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

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
**p
2.5 0.3 8.3 0.7

52.5+ 8.4+ 174.6+ 10.0* 110.1+ 54.5-1: 26.6 + 129.4+

9 weeks 0.9 0.8 17.3 1.3 12.1 5.0 10.2 13.0

** * **

*

58.0* 6.8 8.0+ 0.3 218.5 + 32.9 15.0+ 2.3 150.6+23.9 52.8* 8.3 33.7+ 8.6 160.5 + 22.6

** * **

*

27 by exposure to a lower (Tables 2 and 3).

concentration

of PGF,,

Effect of hypercholesterolemia on relaxation The response to bradykinin in the presence of indomethacin in the vessels from the normal control and hypercholesterolemic pigs appears in Figs. 1 and 2. Bradykinin produced a concentration-dependent relaxation in the vessels with endothelium. In contrast to the response of normal the response to bradykinin was atarteries, tenuated in the hypercholesterolemic arteries. A typical response to bradykinin is shown in Fig. la, b, c. No relaxation was observed in those vessels without endothelium. The average maximum percent relaxation to bradykinin in the C9 vessels was less than in the normal control vessels. The relaxa-

TABLE

3

DEVELOPED TENSION OF NORMAL AND HYPERCHOLESTEROLEMIC PORCINE CORONARY ARTERY TO PROSTAGLANDIN Fzo (g) Data are expressed as means + SEM. Each group consists of 4 experiments on rings with endothelium. There is no significant difference between the tension evoked in arteries by each concentation of PGF,, in the normal vs. cholesterol-fed pigs. PGF,,

Normal

concentration 0.3

1

2.1

0.81i0.22

1.65+0.21

5.41kO.58

19.25k1.05

0.41 kO.09 0.35kO.14

1.68 f0.14 l.OOiO.12

6.10+0.38 4.40+0&l

25.06kO.82 20.86iO.93

Cholesterol 4 weeks 9weeks Tension

(g)

BK

BK lfJ7.5_

Ek! (N)

(PM)

0.1

d E;+i (N) + 101 75mg

prot./dl

PGFzcv 2.7 b E (+I (C9)

PGF*ct 2.7 Fig. 1. Relaxation in response to bradykinin in coronary arteries from normal (N) and cholesterol-fed (C4, C9) pigs in the presence of indomethacin (5 pM). E (+), with endothelium E (-), without endothelium. Strips were first contracted with prostaglandin F&. Concentration of agents added to Krebs solution in organ chamber are expressed as logarithms of cumulative molar concentrations. ‘Wo’ and dots indicate wash-out with Krebs-Ringer bicarbonate solution.

28

BK (-log

M) 12

SubP 11

(-log 10

5HT (-log

M)

M)

9

25

w

Normal

~+--a

Cholesterol

4weeks I”c:j

n----d

Cholesterol

Sweeks (C9)

N=4

t

p< 0.05

tt

p< 0.01

Fig. 2. Cumulative concentration-response curves to bradykinin, substance P, and serotonin during contraction evoked by prostaglandm Fza in control and cholesterol-fed pigs. Relaxation is expressed as a percent decrease in tension from the contraction difference (P < 0.05) between rings obtained from evoked by prostaglandin Fza. Data are shown as means + SEM. * Significant control pigs and cholesterol-fed pigs.

tion of the C4 vessels was significantly attenuated only at a high concentration of bradykinin compared with the normal control vessels (Fig. 2). The cumulative relaxation response to 5HT (in the presence of ketanserin), Substance P, A23187, and nitroglycerin in vessels from each group appears in Figs. 2 and 3. The relaxation induced by substance P was attenuated in the C9 vessels. The relaxation induced by 5HT was significantly attenuated only in the C9 vessels. The relaxation induced by the Ca2+ ionophore A23187 and by nitroglycerin was similar among the 3 groups. Morphology In the coronary arteries examined by light microscopy there were limited small areas of fragmentation of the internal elastic lamina induced by a high cholesterol diet for 4 weeks. Limited small areas of intimal thickening and fragmentation of the internal elastic lamina were observed in those animals given the high cholesterol diet for 9 weeks (Fig. 4). The percent of the lumen’ s cir-

cumference occupied by the lesions was 2.6 + 3.4%. Transmission electron micrography of the left circumflex artery showed that under control conditions, endothelial cells were flat with well-formed junctional complexes (Fig. 5a). There was limited fragmentation of the internal elastic lamina in vessels from the C4 and C9 groups. Cuboidal endothelial cells with prominent vacuoles were seen in vessels of the C9 group (Fig. 5b) with some subendothelial mononuclear cells. No prominent intimal thickening or foam cells were noted. Effect of atherogenic lipoprotein on relaxation We initially investigated the effect of LDL treatment on the endothelium-dependent relaxation induced by bradykinin. TBA reactive substance, LPO reactive substance, agarose gel electrophoresis, and SDS-PAGE of LDL (2 mg/ml containing 60 PM EDTA) did not change significantly during 30 min incubation in an organ chamber with aerated 95% 0,/5% CO,. A typical inhibition of relaxation by LDL at a concentration of 75 mg protein per dl is shown in Fig. Id

29

A23187 9

(-log 8

M) 9

7

C----0 A----C\

Normal Cholesterol Cholesterol

NG (-log 8 7

4 weeks 9 weeks

(NJ (C4) (C9)

M) 6

N=4 n =9

Fig. 3. Cumulative concentration-response curves to A23187 and nitroglycerin during contraction evoked by prostaglandin Fzu in control and cholesterol-fed pigs. Relaxation is expressed as percent decrease in tension from the contraction evoked by prostaglandin between rings obtained from control pigs and cholesterol-fed Fin. Data shown as means f SEM. * Significant difference (P < 0.05) pigs.

and e. Although LDL inhibited the endotheliumdependent relaxation, the contraction response to PGF,, by intact or endothelium-denuded preparations was unaltered by exposure to LDL. The inhibitory effect of LDL was completely abolished after washing and the endothelium-dependent relaxation to bradykinin was observed as before (data not shown). Fig. 6 shows the dose-response curves to cumulative concentrations of bradykinin and serotonin on intact rings precontracted with 2.7 PM PGF,, following preincubation with LDL at concentrations of 25 or 75 mg protein per dl for 30 min. At the low concentration of LDL, a small shift to the right in the dose-response curve to bradykinin is observed, not significant except in the C9 group. A concentration of 75 mg protein per dl of LDL produced a shift to the right and a decrease in relaxation in all 3 groups. In the case of 5-HT, 75 mg

protein per dl of LDL significantly attenuated relaxation especially in the arteries of the cholesterol-fed pigs (Fig. 6). LDL significantly inhibited arterial relaxation in the C9 group induced by substance P similarly. In the case of the Ca2+ ionophore A23187, LDL attenuated the endothelium-dependent relaxation in normal control vessels as well as in the arteries of both groups of cholesterol-fed pigs (C4, C9) (Fig. 7). In the case of nitroglycerin, LDL did not produce any effect on arterial relaxation (Fig. 8). The significance of these observations was evaluated statistically by the analysis of variance. VLDL were used to investigate the specificity of the inhibition (Table 4). VLDL (75 mg/dl) tended to inhibit EDR similar to LDL but the maximal percentage relaxation of EDR exceeded the relaxation preincubation with the same protein concentration of LDL and did not significantly inhibit EC,, evoked by bradykin-

a

Fig ,. 4. Light micrographs of cross-sections of the left descending coronary artery from a control pig (a (top), original magnific :ation, Xl 50) and a pig receiving a cholesterol-rich diet for 9 weeks (b (bottom), original magnification for determination of the thickr less of the int ima and internal elastic lamina. No intimal thickening is noted in Fig. 4a. Note the intimal thickening and fragmentation of the internal elastic lamina in (b).

Fig. 5. Transmission electron micrographs of right coronary artery from a control pig (a (top)) and a pig receiving a cholesterol-rich diet for 9 weeks (b (bottom)), (original magnification, X6000). In the control animals, the endothlium was flat with well-formed junc :tional complexes. Cuboidal endothlial cells with prominent vacuoles with some subendothelial mononuclear cells are seen in 5(b).

32

Normal

(-log

M)

Cholesterol 4weeks (-log

M)

Cholesterol

Sweeks (-log

M)

5HT

I

* --a

Control

-

I LOL25mg/dl

A----A

LDL75mghl

1 :t

I=4

p< 0.05 p< 0.01

Fig. 6. Cumulative concentration-response curves to bradykinin and serotonin with or without preincubation of LDL during a groups. Data are shown as means+SEM. Significant contraction evoked by prostaglandin F,, in control and cholesterol-fed ++ P -c0.01) between rings with and without preincubation of each concentration of LDL. difference (’ P i 0.05,

TABLE

4

ENDOTHELIUM-DEPENDENT WITHOUT PREINCUBATION

RELAXATION WITH VLDL

IN PIG CORONARY

ARTERIES

Data are expressed as means k SEM. There were experiments in each group. response to PGF,, EC50, EC25: effective concentration causing 50 or 25% EC,, was not calculated because the response did not attain the EC,, level. significant difference between rings with or without * P < 0.05, * * P < 0.01 Relaxing

agent

BK control VLDL (75 mg/dl)

5HT control VLDL (75 mg/dl)

133X* 94.3 f

2.7 9.4 *

93.5+ 5.6 63.5 f 10.5

OR 5-HT WITH

VLDL. (x10-‘M)

N4

c9

N

N4

c9

105.6* 7.9 + 74.6111.0

105.4+ 7.9 ++ 74.6kll.O

1.20f0.30 3.43 * 1.13

10.78*6.00 9.201t4.00

12.505 10.31 i

EC,,

M)

90.0f 47.6f

9.0 8.0 *

49.6 f 10.2 + l&l* 2.6 *

OR

Max grelaxation: maximal, relaxation as percent of the inhibition of the contractions to prostaglandin Fza; -, compared with normal control; +P < 0.05, ++ P -e0.01

EC,,

Max. %relaxation N

TO BRADYKININ

(x1O-8

1.20 f 0.45 4.40 * 1.21

0.62 k 0.30

40.6

3.00 + 5.90

k10.1

+

33

Normal

(-log

M)

Cholesterol 4 weeks (-log

Cholesterol

M)

9 weeks (-log

M)

Sub P

A231 87

I

u

Control

@-- a LO1 25mg/dl

,+--A

LO1

75mg/dl

N=4

t tt

P
Fig. 7. Cumulative concentration-response curves to substance P and to A23187 with or without preincubation of LDL during a contraction evoked by prostaglandin Fza in control and cholesterol-fed groups. Data shown as means + SEM. Significant difference (’ P < 0.05, ++ P < 0.01) between rings with and without preincubation of each concentration of LDL.

in. The effect of oxidized LDL was examined in a preliminary experiment (Fig. 9). Oxidized LDL (25 or 75 mg/dl) dialyzed with modified KrebsRinger solution did not decrease the relaxation of EDR evoked by bradykinin (10-7M) significantly. However oxidized LDL dialyzed with KrebsRinger solution without EDTA impaired the EDR to a significant extent. EDTA (60 PM) in organ chamber did not affect the percentage relaxation of EDR induced by exposure to bradykinin significantly. Discussion This study was designed to evaluate the effect of hyperlipidemia on the vascular responsiveness of isolated porcine coronary arteries. The optimal

tension and contraction induced by KC1 or prostaglandin F,, did not differ significantly among the test groups. Various studies have reported that the optimal tension and contraction induced by PGF,, is reduced in atherosclerosis [10,28]. We speculate that the optimal tension and contraction induced by PGF,, may remain within the normal range in the very early stage of atherosclerosis in agreement with the data reported by Shimokawa et al. Ill]. Since the effect of PGF,, may differ from those of other pharmacologic agents that induce muscle contraction, the evaluation of additional agents is indicated to examine more fully the capacity of the arteries to contract. This study demonstrates that a mild degree of hypercholesterolemia impairs the endothelium-dependent relaxation induced by exposure to

34

Normal

NG

(-log

M)

Cholesterol 9

0

6

Sweeks (-log 7

M) 6

.-y=--O 25 .-Cz $ C z : CL

50

25 \\ i\ \

50

75

75

100

100 -

A----A

Control LDL 75mg/dl

N=4

Fig. 8. Cumulative concentration-response curves to nitroglycerin with or without preincubation of LDL during a contraction evoked by prostaglandii Fzti in control and cholesterol-fed pigs. Data shown as means* SEM.

bradykinin, serotonin and substance P but that the hypercholesterolemia must be induced in pigs by feeding a high cholesterol diet for longer than 4 weeks; 9 weeks in this study. The relaxation response to A23187 or nitroglycerin was not impaired. We studied the effects of 5 pharmacological agents: bradykinin, serotonin, substance P, A23187, and nitroglycerin. Bradykinin, serotonin, and substance P are known to stimulate the release of EDRF by receptor-mediated mechanisms, A23187 stimulates the release of EDRF by nonreceptor-mediated mechanisms, and nitroglycerin is an endothelium-independent vasodilator [29]. We speculate from our observations that the first change in endothelium-dependent relaxation in atherosclerosis is not a decrease in the capacity to release EDRF but rather a decrease of endothelial receptor affinity for such vasodilators as bradykinin, serotonin, or substance P. The change in the response to bradykinin by hypercholesterolemia is the earliest and most dramatic. In contrast Simokawa et al. [ll] have reported the change in the response to serotonin to be more dramatic than that to bradykinin. We cannot explain the

difference between our results and those of Simokawa et al. as they did not present cumulative concentration-response curves to bradykinin. One explanation however may be a difference in the extent of hyperlipidemia and atherosclerosis since the change in the response to serotonin was dramatic in the C9 pigs in our experiment. The severity of the hyperlipidemia and of the atherosclerotic lesions was less in our hypercholesterolemic pigs than in those reported by other investigators [10,12,29]. However Simokawa et al. [ll] and Cohen et al. [lo] observed no atherosclerotic lesions by light and electron microscopy in severely hyperlipidemic pigs fed a high-cholesterol diet for 10 weeks. Although the level of serum cholesterol in our experiment was lower than in theirs, using transmission electron microscopy we found intimal changes and fragmentation of the internal elastic lamina in only a very limited area. Recently hyperlipidemia has been reported to cause morphological changes as early as 2 weeks [30,31]. One must be cautious in concluding an absence of atherosclerotic lesions in the presence of hyperlipidemia. These morphologi-

35

t CA Artery Ox-LDLImwdll EOTA

I”

OrganChamber -

CA -

CA

CA

CA

CA

c9

c9

25D-E

25D+E

75D+E

-

75D+E

+

+

25D+E _

+

+

+

+

Fig. 9. Endothelium-dependent relaxation to normal pig coronary arteries to bradykinin (lo-’ M) with or without preincubation with oxidized LDL during a contraction evoked by prostaglandin Fz,. Each column and vertical bar indicates mean f SEM. There are experiments in each group. %relaxation: relaxation as percent of the response to prostaglandin F,, (2.7 PM). Artery, CA: coronary arteries isolated from control pigs, C9 coronary arteries isolated from pigs in C9 group. Oxidized LDL, D + E: oxidized LDL dialyzed with modified Krebs-Ringer solution (containig 60 nM EDTA) after oxidization by exposure to Cuzf; D-E: oxidized LDL dialyzed with modified Krebs-Ringer solution without EDTA EDTA in organ chamber; + : modified Krebs-Ringer solution with EDTA (60 PM) in organ chamber; - : EDTA free modified Krebs-Ringer solution in organ chamber * P -C0.05significant difference between rings with or without oxidized LDL.

cal changes did not seem to present a barrier to the diffusion of EDRFs, or to cause a decrease in EDR, but may precede, or be correlated with, a decrease of EDR [32]. To evaluate the role of lipoprotein in the regulation of vascular tonus, we then investigated the direct effect of atherogenic lipoprotein on vascular responsiveness. A physiological concentration of LDL or VLDL did not alter the arterial contraction induced by KC1 or prostaglandin FZa in the control or hypercholesterolemic pigs. Preincubation with LDL inhibited the relaxations induced

by substance P or serotonin in the arteries of both the C4 and C9 pigs, but not in the control arteries. The degree of inhibition was especially marked in the C9 pigs. A total of 75 mg protein per dl LDL also inhibited the response to bradykinin or A23187 in normal arteries but the decrease was greater in the hypercholesterolemic arteries. As an inhibitory effect of LDL was also observed in the response to A23187, LDL might affect the formation and release of EDRF. This idea is consistent with the report that LDL inhibits two steps in the response of EDRFs [33]. It was reported recently that there may be two relaxing mediators on the nature of EDRF and that the EDRF might not be nitric oxide [34-361. The EDRFs thus might also differ between these agonists: one released from porcine coronary artery on stimulation with bradykinin and A23187 and the other released on stimulation with substance P and serotonin. Jacobs et al. [37] reported that LDL inhibited the relaxation evoked by exogenous NO and suggested that LDL were able to react with or sequestrate endogenous NO. They also reported that the inhibitory effect of LDL differed among pharmacologic agents producing contraction. Further investigation would be required to clarify how LDL further impairs EDR in atherosclerotic arteries from this aspect. In our study the inhibitory effect of LDL was reversible, differing from Andrews’ observation [13] but consistent with those of Jacobs [37] and Tomita [14]. It was confirmed in this experiment that LDL was not oxidized by incubation in Krebs-Ringer solution aerated with a mixture of 95% O,-5% CO,. Furthermore, our preliminary observation suggests that oxidized LDL dialyzed with modified Krebs-Ringer solution containing 60 PM EDTA does not have a remarkable inhibitory effect on EDR by bradykinin. Although it was reported that the half-life of EDRF is shortened by 0, or free radicals, the mechanism of the inhibitory effect of LDL seems to differ from that of the destruction of EDRF released by natural oxidization. This concept is consistent with that reported by_Andrews et al. [13] and Jacobs et al. [37]. Oxidized human LDL, especially lysophosphatidylcholine in oxidized human LDL, was recently reported to inhibit EDR by acetylcholine in

36 rabbit aorta [37,38], while another report showed that oxidized LDL did not affect the EDR in pig coronary artery [39]. Our preliminary findings may in part account for the differences. We found that the effect of Cu2+ was extruded and the character of cytotoxity was changed by dialyzing oxidized LDL with modified Krebs-Ringer solution containing EDTA (personal observation, data not shown). One must also take into account the difference in oxidized LDL used in each experiment, because the character of oxidized LDL (electrophoretic mobility, density, hydrolysis of phosphatidylcholine to lysophophatidylcholine, quantity of hydroperoxide and cytotoxity, etc.) is variable [40-421. There may also be pharmacologic differences among contracting agents and relaxing agents studied as well as differences in the response of various species and arteries. VLDL inhibited EDR similar to LDL but its effect was slightly weaker. We previously reported that rabbit fi-VLDL, the atherogenic lipoprotein in dietinduced atherosclerosis of rabbits, impairs the EDR in rabbits with more advanced aortic atherosclerosis [43]. Combining that finding with our present results, it is tempting to suggest that atherogenic lipoprotein impairs the EDR in normal arteries, and further impairs the already decreased EDR in arteries exposed to hyperlipidemia for prolonged periods. EDRF is reported to be secreted continuously in the absence of added endotheliumdependent relaxing agents [44]. EDRF is also reported to be able to decrease platelet aggregability [7,45]. Atherogenic lipoprotein may contribute to the progression of atherosclerosis in the regulation of vascular tonus and increase platelet aggregability by attenuating the EDR. The demonstration that EDRF and prostacyclin are released concomitantly from the endothelial cells by diverse stimuli [46], and that both inhibit platelet function in a synergistic manner [47], suggests that the release and action of 2 different mediators with similar actions can amplify cellular function [48]. Finally, we would like to consider the course of clinical atherosclerosis relevant to EDR. Strict control of the serum cholesterol level is currently recommended to retard the progression of atherosclerosis and to allow the lesion to regress [49-521. Our experimental results would appear to

support those recommendations since lowering the LDL level would promote circulation by causing vasodilation to some extent, and would also decrease platelet aggregability, even in the presence of mild atherosclerosis. It is of course assumed that the endothelium-dependent relaxation observed in vitro also exists in vivo. Acknowledgements

The authors wish to thank Jyunzo Kito and Tomoyuki Moriya, Experimental Animal Laboratory, Nagoya University School of Medicine, and Tomokazu Yanagihara, Seinoh Health and Sanitation Center, for their advice on animal care. We also wish to thank Sayuri Aoyama, Mari Yosimura and Michiko Doi for their expert technical assistance. This work was partially supported by a grant “Primary Hyperlipidemia” from the Ministry of Health and Welfare, Japan. References Hopkins, P.N. and Williams, R.R., A survey of 246 suggested coronary risk factors, Atherosclerosis, 40 (1981) 1. Yaari, S., Goldbourt, U., Evan-Zohar, S. and Neufeld, H.N., Associations of serum high density lipoprotein and total cholesterol with total, cardiovascular, and cancer mortality in a 7-year prospective study of 10,000 men, Lancet, i (1981) 1011. Igerman-Wojenski, C.M., Sedar, A.W. and Nissenbaum, M., Early morphological changes in the endothelium of a peripheral artery of rabbits fed an atherogenic diet, Exp. Mol. Pathol., 38 (1983) 48. Faggioto, A., Ross, R. and Harker, L., Studies of hypercholesterolemia in the nonhuman primate. I. Changes that lead to fatty streak formation, Arteriosclerosis, 4 (1984) 323. Ross, R., The pathogenesis of atherosclerosis-an update, N. Engl. J. Med., 314 (1986) 484. Sinjinger, H., Feigl, W. and Siiberbauer, K., Prosstacyclin generation in atherosclerotic arteries, Lancet, 11 (1979) 469. Furchgott, R.F. and Zawadzki, J.V., The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine, Nature, 288 (1980) 373. Verbeuren, T.J., Jordaens, F.H., Zomiekeyn, L.L., Van Hove, C.E., Coene, M.C. and Herman, A.G., Effect of hypercholesterolemia on vascular reactivity in the rabbbit. I. Endothelium-dependent and endothelium-independent contractions and relaxations in isolated arteries of control and hypercholesterolemic rabbits, Ciic. Res., 58 (1986) 552. Forstermarm, U., Mugge, A., Alheid, U., Haverich, A. and Frolich, J,C, Selective attenuation of endothelium-mediated

37

10

11

12

13

14

15

16 17

18

19 20 21 22

23

24

25

26

vasodilation in atherosclerotic human coronary arteries, Circ. Res., 62 (1988) 185. Cohen, R.A., Zitnay, K.M., Haudenschlid, C.C. and Cunningham, L.D., Loss of selective endothelial cell vasoactive functions caused by hypercholesterolemia in pig coronary arteries, Circ. Res., 63 (1988) 903. Simokawa, H. and Vanhoutte P.M., Impaired endothelium-dependent relaxation to aggregrating platelets and related vasoactive substances in porcine coronary arteries in hypercholesterolemia and atherosclerosis, Circ. Res., 64 (1989) 900. Friman, P.C., Mitchell, G.G., Heistad, D.D., Armstrong, M.L. and Heistad, D.D., Atherosclerosis impairs endothelium-dependent vascular relaxation to acetylcholine and thrombin in primates, Circ. Res., 58 (1986) 783. Andrews, H.E., Bruckdorfer, K.R., Dunn. R.C., Jacobs, M., Low density lipoproteins inhibit endothelium-dependent relaxation in rabbit aorta, Nature, 327 (1987) 237. Tomita, T., Ezaki, M., Nakamura, K. and Inoue, Y., Rapid inhibition of lipoprotein of the endothelium-dependent relaxation to hemostatic substances in porcine coronary arteries, Circ. Res., 66 (1990) 18. Weiner, B.H., Ockene, IS. and Hoogasian, J.J., Inhibition of atherosclerosis by cod-liver oil in a hyperlipidemic swine model, N. Engl. J. Med., 315 (1986) 841. Allain, C.C., Poon, IS. and Chen, C.S.G., Enzymatic determination of total cholesterol, Clin. Chem., 20 (1974) 470. Noma, A., Simultaneous determination of serum cholesterol in high- and low-density lipoproteins with use of heparin, Ca*+ and an anion exchange resin, Clin. Chem., 24 (1978) 1504. Kajikawa, K., Yamaguchi, T., Katsuda, S. and Miwa, A., An improved electron stain for elastic fibers using tarmic acid, J. Electron Microsc. 24 (1975) 287. Hatch, F.T. and Lees, R.S., Practical methods for plasma lipoprotein analysis, Adv. Lipid. Res., (1968) 1. Noble, R.P., Electrophoretic separation of plasma lipoproteins in agarose gel, J. Lipid Res., 9 (1968) 693. Yagi, K., A biochemical approach to atherogenesis, Trends Biochem. Sci., 11 (1986) 18. Heinecke, J.W., Rosen, H. and Chait, A., Iron and copper promote modification of low density lipoprotein by endothelial cells involves lipid peroxidation and degradation of low density lipoprotein phospholipids, Proc. Natl. Acad. Sci. USA, 81 (1984) 1890. Yagi, K., Yamauchi, T., Kobayashi, S., Komura, S., Ohnishi, N., Kayahara, N. and Tatano, T., Examination of lipid peroxide assay by the hemoglobinmethylene blue method. Lipid Peroxide Res., 11, (1986) 38. Muniz, N., Measurement of plasma lipoproteins by electrophoresis on polyacrylamide gel, Clin. Chem., 23 (1977) 1826. Lowry, O.H., Rosebrough, N.J., Farr, A.L., and Randall, R.J., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 265. Hayashi, T., Naito, M., Kuzuya, M., Funaki, C., Tateishi, T.. Asai, K. and Kuzuya, F., Role of platelet secretory

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

products in modified lipoprotein metabolism in macrophages, Artery, 16 (1989) 248. Ishikawa, T., Inagaki, M. and Hidaka, H., Relaxation of vascular smooth muscle by HA-1004, an inhibitor of cyclic nucleotide-dependent protein kinase, J. Pharacol. Exp. Ther., 235 (1985) 495. Simokawa, H., Aarhus, L.L. and Vanboute, P.M., Porcine coronary arteries with regenerated endothelium have a reduced endothelium-dependent responsiveness to aggregating platelets and serotonin, Circ. Res., 61 (1987) 256. Moncada, S., Radomski, M.W. and Palmer, R.M.J., Endothelium-derived relaxing factor, Biochem. Pharmcol, 37 (1989) 2495. Fish, R.D., Nabel, E.G., Selweyn, A.P., Ludmer, P.L., Mudge, G.H. and Kirshenbaum, J.M. Response of coronary arteries of cardiac transplant patients to acetylcholine, J. Clin. Invest., 81 (1988) 21. Reddick, R.L., Read, M.S., Brinkhous, K.M., Bellinger, D., Nichols, T. and Griggs, T.R., Coronary atherosclerosis in the pig, Arteriosclerosis, 10 (1990) 541. Cocks, T.M., Campbell, G.R. and Angus, J.A., Development of a large fibromuscular intimal thickening does not impair endothelium-dependent relaxation in the rabbit carotid artery, Blood Vessels, 24 (1987) 192. Bassenge, E. and Galle, J., Low density lipoprotein suppresses EDRF-mediated vasodilation by two different mechanisms, FASEB J., 3 (1989) A530 (Abstr.). Gabor, M.R. and Vanhoute, P.M., Nature of endotheliumdependent relaxing factor: are there two relaxing mediators? Circ. Res., 61 (1987). Boulanger, C., Hendrickson, R., Lorenz, R. and Vanhoutte, P.M., Release of different relaxing factors by cultured porcine endothelial cells, Circ. Res., 64 (1989) 1070. Myers, P.R., Minor, R.L., Jr., Guerra, R., Jr., Bates, J.N. and Harrison, D.G., Vasorelaxant properties of the endothelium-derived relaxing factor more closely resemble Snitrosocysteine than nitric oxide, Nature, 345 (1990) 161. Jacobs, M., Plane, F. and Bruckdorfer, K.R., Native and oxidized low-density lipoproteins have different inhibitory effects on endothelium-derived relaxing factor in rabbit aorta, Br. J. Pharmacol., 100 (1990) 21. Kuguyama, K., Kerns, S.A., Morrisett, J.D., Roberts, R. and Henry, P.D., Impairment of endothelium-dependent arterial relaxation by lysolecithin in modified low density lipoproteins, Nature, 344 (1990) 160. Galle, J., Bassenge, E. and Busse, R., Oxidized low density lipoproteins potentiate vasoconstrictions to various agonists by direct interaction with vascular smooth muscle, Circ. Res., 66 (1990) 1287. Wolf, A., Saito, T., Menon, N.K., Zehetgruber, M. and Bing, R.J., Effect of lysophophatidylcholine on atherosclerotic rabbit arteries, Atherosclerosis, 80 (1989) 81. Steinbrecher U.P., Parthasarathy S., Leake D.S., Witztum J.L. and Steinberg, D., Modification of low density lipoprotein by endothelial cells involves lipid peroxidation and degradation of low density lipoprotein phospholipids, Proc. Natl. Acad. Sci. USA, 81 (1984) 3883.

38 42 Jialal, I. and Chait, A., Differences in the metabolism of oxidatively modified low density lipoprotein and acetylated low density lipoprotein by human endothelial cells: inhibition of cholesterol esterification by oxidatively modified low density lipoprotein, J. Lipid Res., 30 (1989) 1561. 43 Hayashi, T., Ishikawa, T., Naito, M., Kuzuya M., Funaki C., Hidaka, H. and Kuzuya, F., /3-VLDL attenuates endothelium-dependent relaxation especially atherosclerotic aorta, Blood Vessels, (1989) 261. 44 Randall, M.D. and Hilley, CR., Detergent and methylene blue effect endothelium-dependent vasorelaxation and pressure/flow relations in rat blood perfused mesenteric arterial bed, Br. J. Pharmacol., 95 (1988) 1081. 45 Glygewski, R.J., Blotting, R.M. and Vane, J.R., Mediators produced by the endothehal cell, Hypertension, 12 (1988) 530. 46 de Nucci, G., Glyglewski, R.J., Warner, T.D. and Vane, J.R., Receptor-mediated release of endothelium-derived relaxing factor and prostacyclin from bovine aortic endothelial cells is coupled, Proc. Natl. Acad. Sci. USA, 85 (1988) 2334. 47 Sneddon, J.M. and Vane, J.R., Endothelium-derived relax-

48

49

50

51

52

ing factor reduces platelet adhesion to bovine endothelial cells, Proc. Natl. Acad. Sci. USA 85 (1988) 2800. Louis, J. and Ignarro, L.J., Biological actions and properties of endothelium-derived nitric oxide formed and released from artery and vein, Circ. Res., 65 (1989) 1. Wissler, R.W. and Vesselinovitch, D., Combined effects of cholestyramine and probucol on regression of atherosclerosis in rhesus monkey aortas, Appl. Pathol., 1 (1983) 89. Basta, L.L., Regression of atherosclerotic stenosing lesion of the renal arteries and spontaneous cure of systemic hypertension through control of hyperlipidemia, Am. J. Med., 61 (1976) 420. Barndt, R., Blankenhom, D.H., Crawford, D.W. and Brooks, S.H., Regression and progression of early femoral atherosclerosis in treated hyperlipoproteinemia patients, Ann. Intern. Med., 86 (1977) 139. Blankenhom, D.H., Neissin, S.A., Johnson, R.L., Sanmarco, M.E., Azen S.P. and Cashin-Hemphill, L., Beneficial effects of combined colestipol-niacin therapy on coronary atherosclerosis and coronary bypass grafts, JAMA, 257 (1987) 3233.