Mice expressing nonpolymerizable fibrinogen have reduced arterial and venous thrombosis with preserved hemostasis

Abstract

Elevated circulating fibrinogen levels correlate with increased risk for both cardiovascular and venous thromboembolic diseases. In vitro studies show that formation of a highly dense fibrin matrix is a major determinant of clot structure and stability. Here, we analyzed the impact of nonpolymerizable fibrinogen on arterial and venous thrombosis as well as hemostasis in vivo using FgaEK mice that express normal levels of a fibrinogen that cannot be cleaved by thrombin. In a model of carotid artery thrombosis, FgaWT/EK and FgaEK/EK mice were protected from occlusion with 4% ferric chloride (FeCl3) challenges compared with wild-type (FgaWT/WT) mice, but this protection was lost, with injuries driven by higher concentrations of FeCl3. In contrast, fibrinogen-deficient (Fga-/-) mice showed no evidence of occlusion, even with high-concentration FeCl3 challenge. Fibrinogen-dependent platelet aggregation and intraplatelet fibrinogen content were similar in FgaWT/WT, FgaWT/EK, and FgaEK/EK mice, consistent with preserved fibrinogen-platelet interactions that support arterial thrombosis with severe challenge. In an inferior vena cava stasis model of venous thrombosis, FgaEK/EK mice had near complete protection from thrombus formation. FgaWT/EK mice also displayed reduced thrombus incidence and a significant reduction in thrombus mass relative to FgaWT/WT mice after inferior vena cava stasis, suggesting that partial expression of nonpolymerizable fibrinogen was sufficient for conferring protection. Notably, FgaWT/EK and FgaEK/EK mice had preserved hemostasis in multiple models as well as normal wound healing times after skin incision, unlike Fga-/- mice that displayed significant bleeding and delayed healing. These findings indicate that a nonpolymerizable fibrinogen variant can significantly suppress occlusive thrombosis while preserving hemostatic potential in vivo.

Graphical abstract

Figure 1.

Fga^WT/EK and Fga^EK/EK mice are protected against the FeCl₃-induced model of AT depending on injury severity. (A) Representative traces of blood flow, (B) vessel occlusion times, and (C) embolic incidence per mouse after a 3-minute exposure of 4% FeCl₃ in Fga^WT/WT, Fga^WT/EK, Fga^EK/EK, and Fga^+/− mice. (D) Representative traces of blood flow, (E) vessel occlusion times, and (F) embolic incidence per mouse after 3-minute exposure of 5% FeCl₃ in Fga^WT/WT, Fga^WT/EK, Fga^EK/EK, and Fga^+/− mice. (G) Representative traces of blood flow, (H) vessel occlusion times, and (I) embolic incidence per mouse after 3-minute exposure of 10% FeCl₃ in Fga^WT/WT, Fga^WT/EK, Fga^EK/EK, Fga^+/−, and Fga^−/− mice. Blue arrows indicate embolic events. Horizontal bars indicate the median. Data were analyzed using Kruskal-Wallis test. (J) Representative images of platelet accumulation (green) in the carotid arteries after injury with 8% FeCl₃ for 1 minute. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001. min, minute; ns, not significant.

Figure 2.

Platelet aggregation and intraplatelet levels of fibrinogen and fibronectin are unaltered in Fga^EK mice. Aggregation traces of platelet-rich plasma collected from Fga^WT/WT, Fga^WT/EK, Fga^EK/EK, and Fga^−/− mice after stimulation with (A) 5 μM adenosine 5′-diphosphate (ADP; n = 4 per genotype) or (B) 250 μM protease-activated receptor 4 activating peptide (Par4p; n = 4 per genotype). (C) Western blot (WB) analyses for fibrinogen, fibronectin, and actin of platelet lysates harvested from Fga^WT/WT, Fga^WT/EK, and Fga^EK/EK mice. Quantification of platelet (D) fibrinogen and (E) fibronectin from the WBs. (F) Western blot analyses for fibrinogen, fibronectin, and albumin of plasma harvested from Fga^WT/WT, Fga^WT/EK, and Fga^EK/EK mice (n = 3 per genotype). Quantification of plasma (G) fibrinogen and (H) fibronectin from the WBs. Data are presented as the mean ± standard error of the mean (SEM), and analyzed using one-way analysis of variance (ANOVA).

Figure 3.

Fga^EK/EK and Fga^WT/EK mice are protected from venous thrombosis. (A) Incidence of thrombus formation and (B) thrombus weights from Fga^WT/WT, Fga^WT/EK, and Fga^EK/EK mice 24 hours after IVC ligation (stasis model). (C) Circulating fibrinogen levels of before and after IVC ligation measured by enzyme-linked immunosorbent assay. (D) Spearman correlation analysis of circulating fibrinogen level after ligation, and thrombus mass. Data are presented as the mean ± SEM and analyzed using Kruskal-Wallis test. (E) Representative images of 5-μm-thick sections of thrombi stained against fibrin(ogen) (green), CD41 (red), and DAPI (4′,6-diamidino-2-phenylindole) (blue). Note that Fga^EK/EK thrombi contained minimal platelets and nucleated cells as well as morphologically distinct aggregates of fibrinogen compared with Fga^WT/WT and Fga^WT/EK thrombi that display web-like fibrin matrixes. ∗P < .05; ∗∗∗P < .001.

Figure 4.

Fga^WT/EK clots have comparable mechanical strength and are resistant to lysis that is not accounted for by differences in clot contraction or FXIII activation. (A) Mean thromboelastography (TEG) curve tracings of whole blood activated with tissue factor and calcium. (B) Maximum amplitude (MA), (C) α angle (degree [deg]), and (D) clot initiation time (R) of whole blood isolated from Fga^WT/WT, Fga^WT/EK, Fga^EK/EK, Fga^+/−, and Fga^−/− mice. (E) Mean TEG curve tracings of whole blood activated with tissue factor and calcium in the presence of 1.5 μg/mL of tPA. (F) MA, (G) α angle, (H) R time, and (I) percentage of lysis 30 minutes after reaching MA (LY30) of whole blood isolated from Fga^WT/WT, Fga^WT/EK, and Fga^+/− mice. (J) Clot mass after clot retraction. (K) Representative western blots and (L) quantification (n = 3 per genotype) of FXIIIa during time course of thrombin-initiated clotting in Fga^WT/WT, Fga^WT/EK, and Fga^EK/EK plasma. (M) Representative scanning electron microscopy images of clots made with platelet-poor plasma of Fga^WT/WT and Fga^WT/EK mice. Arrows indicate fiber termini. (N) Diameters of fibrin fibers of Fga^WT/WT and Fga^WT/EK clots. Large dots indicate biological replicates and small dots indicate total fibers analyzed from all samples. (O) Quantification of the number of fiber terminal ends per fiber in clots made from Fga^WT/WT and Fga^WT/EK mice. Data are presented as the mean ± SEM and analyzed using one-way ANOVA. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001.

Figure 5.

Nonpolymerizable fibrinogen^EK limits fibrin formation and suppresses fibrinolysis. (A-B) Turbidity analysis of Fga^WT/WT, Fga^WT/EK, Fga^EK/EK, Fga^+/−, and Fga^−/− plasma (n = 3) in the (A) absence and (B) presence of 2 μg/mL of tPA. (C-D) Turbidity analysis of plasma containing different ratios of Fga^WT/WT and Fga^EK/EK plasma (n = 3 per ratio) in the (C) absence and (D) presence of tPA. (E-F) Turbidity analysis using increasing amount of purified Fib^WT reconstituted (0.5-1.0 mg/mL) in Fga^−/− plasma (n = 2) in the (E) absence and (F) presence of tPA. (G-H) Turbidity analysis using purified Fib^WT (0.5 mg/mL) and increasing amounts of purified Fib^EK (0-0.5 mg/mL) reconstituted in Fga^−/− plasma (n = 2) in the (G) absence and (H) presence of tPA. OD, optical density.

Figure 6.

Fga^EK/EK mice have preserved hemostatic potential. (A) Time to cessation of bleeding (sustained >30 seconds) of Fga^WT/WT, Fga^WT/EK, Fga^EK/EK, Fga^+/−, and Fga^−/− mice after 3-mm excision of the distal portion of the tail. Horizontal bar indicates the mean with data analyzed using the Kaplan-Meier log-rank test. (B) Time to cessation of bleeding of Fga^WT/WT, Fga^WT/EK, Fga^EK/EK, and Fga^−/− mice after laser-induced saphenous vein injury (n = 18-23 per genotype). Horizontal bar indicates the mean, with data analyzed using one-way ANOVA. (C) Representative images of platelet plug (green), fibrin(ogen) (red), and merged panels 5 minutes after laser-induced saphenous vein injury. (D) Representative 3-dimensional reconstruction of injury sites depicting the side view after injury. Each grid box = 50 μm × 50 μm. Quantification of (E) fibrin(ogen) and (F) platelet accumulation at the site of injury. Data are expressed as the mean ± SEM, and analyzed using one-way ANOVA. ∗P < .05; ∗∗P < .01;  ∗∗∗P < .001. s, second.

Figure 7.

Fga^WT/EK and Fga^EK/EK mice have normal wound healing. (A) Representative images of wound fields and (B) plot of the percentage of mice healed after 1-cm surgical incision in Fga^WT/WT (n = 6), Fga^WT/EK (n = 8), Fga^EK/EK (n = 6), and Fga^−/− (n = 6) mice. Data were analyzed using Kaplan-Meier log-rank test with P < .05 for Fga^WT/WT mice vs Fga^−/− mice and for Fga^WT/EK mice vs Fga^−/− mice. (C) Representative images of hematoxylin and eosin–stained 5-μm-thick sagittal sections of wound sites of Fga^WT/WT, Fga^WT/EK, Fga^EK/EK, and Fga^−/− mice that healed 15 days after incision; scale bar, 200 μm.