Neutrophil elastase-deficient mice form neutrophil extracellular traps in an experimental model of deep vein thrombosis

Abstract

ESSENTIALS: Neutrophil elastase (NE) plays a role in extracellular trap formation (NETosis) triggered by microbes. The contribution of NE was evaluated in mouse NETosis models of sterile inflammation and thrombosis. NE is not required for mouse neutrophil NET production in vitro with non-infectious stimuli. NE deficiency had no significant effect on thrombosis in the inferior vena cava stenosis model.

Background: Neutrophil serine proteases have been implicated in coagulation and neutrophil extracellular trap (NET) formation. In human neutrophils, neutrophil elastase (NE) translocates to the nucleus during NETosis and cleaves histones, thus aiding in chromatin decondensation. NE(-/-) mice were shown not to release NETs in response to microbes. However, mouse studies evaluating the role of NE in NET formation in sterile inflammation and thrombosis are lacking.

Objective: We wished to establish if neutrophils from NE(-/-) mice have a defect in NETosis, similar to peptidylarginine deiminase 4 (PAD4(-/-)) mice, and how this might have an impact on venous thrombosis, a model where NETs are produced and are crucial to thrombus development.

Methods: We performed in vitro NET assays using neutrophils from wild-type (WT), NE(-/-), SerpinB1 (SB1)(-/-) and NE(-/-) SB1(-/-) mice. We compared WT and NE(-/-) animals using the inferior vena cava stenosis model of deep vein thrombosis (DVT).

Results: Neutrophil elastase deficiency resulted in a small reduction in ionomycin-induced NET formation in vitro without affecting histone citrullination. However, NET production in response to phorbol 12-myristate 13-acetate or platelet activating factor was normal in neutrophils from two independent NE-deficient mouse lines, and in NE(-/-) SB1(-/-) as compared with SB1(-/-) neutrophils. NE deficiency or inhibition did not prevent NETosis in vivo or DVT outcome.

Conclusions: Neutrophil elastase is not required for NET formation in mice. NE(-/-) mice, which form pathological venous thrombi containing NETs, do not phenocopy PAD4(-/-) mice in in vitro NETosis assays or experimental venous thrombosis. Our study suggests that NET-targeted therapies need to be highly effective to have an impact on DVT.

Keywords: eukocyte elastase; extracellular traps; neutrophils; serine proteases; venous thrombosis.

Figure 1. Neutrophil elastase-deficient neutrophils form fewer NETs in response to calcium ionophore but citrullination is unaffected

A. Neutrophils isolated from WT (gray bars) and NE^−/− [1] blood (white bars) became hypercitrullinated at histone H3 to a similar degree in response to stimulation with ionomycin or PMA for 3.5 h. Neutrophil elastase inhibition (dashed bars) did not impact histone citrullination. n=5. Results are representative of 3 independent experiments. B. NET formation was similar between WT and NE^−/− neutrophils in response to PMA, but was reduced by about 40% in NE^−/− neutrophils in response to ionomycin. Neutrophil elastase inhibition similarly reduced NET formation in both WT and NE^−/− neutrophils. n=5. Representative of 3 independent experiments. C. Representative micrographs of ionomycin-stimulated cells showing H3Cit-positive NETs forming both by WT and NE^−/− neutrophils. White arrows indicate NETs. White arrowheads indicate H3Cit-negative cells. H3Cit, green; DNA, blue. Representative of n=10, scale bar, 20 μm. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 2. NET production by bone marrow neutrophils from two NE-null mouse lines do not differ from matched NE-expressing neutrophils

The NE-deficient mice in A and B are progeny of unrelated independent gene deletions [1, 9]. Bone marrow neutrophils were isolated and stimulated by PMA (left, 50 nM, 4 h) or PAF (right, 10 μm, 30 min) as previously described [7]. NETs were quantified in fluorescence images after Sytox staining. (A) NE^−/− and WT neutrophils on a 129S6 background [9]. (B) NE^−/−, WT, NE^−/−SB1^−/− and SB1^−/− neutrophils on a C57Bl/6 background [1, 7, 10]. The data (% NETs) representing means ± SEM for 6 or more mice/group in 2 or more experiments, were analyzed by (A) unpaired t test or (B) ANOVA. *P < 0.05; NS, not significant.

Figure 3. NE^−/− mice form NET-containing thrombi in the venous stenosis model of DVT

A,B. NE^−/− mice [1] formed thrombi to a similar extent to WT mice after 6 h of IVC stenosis. C. Plasma DNA was slightly decreased after 6 h in NE^−/− mice compared to WT mice. D,E. At 48 h, NE^−/− mice also presented with thrombi to a similar extent to WT mice. F. Both thrombi collected from WT or NE^−/− mice 48 h after IVC stenosis showed an abundance of NETs as seen by H3Cit immunostaining (green) and extensive diffuse extracellular DNA staining (blue). Scale bar, 25 μm. *P < 0.05.

Figure 4. NE inhibition does not impair venous thrombus formation

Mice were injected with 30 mg/kg of sivelestat or vehicle 20 minutes prior to IVC ligation and then euthanized after 6 h to evaluate thrombus formation. A,B. NE inhibitor did not prevent thrombus formation in WT mice. C. Plasma DNA levels were similar between WT mice treated with vehicle or with sivelestat.