Application of dorsal skin fold chamber microvascular thrombosis model in JAK2V617F mice Free

Individuals with JAK2^V617F+ myeloproliferative neoplasm (MPN) exhibit symptoms of microvascular stasis that include headaches, itch, erythromelalgia, and visual disturbances. JAK2^V617F+ individuals also have increased bleeding and clotting risk. JAK2^V617F+ transgenic mice have increased thrombosis; however, bleeding has limited use of traditional thrombosis models. Dorsal skin fold chambers (DSFC) have been used in sickle cell mouse models to assess microvascular stasis. We sought to determine if DSFC could be applied to model microvascular occlusion in JAK2^V617F+ mice and to characterize the baseline endothelial-related pro-thrombotic state.

Methods: C57BL/6 (control n=10) and JAK2^V617F+ transgenic mice which express one (JAK2^VF, n=22) or two copies (JAK2^VF/VF, n=3) of a human JAK2 (Janus kinase 2) gene carrying the pathological V617F substitution were used.For DFSC implantation, mice were anesthetized, denuded on back, followed by chamber placement and microvessel dissection. Mice were then placed on microscope for vessel enumeration. After recovery, the mice were injected with TNF-α to initiate a pro-inflammatory state. Mice were re-anesthetized and vessel enumeration was completed at 1-4 h post-TNF-α infusion. For n=5 JAK2^VF mice, a subcutaneous injection of 5 mg/mL enoxaparin was added 1 h prior to TNF-α infusion. After 4 h timepoint, mice were euthanized. An additional co-hort of mice, including C57BL/6J (n=4), JAK2^VF (n=18), and JAK2VF/VF (n=4) were sacrificed for blood counts, plasma cytokine analysis, and pathologic organ assessment. For organ assessment, 6 µm sections of inflated mouse lung were stained for CD142 (tissue factor), von Willebrand factor (VWF) and CD31 with immunofluorescence. Five images per animal (n=4) were taken followed by quantitative pixel count analysis using Python-based Napari image analysis.

Results: Ten C57BL6/J mice, 22 JAK2^VF, and 3 JAK2^VF/VF mice underwent DFSC implantation. There was no difference in procedure-related mortality between mouse genotypes, with 2 deaths in the C57BL6/J (22.2%) and JAK2^VF (9.5%) mortality. JAK2^VF/VF had no mortality. 1 JAK2^VF mouse was excluded from analysis due to incidental pregnancy. For bleeding, no C57BL6/J or JAK2^VF/VFmice experienced bleeding; 3 female JAK2^VF mice had minor bleeding. A baseline, there was no difference in microvessel occlusion between genotypes. Starting at 1 hour after TNF-α infusion, there was a significant increase in % microvessels occluded in JAK2^VF and JAK2^VF/VFcompared to C57BL6/J (p<0.05 via Two-Way ANOVA with Tukey's multiple comparisons). Starting at 2 h after TNF-α, compared to respective genotype baseline, JAK2^VFand JAK2^VF/VFmice had significantly higher % microvessels occluded. In comparison, C57BL6/J did not experience any significant change in microvessel flow at any timepoint. Likewise, enoxaparin treatment of JAK2^VF mice prevented microvessel occlusion. Evaluating relationship between microvessel occlusion with age, gender, weight, white blood cell count, hemoglobin, hematocrit, and platelet count variables using least square mean linear regression modeling found that elevated weight was significantly associated increased % occlusion at 2 hours. Thrombin-antithrombin (TAT), a measure of coagulation, was assessed in all mice. Compared to untreated C57BL6/J mice, JAK2^VF and JAK2^VF/VFmice had significantly higher TAT levels. C57BL6/J and JAK2^VF mice treated with TNF-α had significantly increased TAT. Last, quantitative image analysis of mouse lung immunofluorescent staining revealed that compared to C57BL6/J mice, JAK2^VF mice had significant increase in endothelial cell-derived von Willebrand factor (VWF) and pro-coagulant tissue factor (TF).

Conclusions: Overcoming the limitations of other models, JAK2^VFand JAK2^VF/VFmice can undergo DSFC implantation without excessive mortality or bleeding. Weight was found to be associated with % microvascular occlusion in JAK2^VFmice. At 2 h after TNF-α infusion, JAK2^V617F+ mice exhibit increased microvascular occlusion; anticoagulation with enoxaparin prevents occlusion. At baseline, JAK2^VF and JAK2^VF/VFmice exhibit increased pro-coagulant state that includes elevated TAT complexes and increased lung endothelial staining of VWF and TF. Our data confirm that JAK2^VF mice have increased pro-coagulant state and that DFSC can be applied to JAK2^V617F+ mice to model microvascular occlusion.