Genetic and antibody-induced glycoprotein VI deficiency equally protects mice from mechanically and FeCl3-induced thrombosis


Bernhard Nieswandt, University Hospital Würzburg and Rudolf Virchow Center, DFG Research Center for Experimental Biomedicine, Josef-Schneider-Str. 2, 97080 Würzburg, Germany.
Tel.: +49 931 31 80405; fax: +49 931 201 61652.

At sites of vascular injury, platelet adhesion, activation and aggregation occurs on exposed subendothelial collagens and contributes to hemostasis but also to occlusion of diseased vessels leading to a myocardial infarction or ischemic stroke. The activating platelet collagen receptor, glycoprotein (GP) VI [1], is a platelet-specific transmembrane type I receptor that non-covalently associates with the Fc receptor (FcR) γ-chain which contains an immunoreceptor tyrosine-based activation motif (ITAM) [2]. Upon ligand-induced GPVI clustering, the ITAM becomes tyrosine phosphorylated and initiates a series of phosphorylation events finally resulting in cellular activation [3]. GPVI has been proposed as an attractive antithrombotic target, particularly because anti-GPVI antibodies have been shown to induce irreversible downregulation of the receptor through ectodomain shedding in circulating platelets in mice, resulting in long-term antithrombotic protection but overall normal tail bleeding times [4–7]. Different mouse models of GPVI-deficiency (FcR γ-chain-deficient which also lack GPVI [8], GPVI-immunodepleted or Gp6−/− mice) have been used to determine the functional relevance of the receptor in thrombus formation in different experimental thrombosis models (laser-, mechanically and chemically induced injuries). Although GPVI is generally regarded as a critical regulator of pathological thrombus formation [9], partially contradictory results have been reported on the relative importance of GPVI in some of these models. For example, GPVI was found to play no [10,11] or only a minor [12,13] role in models of laser-induced vascular injury where thrombus formation is suggested to be largely dependent on thrombin generation, although different results were reported by others [14].

Ferric(III)chloride (FeCl3)-induced vascular injury is a widely used model in thrombosis research because it allows variable levels of injury in different vascular beds and can be monitored by microscopic visualization or blood flow measurements. Previous studies yielded evidence for a role of GPVI in FeCl3-induced thrombus formation in mice [10,15,16], but this was questioned in a very recent study assessing the mechanisms underlying thrombus formation in that model in detail [17]. The authors demonstrated that in their model, FeCl3 diffuses through all vessel wall layers and severely injures the endothelium without damaging the internal elastic lamina (IEL), and only exposes the basement membrane to the vessel lumen. They further report that in their hands GPVI-immunodepleted mice are not protected from vessel occlusion after FeCl3-induced injury of mesenteric arterioles or carotid arteries by measuring thrombus area and time to occlusion. Based on these data, the authors state that FeCl3-induced vascular injury may not be an appropriate model to study the role of subendothelial adhesive proteins and their platelet receptors in thrombosis [17]. This conclusion, if correct, may have major implications for the interpretation of a large number of previous results obtained with this model and also significantly influence its acceptance as a valid experimental system to assess the mechanisms of arterial thrombosis. Therefore, here we assessed the role of GPVI-mediated platelet activation/adhesion in FeCl3-induced thrombus formation systematically. To do so, we subjected control, Gp6−/− (details of the knockout generation will be published elsewhere) and GPVI-immunodepleted (100 μg JAQ1 i.p., analysis on day 6 post injection) mice to a model of FeCl3-injury of mesenteric arterioles and carotid arteries, and monitored thrombus formation by intravital fluorescence microscopy or blood flow measurements, respectively. In addition, thrombus formation in these mice was tested in a model of mechanical injury of the abdominal aorta.

A drop of 20% FeCl3 was topically applied on small mesenteric arterioles of 4- to 5--week-old mice and thrombus formation of fluorescently labeled platelets was monitored for 40 min or until complete occlusion of the vessel had occurred. Kinetics of platelet adhesion (data not shown) and the beginning of first thrombus formation were similar between control, Gp6−/− and GPVI-immunodepleted mice (Fig. 1A, left). In contrast to control mice, reduced thrombus stability was observed in Gp6−/− and GPVI-immunodepleted mice, evident by permanent embolization of differently sized thrombus fragments (Fig. 1A, right and bottom). Consequently, while all vessels occluded in control mice (mean occlusion time: 17.9 ± 3.2 min), most vessels in GPVI-deficient mice remained open or occlusion was significantly delayed (occluded vessels/total vessels: Gp6−/−: 2/10, GPVI-immunodepleted: 5/11, Fig. 1A). Similar results were obtained when thrombus formation was assessed in carotid arteries that were injured by topical application of a 10% FeCl3-saturated filter paper for 1.5 min and blood flow was monitored with an ultrasonic flow probe. Whereas 11 out of 12 vessels of control mice rapidly occluded within 589 ± 295 s, permanent embolization was observed in both Gp6−/− and GPVI-immunodepleted mice that finally resulted in delayed or no stable vessel occlusion (Fig. 1B). These results clearly demonstrated that GPVI is involved in thrombus formation after FeCl3-induced injury of mesenteric arterioles or the carotid artery in mice. This strongly suggests that FeCl3 produces a type of vascular injury that exposes subendothelial collagens and other extracellular matrix (ECM) components to the flowing blood and that these significantly contribute to platelet recruitment and activation. The present findings are in sharp contrast to the data obtained by Eckly et al. [17], but in agreement with data from other groups [10,15,16]. It may be difficult to explain these discrepancies but different experimental procedures are probably involved.

Figure 1.

 Evidence for a role of glycoprotein (GP) VI in different murine thrombosis models. (A) Mesenteric arterioles were injured with 20% FeCl3, and adhesion and thrombus formation of fluorescently labeled platelets was monitored by in vivo microscopy. Statistical evaluation of the beginning of first thrombus formation (upper left) and time to occlusion (upper right) are shown. The horizontal line indicates a mean value. Note: Delayed beginning of first thrombus formation in GPVI-immunodepleted mice is not significant compared with control mice. Each symbol represents one arteriole. Representative images from different time points of thrombus formation are depicted (lower panel). (B) Carotid arteries were topically injured with a saturated filter paper with 10% FeCl3 for 1.5 min and time to occlusion was determined. Each symbol represents one individual. (C) The abdominal aorta was mechanically injured using forceps (compression for 15 s), and blood flow was monitored for 30 min. Representative curves of the blood flow are shown. Depl., depleted.

Eckly et al. [17] hypothesized that adhesive proteins important for GPVI activation, such as collagen type I, are not exposed to blood after FeCl3-induced injury. However, the basement membrane contains many proteins including collagen type IV and laminin. The latter has been described to stimulate spreading of platelets through integrin α6β1-dependent activation of GPVI [18]. Moreover, very recently, it was speculated that proteins from the tunica media, including collagen type I, could still be exposed [19] as the IEL is fenestrated [20]. Therefore, the mechanisms of how platelets are recruited to sites of vascular injury after application of FeCl3 still require further investigations.

It has been suggested that in the FeCl3-induced injury model after severe injury, thrombus formation is mainly triggered by tissue factor exposure and thrombin generation [10,21–23]. Possibly, in the experimental setting used by Eckly et al. [17] higher concentrations of tissue factor and thrombin were produced which bypassed the collagen-dependent activation pathway. In line with this notion, we also observed a diminished, but not absent, role of GPVI in this model when injury was induced by much higher concentrations (e.g. 15% to injure the carotid artery) or prolonged exposure times (e.g. carotid artery: 3 min) of FeCl3 (data not shown), but this result may not be surprising as such exceedingly strong injury will produce maximal activation of different pathways, each of which by itself may be sufficient to drive occlusive thrombus formation. However, it is questionable whether such massive activation reflects the situation in human disease.

In a last set of experiments, we subjected Gp6−/− and GPVI-immunodepleted mice to a mechanical injury model in which the abdominal aorta was compressed with forceps for 15 s and blood flow was monitored by an ultrasonic flow probe for 30 min [24]. As in the FeCl3-induced injury models, both GPVI-deficient mouse strains were profoundly protected from vessel occlusion (controls: 8/8 occluded, Gp6−/−: 0/6, GPVI-immunodepleted: 1/8) (Fig. 1C). However, in contrast to FeCl3-induced injury of the carotid artery, only an initial reduction of the blood flow and minor embolizations were measured in both groups of GPVI-deficient mice indicating that thrombus growth was severely inhibited in this experimental model.

Taken together, we demonstrate that GPVI-mediated platelet activation/adhesion mechanisms essentially contribute to thrombus formation after FeCl3-induced vascular injury in mice. These findings reconfirm that this injury model is well suited to study the involvement of ECM constituents and respective platelet receptors in thrombus formation. Furthermore, our data provide direct evidence that GPVI-immunodepletion very well mirrors the thromboprotective phenotype of genetic GPVI deficiency in different thrombosis models. This strongly suggests that anti-GPVI treatment does not produce significant undesired side effects on platelet function. This may have important implications for the development of anti-GPVI-based antithrombotic agents.


M. Bender and I. Hagedorn equally performed experiments, analyzed data and contributed to the writing of the manuscript. B. Nieswandt planned the project, analyzed data and contributed to the writing of the manuscript.


This work was supported by the Deutsche Forschungsge-meinschaft (Sonderforschungsbereich 688 to B. Nieswandt) and the Rudolf Virchow Center.

Disclosure of Conflict of Interest

The authors state that they have no conflict of interest.