Atherosclerosis, 87 (1991) 23-38 Q 1991 Elsevier Scientific Publishers ADONIS 002191509100080B

ATHERO

23 Ireland,

Ltd. 0021-9150/91/$03.50

04600

Low level hyperlipidemia impairs endothelium-dependent relaxation of porcine coronary arteries by two mechanisms Functional change in endothelium and impairment of endothelium-dependent relaxation by two mediators Toshio Hayashi ‘, Tomohiko Ishikawa 2, Michitaka Naito *, Masafumi Kuzuya I, Chiaki Funaki ‘, Kanichi Asai I, Hiroyoshi Hidaka 2 and Fumio Kuzuya 1 Department of I Geriatrics and ’ Pharmacology, Nagoya University School of Medirine, Nagoya (Japan) (Received 14 June, 1990) (Revised, received 26 October, 1990) (Accepted 5 November, 1990)

Summary

We evaluated the effect of a low level of hyperlipidemia and the effects of in vitro exposure to atherogenic lipoproteins (LDL, VLDL) on the vascular responsiveness of isolated porcine coronary arteries. Firstly we studied the change in vascular responsiveness induced by feeding a cholesterol-rich diet to pigs for 4 and 9 weeks (C4 and C9 pigs). The serum cholesterol level in pigs fed a cholesterol-rich diet reached 218.5 _+32.9 mg/dl compared with 85.5 f 8.4 mg/dl in the controls. Segments of the left descending coronary artery were examined. The contraction induced by KC1 or prostagfandin Fzu was not altered significantly by hypercholesterolemia nor was the relaxation induced by the Ca2+ ionophore, A23187, or by nitroglycerin. Endothelium-dependent relaxation (EDR) evoked by high, but not low, concentrations of bradykinin was reduced in the C4 pigs as compared with those in normal animals. EDRs evoked by bradykinin, substance P, and serotonin were significantly reduced in C9 pigs. Histologically, as observed by light and electron microscopy, fatty changes or intimal thickenings were not seen in the coronary arteries of the C4 pigs. Minimal changes (intimal thickening and fragmentation of internal elastic lamina) were observed only in parts of arteries of the C9 pigs. Secondly, the direct effects of LDL and VLDL on vascular responsiveness were studied. Although preincubation with LDL inhibited the EDR caused by exposure to bradykinin and A23187 in the coronary arteries of normal and cholesterol-fed pigs, preincubation with LDL inhibited the arterial relaxation induced by exposure to substance P or serotonin in both the C4 and the C9 pigs, but not in the control animals. The degree of inhibition was especially

Correspondence: Toshio Hayashi, Department of Geriatrics, Nagoya University School of Medicine, 65 Tsuruma-cho,

Showa-ku, 741-9326

Nagoya

466, Japan.

Tel.: 052-741-2111;

Fax: 052-

24 marked in the C9 pigs. The inhibitory effect of VLDL on EDR was weaker than that of LDL. Indomethacin (5 PM) did not alter this inhibitory effect of lipoproteins. Neither LDL nor VLDL had any effect on the vascular relaxation induced by nitroglycerin. These results are consistent with the idea that endothelium-dependent arterial relaxation is attenuated even at the very early stage of cholesterol-induced atherosclerosis. Atherogenic lipoproteins may further impair the decreased EDR in the arteries of hyperlipidemic pigs by two factors: one released on stimulation with bradykinin and the calcium ionophore A23187, the other released on stimulation with substance P and serotonin.

Key words: Endothelium-dependent rived relaxing factor

relaxation; Hyperlipidemia;

Introduction

Hyperlipidemia, a major risk factor for atherosclerosis [1,2], can lead to functional and morphologic changes in the endothelial cells [3,4]. Endothelial cellular dysfunction may play a major role in the etiology of atherosclerosis [5]. Since the endothelial cells produce prostaglandin I, (prostacyclin) as well as endothelium-derived relaxing factor(s) (EDRF) [6,7], the endothelium may contribute in several ways to the local regulation of vascular function. The impairment of endothelium-dependent relaxation (EDR) in atherosclerotic arteries has been reported both in experimental animals and humans [8,9]. In porcine coronary arteries exposed to severe hypercholesterolemia, there is an impairment of the relaxation induced by serotonin and substance P [lO,ll]. However, one report found that endotheliumdependent vasodilators produce equivalent degrees of relaxation in arteries removed from nordogs fed mal and hypercholesterolemic cholesterol-rich diets [12]. Information regarding the impairment of EDR in hypercholesterolemia is thus inconsistent. It was reported that a large quantity of human LDL or porcine LDL inhibits EDR in the normal rabbit aorta or normal porcine coronary arteries [13,14]. It is not known whether a physiological concentration of LDL may alter the EDR of coronary arteries that were either atherosclerotic or exposed to hypercholesterolemia. Porcine coronary arteries have been used in studying atherosclerosis of the coronary artery [10,11,14,15]. These studies have described events occurring during very high levels of diet-induced hypercholes-

LDL; Atherosclerosis;

Endothelium-de-

terolemia which seemed to resemble those observed in familial homozygous hypercholesterolemia. Nevertheless there is little information available concerning EDR, especially in the initiation of diet-induced coronary atherosclerosis. This study was therefore designed to examine (1) whether a mild degree of hyperlipidemia might alter porcine coronary artery reactivity; and (2) how atherogenic lipoprotein would affect the vasoactivity of the endothelium of porcine coronary arteries exposed to hyperlipidemia. Material and methods Source of normal and atherosclerotic blood vessels

Fifteen male Land-Yorkshire pigs, about 3 months old and weighing 20-25 kg, were divided into 3 groups. Pigs in the first group were fed a high-cholesterol diet (semi-synthetic diet containing 3.2% cholesterol and 20% lard) for 4 weeks (3.8 f 0.4 weeks). The animals in the second group received that diet for 9 weeks (8.6 f 0.7 weeks), while a third group consisted of control pigs fed regular pig mash for an average of 4.8 f 1.2 weeks. To prevent excessive weight gain, the daily food intake was limited to an amount equal to 3% of the body weight. On the day of study, the animals were sedated with ketamine (500 mg i.m.), anesthesized with pentobarbital (12.5 mg/kg i.v.) and then exanguinated. Plasma cholesterol level was measured by an enzymatic method [16,17]. To study the inhibitory effect of oxidized LDL after this experiment, the hearts of 5 adult Land-Yorkshire pigs of either sex weighing approximately 70 kg were used.

25 Evaluation of fatty streak formation A cross-section of the left descending coronary artery adjacent to each segment used in the experiments on contractility were examined histologically by hematoxylin-eosin staining of the endothelial lining for general observation, and by van Gieson’s elastic staining for the determination of the thickness of the intima. Morphometric determination was performed with a computer-assisted image analyzer (DTlOOO, Watanabe, and PC-9801, NEC, Japan). This system was used to evaluate the percent of the lumen’s circumference occupied by atherosclerotic lesions [15]. The left circumflex coronary artery was studied with an electron stain for elastic fiber and examined by transmission electron microscopy [18]. Lipoproteins Pig LDL (density, 1.019-1.063 g/ml) and VLDL (d < 1.006 g/ml) were isolated from the plasma collected in EDTA (1.5 mg/ml) by differential ultracentrifugation [19]. LDL and VLDL were dialyzed for 24 h against at least 2 changes of modified Krebs-Ringer solution with 60 PM EDTA (content below), and identified by agarose gel electrophoresis [20]. Absence of oxidization was checked by measuring thiobarbituric acid-reacting substances (TBAR) using a lipoperoxide test kit, and by measuring lipid peroxide using hemoglobin-methylene blue [21-231. The absence of fragmentation of apoprotein B was estimated by SDS-PAGE [24]. Protein content was determined by Lowry’s method [25]. Pig LDL was oxidized as follows: 1.0 mg protein from pig LDL which was dialyzed against 0.15 M NaCl with 10 PM EDTA was suspended in 1 ml DMEM containing 25 PM CuSO, and incubated at 37“C for 40 h in a 5% CO, incubator [22,26]. The oxidized LDL was then dialyzed for 24 h against at least 2 changes of modified Krebs Ringer solution (content below). Organ chamber experiments The left descending coronary artery was excised from each animal and trimmed free of adherent fat and connective tissue. Two transverse strips and rings 2 mm wide were cut from each artery using scissors. Intact rings and transverse strips, as well as those denuded of endothelium by a gentle

rubbing of the luminal surface using a swab moistened with control solution, were placed vertically between hooks in an organ bath containing 20 ml modified Krebs solution. It contained (in mM): NaCl, 118; KCI, 4.8; CaCl,, 2.5; MgSO,, 1.2; NaHCO,. 24; KH,PO,, 1.2; disodium EDTA, 0.06; and dextrose, 11. The bath solutions were maintained at 37°C and bubbled with a mixture of 95% 0,/5% CO,. The upper end of the strips was connected by a silk thread to the lever of a force-displacement transducer (TB-612T, Nihon Kohden Kogyo Co., Tokyo, Japan) [27]. Throughout the preparation and mounting of the coronary arteries tissue, special care was taken to avoid unintentional rubbing of the intimal surface either against a foreign surface or itself. Strips were then progressively stretched until the contractile response evoked by 20 mM KC1 was maximal (optimal tension) [28]. Strips were allowed to equilibrate for 90 min. After obtaining a reproducible response to 40 mM KC1 and 2.7 PM prostaglandin FZa (PGF,,), preparations were precontracted with PGF,, and then relaxed by cumulative concentrations of bradykinin (BK), substance P, serotonin (5HT), the CaZf ionophore A23187, and nitroglycerin. After washing, the rings were incubated for 30 min with LDL, VLDL or oxidized LDL. Relaxation was then evaluated during the contraction induced by PGF,, . Furthermore, relaxation was re-examined without preincubation with LDL, VLDL or oxidized LDL. While relaxation was being determined, the strips were incubated with indomethacin (5 x lo-- ’ M) for 30 min to inhibit the synthesis of endogenous prostaglandins. Similarly, when determining the relaxation induced by 5HT, strips were incubated with ketanserin (lop5 M), the 5-hydroxytryptamine, (5-HT,)-serotonergic antagonist, for 30 min to inhibit the direct activating effects of the monoamine on vascular smooth muscle. Statistical analysis Data are expressed as means f SEM. Statistical evaluation was made using Student’s t-test for unpaired comparisons of responses of rings of arteries from normal and hypercholesterolemic animals (N = at least 4; each value was the average of the results obtained with two arterial strips from each animal). The level of confidence chosen

26 for statistical significance was P < 0.05. To examine the effect of hyperlipidemia and the direct effect of lipoprotein, we computed the F ratio by using the analysis of variance (ANOVA procedure, SAS Statistical Programs, version 5, SAS Institute, Cary, NC, U.S.A.). Source of agents tested

The following pharmacological agents were used: bradykinin, substance P, 5-HT creatinine sulfate (serotonin), Ca2+ ionophore A23187, prostaglandin F2a (tris salt) and indomethacin (all from Sigma Chemicals, St.Louis, MO, U.S.A.); nitroglycerin (Nippon Kayaku, Tokyo, Japan) and ketanserin bitartrate (Janssen-Kyowa, Tokyo, Japan). Solutions were prepared fresh daily using distilled water. All concentrations are the final molar concentrations reached in the organ chamber. Results Lipid profile

There were no significant differences among the 3 groups in body weight and serum total protein. The total cholesterol level and the LDL cholesterol level were significantly higher in the C4 and C9 groups than that in the normal group. VLDL cholesterol and phospholipids tended to be higher in the C4 and C9 animals than in the normal controls, but not to a statistically significant extent. No significant alteration in HDL

TABLE

TABLE

2

DEVELOPED TENSION OF NORMAL AND HYPERCHOLESTEROLEMIC PORCINE CORONARY ARTERIES TO POTASSIUM (g) Data are expressed as means + SEM; n = 4. There is no significant difference between the tension evoked in arteries by each concentratioc of potassium in the normal vs. the cholesterol-fed pigs. Potassium

concentration

(mM)

10

20

30

40

Normal

0.69+0.15

1.50+0.19

2.01&0.16

2.59+0.09

Cholesterol 4 weeks 9 weeks

0.54 + 0.07 0.49kO.02

1.51+ 0.15 1.46+0.X

2.25 f 0.16 2.03*0.09

2.88 + 0.17 2.70+0.21

(s)

cholesterol or triglyceride was observed (Table 1). The total cholesterol level and HDL cholesterol level in adult pigs used in the additional experiment for oxidized LDL (92.5 + 9.5 mg/dl, 46.1 & 6.4 mg/dl) were not significantly different from those in the normal group. Baseline characteristics

The optimal resting tension did not differ significantly in the normal control vessels as compared to those of C4 and C9 animals (1.8 g). Contractions caused by potassium (lo-40 mM), and PGF,, (0.1-2.7 PM) did not differ significantly among the 3 groups except for those caused

1

BASELINE

DATA

IN CONTROL

AND

CHOLESTEROL

GROUPS

Mean + SD. Normal

Cholesterol 4 weeks

Body weight (kg) Total protein (g/dl) Total cholesterol (mg/dl) VLDL-cholesterol (mg/dl)

51.0& 8.0* 85.5* 6.0+

LDL-cholesterol (mg/dl) HDL-cholesterol (mg/dl) Triglyceride (mg/dl) Phospholipid (mg/dl)

35.2+ 3.8 44.3* 5.6 19.3c 6.4 93.8 + 14.8

*p