Xinmailong Modulates Platelet Function and Inhibits Thrombus Formation via the Platelet αIIbβ3-Mediated Signaling Pathway

Background: Xinmailong (XML), a bioactive composite extracted from Periplaneta americana, has been widely used to treat cardiovascular diseases such as congestive heart failure. However, it is unclear whether XML has antiplatelet and antithrombotic effects. Methods: The effects of XML on agonist-induced platelet aggregation, adhesion and spreading, granule secretion, integrin α II bβ3 activation, and thrombus formation were evaluated. Phosphorylation of Syk, PLCγ2, Akt, GSK3β, and MAPK signaling molecules was also studied on agonist-induced platelets. In addition, the antithrombotic effects of XML were observed in vivo using an acute pulmonary thrombosis mouse model. Results: XML dose-dependently inhibited in vitro platelet aggregation and granule secretion induced by thrombin, collagen, and arachidonic acid (AA). XML also greatly reduced platelet adhesion and spreading on both collagen- and fibrinogen-coated surfaces. Biochemical analysis revealed that XML inhibited thrombin-, collagen-, and AA-induced phosphorylation of Syk, PLCγ2, Akt, GSK3β, and MAPK. Additionally, XML significantly inhibited in vivo thrombus formation in a collagen–epinephrine-induced acute pulmonary thrombosis mouse model. Conclusions and General Significance: Here, we provide the first report showing that XML inhibits platelet function and that it possesses antithrombotic activity. This suggests that XML could be a potential therapeutic candidate to prevent or treat platelet-related cardiovascular diseases.


INTRODUCTION
Thrombotic diseases, such as stroke, myocardial infarction, and arteriosclerosis have become the primary cause of death in humans, and platelets play crucial roles both in physiological hemostasis and pathological thrombosis (Michelson, 2010). Upon vessel wall damage, platelets are exposed to subendothelial extracellular matrix proteins (ECM) such as collagen, which quickly trigger platelet activation and adhesion (Varga-Szabo et al., 2008). Platelets aggregate with each other and release pro-activator substances, such as adenosine diphosphate (ADP), thromboxane A2 (TxA2), and thrombin. These substances further amplify platelet activation and recruit more platelets to form hemostatic thrombi (Watson et al., 2001). Finally, these events activate integrin α II bβ3, allowing platelets to combine with fibrinogen to mediate adhesion and aggregation (Shattil and Newman, 2004). Integrin α II bβ3 maintains an inactive conformation on the surface of resting platelets. Cytoplasmic signals cause it to adopt an active and high-affinity conformation in a process known as "inside-out" signal transduction, which occurs via G protein-coupled or tyrosine kinase-linked pathways. Activated integrin α II bβ3 can bind soluble ligands such as fibrinogen and von Willebrand factor (vWF) and subsequently mediate "outside-in" integrin signaling, which leads to integrin clustering, changes in platelet morphology, and postoccupancy events (Mehrbod et al., 2013). These signaling events are responsible for stimulating platelet adhesion and thrombus growth via influencing processes such as clot retraction, platelet aggregation, and platelet spreading (Shattil, 1999). The FcγR II A ITAM (immunoreceptor tyrosine-based activation motif)/ Syk/PLCγ2 and the PI3K/Akt/GSK3β signaling pathways play a major role in triggering integrin α II bβ3 "outside-in" signaling (Zhi et al., 2013). ERK1/2, JNK1/2, and p38 are also involved in platelet signal transduction (Adam et al., 2008).
A series of antiplatelet drugs have been used clinically in the treatment of atherothrombotic diseases, such as aspirin and clopidogrel. However, these drugs have some side effects, including gastric bleeding and ulcer development. Thus, a special focus has been placed on natural compounds present in medicinal or dietary plants, especially in Chinese traditional herbs, that exhibit antiplatelet activity (Wang et al., 2014). Xinmailong (XML)-a bioactive composite extracted from Periplaneta americana (a species of cockroach)-was approved by the China State Food and Drug Administration (CFDA) in 2006 for the treatment of heart failure (HF) (Liu, 2016).
Previously studies have reported the chemical constituents of XML using high-performance liquid chromatography (HPLC) and gas chromatograph-mass spectrometry (GC-MS). 29 compounds have so far been identified, including three of polyhydric alcohols (percentage composition: 38.5%), four of organic acids (percentage composition: 18.8%), a variety of alkaloids (pyrroles, piperidines, piperazines, percentage composition: 6.55%), six of fatty acids and their esters (percentage composition: 5.88%), two of unsaturated lactones, two of amines, two of phenolic compounds, and some microconstituents (such as noradrenaline, ketone compounds, and divinyl sulfide). (Jiao et al., 2011;Jiao et al., 2012) It has been reported that XML contains four main active ingredients: adenosine, inosine, protocatechuic acid, and pyroglutamate dipeptide (Figures 1A-D) (Qi et al., 2017). A series of clinical trials performed in HF patients confirmed the effectiveness of XML in improving cardiac function, such as enhanced myocardial contraction and inhibition of ventricular remodeling (Lu et al., 2018). However, no studies have determined whether XML has anti-platelet or anti-thrombotic effects. Therefore, our study focuses on evaluating the effect of XML on platelet activation and thrombosis. Thrombin, collagen, and AA are the most important platelet agonists and can active platelets by inducing signal transduction. In the present study, we used thrombin, collagen, and AA stimulations of platelets to determine if XML differentially regulates platelet function. Here, we provide the first report that XML suppresses thrombin-, collagen-, and AA-induced platelet activation. Moreover, we reveal the potential effects of XML on the thrombin-, collagen-, and AA-stimulated platelet signaling pathway. Based on these studies, we propose that XML inhibits platelet activation through regulation of integrin α II bβ3-mediated signal transduction.

Platelet Preparation
Human blood was drawn from the cubital vein without stasis into silicon vacutainer tubes containing 1:9 (v/v) 3.8% acid citrate dextrose. Platelet-rich plasma (PRP) was obtained by centrifuging whole anti-coagulant blood for 15 min at 300×g after addition of 1 μM prostaglandin E1 (PGE1). After another addition of 1 μM PGE1 and 2.5 mM EDTA, platelets were pelleted by centrifuging the PRP at 300×g for 15 min. Samples were then washed twice with Tyrode buffer (137 mM NaCl, 13.8 mM NaHCO 3 , 5.5 mM glucose, 2.5 mM KCl, 20 mM HEPES, 0.36 mM, NaH 2 PO 4 , pH 7.4) and resuspended in 1× Tyrode buffer for final concentration of 3.0 × 10 8 /ml. All of the platelet preparations were conducted at room temperature. CaCl 2 (1 mM) was added prior to agonist stimulation.
Mouse blood was drawn into a syringe containing 3.8% sodium citrate 1:9 (v/v) from the posterior vena cava of anesthetized Sprague-Dawley rats. Platelets were obtained using the protocol described above.

Platelet Aggregation and ATP Release Assay
Platelet aggregation and ATP release were measured on a light transmission aggregometer (Chrono-Log, Havertown, PA, USA) as described previously (Ming et al., 2011). The PRP or washed platelets (250 μl, 3.0 × 10 8 /ml) were preincubated with either XML (0.1, 1, 10 mg/ml) or vehicle at 37°C for 10 min and then stimulated with different agonists, while stirring for 6 min at 37°C. Platelet aggregation and secretion (measuring the release of ATP) were recorded. Quantification was obtained by the aggregation and release tracings.

Platelet Adhesion and Spreading Assays
Twenty four well glass tissue-culture slides were coated with collagen (5 μg/ml) or fibrinogen (25 μg/ml) overnight at 4°C. The slides were washed twice with phosphate buffer saline (PBS) and then, using 1% albumin from bovine serum (BSA), were blocked for 60 min at room temperature. Washed platelets (2 × 10 7 /ml) pretreated with different concentrations of XML (1 and 10 mg/ ml) or vehicle were allowed to adhere and spread on glass cover slips coated with collagen or fibrinogen for 45 min at 37°C. Nonadherent platelets were removed by PBS, and adherent platelets were fixed with 2% paraformaldehyde, permeabilized with 0.1% triton X-100, and stained with phalloidin-TRITC (1 μg/ml). Adherent platelets were viewed and photographed using a fluorescence microscope (Olympus, Japan). ImageJ software was used to analyze the number of platelets and the spreading area of individual platelets.

Clot Retraction Assessment
Platelet clotting was initiated by adding 10 μl of thrombin (10 U/ml) to 500 μl of platelets (2 × 10 7 /ml) in the presence of 400 μg/ml fibrinogen and 1 mM CaCl 2 . The reaction was allowed to proceed for 30 min at 37°C, and photographic images of the clot retraction were recorded.

Western Blots
Washed platelets were preincubated with vehicle or XML (10 mg/ml) for 10 min. After platelet stimulation with or without agonist for 2 min, reactions were terminated by the addition of an equal volume of 2× lysis buffer (300 mM NaCl, 0.2 mM MgCl 2 , 20 mM EGTA, 30 mM HEPES, 2% Triton X-100, 2× protease, 2× phosphatase inhibitor cocktails). Proteins were electrophoresed on a 10% SDS-PAGE gel and transferred to polyvinylidene difluoride (PVDF) membranes by semi-dry Western blotting. PVDF membranes were blocked using 5% (w/v) BSA, then incubated with the indicated antibodies overnight at 4°C. Membranes were washed with TBST, then exposed to the appropriate secondary antibodies for 1 h at room temperature and detected with chemical luminescence image system.

Tail-Bleeding Assay
Male mice were divided into the following groups: vehicle, 20 mg/kg XML, or 40 mg/kg XML. Then, the mice were anesthetized with chloral hydrate by intraperitoneal injection. Once anesthetized, vehicle or XML (20 mg/kg, 40 mg/kg) solution was injected via the tail vein. After 30 min, tail-bleeding time was measured as previously described (Endale et al., 2012). The distal 5 mm of the tail was transected and immediately immersed into 0.9% saline solution at 37°C. The time between the start of transection to bleeding cessation was recorded as the bleeding time.

Coagulation Assessment
Rats were segregated into the following groups: vehicle, 20 mg/kg XML, or 40 mg/kg XML. Then, the rats were anesthetized with chloral hydrate by intraperitoneal injection. Once anesthetized, vehicle or XML (20 and 40 mg/kg) solution was injected via the tail vein. After 30 min, blood was drawn from the inferior vena cava without stasis into silicon vacutainer tubes containing 1:9 (v/v) 3.8% acid citrate dextrose. Prothrombin time (PT) and activated partial thromboplastin time were detected using an automated coagulation analyzer.

Collagen-Epinephrine-Induced Acute Pulmonary Thrombosis in Rats
To investigate the effect of XML on thrombosis in vivo, the collagen-epinephrine-induced acute pulmonary thrombosis mouse model was adapted. A total of 50 rats were equally divided into five groups as control, vehicle, 50 mg/kg aspirin, 10 mg/kg XML, and 20 mg/kg XML, which were injected via the tail vein saline solution, aspirin (50 mg/kg), or different concentration of XML (10 mg/kg, 20 mg/kg), respectively. 30 minutes later, all groups except for control were injected with a mixture of collagen (1.0 mg/kg) and epinephrine (0.1 mg/kg) in the tail vein to induce thrombus formation. The survival time in each group was recorded, and lung sections were visualized using H&E staining.

Statistical Analysis
All data were analyzed using the Statistical Package for Social Sciences (SPSS, version 20.0) and expressed as the mean ± SEM (standard error of mean). Data were analyzed using one-way ANOVA. Student Newman Keuls (SNK) test was used to determine whether the differences between groups were significant. P values less than 0.05 were considered statistically significant.

XML Inhibits Activation of Platelet
Signaling Pathways by Thrombin, Collagen, and AA Several signaling pathways, including Syk-PLCγ2-MAPK and PI3K-Akt-GSK3β, were evaluated. We found that agonistinduced Syk and PLCγ2 phosphorylation were significantly inhibited by XML (10 mg/ml) treatment (Figure 7). Next, we examined phosphorylation levels of mitogen-activated protein kinase (MAPK) signaling pathways. We found that thrombin, collagen, and AA stimulations led to the activation of MAPKs (JNK, ERK1/2, and p38) in platelets, and that XML (10 mg/ ml) blocked this activation (Figure 7). These results suggest that XML may be a negative regulator of Syk-PLCγ2-MAPK signaling.
Furthermore, Akt phosphorylation was detected after platelets were stimulated by thrombin, collagen, and AA. As shown in Figure 7, XML (10 mg/ml) inhibited this process. Similarly, thrombin-, collagen-, and AA-induced phosphorylation of GSK3β, a substrate of Akt, was attenuated by XML (10 mg/ml) (Figure 7). These findings demonstrate that XML may function as a negative regulator of PI3K-Akt-GSK3β signaling pathways.

XML Negatively Regulates Integrin α II bβ3-Mediated "Outside-In" Signaling
As platelet aggregation is dependent upon the association of activated integrin α II bβ3 in complex with its soluble ligand, fibrinogen, we further study clot retraction and platelet spreading on fibrinogen-coated surfaces.
Clot retraction is an integrin α II bβ3-mediated function involving cytoskeletal reorganization. As shown in Figure 9, a substantial reduction in clot retraction was observed in the presence of XML-treated (1, 10 mg/ml) platelets compared with the vehicle group.

Effect of XML on Coagulation Function
To assess the effect of XML on hemorrhage and hemostasis, tail-bleeding time was measured in mice. Animals were randomly segregated into three groups (vehicle, 20 XML, 40 mg/kg XML). Once anesthetized, vehicle or XML (20 mg/ kg, 40 mg/kg) solution was injected via the tail vein. 30 min after XML administration, tail-bleeding time was assessed. As shown in Figure 11, administration of XML (20, 40 mg/kg) did not extend bleeding time compared to the vehicle group. (P > 0.05, n = 6).
To investigate the effect of XML on coagulation in vivo, rats were segregated into vehicle, 20 mg/kg XML, or 40 mg/kg XML groups. Once anesthetized, vehicle or XML solution was injected via the tail vein. After 30 min, blood was drawn from the inferior vena cava. As shown in Figure 12, we found that neither PT nor activated partial thromboplastin time (APTT) was affected by different doses of XML (20, 40 mg/kg) (P > 0.05, n = 10), indicating that XML has no effect on coagulation in vivo.

XML Inhibited Thrombus Formation In Vivo
To explore the effect of XML on thrombosis in vivo, we developed an acute pulmonary thrombosis mouse model using collagen-epinephrine-induced as described in the "Materials and methods" section. We used vehicle, 10 mg/kg XML, 20 mg/ kg XML, and 50 mg/kg aspirin treatment to examine thrombus formation. As shown in Figure 13A, platelet thrombi formed throughout the pulmonary vasculature in the vehicle group. However, the lungs of mice injected with either 10 mg/kg XML, 20 mg/kg XML, or 50 mg/kg aspirin showed a substantial reduction in thrombus formation. Importantly, aspirin 50 mg/ kg and XML (20 mg/kg, 40 mg/kg) treatment prolonged mice survival time ( Figure 13B). These data demonstrate that XML may alleviate thrombosis in vivo.

DISCUSSION
This study provides the first report of the antiplatelet activity of XML. Using several approaches, we have defined the effects of XML on platelet function, characterized underlying mechanism, and evaluated its ability to regulate platelet-mediated thrombus formation in vivo. This study suggests that XML inhibits agonistinduced platelet activation and integrin α II bβ3-mediated signaling, and that it exerts anti-thrombotic activity in vivo. XML inhibited collagen-, thrombin-, and AA-induced platelet aggregations and also inhibited ATP secretion and P-selectin expression. Furthermore, XML reduced platelet adhesion and spreading when immobilized on both collagen-and fibrinogen-coated surfaces and attenuated platelet clot retraction. These results suggest that XML inhibits integrin α II bβ3-mediated signaling. XML impaired agonist-induced phosphorylation of molecules including those involved in the GPVI-Syk-PLCγ2-MAPK and PI3K-Akt-GSK3β signaling pathways. Finally, XML serves to negatively regulate platelet thrombus formation after collagen-epinephrine-induced acute pulmonary thrombosis in vivo.
Under physiological conditions, maintenance of platelet quiescence requires a balance between activation and inhibition signaling events. Healthy arterial endothelium is able to provide powerful inhibitory signaling to platelets. However, upon vascular injury, platelets are activated by adhesion to adhesive proteins, such as von Wille-brand factor (vWF) and collagen, or by soluble platelet agonists, such as thrombin, TXA 2 , and FIGURE 6 | XML inhibits platelet adhesion on collagen-coated surfaces. (A) Human washed platelets (2.0 × 10 7 /ml) were pre-treated with vehicle or XML (1 mg/ml, 10 mg/ml) for 10 min at 37°C, then allowed to adhere on collagen-coated slides. We selected representative images from three independent experiments. (B) The number of adherent platelets and (C) the average surface area of individual platelets was analyzed. (Mean ± SEM, n = 3, *P <0.05, **P <0.01 vs. vehicle). August 2019 | Volume 10 | Article 923 Frontiers in Pharmacology | www.frontiersin.org  PLCγ2, a downstream molecule of Syk, subsequently induces the formation of the second messengers IP3 and DAG, leading to calcium mobilization, granule secretion, platelet aggregation, and integrin α II bβ3 regulation (Harrison et al., 2018). MAPKs are a downstream target of SFK, and many studies have shown that MAPKs including ERK, JNK, and p38 contribute significantly to platelet activation by various agonists (Chang and Karin, 2001). ERK has been shown to enhance platelet aggregation and secretion, while JNK is thought to decrease integrin α II bβ3 activation and impair granule secretion. p38 is involved in platelet spreading and adhesion (Kuliopulos et al., 2004;Adam et al., 2008;Adam et al., 2010). PI3K signaling contributes to collagen-/GPVI-mediated activation of PLCγ2 by generating PIP3, which induces PLCγ2 translocation to the cell membrane. FIGURE 11 | Effect of XML on hemorrhage and hemostasis. The mouse tails were transected to induce bleeding, and the time to bleeding cessation was recorded. Data are reported as means ± SEM (n = 6). N.S., no significant difference versus vehicle. FIGURE 13 | XML-or aspirin-treated mice were protected from lung vascular thrombosis. Mice were pre-treated with 10 mg/kg XML, 20 mg/kg XML, or 50 mg/kg aspirin for 30 min prior to the intravenous injection of collagen (1.0 mg/kg) and epinephrine (0.1 mg/kg). (A) Representative sections of H&E-stained lung sections from mice treated as labeled. Magnification: ×200. (B) Survival time of collagen-/epinephrine-induced mice injected with XML or aspirin. XML was effective in preventing collagen/epinephrine-induced thromboembolic events (n = 10). **P <0.01 vs. model. N.S., means no significant differences vs. model. This is essential for full activation of PLCγ2 (Ragab et al., 2007). Akt, a downstream molecule of PI3K, and its substrate GSK3β play a vital role in platelet activation as shown by several studies (Li et al., 2008;O'Brien et al., 2011). AA can synthesize TXA 2 in activated platelets through the cyclooxygenase (COX) pathway. Once formed, TXA 2 diffuses across the membrane and activate other platelets via its receptors: TPα and TPβ, which couple to the G q and G 12/13 proteins. Thrombin stimulates platelet activation mainly via G-protein-linked protease-activated receptors (PAR S ), which couple to the proteins G 12/13 , G q , and G i (Davì and Patrono, 2007). Gq protein-coupled receptors stimulated the β 2 and β 3 isoforms of phospholipase C (PLC). PLC activation leads to calcium mobilization and activation of protein kinase C (PKC). Gq proteins also stimulate the PI3K/Akt pathway, resulting in changes in platelet morphology, degranulation, and integrinmediated aggregation (Rivera et al., 2009). Thus, thrombin and AA utilize a common signaling pathway that can activate PLCβ/ PKC. It is well known that most agonists activate platelets through G-protein-coupled receptors, leading to integrin glycoprotein II b/III a (α II bβ3) activation, which is the final step in platelet activation (Davì and Patrono, 2007;Stegner and Nieswandt, 2011).
In our study, we found that platelet adherence to immobilized collagen-coated surfaces was decreased by XML. Similarly, XML inhibited collagen-, thrombin-, and AA-induced platelet aggregations in dose-dependent fashion.
When platelets are activated, intracellular secretory granules such as α-granules, dense granules, and lysosomes are secreted, which is key to platelet activation and thrombus stabilization. The α-granules are the most abundant and contain many bioactive proteins, such as vWF, fibrinogen, and P-selectin (Diacovo et al., 1996;Sharda and Flaumenhaft, 2018). Thus, we examined the expression of immunofluorescence-labeled P-selectin to identify α-granule secretion. XML reversed collagen-, thrombin-, and AA-induced platelet P-selectin expressions. We detected ATP release to show dense granule secretion (McNicol and Israels, 1999). Consistently, XML significantly reduced collagen-, thrombin-, and AA-induced ATP releases from platelets. These observations imply that XML may exert its antiplatelet effect by suppressing secretory granule release.
In our experimental conditions, XML inhibited collagen-, thrombin-, and AA-induced phosphorylations of molecules involved in the GPVI-Syk-PLCγ2-MAPK and PI3K-Akt-GSK3β signaling pathways, indicating that XML may exert its inhibitory effects on stimulated platelets through a common signaling cascade.
Our study shows that XML significantly suppresses collagen-, thrombin-, and AA-induced platelet activations by inhibiting phosphorylations of Syk and PLCγ2. In order to further illuminate the underlying mechanism of XML-mediated inhibition of platelet activation, we examined the phosphorylation of MAPKs (ERK, JNK, and p38) and found that XML inhibited collagen-, thrombin-, and AA-induced phosphorylations of ERK1/2, JNK, and p38. Moreover, XML also inhibited collagen-, thrombin-, and AA-induced phosphorylations of Akt and GSK3β, indicating that XML also functions as a repressor of PI3K-Akt-GSK3β signaling to attenuate platelet activation. Therefore, XML acted as inhibitor of both the GPVI-Syk-PLCγ2-MAPK and PI3K-Akt-GSK3β signaling pathways. Furthermore, the additive inhibitory effect of XML and PP2 or LY294002 on collagen-induced platelet aggregation indicates that XML inhibited simultaneously both SFK-initiated signaling and PI3K-based signaling. Our results demonstrated that agonist-induced integrin α II bβ3 "inside-out" signaling was inhibited by XML.
Moreover, fibrinogen binding to activated integrin α II bβ3 triggers "outside-in" signaling events, inducing integrin clustering that leads to cytoskeletal reorganization and post-occupancy events including platelet spreading, platelet aggregation, and clot retraction (Boylan et al., 2008;Vaiyapuri et al., 2013). We employed these well-accepted assays and found that XML robustly suppressed platelet spreading on fibrinogen-coated surfaces and clot retraction, indicating that XML can block integrin α II bβ3mediated "outside-in" signaling.
We extended the study to the in vivo setting and examined the effect of XML on thrombus formation using an acute pulmonary thrombosis mouse model, which involves the formation of thrombi in the lung vasculature. We found that XML-treated mice were protected from lung vascular thrombosis. Although XML showed anti-platelet and anti-thrombotic effect, it did not prolonged mouse tail-bleeding time; neither PT nor APTT was affected by XML, indicating XML has no effect on the coagulation system. However, our study has limitations. Due to the complex active ingredients of XML and multiple signaling pathways of antiplatelet, we are still far away from fully known the mechanisms of the anti-platelet and anti-thrombotic. Further studies are required to decipher other mechanisms of its function in platelet function and thrombus formation.
In summary, the results of present study demonstrate that XML potently suppresses aggregation and granule secretion induced by several agonists. Additionally, our results show that XML negatively regulates platelet adhesion and spreading on both collagen-and fibrinogen-coated surfaces. Furthermore, we show that XML inhibits platelet activation by inhibiting signaling pathways including the GPVI-Syk-PLCγ2-MAPK and PI3K-Akt-GSK3β pathways, followed by suppression of integrin α II bβ3mediated signaling. Finally, we report that XML protects mice from thrombosis formation in the acute lung thromboembolism model. Collectively, our results suggest that XML may be a potential antiplatelet and anti-thrombotic agent.

DATA AVAILABILITY
All datasets generated for this study are included in the manuscript and the Supplementary Files.

ETHICS STATEMENT
The study protocol was approved by the Ethics Committee of Kunming Medical University in accordance with the Helsinki Declaration for the Use of Human Subjects. All volunteers signed informed consent.
The protocol for the in vivo study with rats was reviewed and approved by the Ethics Committee for experimental animals, Kunming Medical University.