In Vivo Role of Neutrophil Extracellular Traps in Antiphospholipid Antibody-Mediated Venous Thrombosis

Abstract

Objective: Antiphospholipid syndrome (APS) is a leading acquired cause of thrombotic events. Although antiphospholipid antibodies have been shown to promote thrombosis in mice, the role of neutrophils has not been explicitly studied. The aim of this study was to characterize neutrophils in the context of a new model of antiphospholipid antibody-mediated venous thrombosis.

Methods: Mice were administered fractions of IgG obtained from patients with APS. At the same time, blood flow through the inferior vena cava was reduced by induction of stenosis. Resulting thrombi were characterized for size and neutrophil content. Circulating factors and the vessel wall were also assessed.

Results: As measured by both thrombus weight and thrombosis frequency, mice treated with IgG from patients with APS (APS IgG) demonstrated exaggerated thrombosis as compared with control IgG-treated mice. Thrombi in mice treated with APS IgG were enriched for citrullinated histone H3 (a marker of neutrophil extracellular traps [NETs]). APS IgG-treated mice also demonstrated elevated levels of circulating cell-free DNA and human IgG bound to the neutrophil surface. In contrast, circulating neutrophil numbers and markers of vessel wall activation were not appreciably different between APS IgG-treated mice and control mice. Treatment with either DNase (which dissolves NETs) or a neutrophil-depleting antibody reduced thrombosis in APS IgG-treated mice to the level in control mice.

Conclusion: These data support a mechanism whereby circulating neutrophils are primed by antiphospholipid antibodies to accelerate thrombosis. This line of investigation suggests new, immunomodulatory approaches for the treatment of APS.

Figure 1

APS IgG promotes venous thrombosis. A, Schematic of the inferior vena cava (IVC) stenosis model, in which a ligature is fastened just distal to the renal veins to introduce a standard restriction in flow. B, IgG was purified from either APS patients or healthy controls and administered to mice (2 intraperitoneal injections, 48 hours apart). At the time of the second injection, the IVC was narrowed (stenosed), as described in panel A. 48 hours after stenosis, thrombus weight was measured. Each data point represents an individual mouse. Horizontal lines denote group means. APS groups were compared to the control group; *p<0.05, **p<0.01. C, Ultrasound imaging of the IVC at time=0 (left) or 48 hours later (for IgG-injected mice). In the Doppler images, color (either red or blue) indicates active flow. The yellow dashed lines delineate areas in which flow is excluded due to thrombosis. Images are representative of 3 mice per group. D, Representative thrombus sections for APS and control mice stained with H&E (4× images stiched together). For APS, the left two thrombi are from the APS 1 group and the right two from the APS 5 group. Note the mix of RBC-rich (red) and RBC-poor (pink) areas.

Figure 2

APS IgG promotes the release of NETs in vitro and in vivo. A, Mouse neutrophils were stimulated with PMA, control IgG, or APS IgG. NETosis was scored by immunofluorescence microscopy. N=4–6 per condition. Mean and SEM are depicted. **p<0.01, ****p<0.0001. B, Representative microscopy from the experiment of panel A. DNA is stained blue and citrullinated histone H3 (Cit-H3) green. NETs are identified as extracellular areas of blue and green overlap. Scale bar=50 microns. C, 48-hour thrombus sections (as in Figure 1) were stained by H&E, Hoechst 33342 (for DNA), and anti-Cit-H3 (for NETs). Staining is represenative of n=4 per group. The dashed yellow circle indicates an area that is viewed at higher magnification in panel E. D, Quantification of Cit-H3-positive surface area for panel C. Mean and SEM are depicted for n=4 per group. *p<0.05. E, Upon higher magnification, neutrophils are detected in regions that stain prominently for Cit-H3. Three represenative neutrophils are marked by yellow arrowheads (although essentially every cell in this field is a neutrophil). Scale bar=50 microns. F, Western blotting of total thrombus protein for citrullinated histone H3 (Cit-H3) and total histone H3 (H3). Quantification is in arbitrary units. For quantification, N=6 per group (for APS: 3 from the APS 1 group and 3 from the APS 5 group). Mean and SEM are depicted. **p<0.01.

Figure 3

Mice treated with APS IgG have higher levels of neutrophil-bound IgG and plasma cell-free DNA at 6 hours. Both “sham” and “stenosis” mice underwent a laparotomy with exposure of the IVC; however, only stenosis mice had the IVC narrowed. “Control” mice were treated with control IgG. “APS” mice were treated with APS 1 IgG. A, Nucleated cells from peripheral blood were subjected to flow cytometry. Represenative staining is shown for Ly6G (a neutrophil marker) and human IgG, after gating for neutrophils by forward- and side-scatter. B, Quantification of the experiment presented in panel A. C, Cell-free DNA was determined in plasma for the indicated groups. For panels B and C, each data point represents an individual mouse. Horizontal lines denote group means. Statistically-significant comparisons are shown; **p<0.01, ***p<0.001.

Figure 4

Mice treated with control and APS IgG do not differ for numbers of circulating cells or markers of vessel wall activation at 6 hours. Both “sham” and “stenosis” mice underwent a laparotomy with exposure of the IVC; however, only stenosis mice had the IVC narrowed. “Control” mice were treated with control IgG. “APS” mice were treated with APS 1 IgG. A and B, Neutrophil and platelet counts were determined in peripheral blood. C and D, Quantitative PCR analysis of vessel walls, obtained distal to the renal veins (in the region of the vessel where a thrombus formed in some mice). For all panels, each data point represents an individual mouse. Horizontal lines denote group means. Statistically-significant comparisons are shown; *p<0.05, **p<0.01.

Figure 5

Thrombi from APS IgG-treated mice contain NETs at 6 hours. All mice presented in this figure underwent the stenosis procedure. A, Thrombus sections were stained by H&E, Hoechst 33342 (DNA), and anti-citrullinated histone H3 (Cit-H3). At 6 hours, the RBC-rich areas make up the majority of the thrombi, while the RBC-poor areas are largely confined to the periphery. Red arrowheads demonstrate Cit-H3 staining, which localizes to the periphery (vessel wall interface) of the APS thrombus at the 6-hour time point. The dashed yellow circle denotes an area considered at higher magnification in panel B. Staining is representative of n=4 thrombi per group. B, Neutrophils are the predominant nucleated cells infiltrating thrombi at 6 hours. Representative examples are indicated with yellow arrowheads (although essentially all cells in this view represent neutrophils). C, Western blotting of total thrombus protein for citrullinated histone H3 (Cit-H3) and total histone H3 (H3). Quantification is in arbitrary units. For quantification, N=6 per group (for APS: 3 from the APS 1 group and 3 from the APS 5 group). Mean and SEM are depicted. *p<0.05.

Figure 6

DNase administration and neutrophil depletion protect against APS IgG-accelerated thrombosis at 6 hours. Neutrophil depletion was with an anti-Ly6G monoclonal antibody, administered 24 hours before the first control/APS IgG injection. DNase was dosed only once, immediately following stenosis (all mice presented in this figure underwent the stenosis procedure). A–C, Thrombus weight for each of the 9 treatment groups at 6 hours. Each data point represents an individual mouse. Horizontal lines denote group means. D, Neutrophil depletion reduces the NET content of APS thrombi. Western blotting of total thrombus protein is for total histone H3 (H3) and citrullinated histone H3 (Cit-H3). Lanes were loaded with equal amounts of total protein. Mice treated with control IgG have low NET content at baseline (as in Figure 5C) and are not depicted here. E, The efficiency of neutrophil depletion did not differ between groups. The dashed line is the mean neutrophil count for the control-stenosis group of Figure 4A. F, The data of panels A–C is here presented as thrombosis frequency, and organized by treatment group. *p<0.05, **p<0.01, ns=not significant.