Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Amiodarone (AMI¹) {2-butyl-3-[4-(3-diethylamino-1-oxapropyl)-3,5-diiodobenzoyl]-benzofuran} is a potent antiarrhythmic and antianginal drug. In humans, mono-N-desethylamiodarone (MDEA) is the only known metabolite of AMI (Flanagan et al., 1982), and cytochrome P450 3A isoforms are involved in this dealkylation (Ha et al., 1992, 1996; Fabre et al., 1993). During long-term therapy, the plasma level of MDEA is comparable with that of AMI, and this metabolite may contribute to the therapeutic effect of AMI.

Metabolically, it is possible that MDEA is further dealkylated to di-N-desethylamiodarone. The latter has been positively identified in plasma and myocardial tissue of dogs but not in the plasma of humans (Latini et al., 1984). In previous studies, Storey et al. (1969) reported that MDEA concentration in serum and organ tissue of rabbits receiving AMI (peritoneally 40 mg/kg/day for 4 weeks) was very low. Kannan et al. (1985) analyzed the bile of rabbit receiving AMI (20 mg/kg/day for 6 weeks) and found two chromatographic signals other than those corresponding to AMI and MDEA. The identity of these compounds was not investigated. These data suggested strongly that 1) other metabolic pathways of AMI may exist; 2) AMI metabolism is species-dependent; and 3) in rabbits, the clearance of AMI is particularly high, and in Storey's experiments, MDEA is further biotransformed to other unknown products.

These data seem to be in agreement with our own recent observations. Indeed, when AMI was incubated with mammal (e.g., rabbit, rat) liver microsomes in the presence of NADPH, the drug was biotransformed to MDEA and more than three other high-performance liquid chromatography (HPLC)-detectable compounds (submitted for publication). Moreover, the peak height of one signal was directly related to the MDEA concentration in the incubation, suggesting that MDEA was its precursor. Thus, MDEA was selected as the substrate since it provided the most efficient means of generating, isolating, and characterizing the structure of this material. Finally, formation of this new compound was investigated both in vitro and in vivo in the rat.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Chemicals. MDEA was a gift from SANOFI Research Center (Montpellier, France). The chemicals, such as DL-isocitric acid, NADPH, MgCl₂·6H₂O, Na₂HPO₄, and isocitric dehydrogenase (EC 1.1.1.42) from porcine heart, were purchased from Sigma and Fluka (Buchs, Switzerland). All solvents used for solid phase extraction, HPLC, and HPLC-ESI-MS were of HPLC grade (Merck, Darmstadt, Germany).

Biological Materials. Rat liver microsomes were prepared as described by Meier et al. (1981). Rabbit liver microsome fractions were isolated from untreated male New Zealand White rabbits (3-4 kg) using the same method. Protein concentrations were measured using the method of Bensadoun and Weinstein (1976).

Animals and Application of Amiodarone. Two groups of four male Sprague-Dawley rats, weighing 210 ± 20 g, were used to study the detection of new metabolites of AMI after i.p. administration of the drug. The animals were housed in cages and had free access to standard diet and tap water. For 5 days, the commercially available i.v. injection solution (containing 50 mg of amiodarone hydrochloride, 20 mg of benzyl alcohol, and 100 mg of polysorbate 80 per milliliter of solution) was injected daily into four rats (100 mg/kg i.p.). The rats (n = 4) of the control group received (also via i.p. injection) distilled water mixed with 20 mg of benzyl alcohol and 100 mg of polysorbate 80 per milliliter.

Biosynthesis of the Biotransformed Products of MDEA. Empirical conditions were used to generate the MDEA biodegraded products. Rabbit liver microsomes (8 mg of protein) were suspended in 20 ml of 0.1 M sodium phosphate buffer, pH 7.4 (PB7.4). An NADPH-regenerating system (19 ml) consisting of 150 mg of DL-isocitric acid, 0.8 ml of isocitric dehydrogenase, and 0.1 g of MgCl₂·6H₂O was added. MDEA hydrochloride (0.25 mg in 0.2 ml of ethanol) was added, and the mixture was maintained at 37°C for 2 min. The reaction was then started by addition of 80 mg of NADPH dissolved in 1 ml of ice-cooled PB7.4. To monitor the reaction, every 15 min, 50 µl of the incubation was pipetted into 0.5 ml of methanol, vortexed for 30 s, and centrifuged at 1000g for 5 min. An aliquot of 50 µl was injected into the chromatograph (system A, see Instrumentation). The production of the unknown substances was reactivated after 4 h by adding 50 mg of NADP⁺ dissolved in 0.5 ml of PB7.4 and prolonging the incubation time up to 8 h. The final volume of the incubation was 40 ml.

Isolation and Purification of the Major Unknown Product. NMR analysis required at least 0.1 mg of the unknown compound; therefore, liquid extraction and solid phase extraction were used for isolation and purification of products. The incubation medium (40 ml) was centrifuged (10,000g at 4°C for 10 min). The aqueous phase was extracted three times with 20 ml of diethyl ether. The protein pellet was washed five times with 3 ml of methanol. The organic phases were separated by centrifugation, pooled, and evaporated to dryness at 22°C. The residue was dissolved in 0.5 ml of ethanol and 5 ml of n-hexane and applied to a silica cartridge (Chromafix 400 mg; Macherey-Nagel, Oensingen, Switzerland). The stationary phase was then washed successively with 10 ml of n-hexane, 35 ml of n-hexane:isopropanol:aq.NH₃ 25% (30.0:2.5:0.5 w/w), and 50 ml of n-hexane:diethyl ether:acetic acid (52.8:14.2:1 w/w). The unknown products were then eluted with three portions of 2 ml of chloroform:methanol:acetic acid:water (95.5:39.5:5.25:3 w/w). The fractions were pooled and evaporated under reduced pressure at 22°C.

Experiments Performed with Rats.

In vitro Rat liver microsomes were also able to catalyze MDEA hydroxylation. However, to produce HPLC signals comparable with those observed with rabbit liver microsomes, 2 mg/ml microsomal protein and 30 µM MDEA must be used. To characterize the unknown products generated from rat liver microsomes, the incubation medium was extracted with diethyl ether as described above and analyzed using HPLC systems A and C (see Instrumentation).

In vivo. The tissue samples (0.1 g) from liver, heart, lung, and kidney of four rats receiving AMI were minced and homogenized with 2 ml of ice-cooled methanol using a Potter apparatus (Heidolp, Kelheim, Germany). The sample was centrifuged (10,000g) at 0°C for 10 min. The supernatant was pipetted into a clean glass tube. The protein pellet was washed five times with 2 ml of methanol. The organic phases were pooled and evaporated under reduced pressure to dryness at 22°C. The residue was dissolved in 1 ml of methanol. After a brief centrifugation (5000g for 5 min), an aliquot of 25 µl was injected into the chromatograph (system A; see Instrumentation). The fractions from between 4.1 min and 4.4 min were collected, evaporated to dryness, dissolved in 0.1 ml of methanol, and analyzed by means of HPLC-ESI-MS and tandem mass-spectrometry (ESI-MS/MS) direct flow injection technique.

Instrumentation. HPLC separation of compounds was achieved using a LKB chromatograph (Bromma, Sweden) consisting of two HPLC 2150 pumps, a 2152 HPLC controller, and a 2152 UV detector set at 242 nm. Signals were managed by WinChrom software version 1.3 (Dandenong, Victoria, Australia).

HPLC system A. The analytical assay was operated under the following conditions: Nucleosil 100-5 Protect-1 250- × 4-mm column (Macherey-Nagel) maintained at 45°C; mobile phase consisting of methanol:water:aq.NH₃ 25% (300:80:0.25 w/w); flow rate of 1.2 ml/min; and pressure of 190 bar.

HPLC system B. For the semipreparative HPLC assay, system A was modified as follows: Nucleosil 100-7 C₁₈ 125- × 10-mm column maintained at 22°C; mobile phase consisting of methanol:water:aq.NH₃ 25% (300:65:0.50 w/w); flow rate of 2 ml/min; and fractions collector (LKB Superac 2211) setting at 5-s interval between 12 and 14 min. At 15 min, the flow rate was increased to 3 ml/min, and the column was washed with 12 ml of methanol containing 5% aq.NH₃ 25% and re-equilibrated with the mobile phase for 5 min.

HPLC system C. HPLC-ESI-MS experiments were performed on an HP 1100 HPLC system (Hewlett Packard, Palo Alto, CA). The assay was operated under the following conditions: HPLC column (RP-C₈ Waters Symmetry, 150 × 2 mm) maintained at 40°C; variable-wavelength detector setting at 242 nm. The same mobile phase as that used in HPLC system A was pumped through the column at a flow rate of 0.2 ml/min, generating a pressure of 70 bar. The ESI-MS detector was interfaced directly to the output of the UV.

Hydrogen-deuterium (H/D) exchange. The H/D exchange experiments were performed by dissolving MDEA and unknown product in 99.96% CD₃OD (Cambridge Isotope Laboratories, Woburn, MA) and waiting for 10 min to achieve a complete exchange of hydrogen atoms.

HPLC-ESI-MS/MS and ESI-MS/MS analysis using direct flow injection technique. The HPLC eluent or the solutions were introduced continuously through the electrospray interface at a rate of 0.24 ml/h. The MS instrument operated under the following conditions: nitrogen, 15 psi; nitrogen dry gas, 6 ml/min: dry temperature, 250°C; capillary current, 4200 V; endplate, 3700 V; capillary exit, 100 V; and skimmer 1, 35 V. The MS/MS acquisitions were performed at normal resolution (0.6 amu at the half-peak height) in the mass range from m/z 100 to 800 with a m/z 60 cut-off value. The isolation width was monoisotopic (m/z 0.7), the fragmentation cut-off set by "fast calc", and the fragmentation amplitude set at 0.9 V.

NMR. ¹H NMR spectra were recorded at 295 K on a Bruker DRX600 spectrometer equipped with a 5-mm broadband inverse probe with an actively shielded z-gradient coil. Proton and carbon 90° pulse lengths were 7 and 13.4 amu, respectively. Unknown substance (150 µg) was dissolved in 0.2 ml of CD₃OD (NMR-microtube, Shigemi, Tokyo, Japan). All spectra were referred to the solvent peaks (¹H at 3.31 ppm and ¹³C at 49.15 ppm).

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Biosynthesis and Purification of the Major Unknown Product X1. Upon incubation of MDEA with rabbit liver microsomes in the presence of the NADPH-generating system at pH 7.4, at least three unknown compounds, X1, X2, and X3, with respective retention times of 4.26, 5.67, and 6.07 min, were observed (Fig. 1). The substrate MDEA eluted at 7.23 min. The signal at 4.26 min had the highest intensity and may correspond to the major degraded product. The reaction reached steady state after 4 h. After one reactivation cycle (see Experimental Procedures), 5 to 7% of the initial MDEA present was biotransformed into unknown products, as judged by HPLC system A. 

View larger version (38K): [in this window] [in a new window]   Fig. 1.   HPLC separation of the products of MDEA oxidation by liver microsomes. The HPLC chromatogram (system A, UV 242 nm) shows that in rabbit liver microsomes (0.2 mg of protein/ml), MDEA (10 µM) was biotransformed to X1, X2, and X3 products (a). Their retention times were 4.26, 5.67, and 6.07 min, respectively. MDEA was eluted at 7.23 min. To generate an amount of X1 comparable with that observed in experiments performed with rabbit liver microsomes, many more rat liver microsomes (2 mg of protein/ml) and MDEA (30 µM) had to be used (b). The incubation time in both experiments was 15 min. c, blank control.

UV Spectrum of X1. The UV spectrum of the purified X1 in methanol showed the following characteristics: 208 nm, 242 nm (maxima), 223 nm (minimum), and 275 nm, 282 nm (shoulders) and was comparable with that of MDEA·HCl in methanol. Thus, its concentration may be calculated using the molar extinction coefficient of MDEA·HCl at 241 nm (E = 440.000 l/mol/cm) (Plomp, 1991), and the total amount of X1 was estimated to be 0.15 mg. The compound was stable (as assayed by UV spectra measurement) in methanol at 22°C or dry-stored at 20°C for at least 10 days.

MS Results. Since HPLC-ESI-MS was operated using a low flow rate (0.2 ml/min), HPLC system A was modified to system C. When the incubation medium was extracted with diethyl ether and immediately analyzed, the same chromatogram as that shown in Fig. 1 was obtained. The UV signals of X1, X2, X3, and MDEA displaced to 4.4, 5.4, 6.0, and 7.45 min, respectively (data not shown). ESI-MS spectra revealed the quasi-molecular ions ([M + H]⁺) at m/z 618 and 634 for MDEA and X1, respectively (Fig. 2, a and b). Compared with MDEA, the molecular ion of X1 showed a mass shift of +16 amu, suggesting that the MDEA molecule may have been oxygenated. The mass spectra of the minor products X2 and X3 revealed that their quasi-molecular ions were at m/z 591 and 590, respectively (submitted for publication).

View larger version (34K): [in this window] [in a new window]   Fig. 2.   ESI-MS and ESI-MS/MS spectra of MDEA and purified X1. a, ESI-MS of MDEA in methanol and in d4-methanol (insert). b, ESI-MS of X1 in methanol and in d4-methanol (insert). c, ESI-MS/MS of MDEA. d, ESI-MS/MS of X1. e, ESI-MS/MS of MDEA in d4-methanol. f, ESI-MS/MS of X1 in d4-methanol.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Fragmentation pathways leading to the fragment ions observed in Fig. 2f for [d2-X1 + D]+ at m/z 637.

NMR Spectrometric Results. NMR experiments were exclusively performed with the off-line purified X1 isolated from the rabbit liver microsome experiments. The comparison of the ¹H NMR spectra of MDEA and X1 revealed that the chemical shifts and coupling constants of the signals corresponding to protons of the 3,5-diiodo-4-ethyl-aminoethoxybenzoyl moiety and those of the benzene ring of the benzofuran group were unchanged (Table 1, the carbon atoms of the X1 molecule were arbitrarily numerated). However, the proton signals of the n-butyl group of X1 were shifted and split (Fig. 4b). The signals at 1.76 and 2.82 ppm, corresponding to protons attached to the carbon atom at the 3- and 4-positions of MDEA, respectively, were missing in the proton spectrum of X1 (Fig. 4, a and b). Additionally, in the spectrum of X1, a doublet at 1.17 ppm was newly observed, suggesting that the methyl group of n-butyl must now be adjacent to a CH group rather than a -CH₂ group. Unfortunately, due to impurities (phospholipids), the structure for X1 cannot be drawn definitively from these ¹H NMR data (Fig. 4b). Thus, a 1D-TOCSY experiment (selective excitation of the signal at 1.17 ppm) was performed to determine to which spin system the methyl group at 1.17 ppm belongs. Our results showed clearly that the multiplets at 3.70, 3.00, 2.92, and 1.94 to 1.85 ppm built up a new spin system (Fig. 4c). Moreover, the splitting of these multiplets suggested that a chiral center was present in the n-butyl side chain of the X1 molecule.

View this table: [in this window] [in a new window]   TABLE 1 1H and 13C NMR data for MDEA and X1 The carbon atoms of the molecule are arbitrarily numerated. The chemical shifts are in ppm and the coupling constant (J) is in Hz.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   1H NMR spectra (range of 0.8-4.0 ppm) of MDEA (a), X1 (b), and 1D-TOCSY of X1 (the methyl group at 1.17 ppm was selectively excited) (c). The positions of the carbon atoms (see Table 1), to which the hydrogen atoms being measured are attached, are given in parentheses. ?, impurity.

Experiments Performed With Rats.

In vitro experiments When rabbit liver microsomes (0.2 mg of protein/ml; MDEA 10 µM, NADPH 0.1 mM, V_total 0.2 ml) were replaced by those of rats (2.0 mg of protein/ml; MDEA 30 µM, NADPH 0.1 mM, V_total 0.2 ml), comparable HPLC-UV signals at 242 nm (system A) for X1 were obtained (Fig. 1). Unknown products generated from rat liver microsomes were extracted with diethyl ether and analyzed. The MS behavior of the product eluting at 4.4 min (HPLC-ESI-MS/MS) from rat experiments was identical to that of X1 obtained from rabbit microsomes. Based on the protein concentrations, MDEA hydroxylase activity of rat liver microsomes was about 10 times lower than that of rabbits. A detailed report of the quantitative aspect of these reactions is in preparation.

In vivo experiments. Methanolic extracts of organ (liver, heart, lung, and kidney) tissues obtained from rats receiving AMI were analyzed using HPLC system A. Under our HPLC conditions, AMI (retention time 8.9 min) was resolved from MDEA (retention time 7.3 min), and the UV (242 nm) signals at 4.26 and 4.84 min were found. Eluent between 4.1 min and 4.3 min was collected, concentrated, and reanalyzed by means of HPLC-ESI-MS and ESI-MS/MS using the direct flow injection mode. ESI-MS detection revealed signals corresponding to the quasi-molecular ions [M + H]⁺ at m/z 634 and 632. As estimated by peak area measurements, they were 85 and 15%, respectively. The fragmentation of the ions at m/z 634 using the ESI-MS/MS technique was identical to that of X1. These data confirmed that 3OH-MDEA was found in organ tissue of rats receiving AMI. Thus, it may be considered as a secondary metabolite of AMI, at least in rats. Twenty-four hours after the last injection, the concentrations of 3OH-MDEA in all organs studied were 3 to 8.5% of AMI concentrations (based on the integration of peak areas). Using the same techniques, the compounds eluting at 4.6 to 5.2 min showed quasi-molecular ions [M + H]⁺ at m/z between 560 and 662, and might correspond to oxygenated products of AMI (submitted for publication).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

There are probably two main reasons why data on the metabolism of AMI are scarce. First, there is a great difference in the lipophilicity between AMI and its metabolites. Second, the appropriate analytical techniques simply were not available. The available HPLC-UV assays are not selective and sensitive enough to quantify metabolites other than MDEA. While classic gas chromatography GC-MS has excellent sensitivity, it cannot be used for AMI derivatives because the diiodobenzofuran derivatives are thermolabile. In contrast, the new HPLC-ESI-MS and ESI-MS/MS seem to be the analytical methods of choice for studying AMI metabolism: the drug and its derivatives give interpretable and high intensity MS signals. Therefore, in the current studies, our first efforts were designed to find an analytical assay capable of detecting compounds more polar than the known MDEA. This was achieved by chromatographing the AMI analogs synthesized for our previous studies (Ha et al., 2000). Then, HPLC interfaced to an ESI-MS/MS was used to detect and, in combination with NMR, identify metabolites other than MDEA. However, the method was complicated by lipid contamination emanating from the liver microsomes and led to an unstable baseline for the MS detector (ion source must be cleaned after 3-5 injections) or, in the case of NMR, interference with signals in the range of 0.8 to 2.5 ppm (Fig. 4b).

In the literature, the biotransformation of MDEA (and AMI) in rabbit liver microsomes has already been investigated by Young and Mehendale (1986). The authors reported that MDEA was deiodinated to di-deiodinated MDEA. Using HPLC-ESI-MS as an analytical tool, our detector did not reveal any ions that could be related to the presence of di-deiodinated MDEA in the incubation; in place of it, the hydroxylated MDEA was found. This disagreement is understandable because, in Young's investigations, the identity of product was based exclusively on the comparison of retention time of the unknown product with that of the reference substance L32790 obtained from the AMI manufacturer.

Since rabbit liver microsomes were found to metabolize MDEA more rapidly than rat liver microsomes, they were used to produce X1. About 0.15 mg of X1 was obtained, and its structure was characterized as 3OH-MDEA. For drug metabolism investigations, the rat model is more often used than the rabbit; thus, efforts were made to demonstrate that the newly discovered 3OH-MDEA is also found in in vitro and in vivo experiments using rats. HPLC-ESI-MS and ESI-MS/MS techniques confirmed the presence of 3OH-MDEA in the rat model.

In conclusion, the previous results (Flanagan et al., 1982) and the observations in the present study suggest that, in mammals, AMI is dealkylated to MDEA. Thereafter, MDEA (probably together with AMI) is further degraded to other compounds by hydroxylation (Fig. 5). Other studies are being planned to investigate the quantitative aspect of the hydroxylation of AMI and MDEA in laboratory animals and in humans. Only after these evaluations will the importance of the hydroxylation pathway in the metabolism of AMI be known.

View larger version (12K): [in this window] [in a new window]   Fig. 5.   Metabolism pathways of amiodarone.