Serotonergic Neurotoxicity of 3,4 ... - ACS Publications

The data indicate that thioether metabolites of MDA and MDMA contribute to the serotonergic neurotoxicity observed following peripheral administration...
0 downloads 0 Views 100KB Size
Download PDF
Chem. Res. Toxicol. 2001, 14, 863-870

863

Serotonergic Neurotoxicity of 3,4-(()-Methylenedioxyamphetamine and 3,4-(()-Methylendioxymethamphetamine (Ecstasy) Is Potentiated by Inhibition of γ-Glutamyl Transpeptidase Fengju Bai,† Douglas C. Jones, Serrine S. Lau, and Terrence J. Monks* Center for Cellular and Molecular Toxicology, College of Pharmacy, University of Texas at Austin, Austin, Texas 78712-1074 Received January 16, 2001

Reactive metabolites play an important role in 3,4-(()-methylenedioxyamphetamine (MDA) and 3,4-(()-methylenedioxymethamphetamine (MDMA; ecstasy)-mediated serotonergic neurotoxicity, although the specific identity of such metabolites remains unclear. 5-(GlutathionS-yl)-R-methyldopamine (5-GSyl-R-MeDA) is a serotonergic neurotoxicant found in the bile of MDA-treated rats. The brain uptake of 5-GSyl-R-MeDA is decreased by glutathione (GSH), but sharply increases in animals pretreated with acivicin, an inhibitor of γ-glutamyl transpeptidase (γ-GT) suggesting competition between intact 5-GSyl-R-MeDA and GSH for the putative GSH transporter. γ-GT is enriched in blood-brain barrier endothelial cells and is the only enzyme known to cleave the γ-glutamyl bond of GSH. We now show that pretreatment of rats with acivicin (18 mg/kg, ip) inhibits brain microvessel endothelial γ-GT activity by 60%, and potentiates MDA- and MDMA-mediated depletions in serotonin (5-HT) and 5-hydroxylindole acidic acid (5-HIAA) concentrations in brain regions enriched in 5-HT nerve terminal axons (striatum, cortex, hippocampus, and hypothalamus). In addition, glial fibrillary acidic protein (GFAP) expression increases in the striatum of acivicin and MDA (10 mg/kg) treated rats, but remains unchanged in animals treated with just MDA (10 mg/kg). Inhibition of endothelial cell γ-GT at the blood-brain barrier likely enhances the uptake into brain of thioether metabolites of MDA and MDMA, such as 5-(glutathion-S-yl)-R-MeDA and 2,5-bis-(glutathion-S-yl)-R-MeDA, by increasing the pool of thioether conjugates available for uptake via the intact GSH transporter. The data indicate that thioether metabolites of MDA and MDMA contribute to the serotonergic neurotoxicity observed following peripheral administration of these drugs.

Introduction 3,4-(()-Methylenedioxymethamphetamine (MDMA, “ecstasy”, I in Figure 1) and 3,4-(()-methylenedioxyamphetamine1 (MDA, II in Figure 1) produce long-term serotonergic nerve terminal axonal damage, and recreational users of these drugs are at risk of neurotoxicity (1-5). Interestingly, in animal models, direct injection of these amphetamine analogues into the brain fails to reproduce the toxicity observed following peripheral administration (6-8). These data suggest that systemic metabolism is required for the expression of neurotoxicity. Additional studies on the role of inhibitors and inducers of drug metabolism support this scenario (9). However, the precise identity of the metabolites participating in MDA and MDMA-mediated neurotoxicity re* To whom correspondence should be addressed. Phone: (512) 4716699. Fax: (512) 471-5002. E-mail: [email protected]. † Present address: Neuroscience Discovery, Eli Lilly and Company, Indianapolis, IN 46285. 1 Abbreviations: GFAP, glial fibrillary acidic protein; γ-GT, γ-glutamyl transpeptidase; GSH, glutathione; 5-HIAA, 5-hydroxy indoleacetic acid; 5-HT, 5-hydroxyytryptamine (serotonin); MDA, 3,4-(()-Methylenedioxyamphetamine; MDMA, 3,4-(()-methylenedioxymethamphetamine; R-MeDA, R-methyldopamine; 5-(glutathion-S-yl)-R-MeDA, 5-(glutathion-S-yl)-R-methyldopamine; 5-(N-acetylcystein-S-yl)-R-MeDA, 5-(Nacetylcystein-S-yl)-R-methyldopamine; 2,5-bis-(glutathion-S-yl)-R-MeDA, 2,5-bis-(glutathion-S-yl)-R-methyldopamine.

mains unclear. For example, neither R-methyldopamine (R-MeDA, IV in Figure 1) nor 3-O-methyl-R-methyldopamine, major metabolites of MDA, produce neurotoxicity following intracerebral administration (10). Although the 6-hydroxydopamine analogues, 2-(methylamino)-1-(2,4,5-trihydroxyphenyl)propane (2,4,5-trihydroxymethamphetamine) and 2-(amino)-1-(2,4,5-trihydroxyphenyl)propane (2,4,5-trihydroxyamphetamine), putative metabolites of MDMA and MDA, respectively, deplete serotonin (5-HT) concentrations following intracerebroventricular and intrastriatal administration to rats, they also target the dopaminergic system (11, 12), thereby lacking the selectivity of the parent amphetamines. If systemically formed metabolites do indeed contribute to the neurotoxicity of MDA and MDMA, then mechanisms by which such putative metabolites gain access to the brain need to be identified. To address this problem we have been investigating the role of thioether metabolites of R-MeDA in the neurotoxicity of MDA (13-15). R-MeDA and N-methyl-R-MeDA (III, Figure 1) are catecholic metabolites of MDA and MDMA, respectively, that readily oxidize to the corresponding ortho-quinones (V, Figure 1). Quinones are electrophiles and they react readily with the nucleophilic sulfyhdryl group in glutathione (GSH) resulting in the formation of water

10.1021/tx010011l CCC: $20.00 © 2001 American Chemical Society Published on Web 05/31/2001

864

Chem. Res. Toxicol., Vol. 14, No. 7, 2001

Bai et al.

Figure 1. Proposed pathway of MDMA and MDA metabolism to serotonergic neurotoxic metabolites. MDMA (I) and MDA (II) undergo systemic cytochrome P450-mediated (CYP2D, 2B, 3A) demethylenation to N-methyl-R-MeDA (III) and R-MeDA (IV), respectively. The catechols are readily oxidized to the corresponding ortho-quinones (V), which react readily with GSH to form the corresponding GSH conjugates (VI, illustrated for R-MeDA only). Quinol-linked GSH conjugates are also susceptible to oxidation to the quinone-thioether (VII) which undergo a second round of GSH addition to form the bis-GSH substituted catechol (VIII). We hypothesize that systemically formed GSH conjugates are selectively transported across the endothelial cells of the blood-brain barrier via the putative intact GSH transporter.

soluble thioether conjugates. GSH and N-acetylcysteine conjugates of R-MeDA are selective serotonergic neurotoxicants in the rat (14, 15). The serotonergic toxicity of the R-MeDA-thioethers following central administration is manifest as depletions in 5-HT and 5-HIAA concentrations in the striatum, cortex, and hippocampus, suggesting that these reactive metabolites may contribute to the development of the neurotoxicity following exposure to the parent amphetamines. The transport of water-soluble compounds across the blood-brain barrier usually requires the presence of specific transporters on brain microvessel endothelial cell membranes. The fractional uptake of 5-(glutathion-S-yl)R-[3H]-MeDA (VI, Figure 1) into brain following a single pass (7.4 ( 0.5%) is comparable to the uptake of GSH (16). Moreover, brain uptake of 5-(glutathion-S-yl)-R-[3H]MeDA is decreased by coadministration of GSH, suggesting competition between 5-(glutathion-S-yl)-R-[3H]MeDA and GSH for a putative GSH transporter (16). In addition, pretreatment of animals with an inhibitor of γ-GT sharply increases the uptake of 5-(glutathion-S-yl)R-[3H]-MeDA into brain (16). γ-GT is enriched in bloodbrain barrier endothelial cells and catalyzes the first step in the metabolism of GSH and its S-conjugates (17). Inhibition of endothelial γ-GT may therefore enhance the delivery of 5-(glutathion-S-yl)-R-[3H]-MeDA into brain by preventing its metabolic clearance at the blood-brain barrier, thereby increasing the pool of 5-(glutathion-Syl)-R-[3H]-MeDA available for an intact GSH transporter. If R-MeDA-thioethers contribute to MDA and MDMA neurotoxicity, inhibition of γ-GT at the blood-brain barrier should increase the brain uptake of R-MeDA-GSH conjugates, and should potentiate the neurotoxicity of MDA and MDMA. In the present study, we therefore determined the effect of γ-GT inhibition on the neurotoxicity of subcutaneously administered MDA and MDMA. The data confirm a role for R-MeDA-thioethers in MDA/ MDMA-mediated neurotoxicity.

Experimental Procedures Chemicals. (()-MDA and (()-MDMA were kindly provided by the Research Technology Branch, National Institute on Drug Abuse (Rockville, MD). 5-HT, 5-HIAA, dopamine, dihydroxy

phenylacetic acid, homovanillic acid, norepinephrine, and acivicin were obtained from Sigma Chemical Co. (St. Louis, MO). The monoclonal antibody against glial fibrillary acidic protein (GFAP) was obtained from Roche Molecular Biochemicals (Germany). Rationale for Dosage Selection. The extent of 5-HT nerve terminal degeneration is dependent upon the dose, route, and frequency of MDA and MDMA administration (4, 18). Subcutaneous injections are more effective than oral administration at producing neurotoxicity, and multiple doses are more effective than a single dose (4). The most commonly employed dosing regimen for MDA and MDMA used to produce long-term serotonergic toxicity in rats is 20 mg/kg (93 µmol/kg, s.c.) and is frequently administered in a multiple dosing schedule. Administration of MDMA (4 mg/kg, i.p.) to rats once daily for 4 days produces no neurodegenerative effects but, when given twice daily for 4 days, depletes 5-HT by 22-43% (18). Five injections of MDA (10 mg/kg, s.c.) to rats at 6 h intervals decreases 5-HT and 5-HIAA concentrations to less than 30% of control levels (19). Therefore, the degree of MDA and MDMA neurotoxicity varies and is dependent upon the particular dosing regimen used. Single subcutaneous injections of MDA and MDMA result in a steep dose response curve for 5-HT depletion. Thus, in our studies, a single dose of MDA or MDMA (20 mg/ kg, s.c.) depletes brain 5-HT concentrations to 47-64% of control values, consistent with literature reports (20). However, when the dose of MDA or MDMA is reduced 50%, to 10 mg/kg, 5-HT and 5-HIAA concentrations are only modestly depleted. Because of the requirement for an experimental model in which potentiation of MDA and MDMA neurotoxicity could be readily observed, we selected the 10 mg/kg dosing regimen. Animals and Dosing. Male Sprague-Dawley rats (200-250 g) were used for all experiments, and were housed in the Animal Resource Center at the University of Texas at Austin. Animals were maintained on a 12 h light and dark cycle at a constant room temperature (72 °F). Food and water were provided ad libitum. Rats were pretreated with acivicin (18 mg/kg, i.p.) 20 min prior to drug administration. MDA or MDMA (10 mg/kg, subcutaneous) were then administered to either control or acivicin treated rats subcutaneously (outside of right femur). Seven days after the last injection, rats were euthanized by decapitation and their brains quickly removed and placed onto an ice-cold plate. Brain regions were dissected as previously described (13). Brain tissues were stored at -80 °C for no longer than 1 week prior to analysis. For neurotransmitter analyses, tissue was weighed and sonicated with a sonic dismembranator in ice-cold 0.1 N perchloric acid containing 134 µM disodium

Metabolism and Ecstasy-Mediated Neurotoxicity ethylenediamine tetraacetate and 263 µM sodium metabisulfite for 30 s. The sonicated tissues were centrifuged at 13500g (4 °C) for 20 min. Supernatants were centrifuged again under the same conditions, and aliquots (20 µL) of this supernatant used for HPLC analysis. HPLC-Electrochemical Analysis of Neurotransmitters. 5-HT, 5-HIAA, dopamine, dihydroxy phenylacetic acid, homovanillic acid, and norepinephrine concentrations were quantified by HPLC coupled to an eight channel coulometric electrode array system (ESA Inc., Chelmsford, MA). The potentials applied to the electrodes increased in 50 mV increments, starting from 0 mV at the first channel, increasing to +350 mV at the eighth channel. Sample aliquots were loaded onto an ESA HR-80 column (80 × 4.6 mm i.d.; 3 mm particle size). Analytes were separated with a mobile phase consisting of 8 mM ammonium acetate, 4 mM citrate, 54 µM disodium ethylenediamine tetraacetate, 230 µM 1-octanesulfonic acid, and 5% methanol (pH 2.5) at a flow rate of 1 mL/min. Quantitation of monoamine neurotransmitters and their metabolites was achieved by comparing the peak areas with standard curves generated from commercially purchased standards. Determination of Rat Brain Microvessel Endothelial Cell γ-GT Activity. Acivicin (18 mg/kg, s.c.) or vehicle treated rats were euthanized and decapitated 20 min after the injection. Rat brains were removed, alcohol-sterilized, and soaked in Dulbecco’s modified Eagle’s-F12 medium with antibiotics. The meninges and white matter were removed. After 1 h of collagenase/Dispase digestion, microvessels were separated from myelin residues by centrifugation in 50 mL of 13% dextran for 10 min. The microvessel pellets were digested a second time with collagenase/Dispase at 37 °C for 3 h. After enzymatic digestion, the cells were centrifuged for 5 min at 1000g and the sediment was dissolved in 1 mL of Hank’s balanced salt solution. Percoll gradients were prepared by centrifugation of 50% Percoll in Hank’s balanced salt solution for 1 h at 18000g (4 °C). The cell suspension was then layered on top of the Percoll gradient and centrifuged for 15 min at 2000 g (4 °C) to isolate the capillary endothelial cells from the opaque ring, at about the upper one-third of the Percoll gradient. Brain microvessel endothelial cells were sonicated in 0.5 mL of Tris-HCl buffer (pH 7.4; 0.15 M KCl; 20 mM Tris) at 4 °C with a sonic dismembranator. Aliquots (0.15 mL) of the sonicated cell suspension were used for the analysis of γ-GT activity measured as described previously (13), by determination of the rate of p-nitroaniline formation from γ-glutamyl-p-nitroaniline. Protein concentrations were determined by the method of Bradford (21). Determination of GFAP Expression. Rats were euthanized by decapitation 7 days after the last dosing. Brains were quickly removed and areas enriched in 5-HT nerve terminals (cortex, hippocampus, and striatum) were dissected free on ice. Total protein was isolated from these brain regions according to the standard method of Santa Cruz Biotechnology (Santa Cruz, CA) using RIPA (1% NP40, 0.5% sodium deoxycholate, and 0.1% sodium dodecyl sulfate in phosphate buffered saline) containing proteinase inhibitors. Equal amounts of protein from each sample were loaded for gel electrophoresis. A monoclonal antibody against GFAP was used to identify the GFAP band on the membrane. The expression of GFAP was quantified with Kodak Electrophoresis Documentation and Analysis System 120. Protein concentrations were determined by the method of Bradford (21).

Results Acivicin (18 mg/kg, i.p.) administration significantly decreased blood-brain barrier microvessel endothelial cell γ-GT activity from 1.53 ( 0.19 nmol/mg/min in controls to 0.63 ( 0.07 in acivicin treated animals, a decrease of ∼60%. To determine whether inhibition of γ-GT potentiated the neurotoxicity of MDA or MDMA, a dose of the amphetamine analogues was selected at the low end of the dose-response curve. Treatment of rats

Chem. Res. Toxicol., Vol. 14, No. 7, 2001 865

Figure 2. Effect of acivicin (18 mg/kg, i.p.) pretreatment on 5-HT depletion, 7 days following MDA (10 mg/kg, s.c.) administration. Decreases in 5-HT concentrations are expressed as the mean ( SE (N ) 6 in each group). Absolute values for 5-HT concentrations in the striatum (A), cortex (B), hippocampus (C), and hypothalamus (D) in control animals were 2.33 ( 0.07, 1.22 ( 0.06, 1.60 ( 0.08, and 3.96 ( 0.23 pmol/mg tissue, respectively. Vehicle treated controls (error bars only), acivicin-treated animals (hatched bars), MDA-treated animals (open bars), combined acivicin/MDA-treated animals (black bars). A one-way ANOVA followed by Student Newman-Kuels tests were conducted on the data. 5-HT concentrations in the combined acivicin/MDA treated rats were significantly different from those in the control and the MDA treated rats. The F and the P values are [F (3,20) ) 9.026, p ) 0.0006], [F (3,20) ) 16.925, p ) 0.0001], [F (3,20) ) 14.070, p ) 0.0001], and [F (3,20) ) 13.597, p ) 0.0001] for the striatum, cortex, hippocampus, and hypothalamus, respectively. (*) Values significantly different from the control group at p < 0.05; (†) Values significantly different from those in the MDA treated group at p < 0.05.

with MDA at a dose of 10 mg/kg caused relatively modest decreases in both 5-HT (Figure 2) and 5-HIAA (Figure 3) concentrations in the striatum, cortex, hippocampus, and hypothalamus. Decreases in 5-HT concentrations in MDA-treated animals ranged from 11 to 29% and reached statistical significance in the hippocampus and hypothalamus. Following pretreatment with acivicin (18 mg/ kg) the same dose of MDA significantly decreased 5-HT concentrations by 27, 40, 44, and 46% in the striatum, cortex, hippocampus, and hypothalamus, respectively (Figure 2). 5-HIAA concentrations decreased 7-20% in

866

Chem. Res. Toxicol., Vol. 14, No. 7, 2001

Bai et al.

Figure 4. Effect of acivicin (18 mg/kg, i.p.) pretreatment on striatal GFAP expression, 7 days following MDA (10 mg/kg, s.c.) administration. (A) Western blot-ECL analysis of striatal GFAP expression. Each lane contains the same amount of total protein isolated from the striatum of one rat (N ) 3 in each group). The net density of GFAP bands was quantified with Kodak Electrophoresis Documentation and Analysis System 120. To obtain relative densities, each net density value from each treated rat was divided by the mean of the net density value of the control group, a one-way ANOVA followed by Student Newman-Kuels tests were conducted on the data. As shown in panel B, the relative GFAP density in the combined acivicin/ MDA treated rats (black bars) were significantly different from those in the control (shaded bars) and the MDA treated rats (open bars) [F (3,8) ) 8.172, p ) 0.0081].

Figure 3. Effect of acivicin (18 mg/kg, i.p.) pretreatment on 5-HIAA concentrations, 7 days following MDA (10 mg/kg, s.c.) administration. Decreases in 5-HIAA concentrations are expressed as the mean ( SE (N ) 6 in each group). Absolute values for 5-HIAA concentrations in the striatum (A), cortex (B), hippocampus (C), and hypothalamus (D) in control animals were 1.88 ( 0.05, 0.91 ( 0.04, 1.03 ( 0.06, and 1.68 ( 0.12 pmol/mg tissue, respectively. Vehicle treated controls (error bars only), acivicin treated animals (hatched bars), MDA-treated animals (open bars), combined acivicin/MDA treated animals (black bars). A one-way ANOVA followed by Student Newman-Kuels tests were conducted on the data. 5-HIAA concentrations in the combined acivicin/MDA treated rats were significantly different from those in the control and the MDA treated rats. The F and the P values are [F (3,20) ) 9.942, p ) 0.0003], [F (3,20) ) 11.746, p ) 0.0001], [F (3,20) ) 12.097, p ) 0.0001], and [F (3,20) ) 8.444, p ) 0.0008] for the striatum, cortex, hippocampus, and hypothalamus, respectively. (*) Values significantly different from the control group at p < 0.05; (†) Values significantly different from those in the MDA treated group at p < 0.05.

MDA treated animals (Figure 3) and reached statistical significance in all but the striatum. Acivicin potentiated MDA mediated decreases in 5-HIAA concentrations to 22, 28, 37, and 36% in the striatum, cortex, hippocampus, and hypothalamus, respectively. None of the experimental protocols caused changes in the dopaminergic or norandrenergic neurotransmitter levels (data not shown). Increases in GFAP expression are commonly used to identify the presence of neuronal injury following chemical insult (22, 23). Western blot analysis showed a 1.6fold increase in GFAP expression in the striatum of acivicin (18 mg/kg, i.p.) and MDA (10 mg/kg, s.c.) treated rats, 7 days after drug administration (Figure 4). No

changes in striatal GFAP expression were found in rats treated with MDA (10 mg/kg) alone (Figure 4). Acivicin (18 mg/kg, i.p.) had no effect on GFAP expression in rat striatum (Figure 4). Acivicin produced similar potentiations of MDMAmediated serotonergic neurotoxicity (Figures 5 and 6). Thus, MDMA (10 mg/kg) caused modest (18-29%) but significant decreases in brain 5-HT concentrations, decreases which were amplified (52-64%) in acivicin pretreated animals (Figure 5). The striatum (56.2%) and cortex (58.2%) were most sensitive to acivicin/MDMAmediated decreases in 5-HT concentrations. MDMA also caused decreases (19-23%) in brain 5-HIAA concentrations (Figure 6) and pretreatment with acivicin potentiated (41-56%) these decreases. Acivicin treatment alone had no statistically significant effects on either 5-HT or 5-HIAA concentrations, although in the experiment with MDA (Figures 2 and 3), acivicin alone appeared to decrease 5-HT and 5-HIAA concentrations in the hypothalamus. The potential clinical use of acivicin as an antitumor agent was limited by its dose-limiting reversible neurotoxicity (24). However, although some symptoms of acivicin CNS toxicity are shared by humans and higher nonhuman species, such as the monkey and the cat, they are not reproduced in rodents. Since acivicin does cross the blood brain barrier (25), we cannot rule out the possibility that by interfering with glutamine and glutathione metabolism it may produce neurochemical changes within specific regions of the brain that might be exacerbated by either MDA or MDMA. However, the effect of acivicin alone seen in the experiment with MDA was not reproduced in the study of MDMA (Figures 5 and 6) nor did it change GFAP expression (Figure 4).

Metabolism and Ecstasy-Mediated Neurotoxicity

Figure 5. Effect of acivicin (18 mg/kg, i.p.) pretreatment on 5-HT depletion, 7 days following MDMA (10 mg/kg, s.c.) administration. Decreases in 5-HT concentrations are expressed as the mean ( SE (N ) 6 in each group). Absolute values for 5-HT concentrations in the striatum (A), cortex (B), hippocampus (C), and hypothalamus (D) in control animals were 2.44 ( 0.09, 1.17 ( 0.07, 1.53 ( 0.10, and 3.96 ( 0.16 pmol/mg tissue, respectively. Vehicle treated controls (error bars only), acivicin treated animals (hatched bars), MDMA-treated animals (open bars), combined acivicin/MDMA treated animals (black bars). A one-way ANOVA followed by Student Newman-Kuels tests were conducted on the data. 5-HT concentrations in the combined acivicin/MDMA treated rats were significantly different from those in the control and the MDMA treated rats. The F and the P values are [F (3,20) ) 9.026, p ) 0.0006], [F (3,20) ) 16.925, p ) 0.0001], [F (3,20) ) 14.070, p ) 0.0001], and [F (3,20) ) 13.597, p ) 0.0001] for the striatum, cortex, hippocampus, and hypothalamus, respectively. (*) Values significantly different from the control group at p < 0.05; (†) Values significantly different from those in the MDMA treated group at p < 0.05.

Discussion Pretreatment of animals with acivicin, an inhibitor of γ-GT, potentiates both MDA (Figures 2-4)- and MDMA (Figures 5 and 6)-mediated serotonergic neurotoxicity. This potentiation is likely a consequence of increases in the availability of specific metabolites within the capillary lumen that are substrates for this enzyme, and their subsequent increased delivery into brain (see Figure 4 in ref 26 for a model depicting the delivery of polyphenolic-GSH conjugate metabolites of MDA to the brain). Consistent with this view, inhibition of γ-GT substantially increases the brain uptake of 5-(glutathion-S-yl)R-MeDA (16). 5-(Glutathion-S-yl)-R-MeDA and 2,5-bis-

Chem. Res. Toxicol., Vol. 14, No. 7, 2001 867

Figure 6. Effect of acivicin (18 mg/kg, i.p.) pretreatment on 5-HIAA concentrations, 7 days following MDMA (10 mg/kg, s.c.) administration. Decreases in 5-HIAA concentrations are expressed as the mean ( SE (N ) 6 in each group). Absolute values for 5-HIAA concentrations in the striatum (A), cortex (B), hippocampus (C), and hypothalamus (D) in control animals were 2.03 ( 0.10, 1.06 ( 0.08, 1.09 ( 0.07, and 1.64 ( 0.11 pmol/mg tissue, respectively. Vehicle treated controls (error bars only), acivicin treated animals (hatched bars), MDMA-treated animals (open bars), combined acivicin/MDMA treated animals (black bars). A one-way ANOVA followed by Student Newman-Kuels tests were conducted on the data. 5-HIAA concentrations in the combined acivicin/MDMA treated rats were significantly different from those in the control and the MDMA treated rats. The F and the P values are [F (3,20) ) 9.942, p ) 0.0003], [F (3,20) ) 11.746, p ) 0.0001], [F (3,20) ) 12.097, p ) 0.0001], and [F (3,20) ) 8.444, p ) 0.0008] for the striatum, cortex, hippocampus, and hypothalamus, respectively. (*) Values significantly different from the control group at p < 0.05; (†) Values significantly different from those in the MDMA treated group at p < 0.05.

(glutathion-S-yl)-R-MeDA (VIII, Figure 1) are selective and relatively potent serotonergic neurotoxicants (14, 15). Collectively, the data provides evidence that thioether conjugates contribute to the neurotoxicity of MDA and MDMA following the peripheral administration of these drugs. Evidence that systemic metabolism of MDA and MDMA is required for the development of neurotoxicity is well established (6-9). Thus, any model of MDA- and MDMAmediated neurotoxicity is required to provide a mechanism for the uptake of such neurotoxic metabolites into brain. Because drug metabolites are usually more hydrophilic than the parent drug, it is likely that specific

868

Chem. Res. Toxicol., Vol. 14, No. 7, 2001

transporters must be available to facilitate the entry of such water soluble, neurotoxic metabolites into brain parenchyma. The metabolism of 5-(glutathion-S-yl)-RMeDA within the circulation of the brain by microvessel endothelial cell γ-GT and dipeptidases will result in the formation of 5-(cystein-S-yl)-R-MeDA. The transport of 5-(cystein-S-yl)-R-MeDA across the blood-brain barrier into the brain parenchyma may provide one potential route by which systemically formed water-soluble metabolites gain access to the brain. Cysteine conjugates of R-MeDA may therefore be transported into brain tissue across the blood-brain barrier through the system Ltransporter for neutral amino acids (26). Amino acid transporters have been identified within the blood-brain barrier (27), and a variety of neurotoxicants have been shown to be transported into the brain via these carriers (28, 29). Dichlorovinylcysteine, a reactive metabolite of the neurotoxicant, dichloroacetylene, is transported across the blood-brain barrier by a Na+-independent system L-transporter for neutral amino acids, while the uptake of its corresponding GSH conjugate is mediated by an as yet unknown carrier system (30). However, the significant potentiation of MDA and MDMA neurotoxicity by acivicin suggests that the majority of R-MeDA-thioethers are transported into brain across the blood-brain barrier in the form of intact GSH conjugates. Quinone-thioethers inhibit a variety of enzymes that either utilize GSH as a substrate or as a cosubstrate, including γ-GT (31). Such an effect of R-MeDA- or N-methyl-R-MeDA GSH conjugates on γ-GT at the bloodbrain barrier could subsequently increase the brain uptake of neurotoxic metabolites of MDA and MDMA, especially among recreational abusers, increasing the risk of neurotoxicity. Variations in the extent to which the conjugates are transported into brain may therefore be an important determinant of individual susceptibility to MDA and MDMA toxicity. It should be noted that very little is known about human variability in γ-GT. Consistent with the effects of acivicin on MDA and MDMA-mediated neurotoxicity, the saturable low affinity transport of GSH across the blood-brain barrier has been reported (32). Recent studies have demonstrated the apical localization of a sodium-dependent GSH transporter in mouse and human cerebrovascular endothelial cells (33,34). The brain uptake of intact GSH is inhibited by several GSH conjugates, including S-alkyl GSHs, sulfobromophthalein-GSH, GSH monoethyl ester, probenecid, and ophthalmic acid (35), suggesting that GSH S-conjugates and GSH share the same transport mechanism. GSH also inhibits the brain uptake of 5-(glutathion-S-yl)-R-[3H]-MeDA, implying that the two share the same transporter (16). Interestingly, even though acivicin causes a profound glutathionuria by inhibiting renal brush border membrane γ-GT activity, plasma concentrations of GSH increase about 6-fold after 4.5-6 h after acivicin administration (36, 37). We did not measure plasma GSH concentrations 20 min after acivicin (the time of drug administration) but, as noted, we have previously shown that this protocol significantly increases (6.5-fold) the uptake of 5-(glutathion-S-yl)-R[3H]-MeDA into brain (16). The nature of the ultimate neurotoxic metabolite(s) of MDA and MDMA remain unknown. Within brain parenchyma, 5-(glutathion-S-yl)-R-MeDA is rapidly metabolized by γ-GT and dipeptidases to 5-(cystein-S-yl)-RMeDA, concentrations of which decline rapidly with the

Bai et al.

concomitant accumulation of 5-(N-acetylcystein-S-yl)-RMeDA (16). In contrast to 5-(glutathion-S-yl)-R-MeDA and 5-(cystein-S-yl)-R-MeDA, 5-(N-acetylcystein-S-yl)-RMeDA is eliminated relatively slowly from the brain (16). The accumulation and persistence of 5-(N-acetylcysteinS-yl)-R-MeDA in the brain may therefore contribute to the long-term neurotoxicity of MDA and MDMA. Indeed, 5-(N-acetylcystein-S-yl)-R-MeDA is more than 2 orders of magnitude more potent than 5-(glutathion-S-yl)-RMeDA (14, 15). Intrastriatal administration of 5-(Nacetylcystein-S-yl)-R-MeDA, at a dose as low as 7 nmol, produces a 50% reduction in striatal 5-HT concentrations, approximately equivalent to the effects of 23.25 µmol MDA (93 µmol/kg s.c. in rats weighing 250 g). Because this dose represents just 0.03% of the dose of MDA, attempts to demonstrate significant differences in control and acivicin treated animals may be a challenge. Ideally, we would like to demonstrate increases in brain concentrations of thioether metabolites of R-MeDA and N-methyl-R-MeDA in the presence of acivicin. However, in addition to the extremely small fraction of MDA or MDMA that is converted into thioether metabolites, these metabolites are also inherently reactive, and may bind extensively to cellular macromolecules, thereby further decreasing the free concentration available for detection. Nonetheless, ongoing studies are attempting to determine brain concentrations of thioether metabolites of R-MeDA, and N-methyl-R-MeDA in the presence and absence of acivicin. The potential role of dopamine-derived quinonethioethers in Parkinson’s disease and of similar metabolites in a variety of other neurodegenerative disorders has been considered by Dryhurst and colleagues (3841), and their work clearly demonstrates that such novel metabolites possess neurotoxic properties. It is of interest that the antihypertensive drug, R-methyldopa, is decarboxylated to R-MeDA which accumulates in the brain and in the periphery. However, this drug is not associated with degeneration of 5-HT neurons and/ or neurotoxicity. Although R-methyldopa is decarboxylated to R-MeDA, it is subsequently transported into storage vesicles, rich in ascorbate, and metabolized by dopamine β-hydroxylase to the false neurotransmitter, R-methylnorepinephrine. Therefore, the kinetics and distribution of R-MeDA generated from R-methyldopa, either centrally, or within peripheral nerves, are significantly different from R-MeDA generated principally in the liver from MDA. Thus, R-MeDA generated from R-methyldopa will be restricted to the reductive environment of the vesicular compartment. Consistent with this view, direct injection of R-MeDA into the brain fails to reproduce MDA neurotoxicity (16). In contrast, hepatic generation of R-MeDA or N-methyl-R-MeDA from MDA or MDMA exposes the catechol to oxidation from a variety of sources, followed by conjugation of the o-quinone with GSH. Our finding that 5-(GSyl)-R-MeDA is excreted in bile of rats treated with MDA confirm this latter scenario. We also doubt that R-MeDA or N-methyl-R-MeDA produce neurotoxicity following systemic administration given the extent to which such systemically administered catechols would be O-methylated by COMT within the circulation. Thus, the disposition of systemically administered catechols will be substantially different from that of catechols generated in situ within, for example, hepatocytes, and will not reflect the complex pharmcokinetics/ pharmcodynamics of metabolite formation following sc administration of the parent amphetamines.

Metabolism and Ecstasy-Mediated Neurotoxicity

In summary, inhibition of γ-GT within brain microvessel endothelial cells at the blood brain barrier potentiated 5-HT and 5-HIAA depletions following the administration of either MDA or MDMA (10 mg/kg). This effect is probably due to decreases in the metabolic clearance of the GSH conjugates of R-MeDA or N-methyl-R-MeDA within the brain capillary lumen, resulting in increases in intact conjugates available for uptake across microvessel endothelial cells via an intact GSH transporter. The data are consistent with a model in which MDA and MDMA are metabolized systemically to R-MeDA and or N-methyl-R-MeDA and their corresponding thioether conjugates, which are subsequently transported into the brain where they elicit a selective serotonergic neurotoxicity. The mechanisms by which such conjugates produce this selective neurotoxicity are unclear and remains the basis of ongoing investigations.

Acknowledgment. This work was supported in part by an award to T.J.M. from the National Institute on Drug Abuse (DA 108326). Its contents are solely the responsibility of the authors and do not represent the official views of NIDA.

References (1) Bolla, K..I., McCann, U. D., and Ricaurte, G. A. (1998) Memory impairment in abstinent MDMA (“Ecstasy”) users. Neurology 51, 1532-1537. (2) McCann, U. D., Szabo, Z., Scheffel, U., Dannals, R. F., and Ricaurte, G. (1998) Positron emission tomographic evidence of toxic effect of MDMA (“Ecstasy”) on brain serotonin neurons in human beings. Lancet 352, 1433-1437. (3) Parrott A. C. (2000) Human research on MDMA (3,4-Methylenedioxymethamphetamine) neurotoxicity: cognitive and behavioural indices of change. Neuropsychobiology 42, 17-24. (4) Ricaurte G. A., DeLanney L. E., Irwin, I., and Langston, J. W. (1988) Toxic effects of MDMA on central serotonergic neurons in the primate: importance of route and frequency of drug administration. Brain Res. 446, 165-168. (5) Ricaurte, G. A., McCann, U. D., Szabo, Z., and Scheffel, U. (2000) Toxicodynamics and long-term toxicity of the recreational drug, 3, 4-methylenedioxymethamphetamine (MDMA, ‘Ecstasy’). Toxicol. Lett. 112-113, 143-146. (6) Molliver, M. E., O’Hearn, E., Battaglia, G., and DeSouza, E. B. (1986) Direct intracerebral administration of MDA and MDMA does not produce serotonin neurotoxicity. Soc. Neurosci. Abstr. 12, 1234. (7) Schmidt, C. J., and Taylor, V. L. (1988) Direct central effects of acute methylenedioxy-methamphetamine on serotonergic neurons. Eur. J. Pharmacol. 156, 121-131. (8) Paris, J. M., and Cunningham, K. A. (1992) Lack of serotonin neurotoxicity after intraraphe microinjection of (+)-3,4-methylenedioxymethamphetamine (MDMA). Brain Res. Bull. 28, 115119. (9) Gollamudi R., Ali, S. F., Lipe, G., Newport, G., Webb, P., Lopez, M., Leakey, J. E., Kolta, M., and Slikker, W. Jr. (1989) Influence of inducers and inhibitors on the metabolism in vitro and neurochemical effects in vivo of MDMA. Neurotoxicology 10, 455466. (10) McCann, U. D., and Ricuarte, G. A. (1991) Major metabolites of (()-3,4-methylenedioxy-amphetamine (MDA) do not mediate its toxic effects on brain serotonin neurons. Brain Res. 545, 279282. (11) Zhao, Z., Castagnoli, N., Jr., Ricaurte, G. A., Steele, T., and Martello, M. (1992) Synthesis and neurotoxicological evaluation of putative metabolites of the serotonergic neurotoxin 2-(methylamino)-1-[3,4-(methylenedioxy)phenyl]propane[(methylenedioxy)methamphet-amine]. Chem. Res. Toxicol. 5, 89-94. (12) Elayan, I., Gibb, J. W., Hanson, G. R., Foltz, R. L., Lim, H. K., and Johnson, M. (1992) Long-term alteration in the central monaminergic systems of the rat by 2,4,5-trihydroxyamphetamine, but not by 2-hydroxy-4,5-methylenedioxymetamphetamine or 2-hydroxy-4,5-methyl-enedioxyamphetamine. Eur. J. Pharmacol. 221, 281-288.

Chem. Res. Toxicol., Vol. 14, No. 7, 2001 869 (13) Miller, R. T., Lau, S. S., and Monks, T. J. (1996) Effect of 5-(Glutathion-S-yl)-R-methyl-dopamine on dopamine, serotonin and norepinephrine concentrations following intracerebroventricular administration to male Sprague-Dawley rats. Chem. Res. Toxicol. 9, 457-465. (14) Miller, R. T., Lau, S. S., Monks, T. J. (1997) 2,5-bis-(GlutathionS-yl)-R-methyldopamine, a putative metabolite of (()-3,4-methylenedioxyamphetamine, decreases brain serotonin concentrations. Eur. J. Pharmacol. 323, 173-180. (15) Bai, F., Lau, S. S., and Monks, T. J. (1999) Glutathione and N-acetylcysteine conjugates of R-methyldopamine produce serotonergic neurotoxicity. Possible role in methylenedioxy-amphetamine-mediated neurotoxicity. Chem. Res. Toxicol. 12, 11501157. (16) Miller, R. T., Lau, S. S., and Monks, T. J. (1995) Metabolism of 5-(glutathion-S-yl)-R-methyldopamine following intracerebroventricular administration to male Sprague-Dawley rats. Chem. Res. Toxicol. 8, 634-641. (17) Meister, A., and Tate, S. S. (1976) Glutathione and related γ-glutamyl compounds. Biosynthesis and utilization. Annu. Rev. Biochem. 45, 559-604. (18) O’Shea, E., Granados, R., Esteban, B., Colado, M. I., and, Green, A. R. (1998) The relationship between the degree of neurodegeneration of rat brain 5-HT nerve terminals and the dose and frequency of administration of MDMA (‘ecstasy’). Neuropharmacology 37, 919-926. (19) Stone, D. M., Stahl, D. C., Hanson, G. R., and Gibb, J. W. (1986) The effects of 3,4-methylenedioxymethamphetamine (MDMA) and 3,4-methylenedioxyamphetamine (MDA) on monoaminergic systems in the rat brain. Eur. J. Pharmacol. 128, 41-48. (20) Schmidt, C. J., Black, C. K., and Taylor, V. L. (1990) Antagonism of the neurotoxicity due to a single administration of methylenedioxymethamphetamine. Eur. J. Pharmacol. 181, 59-70. (21) Bradford, M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248-254. (22) Streit, W. J. (2000) Microglial response to brain injury: A brief synopsis. Toxicol. Pathol. 28, 28-30. (23) Streit, W. J., Walter, S. A., and Pennell, N. A. (1999) Reactive microgliosis. Prog. Neurobiol. 57, 563-581. (24) McGovern, J. P., Williams, M. G., Tang, A. H., von Voightlander P. F., Piercey, M. F., Einspahr, F. J., and Schreur, P. J. (1989) Animal behavioral and neurochemical effects of the CNS toxic amino acid antitumor agent, acivicin. Res. Commun. Chem. Pathol. Pharmacol. 63, 215-229. (25) Chikhale, E. G., Chikhale, P. J., and Borchardt, R. T. (1995) Carrier-mediated transport of the antitumor agent acivicin across the blood-brain barrier. Biochem. Pharmacol. 49, 941-945. (26) Monks, T. J., and Lau, S. S. (1997) Biological reactivity of polyphenolic-glutathione conjugates. Chem. Res. Toxicol. 10, 1296-1313. (27) Oldendorf, W., and Szabo, J. (1976) Amino acid assignment to one of three blood-brain barrier amino acid carriers. Am. J. Physiol. 230, 94-98. (28) Reichert, D., Neudecker, T., Spengler, U., and Henschler, D. (1983) Mutagenicity of dichloroacetylene and its degradation products trichloroacetyl chloride, trichloroacryloyl chloride and hexachlorobutadiene. Mutat. Res. 117, 21-29. (29) Reichert, D., Spengler, U., Romen, W., and Henschler, D. (1984) Carcinogenicity of dichloroacetylene: an inhalation study. Carcinogenesis 5, 1411-1420. (30) Patel, N. J., Fullone, J. S., and Anders, M. W. (1993) Brain uptake of S-(1,2-dichlorovinyl)glutathione and S-(1,2-dichlorovinyl)-Lcysteine, the glutathione and cysteine S-conjugates of the neurotoxin dichloroacetylene. Brain Res. Mol. Brain Res. 17, 53-58. (31) Monks, T. J., and Lau, S. S. (1998) The pharmacology and toxicology of polyphenolic-glutathione conjugates. Annu. Rev. Pharmacol. Toxicol. 38, 229-255. (32) Kannan, R., Kuhlenkamp, J. F., Jeandidier, E., Trinh, H., Ookhtens, M., and Kaplowitz, N. (1990) Evidence for carriermediated transport of glutathione across the blood-brain barrier in the rat. J. Clin. Invest. 85, 2009-2013. (33) Kannan, R., Mittur, A., Bao, Y., Tsuruo, T., and Kaplowitz, N. (1999) GSH transport in immortalized mouse brain endothelial cells: evidence for apical localization of a sodium-dependent GSH transporter. J. Neurochem. 73, 390-399. (34) Kannan, R., Chakrabarti, R., Tang, D., Kim, K. J., and Kaplowitz, N. (2000) GSH transport in human cerebrovascular endothelial cells and human astrocytes: evidence for luminal localization of Na+-dependent GSH transport in HCEC. Brain Res. 852, 374382. (35) Kannan, R., Kuhlenkamp, J. F., Ookhtens, M., and Kaplowitz, N. (1992) Transport of glutathione at blood-brain barrier of the

870

(36)

(37) (38) (39)

Chem. Res. Toxicol., Vol. 14, No. 7, 2001 rat: inhibition by glutathione analogs and age-dependence. J. Pharmacol. Exp. Ther. 263, 964-970. Hill, K. E., Van Hoff, D. D., and Burk, R. E. (1985) Effect of inhibition of γ-glutamyl transpeptidase by AT-125 (acivicin) on glutathione and cysteine levels in rat brain and plasma. Invest. New Drugs 3, 31-34. Scott, R. D., and Curthoys, N. P. (1987) Renal clearance of glutathione measured in rats pretreated with inhibitors of glutathione metabolism. Am. J. Physiol. 252, F877-F882. Bolton, J. L., Trush, M. A., Penning, T. M., Dryhurst, G., and Monks T. J. (2000) Role of quinones in toxicology. Chem. Res. Toxicol. 13, 135-160. Li, H., Shen, X. M., and Dryhurst, G. (1998) Brain mitochondria catalyze the oxidation of 7-(2-aminoethyl)-3,4-dihydro-5-hydroxy2H-1,4-benzothiazine-3-carboxyli c acid (DHBT-1) to intermediates that irreversibly inhibit complex I and scavenge glu-

Bai et al. tathione: potential relevance to the pathogenesis of Parkinson’s disease. J. Neurochem. 71, 2049-2062. (40) Xin, W., Shen, X. M., Li, H., and Dryhurst, G. (2000) Oxidative metabolites of 5-S-cysteinylnorepinephrine are irreversible inhibitors of mitochondrial complex I and the alpha-ketoglutarate dehydrogenase and pyruvate dehydrogenase complexes: possible implications for neurodegenerative brain disorders. Chem. Res. Toxicol. 13, 749-760. (41) Wrona, M. Z., Waskiewicz, J., Han, Q. P., Han, J., Li, H., and Dryhurst, G. (1997) Putative oxidative metabolites of 1-methyl6-hydroxy-1,2,3,4-tetrahydro-β-carboline of potential relevance to the addictive and neurodegenerative consequences of ethanol abuse. Alcohol 14, 213-223.

TX010011L