Female Rats Fed a High-Fat Diet Were Associated with Vascular Dysfunction and Cardiac Fibrosis in the Absence of Overt Obesity and Hyperlipidemia: Therapeutic Potential of Resveratrol

Obesity represents a growing health problem in industrialized countries contributing to the onset and/or development of various metabolic-associated disease states, including dyslipidemia and type 2 diabetes. Furthermore, in patients diagnosed with metabolic disorders, hypertension and coronary artery disease are prevalent and the subsequent risk of myocardial infarction is increased (Grundy et al., 2006;Kenchaiah et al., 2002; Klein et al., 2004). In the experimental setting, animals provided with a high-fat diet and associated with overt obesity were likewise documented with elevated blood pressure (Dobrian et al., 2000, 2001; Erdei et al., 2006). The study by Dobrian et al. (2000) demonstrated that the increase in systolic blood pressure in hypercholesterolemia obese rats that were fed a high-fat diet may be related to the 2-fold increase of plasma renin. Furthermore, increased oxidative stress was also highlighted in obese patients and in experimental animal models of obesity and delineated as a seminal event contributing to the development of hypertension (Dobrian et al., 2000; Roberts et al., 2001; Erdei et al., 2006; Vachharajani and Vital, 2006; Van Gaal et al., 2006; Yamato et al., 2007). The underlying mechanism attributed to the hypertensive action of oxidative stress is secondary to a reduction of biologically active NO via the interaction with superoxide anions (Dobrian et al., 2001; Erdei et al., 2006).

Despite the unequivocal relationship between obesity and cardiovascular disease, it remains presently unknown whether vascular reactivity is impaired and whether maladaptive remodeling of the heart occurs before an overt gain in body weight or a change in the plasma lipid profile. In this regard, the present study tested the hypothesis that an impairment of endothelial-mediated relaxation and a reactive fibrotic response in the heart are prevalent in normal adult female rats fed a high-fat diet before the manifestation of overt obesity and in the absence of hyperlipidemia. Furthermore, as previously discussed, elevated oxidative stress represents a putative feature of obesity and is implicated in both vascular dysfunction and cardiac fibrosis (Dobrian et al., 2001; Erdei et al., 2006; Van Gaal et al., 2006; Lu et al., 2008). Thus, a second series of experiments tested the hypothesis that the coadministration of the natural antioxidant resveratrol (Baur et al., 2006; Das and Maulik, 2006) would counteract the deleterious effect of a high-fat diet on endothelial-mediated relaxation and reactive fibrosis.

Materials and Methods

Animal Care. Female Sprague-Dawley strain rats (180–200 g; Charles River, St-Constant, QC, Canada) were housed on arrival and had access to food and tap water ad libitum. The environment was controlled in terms of light (12:12-h light/dark cycle starting at 6:00 AM), humidity, and room temperature (20–23°C). All experiments were performed in compliance with recommendations of the Guidelines on the Care and Use of Laboratory Animals issued by the Canadian Council on Animal Research and the Guidelines of the Animal Care and were approved by the Animal Care Committee of the Montreal Heart Institute.

Diet and Treatment Protocol. One week after their arrival, rats were randomly assigned to either a standard or high-fat diet for a period of 8 weeks. The high-fat diet consisted of 42% lipids, 36% carbohydrates, and 22% proteins (kcal) and was provided in small pellets (MP Biomedicals, Irvine, CA). The standard rat diet consisted of 12.5% lipids, 63.2% carbohydrates, and 24.3% proteins (kcal) (Agribrands Purina Canada, Woodstock, ON). To assess the therapeutic benefit of resveratrol, 20 mg/kg/day resveratrol was added directly to rat chow, and this dose was reported to achieve a peak serum concentration of 1.2 μM (Baur and Sinclair, 2006). Furthermore, an equivalent dose of resveratrol was recently shown to significantly prolong the health and survival of mice fed a high calorie diet (Baur et al., 2006). During the 8-week protocol, the quantity of resveratrol added to the rat chow was adjusted according to changes in body weight measured once a week.

Hemodynamic Measurements and Sirius Red Staining. At the end of the study, rats were anesthetized with a mixture of ketamine (50 mg/kg; Rogarsetic, Toronto, ON, Canada) and xylazine (10 mg/kg; Rompun, Cambridge, ON, Canada). Systolic and diastolic blood pressures were measured with a Millar catheter (2F; model SPR-407; Millar Instrument, Houston, TX) following insertion through the right carotid. The Millar catheter was then inserted into the left ventricle (LV) to measure left ventricular contractility. Following baseline measurements, a cannula was inserted into the right jugular vein, and a submaximal dose of dobutamine (5 μg/kg/min) was injected at a rate of for a period of 3 min (Plante et al., 2005). Blood pressure and contractile indices were analyzed with the program IOX, version 1.8.9 (Emka Technologies; Falls Church, VA). Plasma lipid and glucose concentrations were determined by standard techniques in blood samples drawn from the carotid artery before sacrifice and subsequently placed in tubes containing sodium citrate. The heart was removed subsequently, separated into the right ventricle, LV, and septum, weighed, and stored at -80°C.

In a separate series of experiments, collagen content was determined by Sirius red, and staining was detected with a Polarizer-Trans U-P110 filter (Olympus, Carson Group Inc., Markham ON, Canada), as described previously (Mercier et al., 2002b). Ventricular function and cardiac morphology were not determined in these rats. In brief, the heart was excised, immersed in 10% formalin, and subsequently cut halfway between the base and apex. The apex and base sections of the heart were fixed, dehydrated, and embedded in paraffin. Serial cryostat sections (6 μm) of ventricular tissue were prepared. At least three distinct regions from the LV were assessed for collagen α₁ type 1 content (normalized to the surface area; mm²) and subsequently averaged. With regard to perivascular fibrosis, four vessels from each rat were assessed, and the lumen area was subtracted from the surface area. In parallel, vascular wall thickness was determined by measuring four distinct regions of the vessel and subsequently averaged. Following the quantification of collagen α₁ type 1, Sirius red staining was visualized with either a 10×/0.3 PLAN-NEOFLUOR or 63×-oil/1.4 DIC PLAN-APOCHROMAT objective mounted on a Zeiss LSM 510 confocal microscope, and the emitted signal was detected between 560 and 615 nm. Figure 1 represents a maximal projection derived from a z-stack of 0.2 μm slices. The z-stack was deconvolved with the Huygens Professional 3 software (SVI, Hilversum, The Netherlands), and the maximal projection was rendered with LSM 510 software (Carl Zeiss GmbH, Jena, Germany).

Vascular Reactivity. After sacrifice, the aorta was harvested and placed in a modified Krebs bicarbonate solution (118.3 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl₂, 1.2 mM MgSO₄, 1.2 mM KH₂PO₄, 25 mM NaHCO₃, 0.026 mM EDTA, and 11.1 mM dextrose). The aorta was subsequently dissected free of adherent fat and connective tissue and divided into rings (≈4 mm in length). The aortic rings were suspended between two metal stirrups connected to an isometric force transducer in a chamber containing Krebs bicarbonate solution maintained at 37°C and oxygenated with a mixture of 95% O₂/5% CO₂. After 30 min of stabilization, tension was progressively increased to the optimal tension of its active length-tension curve (approximately 3.5 g), as determined by the contractile response to potassium chloride (KCl; 30 mM) at increasing levels of stretch. Maximal contraction was determined with KCl (60 mM), and aortic rings were excluded if the contractile response to potassium chloride was not observed (exclusion rate of less than 5%). After extensive washing, aortic rings were incubated in the presence of indomethacin (10^-5 M) and propranolol (10^-7 M). After a 45-min period of stabilization, phenylephrine (2 × 10^-9 to 10^-8 M) was added to achieve a contraction that was a 100% of the maximal contraction observed with KCl (60 mM). Subsequent endothelial-dependent relaxation of isolated aortic rings was assessed with the addition of acetylcholine (10^-10 to 10^-5 M), and the vascular response was expressed as a percentage of the maximal contraction to phenylephrine. In a separate series of experiments, resveratrol (0.1 μM) was added to aortic rings isolated from both rats fed a high-fat and standard diet 10 min before the addition of acetylcholine. Lastly, endothelial-independent relaxation was assessed via the addition of sodium nitroprusside, and the vascular response was expressed as a percentage of the maximal contraction to phenylephrine.

Isolation of Total RNA, Reverse Transcription-PCR, and Real-Time PCR. Total myocardial RNA was isolated by a modification of the guanidine thiocyanate-phenol-chloroform extraction method, as described previously (Calderone et al., 1995). The reverse transcriptase reaction contained 5 ng/μl total RNA (each sample), Moloney murine leukemia virus reverse transcriptase (800 U), RNAseOUT (40 U), random-hexamer primers (0.04 U), and dNTPs (0.5 mM) and supplied optimal buffers. The reaction protocol consisted of three successive incubation steps at 1) 25°C for 10 min, 2) 37°C for 50 min, and 3) 70°C for 15 min. Real-time PCR was performed on 2.5 ng of cDNA template containing the appropriate primers (300 nM) and SYBR Green PCR master mix. Primers for each gene were obtained from distinct exons that spanned an intron by using the Ensembl Genome Browser program (http://www.ensembl.org). The sequence specificity of each primer was verified with the Blast program derived from the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov). The primers used are as follows: rat atrial natriuretic peptide (ANP), forward, 5′-AGAGCGGACTAGGCTGCAACA-3′, and reverse, 5′-ATTTGGCTGTTATCTTCGGTA-3′; rat sarcoplasmic reticulum calcium ATPase (SERCA2), forward, 5′-TGTATCGACAGGACAGAAAGAGT-3′, and reverse, 5′-TGATGAGCGAGACAGATTCACCTG-3′; rat transforming growth factor-β₁ (TGF-β₁), forward, 5′-CGTGCTAATGGTGGACCGCAACA-3′, and reverse, 5′-AGCTCTGCACGGGACAGCAAT-3′; rat transforming growth factor-β₃ (TGF-β₃), forward, 5′-AGAGATCCATAAATTCGACAT-3′, and reverse, 5′-ACACATTGAAACGGAAAACCT-3′; rat connective tissue growth factor, forward, 5′-AGGCCCTGTGAAGCTGACCTAGA-3′, and reverse, 5′-TTTTAGGCGTCCGGATGCACT-3′; and rat β-actin, forward, 5′-CCCTAAGGCCAACCGTGAA-3′, and reverse, 5′-GAGGCATACAGGGACAACACAG-3′. Appropriate negative controls were used for each experiment.

Drugs. All reagents were prepared daily in Ultrapure distilled water with the exception of indomethacin, and resveratrol was prepared in ethanol and dimethyl sulfoxide, respectively. Acetylcholine, bradykinin, indomethacin, and sodium nitroprusside were obtained from Sigma Chemical Co. (Mississauga, ON, Canada). Phenylephrine was obtained from Cayman Chemical Co. (Ann Arbor, MI), dobutamine came from Sandoz Canada Inc. (Boucherville, QC, Canada), and propranolol was obtained from Biomol Research Laboratories Inc. (Plymouth Meeting, PA). For real-time PCR, Moloney murine leukemia virus reverse transcriptase and RNaseOUT were obtained from Invitrogen (Burlington, ON, Canada), random-hexamer primers came from Amersham Biosciences (Baie-d'Urfe, QC, Canada), dNTPs came from MBI Fermentas (Burlington, ON, Canada), and SYBR Green PCR master mix was from Applied Biosystems (Foster, CA). Resveratrol was obtained from Royalmount Pharma (Montreal, QC, Canada).

Statistical Analysis. All values are expressed as the mean ± S.E.M. The half-maximal effective concentration (EC₅₀) of acetylcholine-mediated relaxation of isolated aortic rings was measured from the individual dose-response curve using a logistic curve-fitting program (Allfit for Windows 2.12; Dr. DeLéan, Université de Montréal, QC, Canada). The pD₂ value, the negative log of the EC₅₀, was likewise obtained with the program Allfit. A one-way analysis of variance was performed to assess differences in collagen content, vessel lumen area, and wall thickness, and a significant difference was determined by the Bonferroni's post hoc test and a p value less than 0.05 was considered statistically significant (StatView; SAS Institute Inc., Cary, NC). The effects of a high-fat diet and resveratrol treatment on body weight, cardiac morphology, left ventricular contractility, TGF-β₃ mRNA, and acetylcholine-mediated relaxation (maximal response) of aortic rings were analyzed by a two-way ANOVA (Statistica; StatSoft, Tulsa, OK), applied to determine whether there was an interaction between the two main effects. In the case of a significant interaction, a Tukey's (honest significant difference for equal or unequal n) post hoc test was used to compare diet (high-fat or standard) and treatment (with or without resveratrol), and a p value less than 0.05 was considered statistically significant.

Results

Morphometric Data and the Plasma Lipid Profile. A modest nonsignificant increase in body weight was observed in female rats fed a high-fat diet for a period of 8 weeks compared with female rats fed a standard diet (Table 1). Despite a modest change in body weight, absolute heart weight and left ventricular weight were similar in rats fed a high-fat diet compared with rats fed a standard diet (Table 1). In rats fed a high-fat diet, plasma cholesterol, high-density lipoprotein, and low-density lipoprotein concentrations were similar to rats fed a standard diet (Table 2). Likewise, plasma glucose levels were comparable between both groups (standard diet, 7.94 ± 1.07 mM; high-fat diet, 7.31 ± 0.79 mM). Lastly, in the high-fat fed rats, plasma triglyceride levels were modestly elevated compared with rats fed a standard diet but did not reach statistical significance (Table 2).

Mean Arterial Pressure and Left Ventricular Function. In rats fed a high-fat diet, systolic and diastolic blood pressure, mean arterial pressure, and left ventricular systolic pressure were significantly (p < 0.05) increased compared with rats fed a standard diet (Table 3). Furthermore, left ventricular rate of contraction (+dP/dt) and relaxation (-dP/dt) were likewise increased but did not reach statistical significance (Table 3). Lastly, to assess myocardial reserve, the contractile response to a dobutamine challenge was examined. In rats fed either a standard or high-fat diet, the increase in contractile response to dobutamine was identical (Table 4).

Cardiac Remodeling. Despite an increase of mean arterial pressure, the steady-state mRNA levels of ANP and SERCA2 in the LV were comparable in rats fed a high-fat and standard diet (Table 5). Employing the Sirius red technique, collagen α₁ protein content in the LV of rats fed a high-fat diet was not significantly increased compared with rats fed a standard diet (Table 6). By contrast, extensive and significant perivascular fibrosis was detected in rats fed a high-fat diet (Table 6; Fig. 1). Vessel lumen area (standard diet, 0.0122 ± 0.0024 mm², versus high-fat diet, 0.0201 ± 0.003 mm²) and wall thickness (standard diet, 0.0128 ± 0.0014 mm; high-fat diet, 0.0156 ± 0.002 mm) were not significantly different between rats fed a standard or high-fat diet. The reactive fibrotic response in rats fed a high-fat diet was not associated with a change in the expression of putative profibrotic peptides as the steady-state mRNA levels of connective tissue growth factor and transforming growth factor-β₁ in the LV were identical to that observed in rats fed a standard diet (Table 5). Moreover, TGF-β₃ mRNA levels were significantly decreased in the LV of rats fed a high-fat diet compared with rats fed a standard diet (Table 5).

Vascular Reactivity and the in Vitro Effect of Resveratrol. Endothelial-mediated relaxation may be compromised in hypertensive rats fed a high-fat diet. Indeed, the maximal relaxation of isolated aortic rings to acetylcholine was significantly decreased (p < 0.05) in rats fed a high-fat diet (22 ± 5%) compared with rats fed a standard diet (53 ± 8%) (Fig. 2). By contrast, vascular sensitivity (pD₂) to acetylcholine was similar in rats fed a standard (7 ± 0.1) and high-fat diet (6.9 ± 0.2). The impaired response to acetylcholine in rats fed a high-fat diet was not related to an alteration in soluble guanylate cyclase responsiveness as the administration of the NO donor sodium nitroprusside promoted an identical dose-dependent relaxation of aortic rings from both groups (Fig. 3).

The impaired acetylcholine-mediated relaxation of aortic rings isolated from rats fed a high-fat diet may be related to either a decrease of NO synthesis and/or a reduction of NO bioavailability secondary to an increase of vascular oxidative stress. In this regard, the antioxidant resveratrol (0.1 μM) reported to also stimulate NO synthase activity was added to aortic rings isolated from rats fed a high-fat diet, and acetylcholine-mediated relaxation was normalized (47 ± 9%) (Fig. 2). By contrast, the exogenous administration of resveratrol to aortic rings isolated from rats fed a standard diet had no effect on acetylcholine-mediated relaxation (53 ± 5%) (Fig. 2). Lastly, vascular sensitivity (pD₂) to acetylcholine of resveratrol-treated aortic rings isolated from rats fed either a standard (7 ± 0.1) or high-fat diet (7 ± 0.1) was similar. Thus, the beneficial in vitro effect of resveratrol on acetylcholine-mediated relaxation of aortic rings isolated from high-fat fed rats provided the impetus to examine whether the in vivo administration of the antioxidant could exert a therapeutic effect.

Morphometric Data and Plasma Lipid Profile: The in Vivo Effect of Resveratrol. The administration of resveratrol to rats fed either a standard or high-fat diet was associated with a modest but nonsignificant decrease of overall body weight and body weight gain compared to the appropriate untreated group (Table 1). The modest decline of body weight was attributed in part to the reduction of food intake in rats receiving resveratrol (Table 1). Lastly, resveratrol administration did not alter either heart weight, LV weight, or the lipid profile in rats receiving either a standard or high-fat diet (Tables 1 and 2).

Blood Pressure and Left Ventricular Function: The in Vivo Effect of Resveratrol. The administration of resveratrol to rats fed a standard diet caused a modest nonsignificant decrease in blood pressure, left ventricular systolic pressure, and LV dP/dt indices compared with untreated rats fed a standard diet (Table 3). By contrast, in rats fed a high-fat diet, the concomitant administration of resveratrol prevented the increase of systolic and diastolic blood pressure, mean arterial pressure, and left ventricular systolic pressure (Table 3). Moreover, the administration of the antioxidant to rats fed a high-fat diet significantly decreased both LV dP/dt indices compared to untreated rats. Lastly, resveratrol administration did not influence the dobutamine response in rats fed either a standard or high-fat diet (Table 4).

Cardiac Remodeling: The in Vivo Effect of Resveratrol. Resveratrol did not alter either vessel lumen area (high-fat diet, 0.0201 ± 0.003 mm², versus high-fat diet + resveratrol, 0.0155 ± 0.003 mm²) or wall thickness (high-fat diet, 0.0156 ± 0.002 mm, versus high-fat diet + resveratrol, 0.0164 ± 0.002 mm) in rats fed a high-fat diet. Furthermore, resveratrol administration to rats fed a high-fat diet failed to prevent the progression of perivascular fibrosis (Table 6). In the LV of rats fed a standard diet, resveratrol therapy did not influence TGF-β₃ mRNA expression (3.14 ± 0.29; n = 5). However, the down-regulation of TGF-β₃ mRNA (1.49 ± 0.36, n = 8) in rats fed a high-fat diet was prevented following the concomitant administration of resveratrol (3.49 ± 0.31; p < 0.05 versus high-fat diet; n = 5 rats).

Vascular Reactivity: The in Vivo Effect of Resveratrol. The administration of resveratrol to rats receiving a standard diet enhanced acetylcholine-mediated relaxation of aortic rings compared to untreated rats (Fig. 4). Moreover, resveratrol treatment of rats fed a high-fat diet significantly improved acetylcholine-mediated relaxation of isolated aortic rings, and the magnitude of relaxation was comparable to that observed in rats fed a standard diet (Fig. 4). The vascular sensitivity (pD₂) to acetylcholine was unchanged in both resveratrol-treated rats groups (resveratrol + standard diet, 7 ± 0.02; resveratrol + high-fat diet, 7 ± 0.12) compared to their respective untreated groups.

Discussion

The relationship between obesity and cardiovascular disease has been unequivocally established clinically and in experimental models. Despite these observations, it remains presently unknown whether vascular reactivity is impaired and maladaptive cardiac remodeling occurs before an overt gain in body weight and in the absence of hyperlipidemia. The present study has demonstrated that normal female rats provided with a high-fat diet for a period of 8 weeks were associated with a significant elevation of blood pressure and extensive perivascular fibrosis in the myocardium despite a nonsignificant increase in body weight and a normal plasma lipid profile and glucose concentration. Furthermore, the blood pressure increase in rats fed a high-fat diet was prevented by the coadministration of the antioxidant resveratrol. However, the therapeutic effect was limited to the vasculature as resveratrol treatment failed to attenuate the perivascular fibrotic response in the heart of rats fed a high-fat diet. An important paradigm established in the present study was that feeding normal female rats a high-fat diet for a period of 8 weeks caused a modest nonsignificant increase in body weight and plasma triglyceride levels without a change in the plasma lipid profile or glucose concentration. The administration of a comparable high-fat diet in a similar time frame to normal male rats led to a significant increase in body weight (Dobrian et al., 2000; Roberts et al., 2001; Erdei et al., 2006). Likewise, a similar sex-dependent relationship with regard to body weight gain was documented in various strains of mice fed a high-fat diet (Nishikawa et al., 2007). Thus, these data further support the premise that the effect of a high-fat diet on body weight gain is temporally delayed in female rats compared to male rats.

Elevated blood pressure was reported in normal male rats fed a high-fat diet characterized by a marked increase in body weight gain and hyperlipidemia (Dobrian et al., 2000; Erdei et al., 2006). In the present study, normal female rats fed a high-fat diet were likewise associated with an increase in blood pressure and left ventricular systolic pressure. However, these changes occurred before a significant increase of body weight and in the absence of hyperlipidemia and hyperglycemia. Previous studies have demonstrated that elevated mean arterial pressure in obese rats was related to an impairment of endothelium-mediated relaxation of vascular tissue (Dobrian et al., 2001; Erdei et al., 2006; Galili et al., 2007). In the present study, acetylcholine-mediated relaxation of aortic rings isolated from rats fed a high-fat diet was significantly compromised, whereas sodium nitroprusside-mediated relaxation was intact. Collectively, these data suggest that NO production was diminished and/or bioavailability was reduced secondary to the increased production of superoxide anions (Dobrian et al., 2001; Erdei et al., 2006; Roberts et al., 2006; Galili et al., 2007). In addition to its well established antioxidant property, resveratrol was also reported to increase NO synthesis (Wallerath et al., 2002; Wang et al., 2007). In the present study, the exogenous administration of resveratrol to aortic rings isolated from rats fed a high-fat diet normalized acetylcholine-mediated relaxation. Thus, these data strongly support the premise that impaired vascular relaxation in female rats fed a high-fat diet was related to an increased oxidative stress and/or decreased synthesis of NO and occurred in the absence of overt obesity, hyperlipidemia, or hyperglycemia.

It has been well established that elevated systemic arterial hypertension promotes a concentric pattern of cardiac hypertrophy characterized by the increased expression of ANP mRNA and concomitant down-regulation of SERCA2 mRNA (Grossman et al., 1975; Calderone et al., 1995). In the present study, despite an increase in blood pressure in rats fed a high-fat diet, neither ANP nor SERCA2 mRNA levels were altered in the LV. These data were consistent with the lack of change of either absolute LV weight or LV/body weight ratio in rats fed a high-fat diet compared with rats fed a standard diet. Thus, at least in female rats fed a high-fat diet for a period of 8 weeks, elevated blood pressure was not associated with a concomitant hypertrophic response. It is possible that the increase in blood pressure was insufficient to promote hypertrophy or a more chronic exposure to a hypertensive state was required. By contrast, an extensive perivascular fibrotic response was observed in the heart of female rats fed a high-fat diet. Likewise, collagen α₁ protein content was modestly increased in the myocardium of female rats fed a high-fat diet but did not reach statistical significance. Collectively, these data were consistent with previous studies demonstrating the presence of a reactive fibrotic response in the myocardium of various experimental animal models of obesity (Zaman et al., 2004; Carroll and Tyagi, 2005). However, the present study has revealed that the reactive fibrotic response in the heart of female rats fed a high-fat diet represents an early maladaptive event independent of overt obesity, hyperlipidemia, and hyperglycemia. Furthermore, the induction of putative profibrotic peptides in the heart of obese animals was implicated in reactive fibrosis (Zaman et al., 2004; Carroll and Tyagi, 2005). Unexpectedly, a significant down-regulation of transforming growth factor-β₃ mRNA was observed in the LV of rats fed a high-fat diet. Presently, the underlying mechanism(s) attributed to the latter paradigm remains unknown. Moreover, neither connective tissue growth factor nor transforming growth factor-β₁ mRNAs were up-regulated in the myocardium of female rats fed a high-fat diet. Despite these findings, we cannot exclude the possibility that these peptide growth factors were selectively increased in the vasculature, thereby directly contributing to the perivascular fibrotic response (Zaman et al., 2004; Ruiz-Ortega et al., 2007). Lastly, the suppression or attenuation of antifibrotic events may likewise contribute to cardiac fibrosis. Indeed, a dysfunction of NO signaling is sufficient to promote both reactive fibrosis and the up-regulation of putative profibrotic peptides in the heart (Goto et al., 1999; Ruetten et al., 2005), and the present study has demonstrated that either a decrease in NO synthesis or bioavailability was reduced in the aorta of female rats fed a high-fat diet.

The efficacy of exogenous resveratrol to normalize acetylcholine-mediated relaxation of aortic rings isolated from rats fed a high-fat diet provided the impetus to assess whether the in vivo administration would likewise attenuate the rise in blood pressure. Indeed, resveratrol administration to rats fed a high-fat diet prevented the hypertensive response. Furthermore, the beneficial in vivo hemodynamic effect of resveratrol was associated with the preservation of acetylcholine-mediated relaxation of aortic rings isolated from rats fed a high-fat diet. Thus, these data support the premise that resveratrol represents an appropriate pharmacological approach to prevent vascular dysfunction in female rats fed a high-fat diet. The scavenging of superoxide anions, increasing NO synthesis, and the phytoestrogenic property of resveratrol represent underlying mechanisms that may have synergistically prevented the rise of mean arterial pressure in rats fed a high-fat diet (Gehm et al., 1997; Wallerath et al., 2002; Baur and Sinclair, 2006; Wang et al., 2007).

The relationship between hypertension and cardiac fibrosis has been well established (Berk et al., 2007). Moreover, increased oxidative stress represents a putative feature of obesity and was further shown to participate in the progression of cardiac fibrosis either directly or indirectly via the suppression of NO synthesis (Dobrian et al., 2000; Ruetten et al., 2005; Erdei et al., 2006; Lu et al., 2008). In this regard, an attenuation of the perivascular fibrotic response would be expected in the heart of resveratrol-treated rats fed a high-fat diet based on the beneficial effect on blood pressure and its established antioxidant property. Furthermore, a direct antifibrotic action was identified in vitro as resveratrol suppressed cardiac fibroblast proliferation (Olson et al., 2005; Wang et al., 2007). Despite these observations, the in vivo administration of resveratrol failed to prevent the progression of perivascular fibrosis in the heart of rats fed a high-fat diet. Nonetheless, cardiac remodeling was sensitive to resveratrol therapy as the decreased expression of transforming growth factor-β₃ mRNA in the heart of rats fed a high-fat diet was prevented. The latter beneficial effect may be related in part to the phytoestrogenic property of resveratrol as 17β-estradiol treatment of either cardiac fibroblasts or osteoclasts increased transforming growth factor-β₃ mRNA expression (Yang et al., 1996; Mercier et al., 2002a). Thus, in female rats fed a high-fat diet, the perivascular fibrotic response was at least independent of a rise in blood pressure. Furthermore, the normalization of acetylcholine-mediated relaxation of aortic rings isolated from resveratrol-treated female rats fed a high-fat diet indirectly suggests that the increased perivascular fibrotic response in these hearts may not be secondary to a decreased bioavailability or synthesis of NO.

In conclusion, the present study has demonstrated that, prior to the development of overt obesity and in the absence of hyperlipidemia and hyperglycemia, female rats fed a high-fat diet were associated with vascular dysfunction and perivascular fibrosis. Resveratrol administration preserved vascular function, whereas the perivascular fibrotic response in the heart of female rats fed a high-fat diet persisted. Thus, these data demonstrate that disparate events are linked to the development of hypertension and perivascular fibrosis in female rats fed a high-fat diet.