Human neutrophil elastase mediates fibrinolysis shutdown through competitive degradation of plasminogen and generation of angiostatin

Abstract

Background: A subset of trauma patients undergo fibrinolysis shutdown rather than pathologic hyperfibrinolysis, contributing to organ failure. The molecular basis for fibrinolysis shutdown in trauma is incompletely understood. Elastase released from primed/activated human neutrophils (HNE) has historically been described as fibrin(ogen)olytic. However, HNE can also degrade plasminogen (PLG) to angiostatin (ANG), retaining the kringle domains but not the proteolytic function, and could thereby compete for generation of active plasmin by tissue plasminogen activator (tPA). We hypothesized that HNE can drive fibrinolysis shutdown rather than fibrinolysis.

Methods: Turbidometry was performed using light scatter (λ = 620 nm) in a purified fibrinogen + PLG system and in healthy citrate plasma clotted with Ca/thrombin ± tPA, ±HNE, and ±ANG to evaluate HNE effects on fibrinolysis, quantified by time to transition midpoint (Tm). ΔTm from control is reported as percent of control ±95% CI. Purified HNE coincubated with PLG or tPA was analyzed by western blot to identify cleavage products. Exogenous HNE was mixed ex vivo with healthy volunteer blood (n = 7) and used in TEG ± tPA to evaluate effects on fibrinolysis.

Results: HNE did not cause measurable fibrinolysis on fibrin clots, clotted plasma, or whole blood as assessed by turbidometry or TEG in the absence of tPA. Upon tPA treatment, all three methods of evaluating fibrinolysis showed delays and decreases in fibrinolysis caused by HNE relative to control: fibrin clot turbidometry ΔTm = 110.7% (CI 105.0-116.5%), clotted citrate plasma (n = 6 healthy volunteers) ΔTm = 126.1% (CI 110.4-141.8%), and whole blood native TEG (n = 7 healthy volunteers) with ΔLY30 = 28% (p = 0.043). Western blot analysis of HNE-PLG co-incubation confirmed that HNE generates angiostatin K1-3, and plasma turbidity assays treated with angiostatin K1-3 delayed fibrinolysis.

Conclusion: HNE degrades PLG and generates angiostatin K1-3, which predominates over HNE cleavage of fibrin(ogen). These findings suggest that neutrophil release of elastase may underlie trauma-induced fibrinolytic shutdown.

Figure 1. Human neutrophil elastase does not cause measurable fibrinolysis in turbidity assays

Panel A: Turbidity assay of purified system containing fibrinogen, thrombin, and -/+ HNE demonstrating no significant differences or fibrinolysis between control and HNE groups. Panel B: Turbidity assay of purified system containing fibrinogen, glu-plasminogen, thrombin, and -/+ HNE demonstrating no significant differences or fibrinolysis between control and HNE groups. Panel C: Turbidity assay of healthy human citrate plasma clotted with 20mM calcium and 5nM thrombin -/+ HNE demonstrating no significant differences or fibrinolysis between control and HNE groups. Results reported as mean ± SEM.

Figure 2. Human neutrophil elastase delays fibrinolysis in tPA-challenged turbidity assays

Panel A and B: Turbidity assay of purified system containing fibrinogen, thrombin, glu-plasminogen, tPA and -/+ HNE demonstrating significant prolongation of fibrinolysis due to HNE compared to the control group as well as increased A_620-Max in the HNE group relative to control. Panel C and D: Turbidity assays of healthy human plasma (n = 6) challenged with tPA and -/+ HNE demonstrating significant prolongation of fibrinolysis due to HNE compared to the control group (representative curve in Panel C) as well as increased A_620-Max in the HNE group relative to control. Results are reported as mean ± SEM on kinetic graphs of A₆₂₀ vs time, and as mean ± 95% confidence intervals (CI) on scatter plots evaluating changes in T_m (ΔT_m) and A_620-Max (ΔA_620-Max).

Figure 3. Human neutrophil elastase causes resistance to fibrinolysis on citrated whole blood native TEG challenged with tPA

Panel A-D: Citrated whole blood native TEG (n=7 healthy volunteers) demonstrating no significant differences in R time, angle, MA, or LY30 between control and HNE groups in the absence of tPA, but a significant reduction in median LY30 in the HNE group upon tPA-challenge compared to controls (no HNE). Results shown as the median and 25^th-75^th percentile range.

Figure 4. Human neutrophil elastase causes fibrinolysis shutdown through plasminogen depletion and generation of the fibrinolysis inhibitor angiostatin

Panel A: Western blot under non-reducing conditions demonstrating that top-normal physiologic levels of PLG (250ug/mL) are rapidly cleaved by HNE at levels similar (or below) those reported in trauma (conc. 80ug/mL, activity 80mU/mL) with formation of angiostatin K1-3 and K1-4 by 30 minutes (lane 2) and near complete degradation of PLG to angiostatin K1-3 by 6 hours (lane 4) at 37°C relative to controls (lanes 1,3). Angiostatin K1-3 is visualized as 3 bands at ∼32kDa, angiostatin K1-4 is visualized as a band at 45kDa, and an intermediate species of angiostatin K1-3 containing the N-terminal 76 amino acids prior to cleavage at 38kDa. Panel B: Turbidity assay using PLG pre-incubated with HNE for 6 hours prior to addition of fibrinogen, thrombin and tPA demonstrating profound delay in fibrinolysis relative to control. Panel C: Non-reducing western blot demonstrating that tPA (250ug/mL) undergoes very little cleavage by HNE (80mU/mL) after 30 minutes (lane 2) or 6 hours (lane 4) at 37°C relative to controls (lanes 1,3). Panel D: Turbidity assay of healthy human plasma + tPA treated with stepwise increases in angiostatin K1-3 and clotted with thrombin + calcium demonstrating significant prolongation of fibrinolysis when angiostatin K1-3 levels are increased beyond reference plasminogen levels.

Figure 5. Mechanistic model for the proposed mechanism of human neutrophil elastase-mediated fibrinolysis shutdown

Elastase released by human neutrophils activated after traumatic injury cleaves PLG to angiostatin K1-3, depleting the pool of PLG substrate that is normally cleaved by tPA to generate plasmin for fibrinolysis (Panel A). Angiostatin K1-3 generated in this fashion then competes with the remaining PLG for lysine binding sites on fibrin clots (Panel B), where PLG binding to lysine residues on fibrin is a requisite step for tPA cleavage of PLG to plasmin. By depleting the pool of PLG and generating angiostatin K1-3, human neutrophil elastase released after traumatic injury may contribute to fibrinolysis shutdown.