Al

HGa

b

c

d

a

ARRAA

KARAC

1

ca(lptpeCer

S

0d

Atherosclerosis 207 (2009) 420–427

Contents lists available at ScienceDirect

Atherosclerosis

journa l homepage: www.e lsev ier .com/ locate /a therosc leros is

new insight into resveratrol as an atheroprotective compound: Inhibition ofipid peroxidation and enhancement of cholesterol efflux

icham Berrougui a,b,c, Guillaume Grenier a,d, Soumaya Loued a,b,eneviève Drouin a,d, Abdelouahed Khalil a,b,∗

Research Center on Aging, CanadaDepartment of Medicine, Geriatrics Service, Université de Sherbrooke, CanadaUniversity of Sultan Moulay Slimane, Department of Biology, Beni Mellal, MoroccoDepartment of Orthopedic Surgery, Faculty of Medicine, Université de Sherbrooke, Sherbrooke, QC, Canada

r t i c l e i n f o

rticle history:eceived 19 February 2009eceived in revised form 16 April 2009ccepted 14 May 2009vailable online 22 May 2009

eywords:therosclerosisesveratrolntioxidant

a b s t r a c t

Resveratrol, a polyphenolic constituent of red wine, is known for its anti-atherogenic properties and isthought to be beneficial in reducing the incidence of cardiovascular diseases (CVD). However, the mecha-nism of action by which it exerts its anti-atherogenic effect remains unclear. In this study, we investigatedthe relationship between the antioxidant effects of resveratrol and its ability to promote cholesterol efflux.We measured the formation of conjugated dienes and the rate of lipid peroxidation, and observed thatresveratrol inhibited copper- and irradiation-induced LDL and HDL oxidation as observed by a reduc-tion in oxidation rate and an increase in the lag phase (p < 0.05). We used DPPH screening to measurefree radical scavenging activity and observed that resveratrol (0–50 �M) significantly reduced the con-tent of free radicals (p < 0.001). Respect to its effect on cholesterol homeostasis, resveratrol also enhanced

2

holesterol efflux apoA-1-mediated cholesterol efflux (r = 0.907, p < 0.05, linear regression) by up-regulating ABCA-1 recep-tors, and reduced cholesterol influx or uptake in J774 macrophages (r2 = 0.89, p < 0.05, linear regression).Incubation of macrophages (J774, THP-1 and MPM) with Fe/ascorbate ion, attenuated apoA-1 and HDL3-mediated cholesterol efflux whereas resveratrol (0–25 �M) significantly redressed this attenuation in adose-dependent manner (p < 0.001). Resveratrol thus appears to be a natural antioxidant that enhancescholesterol efflux. These properties make it a potential natural antioxidant that could be used to preventand treat CVD.

. Introduction

Natural compounds have been used to regulate serum lipidoncentrations to reduce the incidence of hyperlipidemia andtherosclerosis, which are responsible for cardiovascular diseasesCVD) [1]. There has been a recent focus on certain polypheno-ic compounds as possible hypolipidemic agents. Resveratrol, aolyphenolic compound, has been reported to have atheropro-ective properties [2,3]. Since resveratrol is a natural polyphenolresent in red wine, it has been suggested that the antioxidant prop-

rties of resveratrol are responsible for the protective effect againstVD of consuming moderate amounts of red wine. A number ofpidemiological and animal studies have confirmed the ability ofed wine polyphenols to inhibit atherosclerotic progression, even

∗ Corresponding author at: Research Center on Aging, 1036 rue Belvédère Sud,herbrooke, QC, Canada J1H 4C4. Tel.: +1 819 821 1170x45284; fax: +1 819 829 7141.

E-mail address: Abdelouahed.Khalil@USherbrooke.ca (A. Khalil).

021-9150/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.atherosclerosis.2009.05.017

© 2009 Elsevier Ireland Ltd. All rights reserved.

if alcohol intake it self raises high-density lipoprotein (HDL) levels[4]. Animal studies have provided stronger evidence of the positiveeffect of wine polyphenols on plasma lipids. For instance, non-alcoholized red wine increases the plasma concentration of HDLin rats [5]. In addition, red wine polyphenols have been shown toreduce total plasma cholesterol levels in hamsters [6]. Resveratrolhas also been investigated for its antioxidant [7], platelet aggre-gation inhibition [8], smooth muscle cell proliferation inhibition[9], and plasma cholesterol level modulation activities [10]. Resver-atrol also induces LXR-� expression in human monocyte-derivedmacrophages and represses the expression of the lipid uptake genesLPL and SR-AII [11].

Macrophage cholesterol accumulation and foam cell forma-tion are the hallmarks of early atherogenesis [12]. Low-density

lipoprotein (LDL) can be taken up and oxidized by macrophages[13], resulting in a significant increase in macrophage choles-terol mass [12]. Macrophages can also accumulate cholesterol byincreasing the rate of cholesterol biosynthesis and/or decreasingthe rate of HDL-mediated cholesterol efflux. HDL is considered anti-

oscler

a[tmbfatckotpsrmtelftima

roaaoesc[Aotaiilo

2

2

npTe(limc(C(ARIs

H. Berrougui et al. / Ather

therogenic and plays a key role in protecting LDL against oxidation14] and maintaining cholesterol homeostasis by reverse choles-erol transport (RCT). As suggested by Glomset [15], RCT involves the

ovement of cholesterol from peripheral tissues to the liver, whichegins with the transfer of free cholesterol (FC) and phospholipids

rom peripheral tissue cells to lipid-poor or lipid free (unassociated)polipoprotein A1 (apoA-1) and HDL3 [15,16]. This process is ini-iated by cholesterol efflux, a mechanism by which HDL removesholesterol excess from macrophages. FC efflux occurs by threenown pathways: (1) aqueous diffusion, which involves the des-rption of FC molecules from the donor lipid–water interface andhe diffusion of these molecules through the intervening aqueoushase until they collide with and are adsorbed by an acceptor; (2)cavenger receptor class B type I (SR-BI)-mediated FC flux in a bidi-ectional manner. Like the aqueous diffusion mechanism, the net

ovement of FC via SR-BI depends on the direction of the choles-erol gradient [17]; (3) ATP binding cassette-mediated cholesterolfflux. ABCA1 promotes the transfer of cholesterol and phospho-ipids to lipid-poor apoA-1 [18]. In addition to cholesterol effluxrom arterial wall cells, ABCA1 is primarily responsible for the initia-ion of HDL formation, principally in the liver and, to a lesser extent,n the small intestine [19]. ABCG1 promotes cholesterol efflux from

acrophage foam cells and their transfer to HDL particles, but thisctivity has no influence on overall HDL levels [20].

Despite several studies indicating that polyphenols such asesveratrol play a role in reducing and preventing the progressionf atherosclerosis, little is known about their effect on RCT or theirntioxidant mechanisms. Resveratrol was reported to protect LDLgainst ferrylmyoglobin, peroxynitrite, copper or AAPH-inducedxidation [21,22], however, the kinetic of resveratrol-antioxidantffect on the lipoproteins as well as on the cells oxidation systemtill poorly explored. Moreover, effect of resveratrol on RCT pro-ess still poorly investigated, since only one study of Sevov et al.11], have reported that resveratrol modulated LXR, ABCA1 andBCG1 mRNA levels in THP1-derived macrophage. The purposef the present study was to elucidate the mechanism underlyinghe anti-atherogenic properties of resveratrol by investigating itsntioxidant effect on lipoprotein particles, its free radical scaveng-ng activity, and especially its effect on cholesterol homeostasisn several cell lines and the relationship between prevention ofipoproteins and cells from oxidation by resveratrol, and its impactn the cholesterol efflux.

. Materials and methods

.1. Chemicals and cell lines

Acetic acid, sulfuric acid, sodium phosphate, thiobarbituric acid,-butanol, methanol, ethanol, n-isopropanol, and hexane wereurchased from Fisher Scientific (Montreal, QC, Canada). 1,1,3,3-etraethoxypropane, d-�-tocopherol, cupric sulfate (CuSO4),thylenediaminetetraacetic acid (EDTA), 2-�-mercaptoethanol2-ME), phorbol myristate acetate (PMA), 1�, 2�(n)-3H cholesterol,-glutamine, DPPH (1,1-diphenyl-2-picryl-hydrazyl), methylth-azoletetrazolium (MTT), apoprotein-A1, 8-Br-cyclic adenosine

onophosphate (cAMP), bovine serum albumin (BSA), thiogly-ollate, and dimethylsulfoxide (DMSO) were from Sigma–AldrichSt. Louis, MO, USA). Resveratrol was from Calbiochem (La Jolla,A, USA). Dialysis bags were from Spectrum Medical Industries

Houston, TX, USA). The J774 and THP-1 cells were from themerican Type Culture Collection (ATCC) (Manassas, VA, USA). ThePMI 1640 and Dulbecco’s-modified medium (DMEM) were from

nvitrogen Canada Inc (Burlington, Ont. Canada). The fetal bovineerum (FBS) was from Wisent Inc. (St-Bruno, QC, Canada).

osis 207 (2009) 420–427 421

2.2. Measurement of free radical scavenging activity

The free radical scavenging activity of resveratrol was measuredusing the DPPH method as previously described [23]. Briefly, a0.1 mM DPPH solution in ethanol was added to resveratrol dissolvedin DMSO at various concentrations (0–50 �M). Reactions were per-formed at room temperature and the absorbance at 518 nm wasmeasured every 30 min for 3 h. Vitamin E (�-tocopherol, 10 and20 �M) was used a positive control while DMSO was used as anegative control. The antioxidant activity was calculated using thefollowing formula:

AA% = 100 −(

Abssample

Abscontrol

)× 100

2.3. Lipoprotein preparation

Human plasma was collected from healthy volunteers (aged20–25) with normal blood pressure, glycemia, and lipid profiles. Theethics committee of the Sherbrooke Geriatric University Instituteapproved the study, and all subjects provided written informed con-sent. The LDL and HDL3 subfractions were obtained by sequentialultracentrifugation as described previously [24]. Isolated lipopro-teins were dialyzed overnight at 4 ◦C in 10 mM sodium phosphatebuffer (pH 7.0). Protein concentrations were measured using theBradford method according to the manufacturers’ instructions (Bio-Rad. Mississuga, Ont., Canada).

2.4. Lipoprotein oxidation

2.4.1. Copper-mediated lipoprotein oxidationThe induction of LDL and HDL3 peroxidation was performed as

previously described using transition metal ions as oxidizing agents[25]. Briefly, lipoproteins (100 �g/ml of LDL or 200 �g/ml of HDL3)were suspended in 10 mM sodium phosphate buffer (pH 7.0) andwere incubated in a 10 �M CuSO4 solution containing 0–25 �Mresveratrol for 0–8 h. The reactions were stopped at 4 ◦C adding100 �M EDTA.

2.4.2. LDL oxidation by �-radiolysisThe LDL was oxidized by exposure to oxygen free radi-

cals produced by �-radiolysis using a 60cobalt gamma cell 220(Atomic Energy of Canada, Mississauga, Ont., Canada) as previouslydescribed [25]. Water �-radiolysis makes it possible to accuratelyestimate the nature and quantity of the free radicals that react withLDL, unlike commonly used techniques such as incubating cellswith transition metal ions. The dose rate was 0.13 Gy/s [26]. Totalradiation doses varied from 0 to 150 Gy.

2.4.3. Conjugated diene formationLDL or HDL3 oxidized alone or in the presence of various con-

centrations of resveratrol (0–1 �M) were continuously monitoredat 234 nm to detect the formation of conjugated dienes as describedelsewhere [27].

2.4.4. Kinetic profile of LDL oxidationThe kinetic profile of lipid peroxidation was characterized using

mathematical parameters such as the lag phase and the propa-gation phase, i.e., the phase with the maximum oxidation rate(Vmax). These parameters were determined as previously described

by Pinchuk and Lichtenberg [28].

2.4.5. LDL electrophoresisThe electrophoresis mobility of LDL was used as an indication

of Apo-B oxidation and was measured using a Titan gel lipopro-

4 oscler

tUbadm

2

iwM1isaotcaisnmm

2

pwbifc

ptTfwrastmca1msPwci[

2

J(act

22 H. Berrougui et al. / Ather

ein electrophoretic system (Helena Laboratories, Beaumont, TX,SA). Samples (2 �l) were separated using 0.6% agarose gels in bar-ital buffer (pH 8.6) (Helena Laboratories, Montreal, QC, Canada)t a constant voltage (80 V) for 45 min. The gels were then ovenried at 75 ◦C and stained with 0.1% (w/v) Fat Red 7B in 95%ethanol.

.5. Cell cultures

Human THP-1 monocytes and J774 macrophages were grownn RPMI 1640 and DMEM medium, respectively. The media

ere supplemented with 10% heat-inactivated FBS, 50 mM 2-E (for THP-1), 2 mM L-glutamine, 1.5 mg/ml of glucose, and

00 U/ml of penicillin. The differentiation of THP-1 monocytesnto macrophages was induced by plating the cells at a den-ity of 105 cells/cm2 in the presence of 100 nM phorbol myristatecetate (PMA) for 96 h. Primary peritoneal macrophages werebtained from Balb/c mice. Five days after an intra-peritoneal injec-ion of 2 ml of 4% (w/v) thioglycollate medium, the macrophagesells were harvested by injecting 10 ml of cold DMEM into thebdominal cavity [29]. The cells were pelleted and re-suspendedn DMEM supplemented with 5% FBS and plated at a den-ity of 105 cells/cm2. They were allowed to adhere for 4 h andon-adherent cells were removed by rinsing with DMEM. Theacrophage-enriched adherent cells were used for the experi-ents described below.

.6. Measurement of cholesterol efflux

J774 macrophages were cultured for 48 h in RPMI medium sup-lemented with 1% FBS and 3H cholesterol (2 �Ci/ml). They wereashed with serum free DMEM containing 0.2% BSA and then incu-

ated with resveratrol (0–25 �M) or 0.3 mM 8-Br-cAMP for 12 h tonduce ABCA1 cells [30]. The macrophages were then incubatedor 6 h with purified apoA-1 (25 �g/ml), a specific acceptor of freeholesterol, to assess cholesterol efflux.

We also investigated the relationship between the antioxidantroperties of resveratrol and its effect on HDL-mediated choles-erol efflux. In one set of experiments, 3H-cholesterol-labeledHP-1 and cAMP-stimulated J774 macrophages were incubatedor 6 h with 50 �g/ml of native or oxidized HDL3 (the HDL3

as previously oxidized in the presence or absence of 1 �Mesveratrol). In another set of experiments, ABCA1-enriched J774nd mouse peritoneal macrophages were subjected to oxidativetress by incubating them with 0.2 mM iron/ascorbate (Fe/Asc) inhe absence or presence of 0–25 �M resveratrol or 10 �M vita-

in E for 6 h. Cholesterol efflux was assessed by incubating theells with 50 �g/ml of native HDL3 for 6 h at 37 ◦C. The cellsnd medium were then separated by centrifugation (350 × g for0 min) and the cells were lysed in 0.1 M NaOH. The counts perinute (cpm) in the medium and cell lysates were determined

eparately using a liquid scintillation counter (model 1600 TR;ackard Instrument Company, Meriden, CT, USA). Cholesterol effluxas measured by determining the percentage of radiolabeled

holesterol released (% cholesterol efflux) using the follow-ng formula: (cpm in medium/cpm in the cell + medium) × 10031].

.7. Measurement of cholesterol influx

The effect of resveratrol on cholesterol influx was studied using

774 macrophages. The cells were incubated with 3H-cholesterol2 �Ci/ml) for 24 h in the absence or presence of 0–25 �M resver-trol. The medium was then removed from the culture dishes. Theells were washed twice with PBS and lysed. The cholesterol con-ent was measured by determining the percentage of radiolabeled

osis 207 (2009) 420–427

cholesterol incorporated (% cholesterol influx) using the followingformula: (cpm in the cell/cpm in the cell + medium) × 100.

2.8. Statistical analysis

Values are expressed as the means ± SEM. One-way analysisof variance (ANOVA) was used for multiple comparisons. Linearregression analysis was used to assess the association between twocontinuous variables. All statistical analyses were performed usingGraphPad prism-5 softwareTM.

3. Results

3.1. Antioxidant effect of resveratrol

Initial experiments were carried out to assess the antioxidanteffect of resveratrol on LDL peroxidation. LDL oxidation was inducedby copper ions or exposure to oxygen free radicals. Conjugateddiene formation in the lipid fraction was measured to determinethe extent of LDL peroxidation.

3.2. Resveratrol inhibits LDL oxidation

Incubating native human LDL with CuSO4 resulted in the oxi-dation of LDL polyunsaturated fatty acids, as indicated by theformation of conjugated diene (Fig. 1A). The kinetic profile of theoxidation was characterized by an initial lag phase followed bya propagation phase, where the rate of conjugated diene forma-tion was maximal, and then by a decomposition phase. The onsetor lag phase, before appreciable peroxidation could be observed,was measured from the intercept of the initiation and propaga-tion phases. Interestingly, progressively higher concentrations ofresveratrol (0, 0.1, 0.5, and 1 �M) inhibited the oxidation of LDLand reduced its susceptibility to lipid peroxidation, as observed byan increase in the lag phase and a reduction in the oxidation rate(Vmax), respectively (Fig. 1A, upper and lower panels). At longer oxi-dation times (6 and 8 h), the antioxidant effect of resveratrol wassignificant only at 1 �M (p < 0.01 and 0.001, respectively).

We also assessed the antioxidant effect of resveratrol in an ion-free system, where oxidation was induced by exposure to •OH/O2

•−

free radicals produced by �-radioloysis. This method is superior tothe CuSO4 method in that it is quantitative and highly selective interms of free radical production [32], which made it possible topropose a mechanism for the antioxidant effect of resveratrol. LDLoxidized by exposure to •OH/O2

•− free radicals produced using arange of radiation doses (0–150 Gy) resulted in a parallel increase inthe formation of conjugated dienes (p < 0.001), which was inhibitedby resveratrol in a concentration-dependent manner (Fig. 1B).

The protective effect of resveratrol on LDL oxidation was con-firmed by the apoB electronegative charge measurements, whichpointed to the oxidative modification of the LDL-protein moiety.Our results showed that oxidation of LDL alone increased its elec-trophoretic mobility, particularly at 75 and 150 Gy. This effect wasattenuated when the LDL were oxidized in the presence of 5 �Mresveratrol (Fig. 1S).

Taking in account that endothelial cells interact directly withlipoproteins and may exert an oxidative stress especially on theLDL particles, we have demonstrated that resveratrol significantlyinhibited endothelial cells (Eahy926)-induced LDL oxidation byreducing MDA formation (p < 0.05) (Fig. 2S). Moreover, effect of

resveratrol on the glutathione system (glutathione peroxidase andreductase activities) was investigated in the THP-1 macrophages.Our results show that resveratrol protects glutathione systemagainst H2O2-induced oxidation by preserving GPx and GR activ-ities (Fig. 3S).

H. Berrougui et al. / Atherosclerosis 207 (2009) 420–427 423

Fig. 1. Resveratrol possesses antioxidant properties. (A) Kinetics of conjugated diene formation by CuSO4-induced LDL oxidation in the absence or presence of resveratrol(0–1 �M). Conjugated diene formation was followed by monitoring absorbance at 234 nm. The upper panel shows the effect of resveratrol on the lag phase of conjugateddiene formation. The lower panel shows the effect of resveratrol on the maximal rate (Vmax) of conjugated diene formation. (B) Formation of conjugated diene in LDLexposed to free radicals produced by �-radiolysis of ethanol–water mixtures as a function of time (dose: 0.13 Gy/s). The LDL preparations were incubated with 0–10 �Mresveratrol. Preparations without resveratrol were used as negative controls. LDL (0.1 mg/ml) was prepared in oxygenated 10 mM sodium phosphate buffer (pH 7). Vitamin E(10 and 20 �M) was used as a positive control. Conjugated diene formation was followed by monitoring absorption at 234 nm (ε234 nm = 27,000 M−1/cm). (C) DPPH free radicalscavenging activity of resveratrol. DPPH was incubated with 0–50 �M resveratrol for 180 min and the absorbance at 518 nm was monitored every 30 min. Vitamin E (10 and2 esseda

3

Lsadmtopriw2rflas

3

Jd

0 �M) was used as a positive control. The scavenging activity of resveratrol is exprt least three independent experiments.

.3. Free radical scavenging activity of resveratrol

To gain more insight into the ability of resveratrol to preventDL oxidation, we studied the kinetics of resveratrol free radicalcavenging activity by allowing resveratrol to react with DPPH,

stable free radical. The DPPH assay measures the hydrogen-onating ability of antioxidants over a relatively short period byonitoring the decrease in absorbance resulting from the reduc-

ion of the DPPH• free radical form to the DPPH-H form. The effectf increasing concentrations of resveratrol (0–50 �M) was com-ared to that of vitamin E (10 and 25 �M). Both compounds actedapidly to scavenge free radicals since over than 50% of their activ-ty was observed in the first 30 min, and at the remaining 50%,

ithin 3 h. The antioxidant activity (AA%) of resveratrol at 5, 10,5, and 50 �M was 16 ± 3.2, 21.6 ± 1.8, 45.4 ± 1.0, and 69.3 ± 0.8,espectively. As shown in Fig. 1C, the reaction kinetics of the DPPHree radical and resveratrol concentrations were negatively corre-ated (r2 = 0.96, p = 0.0032). Interestingly, resveratrol and vitamin Et identical concentrations exhibited equivalent DPPH free radicalcavenging activity.

.4. Resveratrol enhances cholesterol efflux

We assessed the effect of resveratrol (10−6 to 10−3 M) on774 macrophage viability using the MTT colorimetric assay toetermine the range of resveratrol concentrations that would

as the percentage of remaining DPPH. Results are expressed as the means ± SEM of

be used in the cholesterol homeostasis experiments. The IC50was obtained with 114 ± 1.3 �M of resveratrol while the IC20was obtained with 30 �M (Fig. 4S). Effect of resveratrol on theapoA1-mediated cholesterol efflux was also studied in the Eahy926endothelial cells. Our results show that resveratrol significantlyenhanced the cholesterol efflux (Fig. 5S-A). We next investi-gated the effect of resveratrol on ABCA1-dependent cholesterolefflux pathways in J774 macrophages. Exposing 3H-cholesterol-loaded cells to apoA-1 (cAMP-free) for 6 h (time-range to reactonly with ABCA1) resulted in a low-cholesterol efflux. How-ever, when the macrophages were pretreated overnight withresveratrol (0–25 �M), with 0.3 mM cAMP as a positive control,the apoA1-induced cholesterol efflux was significantly potenti-ated in a resveratrol concentration-dependent manner (r2 = 0.907,p < 0.05, linear regression) (Fig. 2A). To more clearly under-stand the mechanism of action of resveratrol on ABCA1-mediatedcholesterol efflux, we measured the effect of resveratrol on theABCA1 protein expression in J774 macrophages. Incubating themacrophages in the presence of resveratrol (3, 6, 8, and 16 h)resulted in an increase in ABCA1 protein expression (data notshown).

3.5. Resveratrol reduces cholesterol influx

Cholesterol influx is defined as the movement of cholesterolmolecules from the extra-cellular environment to cells. This pro-

424 H. Berrougui et al. / Atherosclerosis 207 (2009) 420–427

Fig. 2. Resveratrol increases cholesterol efflux and reduces influx. (A) Effect ofresveratrol on apoA1-mediated cholesterol efflux. 3H-cholesterol-loaded J774 cellswere treated with various concentrations of resveratrol (0–25 �M) and then incu-bated with 25 �g/ml of apoA1. cAMP (300 �M) was used as a positive control. Thetrace shows a linear correlation between apoA1-mediated cholesterol efflux andresveratrol concentrations. All results are expressed as means ± SEM of at least threeindependent experiments. (B) J774 cells were labeled with 3H-cholesterol in theaice

crowr(acB

Fig. 3. Resveratrol protects macrophages against oxidation and promotes HDL3-mediated cholesterol efflux. J774 macrophages (A) and mouse peritonealmacrophages (MPM) (B) were loaded with 3H-cholesterol and allowed to equili-brate for 16 h. The macrophages were stressed with 0.2 mM iron/ascorbate (Fe/Asc)and cholesterol efflux was assessed using 50 �g/ml of HDL3. The panels show the

bsence or presence of various concentrations of resveratrol (0–25 �M). Cholesterolnflux is expressed as the difference in the cholesterol content of the medium and theell lysates. Results are expressed as the means ± SEM of at least three independentxperiments.

ess, which is mainly observed in macrophages, plays a regulatoryole in cellular cholesterol homeostasis. The effect of resveratroln cholesterol influx was investigated in J774 macrophages ande observed that the influx of 3H-cholesterol in the presence of

esveratrol was significantly reduced in a dose-dependent mannerr2 = 0.89, p = 0.015) (Fig. 2B). Moreover, we show that resver-trol decreased significantly cholesterol uptake by endothelialells (Eahy926) from 3H-cholesterol-enriched media (Fig. 5S-).

linear correlation between HDL3-mediated cholesterol efflux and resveratrol con-centrations. Vitamin E (10 �M) was used as a positive control.

3.6. Resveratrol protects against lipid peroxidation and promotescholesterol efflux

Oxidative damage to macrophages impairs cholesterol effluxas previously demonstrated by an impairment of ABCA1 proteinexpression under Fe/Asc stress Marcil et al. [33]. We investigated theeffect of resveratrol on cholesterol efflux in J774 (Fig. 3A) and mouseprimary macrophages (Fig. 3B) under oxidative stress induced byFe/Asc. We observed that HDL3-mediated cholesterol efflux wassignificantly impaired by Fe/Asc, but that the effect was restoredby resveratrol in concentration-dependant manner (p < 0.05). Thiseffect was also observed with THP-1 macrophages (Fig. 6S). Theoxidation of HDL has been reported to significant decrease theircapacity to mediate cholesterol efflux [34]. We thus assessed thecapacity of resveratrol to preserve the functionality of HDL underoxidative stress conditions. ABCA1-enriched J774 cells previouslyloaded with cholesterol were incubated for 6 h with HDL3 oxi-

dized in the absence (control) or in the presence of resveratrol(treated). Resveratrol protected HDL3 against Cu-induced oxidationin a concentration-dependent manner as shown by the decrease inconjugated diene formation (Fig. 4A). It also maintained the capac-ity of HDL3 to mediate cholesterol efflux (p < 0.05) (Fig. 4B).

H. Berrougui et al. / Atheroscler

Fig. 4. Resveratrol protects HDL3 from copper oxidation and increases cholesterolefflux. (A) Conjugated diene formation in CuSO4-induced HDL oxidation in theabsence or presence of resveratrol as a function of time. HDL3 proteins were incu-bated with CuSO4 for 8 h (in a time course manner) in the absence or presenceof 0–2 �M resveratrol. (B) 3H- cholesterol efflux from human THP-1 macrophagesiei

4

posspnrbrpohnh

[46]. This process, which is in part dependent on the physico-

ncubated with 50 �g/ml of copper-oxidized HDL3 for 0, 2, and 4 h in the pres-nce or absence of 1 �M resveratrol. Results are the means ± SEM of at least threendependent experiments.

. Discussion

Dietary antioxidants have attracted considerable attention asreventive and therapeutic agents because of the “oxidative the-ry” of atherosclerosis [35]. A number of dietary antioxidant intaketudies (epidemiological, case–control, and prospective and retro-pective cohorts) have shown that antioxidants can prevent CVDrogression [36]. Indeed, the consumption of foods rich in phe-ols and polyphenols has been positively correlated with CVDisk reduction by slowing atherosclerosis progression, principallyy protecting lipoproteins from lipid peroxidation [37]. Becauseesveratrol is present in a variety of foods, including grapes, arimary impetus for research on resveratrol was the paradoxical

bservation that a low incidence of CVD could co-exist with aigh-fat diet intake and moderate consumption of red wine, a phe-omenon known as the French paradox [38]. Various mechanismsave been proposed to explain the anti-atherosclerotic proper-

osis 207 (2009) 420–427 425

ties of resveratrol, including inhibition of lipid peroxidation [22],modulation of platelet aggregation [8], inhibition of smooth mus-cle cell proliferation [9], and reduction of macrophage-inducedinflammation [39]. Despite numerous studies showing the effectof resveratrol on the lipid metabolism and the progression ofatherosclerosis, little is known about the antioxidant properties ofresveratrol and their effect on cholesterol homeostasis.

A large body of evidence indicates that oxLDL plays a keyrole in both the early and more advanced inflammatory stagesof atherosclerosis lesions [40]. For example, oxLDL is present inatherosclerotic lesions [41], the progression of atherosclerosis isslowed when oxidation is inhibited [42,43], and there is a cor-relation between the ability of LDL to resist oxidation and theseverity of coronary atherosclerosis [42]. However, LDL oxidationis not the sole factor involved in atherogenesis. In vivo oxidativestress impairs the anti-atherogenic properties of HDL, especiallyby impairing its antioxidant activity and its capacity to enhancecholesterol efflux and promote RCT [34]. Moreover, oxidation canbe directly induced in macrophages, which alters the expression ofreceptors involved in cholesterol flux, which in turn influence foamcell formation and atherosclerosis development [33].

Our results showed that resveratrol prevented copper- and �-radiolysis-induced LDL peroxidation in a concentration-dependentmanner by decreasing the formation of conjugated dienes, thusextending the lag phase and lowering the oxidation rate. It hasalso been reported that resveratrol decreases the rate of vitaminE disappearance and maintains the endogenous vitamin E contentof LDL [44]. A wide variety of natural products exert their antioxi-dant effects by preventing the degradation of endogenous vitaminsin lipoproteins. Resveratrol, which has three phenolic hydroxylgroups, is very lipophilic, enabling it to associate with the lipid moi-ety of lipoproteins and prevent the oxidation of their unsaturatedfatty acids [44].

LDL oxidation is characterized by alterations in the structuraland biological properties of the lipid fraction and the apolipoproteinmoiety (apoB), including the early fragmentation of the pro-tein moiety, which contains sensitive amino acid residues. Thisalteration is followed by cross-linking of reactive aldehydes andoxysterols (end products of lipid peroxidation). In the present study,we showed that resveratrol reduced the �-radiolysis-induced alter-ation of apo-B, as shown by the decrease in the electrophoreticmobility of LDL.

To clarify the mechanism by which resveratrol reduces LDLoxidation, we investigated the free radical scavenging activity ofresveratrol, which may prevent chain-breaking and the alterationof apoB. Resveratrol may interact with free radicals to form rela-tively stable free radicals (SFR•) and non-radicals (NR). It may alsoreduce copper, resulting in the formation of a non-radical productand SFR• that, under conditions of high-oxidative stress, quenchanother free radical. However, in addition to its broad range of activ-ities in the lag phase, resveratrol also reduced the Vmax but had noeffect on the ODmax, except at high concentrations. This antioxidantprofile suggested that resveratrol might also act as an inhibitor ofcopper binding via interactions with apolipoproteins. This observa-tion is in agreement with the results reported by Belguendouz et al.[7], who showed that resveratrol is a potent chelator of free copperions and can also remove copper ions bound to apo-B. Our resultsare also in agreement with those of Zini et al. [45], who suggestedthat resveratrol inhibits the lipid peroxidation induced by Fentonreaction products.

HDL plays an important anti-atherogenic role by mediating RCT

chemical properties of HDL and on the oxidative state of the cells,enables cells to pump out excess free cholesterol. We, and oth-ers, have previously demonstrated that HDL oxidation significantlyaffects the capacity of HDL to promote cholesterol efflux from

4 oscler

mtdhn[

arniciOmotmeoatpp[scwtb

tiwO[actrwdeItHfra

dtCpofiopritt

erm

[

[[

[

[

[

[

[

26 H. Berrougui et al. / Ather

acrophages [27,34]. The oxidation of HDL3 alters the fluidity ofhe phospholipidic layer, modifies the structure of apoproteins, andecreases paraoxonase 1 activity [34,47]. All these modificationsave an effect on the normal interactions between HDL3 compo-ents and membrane receptors, including ABCA1, ABCG1, and SR-BI48].

When J774 macrophages were incubated with resveratrol,poA1-mediated cholesterol efflux increased, suggesting thatesveratrol up-regulates this process by up-regulating ABCA1. Aumber of mechanisms have been proposed to explain this effect,

ncluding direct binding of apoA-1 to ABCA1, ABCA1-mediatedhanges to the plasma membrane that stimulate apoA1 bind-ng, the availability of phospholipids, and cholesterol efflux [49].ver-expression of ABCA1 alters the morphology of the plasmaembrane [18] and increases apoA-1 binding [50] and cholesterol

xidase activity, which point to changes in cholesterol concen-rations or distribution within the external leaflet of the plasma

embrane. Our results showed that resveratrol-stimulated ABCA1xpression in J774 macrophages, which is in agreement with thosef Sevov et al. [11], who showed that resveratrol induces LXR-�nd elevated ABCA1 and ABCG1 mRNA levels. This suggests thathe up-regulation of ABCA1 protein expression may be one of theathways involved in the resveratrol-mediated cholesterol effluxrocess, which would be consistent with the increased HDL levels51] and reduced atherosclerosis seen following polyphenol con-umption. We also showed that resveratrol reduces the influx ofholesterol into J774 macrophages in a dose-dependent manner,hich ties in with the fact that resveratrol down-regulates lipopro-

ein lipase (LPL) and SR-AII, which promote increased lipid uptakey macrophages [11].

Oxidative stress is a continuous process in all physiological sys-ems, and results in the oxidation of various molecules and thempairment of their functions. HDL can also be oxidized in vivo,

hich causes a loss of its anti-atherogenic proprieties [52,53].xidative stress also influences cholesterol efflux in macrophages

33]. In the present study, we also investigated the effect of resver-trol on cholesterol efflux in J774 macrophages stressed by a Fe/Ascomplex, which induces lipid peroxidation [54] and reduces choles-erol efflux. While great care should be taken in extrapolating theesults we obtained with J774 macrophages, they are in agreementith those of a previous study using mouse primary peritoneal-

erived macrophages [33]. Oxidative stress can also reduce thexpression of ABCA1, LXR, and PPAR, but has no effect on SR-BI.ncubating macrophages with Fe/Asc in the presence of resvera-rol significantly restored cholesterol efflux from macrophages toDL3, probably by suppressing the effect of Fe/Asc on the cell sur-

ace receptors involved in this process. This effect has also beeneported with vitamin E and butylhydroxytoluene (BHT), two otherntioxidants [33].

The oxidation of HDL results in increased levels of lipid peroxi-ation markers, including conjugated dienes, lipid hydroperoxides,hiobarbituric acid reactive substances (TBARS), and aldehydes.ompositional changes are associated with alterations in thehysico-chemical properties of HDL, including fluidity, molecularrder, and electric charges [55]. Compositional and structural modi-cations are also associated with changes in the biological activitiesf HDL. For instance, oxidizing HDL in vitro decreases its ability toromote the RCT process [56]. In the present study, we showed thatesveratrol inhibited the oxidation of HDL3 and helped to maintaints capacity to mediate cholesterol efflux. This effect may be relatedo the preservation of the physico-chemical properties of HDL and

he integrity of protein moieties like apoA-1 and PON1.

We contributed to a better understanding of the potential ben-fits of consuming foods rich in polyphenols by showing thatesveratrol protects against lipoprotein oxidation and foam cell for-

ation and promotes cholesterol efflux from macrophages. Our

[

[

osis 207 (2009) 420–427

findings on the intracellular signaling pathways modulating thecholesterol efflux process will help lay the foundation for devel-oping new therapeutic approaches to preventing and treatingatherosclerosis and cardiovascular diseases. Although, our studyis only contribution on some mechanisms by which resveratrolmay exert its beneficial effect, human studies are needed to con-firm the beneficial effect of resveratrol and its impact in clinicalsetting.

Acknowledgements

This work was supported by the Canadian Institute of HealthResearch (CIHR). G. Grenier and A. Khalil are recipients of a Junior 1and 2 investigator award, respectively, from the Fonds de Rechercheen Santé du Québec (FRSQ).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.atherosclerosis.2009.05.017.

References

[1] Superko HR. Drug therapy and the prevention of atherosclerosis in humans. AmJ Cardiol 1989;64:31G–8G.

[2] Das S, Das DK. Resveratrol: a therapeutic promise for cardiovascular diseases.Recent Patents Cardiovasc Drug Discov 2007;2:133–8.

[3] Norata GD, Marchesi P, Passamonti S, et al. Anti-inflammatory and anti-atherogenic effects of cathechin, caffeic acid and trans-resveratrol inapolipoprotein E deficient mice. Atherosclerosis 2007;191:265–71.

[4] Wang Z, Zou J, Cao K, et al. Dealcoholized red wine containing known amounts ofresveratrol suppresses atherosclerosis in hypercholesterolemic rabbits withoutaffecting plasma lipid levels. Int J Mol Med 2005;16:533–40.

[5] Araya J, Rodrigo R, Orellana M, Rivera G. Red wine raises plasma HDL and pre-serves long-chain polyunsaturated fatty acids in rat kidney and erythrocytes.Br J Nutr 2001;86:189–95.

[6] Auger C, Caporiccio B, Landrault N, et al. Red wine phenolic compounds reduceplasma lipids and apolipoprotein B and prevent early aortic atherosclerosisin hypercholesterolemic golden Syrian hamsters (Mesocricetus auratus). J Nutr2002;132:1207–13.

[7] Belguendouz L, Fremont L, Linard A. Resveratrol inhibits metal ion-dependentand independent peroxidation of porcine low-density lipoproteins. BiochemPharmacol 1997;53:1347–55.

[8] Wang Z, Huang Y, Zou J, et al. Effects of red wine and wine polyphenol resveratrolon platelet aggregation in vivo and in vitro. Int J Mol Med 2002;9:77–9.

[9] Araim O, Ballantyne J, Waterhouse AL, Sumpio BE. Inhibition of vascular smoothmuscle cell proliferation with red wine and red wine polyphenols. J Vasc Surg2002;35:1226–32.

10] Avellone G, Di Garbo V, Campisi D, et al. Effects of moderate Sicilian red wineconsumption on inflammatory biomarkers of atherosclerosis. Eur J Clin Nutr2006;60:41–7.

[11] Sevov M, Elfineh L, Cavelier LB. Resveratrol regulates the expressionof LXR-alpha in human macrophages. Biochem Biophys Res Commun2006;348:1047–54.

12] Lusis AJ, Atherosclerosis. Nature 2000;407:233–41.13] Aviram M, Rosenblat M. Macrophage-mediated oxidation of extracellular low

density lipoprotein requires an initial binding of the lipoprotein to its receptor.J Lipid Res 1994;35:385–98.

14] Cherki M, Berrougui H, Isabelle M, et al. Effect of PON1 polymorphism on HDLantioxidant potential is blunted with aging. Exp Gerontol 2007;42:815–24.

15] Glomset JA. The plasma lecithins:cholesterol acyltransferase reaction. J LipidRes 1968;9:155–67.

16] Bortnick AE, Rothblat GH, Stoudt G, et al. The correlation of ATP-binding cas-sette 1 mRNA levels with cholesterol efflux from various cell lines. J Biol Chem2000;275:28634–40.

[17] de La Llera-Moya M, Connelly MA, Drazul D, et al. Scavenger receptor class Btype I affects cholesterol homeostasis by magnifying cholesterol flux betweencells and HDL. J Lipid Res 2001;42:1969–78.

18] Wang N, Silver DL, Costet P, Tall AR. Specific binding of ApoA-I, enhanced choles-terol efflux, and altered plasma membrane morphology in cells expressingABC1. J Biol Chem 2000;275:33053–8.

19] Brunham LR, Kruit JK, Iqbal J, et al. Intestinal ABCA1 directly contributes to HDL

biogenesis in vivo. J Clin Invest 2006;116:1052–62.

20] Kennedy MA, Barrera GC, Nakamura K, et al. ABCG1 has a critical role in medi-ating cholesterol efflux to HDL and preventing cellular lipid accumulation. CellMetab 2005;1:121–31.

21] Zou J, Huang Y, Chen Q, et al. Effects of resveratrol on oxidative modification ofhuman low density lipoprotein. Chin Med J (Engl) 2000;113:99–102.

oscler

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

H. Berrougui et al. / Ather

22] Fremont L, Belguendouz L, Delpal S. Antioxidant activity of resveratrol andalcohol-free wine polyphenols related to LDL oxidation and polyunsaturatedfatty acids. Life Sci 1999;64:2511–21.

23] Mensor LL, Menezes FS, Leitao GG, et al. Screening of Brazilian plant extractsfor antioxidant activity by the use of DPPH free radical method. Phytother Res2001;15:127–30.

24] Sattler W, Mohr D, Stocker R. Rapid isolation of lipoproteins and assessmentof their peroxidation by high-performance liquid chromatography postcolumnchemiluminescence. Methods Enzymol 1994;233:469–89.

25] Khalil A, Fortin JP, LeHoux JG, Fulop T. Age-related decrease of dehy-droepiandrosterone concentrations in low density lipoproteins and its role inthe susceptibility of low density lipoproteins to lipid peroxidation. J Lipid Res2000;41:1552–61.

26] Fricke H. The chemical action of Röntgen rays on dilute ferrosulfate solutionsas measure of dose. Am J Roentgenol Radium Ther Nucl Med 1927;18:429–32.

27] Berrougui H, Cloutier M, Isabelle M, Khalil A. Phenolic-extract from argan oil(Argania spinosa L.) inhibits human low-density lipoprotein (LDL) oxidationand enhances cholesterol efflux from human THP-1 macrophages. Atheroscle-rosis 2006;184:389–96.

28] Pinchuk I, Lichtenberg D. The mechanism of action of antioxidants againstlipoprotein peroxidation, evaluation based on kinetic experiments. Prog LipidRes 2002;41:279–314.

29] Linton MF, Atkinson JB, Fazio S. Prevention of atherosclerosis in apolipopro-tein E-deficient mice by bone marrow transplantation. Science 1995;267:1034–7.

30] Hajj Hassan H, Blain S, Boucher B, et al. Structural modification of plasma HDL byphospholipids promotes efficient ABCA1-mediated cholesterol release. J LipidRes 2005;46:1457–65.

31] Berrougui H, Isabelle M, Cloutier M, Grenier G, Khalil A. Age-related impairmentof HDL-mediated cholesterol efflux. J Lipid Res 2007;48:328–36.

32] Bonnefont-Rousselot D. Gamma radiolysis as a tool to study lipoprotein oxida-tion mechanisms. Biochimie 2004;86:903–11.

33] Marcil V, Delvin E, Sane AT, Tremblay A, Levy E. Oxidative stress influencescholesterol efflux in THP-1 macrophages: role of ATP-binding cassette A1 andnuclear factors. Cardiovasc Res 2006;72:473–82.

34] Girona J, LaVille AE, Sola R, Motta C, Masana L. HDL derived from the dif-ferent phases of conjugated diene formation reduces membrane fluidityand contributes to a decrease in free cholesterol efflux from human THP-1macrophages. Biochim Biophys Acta 2003;1633:143–8.

35] Gugliucci A, Menini T. Three different pathways for human LDL oxidation areinhibited in vitro by water extracts of the medicinal herb Achyrocline sat-ureoides. Life Sci 2002;71:693–705.

36] Kaliora AC, Dedoussis GV, Schmidt H. Dietary antioxidants in preventing athero-

genesis. Atherosclerosis 2006;187:1–17.

37] Xu BJ, Yuan SH, Chang SK. Comparative studies on the antioxidant activitiesof nine common food legumes against copper-induced human low-densitylipoprotein oxidation in vitro. J Food Sci 2007;72:S522–7.

38] Tunstall-Pedoe H, Kuulasmaa K, Mahonen M, et al. Contribution of trends insurvival and coronary-event rates to changes in coronary heart disease mor-

[

[

osis 207 (2009) 420–427 427

tality: 10-year results from 37 WHO MONICA project populations. Monitoringtrends and determinants in cardiovascular disease. Lancet 1999;353:1547–57.

39] Tsai SH, Lin-Shiau SY, Lin JK. Suppression of nitric oxide synthase and the down-regulation of the activation of NFkappaB in macrophages by resveratrol. Br JPharmacol 1999;126:673–80.

40] de Lorgeril M, Salen P. Diet as preventive medicine in cardiology. Curr OpinCardiol 2000;15:364–70.

[41] Haberland ME, Fong D, Cheng L. Malondialdehyde-altered protein occursin atheroma of Watanabe heritable hyperlipidemic rabbits. Science1988;241:215–8.

42] Regnstrom J, Nilsson J, Tornvall P, Landou C, Hamsten A. Susceptibility tolow-density lipoprotein oxidation and coronary atherosclerosis in man. Lancet1992;339:1183–6.

43] Heinecke JW. Oxidants and antioxidants in the pathogenesis of atherosclerosis:implications for the oxidized low density lipoprotein hypothesis. Atheroscle-rosis 1998;141:1–15.

44] Belguendouz L, Fremont L, Gozzelino MT. Interaction of transresveratrol withplasma lipoproteins. Biochem Pharmacol 1998;55:811–6.

45] Zini R, Morin C, Bertelli A, Bertelli AA, Tillement JP. Effects of resveratrol on therat brain respiratory chain. Drugs Exp Clin Res 1999;25:87–97.

46] Fredenrich A, Bayer P. Reverse cholesterol transport, high density lipoproteinsand HDL cholesterol: recent data. Diab Metab 2003;29:201–5.

47] Jaouad L, Milochevitch C, Khalil A. PON1 paraoxonase activity is reduced duringHDL oxidation and is an indicator of HDL antioxidant capacity. Free Radic Res2003;37:77–83.

48] Pirillo A, Uboldi P, Kuhn H, Catapano AL. 15-Lipoxygenase-mediated modifica-tion of high-density lipoproteins impairs SR-BI- and ABCA1-dependent choles-terol efflux from macrophages. Biochim Biophys Acta 2006;1761:292–300.

49] Oram JF, Heinecke JW. ATP-binding cassette transporter A1: a cell choles-terol exporter that protects against cardiovascular disease. Physiol Rev2005;85:1343–72.

50] Vaughan AM, Oram JF. ABCA1 redistributes membrane cholesterol independentof apolipoprotein interactions. J Lipid Res 2003;44:1373–80.

[51] Tsang C, Higgins S, Duthie GG, et al. The influence of moderate red wine con-sumption on antioxidant status and indices of oxidative stress associated withCHD in healthy volunteers. Br J Nutr 2005;93:233–40.

52] Francis GA. High density lipoprotein oxidation: in vitro susceptibility andpotential in vivo consequences. Biochim Biophys Acta 2000;1483:217–35.

53] Panzenboeck U, Raitmayer S, Reicher H, et al. Effects of reagent and enzymat-ically generated hypochlorite on physicochemical and metabolic properties ofhigh density lipoproteins. J Biol Chem 1997;272:29711–20.

54] Trudel K, Sinnett D, James RW, et al. Iron-ascorbic acid-induced oxidant stressand its quenching by paraoxonase 1 in HDL and the liver: comparison between

humans and rats. J Cell Biochem 2005;96:404–11.

55] Ferretti G, Bacchetti T, Negre-Salvayre A, et al. Structural modifications of HDLand functional consequences. Atherosclerosis 2006;184:1–7.

56] Rifici VA, Khachadurian AK. Oxidation of high density lipoproteins: charac-terization and effects on cholesterol efflux from J774 macrophages. BiochimBiophys Acta 1996;1299:87–94.