The present study demonstrates that Dp-3-g, one of the predominantly bioactive compounds of anthocyanins, inhibits platelet activation and aggregation, as well as attenuates thrombosis in vitro
and in vivo
. To the best of our knowledge, this is the first study that clearly demonstrates that an anthocyanin can directly inhibit platelet function, in addition to its potential effects on inflammation 
and lipid metabolism 
that may indirectly alter platelet activities. Compared to the control group, Dp-3-g significantly inhibited platelet aggregation in human and mouse PRP and purified gel-filtered platelets. Dp-3-g also markedly reduced thrombus growth in anticoagulated human and murine blood in perfusion chambers ex vivo
, as well as decreased platelet deposition, destabilized thrombi, and prolonged the time required for thrombus formation and vessel occlusion in intravital microscopy and carotid artery thrombosis models in vivo
. We found that Dp-3-g inhibited platelet activation, which broadly affected platelet α- and δ-granule release and cytosol protein secretion. Dp-3-g decreased active integrin αIIbβ3 expression on the platelet surface, and attenuated the binding of Fg to platelets without directly interfering with αIIbβ3-Fg interaction. Dp-3-g also decreased pAMPK, which may in turn decrease αIIbβ3 activation through its inside-out signaling. Thus, inhibition of platelet activation may be the mechanism that explains the anti-thrombotic effects of Dp-3-g observed in the present study.
The effects of some anthocyanin products on platelet aggregation have been studied in human whole blood and PRP after consumption 
. However, most studies in vivo
have focused on the effects of food containing anthocyanins, instead of pure anthocyanin compounds or its bioactive derivatives, leading to contradictory conclusions regarding their effects on platelet function. In previous studies, it was demonstrated that querecetin did not significantly affect platelet aggregation 
. To date, only one investigation assayed the anti-platelet activity of physiologically relevant concentrations of four purified anthocyanins in vitro
and found that anthocyanins inhibited TRAP-induced platelet aggregation but did not influence platelet reactivity when stimulated by collagen or ADP 
. In the present study, however, we found that the anthocyanin Dp-3-g inhibited platelet aggregation in both human and mouse PRP stimulated not only by TRAP, but also by ADP and collagen. In addition, although there were several studies examining the effect of anthocyanins on platelet function in whole blood and PRP, there is no report describing the direct effects of these chemicals on platelet function. Using gel-filtered platelets, in which other molecules in the blood plasma are removed, we found that the anthocyanin Dp-3-g directly inhibited platelet aggregation ( and S1
). Thus, in addition to possible roles in the inhibition of inflammatory cytokines 
and elevation of high-density lipoprotein 
, which may indirectly affect platelet activities, Dp-3-g can inhibit platelet activation independent of other blood components.
While platelet aggregometry is commonly employed to assess platelet function, it cannot examine platelet aggregation under flowing, pathophysiological conditions. The perfusion chamber is an ex vivo
model of thrombosis that has a number of important advantages over aggregometry, including the ability to assess thrombus formation on a pathophysiologically relevant substrate and under flow conditions with different shear stresses. A previous study used the Gorog Thrombosis Test to examine the anti-thrombotic effect of whole mulberry fruits under flow conditions 
; however, it is difficult to elucidate the effect of anthocyanins using whole mulberry fruits. Moreover, thrombus formation cannot be accurately monitored using the Gorog Thrombosis Test. Therefore, we employed the perfusion chamber model and monitored the effects of the purified anthocyanin Dp-3-g on thrombus formation using a confocal intravital microscopy system 
. We clearly demonstrated that Dp-3-g markedly reduced thrombus growth at both low and high shear rates using both human and murine blood at doses as low as 0.5 µM (
). These data suggest that Dp-3-g may inhibit thrombosis at both venous and arterial shear rates.
A role for red wine in supporting cyclic flow reductions (CFRs) in coronary blood flow has been observed using the Folts model of mechanically stenosed coronary arteries and intimal damage 
; however, the role of anthocyanins in mediating thrombus formation at the site of vascular injury has not been previously studied. Using intravital microscopy, we observed an inhibitory effect of the anthocyanin Dp-3-g on thrombus formation in FeCl3
-injured mesenteric arterioles and carotid arteries (). Therefore, data from our experiments in ex vivo
perfusion chambers and the in vivo
thrombosis models are consistent with our results from in vitro
platelet aggregation assays. It is notable that the inhibition of thrombus growth in these ex vivo
and in vivo
models may not only be due to effects on platelet aggregation, but may also result from the inhibition of platelet adhesion. We found significantly less adherent platelets in the perfusion chambers ( and S2
), and there was a trend toward decreased platelet deposition at the site of vessel wall injury at the early stages of thrombus formation in the intravital microscopy model (). This impairment in platelet adhesion may be due to decreased platelet activation since we did not observe that Dp-3-g interfered with GPIb complex-VWF interaction induced by botrocetin and ristocetin (data not shown), which is an important interaction for platelet adhesion at high shear stress 
. Thus, Dp-3-g inhibited platelet activation, adhesion and aggregation, the key events of thrombosis.
Platelet activation plays an important role in inflammation and the initiation of atherosclerosis, as well as thrombus formation in acute manifestations of atherosclerotic diseases 
. It has been reported that cocoa flavanol and procyanidin supplementation for 28 days significantly decreased P-selectin expression 
, and that cocoa consumption reduced ADP- or epinephrine-stimulated platelet P-selectin and PAC-1 epitope expression and platelet microparticle formation in vivo
, and suppressed unstimulated and stimulated platelet activation in whole blood ex vivo
. Our data showed that Dp-3-g can not only inhibit ADP-, collagen-, TRAP- and thrombin-stimulated P-selectin, CD63, CD40L, and PAC-1 epitope expression, but also suppresses Fg binding to platelets in both human PRP and gel-filtered purified platelets ( and
). The active conformations of integrin αIIbβ3, detected by the conformation sensitive monoclonal antibody JON/A and Fg binding, and cell surface P-selectin expression in ADP- and thrombin-activated mouse platelets, were also inhibited by Dp-3-g (Figure S3
). We found that Dp-3-g did not interfere with Fg binding to immobilized purified αIIbβ3 integrin in vitro
(data not shown). Thus, we propose that the effects of anthocyanins on platelets may be due to inhibition of platelet activation, which subsequently decreases platelet α- and δ-granule release and cytoplasmic factor secretion, as well as inhibits integrin αIIbβ3 activation and ligand binding.
The mechanism by which Dp-3-g inhibits platelet activation is currently unclear. However, several publications suggested that some polyphenolic compounds can affect platelet signaling pathways. Cinnamtannin B-1, a naturally occurring trimeric A-type proanthocyanidin, shows anti-thrombotic actions through inhibition of Ca2+
mobilization and subsequent aggregation in platelets 
. Delphinidin, a dietary anthocyanidin, inhibits platelet-derived growth factor ligand-receptor (PDGF-PDGFR) signaling 
. Genistein affects Ca2+
mobilization, subsequent platelet degranulation and aggregation. Both quercetin and apigenin have been shown to inhibit kinase activation, which may also be involved in the impairment of platelet responses 
, while quercetin alone has been shown to inhibit collagen-stimulated platelet activation through inhibition of multiple components of the GPVI signaling pathway 
. Until the present study, however, there have been no studies to address the effects of a pure anthocyanin compound on platelet signaling pathways.
AMPK, a key sensor of cell energy status, has recently been demonstrated to be involved in platelet signaling. Activated platelets can induce pAMPK generation, which may induce phosphorylation of the cytoplasmic tail of integrin αIIbβ3 via an intracellular signalling cascade 
. This inside-out signaling may lead to conformational changes in the extracellular domains of αIIbβ3 and affect ligand binding, which is required for clot retraction and thrombus stability 
. Our data clearly demonstrated that Dp-3-g can down-regulate the phosphorylation of AMPK in both human and mouse platelets following stimulation of platelets with the pathophysiological agonist collagen (), which may attenuate αIIbβ3 inside-out signaling and inhibit platelet aggregation and thrombus growth. Our data, however, cannot exclude the possibility that other platelet signaling pathways may also be inhibited by Dp-3-g. This question, including the possible tissue-specific signaling pathways mediated by anthrocyanins in platelets versus other cells 
, merit further investigation.
Anthocyanins are natural pigments responsible for the blue, purple, red and orange colors of many fruits, vegetables and other plants. Daily consumption of total anthocyanins has been estimated to be between 3 and 215 mg/day 
. The concentration of anthocyanins in plasma ranged from 0.5–1.0 µM 1.5 hours following consumption 
. Furthermore, anthocyanins have been reported to show different pharmacological properties at high concentrations, including immunomodulatory actions, anti-inflammatory actions, low-density lipoprotein inhibition, and the reduction of CVDs 
. In the present study, the concentrations of the anthocyanins used are within the range between the physiological level and the pharmacological dose (0.5–50 µM) 
. It is conceivable that the dose to attain the anti-thrombotic effects may be lower in those who have sub-optimal hemostatic factors (e.g. conditions of hemophilia or von Willebrand disease) since we found that Dp-3-g more potently inhibited platelet aggregation of β3-integrin heterozygous mice compared to their wild-type littermate controls (data not shown). Therefore, it is likely that the effects we observed will contribute to the protective and potential pharmacological effects of anthocyanins against thrombosis and CVDs.
Collectively, this study, to the best of our knowledge, is the first to show the effect of the anthocyanin Dp-3-g on platelet function. We found that Dp-3-g significantly inhibits human and murine platelet aggregation, including that with purified platelets, and attenuates thrombus formation at both arterial and venous shear stresses. We determined that Dp-3-g can directly inhibit platelet activation induced by various agonists, which likely explains its inhibitory roles in platelet aggregation and thrombus formation. We further demonstrated that Dp-3-g can attenuate platelet AMPK phosphorylation induced by collagen. Our study suggests that daily consumption of anthocyanins may play an important protective role against CVDs, including decreasing platelet-mediated inflammation and atherosclerosis. The mechanisms of this inhibitory effect on platelet activation, particularly on platelet signaling pathways, merit future studies.