Introduction

Top Abstract Introduction Methods Results Discussion References

Numerous studies have demonstrated the role of elevated levels of serum cholesterol in the pathogenesis of atherosclerosis and coronary heart disease. Clinical trials have demonstrated the benefits of lowering plasma cholesterol by diet or drug intervention (The Cholesterol Facts, 1990). Bile acid sequestrants, such as cholestyramine, and HMG-CoA reductase inhibitors, such as lovastatin and pravastatin, have been used successfully to treat this condition (Grundy, 1988; Lipid Research Clinics Program, 1984). The inhibitors of HMG-CoA reductase are well tolerated, having shown little toxicity to date. However, there exists the potential for toxicity due to concomitant blockade of ubiquinone, dolichol and isoprenylated protein synthesis (Folkers et al., 1990; Willis et al., 1990). The enzyme squalene synthase catalyzes the first committed step of the de novo cholesterol biosynthetic pathway. This enzyme catalyzes the reductive dimerization of two molecules of FPP to form squalene at the final branch point of the cholesterol biosynthetic pathway. Inhibition of squalene synthase could therefore avoid any potential adverse effect associated with reduced synthesis of isoprenylated metabolites by the inhibitors of HMG-CoA reductase. Several inhibitors of squalene synthase have been identified. Some bisphosphonates that are used for the treatment of bone disorders are also potent inhibitors of squalene synthase (Amin et al., 1992). Ciosek et al. (1993) discovered other phosphonates with cholesterol-lowering effects in animals. In addition, zaragozic acids or squalestatins are potent inhibitors of squalene synthase, with K_i values of <100 pM (Baxter et al., 1992; Bergstrom et al., 1993). Unfortunately, all of these agents are poorly bioavailable after oral administration. This report describes a novel orally effective squalene synthase inhibitor, RPR 107393 {3-hydroxy-3-[4-(quinolin-6-yl)phenyl]-1-azabicyclo [2-2-2]octane dihydrochloride; fig. 1}, that is hypocholesterolemic in rats. Furthermore, the compound is more effective than lovastatin in marmosets, a small primate species with a lipoprotein profile similar to that of humans (Crook et al., 1990). Preliminary reports of this work have been presented in abstract form (Amin et al., 1995; Needle et al., 1995).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Inhibitors of HMG-CoA reductase and squalene synthase.

	Methods

Top Abstract Introduction Methods Results Discussion References

Materials. [¹⁴C]Mevalonate (NEC-679, 50 Ci/mol) and [³H]FPP (NET-1042, 20 Ci/mmol) were purchased from DuPont-New England Nuclear (Boston, MA). [¹⁴C]HMG-CoA (CFA-577, 50 Ci/mol) was purchased from Amersham, (Arlington Heights, IL). Cholesterol, squalene, NADPH, MgCl₂, potassium phosphate dibasic and cholesterol and triglyceride determination kits (procedures 352 and 339, respectively) were purchased from Sigma Chemical Co. (St. Louis, MO). Petroleum ether was purchased from Fisher Scientific Co. (Fair Lawn, NJ). KOH and Ready Safe scintillation fluid were purchased from VWR Scientific (Westchester, PA). Cholestyramine was a gift from Daniel Elliott (Bristol-Myers Squibb Company, Evansville, IN). Methyl cellulose (Methocel) was a gift from Dow Chemical Co. RPR 107393 and FPP were synthesized at the Medicinal Chemistry Department, Rhône-Poulenc Rorer. RPR 107393 is a racemix mixture (50:50, R and S). Lovastatin was purchased as Mevacor tablets that were ground up before use. Pravastatin was obtained from Dr. Richard Gregg (Squibb Institute of Medical Research). Zaragozic acid was obtained from Dr. James Bergstrom (Merck Co., Rahway, NJ).

Squalene synthase assay. This assay was performed as previously described (Amin et al., 1992). Briefly, the assay was carried out in 50 mM phosphate buffer, pH 7.4, containing 10 mM MgCl₂, 0.5 mM NADPH, rat liver microsomes (30 µg of protein), RPR 107393 (dissolved in distilled water) and substrate [³H]FPP (0.5 µM, 0.27 Ci/mmol). After a 10-min incubation at 37°C, the reaction was terminated by the addition of 1 ml of 15% KOH in ethanol. Synthesized [³H]squalene was extracted in petroleum ether and counted.

HMG-CoA reductase assay. To study the effect on HMG-CoA reductase, a partially purified rat liver microsomal enzyme preparation (70 µg of protein) was incubated with [¹⁴C]HMG-CoA (20 µM, 3.9 Ci/mol) and inhibitor for 10 min as previously described (Amin et al., 1993). The reaction product [¹⁴C]mevalonolactone was quantified.

Inhibition of de novo cholesterol synthesis in rats. Animals were housed and cared for in keeping with the standards set forth in the National Institutes of Health "Guide for the Care and Use of Laboratory Animals" (NIH publication No. 86-23, 1985). Sprague-Dawley rats weighing 60 to 70 g were given rodent diet (#5012, obtained from Purina Mills, Inc.) and kept under reverse-lighting conditions (lights on, 3:00 p.m. to 3:00 a.m.). Cholestyramine was included in the diet for 2 days before the study to stimulate liver cholesterol biosynthetic capacity. Drugs were suspended in 0.5% methyl cellulose or dissolved in saline (zaragozic acid). RPR 107393 was given p.o. by gavage, and zaragozic acid was given s.c.. After a specified time period, the animals received [¹⁴C]mevalonolactone (15 µCi/kg; 40 Ci/mol) by s.c. injection. Fifteen minutes later, the animals were killed with CO₂. The livers were removed, and 0.5 g of the liver was saponified in 2 ml of 15% KOH/ethanol overnight at 80°C. Samples were extracted with petroleum ether in alkaline conditions, and [¹⁴C]cholesterol was quantified by HPLC (see below). The aqueous phase was acidified to pH 1 and reextracted in petroleum ether to give FPP/organic acid fraction (Bergstrom et al., 1993). [¹⁴C]Organic acids were analyzed by HPLC.

HPLC analyses of sterols. Nonsaponifiable lipids were analyzed using a Waters LC Module 1 HPLC instrument according to a modified method of Panini et al. (1991). A Partisil 10 ODS-3 4.1 × 300 mm reverse-phase column (Whatman) was used. The mobile phase consisted of acetonitrile water (94:6) at 2 ml/min at 25°C. The column effluent was monitored at a wavelength of 210 with simultaneous monitoring of radioactivity with a Radiomatic Flo One Beta (Packard Series A500) using Ultima-Flo M cocktail (Packard). Cholesterol was extracted in the alkaline fraction, and diacid products of FPP were found in the subsequent acid fraction. Retention time of [¹⁴C]cholesterol was 19 min. When [¹⁴C]FPP or livers treated with zaragozic acid or RPR 107393 were extracted by this procedure, two radiolabeled products with retention times of 2.7 and 3.5 min were observed in the acid fraction. The sum of these products is presented as [¹⁴C]diacids (table 1). The extraction efficiency of internal standards [¹⁴C]cholesterol was 95%. Of total counts from internal standard [¹⁴C]FPP, 62% were recovered as diacids in the acid fraction.

Effects on serum lipids in rats after repeat administrations. Sprague-Dawley rats were housed under reverse-lighting conditions with access to food and tap water ad libitum. After 1 week of acclimation, rats with body weights of 130 to 150 g were used. Blood (200 µl) was collected from the orbital sinus with the rats under CO₂ anesthesia. Serum cholesterol levels were determined. Serum cholesterol levels were ranked, and the rank was divided into blocks. Animals in each block were then randomly assigned to treatment groups. Where indicated, rats were switched to a meal diet containing 2% cholestyramine 1 day before the initiation of drug treatments. Drugs were dissolved/suspended in 0.5% methyl cellulose (w/v) and administered to rats p.o. either once (8:00 a.m.) or twice (8:00 a.m. and 5:00 p.m.) daily. At the end of the study, blood was collected from the inferior vena cava with the rats under ether anesthesia. Serum was separated, and cholesterol and triglyceride levels were determined with enzyme assay kits. Absorbance was measured at 500 nm using a Beckman Spectrophotometer (model DU640). Serum lipoprotein fractions were separated using HDL precipitation reagent (ReagentSet, Boehringer-Mannheim Biochemicals, Indianapolis, IN) or an ultracentrifugation procedure suggested by Beckman Instruments (procedure DS-693). A TLA-100.2 rotor and Beckman TL-100 table top ultracentrifuge were used for the latter procedure to separate VLDL (d < 1.006), LDL (1.006 < d < 1.063) and HDL (d > 1.063). Cholesterol in these fractions was determined as described above. In the precipitation procedure, the LDL fraction also contained VLDL. Comparable lipoprotein values were obtained by ultracentrifugation or precipitation procedures.

Microsomal preparation to determine HMG-CoA reductase activity. The presence of drug in the liver may interfere with in vitro HMG-CoA reductase determination. To ensure adequate drug wash out from the liver of rats used in the repeat administration studies described above, rat liver collection occurred 16 to 24 hr after the last dose at middark period. Microsomes were prepared according to the method of Amin et al. (1992). During the preparation, microsomes were washed twice with buffer to further remove any drug present in the microsomal preparation. Microsomes were stored at 80°C. HMG-CoA reductase activity was determined in aliquots (30-200 µg of protein) of microsomal preparations by means of the assay described above. HMG-CoA reductase activity is presented as pmol of mevalonate formed/min/mg of microsomal protein.

Hypocholesterolemic effects of inhibitors of HMG-CoA reductase and RPR 107393 in marmosets. Male common marmosets (Callithrix jacchus) housed in small groups with access to food (CPDX primate diet) and tap water ad libitum were used. Animals were bled from the femoral vein after an overnight fast, and the blood was mixed with EDTA as an anticoagulant. Animals were assigned to different treatment groups on the basis of their plasma cholesterol value in the same manner as described above for rats. Plasma cholesterol and triglyceride concentrations were measured with a COBAS Biocentrifugal Analyzer using standard enzymatic methods. Drugs were blended with 0.5% tragacanth mucilage using an Ultra-turrax homogenizer. The resulting suspension was administered orally by stomach tube once or twice daily for 7 days at a dose volume of 2 ml/kg. Lipoproteins were separated by the precipitation method and analyzed as described for rats.

Statistical analysis. Results are presented as mean ± S.E.M. (n) or as percent inhibition compared with the vehicle-treated mean control value. The IC₅₀ or ED₅₀ values were calculated using a computer program (Tallarida and Murray, 1987). For most in vivo studies, drug-treatment values were compared with the vehicle-treatment control values using Dunnett's one-tailed t test (general linear model procedure of SAS Institute Inc., Cary, NC). For HMG-CoA reductase values (tables 2 and 3), due to a heterogeneity of variance (P < .01, according to Levene's test), observations were transformed to normal score (rank procedure of SAS) to provide a homogeneity of variance. Dunnett's t test was performed on transformed values. P < .05 was considered to be significant.

	Results

Top Abstract Introduction Methods Results Discussion References

In vitro assays. RPR 107393 inhibited rat liver microsomal squalene synthase with an IC₅₀ value of 0.8 ± 0.2 nM (n = 4). The two enantiomers of RPR 107393, R and S, were also potent inhibitors of squalene synthase with IC₅₀ values of 0.9 and 0.6 nM, respectively (n = 2). RPR 107393 was inactive (3% inhibition at 1 mM) against HMG-CoA reductase from rat liver microsomes. Lovastatin (Na-salt) was a potent inhibitor of HMG-CoA reductase with an IC₅₀ value of 2 nM but did not reduce squalene synthase activity (3% inhibition at 30 µM).

In vivo inhibition of cholesterol biosynthesis from [¹⁴C]mevalonate in rats. One hour after RPR 107393 (10 mg/kg p.o.), cholesterol biosynthesis was reduced by 92% with an approximate ED₅₀ value of 5 mg/kg (table 1). Six hours after RPR 107393 (10 mg/kg p.o.) administration, cholesterol biosynthesis was reduced by 74% (the time for 50% inhibition was ~7 hr; fig. 2). An 82% inhibition of hepatic cholesterol biosynthesis was observed 10 hr after RPR 107393 (25 mg/kg p.o.), but the effect was no longer apparent at 21 hr. Zaragozic acid was also a potent inhibitor of cholesterol biosynthesis (100% inhibition at 1 mg/kg s.c., 1 hr, table 1). Inhibition of cholesterol biosynthesis by zaragozic acid or RPR 107393 was associated with an accumulation of radiolabeled diacid products in the liver.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Time course for the inhibition of [14C]cholesterol biosynthesis by RPR 107393. Rats were given RPR 107393 p.o. at different times as indicated before termination. [14C]Mevalonate was given s.c. Fifteen minutes later, livers were collected. After extraction in petroleum ether, the amount of [14C]cholesterol in the liver was determined by HPLC. Results are presented as mean ± S.E.M. (n = 5-6).

Lipid-lowering effect of RPR 107393 in rats. In contrast to the ineffectiveness of lovastatin, RPR 107393 is a potent cholesterol-lowering agent in rats. RPR 107393 (30 mg/kg p.o. b.i.d.) lowered serum cholesterol by 35% after 2 days and by nearly 50% after 3 days of treatment (table 2). The reduction in cholesterol was greater in the VLDL and LDL fractions (66-88%) than in the HDL fraction (maximum, 35%). A similar reduction in serum cholesterol was also observed in a second study (table 3). Both lovastatin and RPR 107393 reduced serum triglycerides, although the effect was considerably greater with the squalene synthase inhibitor (tables 2 and 3).

Coadministration of cholestyramine and RPR 107393. Cholestyramine alone (2% in diet) had no effect on serum cholesterol levels; however, coadministration of cholestyramine with RPR 107393 potentiated the cholesterol-lowering effect of RPR 107393. No reduction in serum cholesterol was observed with RPR 107393 (10 mg/kg b.i.d. for 3 days, table 3), whereas RPR 107393 (three doses of 10 mg/kg p.o. at 29, 21, and 5 hr) coadministered with 2% cholestyramine in the diet reduced serum cholesterol by 49% [72 ± 2 mg/dl (n = 8) in control; 37 ± 4 mg/dl (n = 4) after RPR 107393 treatment]. The effect of cholestyramine coadministration was further studied with the R and S enantiomers of RPR 107393 (30 mg/kg p.o. for 7 days; table 4). The R enantiomer administered alone did not lower serum LDL cholesterol, whereas coadministration with cholestyramine resulted in a 30% reduction. The reductions in LDL cholesterol with the S enantiomer in the absence and the presence of cholestyramine were 33% and 61%, respectively. In summary, the presence of resin increased the cholesterol-lowering effect of RPR 107393.

Coadministration of lovastatin and RPR 107393 to rats. Because lovastatin and RPR 107393 inhibit cholesterol biosynthesis at different sites in the cholesterol biosynthetic pathway, the possibility of a synergistic effect on cholesterol lowering was tested. Coadministration of lovastatin with RPR 107393 (both at 30 mg/kg p.o. for 7 days) was no more effective than RPR 107393 alone (fig. 3).

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Coadministration of lovastatin and RPR 107393 on reduction of rat serum cholesterol. Rats were dosed p.o. once a day for 7 days as indicated. At 24 hr after the last dose, serum cholesterol was determined. Results are presented as mean ± S.E.M. (n = 6). *P < .05 compared with control values.

Comparative effects of inhibitors on liver HMG-CoA reductase activity in rats. To study whether squalene synthase inhibition affects the regulation of HMG-CoA reductase, livers were analyzed for HMG-CoA reductase activity. Even though lovastatin did not change serum cholesterol, it increased liver microsomal HMG-CoA reductase activity by 7- to 10-fold (tables 2 and 3). A large reduction in serum cholesterol with RPR 107393 (30 mg/kg) was associated with an even greater induction of HMG-CoA reductase (12-34-fold). To determine whether the induction of HMG-CoA reductase could be separated from the cholesterol-lowering effect of RPR 107393, the compound was administered for different time periods (1-4 days; table 2) or at different doses (3-30 mg/kg; table 3). The reduction in serum cholesterol with RPR 107393 was always associated with increased HMG-CoA reductase activity in the liver.

Lipid-lowering effect in marmosets. The maximum reduction in plasma cholesterol observed in marmosets with two HMG-CoA reductase inhibitors, lovastatin and pravastatin, was ~30% (50 mg/kg b.i.d. for 7 days; table 5). In five similar studies in marmosets,¹ lovastatin (10 mg/kg b.i.d. for 7 days) produced 19% to 24% reductions in plasma cholesterol level, and in one study with 30 mg/kg, a reduction of 27% was observed (data not shown). In three additional studies, no greater reductions were observed after 14 to 28 days of treatment with lovastatin (10 mg/kg b.i.d.). In the present study, a greater reduction (50%) in plasma cholesterol was observed with RPR 107393 (20 mg/kg p.o. b.i.d. for 7 days). A reduction in plasma triglycerides was observed with both RPR 107393 and lovastatin (table 5). The enantiomers of RPR 107393 (20 mg/kg q.d. for 7 days; table 6) were equipotent with respect to plasma cholesterol (27% reduction). The reduction in plasma cholesterol by the enantiomers of RPR 107393 was selectively in the LDL fraction (50%), whereas cholesterol in the HDL fraction was unchanged.

	Discussion

Top Abstract Introduction Methods Results Discussion References

RPR 107393 is a novel, selective squalene synthase inhibitor with subnanomolar potency. When administered orally, the compound inhibits cholesterol biosynthesis in rat liver with an accumulation of diacids derived from FPP, indicating that squalene synthase is the site of inhibition in vivo. RPR 107393 is an orally effective hypocholesterolemic agent in both rats and marmosets. The reduction in serum cholesterol in the rat was greater in the LDL and VLDL fractions, although a small reduction of cholesterol in the HDL fraction was observed (tables 2-4). In marmosets with a lipoprotein profile closer to that of humans than rats (Crook et al., 1990), the reduction in plasma cholesterol was observed selectively in the LDL fraction (50%) without any change in the HDL fraction. These results are in agreement with those reported for another squalene synthase inhibitor, squalestatin 1, which selectively reduced apolipoprotein B but had no effect on apolipoprotein A in marmosets (Baxter et al., 1992). This selectivity of action on the apolipoprotein B-containing lipoproteins is important because an agent that improves the atherogenic index (LDL/HDL) would be of potential value in the treatment of hyperlipidemia and atherosclerosis.

In the rat, RPR 107393 lowered LDL cholesterol by 43% to 77% in 2 to 4 days, whereas lovastatin failed to lower serum cholesterol in any lipoprotein fraction, as has been shown in many previous studies with inhibitors of HMG-CoA reductase (Ciosek et al., 1993; Endo et al., 1979; Low et al., 1992; Tsujita et al., 1986). In the marmoset, the maximum reduction in plasma cholesterol observed with lovastatin and pravastatin was 31% (table 5), whereas RPR 107393 demonstrated a greater reduction in serum cholesterol (50%). Squalestatin 1 also produced a 50% reduction in plasma cholesterol concentration in the marmoset (Baxter et al., 1992). These results suggest that inhibitors of squalene synthase may be more effective agents than inhibitors of HMG-CoA reductase for the reduction of plasma cholesterol levels. Compared with zaragozic acid, which is a poor inhibitor of cholesterol biosynthesis after oral administration (ED₅₀ of 100 mg/kg to mice, Chiang et al., 1993) and has a bioavailability of <1% in mouse (Bergstrom et al., 1995), RPR 107393 inhibited cholesterol biosynthesis in rat with an ED₅₀ of 5 mg/kg (table 1).

Lovastatin and RPR 107393 inhibit different enzymes in the cholesterol biosynthetic pathway. To evaluate the therapeutic potential of coadministering inhibitors of HMG-CoA reductase and squalene synthase, lovastatin and RPR 107393 were coadministered to rats. The treatment regimen of RPR 107393 was chosen to obtain a submaximal reduction in levels of serum cholesterol (30% reduction, fig. 3). Lovastatin did not enhance the cholesterol-lowering activity of RPR 107393. There are at least two possible explanations for a lack of further reduction in cholesterol by lovastatin. First, 30 mg/kg RPR 107393 is expected to produce almost complete inhibition of cholesterol biosynthesis (table 1), and lovastatin, which inhibits upstream than RPR 107393, would not be expected to add significantly to this effect. Second, both lovastatin and RPR 107393 cause a large induction of HMG-CoA reductase (tables 2 and 3), and thus the effect of an inhibitor of HMG-CoA reductase will be reduced. On the other hand, cholestyramine, which acts through sequestration of intestinal bile salts, was found to synergise with RPR 107393. We do not know whether cholestyramine altered the bioavailability of RPR 107393.

HMG-CoA reductase is the rate-limiting enzyme in the cholesterol biosynthetic pathway (Rodwell et al., 1976), the expression of which is regulated by sterols and nonsterol products acting at multiple levels, including transcription, translation, phosphorylation and protein degradation (Correll and Edwards, 1994; Craig et al., 1994; Goldstein and Brown, 1990). In cell culture studies, it has been shown that farnesol and its derivatives (nonsterol products of the cholesterol biosynthetic pathway before squalene synthase) increase degradation of HMG-CoA reductase (Bradfute and Simoni, 1994; Craig et al., 1994). Based on these cell culture studies, a squalene synthase inhibitor is expected to produce less induction of HMG-CoA reductase activity than the inhibitors of HMG-CoA reductase because of a build-up of farnesol analogs. On the contrary, the results presented here show that inhibition of squalene synthase with RPR 107393 resulted in an enhanced induction of HMG-CoA reductase activity (34-fold) compared with lovastatin treatment (7-10-fold). Induction of hepatic HMG-CoA reductase activity was also observed after a single dose of zaragozic acid to rats (Ness et al., 1994). A possible explanation for a greater induction of HMG-CoA reductase after RPR 107393 than lovastatin in our study is the greater reduction in serum cholesterol observed with RPR 107393. It appears that in vivo, a regulatory pool of sterol (or a biosynthetic pathway product occurring later than squalene synthase) plays the major role in the regulation of HMG-CoA reductase activity.

The dual effects of RPR 107393, (inhibition of squalene synthase and induction of HMG-CoA reductase) may result in the accumulation of metabolites of FPP. After the administration of radiolabeled mevalonate, hepatic levels of labeled dicarboxylic acids in RPR 107393-treated rats were shown in the present study, raising the possibility of toxicity with this treatment. It is known, however, that dicarboxylic acids derived from geraniol are readily excreted in urine (Asano and Yamakawa, 1950; Kuhn et al., 1936), so rapid urinary excretion as the dicarboxylic acids may be the fate of the accumulated mevalonate after RPR 107393 treatment, as already suggested for zaragozic acid by Bergstrom et al. (1993). In conclusion, RPR 107393 is a potent and orally effective cholesterol-lowering agent in the rat and marmoset. It demonstrated greater efficacy than lovastatin in both models.