点击显示 收起
1 From the Heart Research Institute, Camperdown, Australia (RS and RAO), and the Centre for Vascular Research, University of New South Wales, Sydney, Australia (RS).
2 Supported by the National Health and Medical Research Council of Australia, the Grape Wine Research Development Council of Australia, and the Australian Wine Research Institute. RRR--Tocopherol and ubiquinone-10 were gifts from Henkel (LaGrange, IL) and Kaneka Corporation (Osaka, Japan). 3 Address reprint requests to R Stocker, Centre for Vascular Research, School of Medical Sciences, University of New South Wales, UNSW Sydney NSW 2052, Australia. E-mail: r.stocker{at}unsw.edu.au..
ABSTRACT
Background: Oxidation of LDL is thought to be important in the development of atherosclerosis. Effective protection against lipoprotein oxidation is achieved by the use of -tocopherol plus coantioxidants—ie, compounds that prevent the prooxidant activity of the vitamin. Wines contain a large number of polyphenols, micronutrients that may act as coantioxidants and may enhance the in vivo antioxidant activity of vitamin E.
Objective: We examined whether wines and wine-derived fractions are able to act synergistically with vitamin E in vitro and whether dealcoholized red wine (DRW) retards the development of atherosclerosis.
Design: Synergy with vitamin E was assessed in vitro by the ability of red and white wines to both attenuate -tocopheroxyl radicals and inhibit in vitro oxidation of LDL in the presence of vitamin E. Female, 6–8-wk-old apolipoprotein E gene-deficient mice were fed a normal nonpurified stock diet for 24 wk to assess the effect on atherosclerosis of DRW at a dose equivalent to 200 mL · 80 kg body wt-1 · d-1.
Results: DRW synergized with vitamin E as effectively as did red and white wine, and phenolic acids accounted for most of this activity. Administration of DRW increased plasma and aortic antioxidants concentrations and the resistance of plasma lipoproteins to ex vivo oxidation. Whereas lipoprotein oxidation in the artery wall was not affected, DRW significantly decreased atherosclerosis in the aortic arch, but not in the root, as assessed by morphometry.
Conclusions: DRW contains polyphenolic compounds capable of synergizing with vitamin E, and long-term moderate consumption of DRW can decrease atherosclerosis in apolipoprotein E gene-deficient mice.
Key Words: Vitamin E • polyphenols • lipoprotein lipid oxidation • tocopheroxyl radicals • mice
INTRODUCTION
A hallmark of early atherosclerosis is the accumulation in the vessel wall of macrophage-derived foam cells—ie, cells laden with lipids derived from LDL. For the lipoprotein to convert macrophages into foam cells, LDL must be modified, and oxidation is thought to be a relevant modification in the pathogenesis of atherosclerosis (1). Many lines of evidence support the "oxidation theory" of atherogenesis (2, 3), although there also are data that are inconsistent with LDL oxidation as an obligatory step in atherosclerosis (4, 5). A potentially important clinical corollary of the oxidation theory is that inhibition of LDL oxidation may also inhibit atherosclerosis independent of lowering plasma cholesterol concentrations. Indeed, some (but not all) antioxidants that attenuate in vitro LDL oxidation also retard atherosclerosis in animal models of the disease (6), although controlled, prospective intervention studies with classic antioxidants, particularly vitamin E, have yielded disappointing results overall (7). A possible reason for these disappointing results is that -tocopherol (the most active form of vitamin E) can exert prooxidant activity on lipoprotein lipids (8), so that supplements with compounds that prevent this prooxidant activity may be more beneficial than supplements with vitamin E alone (9).
Epidemiologic evidence supports the hypothesis that moderate consumption of alcohol reduces mortality due to coronary heart disease (10). The comparatively low incidence of coronary heart disease mortality in France despite a diet high in saturated fat (ie, the French paradox) is commonly explained by the high intake of wine in that country (11). There is also indirect evidence that moderate consumption of wine offers greater cardiovascular protection than does similar consumption of other alcoholic beverages. Part of this protective effect has been attributed to the alcohol component of the wine consumed, although there are conflicting data with regard to the effect of alcohol alone on atherosclerosis in animals (12-14). Both acute and chronic consumption of alcohol increases oxidative stress in vivo (15), and that could counteract a protective effect against atherosclerosis.
Wines contain abundant quantities of polyphenolic compounds that are not necessarily present in other alcoholic beverages (16-18). The concentration of phenolic compounds in red wine is several times greater than that in white wine, and red wine in particular is purported to have a beneficial effect on coronary heart disease. In vitro, red wine polyphenols have several potential antiatherogenic activities. They can protect LDL from oxidation (19-21), inhibit the proliferation of smooth muscle cells (22), and enhance endothelial nitric- oxide synthase activity (23, 24). Unlike alcohol, red wine from which alcohol is removed [ie, "dealcoholized red wine" (DRW)] decreases oxidative stress in vivo, as assessed by plasma and urinary concentrations of isoprostanes (25), and that could have antiatherogenic effects.
In direct support of cardiovascular protection, red wine reduces early atherosclerosis (fatty streak formation) in animals (26-28), although its effect on advanced, clinically more relevant atherosclerosis is less clear. For example, Bentzon et al (29) observed no significant inhibition of atherosclerosis in the aortic root and brachiocephalic trunk of male apolipoprotein E gene-deficient (Apoe-/-) mice by red wine (diluted to 6% ethanol), 6% ethanol, or red wine powder. Apoe-/- mice fed a normal nonpurified stock diet for 6 mo develop atherosclerosis throughout the aortic tree; they had advanced plaques in the aortic root and generally decreasing disease with increasing distance from the aortic origin (30). In the present study, we examined the in vitro effect of wine and wine components on indexes of lipoprotein oxidation, with emphasis on the inhibition of the prooxidant activity of -tocopherol. We also examined the long-term effect of DRW administration, in an amount comparable to the consumption by humans of 1–2 glasses of wine/d, on atherosclerosis in the aortic root and arch in female Apoe-/- mice.
MATERIALS AND METHODS
Materials
The aqueous peroxyl radical generator 2,2-azobis-(2-amidinopropane) hydrochloride was purchased from Wako Pure Chemical Industries (Osaka, Japan), and cholesteryl linoleate (Ch18:2) and arachidonate (Ch20:4) (together referred to as cholesteryl esters), nonesterified cholesterol (NEC), ascorbate, ascorbate oxidase, isoascorbate, sodium hydrogen phosphate, sodium dihydrogen phosphate, formalin, EDTA, glycerol, butylated hydroxytoluene, and chloramphenicol were obtained from Sigma Chemical Company (St Louis). RRR--Tocopherol and ubiquinone-10 were gifts from Henkel (LaGrange, IL) and Kaneka Corporation (Osaka, Japan), respectively. Positively charged cetyltrimethyl ammonium chloride micelles were purchased from Aldrich Chemical (Milwaukee). -Tocotrienol was purified as described (31). Authentic hydroperoxide of Ch18:2 was prepared as described (32) and used as a standard for hydroperoxides and hydroxides of cholesteryl esters (CE-O(O)H). Authentic ubiquinol-10 was prepared fresh (33) and used as a standard for ubiquinol-9 and -10. Protease inhibitor cocktail tablets were from Boehringer Mannheim (Mannheim, Germany), and gentamycin and chloramphenicol were from Gibco BRL (Melbourne). Calcium and magnesium chloride-free Dulbecco's phosphate-buffered saline (Sigma Chemical Company) was prepared from nanopure water and stored over Chelex-100 (BioRad, Richmond, CA) at 4 °C for =" BORDER="0">24 h to remove contaminating transition metals. All buffers were filtered and flushed with argon immediately before use. Organic solvents and all other chemicals were of the highest quality available. Wines were supplied by The Australian Wine Research Institute (Glen Osmond, Australia). The white wine used was a Chardonnay, and the red wine was a Shiraz-Grenache blend. Alcohol was removed from red wine to produce DRW by the "spinning cone column" technique (34). This procedure uses an inert gas to remove volatile alcohol and sulfites from a thin film of wine produced by spinning cones. Other volatile substances removed by the procedure are restored to the wine, and this results in DRW.
Fractionation of wine samples
All wine samples were divided into fractions numbered I–IV by using C18 Sep-pak cartridges (Waters Associates, Milford, MA), as described by Oszmianski et al (35). Accordingly, fraction I contains phenolic acids (caffeoyl tartrate, coumaroyl tartrate, caffeic acid, and coumaric acid); fraction II contains catechins, procyanidins, and anthocyanin monomers; fraction III contains flavonols (myricetin 3-glucoside, quercetin 3-glucoside, and quercetin); and fraction IV contains anthocyanin polymers. Fractions were stored at 4 °C and analyzed within 1 wk of preparation. Preliminary studies showed that such storage did not affect the in vitro activities determined (data not shown). The total phenolic content of each fraction was standardized to gallic acid equivalents by using the Folin-Ciocalteu reagent according to Singleton and Rossi (36). For all 3 types of wine, 75% of the starting gallic acid equivalent was recovered in the 4 fractions.
Synergistic activity of wine and wine-derived fractions with vitamin E
Measurement of the ability of various wines to attenuate the -tocopheroxyl radicals (37) was used to screen selected wines and the fractions derived from them for their ability to synergize with vitamin E in vitro. Briefly, this assay uses a chemically defined method for assessing the ability of a test compound or mixture to attenuate the -tocopheroxyl radical signal generated in positively charged micelles and measured directly by electron paramagnetic resonance spectroscopy. Ascorbate (which readily reduces the -tocopheroxyl radicals) is used at a final concentration of 10 µmol/L as a positive control (37). The rate of decay of the radical signal was expressed as the natural log of the percentage of the initial signal. All wine samples and their respective fractions were diluted serially with 50 mmol phosphate buffer (pH 7.4)/L; the buffer was previously treated with Chelex-100 (38) and analyzed in triplicate. Each sample, at an appropriate dilution, was treated with ascorbate oxidase to completely remove endogenous ascorbate, which was measured directly by using HPLC with electrochemical detection (39). White wine contained 30.2 ± 6.1 mg ascorbic acid/L (n = 3), whereas red wine and DRW contained none.
Antioxidant efficacy of wines
The ability of wine-derived compounds to synergize with -tocopherol in human LDL was examined by exposing the lipoprotein to mild and chemically controlled oxidation initiated by 2,2-azobis-(2-amidinopropane) hydrochloride and by comparing the extent of accumulation of CE-O(O)H in the absence of the putative antioxidant with that in its presence (40). Antioxidant efficacy was assessed at the time point corresponding to 20% consumption of endogenous -tocopherol in the control sample (absence of test compound). The anti-tocopherol-mediated peroxidation index defines the antioxidant efficacy of a test compound by expressing the oxidation in the presence of the compound as a percentage of the oxidation in the absence of the compound. Wine fractions were tested in triplicate at 10 µmol gallic acid equivalents/L with the use of 3 separate LDL preparations, and 10 µmol butylated hydroxytoluene/L was included for every experiment as a positive control (40).
Apolipoprotein E gene-deficient mice
Breeding pairs of C57BL/6J mice that were homozygous for the disrupted Apoe gene (ie, were Apoe-/-) were purchased from Jackson Laboratories (Bar Harbor, ME) and bred at The Heart Research Institute (Sydney, Australia) while consuming a nonpurified stock diet (Lab-Feed, Sydney, Australia). All animals used for the present study were confirmed by genotyping as being Apoe-/-. At 6–8 wk of age, mice (n = 87) were randomly assigned to 2 groups and given water (control group: n = 38) or DRW (treatment group: n = 49). DRW was added to the drinking water at an appropriate dilution, so that the amount of DRW consumed by each mouse was approximately equivalent to the consumption of 200 mL (or 1 glass)/d by an 80-kg person. Food and fluid were changed 3 times weekly, and the consumption of both was monitored, as was body weight, throughout the study. The study was approved by the animal ethics committee of the Central Sydney Area Health Services.
Plasma analyses
Plasma was prepared from blood as described elsewhere (41). Aliquots (50 µL) were extracted immediately into hexane and methanol (42) and stored at -80 °C for analysis of lipids (NEC, Ch20:4, and Ch18:2) and lipid-soluble antioxidants or diluted in 5% metaphosphoric acid (1:1; by vol) and stored at -80 °C for ascorbate analysis (39). Total cholesterol and triacylglycerol concentrations were measured by using the respective enzymatic assay kits (Sigma Diagnostics, St Louis). The residual plasma was pooled, flushed with argon, and stored at -80 °C before ex vivo oxidation. Such storage does not significantly alter the capacity of plasma lipid for oxidation (41).
Ex vivo plasma oxidation
Pooled plasma obtained from animals of the control and treatment groups was exposed to 2,2-azobis-(2-amidinopropane) hydrochloride (final concentration: 10 mmol/L) under air and at 37 °C. The 2,2-azobis-(2-amidinopropane) hydrochloride decomposes to yield aqueous peroxyl radicals at a constant rate (43). Samples (100 µL) of the reaction mixture were taken at the times indicated, extracted, and analyzed with the use of reversed-phase HPLC for antioxidants, nonoxidized lipids, and CE-O(O)H as an index of plasma lipid peroxidation (32, 41, 44). The various compounds were quantified by the comparison of their area with that of authentic standards.
The length of time before the onset of rapid accumulation of CE-O(O)H and the rate of rapid accumulation of CE-O(O)H were 2 additional indexes determined graphically by using this assay. CE-O(O)H accumulation is linear with time only during the rapid phase of lipid peroxidation, which occurs =" BORDER="0">3 h after the initiation of lipid peroxidation and in the presence of -tocopherol. The data for the first 3 h were omitted to allow graphic determination of the values for the rate and lag time. The rate was calculated from the slope of the line graph of time-dependent accumulation of CE-O(O)H, and the lag time was determined as the x intercept of the slope.
Tissue collection for biochemical and histologic analyses
Procedures were carried out largely as described previously (41, 44, 45). Briefly, mice were perfused at near-physiologic pressure with buffer A (phosphate-buffered saline containing 1 mmol EDTA/L and 20 µmol butylated hydroxytoluene/L). For biochemistry, aortas (22 from control mice, 31 from treated mice) were excised, and hearts, ascending aortas, and descending aortas past the femoral bifurcation were cleaned. Aortas were then randomly sorted into 3 groups of 6–8 aortas each and placed immediately in cold buffer B (buffer A containing 1 protease inhibitor tablet/150 mL, 0.008% gentamycin and 0.008% chloramphenicol). Aortas were then stored at -80 °C until they were processed for biochemical analyses.
For histologic analyses, the hearts and aortic arches of individual mice (n = 16 from each group) were excised, placed in 4% (by vol) formaldehyde in saline overnight, and than transferred into 0.1% (by vol) formaldehyde in saline solution. Fixed tissue was used for lesion assessment, which was performed in a blinded fashion by morphometry at 2 sites. Lesion size at the aortic root was calculated as the average of the cross-sectional lesion areas 200 and 300 µm distal to the point where the 3 valve leaflets first appeared (41, 44, 45). Measurement of lesion development at the aortic arch was carried out as described previously by others (46). Briefly, 20–30 longitudinal sections were cut from the arch of each mouse. Lesion size was estimated by identifying a 3-mm-long segment along the inner curvature of the arch. The area and maximal thickness of the inner aortic arch wall of each section were then averaged for each mouse, and the group value was the mean of those values.
Aortic biochemistry
Pooled aortas were pulverized in liquid nitrogen, resuspended in 2 mL buffer B. -Tocotrienol and isoascorbate (final concentrations: 5 µmol/L) were added as internal standards before the mixture was homogenized (41). An aliquot (50 µL) of homogenate was added to an equal volume of 5% (by wt) metaphosphoric acid for ascorbate analysis, and a further 50 µL was removed for use in determining the protein content using the bicinchoninic acid assay kit (Sigma Diagnostics). The remaining homogenate was extracted in 500-µL aliquots by addition to hexane:methanol (5:1 by vol), the sample was then mixed vigorously for 1 min and centrifuged at 4 °C, the hexane phase was removed and dried, and the lipids were redissolved into 400 µL isopropylalcohol (42). This extract was then analyzed for NEC, cholesteryl esters, lipid hydroperoxides (LOOH), -tocopherol, and total coenzyme Q as described above. LOOH, measured as HPLC postcolumn chemiluminescence-positive peaks, represent primary and major lipid peroxidation in lipoproteins from Apoe-/- mice (47). A previous study showed that 70% of [3H]Ch18:2-OOH added to mouse aorta before pulverization is recovered as the hydroperoxide, and 30% is recovered as the corresponding alcohol, and that [3H]Ch18:2 added to aortas before work-up is not converted to [3H]Ch18:2-OOH or [3H]Ch18:2-OH (41). All compounds detected were quantified by peak area comparison with authentic standards run under identical conditions.
Statistical analysis
Data are expressed as means ± SDs. When appropriate, one-way analysis of variance or unpaired Student's t tests were used for group comparisons. Statistical significance was set at P < 0.05. Line graphs were compared by using two-way analysis of variance. Comparisons of the effect of supplementation with DRW on lesion area and thickness were performed by using two-way analysis of variance and Bonferroni post hoc tests with significance set at the 95% CI. SYSTAT software (version 8.0; SPSS Inc, Chicago) was used for calculations.
RESULTS
In vitro studies
All wines contained compounds capable of reducing -tocopheroxyl radicals (Figure 1 A), the 1-electron oxidation product of vitamin E. This activity is a prerequisite for synergization of compoundds with vitamin E (37, 40). The radical-reducing activity of all wines was significantly greater than that of water, and the radical-reducing activity did not differ significantly between DRW and red wine (P = 0.7), which indicated that the activity was independent of alcohol (Figure 1A). DRW and red wine did not contain ascorbate. Their respective radical-reducing activities were not altered by treatment of the sample with ascorbate oxidase (data not shown). White wine showed significantly less radical-reducing activity than did red wine and DRW (Figure 1A). White wine contained ascorbate (data not shown), and treatment with ascorbate oxidase tended to decrease the activity (Figure 1A), but this decrease was not significant. Results from fractionation of wines indicated that phenolic acids contained in fraction I (caffeoyl tartrate, coumaroyl tartrate, caffeic acid, and coumaric acid) made the greatest contribution to the radical-reducing activity of red wine and DRW (Figure 1C and 1D). Thus, fraction I showed significantly more activity than did the corresponding fractions II, III, and IV. For white wine, the radical-reducing activity of fractions I and II (the latter containing procyanides, catechins, and anthocyanin monomers) did not differ significantly from each other, although both fractions showed significantly more activity than did fractions III and IV (Figure 1B).
FIGURE 1.. Mean (±SD) -tocopheroxyl radical-attenuating ability of wines and their fractions for 3 independent experiments in which separate samples were analyzed once. DRW, dealcoholized red wine; RW, red wine; WW, white wine; AO, ascorbate oxidase; FI, fraction I; FII, fraction II; FIII, fraction III; FIV, fraction IV. All wine samples were used at a 1:1000 dilution (A). -Tocopheroxyl radical-attenuating ability of fractionated WW (B), RW (C), and DRW (D) is shown, with ascorbate (10 µmol/L) used as a positive control (data not shown). All wine fractions were used at 4 µmol gallic acid equivalents/L. Results for RW and DRW were significantly different (P < 0.05) from those for controls, WW, and WW+AO (A). For RW (C) and DRW (D), FI contained significantly more -tocopheroxyl radical-attentuating ability than did the corresponding FII, FIII, and FIV (P < 0.05), whereas for WW (B), -tocopheroxyl radical-attentuating ability did not differ significantly between FI and FIII, although both of these fractions showed significantly more activity than did the corresponding FII and FIV (P < 0.05).
We next examined the ability of wine-derived compounds to act synergistically with vitamin E in LDL by using oxidizing conditions under which lipoprotein lipid peroxidation proceeds via tocopherol-mediated peroxidation (40). Antioxidant efficacy is described by the anti-tocopherol-mediated peroxidation index, with low values reflecting high antioxidant activity. Table 1 shows anti-tocopherol-mediated peroxidation indexes of fractions I–IV of each wine. The results obtained confirm that inhibition of in vitro lipid peroxidation in LDL under conditions in which -tocopheroxyl radicals mediate oxidation was observed with and restricted to the wine fractions that contain most of the tocopheroxyl radical-reducing activity. When tested at comparable gallic acid equivalents (ie, at a comparable content of phenols), fraction I of DRW had the highest antioxidant activity (lowest anti-tocopherol-mediated peroxidation index), and fraction I of red wine and white wine had the next highest activity. All other fractions had significantly less antioxidant activity. Therefore, DRW and, to a somewhat lower extent, red wine and white wine contain a compound or compounds associated primarily with phenolic acids that can interact with lipoprotein-associated vitamin E to increase the resistance of lipoprotein lipids to oxidation in the presence of -tocopherol.
View this table:
TABLE 1. Anti-tocopherol–mediated peroxidation (TMP) indexes of wine fractions1
In vivo studies
All animals remained well during the 24 wk of intervention, except for one control mouse that was eliminated early because of failure to thrive. There were no significant differences between control and treatment groups in fluid intake (3.6 ± 0.5 and 3.6 ± 0.6 mL · mouse-1 · d-1, respectively) or weight gain (final body wt: 23.9 ± 0.5 and 23.5 ± 0.5 g, respectively; P > 0.05 for both comparisons).
The effect of 24 wk of DRW supplementation on plasma lipids and antioxidants in female Apoe-/- mice is summarized in Table 2. As can be seen, regular consumption of DRW for 24 wk resulted in a slight (5%) but significant (P = 0.001) decrease in the plasma concentrations of triacylglycerols, whereas concentrations of NEC, Ch18:2, and Ch20:4 were not altered significantly. Of the antioxidants measured, the concentrations of ascorbate and -tocopherol also were not altered significantly, whereas plasma total coenzyme Q (defined as the sum of ubiquinone-9, ubiquinone-10, ubiquinol-9, and ubiquinol-10) increased significantly in animals receiving DRW (P = 0.02).
View this table:
TABLE 2. Plasma lipid and antioxidant concentrations in female apolipoprotein E gene–deficient mice after 24 wk of intervention1
By exposing pooled plasma to peroxyl radicals and following the time-dependent disappearance of antioxidants and accumulation of CE-O(O)H, we next examined the effect of DRW on ex vivo measures of lipoprotein oxidation (Figure 2). Consistent with the results obtained with individual plasma samples, there were no significant differences between the 2 groups of mice in the initial concentrations of ascorbate (Figure 2A) and -tocopherol (Figure 2C), whereas plasma pooled from mice receiving DRW had a significantly (P = 0.03) higher starting concentration of ubiquinols than did plasma from control mice (Figure 2B). Oxidation resulted in the time-dependent consumption of ascorbate and ubiquinols that was slightly but significantly slower in plasma obtained from DRW-treated mice than in that from control mice (P = 0.03 and 0.04 for ascorbate and ubiquinols, respectively; Figure 2 A and B). CE-O(O)H were not measurable during this period of oxidation. Consistent with this observation, the time elapsing before the onset of CE-O(O)H accumulation was significantly (P = 0.05) greater in plasma from treated animals than in that from control animals (3.5 ± 0.2 compared with 4.3 ± 0.2 h; P = 0.05; Figure 2D). CE-O(O)H accumulated initially at a linear rate as -tocopherol was consumed. It was surprising that this rate was significantly (P = 0.03) higher in plasma from treated mice than in that from control mice (6.2 ± 1.9 compared with 3.0 ± 0.3 µmol · L-1 · h-1), and that finding may reflect the ability of polyphenolic acid-derived peroxyl radicals, similar to that of -tocopherol (8), to promote lipoprotein lipid peroxidation (48).
FIGURE 2.. Mean (±SD) 2,2-azobis-(2-amidinopropane) hydrochlorideinduced oxidation of plasma samples from apolipoprotein E gene–deficient mice after 24 wk of intervention. Plasma aliquots from control mice (n = 38; filled symbols) and mice receiving dealcoholized red wine (n = 49; empty symbols) were pooled into 3 pools for each group and then individually exposed to 2,2-azobis-(2-amidinopropane) hydrochloride to initiate oxidation. At the time points indicated, aliquots of the reaction mixture were removed and analyzed for ascorbate, ubiquinol-9 and ubiquinol-10, -tocopherol, and cholesterylestes hydroperoxides (CE-OOH), as described in Materials and Methods. The inset in panel D shows HPLC postcolumn chemiluminescence (CL) chromatograms of plasma extracts from control (upper trace) and dealcoholized red wine–treated (lower trace) mice after 4 h of oxidation. The large positive peak eluting with retention time (RT) 4–5 min represents the solvent front, whereas the CL-negative peak at 5.5 min is due to tocopherols. The arrow indicates elution of hydroperoxides of cholesteryl esters at 9.8 min in the control mouse plasma only. The starting concentration of ubiquinols was significantly (P = 0.03) lower in treated samples than in control samples, and the lines of consumption of ascorbate and ubiquinols differed significantly between the treatment and control groups (P = 0.03 and 0.04, respectively).
The aortic concentrations of lipids and lipid-soluble antioxidants in control and treated animals after 24 wk intervention are shown in Table 3. After treatment with DRW, the aortic concentrations of NEC and the 2 major oxidizable cholesteryl esters, Ch18:2 and Ch20:4, tended to be greater, although this increase was not significant. Similarly, concentrations of all of the antioxidants measured tended to be higher in the treatment group than in the control group, and the results for coenzyme Q were significant (P = 0.01). Between 20% and 50% of aortic coenzyme Q was present as ubiquinol-9 or -10 (data not shown), which indicated that the sample work-up procedure used largely prevented inadvertent oxidation, because ubiquinols are more sensitive to autoxidation than is -tocopherol. Despite this, the tissue concentrations of primary products of oxidized lipoproteins in treated mice were not lower than those in control mice (Table 3). In fact, there was a trend toward greater aortic concentrations of LOOH in tissue from treated animals than in that from control animals, although this difference was not significant.
View this table:
TABLE 3. Aortic lipids and antioxidants in female apolipoprotein E gene–deficient mice after 24 wk of intervention1
As expected, there were lesions of substantial size at the aortic sites examined after 24 wk of stock diet. In the aortic root, all mice had advanced atherosclerotic lesions that contained cholesterol crystals and resembled advanced human lesions (data not shown). There were no differences in gross morphology between the 2 groups (not shown). DRW supplementation of mice decreased the mean values of the measurements used to assess atherosclerosis at both sites by 20% (Figure 3). This difference was not significant (P = 0.41) for the lesion area at the aortic root (312.6 ± 48.6 and 245.4 ± 57.1 µm2 for control and treated mice, respectively). The difference was significant, however, for both lesion area (309.7 ± 55.9 and 257.3 ± 44.1 µm2 in control and treated mice, respectively; P = 0.03) and wall thickness (263.1 ± 38.6 and 211.3 ± 46.1 µm, respectively; P = 0.026; Figure 3) at the aortic arch. Lesions at the arch of control and treated animals are shown in Figure 4.
FIGURE 3.. Atherosclerosis in control and dealcoholized red wine–treated mice. Lesion assessment was carried out as described in Materials and Methods. The lesion area in the aortic sinus of control (n = 15) and treated (n = 15) mice was measured. The wall thickness and lesion area of the aortic arch were measured as described previously (46) in control (n = 13) and treated (n = 15) mice, by using the average of 20–30 sections prepared per animal. Aortic arch lesion areas and wall thickness were significantly (P = 0.03, ANOVA) less in treated mice than in control mice.
FIGURE 4.. Lesions at the aortic arch from control (top) and dealcoholized red wine–treated (bottom) apolipoprotein E gene–deficient mice after 24 wk of intervention (4x objective lens). The arrows point to the inner curvature of the arch, where lesions were assessed as summarized in Figure 3.
DISCUSSION
Epidemiologic studies consistently report a decrease in cardiovascular disease with regular consumption of moderate amounts of alcoholic beverages, such as wine (10), although the scientific literature on the effect of red wine on atherosclerosis in animals is controversial. A beneficial effect of red wine was reported for lesions at early developmental stages (26-28), whereas no effect was found for the formation of complex, more advanced lesions in Apoe-/- mice (29). When a benefit of red wine was shown, the underlying mechanism remained unclear. The present study shows that long-term, regular administration of moderate amounts of DRW slightly but significantly decreased lesion sizes in the aortic arch, whereas it had no significant effect on lesions in the aortic root. Such administration of DRW also had small but measurable antioxidant effects, but it did not decrease the extent of lipoprotein oxidation in the vessel wall. Overall, our study supports the notion of a moderately beneficial effect of regular consumption of DRW on atherosclerosis in Apoe-/- mice. It also indicates, however, that this beneficial effect does not result from the inhibition of lipoprotein oxidation in the vessel wall.
In mouse models of atherosclerosis, assessment of disease severity is often limited to the measurement of the intimal area in the aortic root, where lesions are most advanced and complex. However, the effect of a particular intervention can be site-specific, as is perhaps best exemplified by the case of probucol, which promotes disease in the aortic root but inhibits atherosclerosis with increasing efficacy the further distal from the aortic origin (45). The reason or reasons for these different effects are unclear at present, but the very differences between the effects highlight the potential pitfall of drawing conclusions from observations made at a single aortic site. In the present study, we therefore measured lesions at both the aortic root and arch. We observed a significant, beneficial effect of DRW supplements for 24 wk on atherosclerosis in the arch but not in the root (Figure 3). It is possible that this difference reflects the ability of DRW to attenuate relatively early lesions but not more advanced lesions, which is consistent with existing reports in the literature. Disease progression in Apoe-/- mice is cephalad, so that lesions in the aortic root are more developed than are lesions in the arch. However, a simpler interpretation of our results is that DRW decreased disease at both sites, which is consistent with a 21.5% and 16.9% decrease in the mean lesion area at the aortic root and arch, respectively. We analyzed only 2 sections per mouse at the aortic root, but 20–30 sections per mouse at the arch, so that the accuracy of atherosclerosis assessment at the root may not have been sufficient to ascertain a significant difference between control and DRW-treated animals. In any case, our results show that long-term consumption of moderate amounts of DRW is at best associated with a small decrease in atherosclerosis in Apoe-/- mice.
Wines in general, and red wines in particular, contain a large number of redox-active polyphenols that have been unambiguously shown to exert antioxidant activity in vitro. It remains less clear, however, whether such antioxidant activity can be observed in vivo, and, if so, whether it contributes to potential biological activities, such as those that can be linked to an inhibition of atherosclerosis. In the present study, we focused on the ability of wine components to synergize with lipoprotein-associated vitamin E in the context of the oxidation theory of atherosclerosis. Our results clearly show that the phenolic acid-containing fraction I of wines is enriched with compounds capable of attenuating both -tocopheroxyl radicals and in vitro LDL oxidation in the presence of vitamin E. This activity may be relevant because, even in advanced human atherosclerotic lesions, most of the oxidized lipoprotein lipids are generated in the presence of -tocopherol (49), and vitamin E does not become oxidized to a substantial extent (50). In addition, previous interventions established that agents that inhibit tocopherol-mediated peroxidation in vitro can also inhibit lipoprotein lipid peroxidation in the vessel wall (4, 44, 51). In contrast with these studies, our study did not find that supplementation with DRW inhibited lipoprotein lipid peroxidation in the vessel wall.
We did, however, observe a slight albeit significant antioxidant protection, as seen in the ex vivo plasma oxidation experiments and, as implied, in the greater endogenous concentrations and redox status of coenzyme Q in DRW-treated mice than in control mice. Others have reported that the consumption of DRW by humans decreased oxidative stress, as assessed by decreased concentrations of plasma and urinary isoprostanes (25). These results indicate effective absorption of phenolics, although we did not measure it directly in the present study. Absorption of catechin and quercetin, the major polyphenols of red wine, was shown in Apoe-/- mice (26). The absence in the present study of direct evidence of an antioxidant protection of vessel wall indexes that was related to lipoprotein lipid oxidation suggests that the bioavailability of DRW-derived antioxidants is limited overall. This may be particularly relevant for sites other than blood, such as the vessel wall, and in considerations of the relatively minor contribution these antioxidants may make to the existing endogenous antioxidant defense.
The observed moderate decrease in atherosclerosis along with the lack of effect on the arterial content of LOOH indicates that the protective activity of DRW may be independent of inhibition of lipoprotein oxidation in the vessel wall. Lipoprotein oxidation in the vessel wall is likely more relevant to atherogenesis than is that in the circulation (45, 52), and a large body of literature supports the oxidation theory of atherosclerosis. However, there is increasing evidence that the 2 processes are not causally related (4, 9, 45, 53). The present study with DRW adds to the evidence dissociating arterial lipoprotein oxidation from atherogenesis, and this notion is further supported by the recent finding that some products of aortic fatty acid oxidation—ie, hydroxyeicosatetraenoic acids and F2-isoprostanes—do not reflect the extent of atherosclerosis in Apoe-/- mice receiving a high-fat diet (54). Therefore, potential beneficial activities of DRW other than inhibition of lipoprotein oxidation should be considered. Two recent publications indicated that nitric oxide generation by endothelial cells treated with red wine polyphenols is enhanced (23, 24), although the in vivo relevance of this is presently unclear. Other activities may also be important. We observed that the administration of DRW slightly but consistently decreased plasma concentrations of triacylglycerols (Table 2), and this effect could be beneficial.
A limitation of the present study is that we used DRW rather than alcohol-containing wine. We chose DRW because it had antioxidant activities similar to those of red wine (Figure 1 and Table 1) and because its administration inhibits in vivo lipid peroxidation (25), whereas alcohol consumption increases that process (15) and thus may contribute to atherogenesis (55). Our results, which were obtained with DRW, cannot be directly extrapolated to the effect of red wine consumption, so future studies are needed to establish whether the epidemiologic evidence linking moderate consumption of red wine to decreased cardiovascular disease is based on a direct beneficial effect of the beverage on atherosclerosis.
ACKNOWLEDGMENTS
We thank Gary Martinic for assistance with animal care and maintenance, CoVien Phu for in vitro experiments, and Jing Y Hou for histologic analyses. We also thank Creina Stockley, of The Australian Wine Research Institute, Glen Osmond, Australia, for provision of the wine samples and enthusiastic support throughout the study.
RS was responsible for study design, writing the manuscript, and, in part, for data analysis. RAO was responsible for data collection and analysis. Neither of the authors had a conflict of interest.
REFERENCES