Natural Coumarin Derivative Esculetin Regulates Platelet Activation via Modulating NF-κB Signaling in Cyclic Nucleotide-Independent Manner

Abstract

Esculetin, a natural coumarin derivative, shows exciting biological activities in a variety of cell and animal models. Our recent study demonstrated that esculetin exhibits antiplatelet effects by obstructing the phospholipase C γ2/protein kinase C cascade, hydroxyl radical formation, and Akt activation. In this study, we further examined the involvement of cyclic 3′-5′adenosine monophosphate/, vasodilator-stimulated phosphoprotein (VASP), integrin α_IIbβ₃, and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), since cyclic nucleotides reduce the phosphorylation of VASP and activate NF-κB, subsequently inducing α_IIbβ₃ activation that significantly involves the platelet inhibitory pathways. We found that esculetin (50 and 80 µM) did not significantly affect fibrinogen-induced aggregation of elastase-treated platelets; however, it markedly blocked integrin α_IIbβ₃ activation by interrupting the binding of fluorescein isothiocyanate-labeled PAC-1. In addition, neither ODQ nor SQ22536 significantly reversed esculetin-mediated antiplatelet activity stimulated by collagen. Nitroglycerin and prostaglandin E₁ significantly increased VASP phosphorylation, but esculetin had no effect in this reaction, the values being almost identical with those of normal platelets. Furthermore, esculetin, at its maximum concentration of 80 μM significantly reduced the phosphorylation of IκBα and p65 and reversed IκBα degradation in collagen-induced platelets. These results suggest that the NF-κB-dependent α_IIbβ₃ inhibition of esculetin might represent a novel feedback inhibitory mechanism to regulate platelet functions.

Platelet activation is induced by several compounds, such as adenosine diphosphate (ADP), thromboxane A₂, thrombin, and collagen. Dense granules in the platelets release ADP, which plays a central role in aggregation by increasing platelet activation prompted by numerous factors. ¹ Platelet aggregation is divided into two phases namely inside-out and outside-in pathways. A major secondary messenger molecule, inositol triphosphate (IP₃), together with diacylglycerol, formed during the inside-out pathway increase intracellular calcium concentration and stimulate the elevation of several enzymes such as protein kinase C (PKC). ² The termination of the first phase of platelet aggregation ends with fibrinogen receptor (α_IIbβ₃ integrin) activation and phosphorylation. ² α_IIbβ₃ phosphorylation is enhanced by several enzymes, including c-Src. ³ In contrast, substances like nitric oxide (NO) may adversely control platelet activation. Guanylyl cyclase produced by NO increases cyclic 3′-5′guanosine monophosphate (cGMP), which stops platelet aggregation via either independent-on or dependent-on PKG activity. Several proteins are established as PKG targets such as inositol triphosphate receptor, phosphodiesterase 5, and vasodilator-stimulated phosphoprotein (VASP). ⁴ Besides, NO inhibits ADP-induced platelet adhesion in a cGMP-independent way by reducing α_IIbβ₃ integrin activation. ⁵

Moreover, endothelium-derived prostacyclin (PGI₂) released from the platelets binds to IP receptors which can induce the intracellular accumulation of cyclic 3′-5′adenosine monophosphate (cAMP), ⁶ which initiates AMP-dependent protein kinase (PKA) that phosphorylates VASP at serine residue 157. ⁷ A study has shown that cAMP inhibits platelet aggregation via phosphorylation of VASP following integrin α_IIbβ₃ inactivation. ⁷ Nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) in platelets has been reported to act independently of gene regulation, and, after platelet activation, IκB is phosphorylated and degraded. ⁸ Therefore, compounds that elevate cAMP production, increase VASP phosphorylation, and reduce NF-κB and integrin α_IIbβ₃ activation would be useful for the prevention of platelet-related cardiovascular diseases.

Esculetin, the main active component of Cortex Fraxini, has a similar nucleus to that of warfarin, a well-known antiplatelet drug. This coumarin‐derived antioxidant has anti‐inflammatory, ⁹ anti‐proliferative, ¹⁰ and anti‐tumor ¹¹ activities. Esculetin exerted an anti-lipoperoxidative effect in rats by dampening of isoproterenol-induced myocardial infarction and free radical scavenging properties. ¹² Our recent study showed that esculetin exerts antiplatelet effects by reducing phospholipase Cγ2 (PLCγ2) and the PKC cascade, hydroxyl radical formation, and Akt activation. ¹³ In continuation of this study, here we aimed to investigate the role of NF-κB in the anti-aggregating effects of esculetin in human platelets activated with collagen. We also evaluated the downstream cyclic nucleotides signaling pathway, namely, VASP phosphorylation, and integrin α_IIbβ₃ activation in esculetin treated human washed platelets.

Integrin α_IIbβ₃ classically shows bidirectional signaling. ¹⁴ Agonist stimulated signals result in an upsurge in the affinity of α_IIbβ₃ for extracellular ligands, which, in turn, activates outside-in signaling within the cell initiated by receptor ligation. An initial consequence of α_IIbβ₃outside-in signaling is platelet spreading. The β₃ subunit phosphorylated upon platelet activation via inside-out signaling and the binding affinity to fibrinogen of α_IIbβ₃ increased. ¹⁵ The actions of outside-in signaling, including c-Src phosphorylation and activation, begin after β₃ Tyr773 phosphorylation. In this study, the addition of fibrinogen caused aggregation of elastase-treated platelets and this aggregation was not inhibited by esculetin (50 and 80 µM, Figure 1(b)), indicating that it may not directly interrupt the association between fibrinogen and its receptors on platelet surfaces.

Figure 1

Effect of esculetin on fibrinogen-induced elastase-treated platelet aggregation and α_IIbβ₃ integrin activation in washed human platelets. (a) Chemical structure of esculetin (C₉H₆O₄). (b) The elastase-treated human platelets were preincubated with either esculetin (50 or 80 µM) or 0.1% dimethyl sulfoxide (DMSO) for 1 minute and then fibrinogen (200 µg/mL) was added to trigger platelet aggregation. (c) Washed platelets (3.6  ×  10⁸ cells/mL) were preincubated with either (b) 0.1% DMSO or esculetin (c, 50  µM; d, 80  µM) and fluorescein isothiocyanate (FITC)-PAC-1 (2  µg/mL) for 3  minutes in either the (a) absence or (b–d) presence of collagen (1  µg/mL). ^*** P< 0.001, compared with the resting control; ^### P <  0.001, compared with the 0.1% DMSO-treated group.

The adhesion of fibrinogen to trigger α_IIbβ₃ integrin is the last general pathway in platelet aggregation. Platelet filopodial extension, spreading, aggregation, and granule secretion are known to be regulated by α_IIbβ₃-induced signals. ¹⁶ A study found that the agonists’ thrombin and collagen increased platelet α_IIbβ₃ activation when compared with non-activated platelets. ¹⁶ These authors reported that pre-treatment with BAY 60-2770, a potent sGC activator, significantly a bridged collagen, induced α_IIbβ₃ activation. Here, to examine whether esculetin interrupts integrin α_IIbβ₃ activation, the binding of the fluorescein isothiocyanate (FITC)-conjugated PAC-1 mAb that reacts with the activation-induced conformational epitope of integrin α_IIbβ₃ was analyzed through flow cytometry (Figure 1 (c)). As anticipated, this (PAC-1) showed prominent integrin α_IIbβ₃ activation in the presence of collagen. On the inhibitory effect of esculetin, we found that at 50 and 80 µM, it significantly reduced integrin α_IIbβ₃ activation stimulated by collagen, showing that esculetin may influence the binding of PAC-1 to the activated integrin α_IIbβ₃.

Activation of platelets has been recognized to inhibit cAMP generation and cyclic nucleotide-dependent protein kinase activity. ¹⁷ Hence, the effects of esculetin on platelet cAMP and cGMP levels were examined in this study. When nitroglycerin (NTG) and prostaglandin E₁ (PGE₁) were added to platelets before triggering of aggregation with collagen, they produced substantial inhibition of platelet aggregation, as shown in Figure 2(a). Moreover, ODQ (1H-[1,2,4] oxadiazolo [4,3-a]quinoxalin-1-one, 10 µM), a guanylate cyclase inhibitor and SQ22536 (9-(tetrahydro-2-furanyl)-9H-purin-6-amine, 100 µM), an adenylate cyclase inhibitor, considerably reversed the NTG and PGE₁-mediated inhibition of platelet aggregation stimulated by collagen. In addition, neither ODQ nor SQ22536 significantly reversed esculetin-mediated antiplatelet activity stimulated by collagen. VASP is a substrate of cAMP and cGMP-dependent protein kinases (PKA and PKG), and its stimulation by the increased activity of these kinases prevents platelet activation. ¹⁸ As shown in Figure 2(b), PGE₁ and NTG could significantly increase VASP phosphorylation; however, esculetin did not increase VASP phosphorylation. This result suggested that esculetin inhibited platelet aggregation independently of cAMP/cGMP-mediated VASP phosphorylation.

Figure 2

The influence of cyclic nucleotides on esculetin-mediated platelet aggregation and vasodilator-stimulated phosphoprotein (VASP) phosphorylation. (a) Washed platelets (3.6 × 10⁸ cells/mL) were incubated with 10 µM prostaglandin E₁ (PGE₁), 10 µM nitroglycerin (NTG), and 80 µM esculetin for 3 minutes in the presence of either SQ22536 (100 µM) or ODQ (10 µM), and then collagen (1 µg/mL) was added to trigger platelet aggregation. (b) Washed platelets were preincubated with 10 µM PGE₁, 10 µM NTG, or esculetin (50 or 80 µM) for 3 minutes. ^*** P<0.001, compared with the resting control.

Some part of the catalytic PKA molecules is bound to IκB in an NF-κB–IκB complex. Activation of cells with agonists of NF-κB activity separates NF-κB from IκB, resulting in IκB degradation and release, and cAMP-independent activation of PKA. ¹⁹ The NF-κB complex plays a major role in the differentiation and maturation of megakaryocyte ²⁰ and is also expressed in platelets. ⁸ Collagen activates the PI3K pathway that stimulates PKC and PKB that are elaborated in platelet activation. Conversely, both pathways excite IKK and degradation of an NF-κB–IκB–PKA complex, leading to free active PKA that phosphorylates VASP and other substrates involved in platelet inhibitory pathways. ²¹ Here, we show that collagen significantly induced the phosphorylation of IκBα and p65 at 5 minutes (Figure 3 (a,b)) and also induced IκBα degradation after 10 minutes of stimulation (Figure 3 (c)). However, of the 50 and 80 µM pretreatment with esculetin, the maximal concentration of 80 µM significantly reduced the phosphorylation of IκBα (Figure 4 (a)) and p65 (Figure 4 (b)), and reversed IκBα degradation (Figure 4 (c)) in collagen-induced washed human platelets. These data showed that inhibition of NF-κB by esculetin may interfere with collagen-induced platelet activation and may be a potential target for the treatment of cardiovascular diseases.

Figure 3

Time courses of nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) activation by collagen in washed human platelets. Washed platelets (1.2 × 10⁹ cells/mL) were stimulated by collagen (1 µg/mL) to trigger (a) IκBα phosphorylation, (b) p65 phosphorylation, and (c) IκBα protein degradation for the indicated times (3-10 minutes). ^** P<0.01 and ^*** P< 0.001, compared with the resting group (0 minutes).

Figure 4

Effect of esculetin on nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) signaling pathway in collagen-induced platelet activation. Washed platelets were preincubated with either esculetin (50 µM or 80 µM) or 0.1% dimethyl sulfoxide (DMSO) for 3 minutes and then collagen was added to trigger (a) IκBα phosphorylation, (b) p65 phosphorylation, and (c) IκBα protein degradation. ^* P<0.05 and ^** P<0.01, compared with the resting control. ^# P <  0.05, compared with 0.1% DMSO-treated group.

In conclusion, the results show that platelet activation reduces the phosphorylation of VASP and activates NF-κB, subsequently inducing α_IIbβ₃ activation that is significantly involved in the platelet inhibitory pathways. Esculetin, a coumarin derivative, inhibited collagen stimulated α_IIbβ₃ integrin activation and NF-κB signaling events, independent of cAMP/cGMP. These results suggest that NF-κB-dependent α_IIbβ₃ inhibition of esculetin represents a novel feedback inhibitory mechanism to regulate platelet functions.

Experimental

Materials

Esculetin (Figure 1 (a)), elastase, collagen (type I), heparin, PGE₁, NTG, SQ22536, ODQ, and bovine serum albumin were purchased from Sigma (St. Louis, MO, USA) and NF-κB from Cell Signaling (Beverly, MA, USA). A 0.1% dimethyl sulfoxide (DMSO) solution was used to dissolve esculetin.

Preparation of Platelet Suspension and Platelet Aggregation

The study procedure and concerned subjects were reviewed and approved by the Review Board, Taipei Medical University, Taipei, Taiwan (TMU-JIRB-N201612050). This study conformed to the directives of the Helsinki Declaration. All human volunteers associated with this study were advised to avoid taking any drugs that affect platelet aggregation for at least 14 days before the blood collection. All subjects provided written informed consent. The washed platelet suspensions were prepared as described in our previous study. ²²

Platelet aggregation was determined using a lumi-aggregometer (Payton Associates, Scarborough, ON, Canada), as previously described. ²² The suspensions of platelets (3.6 × 10⁸ cells/mL) were preincubated with either esculetin (50 and 80 µM) or an isovolumetric solvent control (0.1% DMSO) for 3 minutes before agonists were added. The reaction proceeded for 6 minutes, and the level of aggregation was calculated in light transmission units.

Flow Cytometric Analysis of Integrin α_IIbβ₃ Activation

Integrin α_IIbβ₃ activation was detected by flow cytometry by the method described previously in human platelets. ²³ Concisely, esculetin (50 and 80 µM) was added to the washed platelets (3.6 × 10⁸ cells/mL) and subjected to FITC-conjugated PAC-1 mAb (2 µg/mL) for 3 minutes, followed by stimulation with collagen (1 µg/mL) for another 5 minutes. The total incubated suspensions were examined for fluorescein-labeled platelets using a flow cytometer (FACScan system; Becton Dickinson, San Jose, CA, USA).

Immunoblotting

Either esculetin (50 or 80 µM) or 0.1% DMSO was added to washed platelets for 3 minutes, and collagen (1 µg/mL) was subsequently added to induce platelet activation. An equal amount of 80 µg total protein was loaded onto 12% SDS-PAGE, and proteins were separated and electrotransferred to polyvinylidene difluoride membranes using a semidry transfer unit (Bio-Rad, Hercules, CA, USA).

Statistical Analysis

The results are stated as means ± standard error of the means. The differences between the groups were assessed using analysis of variance (ANOVA). If a significant variation was found in the ANOVA results between the group means, the groups were normalized by using the Student-Newman-Keuls method. A P value of <0.05 designated statistical significance.