INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE MODULATION of vascular smooth muscle tone involves the release of potent vasorelaxants such as nitric oxide (·NO) (26) and prostacyclin (27), which counterbalance the vasoconstrictor properties of thromboxane A₂ (38) and the superoxide anion (O·) (15) among other agents. It has also been shown that ·NO has antiatherogenic properties both in vitro and in vivo (18). However, it has been demonstrated at a vascular level that ·NO reacts with O· to give peroxynitrite, reducing its relaxant effect (40).

Fish oils are the main source of human dietary long chain -3 polyunsaturated fatty acids. A fish oil-rich diet, by replacing arachidonic acid by eicosapentanoic acid (EPA) and docosahexanoic acid (DHA) in plasma and in phospholipid membranes (29, 32), impairs the release of free radicals by different types of cells and induces changes in eicosanoid metabolites (3, 24). These oils are used in the prevention of cardiovascular disease such as atherosclerosis (12). Several studies have been carried out on the effect of -3 polyunsaturated fatty acids on the release of relaxing factors (2, 5, 33, 34) and on the reduction of blood pressure in healthy volunteers (35) and in patients with mild hypertension (37). Also, its supplementation increases coronary artery vasodilation in response to acetylcholine infusion in heart transplant patients (10).

Little is known about the production of O· and ·NO, their interaction, or about the impairment of the activity of cyclooxygenase at a vascular level by a fish oil-rich diet. The purpose of this study was to test the effect of a fish oil-rich diet on vascular smooth muscle reactivity, especially through O· and ·NO production and through cyclooxygenase stimulation by acetylcholine in rat aortic rings.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals and diets. After weaning, two groups of male Sprague-Dawley rats were fed for 8 wk with semipurified diets prepared in our laboratory (Table 1), which contained 5% lipids. The lipids were either corn oil (CO), rich in 18:2-6, or menhaden oil (MO), rich in 20:5-3 (EPA) and 22:6-3 (DHA). The content of some of the fatty acids in the diets was evaluated according to the technique of Haan et al. (13) and is shown in Table 2. The oils provided ~2 mg -tocopherol per kilogram diet. Diets were supplemented with 90 mg/kg of all rac--tocopherol acetate (equivalent to 60 IU/kg of -tocopherol). Diets were prepared weekly and stored at 20°C to prevent oxidation. The peroxide value of the MO diet was <10 meq/kg when ready for consumption. Food was provided daily, and uneaten food was also removed daily. Body weight was recorded every week. At the end of the feeding period, rats were anesthetized with sodium urethane (1.5 g/kg ip) and exsanguinated.

Aortic preparation. The thoracic and proximal-abdominal aortas were excised and placed in Krebs-Ringer bicarbonate solution at pH 7.4, which contained (in mM) 118.07 NaCl, 4.70 KCl, 1.77 CaCl₂ · 2H₂O, 1.17 KH₂PO₄, 1.17 MgSO₄ · 7H₂O, 24.04 NaHCO₃, and 12.2 glucose. The proximal-abdominal and thoracic aortas were carefully cleaned of debris and blood, with care taken not to touch the luminal surface, and they were then cut into three to four rings of 3-4 mm in length, respectively. The endothelium of some rings was removed mechanically by gentle rubbing of the intimal surface with a stainless steel wire. Endothelium-intact rings from the abdominal and thoracic aorta were randomized to study O· production. Intact and endothelium-denuded thoracic aortic rings were also randomized for organ bath studies. They were fixed between stainless steel hooks in a bath containing Krebs solution at 37 ± 0.5°C. The hook anchoring the upper end of the ring was connected by a silk thread to the isometric transducer. Rings were equilibrated at an initial tension of 2 g for 60 min, and the solution was changed every 15 min and gassed with a mixture of 95% O₂-5% CO₂. The procedures and the care of rats complied with European Community guidelines.

O· and peroxynitrite production. O· production was measured according to Pagano et al. (25) with some modifications. Test tubes containing endothelium-intact rings in Krebs-Ringer bicarbonate solution were placed in a water bath at 37°C for a 30-min equilibration period, and the solution was bubbled with a mixture of 95% O₂-5% CO₂. Rings were then placed in 1 ml of Krebs-Ringer bicarbonate solution in 1.6-ml polypropylene tubes containing 5 µM lucigenin and put in a luminometer (Bio Orbit; Turku, Finland), the light chamber of which was maintained at 37°C. The luminometer reported arbitrary units of light emitted. The chemiluminescence of lucigenin was taken in the basal condition and after addition of 100 µM NADPH. Two other rings from the same rat were evaluated in the same conditions but in the presence of 60 U/ml Cu/Zn superoxide dismutase (SOD), an O· scavenger, or 1 mM N^G-monomethyl-L-arginine (L-NMMA), an inhibitor of ·NO synthase. Both inhibitors were present during the 30-min incubation period. L-NMMA was used instead of N^G-nitro-L-arginine (L-NNA) because it does not interfere with NADPH-dependent heme reduction and thus does not generate O· (19). We calculated the chemiluminescence by integrating the values obtained over 2.5 min (25). The results obtained with lucigenin were subtracted from the ones obtained in the absence of lucigenin. After the luminometer assay, the rings were opened and extended on a board, and a photo was taken. The images were processed to calculate the surface area of the aortic rings. The results are expressed as arbitrary units per millimeter squared. Peroxynitrite formation was assessed in a similar protocol using 250 µM luminol (28).

Phenylephrine-induced vasoconstriction. By blocking either the O· released by exogenous SOD or the ·NO production by L-NNA, we can modulate the vasoconstriction. This allows us to indirectly assess the production of ·NO and O· in basal conditions, respectively. After the equilibration period, randomized rings were exposed to cumulative concentrations of the ₁-adrenergic agonist phenylephrine (from 10⁹ to 10⁴ M). The phenylephrine-induced contractions took place in some rings in the absence of exogenous SOD and in the presence of 60 U/ml SOD. The organ bath solution was changed three times between the two assays, and rings were allowed to equilibrate for at least 15 min to return to baseline. Other rings were used to study the effect of the addition of 60 U/ml SOD plus 0.1 mM L-NNA. L-NNA was added 20 min before the first concentration of phenylephrine. Endothelium-denuded rings were also exposed to the cumulative concentrations of phenylephrine mentioned above. Phenylephrine-induced contractions of aortic rings are expressed in milligrams of force. The effective molar concentration producing 50% of the maximal response (EC₅₀) was also calculated by linear curve-fitting data and expressed as a negative logarithm.

Acetylcholine-induced vasorelaxation. To assess the effect of MO on NO-mediated vasorelaxation, endothelium-intact randomized rings were precontracted with 2 × 10⁷ M phenylephrine and relaxed with cumulative concentrations of acetylcholine (from 10⁸ to 10⁴ M). This process was repeated in some rings 1) in the absence of exogenous SOD; 2) in the absence of SOD and in the presence of 0.1 mM L-NNA, added 20 min before the precontraction; 3) in the presence of 60 U/ml SOD, added immediately before the precontraction; and 4) in the presence of both SOD and L-NNA. To assess the implication of the cyclooxygenase pathway on acetylcholine-induced vasorelaxation, endothelium-intact randomized rings were precontracted with phenylephrine and relaxed with acetylcholine as decribed above. This process was repeated in the same ring 1) in the presence of 60 U/ml SOD, added immediately before the precontraction; 2) in the presence of both 60 U/ml SOD and 0.1 mM L-NNA, added 20 min before the precontraction; and 3) in the presence of 60 U/ml SOD, 0.1 mM L-NNA, and 0.1 mM indomethacin, added 20 min before the precontraction. The organ bath solution was changed three times after each assay, and rings were allowed to equilibrate for at least 15 min to recover to baseline. Acetylcholine-induced relaxations are expressed as a percentage of the level of precontraction. The EC₅₀ of acetylcholine was also calculated for each concentration-response curve by linear curve-fitting data and expressed as a negative logarithm.

Sodium nitroprusside-induced vasorelaxation in endothelium-denuded rings. The functional removal of the endothelium was assessed by either a precontraction response or absence of relaxation on exposure to 10⁴ M acetylcholine after a precontraction with 2 × 10⁷ M phenylephrine. After the rings were washed, they were exposed to cumulative concentrations of phenylephrine, and when the maximum contraction was attained, rings were exposed to cumulative concentrations of sodium nitroprusside (from 10¹⁰ to 10⁷ M), an exogenous ·NO donor. The sodium nitroprusside-induced relaxation was studied twice in the same ring in the absence and presence of SOD, changing the order of treatment in different rings. The organ bath solution was changed three times after each assay, and rings were allowed to equilibrate for at least 15 min to recover to baseline. Sodium nitroprusside-induced relaxations were expressed as a percentage of the level of precontraction. The EC₅₀ of sodium nitroprusside was calculated for each concentration-response curve by linear curve-fitting data and expressed as a negative logarithm.

Materials. Oils, casein, sucrose, -cellulose, DL-methionine, all-rac--tocopherol acetate, phenylephrine hydrochloride, acetylcholine chloride, L-NNA, L-NMMA, sodium nitroprusside, indomethacin, and SOD were purchased from Sigma (St. Louis, MO). Starch and FeSO₄ · 2H₂O were purchased from Panreac (Barcelona, Spain). Mineral mix (without iron) and vitamin mix (without vitamin E) were obtained from ICN Biomedicals (Aurora, OH).

Statistical analysis. Data are expressed as means ± SE. Data were analyzed by Student's t-test for unpaired or paired observations. Differences were considered to be significant when P values were <0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals. The increase in body weight over the 8-wk period was the same in the three groups of rats, and there was no difference in the final body weight: 343 ± 5 and 343 ± 8 g for CO- and MO-fed rats, respectively.

O· and peroxynitrite production. Aortic rings from rats fed CO or MO produced a similar basal level of O·. Values increased after stimulation with NADPH, and no significant difference was also observed between the two dietary groups (Table 3). SOD reduced O· production by 60% under all conditions. The addition of L-NMMA significantly increased O· production, and this increase was greater in aortic rings from rats fed the MO diet incubated without NADPH. Luminol-mediated chemiluminescence as an indicator of peroxynitrite formation was hardly detectable in the basal condition and after the addition of NADPH.

Phenylephrine-induced vasoconstriction. Phenylephrine induced concentration-dependent contractions in endothelium-intact aortic rings from rats fed CO (Fig. 1A) or MO (Fig. 1B) diets. Incubation with SOD significantly reduced the efficacy of phenylephrine at low concentrations (from 5 × 10⁹ to 2 × 10⁸ M) in the two groups of rats (Fig. 1, A and B). The maximal aortic ring contraction (Table 4) was attained in rats fed the MO diet when incubated with SOD plus L-NNA. The EC₅₀ in the absence of SOD (Table 4) was similar in control aortic rings from rats fed CO and MO diets. Addition of SOD increased the EC₅₀ (log M), by 154% when expressed in nanomolars, in the two dietary groups (P < 0.05), and L-NNA blocked SOD-induced relaxation. A single phenylephrine-induced contraction at a concentration of 2 × 10⁷ M in endothelium-intact rings from the two groups of rats gave the same response (Table 5). Phenylephrine-induced responses in endothelium-denuded rings in the absence of SOD (the EC₅₀ was 7.62 ± 0.05 and 7.54 ± 0.04 for rats fed the CO and MO diets, respectively) were similar to responses in intact rings. The maximal contraction was significantly higher (P < 0.05) in rats fed the MO diet (1,798 ± 124 mg) than in rats fed the CO diet (1,450 ± 100 mg), and, in contrast to intact rings, the addition of 60 U/ml SOD produced no relaxation.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Cumulative phenylephrine-induced contraction in endothelium-intact aortic rings from rats fed corn oil (CO; A) or menhaden oil (MO; B) diets. Rings were incubated with or without superoxide dismutase (SOD; 60 U/ml) in the presence or absence of NG-nitro-L-arginine (L-NNA; 0.1 mM). L-NNA was added 20 min before the first concentration of phenylephrine. Values are means ± SE; n = 8-16 rats. Ctrl, control.

View this table: [in this window] [in a new window]   Table 5.   Phenylephrine (2 × 107 M)-induced contraction in endothelium-intact aortic rings from rats fed CO or MO diets

Acetylcholine-induced vasorelaxation. Relaxation studies took place after precontraction of endothelium-intact aortic rings with 2 × 10⁷ M phenylephrine. Acetylcholine induced cumulative concentration-dependent relaxations in precontracted rings. There was no significant difference in maximal relaxation (46.4 ± 3.9% in rats fed the CO diet vs. 51.9 ± 4.0% in rats fed the MO diet) in the absence of exogenous SOD, but the EC₅₀ (log M) increased in rats fed the MO diet (6.91 ± 0.08) versus the CO diet (6.60 ± 0.08) (P < 0.05), and the addition of L-NNA eliminated this difference (the EC₅₀ was 6.91 ± 0.08 and 7.02 ± 0.1 for rats fed the CO and MO diets, respectively). The acetylcholine-induced relaxation curve in the presence of 60 U/ml SOD shifted leftward in rats fed the MO diet (Fig. 2, A and B), and the maximal relaxation attained was 68.5 and 85.9% for CO- and MO-fed rats, respectively (Table 6). These values represent an increase of 48 and 66%, respectively, when compared with values from rings incubated in the absence of SOD. The blockade of endothelial ·NO synthesis with 0.1 mM L-NNA (Table 6) reduced the maximal relaxation by ~25% in rings from the two groups of rats. This means that the ·NO-dependent relaxation accounted for 40.2 and 62.6% of the maximal relaxation in rings from rats fed CO and MO, respectively. The blockade of endogenous prostanoid-induced relaxation with 10 µM indomethacin gave values of 39.1 and 61.3% in rings from rats fed CO and MO, respectively. Therefore, the prostanoid-dependent relaxation accounted for 29.4 and 24.6% of the maximal relaxation in rings from rats fed CO and MO, respectively. The ·NO-independent relaxation plus prostanoid-independent relaxation equaled the relaxation observed in rings in the control condition. The MO diet increased the EC₅₀ (log M) in aortic rings, a 68% increase (P < 0.05) when expressed in millimolars (Table 6). The inhibitory effect of L-NNA was reversed by 10⁴ M L-arginine (above 85% of the original relaxation).

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Cumulative acetylcholine-induced relaxation in endothelium-intact aortic rings from rats fed CO (A) or MO (B) diets. Rings were incubated with SOD (60 U/ml) in the presence of L-NNA (0.1 mM) or indomethacin (0.1 mM). L-NNA and indomethacin were added 20 min before the precontraction with phenylephrine. Values are means ± SE; n = 8-15 rats.

Sodium nitroprusside-induced vasorelaxation. The studies on sodium nitroprusside-induced relaxation took place in endothelium-denuded aortic rings after precontraction with phenylephrine in the absence or in the presence of 60 U/ml SOD. Sodium nitroprusside induced a concentration-dependent relaxation (Fig. 3), which was potentiated by SOD only at 3 × 10⁹ and sodium nitroprusside at 10⁸ M. The maximal relaxation attained (~100%) and the EC₅₀ (log M) were the same in rings from the two groups of rats.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Cumulative sodium nitroprusside-induced relaxation in endothelium-denuded aortic rings from rats fed CO or MO diets. Rings were incubated with or without SOD (60 U/ml). Values are means ± SE; n = 7 rats.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Early observations have indicated that dietary manipulation with -3 polyunsaturated fatty acids increases endothelium-dependent relaxation in isolated porcine coronary arteries (34) and cerebral arteries (16). The mechanisms underlying the regulation of vascular tone are complex and involve free radicals and products of the cyclooxygenase pathway (15, 26, 27, 38). However, we show here, in aortic rings from rats fed a MO-rich diet and in conditions that prevent interactions between O· and ·NO, the following: 1) there is no impairment in the release of O· by the vascular wall; 2) the endothelium-dependent ·NO relaxation induced by acetylcholine is increased, and there is no facilitated response of the smooth muscle to ·NO; and 3) despite a reduction in arachidonic acid and an increase in EPA in phospholipid membranes, the resultant cyclooxygenase-dependent vascular response to acetylcholine is not modified.

There is increasing evidence of O· production in the vascular wall by NAD(P)H oxidase (22, 25, 41), and the concentration of lucigenin is a critical parameter affecting the validity of the O· measurement because it generates O· at 250 µM but not at 5 µM (36). Our results of both basal and NADPH-stimulated production of O· in aortic rings showed that there were no differences between the two dietary groups. The O· production by aortic vessels is consistent with the observations of Pagano et al. (25). They described an increase in O· release by extracellular NADPH related to a plasma membrane-associated NADPH oxidase. The ·NO depresses the baseline chemiluminescence signal, and the evaluation of this signal in the presence of a NO synthase inhibitor is an indirect probe to estimate ·NO release (25, 36). In this condition, basal ·NO was higher in aortic rings from rats fed the fish oil diet, and, in addition, we observed similar results by electronic paramagnetic resonance (unpublished data).

The release of ·NO by endothelium-intact aortic rings exerts a tonic vasodilator action opposing the effects of vasoconstrictor agents such as O· (15). The favorable kinetics of the reaction between O· and ·NO intrinsically make vascular O· levels an important determinant of ·NO biological activity (23). Thus the reduction of ·NO-induced vasodilation associated with peroxynitrite generation may contribute to pathological situations such as hypertension. Our results with exogenous SOD are indicative of O· production, which would reduce ·NO bioavailability. The absence of SOD effect on endothelium-denuded rings indicates that the vasoconstrictor effect of O· is mainly due to chemical inactivation of ·NO. Therefore, SOD was used in further experiments to reach optimal ·NO-mediated responses. L-NNA potentiation of phenylephrine-induced contraction in rings from rats fed MO (38%) implied a greater ·NO production in animals fed MO than those fed CO.

Relaxation of blood vessels is subject to complex control. There is a basal production of ·NO that exerts a tonic vasodilator action that plays an important role in the local control of vascular tone. This basal production is stimulated by a large number of biological mediators such as acetylcholine (11). The release of ·NO under basal and agonist-stimulated conditions was measured in the presence of SOD in the organ bath, although some authors (1, 21) have observed no SOD effect on the acetylcholine-mediated relaxation in arterial rings. We ruled out that differences in relaxations in precontracted rings were due to a facilitated relaxation of vascular smooth muscle because the concentration-response curves to sodium nitroprusside were similar in rings from the CO and MO groups. Moreover, the endothelium-dependent relaxation to acetylcholine involves ·NO and cyclooxygenase products (4, 39). The incorporation of -3 polyunsaturated fatty acids into phospholipids of cell membranes may affect endothelial ·NO synthase, and -3 polyunsaturated fatty acids decrease the synthesis of series 2 and 4 eicosanoids and increase the synthesis of series 3 and 5. However, Fischer and Weber (9) observed similar effects with prostaglandin I₃ and prostaglandin I₂ (prostacyclin). The ·NO-mediated relaxation increased 58%, whereas prostanoid-mediated relaxation was reduced by 16%, in rats fed the MO diet. Thus we have shown that, in basal and in agonist-stimulated ·NO production, the MO diet elicited a vasorelaxant activity, which was mediated by ·NO.

The imbalance between O· and ·NO production with enhanced activation of ·NO by a fish oil diet may, in addition to the reported beneficial effects in blood vessels, also have lead to vasorelaxant activity. The oxidation of low-density lipoproteins (LDL) is the first step in the development of atherosclerotic lesions. The acceleration of LDL oxidation may be brought about by peroxynitrite production (7), by incorporation of -3 polyunsaturated fatty acids to LDL (20, 24, 42), or by reduction of lipid-soluble antioxidants such as -tocopherol (8) among others. However, in vivo studies (6, 12, 42) have demonstrated that fish oil prevents the development of coronary diseases. It has been observed that ·NO is a potent antioxidant (31) that prevents the oxidation of LDL in vitro (14, 17) by inhibiting radical chain propagation reactions (30) and thus acts as an antiatherogenic agent (18). More work is needed to substantiate the protective role of fish oil on LDL oxidation.

New studies on the effects of dietary fatty acids in the prevention and treatment of cardiovascular disease are emerging. We can conclude from the present study that a long-term MO-rich diet has a beneficial effect on blood vessels by potentiating the production of endothelial ·NO, a potent vasodilator with antiatherogenic properties, without affecting either the production of O· or the relaxation due to the activation of the cyclooxygenase pathway.