Abstract
N-(2,6-dimethylphenyl)-5-methyl-3-isoxazolecarboxamide (D2624) belongs to a new series of experimental anticonvulsants related to lidocaine. This study was undertaken to understand the pharmacokinetics and metabolism of D2624 in rats and humans, with emphasis on the possible formation of 2,6-dimethylaniline (2,6-DMA). After oral administration of stable isotope-labeled parent drug to rats and GC/MS analysis of plasma samples, two metabolites were identified: D3017, which is the primary alcohol, and 2,6-DMA, formed by amide bond hydrolysis of either D2624 or D3017. In urine, three metabolites of D2624 were identified: namely D3017, 2,6-DMA, and D3270 (which is the carboxylic acid derivative of D3017). Based on plasma AUC analysis, D3017 and 2,6-DMA accounted for >90% of the dose of D2624. After oral administration, D2624 was found to be well absorbed (93%), but underwent extensive first-pass metabolism in the rat, thus resulting in 5.3% bioavailability. Rat and human liver microsomal preparations were capable of metabolizing D2624 to D3017 and 2,6-DMA. The formation of D3017 was NADPH-dependent, whereas 2,6-DMA formation was NADPH-independent and probably was catalyzed by amidase(s) enzymes. In a single-dose (25–225 mg) human volunteer study, the parent drug (D2624) was not detected in plasma at any dose, whereas 2,6-DMA was detected only at the two highest doses (150 and 225 mg). D3017 was detected after all doses of parent drug, with approximate dose proportionality in AUC and a half-life of 1.3–2.2 hr. The metabolic behavior observed in humans suggests there is a marked species difference in the oxidative and hydrolytic pathways of D2624.
D26241belongs to a new series of experimental N-aryl isoxazolecarboxamide anticonvulsants. These compounds are structurally related to the local anesthetic lidocaine, which is known to possess anticonvulsant activity (1). The anticonvulsant Drug Development Program at the National Institute of Neurological Disorders and Stroke (Bethesda, MD) found D2624 to have phenytoin-like activity in animal seizure models (2, 3) and to be devoid of the classical lidocaine-like effects, such as local anesthesia and antiarrhythmia (4). After oral administration to rats and intraperitoneal injection to mice, D2624 exhibited an ED50 for maximal electroshock seizures of 5 and 42 mg/kg, respectively (3). However, D2624 had no protective activity toward maximal Metrazol (pentylenetetrazol) seizures in mice (3). Based on its spectrum of activity in animal models, an ability to block sodium channels was suggested as a possible mechanism of action in the same fashion as the established anticonvulsants phenytoin and carbamazepine. The existence of a metabolite of D2624 with anticonvulsant and neurotoxic properties was suggested by the dependence of anticonvulsant potency in rodents on route of administration, with oral administration having a greater TD50/ED50 ratio than parenteral administration (3). As a lidocaine analog, D2624 possesses an amide bond that is susceptible to metabolic hydrolysis to yield 2,6-DMA, a known metabolite of lidocaine (5). Although D2624 was not found to be mutagenic in classical Ames and micronucleus tests, it is known that many simple aromatic amines, such as 2,6-DMA, produce a variety of species-dependent toxicities, such as nasal carcinomas and adenomas in rats (6, 7).
Based on these considerations, the objectives of this study were to:i) identify and quantify the metabolites of D2624, either in rats receiving the drug by different routes or in rat liver microsomal preparations; and ii) determine whether 2,6-DMA is formed in human liver microsomal preparations and whether it can be detected in the plasma of healthy subjects after single-dose administration.
Materials and Methods
D2624 and D3017 were synthesized using the method reported by Lepage et al. (3) and provided by Laboratoires Biocodex (Compiègne, France). [13C2]D2624 (labeled in the xylidine methyl groups; isotopic purity: 99 atom% excess) was synthesized as outlined herein and provided by Laboratoires Biocodex. A sample of D3270 was prepared by synthesis, as outlined. 2,6-DMA was obtained from Aldrich Chemical Co. (Milwaukee, WI). All other chemicals were of analytical grade and were obtained from Sigma Chemical Co. (St. Louis, MO). Solvents were of HPLC grade and were purchased from either Baxter Healthcare Corp. (McGaw Park, IL) or J. T. Baker (Phillipsburg, NJ).
Synthetic Procedures. Synthesis of D3270.
This carboxylic acid derivative was prepared from the corresponding primary alcohol (D3017) in a two-step procedure, involving oxidation first to the aldehyde (8) and then to the acid (9). After these reactions, the methyl ester was analyzed by GC/MS. The mass spectrum of this material, which was fully consistent with the proposed structure, exhibited the following prominent ions: m/z 274 (M+·, 75%), 256 ([M-H2O]+·, 15%), 246 ([M-CO]+·, 38%), 215 ([M-CO2CH3]+, 100%), 187 ([215-CO]+, 20%), 120 (ArNH+, 61%), and 105 (C8H9+, 22%). With respect to purity of the synthesis of D3270, only a single peak was observed on GC analysis of the methyl ester.
Synthesis of [13C2]D2624.
[13C2]D2624 labeled in the xylidine methyl groups was prepared in a three-step procedure. Labeled 2,6-dimethylacetanilide, synthesized from acetanilide and [13C]CH3I according to the procedure of Tremont and Rahman (10), was hydrolyzed with aqueous potassium hydroxide, and the resulting labeled 2,6-DMA was condensed with 5-methyl-3-isoxazolecarboxylic acid to yield 54% [13C2]D2624 labeled in the xylidine methyl groups (99 atom% excess).
Instrumentation. 1H NMR Spectroscopy.
This was performed on a Varian VXR 300 instrument (Varian Associates, Palo Alto, CA), operated at 300 MHz. Samples were analyzed in CDCl3 solution, and chemical shifts are expressed relative to tetramethylsilane, which was added as the internal standard.
MS and GC/MS.
MS was performed on a VG 70-70H double-focusing instrument (VG Instruments, Manchester, UK), operated in the electron impact mode at an ionizing energy of 70 eV and a resolving power of 10,000. Samples were introduced by the direct insertion technique, and data were acquired at a scan rate of 10 sec/decade. GC/MS analyses were conducted on a Hewlett-Packard 5970A MSD instrument, interfaced via a solvent divert valve to a Hewlett-Packard model 5890 GC equipped with a capillary splitless injector and a Hewlett-Packard 7673A autosampler. A fused silica capillary column (30 m × 0.32 mm i.d., 0.25 μm film thickness) coated with bonded stationary phase DB-5 was used for GC separations, and helium (head pressure 20 psi) served as carrier gas. Samples were injected in the splitless mode (injector temperature: 250°C) at a column temperature of 170°C and an interface temperature of 280°C. After a period of 5 min at 170°C, the oven temperature was increased linearly at 14°C/min to 280°C, and maintained at this temperature for 15 min. Metabolites were detected on the basis of the characteristic isotopic “doublets” that result from the presence of unlabeled and 13C2molecules in a 1:1 ratio, and were identified by comparing their GC and MS properties with those of the corresponding authentic reference compounds prepared by synthesis.
Biological Experiments. Metabolite Identification In Vivo.
Adult male Sprague-Dawley rats (230–280 g) from Charles River Laboratories (Wilmington, MA) were housed in individual metabolic cages. Animals were dosed orally either with unlabeled D2624 (10 mg/kg; administered as a 5 mg/ml solution in saline/propylene glycol/ethanol, 5:4:1, v/v) or with an equimolar mixture of unlabeled and [13C2]D2624 (24 mg/kg total dose, administered in a solution of polyethyleneglycol-400/saline, 2:1, v/v). Corresponding control animals received vehicle only. For the identification of metabolites in blood, rats were anesthetized with pentobarbital and provided with a jugular catheter 2–4 days before use and were allowed to recover before drug treatment. Blood samples (0.4 ml) were obtained via the indwelling catheter (using EDTA as anticoagulant) at the following time points: 0, 15, 30, 45, 60, 90, and 120 min postdose. Aliquots of whole blood (100 μl) were extracted with dichloromethane (5 ml), and extracts were evaporated to dryness under a stream of dry nitrogen. The residues then were reconstituted in the mobile phase (acetonitrile/water, 44:55, v/v) and analyzed by HPLC (Zorbax C8 column, 5 μm particle size, 4.6 × 250 mm) at a flow rate of 1.1 ml/min with UV detection at λ = 220 nm. In some cases, the previously described organic extracts were dissolved in dichloromethane and analyzed directly by GC/MS.
For the identification of metabolites in urine, rats were treated orally with the 1:1 mixture of unlabeled and 13C-labeled forms of D2624 (24 mg/kg). Their urine was collected over ascorbic acid (500 mg) for 7 hr postdose. Aliquots of urine (1–3 ml) were treated with either β-glucuronidase (500 μl, 867 units; pH 6.8) or sulfatase (500 μl, 20 units; pH 5.0) and incubated at 37°C for 2 hr. The hydrolyzed urines then were passed through C18-SepPak cartridges (3 ml; Baker), which had been preconditioned by treatment first with methanol (15–20 ml) and then distilled water (18–20 ml). Cartridges were washed with water (6–10 ml), and the aqueous eluates were collected. Lipophilic metabolites were then eluted with methanol (10–15 ml), and the methanol eluates were concentrated to approximately one-half of their original volume under reduced pressure. The latter organic phases were acidified with HCl (12 M) and extracted into dichloromethane (2 × 3 ml). The combined extracts were evaporated to dryness under a stream of dry nitrogen, taken up again in dichloromethane (2 ml), dried over magnesium sulfate, and concentrated to ∼200 μl for analysis by GC/MS. A portion of each extract was methylated by treatment with a freshly prepared solution of ethereal diazomethane to facilitate the analysis of acidic metabolites of D2624. For the identification of 2,6-DMA as a urinary metabolite of D2624, the corresponding pentafluorobenzoyl derivative was prepared using a modification of the method reported by Coutts et al. (11) reconstituted in dichloromethane for analysis by GC/MS.
Metabolite Identification In Vitro.
Hepatic microsomal preparations were obtained from the livers of male Sprague-Dawley rats (250 g body weight) using the procedure previously described by Thummel et al. (12). Incubations (total volume: 1 ml) contained microsomal protein (1 mg/ml), NADPH (1 mM), and D2624 (0.1 mM) in 100 mM phosphate buffer (pH 7.4), and were conducted at 37°C for 15 min. The incubation media then were extracted with dichloromethane (5 ml) and analyzed by HPLC (as previously described) and GC/MS. To analyze the major metabolite of D2624 (which was formed both in vivo and in vitro) by NMR spectroscopy, milligram quantities of the material were isolated by extracting a number of liver microsomal incubations and subjecting the pooled extracts to HPLC, as previously outlined.
In Vitro Metabolism of D2624 by Rat and Human Liver Microsomes.
Three human livers (HL115, HL120, and HL125) were obtained from organ donors through the Solid Organ Transplant Program, University of Washington Medical Center, Northwest Organ Procurement Agency (Seattle, WA). Microsomes were prepared after a procedure similar to that used for rat liver microsomes, except that final resuspension was in 100 mM potassium phosphate buffer at a concentration of 6.6 mg protein/ml. Incubations were conducted at 37°C with a final incubation volume of 1 ml. Metabolite formation was investigated by incubating 1 mg rat or human microsomal protein with D2624 at 10 and 100 μM (rat) or 100 μM (human) for 10 min in the presence or absence of NADPH. Amidase-mediated metabolism of D2624 at 10 μM (rat) or 100 μM (human) was assessed by preincubating 1 mg rat or 0.3 mg human liver microsomal protein with bis(4-nitrophenyl)phosphate (an amidase inhibitor) and 1 mM NADPH for 10 min at 0–4°C. All incubations were performed in triplicate and were stopped either after 10 min by the addition of dichloromethane (4 ml) and 50 μl of internal standard (cyheptamide, 5 μg/ml) for the analysis of D2624 and D3017, or by the addition of acetonitrile (4 ml) and 50 μl of internal standard (cyheptamide, 5 μg/ml) for the analysis of 2,6-DMA. The organic layer was removed and evaporated to dryness under nitrogen. The residues then were reconstituted in a 100 μl mobile phase consisting of either acetonitrile/water (45:55, v/v), for the analysis of D2624 and D3017, or acetonitrile/0.01M potassium phosphate buffer (pH 5; 50:50, v/v), for the analysis of 2,6-DMA.
Pharmacokinetics of D2624 and Its Metabolites.
Male Sprague-Dawley rats (230–300 g) were anesthetized with pentobarbital and equipped with a femoral and a jugular catheter 2–4 days before drug treatment. Drug/vehicle was administered on a 2 ml/kg basis via the femoral vein. The following compounds were given at the doses indicated: D2624 at 4 mg/kg (17.4 μmol/kg;N = 6) or 12 mg/kg (52.2 μmol/kg; N = 9); D3017 at 12 mg/kg (48.8 μmol/kg; N = 6); and 2,6-DMA at 2 mg/kg (16.5 μmol/kg; N = 6). In some experiments, compounds were administered via oral gavage [D2624 at 52.2 μmol/kg (N = 6) or D3017 at 48.8 μmol/kg (N = 6)] as a solution in polyethylene glycol 400/saline (2:1, v/v). Serial samples (N = 10) of blood (300 μl) were taken at the following time points: 0, 3, 10, 20, 40, 60, 80, 120, 180, and 240 min postdose via the indwelling jugular vein catheter into heparinized tubes and stored at −70°C before HPLC analysis of D2624, D3017, and 2,6-DMA.
HPLC Analysis of D2624 and D3017.
Blood samples (100 μl) were extracted with dichloromethane (5 ml) containing 50 μl cyheptamide solution (5 μg/ml; internal standard). After centrifugation at 600g for 15 min, the organic phase was removed and evaporated to dryness under nitrogen. Residues were reconstituted in 100 μl mobile phase (acetonitrile/water, 45:55, v/v) and analyzed by reversed-phase HPLC (Waters Zorbax C8column, 4.6 × 250 mm, 5 μm particle size) at a flow rate of 1.1 ml/min, with UV detection at λ = 210 nm.
HPLC Analysis of 2,6-DMA.
Blood samples (100 μl) were extracted with acetonitrile (4 ml) containing 250 μg cyheptamide (internal standard). After centrifugation at 600g for 15 min at room temperature, the organic layer was removed and concentrated under nitrogen to 100–200 μl at room temperature. Care was taken during the latter step to ensure that samples did not evaporate to dryness because of the volatile nature of 2,6-DMA. Extracts were analyzed by HPLC using a Zorbax C8 column (4.6 × 250 mm, 5 μm particle size), a mobile phase of acetonitrile and 0.01 M potassium phosphate buffer (pH 5; 50:50, v/v) at a flow rate of 1 ml/min, and UV detection at 210 nm.
Human Studies.
The study protocol was approved by the Ethics Committee in the region of Poitou-Charentes in accordance with the French law of December 28, 1988 and the Declaration of Helsinki, Finland, of June 1964, amended at the Third World Medical Assembly (1983), Venice, Italy. Six healthy males aged 28 ± 2 year and weighing 69 ± 11 kg volunteered for the study after giving their written informed consent. Before entering the study, subjects were found to be healthy by clinical examination, and none of the subjects was on any chronic medication. Volunteers fasted overnight before the study. Capsules containing 25, 50, or 100 mg of D2624 were administered with 400 ml water as single doses of 25, 50, 75, 100, 150, or 225 mg (one dose per subject). Blood samples (10 ml) were obtained just before dosing and at 0.25, 0.5, 1, 2, 3, 4, 6, and 8 hr postdose, then analyzed for D2624, D3017, and 2,6-DMA.
Pharmacokinetic Data Analysis.
After intravenous administration of D2624, D3017, or 2,6-DMA to rats, pharmacokinetic parameters were obtained from one- or two-compartment models using the pharmacokinetics software package PCNONLIN V4.2 (SCI Software, Lexington, KY). Half-life was calculated as 0.693/k (where k is the final decay rate constant), and total clearance was obtained from the relationship Dose/AUC [where AUC is (AUC0-∞)]. TheFTot of D2624 and D3017 in rats was calculated from the ratio of the area under the oral dosing curve [(AUC)po], to the area under the intravenous (iv) dosing curve [(AUC)iv], [i.e. Ftot = (AUC)po/(AUC)iv]. The calculation of Fm was determined after intravenous administration of parent drug (P) and metabolite (M), using the AUC of the metabolite (AUCm) according to the general formula: Fm = [(AUCm)P/(AUCm)M] × [(doseM)/(doseP)], where the superscripts P and M denote the compound administered. This approach involves a number of assumptions: i) disposition of metabolite was independent of its origin (i.e. formed in situ or administered); ii) metabolite clearance is constant between studies; and iii) fractions thus calculated are fractions of intravenous clearance, not fractions formed. In the triangular metabolic scheme shown in fig. 1, the fraction of D2624 metabolized to 2,6-DMA via D3017 (Fm4) was calculated from the product of Fm1 (fraction D2624 metabolized to D3017) and Fm3 (fraction D3017 metabolized to 2,6-DMA). The fraction of D2624 metabolized directly to 2,6-DMA (Fm5) was calculated from the difference between Fm2 (total fraction of 2,6-DMA formed from D2624) and Fm4.
The terminal phase of the plasma curve for D3017 and 2,6-DMA in humans after oral administration of parent drug was fitted to a monoexponential equation to determine the final elimination rate constant using the pharmacokinetic software package SAAM II V1.01 (RFKA, University of Washington, Seattle, WA).
Statistical Analysis.
Results are expressed as means ± SD throughout. Statistical significance between groups at the level of p < 0.05 was assessed by ANOVA and Mann-Whitney U/Wilcoxon rank sum test.
Results
Identification of Metabolites of D2624 Formed In Vivo.
The most abundant metabolite of D2624, which was detected in both plasma and urine after oral administration to rats and which also was formed upon incubation of D2624 with rat liver microsomal preparations, afforded a mass spectrum consistent with that of a hydroxylated derivative of the parent drug. Thus, abundant ions were present in the spectrum of the unlabeled metabolite at m/z 246 (M+·, 58%), 218 ([M-CO]+·, 28%), 215 ([M-CH2OH]+, 73%), 187 ([M-CO-CH2OH]+, 18%), 148 ([Ar-NH-CO]+, 20%), 147 ([Ar-NCO]+·, 41%), 120 ([Ar-NH]+, 100%), 105 ([C8H9]+, 52%), 91 ([C7H7]+, 38%), and 77 ([C6H5]+, 51%). Elemental compositions of these ions were verified by their respective accurate masses which, in all cases, were within 10 ppm of the theoretical values. When the metabolite was isolated from rats that had been dosed with [12C/13C2]D2624, all of the previously described ions exhibited the characteristic12C/13C “doublets” consistent with their proposed structure. Based on these mass spectral features, and particularly on the composition of the intense fragment ion atm/z 215, it was concluded that the most likely structure for D3017 was the primary alcohol resulting from hydroxylation of D2624 on the isoxazole methyl group (fig. 1). A sufficient quantity of this metabolite was isolated from incubations of unlabeled D2624 with rat liver microsomes to allow analysis by 1H NMR. This gave a spectrum that exhibited the following resonances (δ): 2.3 (s, 6H, [CH3]2), 4.8 (s, 2H,—CH2—OH), 6.8 (s, 1H, isoxazole C—H), and 7.1–7.3 (m, 3H, Ar-H). This spectrum confirmed that the site of hydroxylation was the isoxazole methyl group, inasmuch as the 3H singlet (at 2.5 ppm) in the spectrum of the parent drug (D2624) had been replaced by the 2H singlet at 4.8 ppm in the spectrum of the metabolite. When a synthetic standard of D3017 became available, it was found to exhibit mass spectral, NMR, and HPLC characteristics identical to those of the biological material, thereby confirming the aforementioned structural assignment.
A second metabolite, which was detected only in methylated extracts of the urine of rats dosed with [12C/13C2]D2624, afforded the following mass spectrum upon GC/MS analysis: m/z 274/276 (M+·, 53%), 256/258 ([M-H2O]+·, 17%), 246/248 ([M-CO]+·, 40%), 215/217 ([M-CO2CH3]+, 100%), 187/189 ([215-CO]+, 22%), 120/122 (ArNH+, 57%), and 105/107 (C8H9+, 20%). On the basis of these mass spectrometric characteristics, and on the coincidence in GC retention times of the methylated metabolite and the methyl ester of synthetic D3270, this metabolite was identified as the carboxylic acid D3270 (fig. 1).
A third metabolite, which was present in both plasma and urine of rats treated with [12C/13C2]D2624, reacted with pentafluorobenzoyl chloride to form a derivative with the following mass spectral characteristics: m/z 315/317 (M+·, 12%), 195 (C6F5CO+, 82%), 167 (C6F5+, 58%), and 120/122 (C8H9NH+, 100%). This spectrum was similar to that of the pentafluorobenzoyl derivative of 2,6-DMA, a reference sample that was found to exhibit the same GC retention time as that of the metabolite. On the basis of these results, the metabolite was identified as 2,6-DMA (fig. 1).
In Vivo Pharmacokinetics of D2624 and Its Major Metabolites.
Figure 2A illustrates semilogarithmic plasma concentration-time profiles for D2624, D3017, and 2,6-DMA after a 52.2 μmol/kg iv bolus dose of D2624. The plasma profile of D2624 was biphasic, and the pharmacokinetic parameters obtained are shown in table 1. There was no statistically significant difference (p > 0.05) between the parameters obtained at 17.4 and 52.2 μmol/kg, suggesting that D2624 exhibits linear pharmacokinetics within this dose range. The concentration-time profiles of the two metabolites were best described by one-compartment models with first-order formation and elimination rate constants. Pharmacokinetic parameters obtained are shown in table 1, from which it may be seen that both metabolites had similar elimination half-lives that were significantly (p < 0.001) longer than that of the parent drug.
The plasma concentration-time curves for D2624 and D3017 (12 mg/kg) were found to be dependent on route of administration. Compared with intravenous administration, significant (p < 0.001) increases in the terminal half-lives of D2624 and D3017 were observed after oral dosing and were attributed to an absorption-rate limitation in elimination kinetics of both parent drug and metabolite (fig. 2B; table 1).
After intravenous administration of D3017 at 48.8 μmol/kg, the plasma profile of D3017 was best described by a one-compartment model, despite some curvature after three half-lives. The similar elimination half-lives of D3017 obtained after intravenous administration of D3017 (24.4 ± 4.1 min) and D2624 (30.7 ± 3.4 min) suggest that D3017 exhibited elimination rate-limited kinetics after D2624 administration. The steady-state volume of distribution of D3017 and its systemic clearance were significantly (p < 0.001) less than those of D2624. The plasma profile of 2,6-DMA following intravenous administration of D3017 was best described by a one-compartment model, with first-order formation and elimination rate constants. Terminal half-life values of D3017 and 2,6-DMA were not significantly different from each other, suggesting formation-rate limited kinetics for 2,6-DMA.
Plasma kinetics of D3017 after oral administration (48.8 μmol/kg) were best described by a one-compartmental model with first-order absorption and elimination rate constants (table 1). A significant (p < 0.001) increase was observed in the terminal half-life of D3017 from 24.4 ± 4.1 min (iv) to 54.1 ± 17.1 min (po) and was attributed to an absorption-rate limitation in plasma kinetics.
The time course of plasma concentrations of 2,6-DMA after intravenous administration of 2,6-DMA (16.5 μmol/kg) was biexponential in four animals, whereas two animals showed a monoexponential decay. Values for elimination half-life of 2,6-DMA after intravenous administration (table 1) were significantly (p < 0.001) shorter than the corresponding values found after administration of either D2624 or D3017, thus suggesting formation-rate limited kinetics.
Signs of neurological side effects were not observed in rats after intravenous administration of D2624 at 17.4 μmol/kg or after oral administration of 52.2 μmol/kg. However, intravenous administration of D2624 at 52.2 μmol/kg caused mild sedation for a short period, and D3017 caused a mild transient muscular incoordination for 5–10 min after intravenous administration of 48.8 μmol/kg.
Oral Bioavailability of D2624 and D3017 in Rats.
The AUC method was used to determine the FTot of D2624 and D3017 after oral administration. TheFTot of D2624 (5.3%) was much lower than that of D3017 (67.1%). Poor bioavailability of D2624 was attributed to either incomplete absorption from the gastrointestinal tract and/or extensive first-pass metabolism of the drug. The AUC of the metabolite (D3017) after oral and intravenous administration of the parent compound (D2624) was used to estimate the fraction of D2624 absorbed from the gut into the portal circulation using the equationFabs = (AUCm)po/(AUCm)iv. This AUC method indicated that absorption of D2624 was almost complete with ∼93% of the dose appearing in the systemic circulation as the metabolite. Thus, the poor bioavailability of D2624 was most likely because of extensive first-pass metabolism. To quantify this first-pass effect, the fraction of the absorbed dose that crosses the liver unchanged (FLiv) was estimated by dividing the oral bioavailability by the fraction of dose absorbed from the gut based on metabolite AUC (FLiv =FTot/Fabs). Using this equation, only 5.8% of the total dose absorbed was able to cross the liver unchanged.
Fractional Metabolism of D2624 and D3017.
The formation of 2,6-DMA, after dosing with D2624, takes placevia either arm of a “triangular metabolic” pathway, because both D2624 and its hydroxylated metabolite, D3017, possess an amide bond that, upon cleavage, can release 2,6-DMA (fig. 1). The fractional metabolism in each part of this triangular metabolic problem was determined using the plasma AUC method for the parent drug and its metabolites. The total fraction of D2624 converted to 2,6-DMA byi) direct hydrolysis to 2,6-DMA and ii) indirect metabolism via D3017 was calculated using the AUC ratio of 2,6-DMA after intravenous administration of D2624 to that obtained by dosing with 2,6-DMA and found to be 0.55 (Fm2). To determine the fraction of D2624 metabolized to 2,6-DMA via D3017 (Fm4), the fraction of D2624 converted to D3017 (Fm1) was estimated to be 0.42. The fraction of D3017 that undergoes metabolism to 2,6-DMA was calculated from the AUC ratio of 2,6-DMA after intravenous administration of both D3017 and 2,6-DMA, and estimated to be 0.16 (Fm3). The product of the fractions Fm1 and Fm3 indicated that only 6.6% of the dose of D2624 was metabolized to 2,6-DMA viaD3017 (Fm4). Therefore, the fraction of D2624 metabolized directly to 2,6-DMA (Fm5) could be estimated by subtracting Fm4 from Fm2 and found to be 0.48. Thus, it seems that D3017 and 2,6-DMA are the main metabolites of D2624, because they account collectively for 90% of an intravenous dose.
In Vitro Metabolism in Rat Liver Microsomes.
Rat liver microsomes metabolized D2624 by NADPH-dependent oxidative biotransformation to the alcohol metabolite D3017. However, the formation of 2,6-DMA was NADPH-independent, but could be inhibited by preincubating microsomes with BNPP, an amidase inhibitor.
In Vitro Metabolism in Human Liver Microsomes.
Human liver microsomes also were found to metabolize D2624 rapidly to D3017 by NADPH-dependent oxidative biotransformation. The formation of 2,6-DMA was NADPH-independent and was inhibited by preincubating microsomes with BNPP.
Human In Vivo Studies.
Plasma obtained from human volunteers receiving single oral doses of D2624 in the range 25–225 mg did not contain measurable levels of D2624, indicating poor total bioavailability. However, plasma samples contained D3017 within 15 to 30 min of administration at all dose levels (fig. 3A). Although the AUCs of D3017 represent single observations, they were approximately proportional to dose. Values for the half-life of D3017 ranged from 1.3 to 2.2 hr. The metabolite 2,6-DMA was detected in plasma (nM levels) only at the highest doses of D2624 studied (150 and 225 mg; fig. 3B).
Discussion
By means of stable isotope labeling and GC/MS techniques, three metabolites of D2624 were identified in the rat. The most abundant metabolite, detected in both plasma and urine after oral administration, was D3017, which is the primary alcohol resulting from hydroxylation of the isoxazole methyl group. A second metabolite, which was detected only in methylated extracts of urine, was D3270, which is the carboxylic acid derivative formed from the corresponding primary alcohol (D3017). A third metabolite, which was present in plasma and urine, was identified as 2,6-DMA, which was formed by the hydrolysis of the amide bond of either D2624 or D3017 (fig. 1).
After intravenous administration of D2624, both metabolites had similar elimination half-lives that were significantly longer than that of the parent drug (fig. 2A). Intravenous metabolite studies were performed to determine whether metabolite disposition after parent drug administration was formation or elimination rate-limited. These studies indicated that D3017 exhibited elimination rate-limited kinetics after parent drug administration. However, the half-life of 2,6-DMA was significantly longer after intravenous administration of either D2624 or D3017 than after intravenous administration of 2,6-DMA, suggesting that 2,6-DMA exhibited formation rate-limited kinetics. 2,6-DMA exhibited a relatively high clearance (58.2 ml/min/kg), suggesting that further metabolism of 2,6-DMA may occur such that the calculated fractions of D2624 and D3017 metabolized to 2,6-DMA may represent underestimates.
After oral administration of D2624, the terminal half-life of D2624 increased 5-fold, which suggests that absorption became the rate-limiting step. This hypothesis was supported by the fact that D3017 half-life was increased 3-fold (fig. 2B). D2624 exhibited a much lower bioavailability than D3017. Using D3017 as a marker for absorption of the parent drug, it was found that D2624 was in fact almost completely absorbed from the gastrointestinal tract. Thus, it seems that D2624 acts as a prodrug for D3017. Because D3017 has potent anticonvulsant properties (3), it must contribute therefore to the activity observed after oral administration of D2624 (3). This may explain why greater efficacy against generalized tonic-clonic seizures was evident when D2624 was given orally, compared with intravenous administration (3).
Because D2624 and its hydroxylated metabolite D3017 possess an amide bond that, upon cleavage, can release 2,6-DMA, the metabolic source of 2,6-DMA was investigated. A stepwise approach, based on plasma AUC, was used to determine the fractional metabolism of D2624 and D3017 (fig.1B). It is interesting to note that methyl hydroxylation of the D2624 structure at a site distant from the amide bond was associated with a marked reduction in susceptibility to hydrolysis. These quantitative metabolic studies also revealed that the two primary metabolites, D3017 and 2,6-DMA, accounted for 90% of a dose of D2624.
Rat and human liver microsomal preparations indicated that the two metabolites of D2624 are formed by different enzyme systems. The hydroxylation of the isoxazole methyl group to form D3017 was NADPH-dependent and therefore probably cytochrome P450-mediated, whereas the formation of 2,6-DMA was NADPH-independent, inhibited by BNPP, and probably catalyzed by amidase enzymes. In human and rat liver microsomes, the formation of D3017 was the major route of metabolism of D2624, whereas 2,6-DMA formation was minor.
The in vitro human microsomal data suggested that D2624 would undergo extensive hepatic metabolism to D3017. Although the enzyme(s) responsible for the formation of D3017 were not identified and the Michaelis-Menten parameters not determined, the in vitro formation velocities of D3017 and 2,6-DMA seem to be in agreement with the in vivo findings [i.e. D3017 was present in plasma in high concentrations, whereas 2,6-DMA was detected in plasma at low concentrations (nM) and only at the highest two doses]. In general, 2,6-DMA concentrations were <3% of those of D3017 and suggest species differences in the oxidative and hydrolytic pathways of D2624. The detection of 2,6-DMA in human plasma and the very short half-life of the active metabolite (D3017) suggest that D2624 does not possess optimal characteristics for an antiepileptic compound. However, these studies have provided an insight into the metabolic fate and pharmacokinetics of a representativeN-aryl isoxazolecarboxamide that should prove useful in the assessment of other candidate drugs from this structural class.
Footnotes
-
Send reprint requests to: Professor René H. Levy, Department of Pharmaceutics, School of Pharmacy, Box 357610, University of Washington, Seattle, WA 98195-7610.
-
This research was supported by a grant from Laboratories Biocodex, Compiègne, France, which is gratefully acknowledged.
- Abbreviations used are::
- D2624
- N-(2,6-dimethylphenyl)-5-methyl-3isoxazolecarboxamide
- ED50
- median effective dose
- TD50
- median toxic dose
- 2
- 6-DMA, 2,6-dimethylaniline
- D3017
- N-(2,6-dimethylphenyl)-5-hydroxymethyl-3-isoxazolecarboxamide
- D3270
- N-(2,6-dimethylphenyl)-5-carboxy-3-isoxazolecarboxamide
- 1HNMR
- proton magnetic resonance
- Fm
- fraction metabolized
- AUC
- area under the plasma concentration-time curve
- FTot
- total bioavailability
- Fabs
- absorbance bioavailability
- FLiv
- hepatic bioavailability
- BNPP
- bis(4-nitrophenyl) phosphate
- Received March 15, 1996.
- Accepted October 10, 1996.
- The American Society for Pharmacology and Experimental Therapeutics