Direct vasoactive and vasoprotective properties of anthocyanin-rich extracts

David R. Bell, Kristen Gochenaur


Reactive oxygen species (ROS) play a critical role in the impairment of nitric oxide-mediated vascular functions and overall pathogenesis associated with cardiovascular disease. Plant pigment anthocyanins are exceptionally potent oxygen radical scavengers that produce beneficial effects in diseases outside the cardiovascular system. We examined for the first time the potential coronary vasoactive and vasoprotective properties of three anthocyanin enhanced extracts prepared from chokeberry (Ck), bilberry (B), or elderberry (E). Coronary arterial rings were isolated from 64 pigs and incubated in sterile tissue culture media overnight for use in one of four separate in vitro isometric force recording studies. Ck and B, but not E, produced dose- and endothelium-dependent vasorelaxation. (%maximal relaxation at 5 mg total anthocyanins per liter: Ck = 68 ± 11, B = 59 ± 10). Coronary vascular tone, endothelium-dependent vasorelaxation to A23187, and vasorelaxation to DEA NONOate were not affected by exposure of rings to any extract at 0.05 mg total anthocyanins per liter for 5 or 30 min. Ck extract at 0.05 mg total anthocyanins per liter showed the greatest protection against loss of A23187 relaxation following exposure to ROS from pyrogallol (Ck, % maximal relaxation and −logED50 to A23187, respectively, means ± SE: Ck alone, 93 ± 5%, 7.91 ± 0.1; pyrogallol alone, 76 ± 7%, 7.46 ± 0.06; pyrogallol + Ck, 98 ± 1%, 7.82 ± 0.06; control: 99 ± 1%, 7.86 ± 0.07; P < 0.05 control vs. pyrogallol alone). Neither the extracts nor pyrogallol affected responses to DEA NONOate. Thus anthocyanin-enhanced extracts produce endothelium-dependent relaxation in porcine coronary arteries. Extract concentrations too low to directly alter coronary vascular tone protect coronary arteries from ROS without altering vasorelaxation to endogenous or exogenous NO. These results suggest that such extracts could have significant beneficial effects in vascular disease.

  • antioxidant
  • endothelium
  • nitric oxide
  • superoxide

the endothelial nitric oxide (NO) system plays a pivotal role in vascular physiology and pathology. NO is a potent vasodilator agent with anti-hypertensive, anti-thrombotic, anti-atherogenic, and anti-smooth muscle proliferative properties (44). However, the endothelial NO system is impaired in conditions associated with atherosclerosis, hypertension, diabetes, and ischemia-reperfusion injury (10). These conditions are also associated with excess vascular and extravascular production of reactive oxygen species (ROS), such as superoxide (O2) (14, 15, 21, 23). ROS impair endothelial NO functions through direct damage to the endothelium as well as by chemical quenching of NO (22). Clearly, factors that can enhance or protect the endothelial NO system, or scavenge and inactivate ROS, have the potential for far-reaching beneficial impacts on cardiovascular disease.

It has been suggested that such protective and anti-oxidant effects may be responsible for the decreased incidence and risk of cardiovascular disease associated with consumption of fruits and vegetables (34). However, extensive studies attempting to link this observation to specific plant nutritional components, such as anti-oxidant vitamins, have failed to demonstrate significant correlations (6, 20). This has led to the suggestion that some previously overlooked, nonnutritional component of plants is the real source of the plant's cardioprotective effect (32). In this regard, recent evidence indicates that plant anthocyanins may have strong potential as cardioprotective agents. Anthocyanins are water soluble, glycosylated, nonacetylated polyphenolic compounds that make up the red, blue, and purple pigments of fruits (7, 13). These pigments are absorbed in intact (glycosylated) form from the human gastrointestinal tract after oral consumption (25, 2729, 47) and have been shown to have tumor suppressive, anti-inflammatory, anti-viral, and anti-diabetic properties (16, 19, 36, 37, 42, 43).

Of potential importance to cardiovascular disease, anthocyanins are potent anti-oxidant and ROS scavengers (11, 18, 33, 35, 40, 46, 48, 52). They have been shown to inhibit oxidation of lipids (35) and LDLs (45) as well as protect cultured endothelial cells from oxidant injury (51). A recent study (45) indicates that the single anthocyanin cyanindin-3-O-glucose chloride has a greater efficacy in protecting against oxidant degradation of LDLs than vitamins C, E, and beta-carotene combined. The anthocyanin concentration of fruit is highly correlated with the oxygen radical absorbance capacity (ORAC) (5, 11, 33, 40, 46, 48, 52), with chokeberry, bilberry, and elderberry having the highest natural intrinsic total anthocyanin (Ta) content and ORAC of dozens of fruits and vegetables tested (48, 52 and Artemis International, Fort Wayne, IN, internal data analysis).

We have obtained novel powdered pigment extracts from chokeberry (Aronia melanocarpa), bilberry (Vaccinium myrtillus), and elderberry (Sambucus nigra) to examine their potential vasoactive and vasoprotective properties in coronary arteries. These extracts have been processed to enrich their Ta content while preserving the naturally different individual anthocyanin profiles within each. We have designed three experiments to test the hypotheses that 1) the extracts are directly vasoactive, 2) the extracts directly modify endothelium-dependent and -independent responses to NO, and 3) the extracts protect coronary arteries from oxidant injury. This report represents the first study of any kind on the vascular properties of these preparations.


In vitro analyses of the direct vasoactive or vasoprotective properties of three berry pigment extracts were performed using isolated coronary artery rings from mature female pigs. Extracts of elderberry (Sambucus nigra), bilberry (Vaccinium myrtillus), and chokeberry (Aronia melanocarpa) were obtained from Atermis International (Fort Wayne, IN). The extracts we received were in the form of lyophilized powders that had their free sugars removed to enhance the Ta content of each extract on a percent dry weight basis. The Ta and total polyphenolic concentration of each extract is listed in Table 1. The extracts also exhibit different anthocyanin profiles. Chokeberry contains four different anthocyanins with ∼90% of its Ta content due to cyanidin-3-galactoside and cyanidin-3-arabinoside. These cyanidin glycosides are not present in elderberry. In contrast, over 90% of the Ta content in elderberry is due to cyanidin-3-glucoside and cyanidin-3-sambubioside, which are not present in chokeberry. Bilberry contains ∼1–15% of 15 different anthocyanins including those in the cyanidin, peonidin, delphinidin, petunidin, and malvidin classes.

View this table:
Table 1.

Total phenolic and anthocyanin composition of 4 berry extracts

Hearts from eight female pigs were obtained from local slaughterhouses immediately after death and placed in physiological salt solution (PSS) at 4°C (PSS in mM: 130 NaCl, 4.7 KCl, 14.9 NaHCO3, 5.5 dextrose, 1.18 KH2PO4, 1.17 MgSO4·7H2O, 1.60 CaCl2·2H2O, 0.03 CaNa2 EDTA). Hearts then were transferred to our laboratories where the left anterior descending and circumflex coronary arteries were carefully dissected free from the surrounding tissue and washed with sterile PSS. With the use of sterile procedures, arteries were cleaned of excess fat and connective tissue and cut into 2- to 4-mm-wide rings. Rings were transferred under sterile conditions into separate petri dishes containing sterile HEPES-buffered serum and phenol red-free Dulbecco's modified Eagle’s media with antibiotic and then placed in an O2/CO2 incubator overnight utilizing standard procedures for cell and tissue cultures as described previously (2). As such, rings were kept in a humidified environment with temperature and CO2 concentration precisely controlled at 37°C and 5.0%, respectively.

After the incubation period, rings were removed from the dishes and washed with PSS. The endothelium was removed from one-half of the rings by gentle rubbing of the intima with a forceps. Next, all rings were suspended between triangular stainless steel hooks and a stationary support rod, transferred in pairs to artery baths containing PSS at 37°C, pH 7.4, bubbled with 95% O2-5% CO2, and connected to FT.03 force transducers (Grass Instruments, Quincy, MA) for isometric force recording (Grass P7 and Gould 2400S recorders). Rings were suspended at a passive tension of 10 g, previously determined to produce maximal force generation in all treatment groups, and allowed to equilibrate for not less than 90 min. Maximal contractile force was determined in all rings by exposure to PSS containing equimolar replacement of NaCl with 130 mM KCl. Next, prostaglandin synthesis was blocked in the rings by exposure to 10−5 M meclofenamic acid for 30 min followed by repeated rinses with PSS. All rings were then contracted to 35–45% of their individual KCl maximum with U46619 (∼1–2 × 10−8 M).

Stock solutions of elderberry (Sambucus nigra), bilberry (Vaccinium myrtillus), or chokeberry (Aronia melanocarpa) extracts were normalized for Ta concentration in double-distilled H2O and used immediately for dose-response determinations in the ring preparations. Rings were exposed to cumulative full log M Ta additions of one of the extracts over a concentration range of 0.005–5 mg Ta/l and any final changes from initial tone recorded. After determination of these relationships, the presence or absence of a functional endothelium in the rings was determined by the presence or absence, respectively, of relaxation of the rings to A23187. In rings from four pigs, the involvement of NO in any observed relaxation instead was tested by exposing rings to 100 uM NO2-l-arginine at the end of the last dose of pigment extract.

In separate experiments with the use of coronary rings from 10 additional female pigs, the effect of the acute presence of berry extracts on endothelium-dependent and independent NO-mediated vasorelaxation was examined. For these experiments, eight artery baths containing two rings each were used with each pig serving as its own control. All rings were incubated in vehicle control media overnight, suspended in artery baths, exposed to KCl and meclofenamate, and precontracted with U46619 as described in the previous section. Next, two baths each were exposed to elderberry, bilberry, and chokeberry extract, respectively, for 5 min at a concentration too low to alter vascular tone (0.05 mgTa/l). Rings in two additional artery baths were not exposed to any agents during this time and served as controls. Rings were then exposed to cumulative −1/2 log M additions of either A23187 or DEA NONOate. A23187 is a direct, receptor-independent activator of endothelial constituitive NO synthase (ecNOS; Ref. 31)]. In porcine coronary artery rings pretreated with meclofenamate, A23187 produces dose-dependent relaxation that is entirely mediated by endothelial NO (2, 3). [Acetylcholine, the most common receptor-mediated activator of ecNOS, is not an appropriate agonist of endothelium-dependent relaxation of coronary arteries from pigs because it produces marked endothelium-independent contraction with little or no relaxation of any type in porcine coronary arteries (8).] DEA-NONOate spontaneously releases NO when exposed to a pH < 8.0 and thus serves as a source of NO for the rings. This source has the advantage of avoiding required intracellular bioconversion to produce NO, as is the case with nitroglycerin, and does not possess the secondary, non-cGMP-mediated vasodilator properties of sodium nitroprusside. In a separate group of 13 female pigs, the procedures used for acute exposure of coronary arteries to berry extracts described above were repeated with the exception that the rings were exposed to the berry extracts for 30 min before the addition of A23187 or DEA NONOate.

In additional studies, coronary arterial rings from 33 additional female pigs were used to examine whether impaired endothelium-dependent or -independent coronary vasorelaxation following exposure to O2 is altered by concurrent exposure to berry extracts. Rings were incubated in vehicle control media overnight, suspended in artery baths for isometric force recording, and exposed to KCl and meclofenamate as described above. Rings were then divided into four treatment groups. Group 1 consisted of rings exposed to O2 through the auto-oxidation of 400 μM pyrogallol (24) added to the artery bath for 15 min. This treatment impairs, but does not abolish, endothelium-dependent, NO-mediated vasorelaxation and thus allows one to test whether any agent exacerbates or attenuates O2-mediated damage to the artery (2). Group 2 consisted of rings exposed to pyrogallol, as in group 1, but with the simultaneous presence of one if the berry pigment extracts at a concentration of 0.05mg Ta/l. Group 3 consisted of rings exposed to the extract only for 15 min, whereas the group 4 received no additional treatments during the 15-min exposures in groups 1–3 and thus served as controls. After these exposures, all rings were rinsed repetitively with PSS and precontracted with U46619. Dose-response relationships to A23187 or DEA NONOate were then determined as described above.

Data analysis.

Responses to all agents were recorded as the average response of two rings per treatment type per pig. Sensitivity of the coronary rings to an agent was expressed as the −log M ED50 value as determined by Probit transformation and linear regression of the dose-response relationship. Differences among the means of treatment groups were tested by a paired t-test with the Bonferroni correction for multiple comparisons. Any alterations in responses to A23187 were compared with responses to DEA-NONOate in coronary artery rings from the same pig. This comparison was made to determine whether any alteration in endothelium-dependent relaxation caused by a treatment was a reflection of altered endothelial function related to NO or, instead, was a reflection of altered arterial smooth muscle responsiveness to NO.


DEA-NONOate used in this study was obtained from Cayman Chemicals (Ann Arbor, MI). All other agents used in these experiments were obtained from Sigma Chemical (St. Louis, MO).


Responses of the rings to direct application of cumulative doses of each berry extract are shown in Fig. 1. Chokeberry and bilberry, but not elderberry, extracts produced dose-dependent relaxation in isolated rings with endothelium. Responses to chokeberry extract produced the greatest degree of relaxation. Rings without endothelium did not relax to any dose of the berry extracts. In rings in which the contribution of NO to this relaxation was examined, the relaxations to either chokeberry or bilberry was abolished by 100 μM NO2-l-arginine. An example of this effect is shown in Fig. 2.

Fig. 1.

Relaxation responses of porcine coronary arterial rings with endothelium (A) and without endothelium (B) to cumulative log doses of chokeberry, bilberry, or elderberry extracts at equal total anthocyanin (Ta) concentrations. Data are means ± SE (n = 8 in all groups) of the mg force remaining following each dose expressed per mg force of initial precontraction to U46619.

Fig. 2.

Examples of the effect of inhibition of endothelial nitric oxide synthase by 100 μM NO2-l-arginine on relaxation of coronary arterial rings with endothelium by chokeberry (top), bilberry (middle), and elderberry (bottom).

The presence of any berry extract at a concentration too low to cause direct changes in arterial tone (0.05 mg Ta/l) did not alter sensitivity of rings to either A23187 or DEA NONOate (Table 2). This lack of effect was seen when the extracts were added either 5 or 30 min before administration of either vasodilator. Precontraction of rings with U46199 did not differ among any groups. Maximum relaxation to each agent was similarly unaffected by acute exposure to any of the extracts.

View this table:
Table 2.

Sensitivity and maximal relaxation of porcine coronary arteries to A23187 and DEA NONOate after 5- and 30-min exposure to 0.05 mg totalanthocyanins/l from berry extracts

The effect of berry pigment extracts on the impairment of endothelium-dependent vasorelaxtion to A23187 following exposure to O2 from pyrogallol is shown in Fig. 3. Exposure of the artery rings to 400 μM pyrogallol for 15 min resulted in a significant decrease in maximal relaxation and sensitivity to A23187 (Fig. 3 and Table 3) without altering responses to DEA NONOate (Table 4). This impairment was abolished by concurrent exposure to 0.05 mg Ta/l chokeberry extract in arterial rings from all 12 pigs used in that determination. Similar impairment in sensitivity was abolished in rings from 9 of 12 pigs used in the bilberry determination and 8 of 12 pigs used in the elderberry determination. Sensitivity of rings exposed to bilberry and pyrogallol simultaneously was not significantly different from controls but was also not significantly different from rings treated with pyrogallol alone. Sensitivity of rings exposed to elderberry and pyrogallol simultaneously was significantly different from controls and from rings treated with pyrogallol alone. Precontraction of rings with U46619 was not different among any groups, nor did prior exposure to pyrogallol affect the concentration of U46619 needed to precontract arteries to 35–45% of their KCl maximal contraction.

Fig. 3.

Effect of chokeberry (A; n = 9), bilberry (B; n = 12), and elderberry (C; n = 12) extract on impaired coronary vasorelaxation from exposure to pyrogallol. Relaxation responses of porcine coronary arterial rings with endothelium were performed to cumulative −log doses of A23187. Rings were precontracted with U46619 to a similar percent of their individual KCl maximum before administration of A23187. ○, Control rings; ▾, rings exposed to 400 μM pyrogallol for 15 min before precontraction with U46619; ▴, rings exposed to 400 μM pyrogallol plus 0.05 mg Ta/l of extract for 15 min before precontraction with U46619; □, rings exposed to 0.05 mg Ta/l extract only for 15 min before precontraction with U46619. Data are means + SE of the mg force remaining following each dose expressed per mg force of initial precontraction to U46619. *Maximum relaxation to A23187 was significantly impaired in rings treated with pyrogallol, P < 0.05.

View this table:
Table 3.

Sensitivity and maximal relaxation of porcine coronary arteries to A23187

View this table:
Table 4.

Sensitivity and maximal relaxation of porcine coronary arteries to DEA NONOate


We observed that anthocyanin-rich extracts from chokeberry and bilberry, but not elderberry, produced dose-dependent relaxation of coronary arteries, with the chokeberry extracts exhibiting the highest potency. These differences in vasorelaxation among the extracts cannot be explained on the basis of a generic, nonselective effect of their Ta content because each dose of an extract was adjusted to the same Ta concentration. Also, it is unlikely that these differences were due to different total phenolic content in the extracts; arteries exposed to elderberry extract were exposed to total phenolic levels similar to those in the other extracts but did not relax. These results suggest that some component or combination of components of the chokeberry and bilberry extracts, but not found in elderberry, may possess vasodilatory properties. These differences could be related to different anthocyanin profiles within each extract. However, the experiments required for a definitive answer to this postulate are beyond the scope of our present study.

Relaxation of coronary arteries to chokeberry and bilberry preparations was totally endothelium dependent. It is unlikely this was mediated through endothelial prostaglandin synthesis because such synthesis was blocked in our arterial rings. In arteries from four of the pigs, we tested for the role of NO in these relaxations and found that such relaxation could be abolished by the application of 100 μM NO2-l-arginine. These observations suggest that the endothelial NO system may be involved in the relaxation response of coronary arteries to chokeberry and bilberry extracts.

The literature on vasoactive properties of anthocyanins is limited. Fruit pigment preparations have been shown to produce endothelium-dependent relaxation of arteries, but these determinations have been largely confined to pigments contained in red wine or grapes (1, 9, 26, 30, 39). Such preparations contain some anthocyanins but also contain significant amounts of other polyphenolics that are minor or nonexistent components of our berry pigment preparations. However, Xu et al. (49) showed that endothelial NO synthase in BAEC is upregulated after a 6-h exposure to 0.1 μM cyanidin-3-glucoside and that a 12-min exposure of these cells to the same compound phosphorylates NO synthase and enhances NO synthase activity (50). Such mechanisms may underlie the endothelial-dependent vasodilatory properties of our extracts. However, it is not certain whether this specific cell culture effect can be extrapolated to NO-mediated relaxation of our coronary arteries because cyanidin-3-glucoside is not present in chokeberry and is a minimal component of bilberry preparations.

It would be premature to try to extrapolate the endothelium-dependent relaxation property of our berry pigment extracts to potential human health benefits. Although anthocyanins are absorbed intact across the human gastrointestinal tract, little is known about the effective plasma concentrations of anthocyanins following oral consumption (25, 2729, 47). Cao et al. (4) have reported that plasma concentrations of Ta are ∼100 nM following oral consumption of elderberry extracts or blueberries and that this amount reflects primarily just two cyanidin glycosides. This concentration is similar to the concentration of cyanidin glycosides in the chokeberry and bilberry extracts that relaxed coronary artery rings. However, no studies have been performed yet to determine the human bioavailability of anthocyanins or other components from the specific berry preparations used in our study.

Results from several studies suggest that anthocyanins and other polyphenolics may have noncardiovascular health benefits (16, 19, 32, 36, 37, 42, 46). However, the health benefit of any compound would be seriously compromised if it interfered with coronary arterial function. Furthermore, agents can exist in plasma in concentrations too low to directly alter vascular tone but nevertheless alter arterial reactivity because they modify the response of arteries to another agent (2, 3). Our experiments indicate that none of the berry preparations, at a Ta concentration too low to directly modify arterial tone, altered endothelium-dependent or -independent NO-mediated coronary vasorelaxation acutely.

NO-mediated vasorelaxation can be enhanced by scavengers of O2 (3, 17, 41). Because anthocyanins and polyphenolics are potent ROS scavengers, one might postulate that these compounds would affect coronary arterial relaxation mediated by NO. However, for such an effect to occur, basal arterial generation of O2 would have to be significant enough to affect the bioavailability of NO and do so to the extent it impaired NO-mediated relaxation. This is an unlikely scenario in our preparations. In previous studies, our laboratory (2) showed that ascorbate, an O2 scavenger, does not alter relaxation to A23187 or DEA NONOate in isolated control coronary arteries even though it readily corrects impaired NO relaxations brought about by exposure to exogenously supplied O2. Therefore, in our study, the failure of normal NO-mediated coronary vasorelaxation to be altered in the acute presence of the berry extracts is not a false negative.

Pyrogallol is an auto-oxidizing source of O2 (24, 31) that has been used to induce damage from ROS in in vitro artery bath and other preparations (2, 31). In our study, acute exposure of isolated porcine coronary arteries to pyrogallol resulted in an impairment of subsequent endothelium-dependent, NO-mediated vasorelaxation by A23187 but not DEA NONOate. This impairment cannot be due to quenching of NO by O2 when the arteries are exposed to A23187 because pyrogallol was not present during A23187 exposure. These results are identical to those our laboratory reported previously (2) and indicate that the loss of endothelium-dependent, NO-mediated relaxation to A23197 following exposure to pyrogallol results from O2-mediated damage that is functionally confined to the coronary arterial endothelium.

It is clear that the berry extracts used in this study significantly attenuated the loss of A23187-mediated relaxation of coronary arteries exposed to pyrogallol. This effect was most prominent in arteries exposed to chokeberry extracts and least prominent in those exposed to elderberry extracts. It is most likely that the beneficial effects of these extracts are due to their ORAC, which would tend to reduce the effective concentration of O2 or ROS from pyrogallol reaching the arterial endothelium. More importantly, this ability to prevent loss of endothelium-dependent, NO-mediated relaxation was achieved at an extremely low Ta concentration that roughly reflects that seen in several studies to exist in human plasma after oral consumption of these compounds. However, because the Ta concentration to which all arteries were exposed was normalized for each of the extracts used in our experiment, the total polyphenolic concentration to which the vessels were exposed had to be slightly higher for the chokeberry and bilberry extracts compared with that obtained with the elderberry preparation. Because the ORAC is also correlated with phenolic concentration in a preparation as well as anthocyanin content, the enhanced efficacy of the chokeberry extract compared with the other preparations tested may be related in part to its total phenolic content (33). Nevertheless our results extend and support recent observations that suggest that red wine or anthocyanins protect vascular endothelium from oxidant injury (12, 38, 51).

It remains to be seen whether our observations, which are limited to a specific type of radical damage and arterial function, are a reflection of a more general vascular protective effect of these berry-derived compounds. Should our observations be representative of a broader ability of these berry preparations to prevent oxygen radical-mediated damage to arteries, the potential implications are significant. Excessive exposure of the vasculature to ROS, radical-mediated damage to arteries, and impairment of the endothelial NO system are key components of the pathogenesis of major cardiovascular diseases. The ability of a common safe component of consumable foodstuffs to attenuate such pathological conditions could have far-reaching implications for individuals suffering from cardiovascular disease. That these same compounds enhance the arterial NO system suggests that they could also provide additional benefits with respect to several coronary arterial functions.

In summary, we have undertaken the first studies to examine the potential vasoactive and vasoprotective properties of three anthocyanin-rich preparations composed primarily of the pigments in chokeberry, bilberry, and elderberry. These preparations possess endothelium-dependent relaxation capacity in porcine coronary arteries. At a concentration too low to directly alter coronary vascular tone, they do not alter coronary responses to endogenous or exogenous NO. However, this same low concentration has a significant ability to prevent loss of endothelial-dependent relaxation caused by exposure of arteries to exogenous ROS.


This research was supported by Grant-In-Aids to David R. Bell from the August Tomusk Foundation, The Lutheran Foundation of Indiana, and an Indiana University School of Medicine Research Enhancement Grant.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract