Background: Myeloproliferative neoplasms (MPNs) are clonal hematopoietic stem cell disorders, characterized by significant increase in one or more myeloid lineages. A mutation in the Janus kinase 2 (JAK2) gene, JAK2V617F, was identified in half of the patients with Philadelphia chromosome-negative (Ph-) MPNs. This activating mutation causes hypersensitivity to certain growth factors, which explain the increased proliferation of myeloid progenitors. It has been reported that patients with MPNs have an increased risk of thrombosis but also increased microvessel density that may reflect angiogenesis in the spleen and bone marrow (Medinger, Br J Haematol, 2009), with no clear physiopathological explanation. Several recent studies have demonstrated the presence of JAK2V617F not only in blood cells but also in endothelial cells (EC) in these patients (Sozer, Blood, 2009; Teofili, Blood, 2011; Rosti, Blood, 2012).

Hypothesis: We hypothesized that the presence of JAK2V617F in EC could change their properties leading to an increased angiogenesis process in MPNs.

Methods: To determine whether the presence of JAK2V617F in EC was responsible for increased angiogenesis, we used an in vitro approach with human JAK2V617F ECs and an in vivo mouse model. We first used HUVECs (human umbilical vein endothelial cells) transduced with GFP lentivirus encoding human JAK2V617F. An empty lentivirus encoding only for GFP was used as a negative control. Proliferation of HUVECs was quantified during 3 days culture in EGM-2 medium and tube formation after culture in Matrigel™ was assessed by microscopy. Cell migration was determined by microscopy after scratch assay. Proteins expression level in cell lysate and supernatant was determined using Proteome ProfilerTM Array (R&D). The intensity of dot blots was determined by imageJ. For the in vivo approach, we crossed JAK2Flex/WT mice with PDGFβcreERT2 mice to generate endothelial-specific JAK2V617F knock-in mouse line (PDGFβERT2-JAK2 V617F/WT). Our lab was pioneer in developing microCT vascular imaging in order to precisely measure arterial vessel density and organization in 3 dimensions. To analyze neoangiogenesis (in the setting of ischemia), we used the model of mouse hind limb ischemia (Oses, ATVB, 2009): 11 days after ligature of the femoral artery, mice are perfused with latex labeled with barium and the hind limb vascular network is visualized with microCT.

Results: We first observed that JAK2V617F HUVECs proliferate more than controls (coefficient rate of 3,53+/-0,18 versus 1,98 +/-0,05), reminding the phenotype of JAK2V617F hematopoietic cells. We then showed that JAK2V617F HUVECs are able to form more tubular structures in Matrigel™. Using an in vitro cell migration assay, we observed that JAK2V617F HUVECs invaded 45% (+/-2.3%) of the total surface area versus 27% (+/-3.3%) for controls. To confirm these results obtained in vitro, we analyzed the vascular network after hind limb ischemia in mice expressing JAK2V617F mutation specifically in endothelial cells (PDGFβERT2-JAK2V617F/WT). Our first results show an increased neoangiogenesis and further results will be presented at the conference. In order to decipher the mechanism responsible for the increased angiogenesis, we then quantified the expression of proteins that regulate angiogenesis, either in transduced HUVECs or in the culture supernatant. Interestingly, we have demonstrated a greater secretion of angiogenin, PDGF-AA, Endostatin, IGFBP-1, MCP-1 and CXCL-16 by HUVEC JAK2V617F.

Discussion: In summary, our data suggest that the presence of JAK2V617F mutation in EC modifies their properties toward a pro-angiogenic profile and can explain, at least in part, the reported increase of angiogenesis in MPN patients. Once the role of JAK2V617F mutation in modifying EC properties will be confirmed, further work will be required to identify the molecular mechanisms underlying these phenotypic changes. It will also be particularly important to investigate if ECs are involved in the pathogenesis of increased angiogenesis observed in other diseases. Thus, they could be a new target in the treatment of pathological neo-angiogenesis.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.