Neutrophil extracellular traps and inflammasomes cooperatively promote venous thrombosis in mice

Abstract

Deep vein thrombosis (DVT) is linked to local inflammation. A role for both neutrophil extracellular traps (NETs) and the assembly of inflammasomes (leading to caspase-1-dependent interleukin-1β activation) in the development of DVT was recently suggested. However, no link between these 2 processes in the setting of thrombosis has been investigated. Here, we demonstrate that stimulation of neutrophils induced simultaneous formation of NETs and active caspase-1. Caspase-1 was largely associated with NETs, suggesting that secreted active caspase-1 requires NETs as an adhesive surface. NETs and their components, histones, promoted robust caspase-1 activation in platelets with the strongest effect exerted by histones 3/4. Murine DVT thrombi contained active caspase-1, which peaked at 6 hours when compared with 48-hour thrombi. Platelets constituted more than one-half of cells containing active caspase-1 in dissociated thrombi. Using intravital microscopy, we identified colocalized NETs and caspase-1 as well as platelet recruitment at the site of thrombosis. Pharmacological inhibition of caspase-1 strongly reduced DVT in mice, and thrombi that still formed contained no citrullinated histone 3, a marker of NETs. Taken together, these data demonstrate a cross-talk between NETs and inflammasomes both in vitro and in the DVT setting. This may be an important mechanism supporting thrombosis in veins.

Figure 1.

Active caspase-1 is present in both thrombi and blood cells in murine DVT and can be induced by histones and NETotic neutrophils in human platelets. (A) FLICA incorporation (active caspase-1) in platelets (from thrombi and blood) and leukocytes 48 hours after IVC stenosis. (B) Quantification of FLICA-positive platelets and leukocytes in both thrombi and blood (n = 5, mean ± standard deviation). (C) Immunofluorescence staining of thrombi sections for FLICA (green), SYTOX Orange (red), and CD41 (blue). Arrows indicate areas of colocalization of FLICA and SYTOX Orange. Scale bars: left = 100 μm, middle = 50 μm, right = 10 μm. (D) Total and cleaved (activated) caspase-1 content in thrombus lysates at either 48 or 6 hours after IVC stenosis (representative gel, n = 4-5). (E) MFI for FLICA incorporation (caspase-1 activation) by washed human platelets stimulated with ATP, nigericin, PAF, and calf histones, at indicated concentrations, or left untreated (n = 3, mean ± standard deviation). One-way analysis of variance, Dunnett’s multiple comparison test, *P ≤ .05. (F) Representative plot of FLICA incorporation by human platelets treated with individual human histones and evaluated by flow cytometry. (G) Immunofluorescent staining of histone-stimulated platelets with FLICA (green) and CD41 (red). Scale bar = 10 µm, n = 3. (H) MFI quantification for FLICA from autologous platelets seeded on NETotic neutrophils after treatment with PMA for 2 hours (n = 3 biological replicates, data points indicate technical replicates). One-way analysis of variance, Tukey’s multiple comparison test, *P ≤ .05.

Figure 2.

NETs and active caspase-1 colocalize both in vitro and in vivo, and inhibition of caspase-1 reduces DVT. (A) Immunofluorescent staining for activated caspase-1 on NETs formed by human neutrophils stimulated with either PMA or glucose oxidase (FLICA, green; ASC, red; myeloperoxidase, blue). Scale bar = 15 μm. Quantification of the percentage of active caspase-1–containing cells that simultaneously release NETs (B) and the percentage of NETotic cells containing active caspase-1 (C) (n = 3). (D) Z-projection of the murine IVC wall 2 hours after stenosis obtained by intravital microscopy (FLICA/active caspase-1, green; SYTOX Orange, red; CD41, blue). Asterisk depicts the site of stenosis; arrow shows the direction of blood flow. Scale bar = 100 μm, n = 3. (E) Thrombus weight and length, and thrombosis incidence in ac-yvad-cmk (n = 12) and vehicle control-treated mice (n = 12). **P ≤ .01; weight and length, Mann-Whitney U test; incidence, Fisher’s exact test. (F) Citrullinated histone 3 (green) and DNA (blue) immunofluorescence staining of control thrombi and of thrombi developed in mice treated with ac-yvad-cmk. Scale bar = 25 μm, n = 4.