Flavonoids inhibit platelet function through binding to the thromboxane A2 receptor


José Rivera, Centro Regional de Haemodonación, C/Ronda de Garay s/n, 30003 Murcia, Spain.
Tel.: +34 968341990; fax: +34 968261914; e-mail: jose.rivera@carm.es


Summary. Background: Dietary flavonoids are known for their antiplatelet activity resulting in cardiovascular protection, although the specific mechanisms by which this inhibition occurs has not been fully established. Objective: The aim of this study was to investigate the interaction of nine flavonoids representative of various chemical classes, with platelet responses dependent on thromboxane A2 (TxA2) generation and on receptor antagonism, and to analyze the structural requirements for such effects. Methods: The effect of several types of flavonoids on platelet aggregation, serotonin release, and TxA2 generation was investigated. Competitive radioligand binding assays were used to screen for affinity of these compounds to TxA2 receptors. Results: Flavones (apigenin and luteolin) and isoflavones (genistein) abrogated arachidonic acid and collagen-induced platelet responses, such as aggregation and secretion, with a less substantial effect on TxA2 synthesis. These compounds were identified as specific ligands of the TxA2 receptor in the µmol L−1 range, this effect accounting for antiplatelet effects related to stimulation with those agonists. Tight binding of flavonoids to the human TxA2 receptor relies on structural features such as the presence of the double bond in C2–C3, and a keto group in C4. Conclusions: The inhibition by specific flavonoids of in vitro platelet responses induced by collagen or arachidonic acid seems to be related, to a great extent, to their ability to compete for binding to the TxA2 receptor. Therefore, antagonism of this TxA2 receptor may represent an additional mechanism for the inhibitory effect of these compounds in platelet function.


Platelets are implicated in hemostasis, thrombosis, and inflammatory processes. Upon vascular damage, platelets adhere to exposed subendothelium, become activated, and secrete biologically active ligands including adenosine diphosphate (ADP), serotonin, and thromboxane A2 (TxA2) [1]. TxA2 is a potent vasoconstrictor and platelet agonist that, at micromolar concentrations, cause platelets to change their discoid shape to spiculated spheres, and leads to the recruitment and deposition of circulating cells at the site of injury and to the expansion of the thrombus.

Development of therapeutic agents inhibiting platelet function more effectively and safely is the aim of intense pharmaceutical efforts. Interestingly, some natural compounds consumed regularly in the diet may inhibit platelet activation pathways. This phenomenon may partially explain the French paradox; that is, the observation of a low cardiovascular mortality rate in Mediterranean populations, regardless of a high saturated fat intake, counterbalanced by the ingestion of polyphenols—flavonoids—of wine [2–4].

Flavonoids, a group of polyphenolic compounds identified in plants, have been used in different fields of medicine. These substances contain two benzene rings (A and B) linked through a pyrone ring (C). They are divided into various classes, depending on their molecular substitutes, as flavones, isoflavones, flavonols, flavanones, and flavan-3-ols. All subtypes can also occur as glycoside flavonoids.

Certain dietary flavonoids have been shown ex vivo to inhibit platelet function [5,6]. In vitro, diverse mechanisms of inhibition of platelet signaling pathways have been hypothesized, and some studies have related the flavonoid structure to their inhibitory potential. Thus, flavonoids have been shown to impair enzymes involved in cellular signaling as cyclo-oxygenases and lypoxygenases [7,8], phosphodiesterases [7], tyrosine kinases [9], and phospholipases [10,11]. They have also been postulated to have anticoagulant activity by inhibition of NAD(P)H:quinine acceptor oxidoreductase [12], an enzyme inhibited by oral anticoagulants, or by interfering with phosphatidylserine exposure [13]. Other mechanisms reported for their antiplatelet effects rely on their antioxidant properties [14,15]. During the screening for in vitro biological actions of flavonoids, quercetin has been shown to inhibit collagen-induced responses of platelets through selective blockade of the glycoprotein VI signaling pathway [16].

To provide further insight into this issue, we have investigated the effect of distinct structural types of flavonoids on agonist-induced platelet responses. We have identified a powerful action of these compounds as antagonists of the TxA2 receptor, exposing a probable common mechanism for their inhibitory actions on platelet function. Study of certain structure–activity relationships reveals key chemical substituents on the flavonoid molecule, greatly affecting their affinity for the thromboxane receptor.

Materials and methods


Selection of flavonoids included representative compounds differing in presumed key chemical substituents of the flavonoid molecule, i.e. number of hydroxyl groups, presence of double bond between carbons 2 and 3 (C2–C3), presence of a keto group on C4, methylation, and glycosylation (Fig. 1).

Figure 1.

Structure and key substituents of flavonoids included in the study.

Apigenin, quercetin, catechin, rutin, rhoifolin, dimethyl-apigenin and diosmetin were provided by Furfural Español (Murcia, Spain). Genistein and luteolin were from Sigma-Aldrich Química (Madrid, Spain). Flavonoids were solubilized in dimethylsulfoxide (DMSO) and stored frozen until use. The anti-TxB2 antibody, dextran, and charcoal were also from Sigma. Collagen and arachidonic acid were from Menarini Diagnostics (Florence, Italy). The TxA2 receptor agonist U46619 (9,11-dideoxy-9α,11α-methanoepoxy-prosta-5Z,13E-dien-1-oic acid), and human thrombin were from Calbiochem-Novabiochem AG (Lucerne, Switzerland). The TxA2 receptor agonist I-BOP ([1S-[1α,2α(Z),3β(1E,3S*),4α]]-7-[3-[3-hydroxy-4-(4-iodophenoxy)-1-butenyl]-7-oxabicyclo [2,2,1]hept-2-yl]-5-heptenoic acid), its iodine labeled form (125I-BOP; specific activity ∼2000 Ci mmol−1, according to the manufacturer's instructions), and the stable synthetic TxA2 receptor antagonist SQ29548 ([1S-[1α,2α(Z),3α,4α]]-7-[3-[[2-[(phenylamino)carbonyl]hydrazino]methyl]-7-oxabicyclo [2,2,1]hept-2-yl]-5-heptenoic acid) were purchased from Cayman Chemical (Ann Arbor, MI, USA). The 5-OH-14C-tryptamine and [3H]-TxB2 were from Amersham Biosciences (Little Chalfont, UK).

Isolation of platelets

To participate in this study, informed consent was obtained from volunteer blood donors, according to the ethical standards of the committee of our institution. Whole blood (450 mL) was collected in conventional triple blood collection systems (Laboratorios Grifols SA, Barcelona, Spain), and platelet-rich-plasma (PRP)-derived platelet concentrates (PCs) were obtained by stepwise centrifugation as described in detail elsewhere [17]. In experiments comparing the effect of the different flavonoids, freshly prepared PCs were washed free of plasma components twice with Ca+2 free Tyrode's buffer (137 mm NaCl, 2.9 mm KCl, 12 mm NaHCO3, 0.42 mm Na2HPO4, 2 mm MgCl2, 5.5 mm glucose, and apyrase 0.5 UI/mL, pH 6.5), according to the method described by Mustard et al. [18], with modifications. Washed platelets for aggregation and secretion studies were resuspended in Tyrode's solution containing 0.35% human serum albumin and 2 mm CaCl2, pH 7.4. For binding studies, washed platelets were in a buffer containing 140 mm NaCl, 5 mm KCl, 10 mm HEPES, 1 mm MgCl2, 10 mm glucose, 2 mm EDTA, pH 7.3 (I-BOP buffer). To assess the effect of flavonoids in the responses of platelets to agonist activation, cells were incubated (15 min, room temperature) before stimulation with variable concentration of flavonoids in DMSO (final concentration < 0.5%).

Aggregation studies

Washed platelets (300 × 109 L−1) incubated with increasing concentrations of flavonoids or with DMSO alone were stimulated with collagen (5 µg mL−1), arachidonic acid (125 µmol L−1) or the TxA2 mimetic U46619 (2 µmol L−1). The aggregation response was followed by an aggregometer (Aggrecorder II, Menarini) as the percentage of light transmission, with washed platelets as the baseline and Tyrode's buffer as 100%. Dose-dependent inhibition curves were analyzed with a non-linear curve-fitting package (Ultrafit, Biosoft, Cambridge, UK), to establish the concentration of flavonoid necessary to obtain half-maximal inhibition of aggregation (IC50).

In some experiments designed to assess the reversibility of effect, platelets were incubated with tested flavonoids at 100 µmol L−1 and then washed and resuspended in Tyrode's buffer. In parallel, the same platelet samples were processed similarly but without flavonoid removal. The aggregation response of these samples to arachidonic acid was then evaluated as above.

Serotonin release

Release of 14C-serotonin was performed as described elsewhere [19]. Basically, washed platelets (300 × 109 L−1) were radiolabeled with 1 µmol L−1 5-OH-14C-triptamine (45 min at 37 °C; 5 µmol L−1 imipramine was added to prevent reuptake of secreted serotonin). Labelled platelets were treated with flavonoids, at a concentration intended for inhibit by at least 50% the aggregating caused by 125 µmol L−1 arachidonic acid or 5 µg mL−1 collagen. The 14C-serotonin secretion was then induced with these agonists for 5 min at 37 °C under stirring conditions (1000 r.p.m.). The release reaction was stopped with 1/6 volume of EDTA 0.05 mol L−1 formaldehyde 0.633 mol L−1, and platelets were pelleted. The radioactivity in the supernatants was measured by a liquid scintillation counter Wallac 1409 (Wallac Oy, AG&G Company, Turku, Finland), and reported as the percentage of the total radioactivity incorporated, after correction for background.

Measurement of TxA2 synthesis

To assess a potential effect of flavonoids on agonist-induced TxA2 synthesis, platelets were preincubated with flavonoids at a concentration similar to that on serotonin release assays. The TxB2 levels were determined in samples stirred on the aggregometer (5 min, 37 °C) upon stimulation with 7.5 µg mL−1 collagen or 125 µmol L−1 arachidonic acid. Generation of TxA2 was stopped by adding 1/6 volume of 0.55 mmol L−1 indomethacin plus 22 mmol L−1 EDTA; samples were centrifuged (2 min at 14 000 g) and supernatants were collected. Levels of TxB2 were measured as described previously [20] by using a dextran–charcoal radioimmunoassay with 3H-TxB2 as the radioactive tracer. Radioactivity was counted by liquid scintillation spectrometry, and results were reported in nanograms of TxB2/108 platelets.

125I-BOP binding assays

Experiments of binding of 125I-BOP to platelets were carried out essentially as described previously [21]. Briefly, platelets (100 × 109 L−1) were incubated with 7.5 pmol L−1 125I-BOP in the presence of increasing concentrations of unlabeled I-BOP, U46619, SQ29548, or flavonoids in a final volume of 1 mL of I-BOP buffer containing 2 mmol L−1 EDTA to prevent aggregation. Non-specific binding was determined in the presence of 10 µmol L−1 I-BOP. After incubation at room temperature for 15 min, platelet-bound and free ligands were separated by centrifugation (12 000 g, 5 min). The supernatants were aspirated, the platelet pellets rinsed in I-BOP buffer, and the amount of platelet-bound 125I-BOP was determined on a gamma counter (LKB, Multigamma Pharmacia, Uppsala, Sweden). The equilibrium binding data for I-BOP were fitted to a single class of sites, using the ‘Cold’ option of the computer program ligand (Elsevier-Biosoft, Cambridge, UK). This computer analysis provides both the number of binding sites per platelet (Bmax) and the affinity constant of the ligand for these sites (Kd). The displacement data of U46619, SQ29548 or the different flavonoids were also analyzed with ligand but using the ‘Drug’ option, which provides the corresponding inhibition or affinity constants (Ki) for each compound (i.e. the concentration of competing ligand that will bind to half the I-BOP binding sites at equilibrium). In kinetic experiments assessing the ability of selected flavonoids to dissociate the binding of 125I-BOP, platelets were first incubated at equilibrium with 7.5 pmol L−1 125I-BOP, and then an excess of apigenin (20 µmol L−1), genistein (20 µmol L−1), or I-BOP (0.3 µmol L−1) was added. Platelet-bound ligand was separated from free compound and plotted against time.

Statistical analysis

Results are reported, unless stated specifically, as mean ± SD from three experiments conducted on different platelet samples. Statistical comparisons of untreated cells and treated platelets with the distinct flavonoids, or with other test compounds, were achieved by two-tailed paired t-test using Prism for Windows version 2.0 (GraphPad Inc., San Diego, CA, USA). When required, log transformation of data was performed to achieve normal distribution. Differences were considered significant when P < 0.05.


Platelet aggregations studies

Because we aimed to investigate the effect of flavonoids on the platelet TxA2 pathway we first assessed, under experimental conditions, the role of TxA2 generation and signaling through the TxA2 receptor on the aggregating response to various agonists. As shown in Table 1, platelets incubated with COX-1 inhibitors (aspirin or indomethacin) or with antagonists of the TxA2 receptor (SQ29548) aggregated normally upon stimulation with thrombin. By contrast, the response to arachidonic acid and collagen was greatly impaired by both inhibition of TxA2 generation, and by blockade of the TxA2 receptor with SQ29548. As expected, the aggregating response to U46619 and I-BOP was essentially dependent on the binding of the analogs to its receptor (Table 1).

Table 1.  Role of the TxA2 pathway on the aggregating response of washed platelets induced by various agonists
AgonistBufferAspirin (1.1 mmol L−1)Indomethacin 50 µmol L−1SQ29548 0.5 µmol L−1
  1. Inhibition of COX-1 (aspirin and indomethacin) or blockade of the TxA2 receptor (SQ29548) caused no variation on the aggregation to thrombin. By contrast, these agents impaired the response to both collagen and arachidonic acid. While COX-1 inhibition did not modify the aggregation to TxA2 analogs (I-BOP and U46619), such a response was abolished by the presence of the TxA2 receptor antagonist SQ29548.

Thrombin, 0.5 U mL−181.8 ± 0.380.2 ± 6.079.8 ± 3.880.3 ± 3.5
Arachidonic acid, 125 µmol L−175.8 ± 2.85.8 ± 2.35.3 ± 5.31.5 ± 1.5
Collagen, 5 µg mL−174.5 ± 1.019.3 ± 3.219.7 ± 2.021.0 ± 0.9
I-BOP, 0.2 µmol L−174.2 ± 6.473.8 ± 5.074.5 ± 4.11.8 ± 1.5
U46619, 2 µmol L−175.3 ± 4.873.8 ± 4.376.0 ± 2.62.2 ± 1.0

We then assessed the effect of flavonoids in platelet aggregation induced by agonists that were found to be dependent on the TxA2 pathway (collagen, arachidonic acid, and the TxA2 receptor agonist U46619). We observed that short incubation of platelets with certain flavonoids caused a concentration-dependent inhibition of aggregation. By contrast, tested flavonoids had no apparent effect on the thrombin-induced responses. Figure 2 is representative of these experiments, while the IC50 values for each flavonoid (performed for 5 µg mL−1 collagen, 125 µmol L−1 arachidonic acid, 2 µmol L−1 U46619, and 0.5 U mL−1 thrombin-stimulated platelets) are in Table 2. Both apigenin and genistein behaved as strong inhibitors of agonist-induced aggregation, with concentrations below 50 µmol L−1 displaying half-maximal impairment. Luteolin, quercetin and, to a lesser extent, catechin also appeared as moderate inhibitors, while rhoifolin and rutin, the glycosylated counterparts of apigenin and quercetin, exhibited a very mild or negligible effect, displaying IC50 values in the millimolar range. We also found that diosmetin, the 4′-methoxylated luteolin, behaved as a strong inhibitor of collagen responses but not for arachidonic acid, which may suggest certain specificity of this compound for TxA2-independent collagen activation pathways in platelets. Dimethylapigenin, the 7, 4′ dimethylated counterpart of apigenin, could be tested only up to a water-soluble concentration of 25 µmol L−1, and showed no inhibition within this range.

Figure 2.

Inhibition of platelet aggregation by flavonoids (IC50 in µmol L−1).This figure shows a representative example of interference of platelet aggregation by some flavonoids. Washed platelets were stimulated with arachidonic acid in absence or presence of flavonoids, and aggregation was recorded as stated in Methods.

Table 2.  Dose-dependent inhibition of platelet aggregation by flavonoids (µmol L−1 IC50)
 Inhibition of platelet aggregation (IC50, µmol L−1)
Collagen, 5 µg mL−1Arachidonic acid, 125 µmol L−1U46619, 2 µmol L−1Thrombin, 0.5 U mL−1
  1. Washed platelets (300 × 109 L−1) were incubated with increasing concentrations of flavonoids, before agonist-induced aggregation. Inhibition curves were analyzed to determine the IC50, i.e. the dose of flavonoid inhibiting aggregation by one-half. Methylated flavonoids (dimethylapigenin and diosmetin) could be tested only up to the highest water-soluble concentration.

Apigenin9.3 ± 0.934.6 ± 13.423.8 ± 5.9> 100
Genistein11.9 ± 1.524.5 ± 7.024.7 ± 10.9> 100
Luteolin40.0 ± 13.356.9 ± 20.847.0 ± 13.3> 200
Quercetin58.6 ± 6.560.6 ± 16.874.5 ± 14.4> 200
Catechin98.4 ± 44.0162.4 ± 6.7233.4 ± 42.2> 600
Rhoifolin425.1 ± 224.6> 1000398.0 ± 165.4> 1000
Rutin1000.3 ± 88.5> 1000> 1000> 1000
Dimethylapigenin> 25> 25> 25> 25
Diosmetin44.4 ± 26.8> 200112.0 ± 38.3> 200

Remarkably we found that, contrary to aspirin, the anti-aggregating effect of these flavonoids was not irreversible, as removal of these compounds from the treated platelet suspension fully restored the capacity of platelets to aggregate (data not shown).

Effects of flavonoids on serotonin release

As reported in Table 3, when flavonoids were added at a concentration intended for inhibition by at least 50%, the arachidonic acid-induced platelet aggregation (all flavonoids except for glycosylated and methylated compounds), a strong inhibition of serotonin release was observed. Similarly, flavonoids, at concentrations inhibiting by at least one-half the collagen-induced aggregation (all tested compounds except for dimethylapigenin), and even the methylated compound dimethylapigenin, induced a significant reduction on serotonin secretion.

Table 3.  Effects of flavonoids on 14C-serotonin release
 % Maximal 14C-serotonin secretion
Collagen (5 µg mL−1)Arachidonic acid (125 µmol L−1)
  1. Platelets loaded with 14C-serotonin were incubated with flavonoids at doses inhibiting aggregation by 50%, except for glycosylated (rhoifolin and rutin) and methylated (dimethylapigenin and diosmetin) flavonoids, and then stimulated under stirring conditions. Released radioactivity was measured in supernatants by liquid scintillation counting. Data are percentage secretion vs. the maximum release (samples without flavonoids). Serotonin release values (% vs. total radioactivity incorporated) in the samples without flavonoids upon stimulation with 5 µg mL−1 collagen and 125 µmol L−1 arachidonic acid, were 45.0% ± 9.3% and 82.1% ± 12.4%, respectively. *P < 0.05 vs. the 14C-serotonin release induced by each agonist in untreated platelets.

ASA 1 mmol L−141.6 ± 3.4*6.3 ± 3.1*
SQ29548 0.5 µmol L−133.1 ± 2.3*0.1 ± 0.1*
Apigenin 100 µmol L−134.6 ± 3.4*8.0 ± 1.6*
Genistein 100 µmol L−142.5 ± 3.3*6.5 ± 2.4*
Luteolin 100 µmol L−135.1 ± 6.4*8.0 ± 2.1*
Quercetin 200 µmol L−143.3 ± 8.9*13.9 ± 2.6*
Catechin 500 µmol L−138.3 ± 1.1*7.2 ± 1.3*
Rhoifolin 1 mmol L−150.0 ± 2.2*74.1 ± 23.1
Rutin 1 mmol L−189.1 ± 2.8*93.0 ± 8.0
Dimethylapigenin 25 µmol L−170.4 ± 10.9*106.7 ± 9.6
Diosmetin 50 µmol L−169.3 ± 9.5*94.1 ± 6.2

Effects of flavonoids on TxA2 generation

As the data above show that some flavonoids impaired platelet aggregation and secretion to agonists with responses highly dependent on TxA2 pathways, and as it has been suggested previously that flavonoids may inhibit enzymes of the arachidonic acid pathway, we investigated whether they influence the agonist-induced platelet synthesis of TxA2. While U46619 and I-BOP did not induce substantial generation of this metabolite (below 1 ng/108 platelets; data not shown), collagen and arachidonic acid caused release of TxB2, a stable metabolite of TxA2. As shown in Table 4, to a great extent aspirin abrogates the TxA2 synthesis induced by these two agonists.

Table 4.  Effect of flavonoids on TxB2 generation
 TxB2 generation (ng/108 platelets)
Collagen (7.5 µg mL−1)Arachidonic acid 125 µmol L−1
  1. Values are given in ng of TxB2/108 platelets. In all cases but glycoside and methylated flavonoids, platelet aggregation was assessed, and was found to be inhibited by more than 50% by the flavonoid. Aspirin was included as a negative control. *P < 0.05. **P < 0.005 vs. values for untreated platelet (0.5% DMSO).

0.2% DMSO16.8 ± 3.5152.7 ± 85.6
1 mm ASA0.2 ± 0.1**2.8 ± 3.5*
Apigenin, 100 µmol L−17.6 ± 3.9*236.4 ± 104.4
Genistein, 100 µmol L−17.9 ± 3.8*175.4 ± 98.4
Luteolin, 100 µmol L−110.1 ± 10.0116.02 ± 30.6
Quercetin, 200 µmol L−13.7 ± 0.8**55.4 ± 37.5
Catechin, 500 µmol L−13.2 ± 0.9**54.7 ± 27.1
Rhoifolin, 1 mmol L−15.4 ± 0.9**222.2 ± 107.7
Rutin, 1 mmol L−111.8 ± 2.8237.5 ± 134.4
Dimethylapigenin, 25 µmol L−111.5 ± 1.4134.4 ± 4.3
Diosmetin, 50 µmol L−16.9 ± 0.9138.8 ± 17.4

Flavonoids, that severely impair the aggregating response to arachidonic acid, did not significantly diminish the agonist-induced synthesis of TxA2. Indeed, only quercetin and catechin decreased mildly the arachidonic acid-induced generation of TxA2. In contrast, to a great extent most flavonoids tested diminished collagen-induced TxA2 synthesis (Table 4). Thus, quercetin and catechin seem to inhibit TxA2 synthesis irrespective of the agonist employed, possibly by a primary effect at the COX-1 or thromboxane synthase level, as has been suggested recently [22].

Effects of flavonoids on 125I-BOP binding

The fact that some flavonoids inhibited platelet aggregation and secretion induced by TxA2 analogs or by agonists using the platelet TxA2 amplification pathway, despite no apparent reduction of the agonist-promoted TxA2 synthesis, prompted us to investigate the potential interaction of flavonoids at the TxA2 receptor level. For that purpose, a cold saturation binding assay was set up in washed platelets by examining the reduction of 125I-BOP binding with increasing concentrations of unlabeled I-BOP. Under our experimental conditions, ligand analysis demonstrated that equilibrium binding data best fitted to a single class of sites, with an apparent Kd of 2.84 ± 0.4 nmol L−1, and a density (Bmax) of 2451 ± 526 sites/platelet. As illustrated in Fig. 3, other well-established TxA2 receptor ligands such as SQ29548 and U46619 were confirmed to be efficient competitors of 125I-BOP binding to washed platelets. In addition, several flavonoids also competed for this binding in a dose-dependent manner. In Table 5 their correspondent Ki values, or affinity/inhibition constant for the I-BOP binding sites, are specified. As shown, apigenin, luteolin, and genistein and, to a lesser extent, quercetin, competed with high affinity for TxA2 binding sites, in good agreement with their effect on the inhibition of platelet aggregation and secretion. However, compared to I-BOP, U46619 or the TxA2 receptor antagonist SQ29548, these flavonoids are still more than 1000-fold less effective in competing with I-BOP for their binding to platelets. Moreover, glycosylated flavonoids (rhoifolin and rutin) have significantly reduced affinity for the TxA2 receptor compared to their non-glycosylated counterparts (apigenin and quercetin). In addition the methylated flavonoids, up to their achievable water soluble concentration (25–200 µmol L−1), displayed no apparent competing effect (data not shown). In kinetics assays we further observed that some flavonoids not only competed for 125I-BOP binding to platelets, but in appropriate excess they were also able to dissociate previously bound ligand (Fig. 4).

Figure 3.

Effect of flavonoids on 125I-BOP binding to platelets. Washed platelets were incubated with 125I-BOP (7.5 pmol L−1) in the presence of increasing concentrations of unlabeled counterpart, other thromboxane A2 receptor ligands (SQ 29585 and U46619), or some flavonoids as competitors. The plot is a representative example of the resultant dose-dependent displacement curves of the specific binding.

Table 5.  Dose-dependent inhibition of 125I-BOP binding to washed platelets by ligand analogs or by flavonoids
 Ki (µmol L−1)
  1. Platelets were incubated with 7.5 pmol L−1125I-BOP, in absence or presence of other TxA2 receptor ligands or flavonoids. The Ki value defines the concentration of competing ligand that binds to half the binding sites at equilibrium in the absence of radioligand or of other competitors. Lower Ki values correspond to higher affinity of compounds for the 125I-BOP binding sites in platelets. The Ki values were derived from inhibition curves analyzed with ligand software (drug option). Results are mean ± SD from at least two experiments with different platelets.

U466190.127 ± 0.007
SQ295480.0040 ± 0.012
Apigenin4.0 ± 2.7
Genistein2.1 ± 0.7
Luteolin3.1 ± 0.6
Quercetin35.6 ± 29.2
Catechin160.5 ± 34.6
Rhoifolin119.3 ± 27.1
Rutin227.0 ± 115.9
Figure 4.

Dissociation of 125I-BOP binding to platelets by flavonoids. Washed human platelets were incubated with 125I-BOP. At the time indicated by the arrow, either buffer (•), or an excess of unlabeled I-BOP (□), 20 µmol L−1 genistein (bsl00066), or 20 µmol L−1 apigenin (bsl00001) were added. At selected time-points platelet bound 125I-BOP was separated from free ligand by centrifugation, counted and plotted against time.


Clinical evidence suggests that inhibition of platelet TxA2 signaling pathway provides a therapeutic basis for the treatment and/or prevention of certain thrombotic disease states [23]. Indeed, the ability of aspirin to inhibit TxA2 synthesis is the primary rationale for its widespread use in recurrent myocardial infarction and thromboembolic stroke. Although there is interest in the development of antagonists of TxA2 synthase inhibitors and/or TxA2 receptor antagonists, to date there is only in vitro evidence of their effect [24,25], and data on their antithrombotic efficacy in clinical settings are still lacking.

Flavonoids have received attention in medicinal use because they are elements in reducing cardiovascular risks. While some flavonoids inhibit platelet aggregation, others do not. Regardless of their potential effects on signaling pathways [7–9], the binding of flavonoids to cell receptors has also been postulated as possible mechanisms for their antithrombotic activity. Thus, flavones have been suggested to block both adenosine receptors [26] and von Willebrand factor binding to platelet glycoprotein Ibα[27]. Red wine inhibits the binding of platelet-derived growth factor (PDGF) to its receptor in vascular smooth muscle cells [28]. Of interest, genistein inhibits the binding of the TxA2-analog [3H]-U46619 to washed platelets [29,30], and this dietary isoflavone inhibits in vivo the progression of atherosclerosis in cholesterol-fed rabbits [31].

In this study we have examined the pharmacological profile of representative flavonoids in terms of their antiplatelet effects. From the platelet aggregation and secretion studies, three flavonoids, apigenin, genistein, and luteolin, effectively inhibited TxA2-mediated responses. Thus, these three flavonoids abrogate aggregation and dense granule secretion elicited by arachidonic acid, resulting in an inhibition of reactivity similar to that caused by aspirin. This effect is, however, reversible and independent of an impaired TxA2 synthesis, suggesting minor effects on cyclo-oxygenase or thromboxane synthase. Furthermore, these flavonoids displace 125I-BOP binding to platelets, consistent with these agents acting as competitors for the TxA2 receptor. It is noteworthy that there is agreement between the strength of these flavonoids in interfering with radioligand binding to TxA2 receptors and their inhibition of platelet aggregation and secretion, thus suggesting a direct relationship between these two effects. Thus, our results confirm the antiplatelet action of certain flavonoids, which selectively inhibit the amplification action of TxA2 at the receptor level.

Much work has been conducted to elucidate domains on the TxA2 receptor critical for ligand binding [32]. However, modeling studies describing the spatial structure of the receptor binding pocket, which would facilitate our understanding of interactions with compounds such as flavonoids, are still lacking.

Our findings, that apigenin, genistein, and luteolin interact with the TxA2 receptor with high affinity, provide evidence that not all flavonoids have similar pharmacological actions. The interaction of selective flavonoids with the TxA2 receptor, a member of the G protein-coupled receptor superfamily, suggests that these three compounds might share antiplatelet actions that are dependent on a similar chemical structure. The study of certain structural characteristics reveals that flavones (apigenin and luteolin) and isoflavones (genistein) with the higher affinity for the TxA2 receptor (lower Ki) share a characteristic conjugation between A, C and B rings with the presence of a lactone structure. Of note is that this lactone-like structure is also present within the TxA2 molecule, which may account for competition between these compounds for the same receptors. Tight binding of flavonoids to the TxA2 receptor also seems influenced by certain features affecting their steric and electrostatic properties. Indeed, glycosylation (rhoifolin and rutin) might enlarge the flavonoid size, thus complicating binding to the receptor, while methylation (dimethylapigenin and diosmetin) might decrease affinity for TxA2 receptors due to change in the electrical charge of the drug. Moreover, the double bond in C2–C3 and/or keto group in C4 are considered important for the binding to the TxA2 receptor, as the lack of such ring substituents (catechin) results in a decreased affinity of these compounds for the receptor.

Thromboxane production, together with ADP generation, is necessary for platelet aggregation induced by low doses of collagen [33]. Consistent with the data above, the three flavonoids exhibiting the highest affinity for the TxA2 receptor (apigenin, genistein, and luteolin) efficiently inhibited collagen induced aggregation. Nevertheless, flavonoids induced a mild decrease, similar to aspirin, in the secretion from platelet dense granules upon activation with collagen, and these compounds also reduced the extent of TxA2 secretion in collagen-stimulated platelets. This is in agreement with previous observations indicating a requirement of TxA2 receptor signaling for aggregation induced by low-dose collagen-induced aggregation; that signaling being, however, only partially required for secretion and TxA2 production but dependent upon αIIbβ3-mediated outside–in signaling [33]. Moreover, the fact that the effect of flavonoids that exhibit higher affinity for the TxA2 receptor in the inhibition of aggregation is stronger in collagen than in arachidonic acid and U46619-stimulated platelets suggests that TxA2 receptor blockade by flavonoids seems to be an additional mechanism to those that have been proposed, through which some of these compounds might inhibit the collagen signaling pathway. Hence, those flavonoids are also inhibitors of tyrosine kinases [16,34,35], an important pathway for collagen signaling leading to platelet aggregation [36], suggesting that this mechanism may also play a role in specificity of inhibition by collagen. Moreover, a methylated flavonoid, diosmetin, seems to specifically inhibit collagen-induced aggregation and secretion, independently of any effect on the TxA2 receptor blockade. The earlier observation, that methylation of flavones increases affinities for adenosine receptors [26], puts forward the hypothesis of additional antagonism of this G protein-coupled receptor as being responsible for this inhibitory effect.

In conclusion, some flavonoids have been found to have biological effects attributable to TxA2 receptor antagonism. Key elements on the flavonoid molecule have been identified as being important to exert this effect. While many of the effects attributed to flavonoids have unexplained mechanisms, our results support that a certain ability of these compounds on their antiplatelet effect is receptor antagonism, mainly on thromboxane receptors. Further studies are expected to analyze the probable capacity of flavonoids to bind to other platelet receptors and determine their activity in the presence of plasma proteins to conclude whether pharmacological supplementation or dietary intake of these compounds might prompt beneficial effects in the prevention or treatment of thrombotic events.


The authors would like to thank Jonathan Gibbins (University of Reading, UK) for helpful discussion and critical reading of this manuscript. This study was supported in part by a research grant from the Fondo de Investigaciones Sanitarias (02/1338).