CS-514, a competitive inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A reductase: tissue-selective inhibition of sterol synthesis and hypolipidemic effect on various animal species

50

Eiochimica et Biophysics Acfa 877 (1986) 50-60 Elsevier

BBA 52250

CS-514, a competitive inhibitor of 3-hydroxy3-methylglutaryl coenzyme A reductase: tissue-selective inhibition of sterol synthesis and hypolipidemic effect on various animal species Yoshio Tsujita a**,Masao Kuroda ‘, Yoko Shimada a, Kazuhiko Tanzawa ‘, Mamoru Arai a, Isao Kaneko b, Minoru Tanaka ‘, Hiroshi Masuda ‘, Chitoshi Tarumi d, Yoshio Watanabe e and Setsuro Fujii f uFe~me~tufio~ Research ~boraforjes,

’ Biascience Laboratories, ’ Ana&fical and metabolic Research Lab~jraforie,~, ’ Laboratory of Animal Science and Toxicology ~bofaforjes, Sunkyo Co. Ltd., T&ya ’ Insfifate for E~~erjmental Animals, Kobe University School of Medicine, Kobe, and Osaka Universrfy, Osaka (Japan)

f

(Received

Key words:

Steroi synthesis;

November

Hydroxymethylgiutaryl-CoA

25 th, 1985)

reductase;

Hypo~ipidemic

drug; Lipoprotein;

(Cell culture)

CS-514 is a tissue-selective

inhibitor of 3-hydroxy3-methylglutaryl coenzyme A reductase, a key enzyme in cholesterol synthesis. For the microsomal enzyme from rat liver, the mode of inhibition is competitive with respect to hydroxymethylglutaryl-CoA, and the Ki value is 2.3 10 -9 M. CS-514 also strongly inhibited the sterol synthesis from [‘4Clacetate in cell-free enzyme systems from rat liver and in freshly isolated rat hepatocytes; the coneen~ations required for 50% inhibition were 0.8 ng/ml and 2.2 ng/ml, respectively. On the other hand, the inhibition by CS-514 was much less in the cells from nonhepatic tissues such as freshly isolated rat spleen cells, and cultured mouse L cells and human skin fibroblasts. In addition, the cellular uptake of “C-labeled CS-514 by isolated rat spleen cells or mouse L cells was less than one-tenth of that by isolated hepatocytes. These differences between hepatic and nonhepatic cells were further confirmed by the fact that CS-514 orally administered to rats inhibited sterol synthesis selectively in liver and intestine, the major sites of cholesterogenesis. CS-514 markedly reduced serum cholesterol levels in dogs, monkeys and rabbits, including Watanabe heritable hyperlipidemic (WHHL) rabbits, an animal model for familial hypercholesterolemia in man, but did not reduce those in rats and mice. In the former case, preferential lowering of atherogenic lipoproteins was observed in all of the animals tested. The biliary neutral sterols significantly decreased, whereas the amount of biliary bile acids was not affected by administration of the drug to dogs. l

Introduction In the development of atherosclerosis and coronary heart disease, which is the major cause of death in western and other developed countries, a high level of cholesterol in blood is considered to be an important risk factor [l]. Since more than

* To whom correspondence 0005-2760/86/$03.50

should

be addressed.

0 1986 Elsevier Science Publishers

70% of the total input of body cholesterol is derived from de novo synthesis in humans [2], it is expected that serum cholesterol can be reduced as a result of in~bition of cholesterol biosynthesis. The most suitable target for this inhibitor is 3-hydroxy-3-methylglutaryl coenzyme A reductase (mevalonate : NADP + oxidoreductase (CoAacylating), EC 1.1.1.34), the rate-limiting enzyme in the pathway of cholesterol biosynthesis [3]. On the basis of this concept, we carried out

B.V. (Biomedical

Division)

51

H3C

R2

Fig. 1. Structure of CS-514 and related compounds. CS-514 (R, = H, R, = OH; molecular weight, 447), ML-236B sodium salt (R, = R, = H) and MB-5308 (monacolin K or mevinolin) sodium salt (R, = CH,, R, = H).

screening of the inhibitor of hydroxymethylglutaryl-CoA reductase and found several compounds in microbial products [4-121, including ML-236B and MB-530B 1121. The latter is an identical compound to monacolin K 1131 or mevinolin 1141. Those compounds were potent competitive inhibitors of this enzyme. Among these microbial products and their chemically synthesized and microbially transformed derivatives, CS-514 (Fig. 1) was selected because of its potency and tissue selectivity. This drug was found at first as a minor urinary metabolite of ML-236B in dogs and later was obtained by microbial transformation of ML-236B [8,9], In this report, biochemical and pharmacological characteristics of CS-514 are described. Materials and Methods Materiels. [l- “C]Sodium acetate (58 Ci/mol), DL-[2-‘4C]mevalonola~tone (27.3 Ci/mol) and DL[3-i4C]hydroxymethylglutaryl-CoA (55.1 Ci/mol) were obtained from New England Nuclear Corp. (Boston, MA, U.S.A.) and [ car&y/-i4C]sodium deoxycholate (52 Ci/mol) was purchased from Amersham International (Amersham, U.K.). CS514 [8,9], ML-236B [5] and i4C-labeled ML-236B [7] were prepared as described previously. 14Clabeled CS-514 was prepared biosynthetically by adding “C-labeled ML-236B as a precursor to cultures of Nocardia aurotrophica grown in Erlenmeyer flasks, and purified extensively as described [S]. MB-530B was prepared from the culture filtrate of Monuscus ruber [12]. As internal standards for gas-liquid chromatography, neutral sterols, (Sa-cholestane, cholesterol, desmosterol, campesterol, stigmasterol, /3-sitosterol and lano-

sterol) and bile acids, (methyl esters of cholic, deoxycholic, lithocholic, chenodeoxycholic and 7ketolithocholic acid) were obtained from Gasukuro Kogyo Co. (Tokyo, Japan) and 9-anthryl diazomethane was purchased from the Funakoshi Pharmaceutical Co. (Tokyo, Japan). Animals. Wistar-Imamichi male rats weighing 120--140 g were used for usual experiments and rats weighing 2.50-330 g for the experiments of Triton-induced hyperlipidemia. The rats were maintained on a commercial rat chow (MM 1, Funabashi Farm Co, Chiba, Japan) for at least 1 week prior to use. Pure-bred beagle dogs weighing 8-10 kg were housed individually and fed a commercial dog food (Type IX, Oriental Yeast Co., Tokyo, Japan) at 200 g/day. Cynomolgus monkeys weighing 4.1-6.2 kg (male) and 3.2-4.1 kg (female) were housed individually and fed a commercial monkey food (Type AB, Oriental Yeast Co., Tokyo, Japan) at 80 g/day. Male Japanese white rabbits aged 11-13 months were housed individually and fed rabbit chow (RC-4, Oriental Yeast Co., Tokyo, Japan) at 150 g/day. Watanabe heritable hyperlipidemic (WHHL) rabbits weighing 3-3.5 kg were used for short-term experiments, and the rabbits at 3 months of age (weighing 1.8-2.2 kg) were used for a long-term experiment of 8 weeks. WHHL rabbits were housed individually and fed a commercial rabbit chow (Type GC, Oriental Yeast Co., Tokyo, Japan) at 120 g/day. In animal experiments with dogs, monkeys and rabbits, blood samples were withdrawn between 9 and 9.30 a.m. on appropriate days before feeding. Ce1I.s.All cultured cells used were maintained at 37°C in a 5% CO, incubator. Mouse L cells were seeded at about 2. lo5 cells per 60 mm dish and cultured for 3 days in Dulbecco’s modified Eagle’s minimum essential medium containing 5% fetal calf serum. Human skin fibroblasts from a normal subject (GM-442) and a homozygous familial hypercholesterolemic patient (GM-486) were obtained from the Institute of Medical Research (Camden, NJ, U.S.A.). These cells were grown in Dulbecco’s modified Eagle’s minimum essential medium containing 1% non-essential amino acids and 10% fetal calf serum. About 1 . lo5 cells were seeded onto 60 mm dishes according to a standard protocol [15] and were cultured for 5 days.

52

Rat hepatocytes were prepared by collagenase digestion according to the method of Mold&us et al. [16]. For preparation of rat spleen cell suspension, the spleen was cut into pieces with scissors and gently pressed on a stainless filter (100 mesh) using a silicon plug. The cells which filtered through were further treated twice on a 150 mesh filter and isolated by centrifugation at 1000 rpm for 5 min. The spleen cells thus obtained were completely disaggregated as single cells. Both hepatocytes and spleen cells were suspended at appropriate cell densities in serum-free Dulbecco’s modified Eagle’s minimum essential medium, and used for experiments within 3 h after preparation. Rabbit aortic fibroblasts were obtained from explants of aortic adventitia and cultured in Dulbecco’s modified Eagle’s minimum essential medium containing 5% fetal calf serum. About 1 . lo5 cells were seeded onto 60 mm dishes and used for experiments after 3 days cultivation. Cellular uptake of CS-514. Suspensions of freshly isolated rat hepatocytes (10’ cells/ml) or spleen cells (2 . lo* cells/ml) were incubated with various concentrations of i4C-labeled CS-514 or 14C-labeled ML-236B sodium salt at 37’C for 30 min in O,/CO, (95 : 5) gas with shaking. 80 ~1 of cell suspension were layered on 250 ~1 of the solvent mixture (butyl phthalate/dinonyl phthalate, 7 : 3, v/v) in a polyethylene tube (1A tube, Sanko Plastic Co., Osaka, Japan) and centrifuged at 3000 rpm for 2 min. The bottom of tube containing the cells was cut and put into a vial for determination of cellular radioactivity. Mouse L cells grown to form a monolayer in plastic dishes received 1 ml of serum-free Dulbecco’s modified Eagle’s minimum essential medium containing various concentrations of 14CCS-514 of r4C-ML-236B sodium salt. After incubation for 30 min at 37”C, cells in the medium were scraped with a rubber policeman. After the cell suspension was centrifuged at 1000 rpm for 5 min, about 900 ~1 of the medium were removed and cells were resuspended with the remaining medium. The cell suspension thus obtained was assayed for cellular radioactivity by the same procedure as that employed for rat hepatocytes. After measurement of cellular radioactivity, the pellet of mouse L cells was taken out from the counting vial, washed with phosphate-buffered saline four

times, dissolved with 4 M NaOH and assessed for protein content. Sterol from

synthesis

at 80°C for 1 h,

in the cell-free

enzyme

system

rat liver, cultured cells and tissue slices from

various organs of rats. [l-‘4C]Sodium

acetate was used as a precursor for sterol synthesis in these experiments. For the measurement of sterol synthesis in the cell-free enzyme system from rat liver, the reaction mixture was incubated for 2 h at 37°C and incorporation of radioactivity into nonsaponifiable lipids was assayed by the method described previously [5]. In the experiments with cultured cells, monolayers of mouse L cells, human skin fibroblasts and rabbit aortic fibroblasts were incubated with [*4C]sodium acetate at 37°C for 2 h. In the experiments with freshly isolated cells from rats, [‘4C]sodium acetate was added to 1 ml of suspension culture containing 2. lo6 hepatocytes or 5 . 10’ spleen cells, and the incubation was carried out at 37°C for 2 or 4 h, respectively. Incorporation of radioactivity into digitonin-precipitable sterols was measured by the method described previously [15]. In the experiment using tissue slices, rats were killed by carotid puncture and various organs were rapidly removed. The reaction mixture containing tissue slices and [r4C]sodium acetate was incubated for 2 h at 37°C in Krebs-Ringer bicarbonate buffer (pH 7.4) and incorporation of radioactivity into digitonin-precipitable sterols was measured as described previously [7]. All experiments were done by duplicate assays, unless stated otherwise. Assay

of hydroxymethylglutaryl-CoA

reductase

activity.

Hydroxymethylglutaryl-CoA reductase activities in rat liver microsome [6] and in cultured cells [15] were measured as described previously. Measurement

of hypolipidemic

in Triton WR-1339-induced

activity of CS-514

hyperlipidemic

rats. The experiment was carried out as described previously [17,18] except that the drug was administered orally at the time of Triton injection. Measurement of lipids and fractionation of serum lipoproteins. Serum lipids were measured by en-

zymatic methods using the following assay kits unless otherwise stated: cholesterol, Determiner TC (Kyowa Hakko Co., Tokyo, Japan); triacylglycerol and phospholipid, Triglyceride G-Test Wako and Phospholipid B-Test Wako, respectively (Wako Pure Chemical Industries, Osaka,

53

Japan). Serum cholesterol levels in a Triton-induced hyperlipidemic rats were assayed by a slightly modified method of Zak (191 and Henly 1201. For the fractionation of serum lipoprotein, 0.5 ml (WHHL rabbit) or 1 ml (other animals) of serum was centrifuged according to the method of Hatch and Lees [21] using a Hitachi 65P ultracentrifuge equipped with an RPS-56T rotor (Hitachi Co., Tokyo, Japan). Each lipoprotein was fractionated as follows: very-low-density lipoprotein (VLDL, d < 1,006); intermediate-density lipoprotein (IDL, 1.006 < d < 1.019); low-density lipoprotein (LDL, 1.019 < d -C 1.063); high-density lipoprotein (HDL, 1.063 < d < 1.21) and infranatant (d> 1.21). Analysis of biliary lipids. Bile samples for analysis were prepared as described previously 1191.For correcting recovery of neutral sterols and bile acids, Sa-cholestane and [‘4C]deoxycholate were added to bile samples, respectively. Neutral sterols were determined by gas-liquid chromatography [22]. Bile acid was measured by high-performance liquid chromatography: the samples treated with St-anthryl diazomethane were chromatographed on a 25 cm ODS C-18 column m~ntained at 35°C in a Hitachi Type 655 ~gh-performance liquid chromatography (Hitachi Co., Tokyo, Japan) equipped with a fluorometric detector FP-110 (Japan Spectroscopic Co., Tokyo, Japan). Under these conditions, the overall recovery of bile acid varied between 65 and 95%. The phospholipid in bile was assayed by phosphorus analysis using a kit, Phospholipid-Test Wako (Wako Pure Chemical Industries, Osaka, Japan). Other assay. Protein content was determined by the method of Lowry et al. [23] with bovine serum albumin as standard. Results Specific synthesis

and

tissue-selective

inhibition

of sterol

Fig. 2 shows the effect of CS-514 on the incorporation of various radiolabeled precursors into digitonin-pr~ipitable sterols in the cell-free system from rat liver. The conversion of [‘4C]acetate and DL-[14C]hydrox~ethylglut~l-CoA was in-

01 0.1

L,

I’

;

n--- n

1.0 CS-514

“,

10

me?

100

J

(ng/mlI

Fig. 2. Effect of CS-514 on cholesterol synthesis from acetate, hydroxymethyl~utaryi-CoA or mevafonate in a cell-free enzyme system from rat liver. The assay method of sterol synthesis is described in Materials and Methods. The concentrations of the precursors were; 1 mM, 0.3 mM and 0.5 mM for [ “C]acetate (O), [‘4C)hydroxymethylglutaryl-CoA (A) and [‘4C]mevalonate (Cl), respectively. Each point is plotted as a percent inhibition to the respective control.

hibited to almost the same extent, i.e., the concentrations required for 50% inhibition (I,,) were 0.8 ng/ml and 0.9 ng/ml, respectively. In the case of ML-236B, the mother compound of CS-514, I,, for the conversion of [t4C]acetate was 10 ng/ml. On the other hand, the conversion of [‘4C]mevalonate was not affected by CS-514 at concentrations up to 100 ng/ml, indicating that this drug inhibited specifically the enzymatic step of the conversion of hydroxymethylglutaryl-CoA to mevalonate catalysed by hydroxymethylglutarylCoA reductase. The mode of inhibition of CS-514 for hydroxymethylglutaryl-CoA reductase was competitive with respect to hydroxymethylglutaryl-CoA CoA and non-~mpetitive with respect to NADPH. The K, values for the reductase from rat liver and mouse L cells were 2.3 . 10m9 M and 2.2 - 10s9 M, respectively. Table I summarizes the values of I,, for incorporation of [14C]acetate into sterols either by CS514 or by ML-236B in freshly isolated and cultured cells. Both CS-514 and ML-236B strongly inhibited sterol synthesis in freshly isolated rat hepatocytes. The values of I,, were 2.2 ng/ml and 7.0 ng/ml, respectively, both of which were close to the values obtained in the cell-free enzyme system from rat liver, suggesting that both drugs were internalized and reached the target enzyme

54

in hepatocytes. In the cells from nonhepatic tissues, such as freshly isolated rat spleen cells and various cultured cells, ML-236B exerted the potent inhibitory activity as in rat hepatocytes. In contrast, the inhibitory activity of CS-514 was much less potent in these cells from nonhepatic tissues than in rat hepatocytes. In order to verify the assumption that the results mentioned above might be explained by poor permeability of CS-514 into the cells from nonhepatic tissues, cellular uptake of “C-CS-514 and 14C-ML-236B sodium salt was compared among freshly isolated rat hepatocytes and spleen cells, and mouse L cells. As shown in Fig. 3, the cellular uptake of “C-CS-514 in hepatocytes occurred to almost the same extent as that of 14C-ML-236B sodium salt. The uptake of i4C-CS-514 in spleen cells and mouse L cells, however, was less than one-tenth of that of 14CML-236B sodium salt. From these results, the less potent inhibitory activity of CS-514 in the cells from nonhepatic tissues can be ascribed to lower uptake of the drug by those cells. After CS-514, ML-236B or MB-530B was administered orally to rats, the activity of sterol synthesis in slices of various organs was measured in vitro and inhibition in each organ was calculated (Table II). CS-514 inhibited the sterol

TABLE

synthesis selectively in liver and ileum (intestine), the major sites of cholesterogenesis, but only weakly inhibited that in other organs, including hormone-producing ones. In the cases of ML-236B and MB-530B, although the inhibition of sterol synthesis in liver and intestine was most potent, that in other organs was significant as well. These results indicate that the inhibitory activity of CS514 is tissue selective, compared to those of ML236B and MB-530B, and this observation is consistent with the results obtained with the isolated and cultured cells.

I

EFFECT OF G-514 AND ML-236B ON STEROL SYNTHESIS IN VARIOUS ISOLATED AND CULTURED CELLS The assay methods of sterol synthesis are described in Materials and Methods. The inhibitory activities of CS-514 and ML-236B on sterol synthesis are expressed by the concentrations required for 50% inhibition (I,,). n.t.. not tested. Cells

I,, (&ml) cs-514

Freshly isolated rat hepatocytes Freshly isolated rat spleen cells Mouse L cells Human skin fibroblasts (normal) (homozygous familial hypercholesterolemia) Rabbit aortic fibroblasts * Result obtained

with sodium

2.2 70 600

salt.

ML-236B 7.0 * 1.3 * 1.4

200

18.0

400 750

12.0 nt.

f4ClCompound added kpm/ml) Fig. 3. Cellular uptake of “C-CS-514 and 14C-ML-236B sodium salt into (A) freshly isolated rat hepatocytes, (B) freshly isolated spleen cells and (C) mouse L cells. The cells were incubated with the indicated concentrations of “C-CS-514 (0) or 14C-ML-236B sodium salt (0) at 37°C for 30 min. Uptake of radioactivity into cells was determined as described in Materials and Methods. Each point expresses the average of duplicate (rat hepatocytes and spleen cells) or quadruplicate (mouse L cells) incubations.

55

TABLE

If

INHIBITORY ACTIVITY OF CS-514, ML-236B AND MB530B (MONACOLIN K OR MEVINOLIN) ON STEROL SYNTHESIS OF VARIOUS TISSUES IN RATS CS-514, ML-236B or MB-530B was suspended in 10% gum arabic and ad~~stered orally at the dose of 25 mg/kg to male Wistar-Imamichi rats. After 2 h, the rats were killed and the sterol synthesis in slices of each organ was measured as described in Materials and Methods. The data represent the average of values obtained from six rats. Organs

Liver Ileum Kidney Lung Spleen Cerebrum Prostate Testis Adrenal Muscle

Inhibition

(%)

as compared to the control value after administration of CS-514 (20 mg/kg) to the rats. The time courses of changes in serum lipid levels in beagle dogs treated with CS-514 are shown in Fig. 4. When the drug was administered orally twice a day at doses of 0.625 and 1.25 mg/kg per day, serum cholesterol and phospholipid levels were reduced dose and time depen-

TABLE

cs-514

ML-236B

MB-530B

93.1 *** 94.8 *** 29.8 * 2.5 0 0 10.4 7.6 7.8 0 0

89.5 86.5 80.6 51.4 54.2 12.2 66.2 54.2 45.0 38.5 27.9

95.8 80.1 71.1 30.8 43.1 20.8 70.7 60.2 78.7 48.1 12.2

Significantly different from control 0.01; *** P~O.001.

*** *** *** ** *** *** ***

value:

*** ** **

*** *** **

* P < 0.05; ** P i

Hypolipidemic effect of CS-514 on oarious animal species As shown in Table III, CS-514 significantly decreased serum cholesterol levels in beagle dogs, cynomolgus monkeys, Japanese white rabbits and WHHL rabbits, an animal mode1 of familial hypercholesterolemia in man [24,25], at doses of O-625-50 mg/kg per day. Serum phospholipid levels also decreased in dogs, rabbits and WHHL rabbits, whereas the levels of triacylglycerol were not reduced significantly in almost all the animals tested. On the other hand, CS-514 had no lowering effect on serum cholesterol levels in rats, even at a dose as high as 500 mg/kg per day. The drug also showed no effect on serum cholesterol levels in mice and spontaneous hyperlipidemic Nagase analbuminemia rats [26] (data not shown). Among rats and mice, CS-514 exhibited an exceptionally hypolipide~~ effect on Triton-induced hyperlipidemic rats which cannot metabolize serum lipoproteins and which show increased cholesterol synthetic activity in liver [17]. The serum cholesterol levels were reduced by 26% (P < 0.01)

EFFECT ANIMAL

III OF CS-514 SPECIES

ON

SERUM

LIPIDS

IN VARIOUS

Each animal group consists of: A and C, 3 males and 3 females; B, 2 males and 2 females; D and E, all males. CS-514 was administered twice a day (9.30 a.m. and 4.30 p.m.) at the indicated doses, except for rats which received the drug once a day (9.30 a.m.). The drug was administered by: A, a gelatin capsule; B, nasogastric intubation; C and E, gastric intubation; D, oral administration, as described previously [27]. The initial values of serum lipids in each animal group were obtained from the average value of at least three point assays.

Dose (mg/kg

Percent of initial vaIue (mean * S.D.) per day) total cholesterol

A. Beagle dog (18 days, n = 6) Control 0.625 1.25

96+ 88i 82i

phospholipid

5 5b 6’

96& 88+ 64k

4 6b 6b

triacylglycerol

103 f 10 83+ Sb 89kl4

B. Cynomolgus monkey (18 days, n = 4) Control 96k 4 20 85k 6a 50 69+11 b

85k 8 87k 9 84+11

C. Japanese white rabbit (18 days, n = 6) Control 96& 9 6.25 78klO b 12.5 68fllb

96+ 84+ 73*

3 7b 7b

94+35 79kl8 77+33

D. WHHL rabbit (12 days, n = 4) Control 100~20 12.5 82+ 5’ 72+ 9’ 50

93+ 88* 84*

4 5 5b

108+13 92511 107*10

E. Wistar-Imamichi rat (14 days, n = 8) 500 118 * Significantly different 0.01; c P c 0.001. * The values represent

101 * from

control

percent

value:

of control.

110+37 96rt 7 94$-15

69 * a P -c0.05;

b P -c

Phospholipid

Triglyceride

Fig. 4. Effect of CS-514 on serum lipid levels in beagle dogs. Experimental conditions are described in Table III. CS-514 was administered orally at doses of 0.625 (a) and 1.25 (0) mg/kg per day for 18 days, respectively, and control animals (0) received placebo capsules. The data represent average values of percent of initial values for six animals at each point. Percent of the initial value of treated groups is significantly different from that of the control group at each point as indicated: * P < 0.05; ** P CC0.01; *** P i Days

dently; the decreases in cholesterol on day 18 were 12 and 188, respectively. The reduced levels of cholesterol and phospholipid recovered to pretreatment levels in about 2 weeks after dicontinuation of drug administration, The effect of CS-514 on serum triacylglycerol levels was unclear because of its inconsistently in control animals throughout the experimental period. In other animals, including monkeys, rabbits and WHHL rabbits, CS-514 also reduced serum cholesterol and phospholipid levels time and dose dependently, and the reduced levels were restored within 2 weeks after termination of the drug administration (data not shown). Table IV shows the effect of oral administration of CS-514 on lipoprotein cholesterol in dogs, monkeys and WHHL rabbits at the indicated doses and periods. In dogs, serum cholesterol levels were lowered by 29% (P < 0.01) and LDL cholesterol levels preferentially decreased by 89% (P < 0.001) compared with those of control animals. Although HDL cholesterol levels were also reduced, the atherogenic index (A.I.) was lowered significantly by 78% as compared with that of the control group. In the case of monkeys, VLDL and LDL cholesterol levels were significantly reduced by 71 and 36%, respectively, while HDL cholesterol levels were not affected. AI. of treated monkeys was lowered by 40% (P < 0.01). In WHHL rabbits, VLDL, IDL and LDL cholesterol levels were significantly reduced by 54, 46 and t7%, respectively, whereas HDL cholesterol

levels were not affected. A.I., which is very high in WHHL rabbits, was also reduced by 38%. In all the animal experiments described above, the effect of CS-514 on the lipid composition of lipoproteins was as follows (data not shown): (i) the ratio of triacylglycerol to whole lipids in VLDL relatively increased by about lo-20%, because the decrease of cholesterol and phospholipid was more prominent than that of t~acylglycerol. The lipid composition of other lipoproteins including IDL, LDL and HDL was not significantly affected. (ii) The ratio of free to total cholesterol in each lipoprotein fraction was not significantly changed, suggesting that this drug did not alter the cholesterol esterification. Effect on biliary lipids in dogs Table V shows the effect of CS-514 on the biliary lipids in beagle dogs which received the drug at doses of 12.5 and 50 mg/kg per day for 5 weeks. Although the amounts of phospholipid and bile acid were not altered significantly, those of cholesterol and total neutral sterols were reduced by 40% (P < 0.01) and 37% (P < 0.02), respectively. The biliary bile acid composition was not greatly affected by the drug treatment except for a slight decrease of chenodeoxycholic acid. The lithogenic index was significantly and dose dependently reduced by 21 and 44% at doses of 12.5 and 50 mg/kg per day, respectively, suggesting that the drug had an antilithogenic activity.

57

TABLE IV EFFECT OF CS-514 ON LIPOPROTEIN CHOLESTEROL LEVELS IN VARIOUS ANIMAL SPECIES The male beagle dogs received CS-514 in a gelatin capsule by gavage twice a day (9.30 a.m. and 4.30 p.m.). The experimental condition of cynomolgus monkeys was the same as in Table III. Three male and three female WHHL rabbits received CS-514 by oral administration once a day as described previously (271.The preparation method of lipoprotein is described in Materials and Methods. Each value represents mean+S.D. AI. (atherogenic index) was calculated from the following formula: ((VLDL cholesterol(C))+(IDL C) + (LDL C))/((HDL C)+(d > 1.21 C)). Lipoprotein cholesterol mg/dI) VLDL

IDL

Beagle dog (20 mg/kg per day, 35 days, n = 5) 1.5* 0.3 control 0.9* 0.6 treated % decrease

Cynomolgus monkey (50 mg/kg per day, 18 days, n control 2.1 Ifr 2.0 treated 0.6+ 1.0 % decrease

% decrease

=

4) -

54

10.1 tt: 1.0 1.1+ 0.9 (PC0.001)

77.2 &-5.6 59.8 rt 8.0 (P < 0.01)

73.11 13.7 46.91 2.8 (P <: 0.01) 36

70

WHHL rabbit (50 mg/kg per day, 8 weeks, n control 118 +31 54 +12 treated (P < 0.001)

HDL

89

40

=

6) 98f28 53*19 (P < 0.01) 46

d > 1.21

LDL

533 tt127 389 + 48 (P < 0.05) 27

Other effects

A concentration of more than 100 pg/ml of CS-514 was required to cause growth inhibition of cultured cells such as mouse L cells and human skin fibroblasts. However, this effect was completely counteracted by addition of a small amount of mevalonic acid, a product of hydroxymethylglutaryl-CoA reductase (data not shown). A 20times higher concentration of CS-514 than the I,, value of sterol synthesis in human skin fibroblasts was required to cause 50% in~bition of coenzyme Q synthesis, one of the branched pathway products of mevalonic acid (Shimada, Y. and Tsujita, Y., unpublished result). In mouse L cells, CS-514 had no effect on the synthesis of DNA, RNA, protein, triacylglycerol, phospholipid and fatty acid even at a 1700-times higher concentration of

23

40.2 rt 9.1 42.1+ 2.6 -5

0

2.2kO.9 2.7rtO.4 -23

6.4 Itr0.9 7.OIt2.3 -9

4.2 + 0.8 4.2 jr 0.4

5.5+ 1.2 6.4+2.1 -16

Total

A.I.

92.7+ 6.1 66.0* 8.2 (P < 0.001)

0.143 + 0.010 0.031 I 0.023 (P
29

117.6+ 20.9 92.3+ 5.3 (P < 0.05) 22

760 +152 509 * 59 (P < 0.01) 33

78

1.79 c 0.25 1.06 + 0.06 (P < 0.001) 40

63 +13 39 +13 (P
I,, for cholesterol synthesis (data not shown). In terms of ketogenesis, the drug neither inhibited hydroxymethylglutaryl-CoA lyase from rat liver at a 4000-times higher concentration of I,, for hydroxymethylglutaryl-CoA reductase, nor affected the blood ketone body levels in rats which were given CS-514 at a dose of 500 mg/kg per day for 14 days (data not shown). In the long-term administration of CS-514 to dogs (2 years), the appearance of estrus in female animals was not affected, in spite of a 30-4096 decrease in serum cholesterol levels (Masuda et al., unpublished result). In all animals administered CS-514, body weight gain and food intake were not affected during the experimental periods.

58

TABLE

V

EFFECT

OF CS-514 ON BILIARY

LIPID

IN BEAGLE

DOGS

The beagle dogs (3 males and 3 females) received (X-514 in a gelatin capsule by gavage once a day (9.30 a.m.) at doses of 12.5 and 50 mg/kg per day for 5 weeks, respectively. Bile lipids were determined as described in Materials and Methods. The Iithogenic index was calculated from the following formula: neutral sterol (mol)/(phospholipid (mol) + bile acid (mol)). n.d., not detected. Dose (mg/kg) control Phospholipid (mg/mI)

53.3

(% of control)

567.5

(% of control) phytosterols (pg/ml)

(100) 103.7

(% of control)

(100) 671.2

total (pg/ml) (S of control)

104.9

(S of control)

(100)

index

(% of control)

5.3 *

acid

4.7 14.0 5.9 71.1 4.3

46.9

50

*

6.8

*131.5

432.7

*

81.6

+ 54.9

(78) 104.5

*

34.1

538.2

+110.8

(80)

+ 33.3

107.8

f

10.5

0.0057 *

2.1 6.2 3.3 6.4 0.9 0.0002

(100)

6.3 21.3 2.8 63.2 6.4

7.5

338.1 k 58.3 (60, P < 0.01) 82.4 f 30.6

420.5 f85.8 (63, P < 0.02)

118.4

f 12.8

(113)

(103)

* f * &*

f

(80)

(101) * 178.8

44.4 (83)

(88)

(100)

Bile acid total (mg/mI)

Lithogenic

f

(100)

Neutral sterol cholesterol (gg/ml)

composition (S) lithocholic acid deoxycholic acid chenodeoxychohc cholic acid others

12.5

+ 2.4 k 8.0 * 5.9 k 18.9 + 3.8

0.0045 * 0.0013 (79, P < 0.05)

5.2 14.5 nd. 75.9 3.9

f 1.1 + 5.6 * f

5.3 1.0

0.0032 + 0.0008 (56, P < 0.001)

* Mean + S.D.

Discussion The present results have shown CS-514 to be a specific inhibitor of hydroxymethylglutaryl-CoA reductase. (i) In the cell-free enzyme system from rat liver, the incorporation of [‘4C]hydroxymethylglutaryl-CoA into sterol was inhibited to almost the same extent as that of [14C]acetate, but that of [‘4C]mevalonate was not affected by CS-514. (ii) In mouse L cells, the drug showed no effect on syntheses of macromolecules and lipids other than cholesterol. (iii) In cultured cells, the growth inhibition by a high dose of CS-514 was completely recovered with the addition of a small amount of mevalonic acid, the product of hydroxymethylglutaryl-CoA reductase. CS-514 showed tissue-selective inhibition of

sterol synthesis. When the drug was administered orally to rats, the sterol synthesis of liver and intestine was markedly inhibited, whereas only weak inhibition was observed in other organs, including steroid hormone-producing ones (Table II). The data obtained from autoradiograms with i4C-CS-514 in rats and dogs also revealed a preferential distribution of the drug to liver and intestine (Tanaka et al., unpublished result), which is consistent with the data of the inhibition study described above. This result is explained primarily by a clear difference in the cell permeability of CS-514 to cells from hepatic and nonhepatic tissues (Fig. 3), and further reinforced by the enterohepatic circulation of CS-514 (Tanaka et al., unpublished result). In the case of labeled 3-hydroxy-3-methylglutaric acid, whose structure is

59

analogous to that of CS-514, far more radioactivity was found in kidney than in liver and intestine [28]. Therefore, the preferential distribution of CS-514 to liver and intestine is probably due to its decaline structure, although minor but significant distribution of CS-514 to kidney (Table II) might be related to its analogous structure with 3-hydroxy-3-methylglutaric acid in its molecule. In addition, ML-236B and MB-530B, both without possessing a hydroxyl group at the 6/3 position, showed less tissue selectivity. It is suggested, therefore, that the hydroxyl group at the 6fl position strongly contributed to the tissue-selective inhibition of sterol synthesis. According to the study of Spady and Dietschy water as a p recursor al[29], using 3H-labeled lowed greater sterol synthesis in nonhepatic tissues than that reported earlier using 14C-labeled substances as precursors, but liver and intestine are still major sites of sterol synthesis. Evidently, sterol synthesis in liver is closely related to the clearance of LDL from the blood. Kovanen et al. [30] reported that the activity of hepatic LDL receptor was increased in young beagle dogs by the treatment of the inhibitor of cholesterol synthesis, mevinolin. More recently, Bilheimer et al. [31] transplanted a normal liver into a familial hypercholesterolemic patient and successfully normalized the serum LDL level. These and other lines of evidence [32] suggest that the liver is the major organ for LDL uptake and that the inhibitors of cholesterol synthesis express its hypolipidemic action via the increase of hepatic LDL receptor activity [33]. In these respects, the selective distribution of CS-514 to liver and intestine is considered to be a desirable feature as a hypolipidemic agent because of its efficient inhibition of hepatic cholesterol synthesis and minimum disturbance of the metabolism of sterols or mevalonic acid in other organs. Hence, this tissue selectivity is one of the advantages of CS-514 over ML-236B, MB-530B or other related compounds. CS-514 decreased serum cholesterol levels in dogs, rabbits, WHHL rabbits (Table III) and Triton-induced hyperlipidemic rats. In the case of dogs, CS-514 reduced serum cholesterol levels significantly by 12% at a dose as low as 0.625 mg/kg per day for 18 days. When ML-236B was administered orally twice a day at doses of 1 mg and 4

mg/kg per day for 3 weeks to beagle dogs, serum cholesterol levels were reduced significantly by 12 and 20%, respectively (Tsujita et al., unpublished results). This suggests that CS-514 is more potent than ML-236B in reducing serum cholesterol levels in dogs. Among the lipoprotein fractions tested, the atherogenic lipoprotein cholesterols were preferentially reduced by the administration of CS-514 in dogs, monkeys and WHHL rabbits (Table IV). Accordingly, the atherogenic indices were markedly decreased. CS-514 also decreased serum phospholipid levels by almost the same degree as cholesterol levels in all the animal tested (Table III). Considering the fact that CS-514 has no inhibitory effect on phospholipid synthesis in mouse L cells, the decrease of phospholipid could be attributed to the decrease in the number of LDL, which contain a large amount of phospholipid as well as cholesterol. In contrast to phospholipid triacylglycerol levels in whole serum were not altered significantly. Concomitantly, the ratio of triacylglycerol to other lipids increases in VLDL in the animals treated with CS-514. It is possible that cholesterol was substituted for triacylglycerol when VLDL is produced. This hypothesis will be proved by the careful comparison of the particle size of VLDL before and after CS-514 treatment. CS-514 exhibited no cholesterol-lowering effect in normal rats, mice and Nagase analbuminemic rats, indicating that the drug possesses a clear species specificity in hypocholesterolemic effect as in the case of ML-236B [18]. Rats administered ML-236B were shown to exhibit a reduction of bile acid excretion and an increase of hydroxymethylglutaryl-CoA reductase activity in liver [18]. Although CS-514 may cause the same responses, these are not considered to give a full explanation for the ineffectiveness of the drug in mice and rats. This problem must be elucidated by further investigations. WHHL rabbit, which lacks functional LDL receptor, is an animal model for familial hypercholesterolemia in man to which many hypolipidemic drugs currently administered are ineffective. Actually, clofibrate, a widely used hypolipidemic drug, failed to reduce serum cholesterol levels in this animal, even at a dose of 165 mg/kg per day for 40 days [34]. On the other hand,

60

B-514 lowered serum cholesterol levels in WHHL rabbits by 18% (P < 0.001) at a dose of 12.5 mg/kg per day for 12 days. In WHHL rabbits, the percent decrease of LDL cholesterol was quite a bit lower than those of VLDL cholesterol and IDL cholesterol (Table IV). This phenomenon might be related to a defect in the hepatic LDL receptor in WHHL rabbits 1351. It has been reported recently that a hypolipide~c drug, probucol, produced a 23% decrease in serum cholesterol levels in WHHL rabbits [36]. However, the administered dose, 1% in diet, is remarkably higher than that of CS-514. In these respects, CS-514 can be expected to reduce serum cholesterol levels in at least some types of homozygous familial hypercholesterolemia in man. Acknowledgments

The authors wish to thank Mr. T. Totori, Mmes. K. Nakazawa-Ishii and I. Suzuki-Oikawa for their excellent technical assistance and Messrs. M. Asai and S. Sudo for their collaborations in part of the experiments. References 1 Page, I.H., Berrettoni,

J.N., Butkus, A. and Sones, F.M., Jr. (1970) Circulation 42, 625-646 2 Dietschy, J.M. and Wilson, J.D. (1970) N. Engl. J. Med. 282,1179-1183 3 Rodwell, V.W., Nordstrom, Adv. Lipid Res. 14, 1-74

J.L. and Mitsche~en, J.J. (1976)

4 Kuroda, M., Hazama-Shimada, Y. and Endo, A. (1977) B&him. Biophys. Acta 486, 254-259 5 Endo, A., Kuroda, M. and Tsujita, Y. (1976) J. Antibiot. 29, 1346- 1348 6 Endo, A.. Kuroda. 72, 323-326

M. and Tanzawa,

7 Endo, A., Tsujita, Y., Kuroda, Eur. J. B&hem. 77, 31-36

K. (1976) FEBS Lett.

M. and Tanzawa,

K. (1977)

8 Serizawa, N., Nakagawa, K., Hamano, K., Tsujita, Y., Terahara, A. and Kuwano, H. (1983) J. Antibiot. 36, 604-607 9 Serizawa, N., Serizawa, S., Nakagawa, K., Furuya, K., Okazaki, T. and Terahara, A. (1983) J. Antibiot. 36,887-891 10 Serizawa, N., Nakagawa. K., Tsujita, Y., Terahara, A, and Kuwano, H. (1983) J. Antibiot. 36, 608-610

11 Serizawa, N., Nakagawa, K., Tsujita, Y., Terahara, A., Kuwano, H. and Tanaka, M. (1983) J. Antibiot. 36.918-920 12 Tsujita, Y., Kuroda, M., Tanzawa, K., lwadoh, S. and Furuya, K. (1980) Belg. Pat. 882325, Sep. 19 13 Endo, A. (1979) J. Antibiot. 32, 852-854 14 Alberts, A.W., Chen, J., Kuron. G., Hunt, V., Huff, J., Hoffman, C., Rothrock, J., Lopez, M., Joshua, H., Harris, E., Patchett, A., Monaghan, R., Currie, S., Stapley, E., Albers-Schonberg, G., Hensens, 0.. Hirshfield, J., Hoogsteen, K., Liesch, J. and Springer, J. (1980) Proc. Natl. Acad. Sci. USA 77, 3957-3961 15 Kaneko, I., Hazama-Shimada, Y. and Endo, A. (1978) Eur. J. B&hem. 87, 313-321 16 Mold&q P., Hogbuer, J. and Orrenius, S. (1978) Methods Enzymol. 52, 60-71 17 Kuroda, M., Tanzawa, K., Tsujita, Y. and Endo, A. (1977) Biochim. Biophys. Acta 489, 119-l 25 18 Endo, A., Tsujita, Y., Kuroda, M. and Tanzawa, K. (1979) Biochim. Biophys. Acta 575, 266-276 19 Zak, B. (1957) Am. J. Clin. Pathol. 27, 5X3-588 20 Henly, A.A. (1957) Analyst 82, 286-287 21 Hatch, F.T. and Lees, R.S. (1968) Adv. Lipid Res. 6, 1-68 22 Tsujita, Y., Kuroda, M., Tanzawa, K.. Kitano. N. and Endo, A. (1979) Atherosclerosis 32, 307-313 23 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275 24 Tanzawa, K., Shimada, Y., Kuroda, M., Tsujita, Y., Arai, M. and Watanabe, Y. (1980) FEBS Lett. 118, 81-84 25 Shimada, Y., Tanzawa, K., Kuroda, M., Tsujita, Y., Arai, M. and Watanabe, Y. (1981) Eur. J. B&hem. 118,557-564 26 Ando, S., Kon, K., Tanaka, Y., Nagase, S. and Nagai, Y. (1980) J. B&hem. 87, 1859-1862 27 Watanabe, Y., Ito, T., Saeki, M., Kuroda, M., Tanzawa, K., Mochizuki, M., Tsujita, Y. and Arai, M. (1981) Atherosclerosis 38, 27-31 P.J. (1975) Can. J. Physiol. 28 Savoie, L.L. and Lupien, Pharmacoi. 53, 638-643 29 Spady, D.K. and Dietschy, J.M. (1983) J. Lipid Res. 24, 305-315 30 Kovanen, P.T., Bilheimer, D.W., Goldstein, J.L., Jaramillo, J.J. and Brown, MS. (1981) Proc. Natl. Acad. Sci. USA 78, 1194-1198 31 Bilheimer, D.W., Goldstein, J.L., Grundy, SM. Starzl, T.E. and Brown, M.S. (1984) N. Engl. J. Med. 311, 165881664 32 Turley, S.D., Andersen, J.M. and Dietschy, J.M. (1981) J. Lipid Res. 22, 551-569 P.T. and Goldstein, J.L. (1981) 33 Brown. M.S., Kovanen, Science 212, 628-635 34 Watanabe, Y. and Ito, T. (1978) Med. Biol. 96, 197-201 Y. and Goldstein, J.L. 3.5 Kita, T., Brown, MS., Watanabe, (1981) Proc. Natl, Acad. Sci. USA 78, 2268-2272 M., Carew. T.E., Pittmann, R.C., Witztum, 36 Naruszewicz, J.L. and Steinberg, D. (1984) J. Lipid Res. 25, 1206-1213