlusion model to compare thrombosis in wild-type, Tie2-Cre/FF1, and Pf4-Cre/FF1 mice. They observed that Tie2-Cre/FF1 mice had a marked increase in time to occlusion compared to wild-type mice. This result was surprising because although increased bleeding is observed in some MPN patients, thrombosis is more common. In contrast, the time to occlusion was normal in Pf4-Cre/FF1 mice. Tail snip bleeding times were also significantly increased in Tie2-Cre/FF1, but not in Pf4-Cre/FF1 mice compared to wild-type mice. 

Platelet function studies were performed to assess possible hemostatic abnormalities. Whole-blood aggregometry revealed an increase rate and extent of aggregation in response to epinephrine, collagen and ADP in Tie2-Cre/FF1 compared to wild-type or Pf4-Cre/FF1 mice. This effect was attributed to thrombocytosis because aggregometry normalized using equal numbers of isolated platelets. No significant differences in expression of platelet integrins β3, αIIb, β1, α2, or activated αIIbβ3 were observed between wild-type and Tie2-Cre/FF1 mice.

The authors then performed reciprocal bone marrow transplantation experiments to determine if ECs were involved in the bleeding/antithrombotic phenotype observed in the Tie2-Cre/FF1 mice. Marrow from Tie2-Cre/FF1 mice was transplanted into irradiated wild-type mice to produce mice in which JAK2 V617F expression was limited to hematopoietic cells. Donor marrow from wild-type mice served as a control. Recipient mice expressing hematopoietic cell–derived JAK2 V617F developed a MPN phenotype characterized by thrombocytosis and neutrophilia. However, the carotid artery occlusion and tail bleeding time assays demonstrated no significant differences between the Tie2-Cre/FF1 and wild-type donor mice. In the reciprocal transplant, wild-type marrow was transplanted into Tie2-Cre/FF1 recipients, producing mice in which only ECs express JAK2 V617F. Wild-type recipients served as a control. Mice expressing EC JAK2 V617F did not develop a MPN phenotype. However, carotid artery occlusion assays showed a significant increase in occlusion time in Tie2-Cre/FF1 recipients, but the tail bleeding time was not prolonged in Tie2-Cre/FF1 recipients. The observation that the hemostatic abnormalities in these mice were less pronounced than in the Tie2-Cre/FF1 mice suggests an interaction between JAK2 V617F-expressing hematopoietic cells and ECs. 

Because ECs express von Willebrand factor (VWF) and acquired von Willebrand syndrome is observed in patients with MPN, the authors determined whether the Tie2-Cre/FF1 mice displayed characteristics of acquired von Willebrand syndrome. Levels of plasma VWF were normal in Tie2-Cre/FF1 mice. However, Tie2-Cre/FF1 mice displayed a significant reduction in ultralarge VWF multimers. Additionally, whole blood ristocetin-induced platelet agglutination was reduced in Tie2-Cre/FF1 mice, consistent with the defect in VWF function observed in acquired von Willebrand syndrome. Levels of ADAMTS13, which catalyzes the proteolysis of VWF into smaller multimers, were not increased in primary lung ECs isolated from Tie2-Cre/FF1 mice. The shift in VWF multimer distribution seen in Tie2-Cre/FF1 mice was not observed in either of the bone marrow transplant models. Additionally, primary lung ECs isolated from mice expressing JAK2 V617F only in ECs showed increased levels of VWF with a normal multimer pattern compared with controls. These results suggest that expression of JAK2 V617F in both hematopoietic and endothelial compartments contributes to the abnormal distribution of VWF multimers. However, transplant mice whose JAK2 V617F expression was limited to ECs showed a defective response to ristocetin. In contrast, the response to ristocetin was normal in mice expressing JAK2 V617F restricted to hematopoietic lineages.

Together, the results suggest that JAK2 V617F mediated changes in EC-derived VWF processing contribute to the changes associated with acquired von Willebrand syndrome. JAK2 has been shown to mediate endothelial nitric oxide synthase activity via AKT activation. eNOS in turn has been shown to be involved in a pathways affecting the exocytosis of Weibel-Palade bodies,²  which store VWF. Additional experiments may reveal other ways in which JAK2 V617F contributes to abnormalities in EC signaling pathways involved in hemostasis.