Abstract
Different pharmacokinetic properties are known for the two enantiomers of the entactogen 3,4-methylendioxy-methamphetamine (MDMA), most likely due to enantioselective metabolism. The aim of the present work was 1) the investigation of the main sulfotransferases (SULT) isoenzymes involved in the sulfation of the main MDMA phase I metabolites 3,4-dihydroxymethamphetamine (DHMA) and 4-hydroxy-3-methoxymethamphetamine (HMMA) and 2) the evaluation of a possible enantioselectivity of this phase II metabolic step. Therefore, racemic DHMA and HMMA were incubated with heterologously expressed SULTs, and quantification of the sulfates by liquid chromatography-high-resolution mass spectrometry was conducted. Because separation of DHMA and HMMA sulfate could not be achieved by liquid chromatography, enantioselective kinetic parameters were determined using the substrate-depletion approach with enantioselective quantification of substrate consumption by gas chromatography-negative ion chemical ionization mass spectrometry. SULT1A1 and SULT1A3 catalyzed sulfation of DHMA, and SULT1A3 and SULT1E1 catalyzed sulfation of HMMA. SULT1A1 and SULT1E1 revealed classic Michaelis-Menten kinetics, whereas SULT1A3 kinetics showed deviation from the typical Michaelis-Menten kinetics, resulting in a concentration-dependent self-inhibition. SULT1A3 showed the highest affinity and capacity of the SULT isoforms. Marked enantioselectivity could be observed for S-DHMA sulfation by SULT1A3 and in human liver cytosol, whereas no differences were observed for HMMA sulfation. Finally, comparison of Km and Vmax values calculated using achiral product formation and chiral substrate depletion showed good correlation within 2-fold of each other. In conclusion, preferences for S-enantiomers were observed for DHMA sulfation, but not for HMMA sulfation.
Introduction
The chiral compound R,S-3,4-methylenedioxy-methamphetamine [R,S-MDMA; R,S-N-methyl-1-(3,4-methylenedioxyphenyl)propane-2-amine], also known as ecstasy, is known as a very popular drug of abuse, but it is also associated with damage of serotonergic neurons (Kalant, 2001; de la Torre and Farré, 2004; Monks et al., 2004; Easton and Marsden, 2006). Metabolism of MDMA may play a role in this neurotoxicity (Miller et al., 1997; Bai et al., 1999; Mueller et al., 2009). As shown in Fig. 1, one major pathway of MDMA includes cytochrome P450 (P450)-catalyzed O-demethylation to 3,4-dihydroxymethamphetamine (DHMA), followed by O-methylation by catechol-O-methyltransferase (COMT) mainly to 4-hydroxy-3-methoxymethamphetamine (HMMA) and conjugation of DHMA and HMMA by sulfotransferases (SULTs) (Maurer, 1996; Maurer et al., 2000; de la Torre et al., 2004). In urine samples of recreational MDMA users, more than 90% of DHMA and HMMA are excreted as conjugates, with sulfates present in higher concentrations (Shima et al., 2008; A. E. Schwaninger, M. R. Meyer, A. J. Barnes, E. A. Kolbrich-Spargo, D. A. Gorelick, R. S. Goodwin, M. A. Huestis, and H. H. Maurer, manuscript in preparation). For the two enantiomers of MDMA, differences in their dose-response curves and in their in vivo kinetics were observed (Fallon et al., 1999; Kalant, 2001; Kraemer and Maurer, 2002; Pizarro et al., 2004; Peters et al., 2005). Enantioselective metabolism is the most likely explanation for the enantioselective pharmacokinetics of MDMA and was observed in vitro for P450, COMT, and UDP-glucuronyltransferase (UGT) metabolism (Meyer et al., 2008; Meyer and Maurer, 2009; Schwaninger et al., 2009).
Enzyme kinetic data for metabolic steps represent important parameters in understanding (chiral) drug clearance. Conventional determination of enzyme kinetic parameters is based on assessing the product formation rate by quantification of the formed metabolite. Therefore, availability of metabolite reference standards and suitable analytic techniques for chromatographic separation of the metabolites are necessary. An alternative was recently described for P450 enzymes by measuring the substrate-depletion rate and was shown to be applicable for various typical P450 probe substrates (Obach and Reed-Hagen, 2002; Youdim and Dodia, 2010).
The aims of the present study were the identification of the human SULT isoforms involved in the formation of DHMA and HMMA sulfates and to elucidate enantioselective enzyme kinetic data in pooled human liver cytosol (pHLC) and heterologously expressed SULT isoforms with the substrate-depletion approach.
Materials and Methods
Hydrochlorides of racemic MDMA, DHMA, and HMMA were obtained from Lipomed (Bad Saeckingen, Germany). 4-Hydroxymethamphetamine (pholedrine), 3,4-dihydroxybenzylamine, and adenosine 3′-phosphate 5′-phosphosulfate lithium salt hydrate (PAPS) were obtained from Sigma-Aldrich (Steinheim, Germany). DHMA 3-sulfate, DHMA 4-sulfate, and HMMA sulfate were synthesized in the authors' laboratory as described previously (Schwaninger et al., 2011b). All other chemicals and reagents used were from Merck (Darmstadt, Germany) and of analytical grade. The following Escherichia coli-derived recombinant human cytosolic SULT isoenzymes (rSULT, 20 μg) were from R&D Systems (Wiesbaden, Germany): SULT1A1, SULT1A3, SULT1B1, SULT1E1, and SULT2A1. pHLC (10 individual adult donors, 20 mg protein/ml) was from BD Gentest (Heidelberg, Germany).
Product Formation Approach.
Incubations were performed as described previously (Schwaninger et al., 2011b). For initial screening experiments, 2.5 μM R,S-DHMA or R,S-HMMA and 10 μg/ml SULT1A1, SULT1A3, SULT1B1, SULT1E1, or SULT2A1 were incubated for 20 min. Kinetic constants of DHMA and HMMA sulfation were derived from incubations (n = 2) with the following substrate concentrations: 5, 10, 25, 50, 75, 100, 250 μM (SULT1A1) and 0.25, 0.5, 0.75, 1, 2.5, 5, 10 μM (SULT1A3) for DHMA; and 0.2, 0.4, 0.6, 1, 1.5, 2.0, 2.5 μM (SULT1A3) and 5, 10, 25, 50, 75, 100, 250 μM (SULT1E1) for HMMA. Incubation times were 30 min (SULT1A1 and SULT1E1) and 5 min (SULT1A3), and protein concentrations were 10 μg/ml for SULT1A1, 0.25 μg/ml for SULT1A3, and 1 μg/ml for SULT1E1 and were within the linear range of sulfate formation. Analysis was performed using liquid chromatography-high-resolution mass spectrometry (LC-HRMS) as described previously (Schwaninger et al., 2011b). Enzyme kinetic constants were estimated by nonlinear curve fitting using GraphPad Prism 5.00 software (GraphPad Software Inc., San Diego, CA). The Michaelis-Menten equation was used to calculate apparent Vmax and Km values.
Calculations of Net Clearance.
The relative activity factor (RAF) approach was used to account for differences in expression between the two enzyme sources. The turnover rates for SULT1A1 (probe substrate nitrophenol, 0.25 μM), SULT1A3 (probe substrate dopamine, 1 μM), and SULT1E1 (probe substrate estradiol, 0.005 μM) were determined in rSULT and pHLC. Analysis was performed by LC-HRMS as described previously (Schwaninger et al., 2011b). The RAFs were calculated according to Crespi and Miller (1999), Venkatakrishnan et al. (2000), and Grime and Riley (2006). The enzyme velocities (Venzyme) for the respective metabolic reactions were calculated at different substrate concentrations and were then multiplied with the corresponding RAF leading to the contribution. From these contributions, the percentages of net clearance by a particular SULT at a certain substrate concentration were calculated.
Substrate-Depletion Approach.
Incubation mixtures (final volume, 275 μl) consisted of 100 mM phosphate buffer (pH 7.4), 5 mM MgCl2, pHLC or SULT1A3, substrate, and PAPS. Aliquots of 50 μl were terminated by the addition of 50 μl of acetonitrile at different time points (0, 5, 10, 15, and 20 min). For determination of enantioselective kinetic constants, the following substrate concentrations were used: 0.25, 0.5, 0.75, 1, 2.5, 5, and 10 μM for DHMA and 0.2, 0.4, 0.6, 1, 1.5, 2.0, and 2.5 μM for HMMA. PAPS concentration was 40 μM and protein concentrations were 0.5 μg/ml (SULT1A3) and 1 mg/ml (pHLC). After termination, 10 μl of internal standard solution (pholedrine and 3,4-dihydroxybenzylamine, 25 μM each) was added and the samples were worked-up and analyzed by gas chromatography-negative ion chemical ionization-mass spectrometry as described previously (Schwaninger et al., 2011a). Analyte versus internal standard peak area ratios were determined, and percentages of remaining substrate were calculated. The slope of the natural log of the percentage remaining over time represented the initial substrate-depletion rates (kdep) for each substrate concentration. If substrate decline demonstrated nonlinearity on ln-percentage remaining versus time curves, only those initial time points wherein ln-linearity was observed were used to determine kdep (Obach and Reed-Hagen, 2002; Youdim and Dodia, 2010). Km values were determined by plotting kdep versus substrate concentration on a linear-log plot using Graph Pad Prism 5 and curve-fitting to eq. 1 (Obach and Reed-Hagen, 2002). where [S] is the substrate concentration, Km is the Michaelis-Menten constant, and kdep([S] = 0) is the theoretical maximum consumption rate constant at an infinitesimally low substrate concentration. In addition, Vmax and Km values were estimated by simultaneous fitting to eq. 2 (Youdim and Dodia, 2010).
Vmax was normalized by dividing the calculated Vmax (μM/min) by the actual incubation protein concentration. Stability of the substrates over the incubation period was determined at two concentrations (0.25 and 2.5 μM for DHMA and 0.2 and 1 μM for HMMA) as described above but without addition of PAPS. Instability would be indicated as a negative slope, significantly different from zero (p < 0.05), by regression analysis plotting calculated concentrations versus time.
Correlation between Product Formation and Substrate Depletion.
General applicability of the substrate-depletion approach was checked by comparison of Km and Vmax values obtained for the two approaches. Therefore, in substrate-depletion experiments, Km and Vmax were additionally calculated, after summarizing the concentrations of both enantiomers, to have comparable values to parameters obtained after achiral product formation measurements. For DHMA sulfation in the product formation approach, DHMA 3-sulfation and 4-sulfation were calculated as the sum of both sulfation reactions to give total DHMA sulfation.
Results and Discussion
Incubation conditions chosen were applicable to check the general involvement of the five major human SULT isoenzymes (Riches et al., 2009) in the sulfation of the MDMA metabolites DHMA and HMMA. Sulfation of DHMA was catalyzed by SULT1A1 and SULT1A3 and sulfation of HMMA was catalyzed by SULT1A3 and SULT1E1, and their Km, Vmax, Vmax/Km values are listed in Table 1. The kinetic data for SULT1A1 and SULT1E1 followed classic Michaelis-Menten plots (Fig. 2A). In contrast, sulfation by SULT1A3 showed deviation from the typical Michaelis-Menten kinetics in which the substrate at higher concentrations had a marked concentration-dependent self-inhibition. To obtain comparable kinetic values, they were calculated by the classic Michaelis-Menten equation within a concentration range in which substrate inhibition was not relevant (0.25–10 μM for DHMA; 0.2–2.5 μM for HMMA). DHMA sulfation could be observed in position 3 and position 4 by SULT1A3, with higher turnover rates for position 3. For SULT1A1, only sulfation in position 3 could be observed. However, it should be considered that sulfation by SULT1A1 was rather low, which might result in formation of DHMA 4-sulfate below the detection limit of the used LC-HRMS method.
SULT1A3 turned out to have the highest affinity and the highest capacity for DHMA and HMMA sulfation, with Vmax/Km values approximately 5000- or 3000-fold higher than those for SULT1A1 or SULT1E1, respectively (Table 1). For assessment of contribution of SULT1A1, SULT1A3, and SULT1E1 in pHLC, the RAF approach was used (Crespi and Miller, 1999; Venkatakrishnan et al., 2000; Grime and Riley, 2006) to account for differences in protein expression. Nitrophenol, dopamine, and estradiol were used as probe substrates to calculate RAFs and were accepted to be selective for the described SULT isoforms (Zhang et al., 1998; Dajani et al., 1999; Wang and James, 2005; Riches et al., 2009). However, some drawbacks such as the involvement of other isoforms at higher substrate concentrations should be considered. Therefore, concentrations of the probe substrates were chosen to be selective for the single SULT isoforms. Although SULT1A3 was claimed to be only scarcely expressed in human liver (Riches et al., 2009), DHMA and HMMA were predominantly catalyzed by SULT1A3 (calculated net clearance >90%) at low substrate concentration. Increasing substrate concentration to 10 μM resulted in decreasing contribution of SULT1A3 to approximately 70% for DHMA, whereas contribution for HMMA sulfation remained more than 90%.
The most likely explanation for the observed different pharmacokinetic properties of R- and S-MDMA is an enantioselective metabolism. Enantioselectivity was observed for CYP2C19 (Meyer et al., 2008), COMT (Meyer and Maurer, 2009), UGT1A9 and UGT2B17 (Schwaninger et al., 2009), always with preferences for the S-enantiomers. Although sulfation of MDMA metabolites was already studied and was found to be the major phase II metabolic step (Schwaninger et al., 2011b), no data concerning enantioselectivity are available. Because stereoselective separation of DHMA and HMMA sulfates could not be achieved, a substrate-depletion approach was used, where the depletion of DHMA and HMMA was measured enantioselectively by gas chromatography-mass spectrometry (Schwaninger et al., 2011a). DHMA and HMMA were stable over the incubation time of 20 min. The Km, Vmax, and Vmax/Km values are listed in Table 1. As shown in the sigmoidal plots in Fig. 2B, marked enantioselectivity (defined as Vmax/Km (S-enantiomer)/Vmax/Km (R-enantiomer) > 1.5), according to Meyer et al. (2008), was observed for DHMA sulfation by SULT1A3 toward S-DHMA. No preferences could be observed for HMMA sulfation by SULT1A3. One reason for this difference in enantioselectivity might be the position for sulfation. DHMA was mainly sulfated in position 3, whereas HMMA could only be sulfated in position 4. Comparable S/R ratios as with SULT1A3 were calculated in pHLC, indicating that in vivo enantioselectivity with preferences for the S-enantiomer could be expected for DHMA 3-sulfate but not for HMMA sulfate. Chiral analysis of urine samples from MDMA users were analyzed after conjugate cleavage and revealed preferences for S-DHMA, but no differences between R- and S-HMMA were observed (Pizarro et al., 2004). This result is in line with the in vitro observation that sulfation of HMMA as the major phase II metabolic step showed no enantioselectivity.
The general applicability for determination of enantioselective sulfation reactions was checked by comparing kinetic parameters calculated using the product formation with the substrate-depletion approach. In the product formation approach, Km and Vmax values for DHMA 3-sulfation and 4-sulfation were calculated after summarizing the concentrations of DHMA 3- and DHMA 4-sulfate to give total DHMA sulfation. For the substrate-depletion experiments, Km and Vmax values were calculated after summarizing the concentrations of both enantiomers to have comparable values to parameters obtained after achiral product formation measurements. As shown in Fig. 2C, good correlation, with Km and Vmax values within 2-fold of each other, was observed between product formation and substrate-depletion experiments using recombinant SULT and pHLC. However, the substrate depletion approach possesses some practical limitations (Obach and Reed-Hagen, 2002), such as compounds with low turnover rates and the lack of differentiation between formation of several metabolites. Therefore, it could not be used for SULT1A1 and SULT1E1, because no substantial amount of the initial substrate concentration was consumed during the incubation period. Furthermore, differentiation between sulfation to DHMA 3-sulfate and DHMA 4-sulfate was not possible by monitoring substrate loss. However, product formation experiments showed that sulfation mainly took place in position 3 and DHMA 4-sulfate was only a minor metabolite that might be neglected.
Authorship Contributions
Participated in research design: Schwaninger, Meyer, and Maurer.
Conducted experiments: Schwaninger.
Performed data analysis: Schwaninger.
Wrote or contributed to the writing of the manuscript: Schwaninger, Meyer, and Maurer.
Acknowledgments
We thank Carsten Schröder, Gabriele Ulrich, Armin Weber, and Carina Wink for assistance and helpful discussions.
Footnotes
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
doi:10.1124/dmd.111.041129.
-
ABBREVIATIONS:
- MDMA
- 3,4-methylenedioxymethamphetamine
- P450
- cytochrome P450
- DHMA
- 3,4-dihydroxymethamphetamine
- COMT
- catechol-O-methyltransferase
- HMMA
- 4-hydroxy-3-methoxymethamphetamine
- SULT
- sulfotransferase
- UGT
- UDP-glucuronyltransferase
- pHLC
- pooled human liver cytosol
- PAPS
- adenosine 3′-phosphate 5′-phosphosulfate lithium salt hydrate
- rSULT
- recombinant human cytosolic SULT isoenzymes
- LC-HRMS
- liquid chromatography-high-resolution mass spectrometry
- RAF
- relative activity factor.
- Received June 9, 2011.
- Accepted July 27, 2011.
- Copyright © 2011 by The American Society for Pharmacology and Experimental Therapeutics