Abstract
Ibogaine is a psychoactive alkaloid that possesses potential as an agent to treat opiate and cocaine addiction. The primary metabolite arises via O-demethylation at the 12-position to yield 12-hydroxyibogamine. In this report, evidence is presented that theO-demethylation of ibogaine observed in human hepatic microsomes is catalyzed primarily by the polymorphically expressed cytochrome P-4502D6 (CYP2D6). An enzyme kinetic examination of ibogaineO-demethylase activity in pooled human liver microsomes suggested that two (or more) enzymes are involved in this reaction: one with a low KMapp (1.1 μM) and the other with a high KMapp (>200 μM). The lowKMapp activity comprised >95% of total intrinsic clearance. Human liver microsomes from three individual donors demonstrated similar enzyme kinetic parameters (meanKMapp = 0.55 ± 0.09 μM and 310 ± 10 μM for low and high KMactivities, respectively). However, a fourth human microsome sample that appeared to be a phenotypic CYP2D6 poor metabolizer possessed only the high KMapp activity. In hepatic microsomes from a panel of human donors, the lowKMapp ibogaine O-demethylase activity correlated with CYP2D6-catalyzed bufuralol 1′-hydroxylase activity but not with other P450 isoform-specific activities. Quinidine, a CYP2D6-specific inhibitor, inhibited ibogaineO-demethylase (IC50 = 0.2 μM), whereas other P450 isoform-specific inhibitors did not inhibit this activity. Also, of a battery of recombinant heterologously expressed human P450 isoforms, only rCYP2D6 possessed significant ibogaineO-demethylase activity. Thus, it is concluded that ibogaineO-demethylase is catalyzed by CYP2D6 and that this isoform is the predominant enzyme of ibogaine O-demethylation in humans. The potential pharmacological implications of these findings are discussed.
Ibogaine (fig. 1) is a psychoactive alkaloid isolated from the root of Tabernanthe iboga, a shrub native to Africa. The plant has been used by native peoples for such purposes ranging from the alleviation of fatigue, thirst, and hunger to its use in greater quantities as a hallucinogen in religious rituals. More recently, ibogaine has been explored as an agent that combats the symptoms of drug withdrawal (Goutarel et al., 1993; Lotsof, 1985; Mash et al., 1996). Additionally, preclinical studies of ibogaine in rodent models of cocaine and opiate self-administration support the notion that it is an anti-addictive agent (Aceto and Harris, 1991; Cappendijk and Dzoljic, 1993; Glick et al., 1992a, 1992b; Sershen et al., 1994, 1997).
The cytochromes P450 (P450)1 constitute a large family of enzymes present throughout the animal and plant kingdoms (Nelson, 1995). The various enzymes function in the metabolism of endogenous and exogenous compounds. For the latter, it is maintained that P450-catalyzed oxidation reactions serve to impart greater polarity to xenobiotics, thereby making them more readily excreted. Also, P450-catalyzed reactions usually result in metabolites that are more amenable to conjugation reactions (e.g.glucuronidation, sulfation, etc.) that result in more polar and more readily excreted compounds. In humans, over 20 isoforms of P450 have been characterized, many of which have been shown to be involved in the oxidative metabolism of drugs and other xenobiotics. A human isoform of interest is CYP2D6. This isoform is involved in the metabolism of numerous neuroleptic agents, β-blockers, tricyclic antidepressants, and opioids (Eichelbaum and Gross, 1990; Fromm et al., 1997). Several investigators have developed a pharmacophore for CYP2D6 substrates, common elements of these being that substrates possess an amino nitrogen (or other cationic center) and a site for P450-catalyzed oxidation 5 to 7 Å away (deGroot et al., 1997; Islamet al., 1991; Koymans et al., 1992; Stroblet al., 1993). An important aspect of the CYP2D6 isoform is that it is subject to polymorphic expression, particularly in Caucasians (Gonzalez and Meyer, 1991). Approximately 5–10% of Caucasians lack a functional copy of the CYP2D6 gene and hence lack the enzyme. Such individuals are termed poor metabolizers, owing to their decreased capacity to metabolize and clear CYP2D6 substrates. Such individuals are often subject to a higher incidence of adverse drug reactions due to elevated drug concentrations. Also, for drugs that require CYP2D6-catalyzed bioactivation to a pharmacologically active metabolite (e.g. codeine → morphine), efficacy can be reduced in poor metabolizer subjects.
As ibogaine represents a potentially useful therapeutic agent in the treatment of opiate and psychostimulant addiction and opiate withdrawal, knowledge concerning the enzymes involved in the metabolism of this compound in humans is important. Thus, these experiments were undertaken to identify P450 isoforms involved in the metabolism of ibogaine to its O-demethylated metabolite, 12-hydroxyibogamine (fig. 1).
Materials and Methods
Reagents and Biological Materials.
Ibogaine and 12-hydroxyibogamine (noribogaine) were obtained from s.a. Omnichem Corp. (Belgium), and the deuterated internal standard of 12-hydroxyibogamine was obtained from the Medications Development Division of NIDA. Commercial sources were as follows: quinidine (Aldrich), sulfaphenazole and furaphylline (Ultrafine Pure Chemicals, Ltd.), and ketoconazole (Janssen Biotech NV). Human liver microsomes were prepared from human liver samples using standard procedures and were characterized for P450 isoform-specific activities using standard methods of catalytic activity measurement: tolbutamide hydroxylase for CYP2C9 (Miners et al., 1988), S-mephenytoin hydroxylase for CYP2C19 (Meier et al., 1985), bufuralol 1′-hydroxylase for CYP2D6 (Kronbach et al., 1987), testosterone 6β-hydroxylase for CYP3A (Sonderfan et al., 1987), and phenacetin O-deethylase for CYP1A2 (Butleret al., 1989). Heterologously expressed P450 isoforms were obtained from either Gentest Corp. (Woburn, MA) or the Molecular Genetics Department, Pfizer Central Research (Groton, CT). Assay for protein was accomplished using the BCA assay kit (Pierce) using bovine serum albumin as a standard, and assay for P450 was conducted using a standard method (Omura and Sato, 1964). Specific content of human liver microsomal preparations were 0.31, 0.11, 0.20, and 0.23 nmol P450/mg microsomal protein for HL-1000–3, HL-1021, HL-1028, and HL-1032, respectively. The pooled human liver microsome preparation consisted of equal mixtures of preparations from ten individual donors (including none of the four individual samples listed above). The P450 content was 0.21 nmol/mg microsomal protein.
Assay of Ibogaine O-Demethylase Activity.
Incubation mixtures contained liver microsomes (0.5 mg protein/ml), ibogaine (0.1–500 μM), MgCl2 (3.3 mM), and NADPH (1.3 mM) in a total volume of 1.0 ml of potassium phosphate buffer (25 mM, pH 7.5). Incubations were commenced with the addition of NADPH and shaken open to air in a water bath set at 37°C. Initial time course experiments demonstrated reaction velocity linearity out to 20 min and reaction velocity linearity with protein concentrations of up to 2.0 mg/ml. Thus, all subsequent experiments were conducted with an incubation time of 20 min or less and protein concentrations below 2.0 mg/ml. Incubations were terminated with the addition of 2 ml of ice-cold Na2CO3, pH 10, and were frozen prior to analysis. 12-Hydroxyibogamine concentrations were determined using a method involving extraction, chemical derivatization, and GC-MS as previously described (Hearn et al., 1995) or using an HPLC-MS method as described below when additional assay sensitivity was needed. Incubations using recombinant heterologously expressed P450 isoforms (rCYP) were conducted in a similar manner except that microsomal protein concentrations were adjusted for each expressed isoform to account for differences in expression level. Protein concentrations ranged between 0.07 mg/ml (for CYP2D6) to 0.62 mg/ml (for CYP2C19). For the rCYP2D6 substrate saturation experiment, the incubation volume was increased to 5.0 ml to permit quantitation of product in incubations containing low ibogaine concentrations. Inhibition experiments were conducted using pooled human liver microsomes in the presence of quinidine (0.1–3.0 μM), ketoconazole (0.1–3.0 μM), sulfaphenazole (0.1–3.0 μM), or furaphylline (1.0–100 μM).
HPLC-MS Assay of 12-Hydroxyibogamine.
Samples were extracted as previously described (Hearn et al., 1995), and the evaporated ethyl acetate extract was reconstituted in 75 μl of HPLC mobile phase. The mobile phase composition was 20 mM CH3COOH (adjusted to pH 4.0 with NH4OH) in 32% CH3CN. Samples (50 μl) were injected onto a Waters Symmetry C18 5μ (3.9 × 150 mm) column at a flow rate of 0.8 ml/min. The effluent was introduced into an APCI source of a Sciex API100 mass spectrometer operated in the positive ion mode. The source temperature was 500°C, and the orifice voltage was 35 V. Other settings and state file parameters were adjusted to optimize the signal. Detection was accomplished by selected ion monitoring of m/z 297 (12-hydroxyibogamine) and m/z 299 ([2H2]12-hydroxyibogamine internal standard). The analytes eluted at 1.8 min. The dynamic range of this assay was from 3.0 to 1000 ng/ml using a 1-ml sample aliquot.
Results and Discussion
Human liver microsomes catalyze the O-demethylation of ibogaine. In initial time course experiments, both the consumption of ibogaine and the formation of 12-hydroxyibogamine were measured. TheO-demethylation reaction was the major route of metabolism. Of the ibogaine consumed, 75–80% was accounted for as 12-hydroxyibogamine (data not shown).
Substrate saturation experiments for ibogaine O-demethylase activity conducted in pooled human liver microsomes suggested the presence of two kinetically distinguishable activities as observed on an Eadie-Hofstee plot (fig. 2). The lowKMapp activity contributed the majority of intrinsic clearance with kinetic parameters of 1.1 μM and 106 pmol/min/mg microsomal protein for KMappand Vmax, respectively (table1). For the highKMapp activity, values of 250 μM and 1260 pmol/min/mg microsomal protein were obtained forKMapp and Vmax, respectively. Scaling the sum of the in vitroCl′int values to reflect an in vivoCl′int (Obach et al., 1997) suggests that ibogaine is a high intrinsic clearance compound (in vivo Cl′int = 90 ml/min/kg) in humans.
Human liver microsomes from four individual donors were also examined to assess interindividual variability; these included one sample from a putative CYP2D6 poor metabolizer (HL-1032). For three of the four, biphasic enzyme kinetics were observed (as assessed through Eadie-Hofstee plots of the data) with low and high meanKMapp values of 0.55 and 310 μM (table1). The fourth donor sample (HL-1032) lacked the low meanKMapp activity.
Upon observation that the low KMappactivity contributed the major portion of ibogaine intrinsic clearance through the O-demethylase pathway, subsequent experiments designed to identify the P450 isoform responsible for this activity were conducted at low substrate concentrations (1.0 μM). Measurement of ibogaine O-demethylase in liver microsomes from a panel of 19 individual donors resulted in activities ranging up to 85 pmol/min/mg protein. Attempts at correlation of these activities with standard P450 isoform-specific activities (CYP1A2 phenacetinO-deethylase, CYP2C9 tolbutamide hydroxylase, CYP2C19 mephenytoin hydroxylase, CYP2D6 bufuralol 1′-hydroxylase, and CYP3A testosterone 6β-hydroxylase) demonstrated a correlation only with CYP2D6-catalyzed bufuralol 1′-hydroxylase (r2 = 0.711; fig.3, table2).
P450 isoform-specific inhibitors furaphylline (CYP1A2), ketoconazole (CYP3A), sulfaphenazole (CYP2C9), and quinidine (CYP2D6) were examined for their ability to inhibit ibogaine O-demethylation at a substrate concentration of 1.0 μM. Of the four inhibitors, only quinidine inhibited this reaction, with an IC50value of 0.2 μM. This inhibitory potency of quinidine is consistent with inhibition of CYP2D6 (Strobl et al., 1993). IbogaineO-demethylase activities at quinidine concentrations above 1 μM were under the limit of detection, indicating that CYP2D6 is exclusively involved in this activity in human liver microsomes at a low ibogaine concentration.
To confirm the involvement of CYP2D6 in the metabolism of ibogaine to 12-hydroxyibogamine, several heterologously expressed rCYP isoforms were examined for this activity. Of the isoforms examined, rCYP2D6, rCYP3A4, and rCYP2C19 demonstrated measurable ibogaineO-demethylase activity (table3). Substrate saturation experiments were conducted with rCYP2D6 to determine whether theKMapp was close to the lowKMapp value measured in human liver microsomes (fig. 4, table 1). TheKMapp value of 0.19 μM was somewhat lower than that measured in human liver microsomes. However, it is not uncommon to measure slightly disparateKMapp values in heterologously expressed rCYP isoforms and liver microsomes. It is interesting to note that theKMapp values measured for ibogaineO-demethylase in this work are close to the reportedKi value for ibogaine on CYP2D6-mediated bufuralol 1′-hydroxylase in human liver microsomes (0.4 μM;Fonne-Pfister and Meyer, 1988). The Vmaxvalue of 12.1 pmol/min/pmol P450 is substantially higher than corresponding low Vmax values in human liver microsomes after normalization to P450 content (range of 0.34 to 0.64 pmol/min/pmol P450). This is consistent with the notion that CYP2D6 comprises a small portion (<5%) of total liver microsomal P450 content.
The evidence presented strongly supports the notion that theO-demethylation of ibogaine is primarily catalyzed by CYP2D6 in human liver microsomes. All three approaches (correlation analysis, P450-specific inhibitors, heterologously expressed rCYP isoforms) provided results that were in agreement. The identity of the highKMapp P450 isoform was not determined but could be CYP3A4 or CYP2C19, as these two heterologously expressed enzymes were observed to catalyze the reaction to a small extent. The importance of CYP2D6 has several important potential implications for the clinical pharmacology of this agent. First, because this major route of ibogaine metabolic clearance is mediated by the CYP2D6 isoform, pharmacogenetic differences in the response to this compound are expected to be observed. CYP2D6 poor metabolizers would be expected to be subject to greater ibogaine exposures, especially after oral administration, than extensive metabolizers. Preliminary experiments in humans suggest that systemic exposure to ibogaine and 12-hydroxyibogamine are substantially different between CYP2D6 extensive and poor metabolizer subjects (D. Mash, unpublished observations). Whether this difference will be great enough to elicit adverse drug reactions in poor metabolizers administered doses deemed efficacious in extensive metabolizers remains to be determined. This is further complicated by the observation that ibogaine and 12-hydroxyibogamine demonstrate some differences in pharmacological profile (Staley et al., 1996). Furthermore, owing to the fact that CYP2D6-catalyzed ibogaine O-demethylase activity is characterized by a low KMapp value, the compound could exhibit oral dose supraproportional exposure due to saturation of first-pass metabolism. Such a phenomenon has been observed with other CYP2D6 substrates such as propafenone (Siddowayet al., 1987) and paroxetine (Sindrup et al., 1992). However, other factors such as the dose, extent of plasma binding, and absorption rate constant also contribute to this phenomenon, so that it cannot be predicted fromKMapp values alone.
Many CYP2D6 substrates are subject to drug interactions. For example, quinidine, a potent CYP2D6 inhibitor, can inhibit the metabolism of the CYP2D6 substrate desipramine in vivo to the extent that extensive metabolizer subjects receiving quinidine demonstrate desipramine pharmacokinetics phenotypically similar to those exhibited in CYP2D6 poor metabolizers (Brosen et al., 1987). The common antidepressant fluoxetine, also a potent inhibitor of CYP2D6 activity, alters the metabolism of dextromethorphan in human subjects (Otton et al., 1993). In consideration that the potential patient population that would benefit from the therapeutic effects of ibogaine are likely to have taken other medications (prescription and/or illicit) that are CYP2D6 substrates and inhibitors, the potential for drug interactions with ibogaine is increased.
The product of ibogaine O-demethylation, 12-hydroxyibogamine, has been demonstrated to possess pharmacologic activity. In vitro radioligand binding assays conducted to identify the potency and selectivity profiles for ibogaine and 12-hydroxyibogamine have demonstrated that the metabolite has a binding profile that is similar, but not identical to, the parent drug (Pablo and Mash, 1998; Staley et al., 1996). 12-Hydroxyibogamine demonstrated the highest potency values at the cocaine recognition site on the serotonin transporter (Mash et al., 1995a; Staleyet al., 1996). Ibogaine and 12-hydroxyibogamine were equipotent at vesicular monoamine and dopamine transporters, whereas the metabolite demonstrated higher affinity at the kappa-1 and mu opioid receptors and lower affinity at the NMDA receptor complex (Mashet al., 1995b; Pablo and Mash, 1998; Staley et al., 1996). The desmethyl metabolite has been shown recently to be a full agonist at the mu opioid receptor (Pablo and Mash, 1998). Thein vivo activity of the metabolite as a full mu agonist may explain the ability of ibogaine to block the acute signs of opiate withdrawal in humans and its suppressive effects on morphine self-administration in rodents. Because the precise mix of molecular targets important for the anti-addictive effects are not definitively known, the relative contributions of ibogaine and 12-hydroxyibogamine to the actions of ibogaine in vivo has yet to be well established at the biochemical level. However, because CYP2D6 has been demonstrated to be present in the brain (Tyndale et al., 1991), it is compelling to hypothesize that some or all of the CNS activity of ibogaine may be the result of 12-hydroxyibogamine generated in situ in the brain. (Such a hypothesis also exists for the CYP2D6-catalyzed metabolism of codeine to the active metabolite morphine [Sindrup et al., 1996]). This effect would have important implications for the pharmacological activity of ibogaine in CYP2D6 extensive and poor metabolizers. Regardless of this hypothesis, the evidence presented in this report suggest that the CYP2D6 phenotype may prove to be an important determinant in the clinical pharmacology of ibogaine and that it may be necessary to determine CYP2D6 metabolizer status in subjects administered this compound.
Acknowledgments
The authors extend their gratitude to Dr. Donald Tweedie and associates (Drug Metabolism Department, Pfizer) for generation and characterization of the human liver microsomes used in these experiments and to Dr. Stafford McLean (Neuroscience Department, Pfizer) for initial discussions of this work. D. C. M. and J. P. are supported by the Addiction Research Fund.
Footnotes
-
Send reprint requests to: Dr. R. Scott Obach, Department of Drug Metabolism, Pfizer Central Research, Groton, CT 06340.
- Abbreviation used is::
- P450
- cytochrome P450
- Received January 19, 1998.
- Accepted April 13, 1998.
- The American Society for Pharmacology and Experimental Therapeutics