No abstract

The tyrosine kinase JAK2 is a critical component of intracellular JAK/STAT cytokine signaling cascades that is prevalent in hematopoietic cells, such as hematopoietic stem cells and megakaryocytes (MKs). Individuals expressing the somatic JAK2 V617F mutation commonly develop myeloproliferative neoplasms (MPNs) associated with venous and arterial thrombosis, a leading cause of mortality. The role of JAK2 in hemostasis remains unclear. We investigated the role of JAK2 in platelet hemostatic function using Jak2fl/fl Pf4-Cre (Jak2Plt−/−) mice lacking JAK2 in platelets and MKs. Jak2Plt−/− mice developed MK hyperplasia and splenomegaly associated with severe thrombocytosis and bleeding. This notion was supported by failure to occlude in a ferric chloride carotid artery injury model and by a cremaster muscle laser-induced injury assay, in which Jak2Plt−/− platelets failed to form stable thrombi. Jak2Plt−/− platelets formed thrombi poorly after adhesion to type 1 collagen under arterial shear rates. Jak2Plt−/− platelets spread poorly on collagen under static conditions or on fibrinogen in response to the collagen receptor GPVI-specific agonist, collagen-related peptide (CRP). After activation with collagen, CRP, or the CLEC-2 agonist rhodocytin, Jak2Plt−/− platelets displayed decreased α-granule secretion and integrin αIIbβ3 activation or aggregation, but showed normal responses to thrombin. Jak2Plt−/− platelets had impaired intracellular signaling when activated via GPVI, as assessed by tyrosine phosphorylation. Together, the results show that JAK2 deletion impairs platelet immunoreceptor tyrosine-based activation motif signaling and hemostatic function in mice and suggest that aberrant JAK2 signaling in patients with MPNs affects GPVI signaling, leading to hemostatic platelet function.

Background Reduced plasma survival of von Willebrand factor (VWF) is characteristic of patients with type 1C von Willebrand disease (VWD). These subjects can be identified by an increased steady-state ratio of plasma VWF propeptide (VWFpp) to VWF antigen (VWF:Ag). A similar phenotype occurs in mice with the Mvwf1 allele. Objectives To (i) determine if the VWFpp/VWF:Ag ratio can be used to identify a 'type 1C' phenotype in mice, (ii) determine the most reliable method for murine blood sampling, and (iii) identify the source of VWF released during problematic blood collection. Methods 'Platelet-VWF' and 'endothelial-VWF' mice were generated by bone marrow transplantation between C57BL/6J and VWF-/- mice. Several blood sampling methods were used and murine VWFpp and VWF:Ag levels determined. Plasma and platelet VWF:Ag and VWFpp, VWF multimers and VWF half-life were examined in mouse strains with and without Mvwf1. Results A single retro-orbital bleed and vena cava collection were found to be the optimal methods of blood collection. Problematic collection resulted in release of VWF from platelets and endothelium. The VWFpp/VWF:Ag ratio identified strains of mice with reduced VWF survival. Conclusion Assay of murine VWFpp and VWF:Ag has utility in determining the acceptability of murine blood samples for coagulation testing and in identification of a reduced VWF survival phenotype in mice.

The tyrosine kinase JAK2 is a critical component of intracellular JAK/STAT cytokine signaling cascades that is prevalent in hematopoietic cells such as hematopoietic stem and progenitor cells (HSPCs), megakaryocytes (MKs), and platelets. Individuals expressing somatic activating JAK2 mutations such as JAK2V617F commonly develop myeloproliferative neoplasms (MPNs) associated with serious complications, including venous and arterial thrombosis, a leading cause of mortality. Here, we investigated the role of JAK2 in hemostasis and thrombosis using Jak2fl/flPf4-Cre (Jak2Plt-/-) mice specifically lacking JAK2 within the platelet lineage. Jak2Plt-/- mice developed severe thrombocytosis with a 5-fold increase in circulating platelet number, MK hyperplasia, and splenomegaly. Jak2Plt-/-platelets were of normal size and the expression of major membrane surface glycoproteins was indistinguishable from controls, except for the integrin β3, which was reduced by 20%. Despite the thrombocytosis, Jak2Plt-/- mice had a severe bleeding diathesis, as evidenced by: 1) prolonged tail bleeding time; 2) failure to occlude in a ferric chloride-induced carotid artery injury model; and 3) failure to form stable thrombi in a laser-induced cremaster muscle injury model. Jak2Plt-/- platelets spread poorly on immobilized collagen or on immobilized fibrinogen following GPVI stimulation with the collagen-related peptide (CRP). Jak2Plt-/- platelets had defective α-granule secretion and integrin αIIbβ3 activation following stimulation with CRP, but not thrombin, and showed aggregation defects with low-doses of CRP. Together, the data support a GPVI-specific impairment in platelets lacking JAK2, a notion that was supported by impaired intracellular signaling following GPVI stimulation, as assessed by protein tyrosine phosphorylation. Jak2Plt-/- platelets adhered poorly to type I collagen under arterial shear rates in whole blood. However, JAK2 deletion in platelets did not alter plasma von Willebrand factor (VWF) levels or botrocetin-mediated binding of plasma VWF to GPIbα. Together, the results underline a critical role for JAK2 in platelet GPVI signaling and hemostatic function, which likely contributes to venous and arterial thrombosis observed in patients with MPNs with the activating JAK2V617F mutation. Disclosures No relevant conflicts of interest to declare.

One mechanism causing type 1 VWD is the reduced survival of VWF in plasma (type 1C VWD), characterized by markedly decreased VWF:Ag and VWF half-life, essentially normal multimers, increased ratio of VWF propeptide (VWFpp) to VWF:Ag, robust response to DDAVP, and normal ratios of VWF:CB, FVIII, or VWF:RCo to VWF:Ag. We enrolled 502 index cases with a pre-existing diagnosis of type 1 VWD through the Zimmerman Program for the Molecular and Clinical Biology of VWD. We confirmed 262 of the index cases as type 1 VWD (VWF:Ag or VWF:RCo ≤ 40 IU/dL). Of these, 58 met the criteria for type 1C VWD with VWFpp/VWF:Ag ≥ 3 and VWF:Ag ≤ 30 IU/dL. Sequence variations were identified in the VWF D3, A1, A2, and D4 domains. Little is known regarding the mechanisms causing type 1C VWD, but it has been assumed that VWF undergoes normal intracellular processing and secretion with rapid clearance upon release into plasma. We hypothesized that defective intracellular processing may contribute to the type 1C phenotype. We studied 10 type 1C variants including C1130Y, W1144G, R1205H, N1231S, R1315C, V1411E, R1527W, N2041S, Y2160C, and S2179F. Variants were expressed alone (homozygously) or with wild-type (WT) VWF (heterozygously) in HEK293T cells and VWF secretion, multimer structure, and binding to collagen (types III and VI), GPIb-alpha, and FVIII was analyzed. To assess regulated storage, variants were expressed homozygously in HEK293 cells where WT VWF forms elongated pseudo-Weibel-Palade bodies (pWPB). Five variants (C1130Y, R1315C, V1411E, N2041S, Y2160C) had severely decreased secretion and defective multimerization when homozygously expressed. These variants did not form pWPB, but appeared to co-localize with the endoplasmic reticulum, consistent with the severely impaired secretion. One variant, W1144G, had mildly reduced secretion, formed only dimeric VWF, and unexpectedly did not form pWPB. These multimer defective variants demonstrated decreased collagen binding and GPIb-alpha binding as would be predicted. The remaining variants (R1205H, N1231S, R1527W, S2179F) were normally secreted, multimerized, stored in pWPB, and had normal binding to FVIII, collagen, and GPIb-alpha. Interestingly, FVIII binding to homozygous VWF D3 variants C1130Y and W1144G was substantially reduced. This result is not entirely unexpected as the FVIII binding region in VWF has been mapped to the D’-D3 region. Co-expression with WT VWF essentially corrected defective secretion, although some variants still had moderately reduced secretion. Multimer structure appeared normal for all heterozygous variants, although staining which discriminates between variant and WT alleles revealed that for some variants, little variant VWF was actually expressed when transfected at a 1:1 ratio with WT. In sum, when variants were homozygously expressed, we observed a constellation of processing and functional defects. Only R1205H, N1231S, R1527W, and S2179F variants demonstrated normal processing and function. Heterozygous expression (consistent with patients) corrected most of the observed defects, although reduced secretion persisted for a subset of variants. We can conclude that while reduced plasma survival of VWF is a major determinant of the type 1C phenotype, additional upstream processing defects may contribute to the severity of the overall VWD phenotype. Disclosures: No relevant conflicts of interest to declare.