Characterization of a tissue factor/factor VIIa-dependent model of thrombosis in hypercholesterolemic rabbits.

Tissue factor (TF) expressed in arterial atherosclerotic plaque plays a key role in activating the extrinsic coagulation pathway and triggering acute coronary syndromes. In this study, we developed and characterized a TF-factor (F)VIIa-mediated thrombosis model in rabbits. Balloon catheter-induced endothelial denudation in the femoral artery and a 4-week high cholesterol diet produced a localized atherosclerotic plaque at the injured site. High levels of TF mRNA and TF protein antigen (152 +/- 25 vs. 49 +/- 12 pg mg-1 protein in normal vessels) were detected in these atherosclerotic plaques. Plasma FVII coagulant activity (FVII:C) was significantly increased in the hypercholesterolemic rabbits (36 +/- 1 s) compared with the normal rabbits (44 +/- 1 s, P < 0.0001). Plaque rupture was induced by balloon angioplasty, which resulted in thrombus formation in the injured vessel segment after a brief period of stasis. FVIIai, a specific TF-FVIIa inhibitor, was administered intravenously to rabbits before plaque rupture at 0.3 and 1.0 mg kg-1. FVIIai dose-dependently reduced thrombus mass (14.7 +/- 2.5 and 5.9 +/- 2.2 mg, respectively, vs. 21.6 +/- 1.9 mg in the control group). PD198961, a novel factor Xa inhibitor, and argatroban, a thrombin inhibitor, also dose-dependently inhibited thrombosis. These results indicate that thrombus formation in this model is initiated by the activation of TF-FVIIa pathway, which is attributed to TF expression in the atherosclerotic plaque and enhanced plasma FVII coagulant activity. This model may be useful for evaluating in vivo efficacy of new antithrombotic drugs, particularly TF-FVIIa inhibitors.

Antithrombotic effect of LB-30057 (CI-1028), a new synthetic thrombin inhibitor, in a rabbit model of thrombosis: comparison with inogatran.

LB-30057 (CI-1028) is a novel, orally bioavailable, direct thrombin inhibitor with a Ki of 0.38 nM against human thrombin. The effects of LB-30057 on thrombus formation and hemostasis were evaluated in a veno-venous shunt model of thrombosis in rabbits, and compared with inogatran, another direct inhibitor of thrombin. Each compound was studied at 5 or 6 different doses with 5 or 6 rabbits in each group. After administration as a bolus i.v. injection followed by continuous infusion, both LB-30057 and inogatran dose-dependently inhibited thrombus formation, which was measured as an increase in time to occlusion (TTO) and a decrease in thrombus weight. Both compounds also improved vena caval blood flow and reduced the overall incidence of thrombotic occlusion. LB-30057 significantly prolonged TTO from 23 +/- 4 min (before dose) to 110 +/- 10 min at the highest dose (0.7 mg/kg + 47 microg/kg/min) (p < 0.001), and reduced thrombus weight from 57 +/- 2 mg to 15 +/- 5 mg (p < 0.001). Occlusive thrombus formed in only one of six rabbits that received the highest dose of LB-30057 (vs. 13/13 in the control group, p < 0.01). At the dose that produced the maximum antithrombotic effect (0.7 mg/kg + 47 microg/kg/min), LB-30057 increased aPTT and bleeding time approximately 2-and 2.5-fold above baseline, respectively. On a gravimetric basis, LB-30057 and inogatran displayed comparable in vivo antithrombotic efficacy. When compared to equally effective anti thrombotic doses of inogatran, LB-30057 caused less prolongation in aPTT, had no effect on PT, and tended to have less of effect on bleeding time. These results indicate that LB-30057 is an effective antithrombotic compound and it appears to have a better benefit/risk profile than inogatran in this experimental model.

The antithrombotic effects of CI-1028, an orally bioavailable direct thrombin inhibitor, in a canine model of venous and arterial thrombosis.

Direct thrombin inhibitors represent a new class of drug that may offer a therapeutic alternative for the treatment and prevention of thrombembolic conditions, especially on the venous side of the systemic circulation. CI-1028 (PD 172524/LB30057) is a potent, highly selective inhibitor of thrombin that is orally bioavailable. The efficacy of this compound has been demonstrated in animal models in which intra-venous administration was used. The objective of this study was to evaluate the efficacy of CI-1028 after oral administration in a canine electrolytic injury model of venous and arterial thrombosis. CI-1028 was administered via oral gavage, and animals received either saline or 10, 15, 20, or 30 mg/kg of drug. Fifteen minutes later, the dogs were anesthetized and a femoral artery and vein were exposed and instrumented to induce electrolytic injury and thrombosis while continuously monitoring blood flow in the vessels. Maximum blood CI-1028 concentrations of 0.88+/-0.27, 1.8+/-0.3, 2.2+/-0.5, and 3.2+/-0.5 microg/mL were generally achieved 15 to 30 minutes after administering the compound in the 10-, 15-, 20-, and 30-mg/kg groups, respectively. Administration of CI-1028 increased the time to occlusion (TTO), the principal efficacy end point, in a dose-dependent manner in both arteries and veins. The TTO in the control group (n=8) averaged 66+/-11 minutes in the arteries and 69+/-6 minutes in the veins. In dogs treated with 10 mg/kg (n=8), the TTO was not significantly different from that of the control group. In the 15-mg/kg group (n=9) TTO averaged 140+/-27 minutes in the arteries (p=not significant) and 125+/-15 minutes (p<0.05) in the veins. In the 20-mg/kg group (n=8), TTO was significantly longer than controls in both types of vessels, averaging 168+/-30 minutes in the arteries (p=0.05) and 155+/-21 minutes (p<0.05) in the veins. Likewise, at 30 mg/kg (n=8) both the arterial (179+/-17 minutes) and venous (188+/-15 minutes) TTO was significantly prolonged compared with controls. Surgical blood loss and template bleeding times tended to increase in a dose-dependent manner but a statistically significant elevation was detected for template bleeding time only at the highest dose. Dramatic changes in thrombin time were detected, consistent with the CI-1028 mechanism of action. Virtually no changes were detected in prothrombin time. Maximum activated partial thromboplastin time (aPTT) and activated clotting time changes were detected approximately 30 minutes after dosing, and they were approximately twofold and fivefold baseline values, respectively, at the highest dose. In conclusion, these results demonstrate that CI-1028 provides dose-dependent antithrombotic efficacy after oral administration in a canine model of venous and arterial thrombosis.

CI-1011 lowers lipoprotein(a) and plasma cholesterol concentrations in chow-fed cynomolgus monkeys.

Lipoprotein(a) (Lp(a)), which is generated through the covalent association of apolipoprotein(a) (apo(a)) and apo B-100-LDL, is an independent risk factor for several vascular diseases. Therefore, there is interest in developing therapies for lowering Lp(a). This investigation was carried out to determine the effect of CI-1011, a potent lipid regulator in rodents, on Lp(a) and other lipid parameters in cynomolgus monkeys (Macaca fascicularis). Nine healthy male monkeys on a normal chow diet were orally treated with CI-1011 at 30 mg/kg per day for 3 weeks. Lp(a) and total cholesterol levels were significantly decreased after 1 week and maximally reduced to 68 and 73% of control levels, respectively, after 3 treatment weeks. The decreases in total cholesterol were mainly due to changes in low density lipoprotein (LDL). The LDL:HDL ratio decreased by 30%. Triglycerides were unaffected by treatment. Lp(a) and total cholesterol levels returned to pretreatment values after stopping treatment suggesting a direct effect of the compound on their inhibition. Further studies demonstrated that CI-1011 was effective at a low dose of 3 mg/kg per day after 1 week of administration. CI-1011 also decreased apo B-100 to 80% of control levels, but this change was not sufficient to account for the Lp(a) lowering. There was also no correlation between the changes in Lp(a) and apo B-100 levels. Treatment of cynomolgus monkey primary hepatocyte cultures with CI-1011 resulted in a dose-dependent inhibition of Lp(a) levels suggesting a direct hepatic effect of the compound. Western blot analysis of the samples showed that changes in Lp(a) were associated mainly with decreased apo(a) (47%), but not apo B-100 (17%). These results demonstrate that CI-1011 effectively decreases Lp(a) levels both in vivo and in vitro.

alpha-Substituted malonester amides: tools to define the relationship between ACAT inhibition and adrenal toxicity.

We prepared a series of alpha-substituted malonester amides that were evaluated for their ability to inhibit acyl-CoA:cholesterol O-acyl transferase activity in vitro and to lower plasma total cholesterol levels in a variety of cholesterol-fed animal models. Compounds of this series were also useful in examining the relationship between adrenal toxicity and ACAT inhibition. One compound from this series, 9f, was a potent inhibitor of ACAT in both the microsomal and cellular assays. It was also bioavailable as determined by both a bioassay and a HPLC-UV assay. This compound was evaluated in both guinea pig and dog models of adrenal toxicity and compared to tetrazole amide 15. In the most sensitive species, the dog, both of these compounds achieved good plasma levels; however, compound 9f caused adrenal necrosis, whereas compound 15 had no effect on the adrenal gland. This adds to the growing body of evidence that the adrenal toxicity observed with ACAT inhibitors may not be mechanism related.

Inhibitors of acyl-CoA:cholesterol O-acyltransferase (ACAT) as hypocholesterolemic agents: synthesis and structure-activity relationships of novel series of sulfonamides, acylphosphonamides and acylphosphoramidates.

Sulfoacetic acid, phosphoramidate, and phosphoramide analogs of the ACAT inhibitors, CI-999 and CI-1011 were synthesized. The structure-activity relationships of these compounds as ACAT inhibitors are described.

Opposite effects of bezafibrate and gemfibrozil in both normal and hypertriglyceridemic rats.

Chow and sucrose-fed rats were used as animal models to study the dose-responses of bezafibrate and gemfibrozil in normolipidemic and hypertriglyceridemic states, respectively. Although both drugs lowered plasma triglycerides (TG) to about the same extent in chow-fed rats, gemfibrozil lowered liver TG as well as plasma total and LDL-cholesterol (LDL-C), but elevated HDL-cholesterol (HDL-C) and plasma apo E concentrations. Bezafibrate produced opposite effects, namely, decreased HDL-C, apo E and liver TG, and tended to increase LDL-C. TG lowering for both drugs in chow-fed rats was not due to changes in TG secretion (production) in normal rats but was associated with enhanced LPL activity. In hypertriglyceridemic rats both drugs modestly reduced TG secretion rates about 40% at a dose producing maximal TG lowering, but again, gemfibrozil elevated and bezafibrate lowered HDL-C and apo E. Unlike gemfibrozil, bezafibrate induced the appearance of LDL-C in hypertriglyceridemic rats which was not detected in control animals, and also tended to increase rather than decrease plasma apo B levels. Finally, changes in liver TG concentration (mg/g) in hypertriglyceridemic rats were opposite for these drugs, resulting in significant drug-related differences in liver TG content (mg/organ). From these data we postulate that, although similar with regard to TG lowering activity and mechanisms thereof, gemfibrozil and bezafibrate produce fundamentally different effects on LDL, HDL and apolipoprotein metabolism (apo B and apo E) in rats which may relate to potential differential effects on reverse cholesterol transport and atherogenesis.

Heterocyclic ureas: inhibitors of acyl-CoA:cholesterol O-acyltransferase as hypocholesterolemic agents.

A series of diaryl-substituted heterocyclic ureas was prepared, and their ability to inhibit acyl-CoA: cholesterol O-acyltransferase (ACAT) in vitro and to lower plasma total cholesterol in cholesterol-fed animal models in vivo was examined. N-(2,6-Diisopropylphenyl)-N'-tetrazole or isoxazole-substituted heterocyclic ureas proved optimal. A carbon chain of 11-14 carbons substituted 1,3 with respect to the amine provided the optimal side chain. Substitution of the alkyl chain generally lowered activity. Tetrazole urea 2i dosed at 3 mg/kg lowered plasma total cholesterol (TC) 67% in an acute, cholesterol-fed (C-fed) rat model of hypercholesterolemia and 47% in C-fed dogs. Tetrazole 2i, dosed at 10 mg/kg, also lowered TC 52% and raised HDL cholesterol 113% in rats with pre-established hypercholesterolemia.

Heterocyclic amides: inhibitors of acyl-CoA:cholesterol O-acyl transferase with hypocholesterolemic activity in several species and antiatherosclerotic activity in the rabbit.

A series of heterocyclic amides were synthesized and evaluated as inhibitors of acyl-CoA: cholesterol O-acyltransferase (ACAT) in vitro and for cholesterol lowering in cholesterol-fed rats. Compounds were evaluated for cell-based macrophage ACAT inhibition, bioactivity, and adrenal toxicity. Candidates were selected for evaluation in cholesterol-fed dogs and, ultimately, the injured cholesterol-fed rabbit model of atherosclerosis. The heterocyclic amides potently inhibited rabbit liver ACAT (IC50's = 0.014-0.11 microM), and the majority of compounds significantly lowered plasma cholesterol (42-68%) in an acute cholesterol-fed rat model at 3 mg/kg. The most efficacious compounds in the rat were evaluated for bioactivity in vivo and arterial ACAT inhibition in a cell-based macrophage ACAT assay. Two highly bioactive analogs, (+/-)-2-(3-dodecylisoxazol-5-yl)-2-phenyl-N-(2,4,6-trimethoxypheny l) acetamide (13a) and (+/-)-2-(5-dodecylisoxazol-3-yl)-2-phenyl-N-(2,4,6-trimethoxypheny l) acetamide (16a), were selected for further study and were found to be nontoxic in a guinea pig model of adrenal toxicity. Compounds 13a and 16a lowered total cholesterol in the cholesterol-fed rat, rabbit, and dog models of pre-established hypercholesterolemia. Compound 13a in the injured cholesterol-fed rabbit model of atherosclerosis was effective in slowing the development of cholesteryl ester-rich thoracic aortic lesions, reducing lesion coverage by 53% at a dose of 1 mg/kg.

Inhibitors of acyl-CoA:cholesterol O-acyltransferase. synthesis and pharmacological activity of (+/-)-2-dodecyl-alpha-phenyl-N-(2,4,6-trimethoxyphenyl)-2H-tetrazole-5- acetamide and structurally related tetrazole amide derivatives.

A series of tetrazole amide derivatives of (+/-)-2-dodecyl-alpha-phenyl-N-(2,4,6-trimethoxyphenyl)-2H-tetrazole-5- acetamide (1) was prepared and evaluated for their ability to inhibit acyl-CoA: cholesterol O-acyltransferase (ACAT) in vitro and to lower plasma total cholesterol in vivo. For this series of compounds, our objective was to systematically replace substituents appended to the amide and tetrazole moieties of 1 with structurally diverse functionalities and assess the effect that these changes have on biological activity. The ensuing structure-activity relationship (SAR) studies identified aryl (7b) and heteroaryl (7f,g) replacements for 2,4,6-trimethoxyphenyl that potently inhibit liver microsomal and macrophage ACAT in vitro and exhibit good cholesterol lowering activity (56-66% decreases in plasma total cholesterol at 30 mg/kg), relative to 1, when compared in the acute rat model of hypercholesterolemia. Replacement of the alpha-phenyl moiety with electron-withdrawing substituents (13e-h), however, significantly reduced liver microsomal ACAT inhibitory activity (IC50 > 1 microM). This is in contrast to electron-donating substituents (13ij,m-q), which produce IC50 values ranging from 5 to 75 nM in the hepatic microsomal assay. For selected tetrazole amides (1, 7b, 13n,o), reversing the order of substituents appended to the 2- and 5-positions in the tetrazole ring (36a-d), in general, improved macrophage ACAT inhibitory activity and provided excellent cholesterol-lowering activity (ranging from 65% to 77% decreases in plasma total cholesterol at 30 mg/kg) in the acute rat screen. The most potent isomeric pair in this set of unsubstituted methylene derivatives (13n and 36a) caused adrenocortical cell degeneration in guinea pigs treated with these inhibitors. In contrast, adrenal glands taken from guinea pigs treated with the corresponding alpha-phenyl-substituted analogs (7b and 36c) were essentially unchanged compared to untreated controls. Subsequent evaluation of 7b and 36c in a rabbit bioassay showed that both compounds and/or their metabolities were present in plasma after oral dosing. Unlike 7b and 36c, compound 1 and related 2,4,6-trimethoxyanilides (13j, 30c,d) showed poor oral activity in the rabbit bioassay. Nevertheless, in cholesterol-fed rabbits, both systemically available (7b, 36c) and poorly absorbed inhibitors (1, 36d) were more effective in lowering plasma total cholesterol than the fatty acid amide CI-976.

Inhibitors of acyl-CoA:cholesterol O-acyltransferase. 17. Structure-activity relationships of several series of compounds derived from N-chlorosulfonyl isocyanate.

Several series of acyl-CoA:cholesterol O-acyltransferase inhibitors were prepared by the stepwise addition of nitrogen, oxygen, and sulfur nucleophiles to N-chlorosulfonyl isocyanate. The (aminosulfonyl)ureas 3-44 were the most potent inhibitors in vitro, with several compounds having IC50 values < 1 microM. Although the other series of compounds were not as potent in vitro, many compounds did display good in vivo activity in cholesterol-fed rats. Several of the oxysulfonyl carbamates (including CI-999, 115) showed excellent lipid-lowering activity in the chronic in vivo screen, demonstrating significant cholesterol lowering in a pre-established hypercholesterolemic state.

Bioisosterism in drug design: identification of and structure-activity relationships in a series of glycine anilide ACAT inhibitors.

To examine the effects of bioisosteric replacement on the biological activity of our previously disclosed disubstituted urea inhibitors of the enzyme acyl-CoA:cholesterol acyltransferase (ACAT), we prepared a series of N'-substituted and N',N'-disubstituted glycine anilides. These compounds were tested for the ability to inhibit ACAT in vitro and lower plasma total cholesterol in cholesterol-fed rats given a single high-fat, high-cholesterol meal. ACAT inhibitory potency was greatest in compounds containing 2,6-diisopropyl substituents in the anilide portion with the glycine nitrogen substituted by a 1,1-diphenylmethyl moiety. Small improvements in potency in vitro were obtained by substitution of electron donating groups in the 2-, 3- or 5-positions of the aryl rings of the 1,1-diphenylmethyl moiety, but not by substitution in the 4-position. In vitro potency was maintained, but not improved by acylation of the glycine nitrogen. Through a QSAR analysis of in vitro ACAT inhibition for this set of compounds, an equation could be derived which accounted for 85% of the variance in the dataset. An optimal clogp of 6.65 was found, comparable to that found for other series of ACAT inhibitors. In general, compounds from this series displayed inhibitory potency against ACAT in vitro and hypocholesterolemic activity in the in vivo rat model of hypercholesterolemia comparable to that found with the ureas.

Inhibitors of acyl-CoA:cholesterol O-acyltransferase. 11. Structure-activity relationships of several series of compounds derived from N-(chlorocarbonyl) isocyanate.

Five series of compounds (4-9) derived from N-(chlorocarbonyl) isocyanate have been synthesized and evaluated for their ability to inhibit the enzyme acyl-CoA:cholesterol O-acyltransferase and lower plasma cholesterol levels in cholesterol-fed rats. Structure-activity relationships indicate that the imino dicarboxylates (6 and 7) and the oxycarbonyl thiocarbamates (8) are the most potent and efficacious series. In these series, the combination of a 2,6-diisopropylphenyl group and an aliphatic alkyl group with a chain length between 6 and 14 carbon atoms gives good activity in vitro and in vivo. In addition, a hydrogen donor is required to maintain good in vitro activity, and the acidic proton on the central nitrogen in these series appears to be important for in vivo activity.

Comparison of lifibrol to other lipid-regulating agents in experimental animals.

In vitro data suggests that lifibrol lowers plasma cholesterol by inhibiting cholesterol synthesis. We report that lifibrol is far less potent in vitro and in vivo than lovastatin for inhibiting 14C-acetate incorporation into sterols. Moreover, several major differences between lifobrol and lovastatin were noted in various animal models. In contrast, lifibrol exhibited several activities in common with gemfibrozil, another phenoxy-acid-type drug. Specifically, in normal rats lifibrol, like gemfibrozil, lowered plasma non-HDL-cholesterol and triglycerides, and increased liver weight and hepatic peroxisomal marker enzyme activities. Lovastatin only lowered plasma triglycerides. In cholesterol-fed rats lifibrol and gemfibrozil lowered non-HDL-cholesterol and elevated HDL-cholesterol while lovastatin was inactive. Finally, lovastatin but not lifibrol exhibited hypocholesterolemic activity in normal guinea pigs and resin-primed dogs. Our interpretation is that these data do not support the notion that lifibrol lowers plasma cholesterol in vivo by inhibiting cholesterol synthesis.

Inhibitors of acyl-CoA:cholesterol acyltransferase (ACAT). 7. Development of a series of substituted N-phenyl-N'-[(1-phenylcyclopentyl)methyl]ureas with enhanced hypocholesterolemic activity.

We recently described our initial structure-activity relationship (SAR) studies on a series of N-phenyl-N'-aralkyl- and N-phenyl-N'-(1-phenylcycloalkyl)ureas as inhibitors of acyl-CoA: cholesterol acyltransferase (ACAT). From this series of analogs, compound 1 (PD 129337) was identified as a potent inhibitor of ACAT with an IC50 value of 17 nM. It was also shown to dose-dependently lower plasma cholesterol in cholesterol-fed rats. However, further investigation led to the suggestion that this compound was poorly absorbed, due to a lack of efficacy when administered by gavage in an aqueous vehicle. To overcome this deficiency, we continued our SAR study on this novel series of ACAT inhibitors using an acute in vivo screen in which the compounds are administered to rats in an aqueous, CMC/Tween suspension vehicle. Modification of the N'-phenyl moiety by incorporating functional groups which were amenable to forming salts and/or polar groups to reduce lipophilicity led to the identification of several inhibitors which displayed excellent efficacy employing this protocol. Overall, substitution on the phenyl ring in the ortho, meta, or para positions led to inhibitors with only a slight decrease in potency in vitro compared to the parent unsubstituted compound. Bulkier groups in the para position tended to lower the ACAT inhibitory activity in vitro. Polar groups, such as carboxyl (33,34), lowered in vitro activity significantly, suggesting that polar-ionic interactions are disfavored for the enzyme activity. From this series, compound 28 was evaluated further in secondary in vivo screens. In a chronic cholesterol-fed rat model of hypercholesterolemia, compound 28 dose-dependently reduced nonHDL cholesterol and significantly elevated HDL cholesterol. It showed significantly greater aqueous solubility than the parent compound 1. However, it was shown to cause adrenal toxicity in guinea pigs. This led us to design a series of homologs (44-51) with increased basicity and lower lipophilicity. Some of these compounds were more potent ACAT inhibitors in vitro and demonstrated excellent hypocholesterolemic activity in vivo. Interestingly, compound 45, unlike 28, did not produce adrenal toxicity in guinea pigs and demonstrated excellent lipid-modulating activity in the chronic model of preestablished dyslipidemia in rats.

Divergent pharmacologic activities of PD 132301-2 and CL 277,082, urea inhibitors of acyl-CoA:cholesterol acyltransferase.

The in vitro potencies and hypocholesterolemic properties of CL 277,082 and PD 132301-2, two urea inhibitors of acyl-CoA:cholesterol acyltransferase (ACAT) were compared. PD 132301-2 was several-fold more potent at inhibiting ACAT in microsomes from rat and rabbit tissues and in cultured cells (murine macrophages and the human HepG2 cell line). This disubstituted urea was also relatively specific for ACAT as other cholesterol esterifying enzymes (e.g., lecithin:cholesterol acyltransferase, pancreatic cholesterol ester hydrolase), as well as intestinal diglyceride synthesis (acyl-CoA:monoglyceride acyltransferase), were unaffected in vitro at relevant concentrations. In normal chow-fed rats, both compounds reduced plasma triglycerides at doses > 50 mg/kg, but only PD 132301-2 reduced plasma cholesterol. In rat and rabbit models of hypercholesterolemia the greater in vitro potency of PD 132301-2 translated into greater in vivo efficacy (i.e., ED50 values 2- to 3-fold higher for CL 277,082 in both acute and chronic rate models). Of particular note was the greater elevation of high-density lipoprotein-cholesterol and parenteral activity of PD 132301-2 compared to CL 277,082 in the chronic rat model. Inhibition of cholesterol absorption in rats was also greater with PD 132301-2. In guinea pigs, in which 77% of plasma cholesterol was transported in low-density lipoprotein, PD 132301-2 potently reduced plasma total cholesterol (lowest significant dose = 1 mg/kg) as well as plasma triglycerides. CL 277,082 only reduced cholesterol at doses > 100 mg/kg in this low-density lipoprotein model. In a canine model of hypercholesterolemia CL 277,082 was inactive at doses up to 50 mg/kg, but PD 132301-2 was active at 3 mg/kg. Moreover, efficacy in dogs with PD 132301-2 was positively correlated with plasma drug concentration, an observation not previously demonstrated for other hypolipidemic drugs. The combined data illustrate that pharmacologic activities can vary widely among ACAT inhibitors of the same general class. In addition, the unique observation of proportionality between efficacy and blood drug levels in nonrodent animal models may not only help to simplify early stages in drug development but also may help to predict or monitor a direct action of the drug on vascular disease.

Inhibitors of acyl-CoA:cholesterol acyltransferase. 5. Identification and structure-activity relationships of novel beta-ketoamides as hypocholesterolemic agents.

A study of structure-activity relationships of substituted beta-ketoamide ACAT inhibitors I and II was performed. The results of this study suggest that whereas the beta-keto group was tolerated with no loss in activity, beta-hydroxy and oxime moieties led to significantly reduced activity in vitro and in vivo. The most potent inhibitor from the acyclic series (I) (11, IC50 = 0.006 microM) contained a C-13 alkyl chain. This compound reduced plasma total cholesterol by 38% and 66% at 3 and 30 mg/kg, respectively, in cholesterol-fed rats. Dimethylation alpha to the anilide core (5) and subsequent N-methylation of the amide NH (6) decreased in vitro potency significantly. It was also found that high potency was retained with inhibitors which incorporated the carbonyl into a lactam ring (II).

In vivo evidence that the lipid-regulating activity of the ACAT inhibitor CI-976 in rats is due to inhibition of both intestinal and liver ACAT.

CI-976, a new trimethoxy fatty acid anilide, is a potent and specific inhibitor of liver and intestinal acyl coenzyme A:cholesterol acyltransferase (ACAT) in vitro. Several in vivo approaches were used to determine the efficacy and sites of action of this compound in rats. CI-976 decreased non-high density lipoprotein (HDL)-cholesterol and increased HDL-cholesterol in rats with pre-established dyslipidemia. High performance gel chromatographic separation of plasma lipoproteins also revealed that CI-976, but not CL 277,082, lowered low density lipoprotein (LDL)-cholesterol and elevated HDL-cholesterol. Bay o 2752, octimibate, melinamide, and SaH 58-035 were all less potent in vivo compared to CI-976 and CL 277,082, and CI-976 produced the greatest decrease in liver cholesteryl esters. Subcutaneous (SC) administration of CI-976 was also efficacious in cholesterol-fed animals. In sucrose-fed rats, oral and SC CI-976 administration potently lowered plasma triglycerides. Hepatic cholesteryl ester accumulation in the ethinyl estradiol-treated rat was also diminished by orally administered CI-976. ACAT activity and cholesteryl ester mass were dose-dependently decreased in the livers from cholesterol-fed rats treated with CI-976, suggesting a direct effect on the liver. In both hypercholesterolemic and hypertriglyceridemic models, CI-976 also decreased plasma apoB concentrations. In other experiments radiolabeled CI-976 accumulated in the liver after multiple doses. Time-dependent changes in biliary lipid and bile acid secretion suggested that free cholesterol did not accumulate in the liver but instead was excreted as such or as bile acid. Finally, inhibition of endogenous and exogenous intestinal cholesterol absorption was demonstrated using several in vivo techniques. The combined data strongly supports the hypothesis that orally administered CI-976 inhibits both intestinal and hepatic ACAT, and that both of these enzymes may be determinants of plasma lipid concentrations in the rat.