Regulation of macrophage fibrinolysis during venous thrombus resolution

Abstract

Background: Venous thromboembolism (VTE), which includes pulmonary embolism (PE) and deep vein thrombosis (DVT), is a serious cardiovascular disease with significant mortality and morbidity. Clinically, patients with faster resolution of a venous thrombi have improved prognosis. Urokinase-plasminogen activator (uPA), produced by macrophages, is a key mediator of fibrinolysis required for resolving venous thrombi and restoring vascular integrity. The major macrophage protein, plasminogen activator inhibitor type-2 (PAI-2), was originally identified as an inhibitor of uPA and is implicated in the modulation of pathways affecting fibrinolytic uPA activity, however its direct role in blocking uPA-mediated clot lysis is not known.

Objective: To determine the contribution of macrophage PAI-2 in inhibiting uPA-mediated fibrinolysis during resolution of DVT.

Methods: Using a murine model of venous thrombosis and resolution, we determined histological changes and molecular features of fibrin degradation in venous thrombi from WT mice and mice genetically deficient in PAI-2 and PAI-1, and determined the fibrinolytic activities of macrophages from these genotypes ex vivo.

Results: Acceleration of venous thrombus resolution by PAI-2^-/- mice increases fibrin degradation in venous thrombi showing a pattern similar to genetic deficiency of PAI-1, the major attenuator of fibrinolysis. PAI-2 deficiency was not associated with increased macrophage infiltration into thrombi or changes in macrophage PAI-1 expression. uPA-initiated fibrinolysis by macrophages in vitro could be accelerated by PAI-1 deficiency, but not PAI-2 deficiency.

Conclusion: PAI-2 has an alternate anti-fibrinolytic activity that is macrophage uPA independent, where PAI-1 is the dominant uPA inhibitor during DVT resolution.

Keywords: Fibrin; Fibrinolysis; Macrophage; PAI-1; PAI-2; Plasminogen activator inhibitor; Venous thrombosis.

Fig. 1.. PAI-2 deficiency impairs thrombus resolution.

(A) Thrombus weight and (B) thrombus weight normalized to body weight at day 12 after IVC ligation of PAI-2^−/− mice compared to WT and PAI-1^−/− mice. All values represent the mean ± standard error of the mean (SEM) (n = 7 per group) **P < 0.005, ***P < 0.0005, ns = not significant.

Fig. 2.. Histological analysis of WT and PAI-2^−/− venous thrombi over the course of venous thrombus resolution.

(A) Representative images of thrombus sections stained with Hematoxylin and Eosin (H&E), depicting nucleated cells in purplish-blue and thrombotic structural elements such as fibrin in pink, and red blood cells in red, (B) Martius Scarlett Blue (MSB), illustrating fibrin in red (arrows), RBCs in yellow and collagen fibers in blue, (C) Picrosirius Red staining for collagen content in red (arrows). Scale bar = 500μm. Quantitation of (D) fibrin (red) and (E) collagen content (red) performed on 10x magnification images using the Image J color deconvolution plug-in over the time course of venous thrombus resolution. All values represent the mean ± SEM. WT (Day 2, n=4; Day 4, n=3; Day 7, n=3; Day 12, n=3) and PAI-2^−/− (Day 2, n=3; Day 4, n=4; Day 7, n=4; Day 12 n=4). All values represent the mean ± SEM, *P < 0.05, ns = not significant.

Fig. 3.. PAI-2 deficiency accelerates the appearance of fibrin degradation products (FDPs) in isolated venous thrombi.

(A) Immunoblot analysis of intrathrombotic fibrinolytic products present in venous thrombus samples from WT and PAI-2^−/− mice using anti-fibrin(ogen) polyclonal antibody under non-reducing (upper panel) and reducing (lower panel) conditions. Data is representative of 3 independent experiments using 3 mice of each genotype. (B) Quantitation of intrathrombotic FDPs by densitometry. Data is the average of 3 mice per genotype per time point normalized to β-actin. All values represent the mean ± SEM. *P < 0.05, ns = not significant. (C) Immunoblot analysis of intrathrombotic FDPs at day 4 in venous thrombus samples from 3 WT, PAI-2^−/− and PAI-1^−/− mice using anti-fibrin(ogen) polyclonal antibody under reducing conditions. (D) Quantitation of intrathrombotic FDPs by densitometry in 3 venous thrombus samples from WT, PAI-2^−/− and PAI-1^−/− mice at day 4. Data is normalized to β-actin. All values represent the mean ± SEM, *P < 0.05, ns = not significant.

Fig. 4.. Absence of PAI-2 in macrophages does not enhance uPA-dependent fibrinolysis.

(A) Cell-based fibrinolysis assay using nonactivated (left panel) and LPS activated (right panel) WT and PAI-1^−/− macrophages. (B) Quantitation of fibrinolysis activity in (A) at endpoint. (C) Cell-based fibrinolysis assay using LPS-activated PAI-2^−/−, PAI-1^−/− and WT macrophages in the absence and presence of the uPA inhibitor BC11 (10nM). (D) Quantitation of fibrinolysis activity in (C) at endpoint. (E) Cell-based fibrinolysis assay using LPS-activated PAI-2^−/−, PAI-1^−/− and WT macrophages in the absence and presence of uPA inhibitor amiloride (20nM). (F) Quantitation of fibrinolysis activity in (E) at endpoint. All graphs show percent (%) of maximal turbidity, averaged from macrophages isolated from 3 individual mice of each genotype performed in quadruplicate and normalized to the maximal clot formed. Values represent mean ± SEM. *P <0.05, **P < 0.005, ***P < 0.0005, ****P < 0.0001, ns = not significant.

Fig. 5.. Macrophage infiltration into venous thrombi is unaffected by PAI-2 deficiency.

(A) Immunohistochemical analysis of intrathrombotic macrophage accumulation using anti-CD68 antibody in venous thrombus samples from WT and PAI-2^−/− mice. Scale bar = 100μm. (B) Quantification of intrathrombotic CD68 positive macrophages determined from five randomly selected high-power fields (h.p.f.) at 20x magnification using the Image J color deconvolution plug-in. Values represent mean ± SEM. (Day 4, n=4 WT and 4 PAI-2^−/−; Day 7, n=4 WT and 3 PAI-2^−/−; Day 12, n=4 WT and 3 PAI-2^−/−). ns = not significant. (C) Immunoblot analysis of intrathrombotic macrophage content in venous thrombus clots from WT, PAI-2^−/− and PAI-1^−/− mice using anti-Mac-2/Galectin-3 antibody. (D) Quantitation of intrathrombotic Mac-2 expression. Graph shows densitometric analysis of Mac-2 expression after normalizing to β-actin. All values represent the mean ± SEM. ns = not significant.

Figure 6.. Macrophage PAI-2 deficiency does not affect PAI-1 expression.

(A) PAI-1 protein levels in conditioned media from untreated WT and PAI-2^−/− macrophages at 24 hr determined by ELISA. Values represent mean ± SEM. (n = 3 WT and 3 PAI-2^−/−). ns = not significant. (B) Immunoblot analysis of PAI-2 expression in untreated WT, PAI-1^−/− and PAI-2^−/− macrophages. (C) PAI-1 protein levels in conditioned media from WT and PAI-2^−/− macrophages treated for 24 hr with LPS (100ng/mL) determined by ELISA. Values represent mean ± SEM. (n = 3 WT and 3 PAI-2^−/−). ns = not significant. (D) Immunoblot analysis of PAI-2 expression in WT, PAI-1^−/− and PAI-2^−/− macrophages treated for 24 hr with LPS (100ng/mL). (E) Quantitation of PAI-2 expression levels by densitometry. Graph shows densitometric analysis of PAI-2 expression after normalizing to β-actin and the same RAW264.7 cell lysate (RAW). All values represent the mean ± SEM. ns = not significant.

Fig. 7.. PAI-1/PAI-2 double deficient macrophages have similar fibrinolytic activity as PAI-1 deficient macrophages.

(A) Cell-based fibrinolysis assay using LPS-activated PAI-2^−/−, PAI-1^−/−, WT and PAI-1/2^−/− macrophages. Values represent the mean ± SEM from quadruplicate wells. Graph shows percent (%) of maximal turbidity, averaged from macrophages isolated from 3 individual mice of each genotype performed in quadruplicate and normalized to the maximal clot formed. (B) Quantitation of fibrinolysis activity from WT, PAI-2^−/− and PAI-1^−/− mice in (A) at endpoint. Graph shows data averaged from 3 independent experiments performed in quadruplicate. Values represent mean ± SEM. *P < 0.05, **P < 0.005, ***P < 0.0005, ns = not significant.