Introduction

Top Abstract Introduction Methods and Materials Results Discussion References

ABT-200¹ (I; Fig. 1) is a novel pyrrolidine derivative that exists as a racemic mixture of the S,S()- and R,R(+)-enantiomers. Preliminary studies demonstrated that the compound exhibited nonlinear pharmacokinetics during oral administration in the dog. Whereas initial studies with dog and human liver microsomes generated only the NADPH-dependent catechol metabolite of I, liver slices from these species formed extensive amounts of secondary metabolites resulting from further biotransformation of the catechol derivative via O-methylation, sulfation, and glucuronidation (Ferrero et al., 1997). These results suggested that oxidation of the methylenedioxy bridge carbon to form the catechol was a major initial metabolic step in the metabolism of I.

View larger version (10K): [in this window] [in a new window]   Fig. 1.   Chemical structure of racemic I.

Certain methylenedioxyphenyl compounds interact with the mixed-function oxidases to form metabolite complexes with cytochrome P450 (P450) (Franklin, 1971; Philpot and Hodgson, 1972; Hodgson and Philpot, 1974; Fennell et al., 1980; Murray et al., 1985). Treatment of rats with isosafrole followed by isolation of hepatic microsomes results in the formation of a P450-metabolite complex, which exhibits absorption maxima at 455 nm in the dithionite-reduced difference spectrum (type III spectra) (Franklin, 1971). Formation of these complexes is believed to occur by NADPH-dependent oxidation of the methylenedioxy ring to yield a carbene and is accompanied by the inhibition of monooxygenase activity (Ullrich, 1977; Dickins et al., 1979; Mansuy, 1980).

The objective of these studies was to evaluate the metabolic fate of I in the dog and the potential relationship between metabolism and the nonlinear pharmacokinetic behavior. This report summarizes the results of a series of in vivo and in vitro studies, which examined the metabolism of each enantiomer of I (as the respective ¹⁴C-pseudoracemate), the putative formation of a P450-metabolite complex, and the potential involvement of a P450-metabolite complex in the nonlinear pharmacokinetics in the dog.

	Methods and Materials

Top Abstract Introduction Methods and Materials Results Discussion References

Chemicals. The carbon-14 labeled and unlabeled R,R(+)- and S,S()-enantiomers of I were synthesized by Abbott Laboratories (Abbott Park, IL). The structures of the enantiomers of I and the position of the carbon-14 label are depicted in Fig. 1. The authentic catechol reference standard and the two O-methyl reference standards, in which the methyl moiety is attached at the hydroxy position of the catechol either meta or ortho to the six-membered saturated ring, were synthesized by Abbott Laboratories. Sulfatase, -glucuronidase and Krebs-Henseleit buffer were purchased from Sigma-Aldrich (St. Louis, MO).

Animals. Fifteen male beagle dogs were used in these studies (Marshall Research Animals, North Rose, NY). The dogs weighed 6 to 10 kg and were 1 to 3 years of age. The animals were fed 12 h after drug administration and daily thereafter. During the study, the dogs were housed individually in stainless steel metabolism cages.

In Vivo Metabolism Studies. Metabolism and disposition of the ¹⁴C-pseudoracemates of I. This was a multicrossover study in which the same animals were used for both the oral and intravenous legs with the ¹⁴C-pseudoracemates of I.

Dosing of the ¹⁴C-pseudoracemates of I and sample collection. The carbon-l4 labeled R,R(+)-enantiomer of I was mixed with an equal amount of unlabeled S,S()-enantiomer (to form the ¹⁴C-R,R(+)-pseudoracemate). Likewise, to form the ¹⁴C-S,S()-pseudoracemate, carbon-14 labeled S,S()-enantiomer of I was mixed with an equal amount of unlabeled R,R(+)-enantiomer. Both pseudoracemates were dissolved in 5% dextrose in sterile water at a concentration of 2.0 mg/ml (2.6 mg mesylate salt/ml). The ¹⁴C-R,R(+)-pseudoracemate of I was administered to three dogs orally by gavage at a dose of 2 mg/kg, corresponding to 1 ml/kg, and about 50 µCi/animal. Two weeks later, these dogs were given a 2 mg/kg dose of the ¹⁴C-R,R(+)-pseudoracemate (corresponding to 1 ml/kg and about 50 µCi/animal) intravenously through the cephalic vein. Approximately three weeks after the intravenous segment of the ¹⁴C-R,R(+)-pseudoracemate study was completed, the same three dogs were given a 2 mg/kg oral dose of the ¹⁴C-S,S()-pseudoracemate, corresponding to 1 ml/kg, and about 50 µCi/animal. Two weeks later, these dogs were given a 2 mg/kg dose of the ¹⁴C-S,S()-pseudoracemate (corresponding to 1 ml/kg and about 50 µCi/animal) intravenously through the cephalic vein. After dosing was completed, the radiochemical purity of the dose solution was determined to be at least 98%. After each dosing segment was completed, the radiochemical purity of the dose solutions were determined to be at least 98%.

Metabolism and disposition of the ¹⁴C-enantiomers of I. Since only one of the enantiomers was radiolabeled in each pseudoracemate formulation, only the metabolites from the respective radiolabeled enantiomer were detected by HPLC/radioassay. The samples taken from these animals, however, necessarily contained a mixture of both S,S() and R,R(+) metabolites. Consequently, the resolution of metabolites from these preparations by the present nonchiral HPLC method would result in the isolation of a single metabolite peak, which would likely contain the metabolites of both enantiomers. Therefore, to more accurately characterize the metabolites of each enantiomer of I, separate dogs were dosed orally (by gavage) with either the single ¹⁴C-S,S()- or ¹⁴C-R,R(+)-enantiomer orally by gavage at a dose of 2 mg/kg (corresponding to 1 ml/kg, and about 50 µCi/animal). The urine of these animals was collected for metabolite isolation and mass spectral characterization.

Pharmacokinetic Study and In Vivo P450-Metabolite Complex Formation of Racemic I. Dogs were given a single or multiple doses of racemic I, and the pharmacokinetic parameters were examined after the first and last dose. The animals were sacrificed after the last dose, and the livers were excised. Microsomes were prepared from the liver of each animal and the amount of P450-metabolite complex was determined as described below.

Dose and sample collection. In the multiple dose group, racemic I was administered to two dogs orally in capsules at a dose of 12.0 mg base/kg/day for 14 days. The corresponding two single dose animals were given empty gelatin capsules on days 1 through 13 followed by a single 12.0 mg base/kg oral dose of I on day 14. The control animal received empty gelatin capsules for 14 days.

Measurement of P450-metabolite complex formed in vivo. Livers were homogenized and microsomes prepared according to standard procedures (Omura and Sato, 1964). Measurement of the P450-metabolite complex was determined using a Shimadzu model UV-2101PC UV/visible scanning spectrophotometer (Scientific Instruments Inc., Hawthorne, NY). The microsomes were diluted in 0.1 M potassium phosphate buffer (pH 7.4) containing 0.1 mM EDTA and divided equally into sample and reference tubes. Formation of the P450-metabolite complex was monitored by recording the reduced minus oxidized spectrum between 400 and 500 nm. Reduction was effected by addition of sodium dithionite to the sample cuvette. Quantitation of the P450-methylenedioxyaryl complex was determined by A_{455-490 nm} using the extinction coefficient of 75 mM¹cm¹ (Elcombe et al., 1975; Fennell et al., 1980; Murray et al., 1985). The concentration of protein in the liver microsomes was determined by the method of Smith et al. (1985).

Plasma analytical method. The concentration of racemic I in each sample was determined by reverse-phase HPLC with electrochemical detection following liquid-liquid extraction (ethyl acetate/hexane) of the plasma matrix. The assay was linear (correlation coefficient >0.99) over the concentration range 0 to 220 ng/ml with a pooled standard deviation of <5%. Standards were analyzed in triplicate at six different concentrations with an estimated detection limit of 0.1 ng/ml. Samples with I concentrations in excess of 30 ng/ml were subjected to a second analysis to determine the enantiomeric distribution. The chiral analysis used liquid-liquid extraction with ethyl acetate/hexane to separate the compounds of interest from the plasma contaminants. Chiral separation was achieved on a 10 cm × 4.0 mm 5 µm 1-acid glycoprotein column with an acetonitrile/0.01 M KH₂PO₄ pH 5.5 (12:88, by volume) mobile phase at a flow rate of 0.7 ml/min and with detection at 205 nm. Column temperature was maintained at 35°C.

Pharmacokinetic analysis. The area under the plasma concentration-time curve from 0 to 24 h (AUC_0-24h) after dosing was calculated using the linear trapezoidal rule. Since the bioavailability of racemic I was not known for the animals in this study, the apparent plasma clearance of I was determined using the equation of Cl = dose/AUC.

Radioassay. In the radiolabeled metabolism studies, duplicate aliquots of the bile, urine, and cagewash samples were radioassayed directly in Insta-Gel (PerkinElmer Life Sciences, Boston, MA) scintillation cocktail. The feces were homogenized in water using a motor-driven high frequency homogenizer. Duplicate aliquots of the whole blood samples and fecal homogenates were placed in Combusto-Cones (PerkinElmer Life Sciences) containing cellulose powder (Whatman, Clifton, NJ) and burned in a Tri-Carb (PerkinElmer Life Sciences) model 307 sample oxidizer. The resulting ^l4CO₂ was trapped in Carbosorb and radioassayed in Permafluor V (PerkinElmer Life Sciences) liquid scintillation fluid and counted in a Tri-Carb (PerkinElmer Life Sciences) model 2500TR liquid scintillation spectrometer. All samples were corrected for quenching with an automatic external standardization technique.

HPLC/Radioassay. Analyses of I and metabolites in biological samples, as well as incubations from the in vitro studies with dog liver microsomes, were accomplished using an IBM Instrument, Inc., LC 9560 liquid chromatograph and a diode array detector (Applied Biosystems, Inc., Foster City, CA) operated at 215 nm. Separations were achieved in a reversed phase mode using a C₁₈ Altima 5-µ column (4.6 × 250 mm; Alltech Associates, Deerfield, IL) at ambient temperature. I and its metabolites were resolved by gradient system using two mobile phases. Mobile phase (A) was composed of 0.01 M triethylammonium formate, whereas mobile phase (B) contained 75% acetonitrile in 0.01 M triethylammonium formate. The gradient was run from 80% A/20% B at 0 min, 65% A/35% B at 15 min, 60% A/40% B at 30 min, 100% B at 40 min and then isocratically with 100% B for 10 min thereafter. The flow rate was 0.9 ml/min. The column effluent was passed through a Flo-One Beta radioactive flow detector (Radiomatic Instruments, Inc., PerkinElmer Life Sciences) where it was combined with scintillation cocktail (Flo-Scint III, Radiomatic Instruments, Inc., PerkinElmer Life Sciences) at a ratio of 1:3.6. The radioactivity eluting from the column was monitored at 6-s intervals to obtain the metabolic pattern.

Isolation of Metabolites. Urine samples from the radiolabeled enantiomer metabolism study were carefully applied to an ion-exchange column and washed with several column volumes of distilled water to remove endogenous urinary components yet retain metabolites. Metabolites of I were eluted with methanol, and the methanol was evaporated under vacuum at room temperature. The metabolites were dissolved in HPLC mobile phase prior to chromatography. Samples containing metabolites of the individual S,S()- and R,R(+)-enantiomers were chromatographed using the system previously described. Fractions were collected at 0.5-min intervals, and a small aliquot of each fraction was counted to obtain the overall metabolic pattern. Fractions containing a radioactive peak were evaporated to a small volume (<100 µl) and analyzed by electrospray ionization and MS/MS techniques.

Characterization of Metabolites by ESI MS/MS. Electrospray-ionization mass spectral analysis (ESI-MS) was done using a Finnigan-MAT TS700 (Thermo Finnigan, San Jose, CA) with 2-methoxy ethanol as a sheath liquid. The ESI MS/MS analysis was done on the TS700 using argon gas for ESI analysis and MS/MS to verify the metabolite structure.

In Vitro Studies. Metabolism of I by control dog liver microsomes. Dog livers were homogenized and microsomes prepared according to standard procedures (Omura and Sato, 1964). For generation of oxidative metabolites of I, the microsomes were diluted to a concentration of about 1 mg/ml in 0.1 M potassium phosphate buffer (pH 7.4) containing 0.1 mM EDTA and radiolabeled I (100 µM). Reactions were initiated with NADPH (1 mM). In some experiments, unlabeled racemic (final concentration 100 µM) was incubated with control dog liver microsomes in the presence or absence of NADPH for 20 min followed by the addition of a trace amount of the radiolabeled compound (and NADPH was added to incubations not yet containing this cofactor) to initiate radiolabeled metabolism and the reactions were stopped at various times with acetonitrile/methanol (9:1, v/v). Extracts were analyzed by HPLC/radioassay. Radiolabeled metabolite conversion over the time course averaged less than 12%.

Formation of the P450-metabolite complex of I by control dog liver microsomes. For in vitro formation of P450-metabolite complex, liver microsomes from control dogs were diluted with 0.1 M potassium phosphate buffer (pH 7.4) containing 0.1 mM EDTA and divided into two tubes (sample and reference). Racemic I, or the respective S,S- and R,R-enantiomers of I, were dissolved in dimethyl sulfoxide and added to the sample cuvette while dimethyl sulfoxide alone was added to the reference cuvette. The final concentration of drug was 100 µM. Preliminary experiments indicated that P450-metabolite complex formation was linear from 50 to 100 µM. Measurement of complex formation at concentrations of I below 50 µM, however, resulted in poor signal-to-noise. Therefore, 100 µM was used. Complex formation was initiated with NADPH (1 mM, added to both sample and reference cuvette), and the spectrum was scanned from 400 to 500 nm at zero time and approximately every 5 min thereafter until complex formation was maximal (20 min). Incubations were run at 37°C. Quantitation of the P450-metabolite complex was determined by A_455-490nm using the extinction coefficient of 75 mM¹cm¹.

	Results

Top Abstract Introduction Methods and Materials Results Discussion References

Characterization of Metabolites of I by ESI MS/MS. Identification of the metabolites of I by mass spectral analysis of the urinary metabolites isolated from dogs given the ¹⁴C-R,R(+)- or ¹⁴C-S,S()-enantiomer is presented in the metabolic pathway in Fig. 2. In many cases, the metabolites were present as triethylammonium adducts resulting from the paired ion buffer contained in the mobile phase HPLC used during metabolite isolation. An estimate of the relative positions of the mono-conjugates or the O-methyl sulfate and O-methyl glucuronide moieties on the respective hydroxyl groups of the catechol derivative of I were determined by the HLPC characteristics of each aglycone after hydrolysis compared with that of the authentic O-methyl reference standards, which were methylated at either the meta or ortho hydroxyl positions of the catechol.

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Proposed metabolic pathways of I in dog.

Metabolite Profiles in Urine After Administration of the ¹⁴C-Enantiomers of I. The urinary metabolite profiles from dogs given the ¹⁴C-enantiomers of I were virtually identical to those observed in dogs given the ¹⁴C-pseudoracemates of I. In addition, the effect of various hydrolysis treatments on the urinary metabolite profiles were similar whether the animals were dosed with the ¹⁴C-enantiomers or the ¹⁴C-pseudoracemates of I. These results indicated that the presence of either enantiomer did not affect the metabolite product specificity of the antipole.

Metabolism and Disposition of the ¹⁴C-Pseudoracemates of I. Excretion of Radioactivity.

Metabolite Profiles in Urine, Feces, and Bile. A summary of the amounts of parent drug and metabolites excreted in the urine and feces is presented in Table 1.

Pharmacokinetic Study of Racemic I After Single and Multiple Dosing. A comparison of the pharmacokinetic parameters in dogs after single and multiple doses of I are presented in Table 2 and Figs. 3, A and B. 

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Plasma concentration-time curves after single or multiple doses of I in dogs; curves A and B correspond to dogs 5 and 4, respectively.

Formation of a Liver P450-Metabolite Complex of I. In vivo.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   In vivo formation of a P450-metabolite complex in dog liver after single or multiple doses of I. Livers were removed from dogs after a single or 14 daily doses of I. Microsomes were prepared and diluted in 0.1 M potassium phosphate buffer (pH 7.4) containing 0.1 mM EDTA and divided equally into sample and reference tubes. Formation of the P450-metabolite complex was monitored by recording the reduced minus-oxidized spectrum between 400 and 500 nm.

View larger version (38K): [in this window] [in a new window]   Fig. 5.   Amount of complexed, uncomplexed, and total P450 in the liver of dogs given single or multiple doses of I. Dose, 12.0 Mg/Kg. Animal treatment abbreviations are C-1 (control, dog-1), SD-1 (single dose, dog-1), SD-2 (single dose, dog-2), MD-1 (multiple dose, dog-1), and MD-2 (multiple dose, dog-2).

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Correlation between the amount of liver P450-metabolite complex and the plasma clearance of I in dogs given single or multiple doses of I.

In vitro. Incubation of I with control dog liver microsomes in the presence of NADPH demonstrated the formation of a P450-metabolite complex (155-262 pmol/mg). No P450 complex was formed in the absence of NADPH indicating that I does not form a complex with P450 directly but requires metabolism. Formation of the P450-metabolite complex was maximal by 20 to 25 min. Similar amounts of the P450-metabolite complex were formed after incubation of the individual R,R- and S,S-enantiomers (196 and 204 pmol/mg, respectively), indicating no stereoselectivity regarding formation of the P450-metabolite complex. Addition of 10 mM 2-methylbenzimidazole to the cuvette following complex formation resulted in the displacement of 70 to 80% of the metabolite from P450 (data not shown). These results indicated that I can form a metabolite complex with liver P450 both in vivo and in vitro and that the binding of the metabolite to P450 is reversible and not stereoselective.

Metabolism of ABT-200 by Dog Liver Microsomes. Incubation of I with dog liver microsomes in the presence of NADPH generated only the catechol metabolite. There was a significant decrease in the rate of catechol formation during the 30-min incubation period, decreasing by an average of 62%. The rate of catechol formation was also compared in dog liver microsomes preincubated with unlabeled I in the presence or absence of NADPH for 20 min prior to the addition of a trace amount of radiolabeled I (to initiate radiolabeled metabolism) and NADPH (to incubations not yet containing the cofactor). Preincubation of I with NADPH and microsomes decreased the initial rate of catechol formation by 48 to 60% compared with preincubation without NADPH.

	Discussion

Top Abstract Introduction Methods and Materials Results Discussion References

This study was undertaken to test the hypothesis that the considerable decrease in the clearance of I during chronic dosing was due to the formation of an inhibitory methylenedioxyaryl metabolite complex with the P450s that metabolize the compound. In support of this, a small amount of the P450-metabolite complex was formed in dogs within 24 h after a single dose of I, whereas considerably more of the complex was found after 14 days of treatment. Since the study design enabled the determination of the plasma pharmacokinetics of I and the formation of a P450-metabolite complex from the same animals, it was possible to evaluate the correlation between these two parameters. Although there was a 30-fold variation in the plasma AUC or clearance of I between animals and a 12-fold variation in the amount of the liver P450-metabolite complex formed between dogs, there was a positive correlation between these pharmacokinetic parameters and the amount of the P450-metabolite complex formed in the livers of each animal (Fig. 6). Studies with control dog liver microsomes indicated that the in vitro formation of the P450 complex with I was NADPH-dependent and reversible with 2-methylbenzimidazole. This indicated that I did not form a complex with P450 directly but required oxidative metabolism and that the binding was not covalent but reversible. The rate of conversion of I to the catechol decreased with time in control dog liver microsomes. In a separate set of experiments, preincubation of unlabeled I with control dog liver microsomes and NADPH for 20 min followed by the addition of a tracer amount of radiolabeled I resulted in a significant decrease in the rate of catechol formation compared with the rate without preincubation. These results indicated that I appears to be a mechanism-based inhibitor and that formation of the complex inhibits the rate of catechol formation. Currently, however, it is not known whether the mechanism of catechol formation is tightly coupled to P450-metabolite complex formation, but the data suggest that other mechanisms may be involved. Preliminary in vitro experiments indicated that even when formation of the P450-metabolite complex is maximal, the rate of catechol formation is still about 40 to 50% of the initial rate. The fact that both the R,R- and S,S-enantiomers formed similar amounts of the P450-metabolite complex in vitro was in agreement with the nonstereoselective accumulation of I during multiple dosing in dogs. Whereas the S,S()-enantiomer predominated in the plasma after administration of racemic I, the nonlinear accumulation was observed with both enantiomers. Preliminary studies in human liver microsomes suggest that CYP3A4 is the major isozyme involved in metabolite complex formation of I.

The pharmacokinetic results in this study were consistent with previous preliminary investigations of I in dogs, which demonstrated considerably higher plasma levels of I after multiple dosing compared with a single dose. In the present study, the plasma area under the curve values in dogs given multiple doses of I averaged 4-fold greater than in the same animals given a single dose. Correspondingly, there was a 9-fold decrease in the apparent plasma clearance after multiple dosing compared with a single dose of I. Since the clearance of I in the dog was due entirely to metabolism, the large decrease in drug clearance during chronic dosing was predominantly due to a decline in the rate of metabolism.

Figure 2 depicts the proposed metabolic pathways of I. The first metabolic step involving the metabolism of I was oxidation at the methylenedioxy region to form the catechol derivative. The catechol then underwent O-methylation, sulfation and/or glucuronidation. Some stereoselectivity in the metabolic patterns was also apparent, with the O-methyl/sulfate metabolite formed only by the S,S()-pseudoracemate of I. In addition to the formation of mono-glucuronide and sulfate conjugates, di-conjugates (glu/glu and glu/sulfate) were also confirmed by mass spectral analysis. The in vivo formation of di-conjugated metabolites has been demonstrated with steroids and bile acids, including glucuronide/glucuronide, sulfate/sulfate and sulfate/glucuronide conjugates (Billing et. al., 1957). The di-glucuronic acid conjugate of p-hydroxybenzoic acid has also been reported in the urine of dogs after administration of the aglycone (Quick, 1932).

The mass spectral data demonstrating the formation of two di-glucuronic acid conjugates of the catechol derivative of I was somewhat perplexing. These two conjugates, designated as metabolites "A" and "B", exhibited different chromatographic retention times. Thus, these metabolites were clearly distinct entities with no apparent cross contamination prior to mass spectral analysis. Metabolite B was hydrolyzable by -glucuronidase, whereas metabolite A was resistant to hydrolysis by -glucuronidase, acid or base. A possible explanation for these results may be that metabolite A is a intramolecular rearrangement of the glucuronic acid sugar moieties. This has been shown to occur with some ester glucuronides in which the glucuronic acid attached to the aglycone at the control dog-1 position can rearrange or migrate and attach to the 2-, 3-, or 4-position of the sugar (Dickinson and King, 1991). These positional isomers generally have shorter retention times on reverse-phase HPLC and are resistant to -glucuronidase hydrolysis, as was metabolite A for both pseudoracemates in the present study. Although the intramolecular rearrangement of ester glucuronides is well documented, a similar occurrence for ether glucuronides has not been reported.

The effects of methylenedioxyaryl compounds that form metabolite complexes with P450 are frequently biphasic when administered to animals in vivo. Their inhibitory effects on drug metabolism are often followed by an increase in mixed function oxidase activity due to the induction of liver P450 (Karmienski et al., 1971; Kumagai et al., 1994). In the present study, one of the two dogs dosed chronically with I demonstrated P450 levels that were significantly above those in the control animal. Although this was not investigated, there may exist the potential for induction of P450 by I.

Several, but not all, methylenedioxyaryl-containing compounds have been shown to form metabolite complexes with P450 (Franklin, 1971). Structural requirements appear to include both a methylenedioxyaryl configuration and compound lipophilicity. I fits these criteria. The NADPH-dependence of complex formation indicated that oxidative metabolism of I is required. Based on previous mechanistic studies of P450 complex formation with isosafrole, another lipophilic methylenedioxyphenyl compound, the formation of a P450-metabolite complex with I may involve oxidation of the methylenedioxy bridge carbon of ABT-200, possibly through a carbene intermediate, which then binds to the heme iron of P450 forming a complex as proposed by Dickins et al. (1979).

Collectively, these results demonstrate that some of the liver P450s that are responsible for the metabolism of I form a metabolite complex after administration in dogs. Formation of these complexes increases during chronic dosing and inhibits the enzyme(s) from further metabolism of I, resulting in a marked decrease in drug clearance and nonlinear accumulation in the plasma.