INTACT TOLL-LIKE RECEPTOR 9 SIGNALING IN NEUTROPHILS MODULATES NORMAL THROMBOGENESIS IN MICE

Abstract

Background

Deletion of Toll-like receptor 9 (Tlr9) signaling, which is important for sterile inflammatory processes, results in impaired venous thrombosis (VT) resolution in mice. The purpose of this study was to determine if deletion of Tlr9 affected sterile necrosis, apoptosis, and neutrophil extracellular trap (NET) production in VT.

Methods

Stasis and non-stasis murine models of VT were used in wild type (WT and Tlr9^−/− mice, with assessment of VT size, and determination of neutrophil extracellular traps (NETs), necrosis and apoptosis markers. Anti-PMN and anti-platelet antibody strategies were used to determine the cellular roles and their roles in WT and Tlr9^−/− mice.

Results

At 2d, stasis thrombi in Tlr9^−/− mice were 62% larger (n = 6–10) with 1.4 fold increased uric acid levels, 1.7 fold more apoptotic cells, 2 fold increased citrullinated histones (cit-H₃), 2 fold increased peptidylarginine deiminase – 4 and 1.5 fold increased elastase, with a 2.4 fold reduction in tissue factor pathway inhibitor (TFPI) as compared with WT (all n = 4–7; P < .05). In contrast, non-stasis VT sizes were not significantly different in Tlr9^−/− mice (n = 4–6), and did not have elevated necrosis or NET markers. Stasis VT size was not reduced at the 2d time-point in WT or TLR9−/− mice that received treatment with DNAse-I, or in PAD4−/− mice, which are incapable of forming NETs. Stasis VT size was reduced 18% inTlr9^−/− mice undergoing PMN depletion (n = 8–10), and was associated with 29 fold decreased cit H3, 1.3 fold decreased elastase, and 1.5 fold increased TFPI (all n = 6; P < .05). Lastly, platelet depletion (>90% reduction) did not significantly reduce stasis VT inTlr9^−/− mice.

Conclusions

These data suggest the thrombogenic model impacts Tlr9 thrombogenic mechanisms, and that functional Tlr9 signaling in PMN, but not platelets or NETs, is an important mechanism in early stasis experimental venous thrombogenesis.

Introduction

Deep vein thrombosis (DVT) is a major clinical problem with an estimated incidence of more than 900,000 cases per year in the U.S,¹ yet the mechanisms of venous thrombosis (VT) are only now being better defined. Using this knowledge, new therapies for DVT that do not increase the risk of bleeding could fundamentally improve treatment.

Venous thrombogenesis and resolution is an inflammatory process that resembles sterile wound healing.² Dying cells in venous thrombi (VT) may release endogenous danger signals that trigger the immune response promoting further thrombosis. Endogenous “danger signals”, such as nucleic acids,³ uric acid,⁴ and intracellular peptides,⁵ are known to be released from damaged and dying cells and to trigger an innate immune response. Dying leukocytes such as PMNs release neutrophil extracellular traps (NETs), dependent on peptidyl arginine deaminase-4 (PAD-4),⁶ which may occur in sterile inflammatory settings, including venous thrombosis.^{7, 8} Moreover, elastase and other proteases may degrade tissue factor pathway inhibitor (TFPI), a key anti-thrombotic factor.^9–11 Many of these extracellular signals are recognized by pattern recognition receptors, including Toll-like receptors, present on leukocytes including neutrophils (PMN) and monocytes/macrophages.¹²

Toll-like receptor (Tlr) 9 recognizes both foreign and host DNA in sterile inflammation, evidenced by critical roles in liver injury, pulmonary injury, and atherosclerosis.^13–15 Tlr9 activation leads to release of interleukin (IL)-1, a critical mediator of sterile inflammation. Previous work from our group has shown that Tlr9^−/− mice have impaired resolution of experimental VT and altered markers of sterile inflammation.¹⁶ Furthermore, administering an exogenous Tlr9 ligand increased VT resolution in WT mice. Clinically, patients with autoimmune diseases such as systemic lupus erythematous (SLE) have an increased risk of DVT and also have aberrant processing of self-antigens via Tlr9, including NETs.^{17, 18}

The purpose of this study was to specifically examine the early inflammatory thrombogenic mechanism of sterile necrosis, apoptosis, and NETs in relation to Tlr9 signaling in experimental VT.

Materials and Methods

Mice

Male Balb/c WT mice were purchased from Harlan Laboratories (Indianapolis, IN). Male Tlr9^−/− mice on Balb/c background were purchased from S. Akira (Coley Pharmaceutical, Wellesley, MA). Genetic deletion was confirmed using real time PCR on tail samples by Transnet (Cordova, TN). Mice unable to produce NETs, utilizing (PAD-4^−/−; kindly donated by Dr. Y. Wang), on 129×C57BL/6 × Balb/c mixed background or WT mixed background littermates were used for selected experiments.⁶ All mice that underwent survival surgeries were used between 7 and 12 weeks of age (20–30 g). Mortality for mice was < 5%, and 95% of mice form thrombus. Mice underwent general anesthesia with isofluorane/O₂ for all surgical procedures and all animal experiments were approved by the University of Michigan Committee for the Care and Use of Animals.

Stasis model of VT

WT, Tlr9^−/−, and PAD4^−/− mice underwent surgical ligation of the inferior vena cava (IVC) and visible contributing vessels below the renal veins, producing full stasis and subsequent VT formation. This model is well-characterized and consistently (>97%) produces a VT.^{16, 19–22} IVC and thrombus were harvested at 2d after induction of stasis for tissue analysis. Thrombus and IVC were left intact for immunohistochemistry or carefully separated for the molecular and immunologic assays.

Non-Stasis Model of VT

An endothelial injury model of VT²³ was used to establish thrombus in the setting of continued blood flow. WT and Tlr9^−/− mice underwent intraluminal electrolytic injury to their IVC. A 25 gauge needle was inserted into the IVC and put in contact with the anterior IVC wall between the renal veins and iliac bifurcation. A current of 25 mAmp was applied for 15 minutes, consistently producing non-occlusive, VT. Vein wall and thrombi were harvested and separated at 2d.

DNAse administration

DNAse I (Pulmozyme, Genentech, San Francisco, CA) was given to each mouse both intravenously (10 ng) just prior to VT and intraperitoneally (50 ng) shortly after VT and then every 12 hours as described.²⁴ At 2d, mice were euthanized and thrombus was quantified for size.

Histological Analysis/Immunohistochemical Staining

Staining of Ly6G (eBiosciences, San Diego, CA) and TUNEL (Promega, Madison, WI) + cells were done as described, on paraffin embedded tissue sections (5 µm).^{20, 21, 25} After processing, the slides were counterstained with hematoxylin or DAPI and cover-slipped. TUNEL and Ly6G, + cells in thrombi were counted in 5 high powered fields (1000×) using Image J software.

To visualize NETs in thrombus sections, we stained for citrullinated histone H₃ (cit-H₃;1:500, Abcam) and Ly6G (1:500, BD Pharmingen, San Jose, CA), utilizing FITC and Texas Red fluorescently tagged secondary antibodies (Invitrogen, Grand Island, NY). Slides were cover slipped with ProLong Gold with DAPI mounting medium (Invitrogen). Pictures were taken using a Nikon Eclipse E400 microscope at 1000× equipped with a Nikon Digital Sight DS-U3 camera using the DAPI (nuclei), FITC (Ly6G) and Texas Red (Cit-H₃) channels.

Neutrophil and platelet depletion

WT and Tlr9^−/− were given 0.5 ml anti-PMN serum (Accurate Chemical, Westbury, NY, AIAD31140) or 0.5 ml Rabbit non-specific serum (Accurate Chemical, JNZ000001) IP once daily for two days prior to IVC ligation and then again at 1d post IVC ligation.^{21, 26} In separate experiments, WT and Tlr9^−/− were given 50 ul anti-Thrombocyte serum (Accurate Chemical, AIA31440) or 50 ul normal rabbit serum (Accurate) IP, approximately 1 hour prior to IVC ligation as described.²⁷

Antigen and enzyme activity analysis

Thrombus was homogenized and sonicated in 1 mL of protease inhibitor buffer (Roche, Basel, Switzerland). IL-1α, IL-1β, IL-18 antigen levels were quantified using Bio-Plex (Biorad, Hercules, CA) according to manufacturer’s instructions. Uric acid concentration was determined using the Quantichrom Uric Acid Assay (Bioassay Systems, Hayward, CA). All measurements were normalized to total protein determined by Pierce BCA assay (ThermoScientific, Waltham, MA).

Antigen Markers by Western Immunoblotting

Stasis and non-stasis VT from 2d WT and Tlr9^−/− were processed for antigen quantification including cit-H₃ (1:500; Abcam, Cambridge, MA), Elastase (1:1000, Abcam), PAD4 (1:500, Abcam), TFPI (2 ug/mL Novus Biologicals, Littleton, CO), and B-Actin (1/50,000; Santa Cruz, Dallas, TX) as described.^{25, 28} Primary antibodies were diluted in TBST, added to the membrane and incubated at 4°C overnight while gently shaking. For normalization of proteins on the western blots, the membranes were stripped and probed with anti-B-actin antibodies conjugated with HRP (Santa Cruz). The membranes were developed with the West-Pico ECL kit (Pierce, Rockford, IL). Densitometry was performed using Image J program.

Statistical analysis

Data were analyzed and graphed using the Prism 6.0 software program for Windows (GraphPad Software, San Diego, CA). Statistical significance was calculated using two-tailed t test and defined as P<0.05. Data are presented as mean ± SD.

Results

Tlr9^−/− thrombi contain more necrotic and apoptotic cell markers in a model dependent fashion

In experimental models of VT, thrombus size (measured by weight per length) can be used to reflect the sum of thrombogenesis and thrombolysis over time, with thrombogenesis considered as ≤ 2d after the induction of thrombosis.^{8, 16, 20, 21, 29} We have previously shown that Tlr9^−/− mice have impaired VT resolution (i.e. larger thrombi) up to 8d in the stasis model.¹⁶ In this experiment, we again found that Tlr9^−/− mice had larger stasis thrombi at 2d and now the early time point of 6 hours (Fig. 1a). We used the non-stasis model in Tlr9^−/− and WT mice to determine if a similar increase in VT was observed as with the stasis model. In non-stasis VT, there was no significant difference in VT size between Tlr9^−/− and WT at 2d (Fig. 1a). To determine if any baseline hemostatic differences were present, we tested the propensity of blood samples from Tlr9^−/− and WT mice to clot by tail vein bleeding time. We found no difference in bleeding time between strains (141 ± 40 vs. 141 ± 31 s; N = 4 – 6). No differences between Tlr9^−/− and WT in circulating PMNs, monocytes, or platelets counts were found at baseline (not shown).

Figure 1.

A). Larger VT are present in stasis VT at 6 hours and 2d in Tlr9^−/− as compared with WT (n = 6 – 10) but not in non-stasis (EIM = electrolytic induction model) Tlr9^−/− as compared with (n = 4 – 6); B). Thrombus IL-1α was reduced in Tlr9^−/− as compared with WT controls (n = 5); C). Thrombus uric acid concentration was increased in Tlr9^−/− mice as compared with WT controls (n = 6 – 7); D). Thrombus TUNEL+ cell counts in 5 high power fields is shown. At 2d, Tlr9^−/− thrombi contained an increased number of TUNEL+ cells compared to WT controls (n = 4). Histology representative photos with TUNEL+ cells fluorescing green and DAPI fluorescing blue (100×, scale = 100 µm) in WT and Tlr9^−/− thrombus (arrows denote dual staining cells); E). Activated PMN (Ly6G+) were increased in Tlr9^−/− thrombi as compared with WT (n = 5); F). Two histology image examples show greater Ly6G+ cells labeled in brown (arrows) in Tlr9^−/− and WT (100 ×; scale = 100 µm). Data are mean +/− SD. *p = < .05.

The inflammatory milieu was altered in Tlr9^−/− mice as well, with thrombus IL-1α, but not IL-1β (not shown) or IL-18 (not shown), reduced in stasis Tlr9^−/− VT as compared with WT (Figure 1b). Uric acid in tissues induces a strong inflammatory response,^{30, 31} and was increased in the Tlr9^−/− VT at 2d as compared to WT controls (Figure 1c).

As a marker of apoptosis, TUNEL staining was greater in 2d Tlr9^−/− stasis VT as compared with WT (Figure 1d). The majority of TUNEL positive cells had PMN morphology, with Tlr9^−/− VT containing about 60% more TUNEL+ cells than WT controls. As activated PMN may be more prothrombotic,⁸ we assessed for PMN activation state by Ly6G+ cellular expression. We found a greater number of activated PMNs present in Tlr9^−/− as compared to WT (Figure 1e, f).

No differences in thrombus uric acid between WT and Tlr9^−/− was found in non-stasis VT (.007 +/− .0001 vs. .0048 +/− .0002 mcg/mg protein; N = 7, p = .37). We also found that Tlr9^−/− non-stasis VT had no difference in PMN counts as compared with WT at 2d (202 +/− 58 vs. 150 +/− 51 cells/5 hpf; N = 3 – 5; P = .54), contrasting with the previously documented elevated PMN in Tlr9^−/− stasis VT.¹⁶

Stasis VT in Tlr9−/− mice have increased thrombus NET markers and decreased TFPI

To determine whether PMN mediated thrombo-inflammatory processes were altered in Tlr9^−/− mice, we assessed thrombus levels of NET markers cit-H₃, PAD-4,³² elastase^{24, 32} and the coagulation inhibitor TFPI in stasis VT.^{10, 11} We found in Tlr9^−/− mice, as compared with WT, thrombus cit-H₃, PAD4, and elastase were significantly increased, and TFPI was reduced (Figure 2a – d). We also evaluated a very early time point at 6 hours post thrombosis. Here, we found that Tlr9^−/− thrombi with increased elastase and decreased TFPI, but no difference in cit-H₃ as compared with WT (Supplemental Figure). Consistent with no significant difference in VT size in the non-stasis model, thrombus cit-H₃ (OD = 1.56 +/−.29 vs 1.41 +/− .18, ratio to B-actin, n = 4–5, P = .67) and TFPI (OD = .48 +/− .09 vs .58 +/− .04 ratio to B-actin, n = 5–6, P = .37), were not different, although elastase antigen was elevated in the non-stasis model between WT as compared to Tlr9^−/− (OD = .26 +/− .07 vs .87 +/− .19, ratio to B-actin, n = 4, P = .02).

Figure 2.

Tlr9^−/− VT have increased NET markers by Western immunoblotting and less anti-thrombogenic inhibitor, as compared with WT controls: A). Citrullinated histones (Cit-H₃); B). PAD-4; C). Elastase; D). TFPI; (n = 4 – 6). Data are mean +/− SD. * = p <.05.

Given that we found increased NET markers in Tlr9^−/− stasis thrombi compared WT, we further examined the role of NETs in stasis thrombogenesis. We found greater NET positive cells (Ly6G+/cit-H3+) in the VT of Tlr9−/− mice (Fig 3a, b). As others have shown that early venous thrombogenesis in mice is NET dependent,²⁴ WT mice were treated with DNAse I, immediately after inducing stasis VT. We found that the stasis VT size in mice treated with DNAse I was similar to those of controls at 2 d (Fig 3c). Moreover, there was no difference when Tlr9−/− mice were treated with DNAse I, with a slight trend to actually larger VT. To more fully establish that NETs were not playing a role in stasis VT, we used PAD4^−/− mice, documented to not be able to produce NETs,⁶ and associated with smaller VT in a stenosisVT model.³² Consistent with the lack of effect with DNAse-I treated mice, no difference in stasis VT size was found in PAD-4^−/− mice as compared with controls (Figure 3d).

Figure 3.

A.) The number of NET + PMNs were increased in Tlr9^−/− as compared with WT (n = 4 – 6); B). NET immunofluorescence histograph examples show co-staining with PMN of DAPI/cit-H₃/Ly6G (arrows) at 1000×; C). DNAse I (Tx = treatment) did not affect stasis VT in WT or Tlr9^−/− at 2d, (n = 6 – 8); D). Thrombus size was not altered in stasis VT in PAD4^−/− mice as compared with WT (n = 6 – 8). Data are mean +/− SD. * = p <.05.

Functional leukocyte contributions in Tlr9^−/− and WT stasis VT

Given that Tlr9^−/− had larger stasis VT with increased thrombus necrosis and apoptosis, but was not directly NET dependent, this suggested that other PMN activities may be involved. The timing of PMN predominance in VT is between 1 and 3 days.^{8, 21} We used an anti-PMN strategy,^{21, 26} confirming a significant reduction in circulating PMN prior to thrombus induction (>70% from baseline). PMN depletion decreased stasis VT size at 2d in Tlr9^−/− mice to near control Tlr9^+/+ size. PMN depletion did not affect 2d stasis VT size in WT controls, and is consistent with our prior work (Figure 4a).²¹ We found significantly reduced cit-H₃, elastase, and reciprocally increased TFPI in Tlr9^−/− with PMN depletion, as compared with Tlr9^−/− serum controls (Figure 4b – d). No change in uric acid was found in Tlr9^−/− as compared with WT with PMN depletion (not shown). In histological section analysis, we also found intra-thrombus PMN were significantly reduced and confirmed the systemic depletion (Figure 4e).

Figure 4.

Cellular PMN processes in thrombogenesis: A). PMN antibody depletion (Tx) significantly reduced VT size in Tlr9^−/− as compared to IgG control (n = 5). Thrombus elastase; (B) and cit-H₃ (C) were reduced in Tlr9^−/− with PMN depletion, while TFPI was increased (D) (n = 4 – 6); Intra-thrombus PMN were reduced in Tlr9^−/− with Tx; (E) (n = 3 – 5). Example of H and E stained thrombus sections, at 200× (arrows indicated PMNs). Data are mean +/− SD. *p < .05.

As platelet Tlr9 engagement may be important for early thrombosis,³³ we used an anti-platelet serum (>90% circulating platelet depletion from baseline) strategy.²⁷ However, as compared with control serum, we found no difference in VT size between WT (control = 18+/− 1 vs. anti-platelet = 15+/− 1 mg/cm; N = 4 – 5, P = .14) and Tlr9^−/− (control = 24.8+/− .2 vs. anti-platelet = 21+/− 1 mg/cm; N = 5 – 7, P = .12) at 2d in thrombocytopenic mice.

Discussion

Venous thrombogenesis is a sterile inflammatory process, mediated by leukocytes, chemokines, and coagulation factors.² We previously reported that loss of Tlr9 signaling in mice results in impaired VT resolution,¹⁶ but the inflammatory cell necrosis products were not fully delineated, the PMN cellular responses were not defined, and only the stasis model was examined. The current report shows: 1). The Tlr9 role in VT is dependent on the model of VT; 2). Lack of Tlr9 is associated with increased stasis VT necrosis, apoptosis, NET markers, and decreased TFPI; 3). Early stasis VT is not NET dependent; 4). Dysfunctional Tlr9^−/− PMNs, but not platelets, contribute to greater thrombogenesis.

A novel finding in this study is that Tlr9 signaling only affects thrombogenesis in the stasis model of VT, but not in the non-stasis model. The main physical difference between the non-stasis and stasis models of VT is that the non-stasis induced thrombus is continuously exposed to peri-thrombus blood flow and the stasis induced thrombus is not; although recanalization of the thrombus does occur over time.²⁹ These two models are replicative of what is believed to occur in humans, with areas of stasis and non-stasis,³⁴ and is why we typically use both models to more closely model human DVT.²⁵ Consistent with a lack of difference in VT size in the non-stasis Tlr9^−/− and WT mice were findings of no increase in concentrations of uric acid and cit-H₃, and no decrease in TFPI. These data suggest that peri-thrombus blood flow abrogates Tlr9 mediated cellular thrombogenic activities, possibly by removing Tlr9 ligands that accumulate in a stasis milieu.

The Tlr9^−/− stasis thrombi contained more uric acid, reflective of necrosis, and apoptotic cells, as compared with WT. Uric acid is formed in settings of cell lysis,³⁰ and its presence in thrombi strongly suggests necrotic cell death. Altered clearance of uric acid may promote thrombosis in our model. However, this is only an association, as we were unable to alter the thrombus uric acid concentration in 2d stasis thrombus despite very high doses of uricase (pegloticase, unpublished data). Prior studies have confirmed apoptosis occurs in late stasis VT.³⁵ Greater apoptosis was observed in Tlr9^−/− thrombi as compared with WT, with more TUNEL+ cells. This could result from altered apoptotic signaling kinetics in the Tlr9^−/− thrombus milieu, or decreased clearance of apoptotic cells due to impaired sterile processing, which we believe is more likely. Decreased IL-1 is consistent with TLR9 genetic deletion as this is upstream of IL-1 processing.³⁶ This mechanism may account in part for the observed increased apoptosis in the Tlr9^−/− VT, as this cytokine has apoptotic inhibitory activity in PMNs.³⁷ Together, in stasis VT, Tlr9 signaling is critical for clearance of necrosis and apoptosis.

NETs have been shown to induce a sterile inflammatory response through Tlr9 signaling, as shown in models of systemic lupus erythematosus.^{17, 18} NETs are prothrombotic,¹¹ are present in experimental VT in baboons and mice,^{7, 24} and are another cell death related process. We found that the Tlr9^−/− thrombi contained increased total extracellular cit-H₃ and PAD-4, as well as an increased number of NET positive PMNs. We were thus surprised to find that DNAse-I did not reduce 2d stasis VT size, in contrast to others who have shown DNAse-I administration reduces NETs and VT size in a stenosis (flow) model.²⁴ This also contrasts with Martinod, et al, who found non-stasis VT dependent on PAD-4 activity.³² The lack of DNAse I effect in stasis VT may be due to the solid consistency of thrombi, which may impede the exogenous enzymes from coming into contact with the NETs, or that breakdown of NETs in a static environment promotes greater histone release, which are directly prothrombotic.³⁸ Consistently, recent data suggests that histones may directly promote thrombogenesis, independent from the NET complex.³⁹ However, the direct role of histones is not directly answered here, and is not entirely consistent given that the PAD4−/−, unable to produce cit-H3, did not have larger VT. We postulate that the Tlr9^−/− mice are unable to process or breakdown the NETs in the normal manner,¹⁷ and this may be one mechanism whereby NET markers (e.g. cit-H3) may accumulate in the thrombus.

Prior experimental data suggested no differences in baseline coagulation parameters in Tlr9^−/− mice as compared with WT,¹⁶ and in this report, no differences in bleeding time were found. Given that NETs are unlikely mechanistically important in stasis VT, the fact that Tlr9−/− mice had increased thrombogenesis suggests another mechanism may be playing a role. One plausible pathway is elastase mediated breakdown of TFPI, as shown in arterial thrombosis.¹¹ Our data suggest that the Tlr9 deletion affects PMN function such that increased elastase is released, which may degrade TFPI, and promote thrombosis. As TFPI affects the Factor VII-tissue factor and thrombin production, this may also explain why we previously found elevated thrombin-antithrombin complexes in Tlr9−/− VT.¹⁶ These data suggest that the Tlr9^−/− PMNs are dysfunctional and directly contribute to the exaggerated thrombosis in different ways.

Since it was less likely that NETs were playing a mechanistic role in stasis thrombogenesis, we examined the direct role of PMN in the Tlr9^−/− mice. PMN are present in large numbers in early VT,²¹ and express Tlr9. We reduced circulating and intra-thrombus PMN using an antibody depletion strategy,^{21, 26} and found that depletion of Tlr9^−/− PMN, but not WT PMN, reversed the increased thrombogenic response. Specifically, depleting Tlr9^−/− PMNs was associated with normalization of VT sizes comparable to IgG controls, and consistently, a reduction in Cit-H₃, elastase, and an increase in TFPI. Given that the PMN were depleted prior to the thrombotic stimulus, this suggests that the PMN are playing an active role in the venous thrombogenesis rather than a bystander ‘trapped’ in the thrombus. Interestingly, PMN depletion in WT stasis model did not reduce VT size at 2d; consistent with prior work in our lab,²¹ but not others who used a stenosis VT model.⁸ Thus, a threshold PMN function in VT likely exists. Although platelets play a role in stenosis derived VT,⁴⁰ we did not find that platelet depletion in Tlr9^−/− or WT mice affected stasis VT size. Despite documented Tlr9 receptors on platelets and aggregatory activity with certain ligands that induce reactive oxygen species,³³ no effect was found with platelet depletion, suggesting that the platelet Tlr9 role is not significant as compared with the PMN’s role.

Conclusion

This study demonstrates a link between endogenous danger signals and Tlr9 signaling in experimental stasis venous thrombogenesis. These observations may also be relevant for the hypercoagulable states that occur with SLE and markedly increase those patients’ risk for DVT. Prevention of DVT by immunomodulation may be a promising approach to decrease bleeding risk of anticoagulants, as clinical trials of therapies that act by manipulation of Tlr signaling are currently underway in a variety of other diseases.⁴¹

Clinical Relevance.

Deep vein thrombosis is a sterile inflammatory process. The early neutrophil medicated events that drive thrombosis are beginning to be better clarified. This study using mouse models of venous thrombosis shows the importance of neutrophils in clearance of sterile inflammatory mediators; in particular neutrophil extracellular traps and uric acid, and in modulating apoptosis. These experimental may inform therapies that may be able to prevent DVT formation or accelerate its resolution without the risks of bleeding.