A Naphthalenic Derivative ND-1 Inhibits Thrombus Formation by Interfering the Binding of Fibrinogen to Integrin αIIbβ3

Integrin αIIbβ3 plays a crucial role in the process of platelet aggregation. Three integrin αIIbβ3 antagonists (abciximab, eptifibatide, and tirofiban) have been approved by FDA for clinical use. Unfortunately, they all showed severe side effects such as thrombocytopenia and bleeding risk. Thus, researches on the development of more effective and safer antiplatelet agents are needed. In this manuscript we reported a novel naphthalenic derivative compound ND-1 with potent antithrombotic effect and lower bleeding risk. ND-1 inhibited ADP-, collagen-, thrombin-, and U46619-induced platelet aggregation with IC50 values of 1.29, 14.46, 12.84, and 40.24 μM, respectively. Mechanism studies indicated that ND-1 inhibited the binding of fibrinogen to integrin αIIbβ3 in a dose-dependent manner with an IC50 value of 3.12 μM. ND-1 inhibited P-selectin expression induced by ADP, collagen, thrombin, and U46619 on the surface of platelets. Additionally, this compound reduced platelets spreading to the immobilized fibrinogen. In vivo, ND-1 potently decreased thrombus formation in an arteriovenous shunt thrombosis model in rats and slightly prolonged bleeding time in a tail cutting model in mice. Taken together, our results reveal that ND-1 is a novel antagonist of αIIbβ3 with strong antithrombotic effect and lower bleeding risk.


Introduction
Cardiovascular diseases (CVDs), including acute coronary syndrome, myocardial infarction, deep-vein thrombosis, and pulmonary embolism, are the leading cause of morbidity and mortality worldwide [1]. The proximal cause of CVDs is intravascular thrombus formation [2,3]. In the process of thrombosis, platelets play an important role as they can adhere to the exposed subendothelial matrix and then change shape and release ADP, thromboxane A2, thrombin, and epinephrine, which recruit additional platelets from the blood flow to the injury sites and form plugs [4,5]. Accordingly, antiplatelet agents have been widely developed as an important tool for preventing thrombotic events [6].
Integrin IIb 3 is a crucial molecule in the process of platelet aggregation. It has extracellular region for ligands binding and intracytoplasmic tail mediating intracellular signal transduction. Both inside-out and outside-in signaling in platelet involve this molecule [7,8]. So the activation of integrin IIb 3 was recognized as the final common pathway of platelet aggregation [9,10]. Three integrin IIb 3 antagonists (abciximab, eptifibatide, and tirofiban) have been approved by FDA for clinical use. Unfortunately, they all showed severe side effects such as thrombocytopenia and bleeding risk [11][12][13][14]. Thus, studies on the development of more effective and safer antiplatelet agents are needed.
In the search for new antiplatelet agents, we screened more than 500 synthetic compounds and natural products and found that ND-1 (Figure 1(a)), a compound derived from naphthalene, had potent antiplatelet aggregation effect. The aim of this study was to investigate the effect of ND-1 on thrombus formation, as well as the underlying mechanisms. to (e)) Effect of ND-1 on platelet aggregation. Platelet-rich plasma was preincubated with indicated concentrations of ND-1 or vehicle for 5 min at 37 ∘ C. Aggregation was initiated by the addition of 20 M ADP (b), 1 g/mL collagen (c), 0.25 U/mL thrombin (d), or 2 M U46619 (e). Data are presented as mean ± SD ( = 3). * * < 0.01, * * * * < 0.0001 compared with vehicle.

Animals.
Institute of Cancer Research (ICR) mice (male, 18-22 g) and Sprague-Dawley rats (male, 180-220 g) were purchased from Nanjing Qinglongshan Animal Center (Nanjing, Jiangsu, China). All experiments were carried out in accordance with the guidelines and the regulations of the Ethical Committee of China Pharmaceutical University. The protocols were approved by the Institutional Animal Care and Use Committee of China Pharmaceutical University.

Human Blood Samples and Washed Platelet Preparation.
Human platelet-rich plasma (PRP) was obtained from Jiangsu Province Blood Center. Washed platelets were prepared as described previously [15]. Briefly, PRP containing 5 mM EDTA, 2 U/mL apyrase, and 0.1 g/mL PGE1 was centrifuged at 1100 ×g for 10 min, and then pellets were resuspended in Tyrode buffer and the concentration of platelets was adjusted to approximately 3 × 10 7 platelets/mL.

Platelet Aggregation Assay.
Platelet aggregation was measured as previously described using a four-channel aggregometer (LBY-NJ4, Pulisheng Science Instrument Company, Beijing, China) [9,16]. Briefly, PRP was centrifuged at 1580 ×g for 10 min to obtain platelet-poor plasma (PPP). PRP was preincubated with or without ND-1 for 5 min at 37 ∘ C, and aggregation was induced by the addition of ADP (20 M), collagen (1 g/mL), thrombin (0.25 U/mL), or U46619 (2 M). The maximum platelet aggregation rate was determined within 5 min by continuously stirring.

Fibrinogen/Integrin IIb 3 ELISA.
Similar to the previous method [17,18], the 96-well microplate (Corning Incorporated, Corning, NY, USA) was coated with human fibrinogen (10 g/mL) at 4 ∘ C overnight. After washing with TACTS (20 mM Tris, 0.15 M NaCl, 2 mM CaCl 2 , and 0.05% Tween 20, pH 7.5), wells were blocked with 1% BSA in TACTS for 1 h at 37 ∘ C and then washed with TACTS three times. Purified human platelet integrin IIb 3 receptor (20 g/mL) and ND-1 or vehicle were added and incubated for 2 h at 37 ∘ C. Wells were washed and then mouse anti-human integrin 3 antibody (1 : 2000) was added. After incubation for 1 h at 37 ∘ C, wells were washed and anti-mouse IgG-conjugated alkaline phosphatase (1 : 2000) was added and incubated for 1 h at 37 ∘ C. Disodium 4-nitrophenyl phosphate substrate was added and incubated for 30 min, and the reaction was stopped by adding 3 M NaOH. The OD values of samples were estimated at 405 nm.

P-Selectin Expression on Platelets.
The expression of P-selectin on the surface of platelets was studied using a flow cytometric method as described previously with some modifications [19,20]. Briefly, PRP was incubated with or without compound ND-1 for 10 min at 37 ∘ C. The platelets were then separately stimulated with ADP (20 M), collagen (1 g/mL), thrombin (0.25 U/mL), and U46619 (2 M) at 37 ∘ C within 10 min, and then PE-conjugated anti-human CD62P and FITC-conjugated anti-human CD42a were added and incubated for 10 min at 4 ∘ C in the dark. After that platelets were fixed with 1% paraformaldehyde at 4 ∘ C and analyzed with a BD FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA). For each sample, 20000 platelets were analyzed.

Platelet
Spreading on Immobilized Fibrinogen. Platelet spreading was measured as previously described [21,22]. Glass coverslips were coated with 40 g/mL fibrinogen in 0.1 M NaHCO 3 (pH 8.3) at 4 ∘ C overnight. Washed platelets incubated with ND-1 (50, 25, and 10 M) or vehicle for 5 min at 37 ∘ C were allowed to spread on the fibrinogen-coated surfaces for 60 min at 37 ∘ C. After three washes, the platelets were fixed with 4% paraformaldehyde and permeabilized with 0.2% triton X-100 solution and stained with FITCconjugated phalloidin. Adherent platelets were viewed by an upright fluorescent microscope AXIO ScopeA1 (ZEISS Group, Jena, Germany). Three images were chosen randomly per experiment. The spreading area of individual platelet was assessed using ImageJ software (National Institutes of Health, Bethesda, MD).

Arteriovenous Shunt Model in Rats.
The arteriovenous shunt thrombosis in rat was measured with a modified method [23]. Male Sprague-Dawley rats were randomly divided into five groups, six for each group. ND-1 (1, 3, and 7 mg/kg), aspirin (50 mg/kg), or vehicle was injected through caudal vein. After 15 minutes, rats were anesthetized with intraperitoneal 10% chloral hydrate (3 mL/kg). The right jugular vein and the left carotid artery were isolated, and two 4 cm polyethylene 50 (PE50) catheters were inserted separately. The two PE50 catheters were linked by a 12 cm PE60 catheter containing 10 cm long silk thread soaked with saline to induce thrombosis. After 20 min of blood circulation, the silk thread carrying thrombus was pulled out. The weights of the thrombus were measured after the thread was dried for 30 min at 60 ∘ C.

Bleeding Time Assay.
Bleeding time in mice was measured according to previous method [24]. Male ICR mice were randomly divided into five groups, ten for each group. ND-1 (2, 6, and 14 mg/kg), aspirin (50 mg/kg), or vehicle was injected through caudal vein. After 15 min, mice were anesthetized with intraperitoneal 5% chloral hydrate (6 mL/kg), followed by a cut of 3 mm from the tail tip. The tail was immediately immersed into 12 mL saline at 37 ∘ C and the accumulated bleeding time was recorded within 20 min.

Statistical Analysis.
Statistical analysis was performed using GraphPad Prism Statistical Software (version 6.0, San Diego, CA). The data are presented as means ± SD or means ± SEM. Differences between the sample-treated group and vehicle were analyzed by one-way analysis of variance assay (ANOVA) followed by Tukey's multiple comparison test. Statistical significance was considered as < 0.05.

Effect of ND-1 on the Binding of Fibrinogen to Integrin
IIb 3. To verify the inhibition of ND-1 for the binding of fibrinogen to integrin IIb 3, fibrinogen/integrin IIb 3 solid-phase ELISA was performed. As shown in Figure 2(a), ND-1 strongly inhibited the binding of fibrinogen to purified integrin IIb 3 in a dose-dependent manner with an IC 50 value of 3.12 M (95% CI, 2.75-3.56 M).

Effect of ND-1 on Platelet P-Selectin Expression.
P-selectin is a major marker protein of platelet activation, which appears on the surface of platelets during platelet activation. To test whether ND-1 affected platelet activation, the level of P-selectin expression was assessed by flow cytometry. As shown in Figures 2(b)-2(e), ND-1 at 50 M inhibited U46619-, ADP-, collagen-, and thrombin-induced P-selectin expression by 96.55%, 86.61%, 75.53%, and 41.16%, respectively.

Effect of ND-1 on Platelet
Spreading. Platelet spreading on immobilized fibrinogen is dependent on cytoskeletal reorganization driven by integrin IIb 3 outside-in signaling. To investigate the role of ND-1 in the outside-in signaling, the platelets' spreading on immobilized fibrinogen was assessed. The results in Figures 3(a) and 3(b) showed that ND-1 significantly inhibited platelets' spreading in a dose-dependent manner with inhibition rates of 66,06 ± 5.45%, 50.64 ± 4.48%, and 34.55 ± 4.92% at concentrations of 50, 25, and 10 M, respectively. Therefore, ND-1 is involved in the regulation of integrin IIb 3 outside-in signaling.

Antithrombotic Activity of ND-1 In Vivo.
To examine the antithrombotic activity of ND-1, the arteriovenous shunt thrombosis model was used. ND-1 inhibited the formation of thrombus in a dose-dependent manner as shown in Figure 4(a). ND-1 at 7 mg/kg inhibited thrombus formation by 31.84 ± 6.76% ( = 6), whereas aspirin at 50 mg/kg inhibited thrombus formation by 26.32 ± 9.5% ( = 6), indicating that this compound at a low dose showed similar effect to that of aspirin at a high does.

Bleeding Risk of ND-1 In Vivo.
To assess the bleeding risk incurred by ND-1, we measured the bleeding time of ND-1-treated mice by a tail cutting assay at ND-1 doses of 2, 6, and 14 mg/kg. As shown in Figure 4(b), ND-1 at 2 and 6 mg/kg did not significantly prolong bleeding time, compared with vehicle group. But when the dose of ND-1 increased to 14 mg/kg, which represented two times the high dose used in the arteriovenous shunt thrombosis model, it significantly prolonged bleeding time (13.67 ± 2.67 min), while the positive control aspirin prolonged bleeding time more (17.18 ± 1.47 min).

Discussion
In this study, we found that ND-1 inhibited platelet aggregation via blocking the binding of fibrinogen to integrin IIb 3. In vivo, ND-1 potently decreased thrombus formation in a rat arteriovenous shunt thrombosis model and slightly prolonged bleeding times in mice, suggesting that ND-1 is a lead compound for development of cardiovascular drug.
The diverse platelet activation signaling pathways stimulated by various agonists converge into common signaling events, which induce the inside-out signaling transduction, leading to activation of integrin IIb 3, which is a final effector for platelet aggregation [25]. Our studies showed that ND-1 inhibited platelet aggregation induced by various agonists (ADP, collagen, thrombin, and U46619), suggesting a possible inhibition of integrin IIb 3. Fibrinogen/integrin IIb 3 ELISA was conducted and the result showed that ND-1 inhibited the binding of fibrinogen with IIb 3. The activation of platelets leads to degranulation, which results in exposure of platelet granule membranes and their associated antigens. The CD62P (P-selectin), one of -granule membrane antigens, is highly expressed on the activated platelets surface [26]. The present studies showed that ND-1 inhibited platelet activation (indicated by P-selectin expression) induced by ADP, collagen, thrombin, and U46619. These agonists have different receptors on the platelet surface and activate platelets via different pathways [27]. So, we speculated that ND-1 can act on integrin IIb 3, the "final common pathway" of platelet activation. This result further confirms our conclusion that ND-1 affects the binding of fibrinogen to integrin IIb 3.
Integrin IIb 3, a dual-direction effector, not only mediates inside-out signaling but also affects outside-in pathway. Ligands such as fibrinogen bind to integrin IIb 3 and mediate outside-in signaling transduction, resulting in platelet spreading. ND-1 reduced platelet spreading on immobilized fibrinogen, demonstrating its inhibition on integrin IIb 3mediated outside-in signaling.
Three agents targeting IIb 3 have been approved for human use in the USA, starting with abciximab, a monoclonal antibody fragment, followed by eptifibatide, a cyclic peptide, and tirofiban, a peptidomimetic molecule [28]. These drugs have shown efficacy in the adjunctive therapy to ischemic complications of percutaneous coronary interventions in a large number of randomized studies. However, they are involved in an increased bleeding risk and thrombocytopenia, which restrict their use clinically [29]. Thus, numerous compounds have been designed and evaluated to overcome these drawbacks for the past several years. RUC-1 (a small molecular inhibitor of integrin IIb 3) only binds to the IIb subunit but not to 3 and does not induce conformational changes of IIb 3. By avoiding such a change, it is anticipated that associated thrombocytopenia, induction of fibrinogen binding, and paradoxical platelet activation would be prevented [28,30]. RUC-2 and RUC-4 were developed following RUC-1. Currently RUC-4 is under development for the prehospital therapy of ST segment elevated myocardial infarction [31,32]. Ur-3216/2922, binding to IIb D224 and not engaging the MIDAS metal ion, is a high affinity IIb 3 antagonist without ligand-induced binding site (LIBS) expression. In vivo UR-3216 showed antithrombotic effect without bleeding at effective doses [33]. We found that ND-1 potently inhibited thrombus formation in a dose-dependent manner in an arteriovenous shunt thrombosis model in rat and slightly prolonged bleeding time in a tail cutting model in mice.
In conclusion, ND-1 inhibited platelet aggregation via interfering the binding of fibrinogen to integrin IIb 3 and significantly attenuated the thrombus formation with lower bleeding risk. Thus, ND-1 can be a lead compound for antithrombotic agents development.