Abstract
Methamphetamine (MeAmp) abuse has recently experienced a resurgence and approaches to the treatment of its addiction similar to those used with cocaine have been considered. As the treatment regimes are likely to use drugs whose metabolism is related to that of MeAmp, studies were initiated to establish the enzymology of the fate of MeAmp.
This report describes investigations of the role of CYP2D6, the human isoform of the enzyme that catalyzes debrisoquine hydroxylation, in the 4-hydroxylation and N-demethylation of MeAmp. The results of studies with human liver microsomes including those from a genetically poor metabolizer with respect to CYP2D6, showing correlation between MeAmp and metoprolol hydroxylation and MDMA demethylenation, were consistent with a major involvement of CYP2D6 in the aromatic 4-hydroxylation of MeAmp. This was confirmed by studies with recombinant CYP2D6 expressed in yeast, which was also shown to effect the N-demethylation of MeAmp. The rate of the 4-hydroxylation reaction was substantially slower than the demethylenation of MDMA. In contrast to MeAmp, MDMA was notN-demethylated by CYP2D6. Since CYP2D6 participates in the major steps of MeAmp metabolism, pharmacokinetic interactions are likely with other drug substrates proposed for the treatment of MeAmp addiction. Furthermore, the genetic polymorphism associated with the enzyme could manifest itself in abnormal responses to MeAmp.
MeAmp (fig. 1) has recently reemerged as a significant drug abuse problem in the United States (1, 2). As approaches to treatment of addiction to this compound and its pharmacological congeners are being considered, characterization of the enzymatic nature of its metabolism becomes an important consideration. In early studies in humans, Caldwellet al. (3) showed that the urinary excretion products of this compound included the 4-hydroxy derivative (4-OH MeAmp, fig. 1) and the N-demethylation product, amphetamine (Amp, fig. 1), with 4-OH MeAmp as the dominant metabolite, representing almost 50% of all metabolites excreted (3). Metabolic oxidation of amphetamine derivatives has led to pharmacologically active species in the case of Amp (4), 4-Allyloxy MeAmp (5), and MDMA (fig. 1) (6) so the enzymatic nature of the reaction is of interest in terms of individual differences in drug effects and in potential drug interactions. As studies of the metabolism of MDMA (7), Amp (8), 4-Allyloxy MeAmp (5), and 4-Methoxy Amp (9) had shown that the oxidation of the aromatic ring and its substituents were catalyzed by CYP2D isozymes, studies investigating the role of the human isoform of this enzyme in MeAmp metabolism were initiated and the results are reported here.
Metabolic reactions of MDMA and MeAmp.
The studies used human microsomes in correlative studies with established substrates and recombinant CYP2D6, expressed in Saccharomyces cerevisae (10). The results show that 4-hydroxylation of MeAmp is indeed catalyzed by CYP2D6, the human isoform of CYP2D, and that this enzyme effects itsN-demethylation as well. Kinetic studies with the expression system preparation indicated that while both reactions are catalyzed by CYP2D6, the 4-hydroxylation of MeAmp is much slower than MDMA demethylenation.
Materials and Methods
Chemicals.
(+) and (−) MDMA hydrochloride and (+) MeAmp hydrochloride were obtained from the Research Technology Branch of the National Institute on Drug Abuse (Rockville, MD). (−) MeAmp hydrochloride was prepared from (−) ephedrine using the procedure of Kishi et al.(11). DHMA, the product of MDMA demethylenation, was prepared by modifying the procedure of Smissman and Borchardt (12).N-formyl 4-hydroxyamphetamine, prepared by procedures described earlier (5), was reduced toN-[2H2]-Methyl-4-hydroxy-phenylisopropylamine (2H2-4-OH MeAmp) with LiAl2H4. [2H3]-Amp was synthesized by the procedure of Gal (13). All other chemicals were obtained from commercial sources.
Source of CYP2D Enzymes and Preparation of Microsomes.Human livers.
Samples of seven human livers were used with the approval of the Hospital Ethics Committee. Six of the livers (HL4, 5, 6, 7, 9, and 10) were from extensive CYP2D6 metabolizers and the seventh (HL3) from a poor CYP2D6 metabolizer. Clinical details of the donors are described elsewhere (14).
Yeast.
The heterologous expression of human CYP2D6 in the yeastSaccharomyces cerevisiae and the preparation of microsomes have been described previously (10). The CYP2D6 wild-type cDNA employed has a valine and not methionine at position 374 (15).
Incubation Conditions.
Incubation conditions for MDMA and MeAmp differed because of the instability of the MDMA metabolite, DHMA, in the presence of reactive oxygen species (6). The incubations of MDMA therefore used a reaction mixture consisting of 100 mM HEPES buffer (pH 7.4), superoxide dismutase (100 units), an NADPH generating system (0.5 mM NADP+, 8mM glucose-6-phosphate, 5 mM MgCl2, and 1 unit of glucose-6-phosphate dehydrogenase), the microsome preparation (containing 33-65 pmol cytochrome P450), and substrate in a final volume of 1.0 ml (16). The reactions were initiated by the addition of the microsomal preparation and conducted at 37°C. Reactions were terminated by the addition of 0.5 ml of 7.5% (w/v) perchloric acid (HClO4). The mixture was centrifuged and the supernatant refrigerated until HPLC analysis for DHMA.
The methamphetamine incubation mixture contained substrate, microsomal preparation, an NADPH-generating system consisting of 0.5 mM NADP+, 8 mM glucose-6-phosphate, 5 mM MgCl2, and 1 unit of glucose-6-phosphate dehydrogenase in a final volume of 1 ml of 0.1 M HEPES, pH 7.4. The reaction was initiated as indicated above by the addition of the enzyme preparation and the incubations were conducted for 5 min at 37°C. The metabolic reactions were terminated by addition of 0.5 ml of 7.5% (w/v) HClO4, containingN-[2H2]-methyl-4-hydroxy-phenylisopropylamine ([2H2]-4-OH MeAmp, 99.9% isotopic purity) and [2H3]amphetamine ([2H3]Amp, 99.9% isotopic purity) as internal standards.
Interference of the analysis of 4-OH MeAmp by the high concentrations of MeAmp that occur under the conditions used in the study required that a two-stage extraction and analysis procedure be used in which MeAmp and Amp were first extracted with dichloromethane and the 4-OH MeAmp in the aqueous residue extracted subsequently with an isopropanol-dichloromethane mixture. The two extracts were processed separately and analyzed in two different assays. In the first extraction, the reaction mixtures were centrifuged (13,500 g for 5 min) and 1 ml aliquots of the supernatants extracted by addition of 1 ml of 10% Na2CO3, 5 ml of dichloromethane, and shaking. The dichloromethane layer was collected, evaporated to about 100 μL, and treated with 100 μL of TFAA at room temperature for 15 min. Then excess TFAA was evaporated under nitrogen and the residue reconstituted in 120 μL of acetonitrile for GC/MS analysis of Amp.
To assay for 4-OH MeAmp, 1–2 g of NaCl was added to the aqueous residue from above which was then re-extracted with 6 ml of isopropanol: CH2Cl2 (volume ratio 1:4). The organic layer was evaporated to dryness under a stream of nitrogen and the residue reconstituted with 50 μL acetonitrile, then treated with 100 μL of TFAA for 15 min at 60°C. The residual TFAA was removed by evaporation under nitrogen and the residue reconstituted in 150 μL of acetonitrile.
GC/MS Analysis of Amp and 4-OH MeAmp.
A 1 μL aliquot of the acetonitrile solution containing the dichloromethane extracts was injected into the GC/MS, which consisted of a Hewlett Packard 5971A system equipped with a methyl silicone (0.2 mm i.d., 0.33 micron film thickness, and 12 m length) capillary column. The gas chromatograph temperature program began at 65°C and was increased to 195°C at a rate of 25°C/min. Under these conditions, the retention time of the N,O,-bis trifluoroacetyl derivative of 4-OH MeAmp was 5.6 min and that ofN-trifluoroacetylamphetamine was 4.0 min. The mass spectrometer was operated in its specific ion detection mode. Quantitation of Amp was achieved by comparing the mass spectrometer responses at M/Z = 140 and 143 (internal standard). The internal standard used in the assay was [2H3]Amp of 99.9% isotopic purity.
Assay of 4-OH MeAmp used the acetonitrile solution from the isopropanol-dichloromethane extract, and quantitation was achieved by comparing the mass spectrometer responses at m/z = 154 and 156 (internal standard, [2H2]-4-OH MeAmp). As these M/Z values are the base peaks for MeAmp as well, at high substrate concentrations, the signal from MeAmp was found to interfere with the measurement of low levels of 4-OH MeAmp. For this reason, the MeAmp was initially removed in an initial extraction. When performed in this manner, the assay was linear over a range of 25 pmoles to greater than 2 nmoles (r2 > 0.99, with 95% confidence limits of less than 10% over the range examined). The variation in slopes between assays was less than 1%.
Analysis of DHMA.
The concentration of DHMA, the demethylenation product of MDMA, was determined in a 20 μL aliquot of the quenched incubation mixture by HPLC analysis using a Biophase ODS column, 4.6 × 250 mm (Bioanalytical Systems Inc., West Lafayette, IN) and an electrochemical detector (LC-4; Bioanalytical Systems Inc.). The mobile phase was 0.1 M citrate buffer (pH 3.5) containing 1 mM octyl sodium sulfate, acetonitrile and methanol in a volume ratio of 8:1:1 at a flow rate of 0.7 ml/min. The working electrode (LC-4) was set at 0.7 V relative to a Ag/AgCl reference electrode. The retention time for the catechol was 11.3 min. Concentrations were determined by comparison with a standard curve generated from authentic compound carried through the analysis (16).
Analysis of α-Hydroxymetoprolol (α-HM).
The preparation of samples and the chromatography were carried out using modifications of previously described methods (17). After centrifugation of the precipitated microsomal protein, 0.45 ml of the supernatant was transferred to glass tubes containing 0.5 ml of 4 M NaOH and 5 ml of methyl-t-butyl ether. The tubes were vortex mixed for 2 min and then centrifuged at 1,500 rpm for 10 min. The organic layer was transferred to a glass test tube and evaporated to dryness under a nitrogen stream at room temperature. The residue was dissolved in 200 μL of mobile phase which had been adjusted to pH 3, of which 100 μL were injected onto the HPLC column.
Metabolites were separated by reversed phase chromatography using a Biophase ODS 5 μm column (4.6 × 250 mm, Bioanalytical Systems Inc.) and a mobile phase consisting of water/acetonitrile (88:12, v/v) containing 1% (w/v) triethylamine adjusted to pH 3 with orthophosphoric acid at flow rate of 1.0 ml/min. The metabolites were detected using a Hewlett Packard (Wilmington, DE) 1064 fluorescence detector with an excitation wavelength of 193 nm and an emission wavelength of 280 nm. The concentration of α-HM was estimated on the basis of peak area ratio from the standard calibration curve. Retention times for α-HM and internal standard, guanoxane, were 8.2 min, and 16.5 min, respectively.
Kinetics.
Estimates of the enzymatic parameters for MeAmp and MDMA metabolism were obtained by nonlinear regression procedures using the AR program, a regression program, of the BMDP Statistical Software System (18).
Results
Six preparations of microsomes from human livers were compared for their ability to catalyze MeAmp metabolism (4-hydroxylation andN-demethylation) at a substrate concentration of 100 μM (fig. 2). The activities of MDMA demethylenation, at a substrate concentration of 10 μM (7), and metroprolol α-hydroxylation (19), at a substrate concentration of 40 μM, of these preparations were determined also. The correlations between the amphetamine reactions and metroprolol α-hydroxylation are shown in fig. 3, and values of the correlation coefficients for all of the reactions are shown in table1. Both MDMA demethylenation and MeAmp 4-hydroxylation (the (+) isomer more than the (−)) correlated with metoprolol α-hydroxylation activity, and MDMA demethylenation and MeAmp hydroxylation correlated highly with each other. However, theN-demethylation of (+) MeAmp exhibited a lower but significant correlation with MDMA demethylenation and with MeAmp 4-hydroxylation.
Metabolism of MDMA and MeAmp by human liver microsomes.
Reaction conditions are described in Methods. The substrate concentrations were 10 μM for MDMA demethylenation and 100 μM for MeAmp N-demethylation and 4-hydroxylation. The values represent the means ± SD of three determinations.
Correlation between metoprololα-hydroxylation and MDMA, MeAmp oxidation by human livers.
The rates shown are expressed as pico mols/min·(mg protein)−1 with substrate concentrations of 10 μM (MDMA), 100 μM (MeAmp), and 40 μM (metoprolol).
Correlations between metoprolol α-hydroxylation and MeAmp and MDMA oxidation by human liver microsomes
The demethylenation of MDMA by human EM liver microsome preparations was also inhibited by quinidine (1μM), a selective and potent inhibitor of CYP2D6, to a level similar to that observed in the PM (HL3) preparation (fig. 4), consistent with previous work (7) and the metoprolol data that indicated participation by this enzyme in the reaction. These observations, indicated that CYP2D6 in human liver catalyzes MDMA demethylenation and, based on the correlative studies, the 4-hydroxylation of MeAmp. The correlation suggested also that CYP2D6 may also catalyze theN-demethylation of the (+) enantiomer of this substrate.
Effect of quinidine on MDMA demethylenation.
MDMA (10 μM) was incubated with liver microsomes from four human subjects in the presence or absence of 1 μM quinidine according to the procedures described in Methods. The values are the means ± SD of three determinations.
Subsequent incubation of MeAmp with microsomes from yeast expressing CYP2D6 revealed formation of both 4-OH MeAmp and Amp. When the kinetic behavior of (+) and (−) MeAmp was examined over a substrate range of 0.005 to 0.5 mM, kinetic patterns consistent with a single enzyme were observed (fig. 5) with the parameters shown in table 2. TheKm values for the two reactions were in the micromolar range and the Vmax values showed a different stereoselectivity, i.e. the value forN-demethylation was higher for the (+) enantiomer, whereas that for 4-hydroxylation was higher for the (−) enantiomer. Kinetic parameters for MDMA demethylenation, obtained in the same preparation, are also shown. N-demethylation of this substrate was not detected under these conditions.
Eadie-Hofstee plots of the 4-hydroxylation and N-demethylation of MeAmp enantiomers by microsomal preparations of yeast expressing CYP2D6.
The reactions were carried out as described in Methods with substrate concentration varying from 5–500 μM. The lines are generated from parameter estimates obtained by nonlinear regression analysis of the kinetic data.
Kinetic parameters of product formation from MDMA and MeAmp in microsomes from yeast expressing CYP2D6
Discussion
The studies correlating MeAmp metabolism with MDMA demethylenation and metoprolol α-hydroxylation, together with direct experiments with microsomes from yeast preparations expressing CYP2D6, show that MeAmp is a substrate that undergoes 4-hydroxylation andN-demethylation by this enzyme.
The hydroxylation of the aromatic ring of Amp is pharmacologically important, as it leads to 4-hydroxynorephedrine, a compound implicated in the tolerance to the peripheral actions of Amp (20, 21). The role of 4-hydroxylation in the central actions of Amp and MeAmp is not well established, but 4-OH Amp (22) and 4-OH MeAmp (23) accumulate in the striatum after administration of the parent drugs. The genetic polymorphism associated with CYP2D6 (17) and its role in the metabolism of other drugs such as the tricyclic antidepressants (24), serotonin uptake blockers (25), and beta blockers (17, 26) has resulted in significant individual differences in responses to some of these compounds. In addition, however, therapy with two or more CYP2D6 substrates can lead to significant bioavailability differences of each. As tricyclic antidepressants (27) and serotonin uptake blockers have been suggested as potential therapeutic approaches to MeAmp addiction, evaluation of the effectiveness of the treatment will require that pharmacokinetic interactions be considered.
Catalysis of N-dealkylation reactions by CYP2D6 has been reported for imipramine (28) β-blockers (26, 29) and the MeAmp analog, deprenyl (30), but was not a significant reaction with the ring substituted analog, MDMA. The lower correlation of CYP2D6 activity with MeAmp N-demethylation compared with MeAmp 4-hydroxylation and MDMA demethylenation (table 1) is consistent with the participation of additional cytochromes P450 in MeAmp N-demethylation. The variable nature of CYP2D6 contribution to propranololN-dealkylation has been reported earlier. The contribution of CYP2D6 to the catalysis of MeAmp N-demethylation in the liver preparations examined appears to be stereoselective, as the metabolism of the (+)-isomer correlates much more strongly with CYP2D6 activity than does the (−)-isomer. This difference is not reflected in the kinetics for N-demethylation by recombinant CYP2D6, although the reaction does favor the (+)-isomer.
N-demethylation of MeAmp and deprenyl, which have single basic nitrogen atoms, appears to be a reaction inconsistent with the observation that most reactions catalyzed by CYP2D6 occur on carbon atoms at sites 5-7Å from the basic nitrogen atom (31). To account for this inconsistency, Grace et al. (30) suggest the possibility of alternative positioning in the active site, together with significant concentrations of the free base which undergoes electron transfer (30). Since deprenyl, which is N-propargyl MeAmp, is a weaker base than MeAmp, its free base would be present at higher concentrations than that of MeAmp. Consistent with this notion, the kcat reported by them is about 10 times that found here for MeAmp. However, the stereoselectivity observed for deprenyl N-demethylation is opposite that found for MeAmp, so the positioning of the two substrates may be different. Alternatively, Guengerich (32) has suggested anN-dealkylation mechanism involving single electron transfer from the nitrogen atom, which may be operative over the range of about 5Å. A related question requiring further study is why MeAmp but not MDMA (7) undergoes significant N-demethylation by CYP2D6. One possibility may relate to the relative rates of oxidation. The high turnover of MDMA demethylenation relative to 4-hydroxylation of MeAmp may preclude alternate orientation within the active site. Although MeAmp and MDMA are both substrates for the same cytochrome P450 enzyme, the rates by which they are oxidized differ markedly, with MDMA oxidation taking place at almost 100 times the rate of MeAmp oxidation (see Vmax/Km , table 2). The basis for the difference is not clear, but oxidation of alkyl groups of ethers on the aromatic ring by CYP2D seems to be faster; we have observed that 4-alloxy MeAmp, also a substrate for CYP2D enzymes, is cleaved rapidly to 4-OH MeAmp in rat liver microsomes. The slow oxidation of the aromatic ring of MeAmp may contribute to the N-demethylation of this substrate by allowing its reorientation in the catalytic site, i.e. a form of “metabolic switching.”
Footnotes
-
Send reprint requests to: Dr. Arthur K. Cho, Department of Molecular and Medical Pharmacology, UCLA School of Medicine, 23-120 CHS, Box 951735, Los Angeles, CA 90095-1735
-
L.Y.L. was supported by a Pharmaceutical Manufacturer Foundation Predoctoral Fellowship during the course of this work.
- Abbreviations used are::
- MeAmp
- methamphetamine
- MDMA
- methylendioxymethamphetamine
- DHMA
- dihydroxymethamphetamine
- CYP2D6
- cytochrome P-450 2D6
- TFAA
- trifluoroacetic anhydride
- 4-OH MeAmp
- 4-hydroxymethamphetamine
- Received January 21, 1997.
- Accepted April 30, 1997.
- The American Society for Pharmacology and Experimental Therapeutics