P-selectin can promote thrombus propagation independently of both von Willebrand factor and thrombospondin1 in mice

Abstract

Background

P-selectin and thrombospondin1 (TSP1) have been suggested as counter ligands that may mediate GPIbα-dependent thrombus growth independently of VWF in vitro. However, residual thrombus formation still persists in Vwf ^−/−Tsp1^−/− mice suggesting existence of other mechanisms that modulates thrombus propagation.

Objective

We determined whether P-selectin modulates thrombus propagation in injured arterioles independently of TSP1 and VWF.

Methods

CD-62P blocking antibody in Vwf ^−/−Tsp1^−/− mice was used to inhibit P-selectin. We determined thrombus growth kinetics in two models of thrombosis: FeCl₃ injury-induced and laser injury-induced thrombosis.

Results

In a 10% FeCl₃ injury-induced thrombosis model, the initial platelet adhesion, time to form 1^st thrombus, and non-occlusive residual thrombus growth kinetics were comparable between P-selectin-blocking antibody-treated Vwf ^−/−Tsp1^−/− mice and control-IgG treated Vwf ^−/−Tsp1^−/− mice. On the other hand, in a laser injury-induced thrombosis model, residual thrombus growth kinetics was significantly decreased in P-selectin-blocking antibody-treated Vwf ^−/−Tsp1^−/− mice (P< 0.05 versus control-IgG treated Vwf ^−/−Tsp1^−/− mice). Because P-selectin has been suggested as a counter ligand for platelet GPIbα, we determined the role of GPIbα in laser injury-induced thrombosis model. Surprisingly, in a laser injury model, unlike FeCl₃ injury model, thrombus formation was not completely inhibited in IL4Rα/GPIbα-tg mice. Residual thrombus growth kinetics were comparable between P-selectin-blocking antibody-treated IL4Rα/GPIbα-tg mice and control-IgG-treated IL4Rα/GPIbα-tg mice. Comparison of slopes over time showed that residual thrombus growth kinetics were comparable in P-selectin-blocking antibody-treated Vwf ^−/−Tsp1^−/− and control IgG-treated IL4Rα/GPIbα-tg mice

Conclusion

In laser injury-induced thrombosis model, P-selectin modulates thrombus propagation independently of VWF and TSP1.

Introduction

The interaction between platelet GPIbα and its main ligand VWF plays an important role in mediating platelet adhesion/thrombus propagation at sites of vascular injury [1–4]. For several years the conventional view was that GPIbα binding to VWF in the extracellular matrix is the first step required for initial tethering of platelets to the injured vessel wall, particularly under arterial shear conditions [5, 6]. However, platelet adhesion under arterial shear conditions was completely absent in mice lacking platelet GPIbα, but not in mice lacking VWF in the FeCl₃-injury induced thrombosis model, suggesting that other collagen receptors, such as GPVI, α2bβ1, cannot support platelet tethering/adhesion to the injured vessel wall in absence of GPIbα [3]. These findings suggest the existence of alternative ligands for GPIbα other than VWF that mediate initial platelet tethering/adhesion and subsequent thrombus formation. TSP1 was suggested as an adhesive substrate that may mediate GPIbα-dependent platelet adhesion/thrombus formation independently of VWF [7]. In contrast, utilizing Vwf ^−/−Tsp1^−/− mice, we found that TSP1 does not contribute to initial platelet adhesion/thrombus formation independently of VWF [8]. It has been suggested that P-selectin, which is expressed on activated platelets and endothelial cells, can act as a counter ligand for GPIbα, and, thereby may contribute to platelet adhesion/thrombus formation independently of VWF [9, 10]. Herein, we determined whether P-selectin modulates thrombus propagation in injured arterioles independently of both VWF and TSP1.

Materials and Methods

Animals

Vwf ^−/− [1], Tsp1^−/− [11], Vwf ^−/−/Tsp1^−/− [8], IL4Rα/GPIbα-tg [12] and WT littermate mice (4–5 weeks in age) on the C57BL/6J background were used in the study. The University of Iowa Animal Care and Use Committee approved all experiments.

Intravital microscopy of FeCl₃ injury and laser injury-induced mesenteric thrombosis

Intravital microscopy was done as described [8, 13, 14]. Briefly, fluorescent platelets labeled with calcein green (1.25 X 10⁹ platelets/kg) were infused in anaesthetized mice through the retro-orbital plexus. Platelets for infusion were isolated from adult (4–6 months) donor mice of the same genotype. Arterioles of diameter approximately 60–100 μm with shear rates of ~1300–1800 s⁻¹ were used for the experiment. FeCl₃ injury-induced thrombosis: Whatman paper saturated with FeCl₃ (10%) solution was applied topically for 3 minutes and thrombus formation in the injured vessel was monitored in real time by using Nikon inverted microscope. Each injured vessel was recorded by using high-speed EM camera for 40 minutes or until occlusion. Laser injury-induced thrombosis: The injury was initiated in the arterioles using a micropoint laser ablation system. The precise illumination of the area of interest was done through the microscope eyepiece. The wavelength of light used for illumination was in the range of 365–400 nm with maximum output of 50–500 μJ. The power and frequency of bursts were controlled by software and empirically defined. Thrombus formation in the injured vessel (4) was monitored in real time by using a Nikon upright microscope with a Plan Fluor 10X/0.3 objective and thrombus growth was recorded using a high-speed EM camera for 4 minutes. In experimental conditions set up in our lab, in the laser injury-induced thrombosis model, thrombus grows to its maximum size in approximately 1 minute, and then gradually disintegrates in the following 3–4 minutes. Videos were evaluated offline using a Nikon computer-assisted image analysis program.

Statistics

For statistical analysis, Graph Pad Prism (version 7.0) was used. Statistical comparisons were performed using either unpaired, two-tailed, Student’s t test or non-linear regression analysis (comparison of two slopes). P<0.05 was considered statistically significant.

Results and Discussion

We utilized FeCl₃ injury-induced thrombosis and laser injury-induced thrombosis models to determine whether P-selectin modulates thrombus propagation at arterial shear rates independently of both VWF and TSP1. P-selectin was inhibited in Vwf ^−/−Tsp1^−/− mice by intravenous injection of anti-mouse CD-62P blocking antibody (RB40.34, no azide; 2 μg/g of body weight) as described [13, 15]. Identical dose of rat immunoglobulin (IgG) was used as control. Inhibition of P-selectin did not affect complete blood counts (not shown).

In 10% FeCl₃ injury-induced thrombosis model, we found that initial platelet adhesion to injured arterioles was similar in P-selectin-blocking antibody-treated Vwf ^−/−Tsp1^−/− mice compared with control-IgG-treated Vwf ^−/−Tsp1^−/− mice (Figure 1A). We speculated that P-selectin might support thrombus propagation, but not platelet adhesion, via other mechanisms independently of VWF and TSP1 based on following studies. These include: 1) recruitment of tissue bearing microparticles in growing thrombus [16], 2) shear-dependent platelet aggregation [17], and 3) via its soluble form [18]. We found that time to form 1^st thrombus, and non-occlusive thrombus growth kinetics were similar in P-selectin-blocking antibody-treated Vwf ^−/−Tsp1^−/− mice when compared with control-IgG-treated Vwf ^−/−Tsp1^−/− mice (Figure 1BC). As expected none of the vessels occluded in P-selectin-blocking antibody-treated Vwf ^−/−Tsp1^−/− mice or control-IgG-treated Vwf ^−/−Tsp1^−/− mice (Figure 1D) because it is known that VWF is required for occlusive thrombus growth under arterial and venous shear conditions [1, 2, 4, 19]. Similarly, P-selectin inhibition did not affect initial platelet adhesion and thrombus growth rate in WT mice (Figure 2). Together these results suggest that, in a 10% FeCl₃ injury model, P-selectin is not essential for thrombus propagation.

Figure 1.

Figure 2.

In laser injury-induced thrombosis model, we found that inhibition of P-selectin in WT mice significantly decreased platelet mean fluorescence intensity (MFI) and rate of thrombus growth (Figure 3AB). Our findings are consistent with previous studies that have suggested that P-selectin was incorporated within the thrombus and modulates thrombus propagation in laser injury-model [16]. Furthermore, in the Vwf ^−/−, Tsp1^−/−, and Vwf ^−/−Tsp1^−/− mice, we found that platelet MFI was markedly, but not completely, reduced when compared to WT mice (Figure 3) suggesting that both VWF and TSP1 contribute to thrombus propagation in laser injury-induced thrombosis model. Because residual thrombus was present in Vwf ^−/−, Tsp1^−/−, and Vwf ^−/−Tsp1^−/− mice, it was feasible to determine whether P-selectin modulates thrombus propagation independently of VWF or TSP1 or both. Platelet MFI and thrombus growth rate were decreased in P-selectin-blocking antibody-treated Vwf ^−/− mice (P<0.05 versus control-IgG-treated Vwf ^−/− mice; Figure 3CD), P-selectin-blocking antibody-treated Tsp1^−/− mice (P<0.05 versus control-IgG-treated Tsp1^−/− mice; Figure 3EF), and P-selectin-blocking antibody-treated Vwf ^−/−Tsp1^−/− mice (P<0.05 vs. control-IgG-treated Vwf ^−/−Tsp1^−/− mice; Figure 4AB). These findings suggest that, at arterial shear rate, P-selectin can contribute to thrombus propagation on the activated endothelium independently of VWF and TSP1.

Figure 3.

Figure 4.

Because P-selectin has been suggested as a counter ligand for platelet GPIbα[9], we next determined whether GPIbα is absolutely required to form thrombus in laser injury-induced thrombosis model. We utilized GPIbα/human IL4R transgenic (IL4Rα/GPIbα-tg) mice, which lack murine GPIbα, but express the extracellular domain of the human interleukin-4 (IL-4) receptor fused to the transmembrane and cytoplasmic domains of human GPIbα[12]. IL4Rα/GPIbα-tg mice have a 2-fold increase in circulating platelet count and a 50% reduction in platelet size when compared to platelets from GPIbα^−/−, a mouse model of Bernard-Soulier syndrome [12]. Surprisingly, unlike FeCl₃ injury model, in a laser injury model thrombus formation was not completely inhibited in IL4Rα-GPIbα-tg mice (Figure 4AB). Residual thrombus growth kinetics were comparable between P-selectin-blocking antibody-treated IL4Rα/GPIbα-tg mice and control-IgG-treated IL4Rα/GPIbα-tg mice (Figure 4CD). This result suggests that ligands other than P-selectin and GPIbα initiate thrombus formation in laser injury model (Figure 4CD). One of the possible mechanisms could be neutrophil interaction with activated endothelial cells, through LFA-1/ICAM1, independently of platelet GPIbα and P-selectin. Such a mechanism has been proposed as the first step that initiates fibrin generation before propagation of platelet thrombus in a laser injury model [20]. When we compared slopes over time, we found that thrombus growth kinetics were comparable in P-selectin-blocking antibody-treated Vwf ^−/−Tsp1^−/− mice (Slope: 0.00223 ± 0.00064) and control IgG-treated IL4Rα/GPIbα-tg mice (Slope: 0.00246 ± 0.00062; Figure 4B). Previous studies suggest that platelets GPIbα-P-selectin interaction can support platelet interaction on activated endothelium at venous shear rate [9, 10]. Although not proved in this study, it is possible that GPIbα-P-selectin interaction may support thrombus propagation independently of both VWF and TSP1. Further studies using a specific inhibitor of P-selectin/GPIbα interaction or mutant mice are warranted to confirm these findings. The possibility that there may be additional mechanisms by which P-selectin may modulate thrombus propagation independently of VWF and TSP1 could not be ruled out. For example, a study in baboon has shown that blocking platelet-leukocyte interactions by anti P-selectin antibodies inhibits fibrin deposition in growing thrombus [21]. Another study showed that P-selectin interaction with an unknown ligand different from PSGL-1 or GPIbα could stabilize initial αIIβ3-fibrinogen interactions resulting in formation of stable platelet aggregates in vitro [22].

Our studies have few limitations. First, our conclusions that P-selectin modulates thrombus propagation independently of VWF and TSP1 are based on laser injury-induced thrombosis model. Second, we did not define the role of endothelial P-selectin versus platelet P-selectin in thrombus propagation independently of VWF and TSP1. Further studies utilizing bone-marrow experiments from P-sel^−/−Vwf ^−/−Tsp1^−/− are warranted. Third, under what pathological conditions, such mechanism exists remain unclear. It is possible that P-selectin may contribute to platelet aggregation independently of VWF and TSP1 in certain conditions such as Helicobacter pylori infection [23].

In conclusion, our studies provide the first in vivo evidence that P-selectin on activated platelets and/or endothelial cells can provide a backup mechanism to VWF and TSP1 in recruiting and activating quiescent platelets, thereby promoting thrombus propagation.

Essentials.

• The main receptor for platelet glycoprotein (GP) Ibα is von Willebrand factor (VWF).

• P-selectin and thrombospondin1 (TSP1) have been suggested as counter receptors for GPIbα.

• In laser injury model, P-selectin promotes thrombus propagation independently of VWF and TSP1.

• In laser injury model, thrombus persists in interleukin-4 receptor α (IL4α)/GPIbα-transgenic mice.