Journal of Applied Physiology

A high-fat, refined-carbohydrate diet induces endothelial dysfunction and oxidant/antioxidant imbalance and depresses NOS protein expression

Christian K. Roberts, R. James Barnard, Ram K. Sindhu, Michael Jurczak, Ashkan Ehdaie, Nosratola D. Vaziri


We tested whether consumption of a high-fat, high-sucrose (HFS) diet can affect endothelium-dependent relaxation, whether this precedes the development of diet-induced hypertension previously noted in this model, and whether it is mediated, in part, by changes in nitric oxide synthase (NOS) and/or NOS regulatory proteins. Female Fischer rats were fed either a HFS diet or standard low-fat, complex-carbohydrate chow starting at 2 mo of age for 7 mo. Vasoconstrictive response to KCl and phenylephrine was similar in both groups. Vasorelaxation to acetylcholine was significantly impaired in the HFS animals, and there were no differences in relaxation to sodium nitroprusside, suggesting that the endothelial dysfunction is due, at least in part, to nitric oxide deficiency. HFS consumption decreased protein expression of endothelial NOS in aorta, renal, and heart tissues, neuronal NOS in kidney, heart, aorta, and brain, and inducible NOS in heart and aorta. Caveolin-1 and soluble guanylate cyclase protein expression did not change, but AKT protein expression decreased in heart and aorta and increased in kidney tissue. Consumption of HFS diet raised brain carbonyl content and plasma hydrogen peroxide concentration and diminished plasma total antioxidant capacity. Because blood pressure, which is known to eventually rise in this model, was not as yet significantly elevated, the present data suggest that endothelial dysfunction precedes the onset of diet-induced hypertension. The lack of a quantitative change in caveolin-1 and soluble guanylate cyclase protein content indicates that alteration in these proteins is not responsible for the endothelial dysfunction. Thus nitric oxide deficiency combined with antioxidant/oxidant imbalance, appears to be a primary factor in the development of endothelial dysfunction in this model.

  • reactive oxygen species
  • hypertension
  • oxidative stress
  • blood pressure
  • nitric oxide synthase

endothelial dysfunction is known to be associated with atherogensis and hypertension (18). Recent evidence suggests that a decrease in nitric oxide (NO) availability may contribute to endothelial dysfunction, leading to hypertension. Additionally, Western diets have been reported to adversely affect blood pressure in both humans (1) and animals (40, 41). Recently, our laboratory has documented that a high-fat, refined-carbohydrate (HFS) diet fed to rats and designed to be analogous to that consumed in Westernized societies induces insulin resistance and hypertension (4). Diet-induced hypertension was first noted after 1 year and was induced, at least in part, by oxidative stress (29), which may have diminished NO production and/or enhanced NO sequestration (29). Regarding depressed NO production, our laboratory's previous study suggested that a decrease in l-arginine availability was not responsible for diet-induced hypertension (29), and an alternative candidate, quantitative NO synthase (NOS) deficiency, was suggested. Additionally, several proteins regulate NO production and action including caveolin-1 (Cav-1), soluble guanylate cyclase (sGC), and AKT. Thus the purpose of the present study was to determine whether HFS diet consumption induces endothelial dysfunction in female rats before the development of hypertension and to identify the possible mechanism(s) that may contribute to endothelial dysfunction.

Having used female rats in our laboratory's previous studies, we examined the effects of diet on endothelial function in female rats fed either a HFS or standard low-fat, complex-carbohydrate (LFCC) rat chow diet for 7 mo, a time point before the development of hypertension in this model (29). Our experimental design allowed for the measurement of endothelial function before the development of hypertension. Four hypotheses were tested: HFS diet consumption would 1) induce prehypertensive endothelial dysfunction, which would be due in part to altered NO availability; 2) increase reactive oxygen species (ROS) and decrease antioxidant capacity; 3) reduce protein expression of NOS isotypes; and 4) alter expression of Cav-1, sGC-β, and/or AKT, contributing to endothelial dysfunction. Endothelial function was measured using vascular rings incubated with acetylcholine (ACh) or sodium nitroprusside (SNP). Plasma and tissues were used to determine protein expression of the aforementioned proteins and indexes of oxidative stress.


Animals and Diet

All protocols were approved and conducted in accordance with the University of California, Los Angeles, Animal Research Committee. Two-month-old female Fischer 344 rats were obtained from Harlan Sprague Dawley (San Diego, CA). We have used this rat model in our previous studies, because the female Fischer rat normally shows little weight gain after its maturation phase (3, 4). The animals (8 per group) were housed four animals per cage with a 12-h light cycle starting at 0700 at 75–76°F and were allowed to acclimatize to their environment for 1 wk, consuming standard rat chow (Purina 5001), before the dietary intervention was initiated. The animals were randomly assigned to either the LFCC or HFS diet and fed the diets and water ad libitum. The diets were prepared in powder form by Purina Test Diets (Richmond, IN) and contained a standard vitamin and mineral mix and all essential nutrients. The LFCC diet (Purina 5001) is low in saturated fat and contains mostly complex carbohydrates, whereas the HFS diet is high in saturated and monounsaturated fat (primarily from lard plus a small amount of corn oil) and high in refined-sugar (sucrose) as previously published (26). Energy intake was determined by measuring daily consumption of food in each cage and dividing by the number of animals per cage. Animals were anesthetized and exsanguinated by cardiac puncture, and plasma was stored at −80°C until processed. The tissues were flash frozen in liquid nitrogen and stored at −80°C until processed.

Blood Pressure

Blood pressure was measured by tail-cuff plethysmography as previously described (29).

Measurement of Endothelial Function

The animals were anesthetized by intraperitoneal injection of ketamine (100 mg/kg) and xylazine (5 mg/kg). After adequate anesthesia, a thoracotomy was performed to expose the heart for blood sampling. The descending thoracic aorta was then carefully removed and placed in chilled Krebs-Henseleit physiological solution containing (in mM) 131.5 NaCl, 5.0 KCl, 1.2 MgCl2·6H2O, 2.5 CaCl2·2H2O, 1.2 NaH2PO4·H2O, 11.2 glucose, and 20.8 NaHCO3. All remaining fat and connective tissue were gently removed, and the aorta was sectioned into 5-mm segments. The segments were mounted on standard tungsten wire triangles (AM Systems, Everett, WA), attached to isometric force displacement transducers (FTO3C, Grass Instrument, Quincy, MA), and placed into tissue baths. The transducer output was amplified and recorded continuously on a portable computer with digital analysis software (Femto Tek, Mt. Laurel, NJ). The tissue baths were temperature controlled via a heated water jacket at 37°C. A mixture of 95% oxygen-5% carbon dioxide gas mixture was bubbled into the tissue baths. Preload (2 g) was applied to the arterial rings, and the vessels were allowed to equilibrate for 45 min. After equilibration, the arteries were constricted using a high-potassium Krebs solution (18 mM KCl) and allowed to reequilibrate. This represented the maximal contractile force for the artery (Fmax). The baths were then emptied and rinsed three times with Krebs solution and again allowed to equilibrate. The arteries were then constricted to 75% of Fmax with phenylephrine (10−5 M, Sigma, St. Louis, MO). ACh (Sigma) was then added in incremental log concentrations from 10−8 to 10−4 M for determination of endothelium-dependent relaxation. Endothelium-independent relaxation was measured in a similar fashion using SNP (Sigma) in incremental log concentrations from 10−9 to 10−5 M. Baths were rinsed and arterial segments were brought to 75% Fmax with phenylephrine between each reagent. Tension was measured in grams, and contraction and relaxation were recorded as percentages of 75% Fmax for each incremental dose of reagent.

Plasma Hydrogen Peroxide and Total Antioxidant Power

Plasma hydrogen peroxide concentration was determined by the quantitative hydrogen peroxide assay kit (catalog no. 21024, OXIS International, Portland, OR). In brief, this assay is based on the oxidation of ferrous ions (Fe2+) to ferric ions (Fe3+) by hydrogen peroxide under acidic conditions. Ferric ions bind with the indicator dye xylenol orange {3,3′-bis[N,N-di(carboxymethyl)-aminomethyl]-o-cresolsulfone-phthalein, sodium salt} to form a stable colored complex and measured at 560 nm. Total antioxidant potential was quantified by assaying the reduction of Cu2+ to Cu+ by the combined action of all antioxidants present in plasma using the chromogenic reagent bathocuproine (2,9-dimethyl-4,7-diphenyl-1,10-phenanthroline), which selectively forms a 2:1 complex with Cu+, which has a maximum absorbance at 490 nm (BIOXYTECH AOP-490, catalog no. 21052, Oxis International). This assay has a coefficient of variation of 2.2% for intra-assay and 4.2% for interassay.

Measurement of Protein Carbonyls

Protein carbonyls were measured in the brain using an OxyBlot protein oxidation detection kit (Intergen, Purchase, NY), which can detect protein carbonyls in the femtomolar range. After denaturation, 15–20 μg of cytosolic proteins were treated with 2,4-dinitrophenylhydrazine for 15 min, followed by electrophoresis on 4–20% SDS-polyacrylamide gels. Standard Western blotting procedures were followed thereafter using primary (1:150) and secondary (1:500) antibodies supplied with the kit. Enhanced chemiluminescence detection reagents (Amersham Biosciences, Piscataway, NJ) were used to generate a chemiluminescent signal, and bands were visualized by exposing the membranes to autoradiography film.

Immunoblot Analyses

Immunoblotting was performed on thoracic aorta, heart, and kidney of both diet groups to detect endothelial NOS (eNOS), neuronal NOS (nNOS), Cav-1, sGC-β, and AKT protein levels as previously described (37). nNOS was also determined in brain cortex and inducible NOS (iNOS) was determined in aorta and heart.

Homogenates (25% wt/vol) of kidney, heart, and thoracic aorta were prepared in 10 mM HEPES buffer, pH 7.4, 1 mM EDTA, 1 mM DTT, 10 μg/ml leupeptin, 2 μg/ml aprotinin, and 1 mM phenylmethylsulfonyl fluoride at 0–4°C with a Polytron tissuemizer. Homogenates were centrifuged at 9,000 g for 10 min at 4°C to remove nuclear fragments and tissue debris without precipitating plasma membrane fragments. A portion of the supernatant was used for the determination of total protein concentration by using a Bio-Rad kit (Hercules, CA).

Total cellular protein (20 μg each) was electrophoresed in 4–20% Tris-glycine sodium SDS polyacrylamide gels (Novex). Proteins were transferred onto polyvinylidene fluoride membranes (Millipore, Bedford, MA), blocked in 5% dry milk in Tween 20 TBS (TTBS; 0.02 M Tris/0.15 M NaCl, pH 7.5, containing 0.1% Tween 20) at room temperature for 3 h, washed three times with TTBS, and incubated with the following primary antibodies for 3 h at room temperature: anti-eNOS (catalog no. N30030), anti-iNOS (catalog no. N39120), and anti-nNOS (catalog no. 610309) monoclonal antibodies (Transduction Laboratories, Lexington, KY, 1:1,000), anti-sGC-β subunit (catalog no. 371712, Calbiochem, 1:1,000), anti-Cav-1 (catalog no. PA1–064, Affinity Bioreagents, Golden, CO, 1:1,000), and anti-AKT (catalog no. 610861, BD Biosciences, 1:2,000). After being washed five times with TTBS, the blots were incubated with secondary antibodies (anti-rabbit, catalog no. NA 934V, 1:1,000 for sGC and Cav-1) and anti-mouse [catalog no. NA 931V, 1:1,000 for eNOS, iNOS, and nNOS, 1:2,000 for AKT] from Amersham Biosciences, Piscataway, NJ, conjugated with horseradish peroxidase at room temperature for 2 h. After washing five times with TTBS, the membranes were developed using enhanced chemiluminescent reagent (Amersham Biosciences) and subjected to autoluminography for 1–5 min. The autoluminographs were scanned with a laser densitometer (model PD 1211; Molecular Dynamics) to determine the relative optical densities of the bands. Normalization to β-actin (anti-β-actin, 1:5,000, catalog no. MAB 1501 R, Chemicon International, Temecula, CA) or glyceraldehyde-3 phosphate dehydrogenase (anti-glyceraldehyde-3 phosphate dehydrogenase, 1:2,500, catalog no. MAB 374, Chemicon International) was used to verify the uniformity of protein load and transfer efficiency across the tested tissues (data not shown).

Statistical Analysis

Data were analyzed using paired Student's t-tests. Vascular ring experiments were analyzed using an ANOVA, and post hoc analyses were performed when significant differences were noted using a Newman-Keuls multiple comparison test. Differences were considered statistically significant at P < 0.05. Values reported are means ± SE with six rats per group unless otherwise indicated.


Blood Pressure, Hydrogen Peroxide, Total Antioxidant Power, and Carbonyl Content

Although energy intake did not differ significantly between the two groups (data not shown), body weight was significantly higher in the HFS group relative to that found in the LFCC group (data previously published in Refs. 23 and 25). Blood pressure was elevated in HFS, but not significantly, after 7 mo of HFS diet consumption compared with the control diet (120 ± 2 vs. 131 ± 6 mmHg, LFCC vs. HFS, P > 0.05), in agreement with our laboratory's previous study demonstrating that, in female rats, no significant differences in blood pressure were noted at 6 mo and only became significant after 1 yr (4). Plasma hydrogen peroxide, measured as an indicator of ROS generation, was elevated in the HFS group compared with the LFCC group (Fig. 1A, P = 0.02). To test free-radical scavenging capacity, plasma total antioxidant power was determined, and HFS diet consumption resulted in a marked reduction in the antioxidant capacity of the plasma (P < 0.001; Fig. 1B). Additionally, protein oxidation was quantified by protein carbonyl determination using 2,4-dinitrophenylhydrazine. HFS animals had an ∼50% increase in protein carbonyls (P < 0.01; Fig. 1C), denoting an increased protein oxidation.

Fig. 1.

Effect of diet on hydrogen peroxide, total antioxidant capacity, and carbonyl content. Top: plasma concentration of hydrogen peroxide was significantly elevated in the high-fat, high-sucrose (HFS) group (*P = 0.02). Middle: HFS animals exhibited a marked reduction in antioxidant capacity of the plasma (*P < 0.01). Bottom: relative levels of oxidized protein in brain determined by a quantifiable Western blot analysis of 2,4-dinitrophenylhydrazine (DNPH)-derivatized carbonyls. Levels of protein oxidation were increased in HFS-fed animals (*P < 0.01). Values are means ± SE (n = 6 animals/group). LFCC, low fat, complex carbohydrate.

Vascular Responses

Arterial smooth muscle contraction.

There were no significant differences in maximal contraction to KCl between the two groups. Ring segments were also constricted with phenylephrine, and again there were no statistical differences noted between the groups in the amount of phenylephrine required to elicit 75% Fmax.

Endothelium-dependent relaxation.

Arterial segments from rats fed the HFS diet showed a significantly decreased endothelium-derived relaxation to ACh at concentrations from 10−8 to 10−4 M compared with rats fed the LFCC diet (P < 0.01 at all concentrations starting at 10−7 M; Fig. 2).

Fig. 2.

Effect of diet on dose-dependent endothelium-dependent and endothelium-independent relaxation to ACh (top) and sodium nitroprusside (SNP; bottom). Results are presented for thoracic aortas isolated from 6 HFS and 6 LFCC animals. Top: relaxation response to ACh. HFS exhibited endothelial dysfunction relative to controls from 10−8 to 10−4 M (*P < 0.01). Bottom: relaxation responses to SNP. Except for 10−8 M (*P < 0.05), no differences were noted. Values are means ± SE.

Endothelium-independent relaxation.

Ring segments were tested for endothelium-independent relaxation using SNP. Both groups relaxed to SNP in a dose-dependent manner (Fig. 2). At maximal SNP concentration (10−5 M), the HFS group relaxed to 94 ± 6% and the LFCC group relaxed to 95 ± 4%. Other than at 10−8 M, there were no significant differences between groups.

Immunoblotting for NOS, Cav-1, sGC-β, and AKT

Measurements of eNOS protein expression demonstrated reduction of eNOS protein content in aorta, heart, and kidney tissue in the HFS group compared with the LFCC (Fig. 3). Similarly, protein content of both iNOS and nNOS were reduced in the tissues tested in a uniform manner (P < 0.05; Fig. 3) in animals consuming the HFS diet. Protein expression of AKT, Cav-1, and sGC-β were also measured, because these proteins play vital roles in regulating NO production and activity. HFS diet consumption induced a downregulation in AKT abundance in aorta and heart and an upregulation in renal tissue (Fig. 4). However, there were no significant differences between the two groups in Cav-1 and sGC-β protein expression in any of the tissues tested (Fig. 4).

Fig. 3.

Effect of diet on nitric oxide synthase (NOS) isotypes. Left: representative Western blot of endothelial NOS (eNOS) in kidney (top), heart (middle), and aorta (bottom). Middle: neuronal NOS (nNOS) in kidney, heart, aorta, and brain (top to bottom, respectively). Right: inducible NOS (iNOS) in heart (top) and aorta (bottom). Corresponding group data are illustrated below immunoblots. y-Axes are relative optical density units. Bars show means ± SE; n = 6 observations/group. *P < 0.05 vs. LFCC.

Fig. 4.

Effect of diet on nitric oxide activity-related protein content. Left: representative Western blots of soluble guanylate cyclase (sGC)-β in kidney (top), heart (middle), and aorta (bottom) tissues. Middle: caveolin-1 (Cav-1) in kidney (top), heart (middle), and aorta (bottom) tissues. Right: AKT in kidney (top), heart (middle), and aorta (bottom) tissues. Group data are below immunoblots, illustrating relative optical densities of protein bands in the study animals. y-Axes are relative optical density units. Bars show means ± SE; n = 6 observations/group. **P < 0.01 relative to other groups.


Endothelial dysfunction is a common feature of atherosclerosis and chronic hypertension. Evidence also suggests that increased ROS generation is an important aspect of vascular dysfunction, and increases in ROS activity accelerate NO inactivation, resulting in decreased bioactive NO and increased vascular resistance. Several recent studies (6, 32, 40) including our laboratory's (26, 29) have indicated that dietary fat, fructose, and sucrose can induce endothelial dysfunction and hypertension. In a series of recent studies (4), our laboratory documented that an HFS diet fed to rats and designed to be analogous to that consumed in Westernized societies induced insulin resistance within 2 wk and hypertension after 1 yr. This diet-induced hypertension was accompanied by a marked reduction in urinary NO metabolites and appeared to be caused by oxidative stress (29). Accordingly, the reduction of urinary NO metabolites could be due to diminished NO production and/or enhanced ROS-mediated NO sequestration. In regard to the latter possibility, superoxide reacts with and inactivates NO to produce peroxynitrite, a potent cytotoxic reactive nitrogen species that subsequently reacts with proteins, lipids, and DNA (13). Peroxynitrite can react with free tyrosine or tyrosine residues in protein molecules to form nitrotyrosine, a stable footprint of ROS-mediated inactivation of NO. Our laboratory previously demonstrated elevated nitrotyrosine abundance in various tissues of rats maintained on the HFS diet for 2 mo (29), reflecting ROS-mediated NO inactivation and sequestration in this model (8, 9, 13).

Previously, our laboratory documented the effect of diet on endothelium-dependent relaxation in male rats with hypertension (22). In this study, we used female rats that had not yet developed diet-induced hypertension, and, accordingly, we asked the question of whether endothelial dysfunction precedes hypertension and whether NO deficiency may play a role. The primary findings of this study are that 1) HFS consumption induces endothelial dysfunction in female rats before the onset of hypertension; 2) endothelium-independent relaxation is not impaired in HFS-fed rats; 3) HFS diet consumption increases hydrogen peroxide and protein carbonyl content, while decreasing plasma antioxidant capacity, pointing to oxidative stress and its possible contribution to the noted endothelial dysfunction; 4) the defective vasorelaxation response to ACh may be mediated, in part, by decreased expression of NOS protein, particularly eNOS; and 5) the defects in endothelium-dependent relaxation are apparently not the result of altered expression of Cav-1 and sGC-β but are associated with decreased content of the protein kinase AKT, which phosphorylates NOS in endothelial cells to augment NO production and vasodilation.

Effect of Diet on Endothelial Function

ACh was used to assess the effects of diet on endothelial function, and HFS diet consumption induced endothelial dysfunction in female rat aorta, in agreement with responses noted in male aorta (22). Thompson et al. (33) using coronary arteries and Woodman et al. (39) using brachial arteries demonstrated that high-fat feeding alone may induce endothelial dysfunction in male and female pigs, respectively. The HFS diet included sucrose as its main carbohydrate source, which may contribute to the observed abnormalities, because it has been shown that high-sugar diets containing fructose induce endothelial dysfunction (16). Additionally, Shinozaki et al. (32) demonstrated that a high-fructose diet increased NO inactivation, secondary to enhanced formation of superoxide, and decreased vascular relaxation through impaired eNOS activity caused by relative deficiency of tetrahydrobiopterin in endothelial cells. Given that sucrose contains high amounts of fructose, the ROS-induced tetrahydrobiopterin depletion may contribute to endothelial dysfunction noted in the present study, and further studies should address this possibility. Additionally, SNP was used to assess the effects of diet on vascular smooth muscle function, and SNP-induced relaxation was similar in both HFS and LFCC fed animals, indicating that the ability of vascular smooth muscle to relax in response to exogenous NO was not impaired in HFS-fed animals and that diet selectively impaired endothelium-dependent vasodilation. Interestingly, when we previously blocked NO production using NG-nitro-l-arginine methyl ester, blood pressure increased in LFCC rats to the level of HFS rats (29). It is possible that a portion of the defects in vasorelaxation response to ACh may be due to reduction of endothelium-dependent hyperpolarizing factor (31). However, endothelium-dependent hyperpolarizing factors normally play a minor role in the vasodilatory response in the aorta. Limitations in our study were that we did not assess endothelial function in resistance arteries, which are more important for control of blood pressure, nor did we assess endothelial function in the presence of an antioxidant to confirm that oxidative stress was a cause of endothelial dysfunction.

Effect of Diet on Oxidative Stress

To investigate the effect of diet on oxidative stress, we measured plasma hydrogen peroxide and antioxidant capacity. In agreement with our hypothesis, there was an increase in hydrogen peroxide and a decrease in antioxidant capacity of the plasma. Additionally, protein carbonyl levels were increased in the HFS animals, providing evidence for protein oxidation. These data corroborate our previous findings of increased nitrotyrosine abundance (29) and malondialdehyde (27), documenting diet-induced antioxidant/oxidant imbalance in this model as evidenced by several indexes of oxidative stress. The associated oxidative stress must have, in part, contributed to endothelial dysfunction in HFS-fed animals (36).

Effect of Diet on NOS Protein Content

The results demonstrated that the reduction in endothelial function may be attributed, in part, to a quantitative eNOS deficiency, since HFS diet consumption reduced the protein content of eNOS in aorta. In fact, widespread downregulation of eNOS, iNOS, and nNOS were noted. Dobrian et al. (7) reported the occurrence of hypertension, oxidative stress, and compensatory upregulation of aortic and renal eNOS mRNA in a subgroup of rats designated as obesity prone. Additionally, our laboratory previously noted that long-term HFS diet consumption results in a compensatory upregulation of renal NOS expression (28) in agreement with Dobrian et al., which suggests that there may in fact be a time-dependent alteration in NOS protein expression. Woodman et al. (39) noted that high-fat feeding induced a reduction in brachial artery eNOS immunohistochemical staining, and this was ameliorated by exercise. Thus we hypothesize that in the early phase, before onset of hypertension, diet-induced oxidative stress and eNOS deficiency promote endothelial dysfunction, whereas in the latter phase (2 yr), when hypertension has developed, eNOS may be upregulated possibly by increased shear stress and/or as a compensatory attempt to reestablish blood pressure homeostasis (17, 30, 35).

nNOS is normally expressed in various regions of the brain and is thought to be involved in neurogenic control of blood pressure by inhibiting central sympathetic outflow (14, 34). Consequently, nNOS-derived NO in the brain is considered to exert a blood pressure-lowering influence. If true, downregulation of brain nNOS content as noted in the present study may potentially contribute to subsequent blood pressure elevation in this model. Additionally, HFS diet consumption caused a reduction in renal nNOS expression, which is normally expressed in different parts of the kidney, particularly in tubular structures, macula densa cells, and endothelium of efferent arterioles. This may also contribute to the development of hypertension by augmenting the tubuloglomerular feedback response and modulating renal microvascular function (2, 38).

Although the role of eNOS in the regulation of cardiovascular function is well characterized, the role of iNOS is less clear. It should be noted that, contrary to the conventional view, iNOS may be constitutively expressed in several tissues, such as heart, vascular smooth muscle, and kidney (19, 20). The present data are in agreement with these reports, as iNOS was detected in aorta and heart. This constitutive expression of iNOS suggests a homeostatic role separate from its immunologically mediated induction.

Effect of Diet on NOS Regulatory Proteins

We tested whether altered Cav-1 expression might contribute to endothelial dysfunction. Incubation of cultured bovine aortic endothelial cells with serum from individuals with high cholesterol increases Cav-1 protein and decreases NO production (10). Contrary to our hypothesis, Cav-1 protein content in aorta, heart, and renal tissues was not altered by HFS consumption. Thus our results suggest that defects in endothelium-dependent relaxation may be due to changes in NOS expression as opposed to eNOS-Cav-1 interactions (12). Woodman et al. (39) reported that high-fat feeding did not alter expression of Cav-1 in brachial artery rings. Although Cav-1 did not increase, we cannot rule out the possibility that diet affected the interaction between existing Cav-1 and eNOS content (15).

AKT has been documented to phosphorylate and activate eNOS (5, 11) independent of classical Ca2+/calmodulin-dependent mechanisms. To the best of our knowledge, this is the first study to investigate the effect of diet on vascular AKT protein content. As hypothesized, AKT protein expression in aorta and cardiac tissue was decreased in HFS-fed animals. On the other hand, renal AKT protein expression was increased. This indicates that eNOS activation by AKT may be modulated by dietary fat and sugar. The aortic reduction in AKT protein content is especially of interest, given that AKT phosphorylation of eNOS can modulate its enzymatic activity, which may contribute to reducing NO availability in diet-induced endothelial dysfunction and hypertension. However, further studies are needed to explore the effect of diet on eNOS phosphorylation by AKT.

Finally, by reacting with the heme moiety of sGC-β, NO activates this enzyme, which catalyzes the formation of the second messenger, cyclic guanosine monophosphate. The present study suggests that the defects in endothelial function are not related to diminished sGC-β protein expression, and normal vascular relaxation to SNP supports this contention.


Consumption of HFS diet leads to endothelial dysfunction, which is due to a combination of oxidant/antioxidant imbalance and downregulation of vascular NOS, reducing NO production capacity. HFS diet-induced oxidative stress would result in NO sequestration and inactivation, contributing to endothelial dysfunction and subsequent development of hypertension (29). Additionally, reduction in NOS isotypes was noted in various tissues, suggesting a systemic effect of this diet. The effects of diet do not appear to be due to altered vascular smooth muscle responsiveness to NO or to quantitative alterations in Cav-1 or sGC-β protein content.


Given the high incidence of cardiovascular morbidity and mortality in developed nations, therapies for optimal cardiovascular health are needed. The prevailing epidemic of cardiovascular diseases is, in part, linked to dietary factors. Oxidative stress and endothelial dysfunction are critical in the pathogenesis of atherosclerosis and hypertension and may be mitigated by diet modification. In this context, Reil et al. (22) noted that diet modification for 1 mo in male rats mitigated endothelial dysfunction and Plotnick et al. (21) noted that antioxidant vitamins could ameliorate endothelial dysfunction induced by high-fat meal consumption, whereas Roberts et al. (24) demonstrated that diet and physical activity for 3 wk ameliorated oxidative stress and increased urinary NO metabolites in humans with hypertension. The present study confirms the ability of a Westernized diet to produce endothelial dysfunction, which was related to oxidative stress and downregulation of NOS and its activator AKT in the aorta, a conduit vessel frequently affected by atherosclerosis in humans.


C. Roberts was supported by a National Research Scholarship Award postdoctoral fellowship, National Heart, Lung, and Blood Institute Grant F32 HL-68406-01 during this project.


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


View Abstract