Amiodarone inhibits arterial thrombus formation and tissue factor translation.

BACKGROUND: In patients with coronary artery disease and reduced ejection fraction, amiodarone reduces mortality by decreasing sudden death. Because the latter may be triggered by coronary artery thrombosis as much as ventricular arrhythmias, amiodarone might interfere with tissue factor (TF) expression and thrombus formation. METHODS AND RESULTS: Clinically relevant plasma concentrations of amiodarone reduced TF activity and impaired carotid artery thrombus formation in a mouse photochemical injury model in vivo. PTT, aPTT, and tail bleeding time were not affected; platelet number was slightly decreased. In human endothelial and vascular smooth muscle cells, amiodarone inhibited tumor necrosis factor (TNF)-alpha and thrombin-induced TF expression as well as surface activity. Amiodarone lacking iodine and the main metabolite of amiodarone, N-monodesethylamiodarone, inhibited TF expression. Amiodarone did not affect mitogen-activated protein kinase activation, TF mRNA expression, and TF protein degradation. Metabolic labeling confirmed that amiodarone inhibited TF protein translation. CONCLUSIONS: Amiodarone impairs thrombus formation in vivo; in line with this, it inhibits TF protein expression and surface activity in human vascular cells. These pleiotropic actions occur within the range of amiodarone concentrations measured in patients, and thus may account at least in part for its beneficial effects in patients with coronary artery disease.

Interaction with the hERG channel and cytotoxicity of amiodarone and amiodarone analogues.

BACKGROUND AND PURPOSE: Amiodarone (2-n-butyl-3-[3,5 diiodo-4-diethylaminoethoxybenzoyl]-benzofuran, B2-O-CH(2)CH(2)-N-diethyl) is an effective class III antiarrhythmic drug demonstrating potentially life-threatening organ toxicity. The principal aim of the study was to find amiodarone analogues that retained human ether-a-go-go-related protein (hERG) channel inhibition but with reduced cytotoxicity. EXPERIMENTAL APPROACH: We synthesized amiodarone analogues with or without a positively ionizable nitrogen in the phenolic side chain. The cytotoxic properties of the compounds were evaluated using HepG2 (a hepatocyte cell line) and A549 cells (a pneumocyte line). Interactions of all compounds with the hERG channel were measured using pharmacological and in silico methods. KEY RESULTS: Compared with amiodarone, which displayed only a weak cytotoxicity, the mono- and bis-desethylated metabolites, the further degraded alcohol (B2-O-CH(2)-CH(2)-OH), the corresponding acid (B2-O-CH(2)-COOH) and, finally, the newly synthesized B2-O-CH(2)-CH(2)-N-pyrrolidine were equally or more toxic. Conversely, structural analogues such as the B2-O-CH(2)-CH(2)-N-diisopropyl and the B2-O-CH(2)-CH(2)-N-piperidine were significantly less toxic than amiodarone. Cytotoxicity was associated with a drop in the mitochondrial membrane potential, suggesting mitochondrial involvement. Pharmacological and in silico investigations concerning the interactions of these compounds with the hERG channel revealed that compounds carrying a basic nitrogen in the side chain display a much higher affinity than those lacking such a group. Specifically, B2-O-CH(2)-CH(2)-N-piperidine and B2-O-CH(2)-CH(2)-N-pyrrolidine revealed a higher affinity towards hERG channels than amiodarone. CONCLUSIONS AND IMPLICATIONS: Amiodarone analogues with better hERG channel inhibition and cytotoxicity profiles than the parent compound have been identified, demonstrating that cytotoxicity and hERG channel interaction are mechanistically distinct and separable properties of the compounds.

Comparison of the pharmacodynamic effects of intravenous and oral propafenone.

The effect of propafenone and its major metabolite 5-hydroxy-propafenone on ECG intervals was investigated in eight healthy extensive metabolizers after single oral (300 to 450 mg) and intravenous (35 to 50 mg) doses of propafenone in a single-blind randomized trial. Peak serum concentrations were 278 +/- 233 ng/ml (oral) and 295 +/- 131 ng/ml (intravenous). After oral administration peak 5-hydroxy-propafenone levels were 194 +/- 65 ng/ml, whereas after intravenous dosing no metabolite was detected, except in one subject. Serum concentrations were related to effects by linear regression including a hypothetical effect-site compartment in a pharmacokinetic-pharmacodynamic model. Significant prolongations of ECG intervals were found in both sequences. Comparison of the two concentration-effect data sets (intravenous, oral) revealed an additive effect of 5-hydroxy-propafenone in four of eight subjects for PQ interval and seven of eight subjects for QRS duration. We conclude that 5-hydroxy-propafenone exerts pharmacologic activity and could thus contribute to the antiarrhythmic effect of propafenone.

Kinetics and electrocardiographic changes after oral 3-OH-quinidine in healthy subjects.

3-OH-quinidine, a major quinidine metabolite, has been reported to have antiarrhythmic activity in animals and is suspected to contribute to the effect of quinidine in man. Four healthy subjects received 3-OH-quinidine in increasing oral doses (35, 100, 300 mg) to achieve serum concentrations in the range of those after quinidine dosing in patients. Blood and urine were collected up to 48 hours and blood pressure, heart rate, and averaged ECG complexes were recorded during 12 hours after dosing. Kinetic analysis revealed differences from published data for the parent drug. Renal clearance was 16 L/hr. The elimination t1/2 was 12.4 hours, substantially longer than that of quinidine. No systematic ECG changes were observed in two subjects with maximum concentrations of 55 and 215 micrograms/L. In the other two subjects who achieved higher maximum concentrations (447 and 918 micrograms/L), there was a significant relationship between the length of the corrected QT interval and the serum concentration of 3-OH-quinidine. These first dynamic results indicate that 3-OH-quinidine exerts effects in man resembling those of quinidine and may contribute to the antiarrhythmic activity of quinidine.

Kinetics and dynamics of quinidine-N-oxide in healthy subjects.

The pharmacokinetics of a major metabolite of quinidine in humans, quinidine-N-oxide, were investigated after single oral doses (3 to 15 mg) in four healthy subjects. The concentration in serum and urine was determined by an HPLC assay. Because of a small volume of distribution, the elimination half-life of quinidine-N-oxide was only 2.5 +/- 0.28 hours (mean +/- SD), considerably shorter than that of quinidine. The renal clearance was 1.3 +/- 0.3 L/hr. Only 13.9% +/- 3.7% of the dose was recovered in urine as unchanged compound for up to 12 hours. Two unidentified compounds with the same retention time as quinidine and 3-hydroxyquinidine were found in the urine samples of two subjects. The free fraction in serum was 3.3% +/- 0.83%. No systematic changes in heart rate-corrected QT interval were observed up to concentrations of 500 ng/ml. The results indicate that quinidine-N-oxide, in contrast to 3-hydroxyquinidine, does not possess quinidine-like pharmacologic activity.

Interaction between grapefruit juice and midazolam in humans.

OBJECTIVE: To investigate the effects of grapefruit juice on the pharmacokinetics and dynamics of midazolam. METHODS: Eight healthy male subjects participated in this open crossover study. Intravenous (5 mg) or oral (15 mg) midazolam was administered after pretreatment with water or grapefruit juice. We measured the pharmacokinetics and pharmacodynamics (reaction time, Digit Symbol Substitution Test [DSST], general impression judged by the investigators, and drug effect judged by the subjects) of midazolam and the pharmacokinetics of alpha-hydroxymidazolam. RESULTS: In comparison to water, pretreatment with grapefruit juice did not change the pharmacokinetics or pharmacodynamics of intravenous midazolam. After oral administration, pretreatment with grapefruit juice led to a 56% increase in peak plasma concentration (Cmax), a 79% increase in time to reach Cmax (tmax), and a 52% increase in the area under the plasma concentration-time curve (AUC) of midazolam, which was associated with an increase in the bioavailability from 24% +/- 3% (water) to 35% +/- 3% (Grapefruit juice; mean +/- SEM, p < 0.01) After oral administration of midazolam, pretreatment with grapefruit juice was associated with a 105% increase in tmax and with a 30% increase in the AUC of alpha-hydroxymidazolam. For oral midazolam, pretreatment with grapefruit juice led to significant increases in tmax for all dynamic parameters and in the AUC values for the reaction time and DSST, whereas the maximal dynamic effects remained unchanged. CONCLUSIONS: Pretreatment with grapefruit juice is associated with increased bioavailability and changes in the pharmacodynamics of midazolam that may be clinically important, particularly in patients with other causes for increased midazolam bioavailability such as advanced age, cirrhosis of the liver, and administration of other inhibitors of cytochrome P450.

Pharmacodynamics of 3-hydroxyquinidine alone and in combination with quinidine in healthy persons.

The relation between serum concentration of 3-hydroxyquinidine (3-OHQ), a major metabolite of quinidine in humans, and the pharmacologic effect alone and in combination with the parent drug was studied. The heart rate-corrected, computer-averaged QT interval (QTc) was used as the pharmacologic endpoint. In a randomized, double-blind study, 5 healthy subjects received, on 3 separate days 1 week apart, either (1) 300 to 400 mg 3-OHQ orally or (2) 150 mg quinidine base intravenously or (3) a combination in the same doses. Blood samples and electrocardiographic recordings were obtained over the following 10 hours. Serum concentrations of 3-OHQ and quinidine were determined by high-pressure liquid chromatography and the free fraction by ultrafiltration. Peak concentrations of 3-OHQ varied between 1,362 and 3,480 ng/ml after oral 3-OHQ ingestion, but were negligible after intravenous quinidine infusion. The free fraction was 49% +/- 4.8 (mean +/- standard deviation) for 3-OHQ and 20% +/- 4.3 for quinidine. In all 5 subjects a statistically significant correlation was found between serum concentration and QTc prolongation for both quinidine and 3-OHQ (largest p value less than 0.025). The mean slope of the regression line was 0.0184 +/- 0.0128 for 3-OHQ and 0.0297 +/- 0.0111 for quinidine. Multiple linear regression revealed in each subject a significant additive effect of 3-OHQ when administered together with quinidine (largest p value less than 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)

Concentration-effect relations of 5-hydroxypropafenone in normal subjects.

To evaluate the pharmacologic activity of 5-hydroxypropafenone, electrocardiographic changes (PQ and QRS duration) and blood pressure levels were measured in 6 healthy extensive metabolizers of debrisoquine after a single oral dose of 300 mg of this metabolite as a solution in a placebo-controlled, double-blind crossover study. Well-absorbed, with a lag time of 4.4 to 9.8 minutes, 5-hydroxypropafenone reached peak concentrations of 153 to 337 ng/ml after 20 to 51 minutes. The terminal half-life was 506 to 963 minutes. To describe the temporal aspects of the concentration-effect relation, a pharmacokinetic-pharmacodynamic model with a hypothetical effect compartment was applied. The relation between electrocardiographic changes and drug concentration at the effect site could be described by a linear regression model. Significant prolongations of PQ and QRS duration were found in 5 of 6 subjects. There were no changes in QTc interval, blood pressure measurements and heart rate in the supine position. However, blood pressure measurements in the upright position revealed a greater percent decrease of systolic blood pressure than with placebo (mean +/- standard deviation -25.6 +/- 13.8% vs -3.4 +/- 13.1%, p less than 0.05). It is concluded that 5-hydroxypropafenone exerts significant pharmacologic activity in humans as well as animals. Because QRS prolongation in patients treated with class IC antiarrhythmic drugs correlates with the antiarrhythmic effect, our data suggest that 5-hydroxypropafenone may contribute to the therapeutic activity of propafenone in humans.

Determination of piroximone in human plasma by high-performance liquid chromatography.

A rapid and selective high-performance liquid chromatographic (HPLC) method using solid-phase extraction (SPE) for measuring piroximone in plasma samples has been developed. Human plasma and internal standard were pipetted onto a Bond Elut C18 extraction cartridge conditioned with methanol and water. Impurities and proteins were washed out with water. Piroximone and internal standard were eluted with methanol. After evaporation, the residue was dissolved in the mobile phase and the aliquot injected into the HPLC system. Piroximone and its related compounds were separated on a reversed phase C18 HPLC column maintained at 50 degrees C using a mobile phase consisting of phosphate buffer and methanol. Piroximone-N-oxide, piroximone, and internal standard were detected spectrophotometrically at 340 nm. The extraction recovery for piroximone was 94%. The within- and between-run coefficients of variation were < 3% in the concentration range of 320-5,000 ng/ml and < 17.5% at lower concentrations, e.g., 20 ng/ml. The limit of detection was 10 ng/ml.

The antiarrhythmic effects of controlled release disopyramide phosphate and long acting propranolol in patients with ventricular arrhythmias.

The antiarrhythmic effect of slow-release disopyramide phosphate (DR) 300 mg twice daily and of long-acting propranolol (PR) 1 X 160 mg daily was compared in a randomized cross-over study in patients with premature ventricular beats (PVB). 12 patients with PVB (Lown Classes II-V) were given: placebo I for 3 days, DR or PR for 7 days, placebo II for 5 days and PR or DR for 7 days. During each study phase Holter-ECG recordings were taken over a period of 24 h. With DR 6 patients showed a positive qualitative effect, improving by at least one Lown class, whereas only 2 patients did so with PR. With DR reduction of PVB greater than 80% occurred in 7 patients, and with PR in 2 patients. In all patients with any reduction in PVB, the median decrease was 85% with DR and 59% with PR. The overall results suggest that the antiarrhythmic effect of disopyramide phosphate in the slow-release preparation is at least satisfactory and comparable to that of disopyramide phosphate in the standard capsule formulation given in the usual and more complicated regime of four divided doses. The antiarrhythmic effect of PR in the recommended dose as given was not convincing.

Chronic verapamil administration lowers portal pressure and improves hepatic function in rats with liver cirrhosis.

The effects of verapamil on portal pressure, microsomal liver function and extravascular albumin space were investigated in rats rendered cirrhotic by chronic exposure to phenobarbital and carbon tetrachloride. Verapamil significantly decreased splenic pulp pressure by 28% (P less than 0.05). In cirrhotic animals it improved liver function, measured by the aminopyrine and caffeine breath tests, by 36% (P less than 0.025) and 53% (P less than 0.05), respectively. The extravascular albumin space, an important determinant of drug clearance, was measured by a multiple indicator dilution technique. It was significantly larger in verapamil treated than in untreated cirrhotics (4.41 +/- 1.06 vs 2.73 +/- 0.79 ml/g; P less than 0.01). We conclude that verapamil has significant potential as a portal antihypertensive agent and its value in treating cirrhosis in man should be explored by controlled studies.

Pharmacokinetics of conventional and slow-release verapamil.

Verapamil is a racemic mixture of two optical isomers, the (-)-form being the more active component. Recent studies indicate a rapid hepatic transformation of (-)-verapamil, which results in different concentration-effect relationships after oral and intravenous administration. In practice the important pharmacokinetic properties of verapamil are low bioavailability (20%), predominant elimination by metabolism (greater than 95%) and a relatively short half-life (t1/2, beta is 3-5 h). After repeated dosing, the rate of hepatic drug clearance seems to decrease. Slow release (SR) formulations of verapamil may offer certain therapeutic advantages during long-term treatment. A comparison of conventional (C) and SR tablets in a 1-week treatment of eight cardiac patients showed a relative bioavailability (AUCSR/AUCC) of 90 +/- 30%. More stable serum drug levels were maintained by 12-hourly administration of SR verapamil. A further study using a new 240 mg SR preparation in patients with arterial hypertension showed that even a single daily dose can be sufficient for adequate blood pressure control over 24 h.

Biotransformation of caffeine by cDNA-expressed human cytochromes P-450.

OBJECTIVES: The biotransformation of caffeine has been studied in vitro using human cytochrome P-450 isoenzymes (CYPs) expressed in human B-lymphoblastoid cell lines, namely CYP1A1, 1A2, 2A6, 2B6, 2D6-Val, 2E1 and 3A4, and microsomal epoxide hydroxylase (EH). In addition, CYP2D6-Met was also studied, in which a valine in the wild type (CYP2D6-Val) has been replaced by a methionine due to a G to A mutation in position 112. RESULTS: At caffeine 3 mmol center dot l-1, five CYPs (1A1, 1A2, 2D6-Met, 2E1 and 3A4) catalysed the biotransformation of caffeine. Among the enzymes studied, CYP1A2, which predominantly catalysed paraxanthine formation, had the highest intrinsic clearance (160 l center dot h-1 center dot mmol-1 CYP). Together with its high abundance in liver, it should be considered, therefore, to be the most important isoenzyme in caffeine metabolism. The affinity of caffeine for CYP1A1 was comparable to that of its homologue 1A2. CYP2D6-Met, which catalysed caffeine metabolism by demethylation and 8-hydroxylation, also had a relatively high intrinsic clearance (3.0 l center dot h-1 mmol-1 CYP), in particular for theophylline and paraxanthine formation, with kM values between 9-16 mmol center dot l-1. In contrast, the wild type, CYP2D6-Val, had no detectable activity. In comparison, CYP2E1 played a less important role in in vitro caffeine metabolism. CYP3A4 predominantly catalysed 8-hydroxylation with a kM value of 46 mmol center dot l-1 and an intrinsic clearance of 0.60 l center dot h-1 center dot mmol-1 CYP. Due to its high abundance in human liver, the latter CYP may contribute significantly to the in vivo formation of TMU. CONCLUSION: The findings of this study indicate that i) microsomes from transfected human B-lymphoblastoid cell lines give results close to those obtained with microsomes isolated from human liver, ii) at least four CYP isoforms are involved in caffeine metabolism, iii) at a substrate concentration <0.1 mmol center dot l-1, CYP1A2 and 1A1 are the most important isoenzymes, iv) at higher concentrations the participation of other isoenzymes, in particular CYP3A4, 2E1 and possibly also CYP2D6-Met, are important in caffeine metabolism, and v) the nucleotide composition at position 1120 of CYP2D6 determines the activity of this isoenzyme in caffeine metabolism.

Determination of saquinavir in human plasma by high-performance liquid chromatography.

We developed and characterized a high-performance liquid chromatography (HPLC) assay for the determination of saquinavir, an HIV protease inhibitor, in human plasma samples. Extraction of plasma samples with diethyl ether resulted in quantitative recovery of both saquinavir and its stereoisomer Ro 31-8533 which was used as an internal standard. The assay was performed isocratically using 5 mM H2SO4 (pH 3.5) and acetonitrile (75.5:24.5, v/v) containing 10 mM tetrabutylammonium hydrogen sulfate (TBA) as a mobile phase, a Nucleosil 3C8 column kept at 45 degrees C and UV detection at 240 nm. Using this method, saquinavir and Ro 31-8533 can be separated from endogenous substances, and in the concentration range of 5-110 ng/ml the relative standard deviations for the determination of saquinavir were below 5%. The detection limit of saquinavir in human plasma was 1 ng/ml. The usefulness of the method was demonstrated by quantification of saquinavir in plasma of human subjects treated with 600 mg of saquinavir per os or 12 mg intravenously.

Evaluation of a rapid ultrafiltration technique for determination of quinidine protein binding and comparison with equilibrium dialysis.

The free level ultrafiltration (UF) assay by the enzyme multiplied immunoassay technique (EMIT) for determination of unbound quinidine concentration in serum (Qf) was evaluated in 50 samples obtained from cardiac patients treated with quinidine for ventricular arrhythmias. Equilibrium dialysis (ED) at 37 degrees C and high performance liquid chromatography (HPLC) served as standard methods for comparison. A good agreement was found between EMIT and HPLC at the low range of free quinidine concentration (0.1-0.7 mg/L) observed in our patients (r = 0.959). Although the correlation between UF and ED was high (r = 0.972), Qf was systematically underestimated by UF. This bias was due to the fact that UF was performed according to the recommendations of the manufacturer at 25 degrees C. No systematic differences were found when 20 additional samples were assayed by the two methods at the same temperature (25 degrees C; r = 0.992). The quinidine binding ratio showed a correlation with the serum concentration of alpha 1-acid-glycoprotein (r = 0.61). The metabolites 3(S)-hydroxyquinidine and quinidine-N-oxide did not influence the protein binding of the parent drug. The importance of adjusting the serum pH to physiological values before measurement of Qf was confirmed in this study. Our results show that, if performed under the same conditions, ED and UF yield practically identical values. Because of easy handling, the EMIT Free Level System II should be applicable under clinical conditions.

The relative potency of major metabolites and enantiomers of propafenone in an experimental reperfusion arrhythmia model.

We used isolated rat hearts subjected to coronary artery ligation and reperfusion to study the antiarrhythmic activity of 5-hydroxypropafenone (5OHP) and N-depropylpropafenone (NDPP), major metabolites of propafenone (P) in humans, and of the two enantiomers (R)- and (S)-propafenone. 5OHP suppressed reperfusion arrhythmias similar to the parent drug in a concentration-dependent manner. The concentration of 5OHP needed to prevent ventricular fibrillation in 50% of experiments (EC50) was significantly higher than that of P (0.186 +/- 0.05 vs. 0.153 +/- 0.005 mg/L, mean +/- SEM, p less than 0.05). 5OHP had a relative potency of 80% compared to P. When 5OHP and P were administered together, their antiarrhythmic effect appeared to be supra-additive. The NDPP metabolite showed very little antiarrhythmic potency and was about four times less active than P. The two enantiomers (R) and (S) were equipotent and showed antiarrhythmic activities similar to racemic P.

Method for stable measurement of the electrocardiogram in isolated guinea pig heart. Evaluation of the RR-QT relationship and the effect of quinidine.

A method for stable measurement of ECG in an excised guinea pig heart was presented. The effects of quinidine and its major metabolite, 3-OH quinidine, on QT interval were compared using excised guinea pig hearts with A-V nodal rhythm paced at a constant frequency (190 beats/min). The relative potency of 3-OH quinidine to its parent drug was about one-fourth. Taking into account the fact that the free fraction of 3-OH quinidine in blood is larger than quinidine, this metabolite may contribute to the effects of quinidine in man. Changing the frequency of electrical pacing in excised hearts with A-V nodal rhythm, the RR-QT relation was determined to be rather linear, and this relation became more curvilinear after addition of quinidine (2 mg/liter), suggesting that the Bazett's formula for correcting QT interval with the heart rate is not applicable to guinea pig heart and/or after an addition of drug.

Pharmacokinetics and metabolism of quinidine in extensive and poor metabolisers of sparteine.

The pharmacokinetics and metabolism of quinidine were investigated in extensive and poor metabolisers of sparteine. No differences in plasma clearance, terminal half life, volume of distribution or cumulative urinary excretion of quinidine, 3-hydroxyquinidine and quinidine-N-oxide were observed between phenotypes. Thus, it is unlikely that quinidine metabolism is controlled by the sparteine/debrisoquine gene locus.

Population pharmacokinetics of quinidine.

1. Population pharmacokinetic parameters of quinidine were determined based on 260 serum drug concentration measurements in 60 patients treated for arrhythmias with quinidine sulphate or quinidine bisulphate (Kinidin duriles) orally. 2. Quinidine kinetics were best described by a two compartment model with zero order absorption from the gastrointestinal tract. The pharmacokinetics are influenced by severe heart or liver failure and renal function impairment. No effect was found for mild or moderate heart failure, for age, for body weight or for coadministration of nifedipine. 3. Population pharmacokinetic parameters of quinidine (assuming 100% bioavailability of oral quinidine sulphate) were: nonrenal clearance for patients without severe heart and liver failure 12.6 l h-1, reduction in patients with severe heart or liver failure to 6.8 l h-1, renal clearance (l h-1) related to creatinine clearance (ml min-1), proportionality constant 0.0566, volume of distribution of the central compartment 161 l, maximum serum drug concentration 1.4 h after administration of quinidine sulphate and 6.0 h after administration of quinidine bisulphate. 4. The results were validated by predicting the serum drug concentration in a separate group of 30 patients. The model reliably predicted both the population average and the variability of the serum concentration of quinidine. 5. Using Monte Carlo computer simulations, an a priori dosing regimen was derived that should maximize the proportion of patients having quinidine serum concentrations within the recommended range (2-5 mg l-1): initial dose of 600 mg quinidine sulphate in all patients, 3 h later first maintenance dose of quinidine bisulphate.(ABSTRACT TRUNCATED AT 250 WORDS)

Toxicity of amiodarone and amiodarone analogues on isolated rat liver mitochondria.

BACKGROUND: Amiodarone is a well-known mitochondrial toxin consisting of a benzofuran ring (ring A) coupled to a p-OH-benzene structure substituted with 2 iodines and a diethyl-ethanolamine side chain (ring B). AIM: To find out which part of amiodarone is responsible for mitochondrial toxicity. METHODS: Amiodarone, ring A and B without the ethanolamine side-chain and iodines (B0), ring A and B with iodines but no ethanolamine (B2), ring B with 1 iodine and no ethanolamine (C1) and ring B with ethanolamine and 2 iodines (D2) were studied. RESULTS: In freshly isolated rat liver mitochondria, amiodarone inhibited state 3 glutamate and palmitoyl-CoA oxidation and decreased the respiratory control ratios. B0 and B2 were more potent inhibitors than amiodarone and B2 more potent than B0. C1 and D2 showed no significant mitochondrial toxicity. After disruption, mitochondrial oxidases and complexes of the electron transport chain were inhibited by amiodarone, B0 and B2, whereas C1 and D2 revealed no inhibition. Beta-oxidation showed a strong inhibition by amiodarone, B0 and B2 but not by C1 or D2. Ketogenesis was almost unaffected. CONCLUSIONS: Amiodarone, B0 and B2 are uncouplers of oxidative phosphorylation, and inhibit complexes I, II and III, and beta-oxidation. The benzofuran structure is responsible for mitochondrial toxicity of amiodarone and the presence of iodine is not essential.