INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ATHEROSCLEROTIC PLAQUES prone to rupture were found to be rich in lipids, and fibrous plaques were found to be difficult to rupture (1, 20). Furthermore, ~70-80% of myocardial infarctions was found to occur in nonsignificant stenotic vessels (7, 33). Therefore, decreasing the amount of lipid and inflammatory components and increasing the fibrous components in a vessel seems to be more important than morphological regression of the atherosclerotic area to a normal range wherein cell components do not change. These changes are consistent with the concept of lesion stabilization, which is thought to be responsible for the reduction in acute coronary events associated with cholesterol reduction in humans (6, 12). Recently, Hayashi and colleagues (18) studied the effects of removing cholesterol from the diet of rabbits with atherosclerotic arteries. The results were compared with normal control rabbits, rabbits fed a high-cholesterol diet (HCD) for 9 wk (atherosclerotic group), and rabbits fed a normal diet for 9-36 wk after 9 wk of an HCD (regression group). Thoracic aortas revealed more atheromatous lesions in the regression groups than in the atherosclerotic animals. The vascular responses in the 9-wk, HCD group did not return to normal after 36 wk of normal diet. There also have been many other studies about the regression and stabilization of atherosclerosis after removal of cholesterol from the diet; however, the morphological regression and restoration of vascular responses needs a much longer normolipidemic term, one of many years (14, 18, 30). There have been a few studies of regression or stabilization using lipid-lowering drugs in rabbits with advanced atherosclerosis (5). Simvastatin, an 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA) reductase inhibitor, has been reported to increase endothelial nitric oxide synthase (eNOS) activity in vitro (24). From these findings, the present study focuses on the effect of HMG-CoA reductase inhibitor on the regression and stabilization of atherosclerosis following the removal of dietary cholesterol in rabbits. The present study also focuses on the role of nitric oxide (NO) in both the regression and stabilization of atherosclerosis and in the effects of simvastatin. To analyze factors that regulate NO, eNOS mRNA expression and O release were measured in aortas with competitive reverse transcription-polymerase chain reaction (RT-PCR) and with lucigenin analog 2-methyl-3,7-dihydroimidazol[1,2-a]pyrazine-3-one (MCLA) chemiluminescence methods (25).

ONOO, which is produced by the reaction NO and O, has been found to be more cytotoxic to cells than NO itself, and O has been shown to react with NO faster than with superoxide dismutase by forming peroxynitrite (3, 22). The reaction rate is 6.7 ± 0.9 × 10⁹ M¹ · S¹, which is approximately six times faster than the scavenging of superoxide by copper/zinc superoxide dismutase at physiological ionic strength (8, 27). Esaki and colleagues (10) reported the presence of inducible NOS (iNOS) in advanced atherosclerotic plaques, but not in normal vessels or the early stage of atherosclerosis. Furthermore, a large amount of NO is believed to be released from iNOS compared with that from eNOS in vitro (35). O has been shown to increase in atherosclerotic vessels; however, it is unclear whether it increases or decreases in regression or stabilization states (15). ONOO may be produced in atherosclerotic and regressive vessels. In the present study, we determined the effect of an HMG-CoA reductase inhibitor on the regression and stabilization of atherosclerosis and the role of NO and oxidative stress.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Chemicals and solutions. Acetylcholine chloride (ACh), phenylephrine, calcium ionophore A-23187, hemoglobin, indomethacin, and N^G-monomethyl-L-arginine (L-NMMA) were all purchased from Sigma (St. Louis, MO). Simvastatin was obtained from Merck-Banyu Japan (Tokyo, Japan). Monoclonal anti-eNOS, anti-iNOS, anti-rabbit T cells (P8022), anti-rabbit macrophages (RAM11), anti-smooth muscle cells (HHF35), and anti-nitrotyrosine antibodies were purchased from Transduction (Lexington, KY). Nitroglycerin (NTG) was obtained from Nihon Kayaku (Tokyo, Japan). Krebs-Henseleit solution (in mM: 118 NaCl, 4.7 KCl, 1.5 CaCl₂, 1.2 MgSO₄, 1.2 KH₂PO₄, 25 NaHCO₃, 11 glucose, and 0.002 disodium ethylenediaminetetraacetic acid; pH 7.4) was saturated with 95% O₂-5% CO₂. The depolarizing KCl solution was similar in composition to the Krebs-Henseleit buffer, except that NaCl was replaced by an equimolar amount of KCl. All concentrations indicated for the in vitro studies were final bath values.

Animals. Male New Zealand White rabbits (n = 30; 2.5-3 kg body wt) were fed a regular rabbit chow (Oriental Yeast, Tokyo, Japan) for 2 wk, and then they were randomly divided into three groups of 10 animals. Treatments were as follows: group 1, 0.5% cholesterol plus regular rabbit chow (atherogenic diet) for 8 wk; group 2, atherogenic diet for 8 wk and then regular chow for 12 wk; and group 3, atherogenic diet for 8 wk and then regular chow with simvastatin (5 mg · kg¹ · day¹) for 12 wk. A separate group of animals was fed a regular chow for 20 wk as a long-term control (LC, n = 6). Furthermore, the effect of simvastatin was investigated in the rabbit with a regular chow with or without simvastatin (5 mg · kg¹ · day¹) for 4 wk (NC, NC_simva, each n = 6). The rabbits were housed individually at 20 ± 3°C with a 12-h:12-h light/dark cycle and with free access to water. Feeding was restricted to 120 g/day. Blood samples were collected 24 h after feeding. The general appearance of the rabbits was observed daily. Body weights were determined every 4 wk. All experiments were conducted in accordance with institutional guidelines for animal studies. All experiments were performed on the same set of rabbits.

Assay for lipids. Total cholesterol and triglyceride levels were measured by enzymatic techniques as described previously (2). High-density lipoprotein (HDL) cholesterol was measured after precipitation with phosphotungstate-MgCl₂ (25).

Histological evaluation of atherosclerosis and assays for tissue cholesterol content. Cross sections of the descending thoracic aorta were stained with hematoxylin-eosin to examine the endothelial lining and with van Gieson's elastic stain to determine the thickness of the intima. Morphometric analysis was performed as described by Weiner and colleagues (34). Briefly, six samples from each rabbit aorta were analyzed with the objective lens. To determine the surface involvement of atherosclerotic lesion (fatty streaks and fibrous plaques) and area occupied by the atherosclerotic lesion defined as below, the first complete section of each block was projected onto a vertical surface with a projecting microscope. The contours of the lumen and the internal elastic lamina were traced, and the tracings were digitized (PC-9801 ES, NEC; Tokyo, Japan) with a graphics tablet. The mean surface involvement by atherosclerotic lesion per vessel per animal (n = 6 for one vessel) was calculated, and then mean value of the groups was calculated (n = 10 for each group). Circumferences of lesion and normal part were defined as circumferences of internal elastic lamina where intimal thickening was observed and that where normal intima was observed. The area occupied by atherosclerotic lesions was defined as the percent area bounded by the lumen and the internal elastic lamina for the ideal luminal area. The mean area occupied by the lesions per vessel per animal (n = 6 for one vessel) was calculated, and then the mean value of the groups was calculated (n = 10 for each group). To assay for free and esterified cholesterol content, the segment of the aortic arch (2 cm distal to the aortic valve) down to the bifurcation of the left subclavian arteries was weighed, minced, and homogenized in 10 volumes of sucrose-Tris buffer. The homogenates were assayed for free and esterified cholesterol (21).

Measurement of nitrite and nitrate. Nitrite and nitrate (NO/NO) in plasma were measured with an automated NO detector-HPLC system (ENO10: Eicom; Kyoto, Japan) as previously reported (17). In brief, samples were collected in an automated sample injector connected to an automated NO detector, and then NO and NO were separated from the samples by a reverse-phase separation column (4.6 × 50 mm, NO-PAK, Eicom). NO was then reduced to NO in a reduction column (NO-RED, Eicom). The absorbence of the product dye formed by NO and a Griess reagent at 540 nm was measured by a spectrophotometer (NOD-10). The mobile phase, which was delivered by a pump at a rate of 0.33 ml/min, was 10% methanol containing 0.15 M NaCl/NH₄Cl and 0.5 g/l 2Na-EDTA. The Griess reagent (0.5% sulfonamide, 0.025% N-naphthylethyl-ethylenediamine dihydrochloride, and 1.25% HCl) was employed.

Preparation for isometric tension measurement. Rabbit aortic rings were prepared as described by Furchgott and Zawadzki (11). Briefly, after being anesthetized with pentobarbital (50 mg/kg iv), the rabbits were exanguinated; then their thoracic aortas were carefully removed to protect the endothelial lining, cleared of adhering fat and connective tissue, and cut into transverse rings 3-mm wide. The rings were bathed in Krebs-Henseleit solution. They were then stretched to their previously determined optimal tension, which is the contractile response to 122 mM KCl, pH 7.4 at 37°C for 1 h.

7

6

Measurement of eNOS mRNA. We quantitated eNOS mRNA of the arterial wall as copies using competitive RT-PCR methods. Briefly, to make a DNA competitor, we first designed and synthesized two primers [5'-ATTTAG-GTGACACTATAGAATACCAGTGTCCAACATGCTGCTGGAAATTGGTACGGTC-ATCATCTGACAC-3' (sense primer), 5'TAAAGGTCTTCTTCCTGGTGATGCCAAT-ACATCAAAC-GCCGCGAC-3' (anti-sense primer)] based on the sequences of human eNOS cDNA and DNA and using a competitive DNA construction kit (Takara Shuzo; Otsu, Japan). We then synthesized an RNA competitor using the DNA competitor and a competitive RNA transcription kit (Takara Shuzo). Rabbit aorta total RNA was extracted using TRIzol reagent (GIBCO) following the manufacture's protocol and was quantified by a spectrophotometer. Competitive RT-PCR was performed using the RNA competitor and an RNA PCR Kit Ver2.1 (Takara Shuzo). Part of the PCR reaction mixture was electrophoresed through a 3.5% agarose gel. eNOS cDNA primers amplify a product with a predicted length of 486 bp, and the competitor was produced at a length of 558 bp. The same amount of mRNA was corrected using a ß-actin competitive PCR Kit (Takara Shuzo).

Immunocytochemical study. Tissue sections were deparaffinized with xylene and rehydrated with graded alcohol. The specimens were preincubated for 30 min with methanol containing 0.3% hydrogen peroxidase, washed with phosphate-buffered saline (PBS), permeabilized with 0.1% Triton X100 in PBS for 20 min, and then washed with PBS. They were incubated for 60 min with primary monoclonal antibody (for anti-eNOS, anti-macrophage iNOS, P8022, RAM11, HHF35, or anti-nitrotyrosine) diluted in PBS with horse serum and then washed again with PBS. A biotinated rabbit anti-mouse IgG (1:500 dilution) was applied for 30 min, followed by an avidin-biotin peroxidase complex (ABC Kit, Vector Laboratories; Burlingame, CA). The result was a brown peroxidase reaction product. Negative controls included substitution of primary antiserum/antibody with either PBS or irrelevant antibodies. As a control experiment, treatment with sodium dithionate (100 µM in 100 mM Na₂CO₃ buffer, pH 9 for 5 min) before the antibody incubation abolished the staining for nitrotyrosine. Each field was scored for the number of target antibody such as nitrotyrosine-positive cells on a slide and analyzed statistically as described previously (13, 17). Five samples were prepared from each rabbit.

Detection of aortic superoxide generation. Formation of O was assayed by measuring the intensity of chemiluminescence (CL) probes in the presence of MCLA. O generation in an aorta was measured in PBS at physiological pH 7.4. The O generation signal of CL was detected by a luminescence reader (BLR-201, Aloka; Tokyo, Japan). To ensure the specificity of MCLA to detect O in vessels, increasing concentrations of superoxide distmutase (1-50 U/ml) were added to the vascular tissues (29). The O generation signal of CL was defined as the inhibitory signal by superoxide dismutase (100 U/ml).

Data analysis. Relaxation was defined as the percent decrease in tension below that elicited in arterial rings precontracted with phenylephrine. Contraction was measured as the percent increase in tension above that elicited in arterial rings precontracted with phenylephrine. Data were expressed as means ± SE. Statistical significance was assessed by the Students' t-test for paired values. When more than two means were compared, an analysis of variance with repeated measurements was used. A P value <0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Blood chemistry. No rabbits appeared to suffer from any generalized disorders. No significant differences in body weight, serum triglyceride, HDL cholesterol, or serum total protein were observed among the three groups during the study (Table 1). No significant change in total cholesterol or triglyceride was observed in the control animals (LC, data not shown). Adding 0.5% cholesterol to the diet (groups 1-3) increased total cholesterol in the rabbits (Table 1). Removal of cholesterol from the diet reversed the lipid level to within the normal range (groups 2 and 3), and the simvastatin-treated groups showed a tendency of lowering concentration of plasma total cholesterol without significant differences.

Histological evaluation of atherosclerosis and assays for tissue cholesterol content. The atherosclerotic area of the thoracic aorta was indicated by the mean lesion area (% occupied lesion). There were more atheromatous lesions observed in the regression group (group 2, 18.2±5.2%) than that in the hypercholesterolemic group (6.9 ± 1.2% in group 1) (Fig. 1). However, lesions in group 3 were reversed to levels compatible with those of the group 1 (11.1±3.1% in group 3). The surface involvement showed the same tendency (39.5 ± 4.6% in group 1, 57.1 ± 9.2% in group 2, and 44.1 ± 5.8% in group 3). The amount of total and esterified cholesterol in the vessels of group 3 significantly decreased than found that in group 2 (total cholesterol: group 1, 32.8 ± 3.4 mg/wet g; group 2, 46.6 ± 5.7 mg/wet g; group 3, 37.1 ± 3.5 mg/wet g; *P < 0.05 vs. group 1 or group 3).

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Histological evaluation of atherosclerotic area of the thoracic aorta as indicated by the mean percentage of the lesion area (% occupied lesion). Atherosclerotic areas from 3 groups of rabbit aorta (group 1: fed with atherogenic diet for 8 wk; group 2: atherogenic diet for 8 wk then regular chow for 12 wk; and group 3: atherogenic diet for 8 wk then regular chow with 5 mg/kg simvastatin for 12 wk) were measured. Each value is mean ± SE (n = 7). *P < 0.05.

Measurement of plasma NO/NO. Removal of cholesterol tended to decrease the plasma level of NOx (sum of NO and NO), especially in group 3; however, it did not attain statistical significance (data not shown).

Measurement of NO-related vascular responses. Tone-related basal NO-dependent response by L-NMMA, which abolishes the activity of NOS by inhibiting the coupling of L-arginine with NOS, was concentration dependent. The magnitude of vasocontraction decreased in the aortas of group 2 compared with those of group 1 (Fig. 2). However, vasocontraction increased significantly in the aortas of group 3 compared with those of group 1 and 2. In regular chow groups, there is no difference between LC and NC; however, the value was higher in NC_simva than that in LC or NC (% of contraction by 100 µM L-NMMA was 31.6 ± 6.5, 38.9 ± 7.4, and 62.1 ± 8.4% in LC, NC, and NC_simva, *P < 0.05 vs. NC). The relaxation in samples from group 1 in response to ACh was clearly diminished compared with that in samples from control rabbits on a regular diet (LC group). The magnitude of this relaxation was further diminished in group 2 and was slightly recovered in group 3 (Fig. 3A). In regular chow groups, there were no significant differences among LC, NC, and NC_simva (max% of relaxation was 93.6 ± 2.5, 94.9 ± 3.4, and 98.9 ± 2.4% in LC, NC, and NC_simva). Calcium ionophore A23187-induced NO-stimulated response showed the same tendency; however, statistical significance was not achieved because the relaxation was not severely impaired compared with that induced by ACh (Fig. 3B). NTG-induced, concentration-dependent relaxation did not differ among the three groups (Fig. 4). In regular chow groups, there were no differences among LC, NC, and NC_simva (Max% of relaxation was 98.6 ± 3.1, 98.9 ± 2.4, and 99.1 ± 1.4% in LC, NC, and NC_simva). Indomethacin did not affect these responses (data not shown).

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Cumulative concentration-response curves to NG-monomethyl-L-arginine (L-NMMA) during contraction evoked by prostaglandin F2 (0.8 × 106 M) in the thoracic aortas of groups 1, 2, and 3 rabbits. Data are shown as means ± SE. *Significant difference (*P < 0.05, **P < 0.01) vs. group 1. +Significant difference (+P < 0.05) vs. group 2.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   A: cumulative concentration-response curves to acetylcholine during contraction evoked by phenylephrine in the thoracic aortas of rabbits fed with 0.5% cholesterol (high-cholesterol diet, HCD) for 8 wk (group 1); HCD for 8 wk then regular chow for 12 wk (group 2); HCD for 8 wk then regular chow for 12 wk with simvastatin (5 mg/kg; (group 3), Group LC, regular chow for 20 wk. B: cumulative concentration-response curves to Ca ionophore A23187 during contraction evoked by phenylephrine in the thoracic aortas of rabbits fed with HCD for 8 wk (group 1); HCD for 8 wk then regular chow for 12 wk (group 2); HCD for 8 wk then regular chow for 12 wk with simvastatin (5 mg/kg; group 3); Group LC, regular chow for 20 wk.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Cumulative concentration-response curves to nitroglycerin during contraction evoked by phenylephrine in the thoracic aortas of rabbits fed with the same diets as indicated in Fig. 3. Each value is mean ± SE (n = 7).

Detection of mRNA for eNOS in the rabbit aorta. After electrophoresis on agarose, the ethidium bromide-stained bands were quantified by densitometry from a photograph of the gel (Fig. 5). The signal for eNOS increased about 80% in reverse-transcribed RNA samples from aortas of hypercholesterolemic rabbits (group 1) compared with those from control rabbits (LC)(Fig. 5). In contrast, the yield of PCR products of the predicted size for eNOS did not change in aortas from the regression group on a regular diet (group 2) and increased in aortas from the regression group treated with simvastatin (group 3) compared with that of hypercholesterolemic rabbits (Fig. 5). The yield of PCR products also increased in aortas from the NC_simva group, compared with that of NC groups (NC_simva 1.02 ± 0.06, NC 1.11 ± 0.10 × 10⁵ copies, *P < 0.05 vs. NC). These findings were compatible with the data from tone-related basal NO release investigated in vascular responses. These findings indicate that eNOS is expressed in increased levels in the aorta after simvastatin administration in vivo.

View larger version (46K): [in this window] [in a new window]   Fig. 5.   Determination of mRNA for endothelial NO synthase (eNOS) in abdominal aortas using competitive RT-PCR methods. Samples containing a fixed amount of target cDNA and increasing amounts of competitor (-actin). Top: photograph of ethidium bromide-stained gel after electrophoretic resolution of the internal standard and eNOS target bands (C, control without competitor). Signal for eNOS increased in aortas of hypercholesterolemic rabbits (group 1) compared with those of control rabbits (LC), because eNOS signal in group 1, not in group LC, was seen in 105 copies. Yield of PCR products decreased in aortas of the regression group on a regular diet (group 2); that in aortas from the regression group that was administered simvastatin (group 3) did not decrease. Bottom: effect of HCD (group 1), removal cholesterol from diet (group 2), and simvastatin treatment (group 3) on eNOS mRNA. mRNA for both eNOS and -actin was quantitated by RT-PCR. Relative amounts of eNOS mRNA to competitors. Values are means ± SE. *P < 0.05.

Immunocytochemical study. iNOS was apparent in T cells and some macrophages of the advanced lesions in the atherosclerosis and the regression groups (groups 1-3). Peroxynitrite shown by staining with nitrotyrosine was distributed over larger areas than those occupied by iNOS-positive cells. These cells were observed not only in necrotic cores of fibrous plaques but also in subintimal areas of fibrous plaques.

View larger version (78K): [in this window] [in a new window]   Fig. 6.   Representative immunohistochemical analysis with anti-nitrotyrosine monoclonal antibody of thoracic aortas of New Zealand White rabbits. A: nitrotyrosine, the marker of ONOO, was distributed over larger areas in vessel from group 1. B: nitrotyrosine was also distributed over larger areas in vessel from group 2. C: nitrotyrosine was not significantly stained in vessel from group 3. Original magnification ×100.

Aortic superoxide anion production. We measured superoxide anion production in arterial walls with a lucigenin analog (MCLA). The CL signals as O production, which was determined by 100 U/ml superoxide dismutase inhibitable signals, were greater in aorta from the cholesterol-fed group compared with aorta from the regular-diet fed group (Fig. 7). CL signals from vascular tissue with endothelium showed a decrease in the group 3 compared with the group 2. In aortas without endothelium, CL signals also decreased. The endothelium-dependent chemiluminescence was defined by the mean of CL signals with endothelium reduced from that without endothelium. It decreased in group 2 than that in group 3. In other words, the relative difference between the group 2 and group 3 was higher in the endothelial part than in the nonendothelial part of the vessels. CL signals from vascular tissue with endothelium showed a decrease in the NC_simva group compared with the NC group (NC_simva, 0.201 ± 0.036, NC 0.321 ± 0.110; P < 0.05). Similarly, CL signals from endothelium showed a decrease in the NC_simva group compared with the NC group (NC_simva 0.003 ± 0.004, NC, 0.071 ± 0.061; **P < 0.01 vs. NC).

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Effects of simvastatin on superoxide production from rabbit aortas. With EC: data of vessel with endothelium. Without EC: data of vessel without endothelium. EC: change in the intensity of MCLA chemiluminescence with or without endothelium. Data are shown as means ± SE. *Significant difference (P < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This study revealed that treatment with simvastatin, a HMG-CoA reductase inhibitor, with a regular diet during a period of regression significantly reduced total and esterified cholesterol concentrations in the aorta and significantly diminished the atherosclerotic area compared with the animals given a diet only of regular chow (group 2). They showed no regression but rather progression of atherosclerosis. This is consistent with our previous finding and other reports (18, 28). In humans, diet control without exercise or other drugs cannot stop the progression of coronary atherosclerosis (23). Therefore, we speculate that this phenomena in rabbits can be applied to human coronary atherosclerosis. Although other reports showed a regular chow could decrease O and restored ACh response, the experimental condition is different from ours and the term of HCD in theirs is only 4 wk (26). Their study was aimed to observe the effect of hyperlipidemia, not of atherosclerosis. In this study, plasma total cholesterol level was slightly lower in group 3 compared with those in group 2 during the regression term. In humans, the restoration of impaired NO-dependent relaxation was reported in atherosclerotic coronary arteries as a result of lowering plasma lipid levels. In the present study, the effect of simvostatin on lowering lipid levels may have contributed to stabilization of atheroma. However, the lack of totally significant difference in plasma lipid levels between group 2 and group 3 during the regression period may suggest that simvastatin had other effects and may have had a direct effect on vessels.

Functional changes due to atherosclerosis in properties such as vascular reactivity and morphological regression due to lipid-lowering therapy may occur at different rates and to different degrees in different parts of the vascular bed (4). No morphological regression of atherosclerosis has been observed despite removal of lipids from the diet (group 2) or addition of HMG-CoA reductase inhibitor (group 3). Nevertheless, lesions in group 3 were reversed to levels compatible with those of the group 1, and there were more atheromatous lesions observed in the group 2 than those found in the hypercholesterolemic group (group 1). This means a very strong antiatherosclerotic and probably regressive effect of simvastatin. Atherogenic lipids such as oxidized low-density lipoprotein (LDL) or very LDL (-VLDL) are known to inhibit endothelium-dependent relaxation (EDR) and are known to exist in subintimal areas in the early stage of atherosclerosis (20, 36). The lipid concentrations in the vessel wall in the regular diet group (group 2) in the regression term increased compared with the values in the atherosclerosis group (group 1), and the grade of impairment of EDR did not greatly differ. However, not only restoration of EDR, but also prevention of progression and stabilization of atherosclerosis, were observed in the simvastatin treatment group. NO-related functions, such as tone-related basal NO and eNOS mRNA, increased after treatment with simvastatin. Although plasma NOx did not change statistically, it may be inevitable because many factors, such as renal function and urea cycle activity, may affect the level. O decreased with simvastatin treatment, especially in the endothelium. Therefore, we speculated that this effect of simvastatin on regression was due to an improvement in endothelial functions such as eNOS activity and O release.

The proliferation of vascular smooth muscle cells, monocyte adherence, and infiltration are key processes involved in atherogenesis. Because NO inhibits each of these processes, NO is believed to be an endogenous anti-atherogenic molecule. In the present study, eNOS mRNA and O release increased in atherosclerotic vessels compared with those in control rabbit aortas (LC). This implies that hypercholesterolemia induces an abnormality in the enzyme NOS. Increase of eNOS mRNA in atherosclerosis was thought to have an intrinsic protective effect on the vessel wall against atherosclerotic change, and that it might be feedback from impairment of the effect of NO by O (13). eNOS mRNA decreased significantly after the animals were placed on a regular diet for 12 wk and release of O did not decrease. Simvastatin treatment in the regression term significantly increased eNOS mRNA and significantly decreased the release of O in vessels, especially in the endothelium. Because a reduction in NO activity promotes atherogenesis, a restoration of NO activity would be expected to contribute to the regression of the disease. An HMG-CoA reductase inhibitor (simvastatin, 1 mmol/l and lovastatin, 10 mmol/l) was recently reported to upregulate eNOS in vitro (24), and we observed that an HMG-CoA reductase inhibitor in chow upregulated eNOS mRNA in normal rabbit aortas. Namely, the results of this study are compatible with those in vitro. On the other hand, a HMG-CoA reductase inhibitor lovastatin was shown to increase O; however, the experiment was done in a HCD feeding for rabbits fed a regular diet, which was a completely different experimental condition of ours (5). In other words, we investigated whether HMG-CoA reductase can restore the impaired endothelial function in previously existed atherosclerosis, and they investigated the effect of lovastatin in the preventing the progression of atherosclerosis.

The decrease in the release of O in rabbits fed a regular chow brought about by treatment with a HMG-CoA reductase inhibitor substantiated the first observation in vivo as much as our knowledge. Although reduction in serum cholesterol is known to reduce superoxide levels, this study certified that simvastatin itself scavenges O, and it may indirectly reduce NO degradation (26). This mechanism would explain our observation as described above. The effect of simvastatin on the synthesis of enzymes in endothelial cells, such as NADPH oxidase or xanthine/xanthine oxidase may be relevant, and it is possible that simvastatin has a direct effect through the impairment of Rho kinase and NADPH oxidase. Furthermore, the decrease of O release in group 3 may reflect an increase of eNOS mRNA or the result of regressed atherosclerosis. More elucidation may be necessary concerning the mechanism of the decrease in O release in the simvastatin treatment groups.

Hayashi and colleagues previously reported the role of peroxynitrite in blocking the restoration of EDR (18). We (18) found that certain physiological concentrations of peroxynitrite impair EDR. Furthermore, nitrotyrosine, a marker of ONOO, existed in the advanced stage of atherosclerotic areas but not in the early stage. In the present study, nitrotyrosine was stained in the atherosclerotic and regression groups (groups 1 and 2) and restoration of EDR was not seen in these groups. Our results are consistent with the report of Beckman et al. (3) that NO is scavenged by O, which is known to increase in atheroma, and thus ONOO is produced. ONOO spontaneously reacts with tyrosine residues to yield the stable product, 3-nitrotyrosine, which could be a footprint left by peroxynitrite in vivo (3). Nitrotyrosine was known to be produced not only by the reaction of tyrosine residues and ONOO but also by the reaction of hypochlorous acid (HOCl) and NO (32). HOCl can be produced by myeloperoxidase, which has been observed in human atherosclerotic lesions (9). This reaction was competitive with the reaction of ONOO (31). However, there have been no reports about HOCl in atherosclerosis, and myeloperoxidase is less abundant in monocytes and is generally absent from macrophages that were stained for nitrotyrosine and iNOS in our study. The staining areas of myeloperoxidase, was reported to be transitional lesions near the shoulder lesion and an area adjacent to cholesterol clefts, were different from staining areas of nitrotyrosine in this study (9).

We hypothesized that the restoration of EDR relates to ONOO production throughout the atherosclerotic area and the production of O, which increased in endothelial cells in atherosclerosis (16). Our data suggested that removal of dietary cholesterol alone could not reduce the amount of O released and that simvastatin treatment could reduce it. The difference in nitrotyrosine-positive areas among the three groups of vessels was consistent with the data for O.

Basic insights regarding the mechanisms of atherogenesis and the regression should lead to new therapeutic strategies to induce the regression of atherosclerosis. Despite the clinical efficacy of cholesterol reduction in reducing cardiovascular events, novel therapeutic approaches to atherosclerosis are still needed. Upregulation of eNOS is an attractive target for therapeutic intervention and could be important for the regression and stabilization of plaque. In conclusion, the HMG-CoA reductase inhibitor simvastatin caused a case of cholesterol diet-induced atherosclerosis to stabilize, and a NO-mediated system may have played a role in its effect.