Introduction

Diet, one of the most important lifestyle risk factors, can strongly influence the incidence of cardiovascular diseases [10, 96]. Indeed, accumulating data from numerous epidemiological studies indicate that regular intake of polyphenol-rich beverages and foods such as red wine, tea, and chocolate, and fruit and vegetables is associated with an improved cardiovascular prognosis [29, 48, 75, 95].

Studies aiming to characterize the beneficial effects of polyphenols on the cardiovascular systems have started several decades ago [14, 32]. The beneficial effects of polyphenols on the cardiovascular system have been attributed to their ability to reduce vascular oxidative stress, in particular, not only through their direct superoxide anion scavenging properties and interaction with other reactive oxygen species such as hydroxy and peroxy radicals [50, 81, 88, 102] but also through their stimulatory effect on endogenous antioxidant enzymes [81] and their inhibitory effect on xanthine oxidase and NAD(P)H oxidase, two major enzymes generating large amounts of reactive oxygen species [81, 82]. In addition to the antioxidant effects of polyphenols, both experimental and clinical studies indicate that polyphenols might also protect the cardiovascular system by improving the endothelial function. As it has been first demonstrated by Furchgott and Zawadzki [45], the endothelium plays a key role in the control of vascular tone by releasing several vasorelaxing factors, which have been identified later on as nitric oxide (NO) and endothelium-derived hyperpolarizing factor (EDHF) [19, 38, 51, 83, 101]. Therefore, the aim of this review is to focus on the polyphenol-induced endothelial formation of NO and stimulation of EDHF-mediated responses, two major vasoprotective factors. In addition, this review will summarize both the experimental and the clinical evidences indicating that polyphenols might improve the endothelial function in health and disease.

Polyphenols cause NO-mediated endothelium-dependent relaxations

It was initially shown that various grape products including wines, grape juices, and grape skin extracts are able to induce concentration-dependent relaxations in rat aortic rings with endothelium but only minor relaxations in rings without endothelium [41]. The fact that the polyphenol-induced relaxation is associated with an increase in the cyclic GMP content in intact aortic rings and that both the relaxation and the formation of cyclic GMP are prevented by NO synthase inhibitors indicates that grape-derived products increase the endothelial NO synthase activity leading to the formation of NO, which subsequently relaxes the vascular smooth muscle via the cGMP-mediated pathway. In addition, the endothelium-dependent relaxation appears to be strongly correlated with the concentration of polyphenols in red wines [17].

Thereafter, NO-mediated endothelium-dependent relaxations induced by polyphenols from grape-derived products, including wines and grape seeds, have been observed in various types of blood vessels including the rat and rabbit aorta, the perfused rat mesenteric artery, and in the porcine and human coronary arteries [4, 24, 40, 43, 78, 94]. Electron paramagnetic resonance spectroscopy and NO-selective microsensor provided direct evidence that grape-derived polyphenols cause an increase in the endothelial formation of NO in intact segments of rat aorta and porcine coronary artery and in cultured endothelial cells [77, 98, 99]. In addition, polyphenols from various sources such as cocoa, tea, hawthorn, maritime pine bark, honey, and propolis have also been shown to induce endothelium-dependent NO-mediated relaxations in arteries [6, 21, 33, 39, 42, 57, 60, 67, 99].

Berries and red fruits are another interesting rich source of polyphenols, mainly anthocyanins and ellagitannins. Indeed, blackcurrant juice induced pronounced endothelium-dependent relaxations of isolated porcine coronary arteries (Fig. 1). Similarly, red raspberry caused endothelium-dependent relaxations of rabbit aortic rings, and the biological activity was predominantly associated with the fraction enriched in ellagitannins [76]. Endothelium-dependent relaxations have also been observed in response to anthocyanin-enriched extracts of chokeberry and of bilberry whereas that of elderberry had only minor effects [15] suggesting that the phenolic composition of berries is determinant for their vasorelaxant activity. In addition, a procyanidin-rich extract from apples also induced endothelium-dependent NO-mediated relaxations in rat aortic rings [71].

Fig. 1
figure 1

Original tracings indicating that blackcurrant juice causes concentration-dependent relaxations of porcine coronary artery rings with endothelium but only minor effects in rings without endothelium

Full size image

Endothelium-dependent relaxations have also been observed in response to several authentic polyphenolic compounds including curcumin [111], apigenin [54], resveratrol [65, 87], and soy isoflavones [103].

Polyphenols increase endothelial NO synthase activity

Role of intracellular calcium

Activation of endothelial NO synthase in response to circulating hormones, local autocoids, substances released by platelets, by the coagulation cascade, and by the autonomic nervous system is mostly dependent upon an increase in the free cytosolic calcium concentration ([Ca2+] i ) in endothelial cells [74]. A polyphenol extract from red wine and delphinidin, a major anthocyanin present in red wine, at a concentration of 10 mg/l induced a significant increase in [Ca2+] i leading to the endothelial formation of NO [69, 98] (Fig. 2). Nevertheless, the amplitude of the calcium signal, less than 100 and 200 nM for the polyphenol extract and delphinidin, respectively, is relatively low in comparison to that induced by physiological agonists such as bradykinin. Therefore, although an increase in [Ca2+] i in endothelial cells is an important pathway leading to eNOS activation, it is likely that additional mechanisms contribute to mediate the stimulatory effect of polyphenols on endothelial NO synthase.

Fig. 2
figure 2

Schematic indicating that polyphenols are potent inducers of the endothelial formation of NO involving different intracellular signaling pathways. ROS reactive oxygen species; PI3K phosphatidylinositol 3-kinase; PDK1 phosphoinositide-dependent kinase 1; eNOS endothelial NO synthase; NO nitric oxide; ER estrogen receptor; Cav-1 caveolin-1; sGC soluble guanylyl cyclase

Full size image

Role of the PI3-kinase/Akt pathway

Besides the calcium signal, the PI3-kinase/Akt pathway is another important signal pathway leading to the activation of endothelial NO synthase such as in response to the shear stress induced by blood flow, estrogens, and vascular endothelial growth factor [18, 30, 72]. The PI3-kinase/Akt causes a rapid activation of endothelial NO synthase through its phosphorylation at Ser1177 [30].

It has been shown that red wine polyphenols (RWPs) induce the activation of the PI3-kinase/Akt pathway in endothelial cells, which, in turn, causes phosphorylation of eNOS at Ser1177 (an activator site) and dephosphorylation of eNOS at Thr495 (an inhibitor site), leading to an increased formation of NO [77] (Fig. 2). The PI3-kinase/Akt signaling pathway is triggered by the polyphenol-induced intracellular formation of reactive oxygen species predominantly superoxide anions in endothelial cells [5, 68, 77]. A similar mechanism of eNOS activation has been observed in response to a grape seed extract [35], grape skin extract [68], purple grape juice [5], strawberry powder rich in polyphenols [36], and epigallocatechin-3-gallate [67], a major flavonoid from the flavan-3-ol subclass found in green tea. In addition, polyphenols cause within minutes the phosphorylation of eNOS, and this effect persists for several hours indicating a sustained activation of eNOS. In contrast, activation of G protein-coupled receptors by bradykinin or thrombin caused only a transient formation of NO, which reached a peak value within 1 min and thereafter vanished rapidly [27, 91].

Additional investigations identified Src kinase as a redox-sensitive mediator, which acts upstream of the PI3-kinase/Akt pathway leading to eNOS activation in response to grape-derived polyphenols [5, 6, 68] (Fig. 2). Moreover, Fyn, a member of the Src family, mediated the epigallocatechin gallate-induced PI3-kinase/Akt-mediated activation of eNOS [59].

The polyphenolic-rich fraction of black tea acutely enhanced NO formation subsequently to the phosphorylation eNOS at Ser1177 and the dephosphorylation of eNOS at Thr495 in porcine aortic endothelial cells [8]. This stimulatory effect is calcium-dependent involving both intracellular and extracellular calcium and involves the p38 MAPK upstream of the PI3-kinase/Akt pathway [8]. A calcium-dependent activation of eNOS has been shown in response to the tannin 1-α-O-galloylpunicalagin which is associated with the PI3-kinase/Akt pathway [20]. Therefore, changes in [Ca2+] i in endothelial cells are likely to contribute to the redox-sensitive activation of eNOS in response to polyphenols via the PI3-kinase/Akt-dependent pathway.

Role of estrogen receptors

It has been shown that low concentrations (in the nanomolar range) of resveratrol, a polyphenolic phytoestrogen found in grapes and wine, as well as black tea polyphenols are able to activate estrogen receptors resulting in activation of p38 MAPK and eNOS in endothelial cells [7, 61] (Fig. 2). In contrast, the estrogen receptor antagonist ICI 182,780 did not alter relaxations to RWPs in intact aortic rings from both male and female rats [55]. In addition, the RWP-induced phosphorylation of Akt and eNOS in porcine coronary artery endothelial cells was not affected by ICI 182,780 [55]. Similar results were observed in the isolated porcine coronary arteries with the Crataegus (Hawthorn species) special extract WS 1442 [6]. Therefore, these data indicate that an estrogen receptor-dependent pathway mediates activation of eNOS in response to some polyphenols including resveratrol and black tea polyphenols, but not to red wine polyphenols and Crataegus polyphenols.

Role of caveolin-1

A recent study has indicated that green tea polyphenols downregulate caveolin-1 protein expression and mRNA levels in both time- and concentration-dependent manners via activation of ERK1/2 and inhibition of p38 MAPK signaling pathways in bovine aortic endothelial cells [66] (Fig. 2). As caveolin-1 is a major negative regulator of eNOS activity, such an effect might contribute to increase eNOS activation by polyphenols.

Polyphenols increase endothelial NO synthase expression

In addition to causing a rapid and sustained activation of eNOS, polyphenols have been shown to increase the expression level of eNOS in cultured endothelial. Indeed, exposure of endothelial cells to red wine increased eNOS expression at both the mRNA and protein levels [105, 107]. The upregulation of eNOS is attributed to polyphenols contained in red wine because ethanol alone has no such effect [107]. Moreover, eNOS upregulation is also observed in response to a red wine extract without alcohol [63]. However, it must be underlined that large concentrations of red wine are required to increase eNOS expression (1% v/v in culture medium for 10 days, 3% v/v for 24 h, and 10% v/v for 12 h) [107]. The analysis of the active polyphenols in red wine indicated the involvement of several polyphenolic compounds, in particular, trans-resveratrol and also to a lesser extent cinnamic and hydroxycinnamic acids, cyanidin, and several phenolic acids [106].

Resveratrol upregulated eNOS expression at concentrations ranging from 10 to 100 µM [49, 105]. The stimulatory effect is mainly mediated by an increase in the activity of the eNOS promoter (transcriptional effect) and a stabilization of eNOS mRNA (post-transcriptional effect) [105]. In addition, lower concentrations of resveratrol such as 0.1 µM have also been shown to increase the expression of eNOS mRNA in cultured human umbilical vein endothelial cells [80]. Moreover, it has been reported that in cultured human coronary arterial endothelial cells, resveratrol upregulates the expression of SIRT1, a NAD+-dependent protein deacetylase, and induces its enzymatic activation leading to an upregulation of eNOS mRNA expression and mitochondrial biogenesis; all these events might contribute to improve the endothelial function [25, 71].

The evaluation of the effectiveness of 33 different polyphenols (procyanidins, monomeric flavan-3-ols, flavonols, a flavone, a flavanone, a chalcone, a stilbene, and phenolic acids) to enhance eNOS mRNA expression at 100 µM in a hybrid human endothelial cell line EA.hy926 revealed a significant stimulatory effect only in response to four polyphenols including resveratrol, quercetin, epicatechin-gallate, and epigallocatechin-gallate. No clear relationship between the structure of the active polyphenols and the stimulatory activity was observed [9].

Polyphenols induce EDHF-mediated responses in arteries

The component of arterial endothelium-dependent relaxations, which is resistant to inhibitors of NO synthase and cyclooxygenases, has been attributed to EDHF. It is now commonly admitted that the importance of the EDHF phenomenon increases as the vessel size decreases [93]. Several mechanisms have been proposed to explain the nature of the EDHF phenomenon (see related chapters in the present issue), but one hallmark of EDHF-mediated responses is that these relaxations are associated to an endothelium-dependent hyperpolarization of the vascular smooth muscle cells.

The initial demonstration of a participation of EDHF to the endothelium-dependent relaxations to polyphenols has been observed in isolated porcine coronary arteries [78] (Fig. 3). Indeed, RWPs caused concentration-dependent relaxations and hyperpolarizations of vascular smooth muscle cells, at concentrations ranging from 1 to 100 mg/l, with a maximal effect at 100 mg/l. Thereafter, it was shown that Concord grape juice, a nonalcoholic rich source of grape-derived polyphenols, is also able to induce endothelium-dependent EDHF-mediated relaxations of porcine coronary arteries [5]. EDHF-mediated endothelium-dependent relaxations have also been observed in the isolated, perfused or not, mesenteric arterial bed in response to alcohol-free lyophilized Brazilian red wine [28], an extract of Eucommia bark (a traditional Chinese medicinal herb containing polyphenols) [62], RWPs [26], and açai (Euterpe oleracea Mart.) [89]. Moreover, intercellular communication through gap junctions is involved in the RWP-induced EDHF-mediated relaxation in the rat mesenteric artery [26].

Fig. 3
figure 3

Schematic indicating that polyphenols are potent inducers of the endothelial formation of EDHF via the PI3-kinase/Akt pathway; ROS reactive oxygen species; PI3K phosphatidylinositol 3-kinase; PDK1 phosphoinositide-dependent kinase 1; EDHF endothelium-derived hyperpolarizing factor

Full size image

Although resveratrol has been shown to activate IKCa channels (24 pS) in MS1 cell line, a pancreatic islet endothelial cell line, by increasing their open probability [64], other mechanisms have been involved in the polyphenol-induced EDHF-mediated relaxation. Indeed, the RWP-induced EDHF-mediated relaxation and hyperpolarization are strictly dependent on an endothelial redox-sensitive mechanism involving the intracellular formation of superoxide anions in porcine coronary arteries [78] (Fig. 3). EDHF-mediated relaxations are reduced by antioxidants, such as N-acetylcysteine, and membrane permeant analogs of superoxide dismutase (SOD), such as Mn(III)tetrakis(1-methyl-4-pyridyl)porphyrin (MnTMPyP) and polyethylene glycol-SOD (PEG-SOD). In addition, exposure of cultured porcine coronary artery endothelial cells to RWPs induced the MnTMPyP-sensitive formation of superoxide anions in the presence of L-NA and indomethacin [78]. Further investigations indicated that the RWP-induced EDHF-mediated relaxation of porcine coronary arteries involves the redox-sensitive activation of PI3-kinase leading to Akt phosphorylation in endothelial cells [79] (Fig. 3). The possibility that the PI3-kinase/Akt pathway modulates myo-endothelial gap junctions and/or potassium channel activity remains to be investigated.

Polyphenols and endothelial function in experimental models of cardiovascular diseases

Since grape-derived products are able to strongly stimulate the endothelial formation of NO and also the EDHF-mediated responses, two potent vasoprotective factors, several studies have assessed their potential to delay the development of major types of cardiovascular diseases characterized by an endothelial dysfunction. These studies have indicated that intake of grape-derived products in the drinking water reduced blood pressure in several experimental models of hypertension including the spontaneously hypertensive rat, the NG-nitro l-arginine-induced hypertension, the DOCA salt-induced hypertension and the angiotensin II-induced hypertension in rats [16, 53, 85, 90, 93]. Besides normalizing blood pressure in angiotensin II-treated rats, red wine polyphenols prevented the development of an endothelial dysfunction by normalizing the excessive vascular formation of superoxide anions, which react with NO to form peroxynitrites [90]. The protective effect of polyphenols on the endothelial function is explained by their ability to prevent the increased vascular expression of NADPH oxidase, a major vascular source of superoxide anions, and the cyclooxygenase-dependent formation of endothelium-derived contracting factors [56, 90] (Fig. 4). Alternatively, the beneficial effect might also be due to the downregulation of angiotensin II type I receptor (AT1) as shown in vascular smooth muscle cells treated with resveratrol [73] (Fig. 4). In the spontaneously hypertensive rat, ingestion of grape-derived polyphenols also reduced blood pressure, and this effect is associated with a reduced vascular formation of superoxide anions, improved cognitive function, and reduced cardiac and ventricular hypertrophies [85]. Moreover, grape-derived extracts prevented the development of hypertension, cardiac hypertrophy, and vascular oxidative stress by reducing NADPH oxidase expression in fructose-fed rats [2, 3]. In addition, intake of grape-derived polyphenols prevented the development of aortic fatty streaks, early lesions of atherosclerosis, in hamsters receiving a high-fat diet, an experimental model of atherosclerosis [11, 12].

Fig. 4
figure 4

Schematic summarizing the protective effect of polyphenols on the arterial wall in major types of cardiovascular diseases; polyphenols are able to strongly stimulate the redox-sensitive endothelial formation of NO and EDHF, two potent vasoprotective factors, and to prevent the endothelial formation of cyclooxygenase-derived metabolites of arachidonic acid that act on thromboxan-protanoid receptors at the vascular smooth muscle to cause vasoconstriction. In addition, polyphenols also prevent the excessive vascular formation of reactive oxygen species mostly by normalizing the expression of NADPH oxidase and the angiotensin AT1 receptor in the arterial wall. Both the effect at the endothelial cell and the vascular smooth muscle cell will promote vasodilation and prevent inflammatory responses and activation of platelets. ROS reactive oxygen species; COXs cyclooxygenases; PI3K phosphatidylinositol 3-kinase; EDCF endothelium-derived contracting factors; EDHF endothelium-derived hyperpolarizing factor; NO nitric oxide; Ang II angiotensin II

Full size image

A polyphenol-rich cocoa powder (up to 300 mg/kg bodyweight) also reduced blood pressure similarly to 50 mg/kg of captopril, an angiotensin converting enzyme inhibitor, in the spontaneously hypertensive rat [23].

Several studies have also reported an antihypertensive effect in response to several purified polyphenols from fruits and vegetables. Indeed, quercetin, a flavonol found widely in fruits and vegetables, reduced blood pressure in several experimental models of hypertension including spontaneously hypertensive rats, NG-nitro l-arginine methyl ester-treated rats, two-kidney one clip Goldblatt rats, DOCA-salt, and Dahl salt-sensitive rats [86]. Similarly, genistein, an isoflavone found mainly in vegetables, as well as hesperitin and glucosyl-hesperidin, two flavonones, decreased blood pressure and improved endothelium-dependent relaxations of aortic rings of spontaneously hypertensive rats [104, 112, 113]. Furthermore, catechin prevented endothelial dysfunction in the prediabetic stage of Otsuka Long-Evans Tokushima Fatty rats by reducing vascular NADPH oxidase activity and expression [52].

Altogether, these experimental studies indicate that ingestion of polyphenol-rich products has a beneficial effect in several experimental models of major cardiovascular diseases including hypertension, diabetes, and atherosclerosis. The beneficial effect is associated with the restoration of the vasoprotective effect of endothelial cells most likely by reducing the excessive NADPH oxidase-dependent vascular oxidative stress. In addition, the ability of polyphenols to increase the endothelial formation of NO and the EDHF-mediated responses might also contribute to the vasoprotective effect either directly or indirectly by preventing the oxidative stress-induced inactivation of NO.

Polyphenols and endothelial function in healthy and diseased humans

In healthy subjects, ingestion of 250 to 500 ml of red wine or de-alcoholized red wine increased, in some but not all studies, flow-mediated vasodilatation as assessed in the brachial artery by plethysmography [1, 31, 47]. Intake of 5–7 ml/kg of purple grape juice by healthy subjects for 7–14 days reduced ex vivo platelet aggregation and increased platelet-derived NO formation [44, 58]. In patients with coronary artery diseases, flow-mediated dilatation was improved after intake of 4 ml/kg of either white or red wine during a light meal [109] and also after intake of about 8 ml/kg of purple grape juice [22, 97]. The beneficial effect of purple grape juice on the endothelial function was associated with a reduced susceptibility of LDL particles to oxidation [97]. In addition, intake of 5.5 ml/kg of purple grape juice daily by hypertensive Korean patients for 8 weeks reduced both systolic and diastolic blood pressure by, respectively, 7.2 and 6.2 mmHg [84]. Besides grape-derived products, an improved endothelial function in patients with coronary artery diseases has been observed in response to intake of short- and long-term black tea consumption [34] and after dietary supplementation with epigallocatechin gallate, a major catechin in tea (300 mg) [110]. In addition, consumption of a flavonoid-rich dark chocolates and cocoa beverages improved endothelial function in healthy subjects, hypercholesterolemic postmenopausal women, and in diabetic patients associated to an increased circulating level of NO species [13, 37, 92, 108]. Moreover, an improved endothelial function was observed after administration of (–)-epicatechin (1 mg/kg, a major flavanol of cocoa) to healthy humans [92]. Chronic intake of flavanol-rich dark chocolate reduced blood pressure in patients with upper-range hypertension or stage 1 hypertension without concomitant risk factors [46, 100].

Conclusion

Altogether, these experimental and clinical studies highlight the potential of polyphenol-rich sources to ameliorate or restore vascular protection in health and disease by enhancing the two major endothelial vasoprotective mechanisms, i.e., the formation of NO and EDHF-mediated responses and also by reducing oxidative stress in the arterial wall which promotes pro-inflammatory and pro-thrombotic responses (Fig. 4).