The recreational drug ecstasy (3,4-methylenedioxymethamphetamine; MDMA) is often used at “raves” or dance parties (Irvine et al., 2006), which are characterized by marathon, vigorous group dancing with laser light displays, hypnotic electronic music, hot environmental conditions, and concurrent use of other illicit drugs (Schwartz and Miller, 1997; Weir, 2000). In this setting, ecstasy can induce a hyperthermic response that is life-threatening (Mallick and Bodenham, 1997). Ambient temperature, which is often elevated at these dance parties, is known to alter the thermogenic (Gordon et al., 1991; Von Huben et al., 2007), neurochemical (O'Shea et al., 2005), behavioral (Cornish et al., 2003), and neurotoxic effects of MDMA (Malberg and Seiden, 1998). However, the interaction of ambient temperature and MDMA pharmacokinetics has not been investigated.

MDMA thermodysregulation in laboratory animals is dependent upon the ambient temperature. That is, a hyperthermic response occurs at ambient temperatures greater than 24°C, and a hypothermic response occurs at ambient temperatures of 20°C or lower (Gordon et al., 1991; Taffe et al., 2006; Von Huben et al., 2007). Although a hyperthermic response to MDMA is not essential for subsequent serotonergic neurotoxicity (Shankaran et al., 1999; Falk et al., 2002; Darvesh and Gudelsky, 2004; Jones et al., 2005), experimental manipulations such as pharmacological agents (Farfel and Seiden, 1995; Malberg et al., 1996), surgical removal of pituitary and thyroid glands (Sprague et al., 2003), and changes in ambient temperature (Malberg and Seiden, 1998) that attenuate MDMA-induced hyperthermia also attenuate the subsequent neurotoxic effects. A better understanding of MDMA disposition in nonhuman primates at cool, room, and warm ambient temperatures may provide insight into any relationship between ambient temperature, thermoregulation, MDMA metabolism, and subsequent neurotoxicity.

MDMA disposition has been assessed previously in rodents, nonhuman primates, and humans (de la Torre and Farré, 2004; de la Torre et al., 2004; Monks et al., 2004). In brief, the major metabolic pathway in humans involves O-demethylenation to 3,4-dihydroxymethamphetamine (HHMA) by the cytochrome P450 isozyme 2D6 (CYP2D6; Tucker et al., 1994; Kreth et al., 2000; Fig. 1). A minor pathway in humans involves N-demethylation to 3,4-methylenedioxyamphetamine (MDA) by CYP2B6, which can be further O-demethylenated to 3,4-dihydroxyamphetamine (HHA; Maurer et al., 2000; Segura et al., 2001; Fig. 1). HHMA also can be N-demethylated to HHA by CYP2B6. HHMA and HHA have been postulated by some to be the neurotoxic metabolites of MDMA (de la Torre and Farré, 2004; Jones et al., 2005). Species differences in MDMA metabolism are apparent between rodents and humans. For instance, the major pathway of MDMA metabolism in rodents is N-demethylation to MDA compared with O-demethylenation to HHMA in humans (Lim and Foltz, 1988; de la Torre et al., 2004). In contrast, little is known about MDMA metabolism in nonhuman primates. To date, studies in nonhuman primates have assessed only plasma MDMA and MDA concentrations (Bowyer et al., 2003; Mechan et al., 2006). Thus, there is a critical gap in our knowledge of the metabolic disposition of MDMA in nonhuman primates, regardless of ambient temperature.

The aims of the present study were to characterize MDMA disposition in nonhuman primates, to examine the relationship between the thermodysregulatory effects of MDMA (1.5 mg/kg i.v.) and three ambient temperatures (18°C, 24°C, and 31°C), and to measure plasma concentrations of MDMA and its metabolites (HHMA, MDA, or HHA). We hypothesized that MDMA would induce a hypothermic response at 18°C, whereas at 24°C, a hyperthermic response would be observed that was potentiated at 31°C. We also hypothesized that MDMA metabolism would be similar to humans and that HHMA would be the primary metabolite. One potential mechanism by which lowering the ambient temperature protects, and elevating the ambient temperature potentiates, the serotonergic neurochemical consequences of MDMA could be changes in metabolism and plasma concentrations of the putative neurotoxic metabolites HHMA and HHA. We hypothesized that lowering the ambient temperature would decrease HHMA and HHA plasma concentrations, whereas elevating the ambient temperature would increase HHMA and HHA plasma concentrations.

Materials and Methods

Subjects. Adult male rhesus monkeys (Macaca mulatta; n = 10) served as subjects. Four monkeys were drug-naïve, and six monkeys had previous histories of cocaine self-administration (Martelle et al., 2007) but had not self-administered cocaine or any other psychostimulant for at least 3 months before the start of the study. Four subjects were randomly selected and first exposed to MDMA at room temperature, and three subjects were first exposed to MDMA at the warm ambient temperature. After both room and warm ambient temperature studies, five subjects (including some of the earlier subjects) were exposed to MDMA at the cool ambient temperature. Approximately 21 days separated MDMA administrations. Each monkey was surgically prepared with an indwelling intravenous catheter into the femoral vein and a subcutaneous vascular access port (VAP; Access Technologies, Skokie, IL) as described previously (Martelle et al., 2007). All subjects were trained to sit calmly in a primate restraint chair (Primate Products, Redwood City, CA) during the monitoring period. Monkeys were weighed weekly and fed sufficient food daily (LabDiet High Protein Primate Chow no. 5045 and fresh fruit) to maintain 95% of free-feeding weights; water was available ad libitum in the home cage, where monkeys had visual and auditory contact with each other. The facilities for housing and care of the animals are accredited by the American Association for the Assessment and Accreditation of Laboratory Animal Care. All procedures were approved by the Animal Care and Use Committee of Wake Forest University and performed in accordance with established practices as described in the National Research Council (2003) Guidelines for the Care and Use of Mammals in Neuroscience and Behavioral Research, except the elevation of ambient temperature that was outside the Guide's recommended range. During these experiments, no monkey had a core temperature above 41°C that required intervention or termination of the experiment. Environmental enrichment was provided as outlined in the Animal Care and Use Committee of Wake Forest University Nonhuman Primate Environmental Enrichment Plan.

Ambient Temperature Manipulation. While in the restraint chair, subjects were placed in a ventilated and sound-attenuating behavioral chamber (150 × 74 × 76 cm; Med Associates, East Fairfield, VT). An indoor/outdoor thermometer (Oregon Scientific, Tualatin, OR) was threaded through a hole in the chamber and attached to the chamber wall. The digital readout was placed on top of the chamber to allow for continuous recording and adjustments to maintain the ambient temperature within ±1°C of target temperature without opening the chamber door. To elevate the ambient temperature, a commercially available space heater was placed outside one of the vent holes and warm air forced into the chamber. To lower the ambient temperature, a commercially available air conditioner was placed outside the vent holes and cool air forced into the chamber. Monkeys were placed into the chamber and allowed to acclimate to the ambient temperature for approximately 20 min before the start of the study.

Temperature Monitoring. Rectal temperatures (RET-1 rectal probe attached to a TH-3 thermocouple; Physitemp, Clifton, NJ) were recorded immediately before MDMA administration and each time point corresponding with a blood sample (see below).

Blood Sampling. Blood samples were taken from the subcutaneous VAP immediately before MDMA administration (1.5 mg/kg i.v.) and at 5, 10, 15, 30, 60, 90, 120, 180, and 240 min post-MDMA administration. After MDMA, the VAP was flushed with 5 ml of sterile saline. At each time point, approximately 3 ml of blood was withdrawn using a disposable, sterile, 3-ml syringe attached to a Huber needle. The first milliliter was discarded to prevent dilution from heparinized saline (100 U/ml) in the VAP and catheter. Following each blood sample, approximately 2 ml of heparinized saline was flushed to maintain patency. Blood samples were immediately dispensed into chilled 2-ml Vacutainer tubes containing K2 EDTA 3.6-mg solution (Becton-Dickinson, Franklin Lakes, NJ) and stored on ice for up to 20 min until centrifugation. Samples were centrifuged at 3000 rpm for 5 to 10 min at 4°C; plasma was withdrawn using a disposable pipette and decanted into labeled 1.5-ml micro-centrifuge tubes. Next, 250 mM sodium metabisulfite was added (30 μl) to the plasma sample as described in Segura et al. (2001) to minimize oxidation of MDMA and its metabolites. Samples were immediately vortexed and stored at -30°C until analysis.

Materials and Reagents. Racemic MDMA (HCl and base) and MDA base were provided by the National Institute on Drug Abuse Drug Supply Program (Research Triangle Park, NC). MDMA HCl was dissolved in 0.9% saline at a concentration of 50 mg/ml for thermoregulation and metabolism studies. MDMA and MDA were dissolved in methanol for stock and working solutions for analysis. HHMA and HHA were synthesized as described previously (Ensslin et al., 1996; Pizarro et al., 2002) and prepared in methanol for stock and working solutions. Sodium metabisulfite (Sigma, St. Louis, MO) was dissolved in sterile water.

Analytical Procedure. MDMA, HHMA, HHA, and MDA concentrations were analyzed using a modification of previously described methods (Segura et al., 2001, 2002) using acid hydrolysis, solid-phase extraction (SPE) using strong cation exchange (SCX) SPE columns, and liquid chromatography/tandem mass spectrometry analysis.

Calibration, Working, and Control Solutions. For each analytical run, calibration standards ranging from 1 to 1000 ng/ml were prepared in clean, separate, silanized 16 × 100 mm glass culture tubes. The appropriate amount of standard working solution was added to 1 ml of 4-fold diluted blank rhesus monkey plasma [0.25 ml of monkey plasma plus 0.75 ml of Milli-Q water (Millipore Corporation, Billerica, MA)] that also contained 50 μl of 250 mM ascorbic acid, 50 μl of 250 mM sodium EDTA, 50 μl of 250 mM sodium bisulfite, and 50 μl of 4-methyl catechol. The control working solutions were used to prepare quality control samples at 35, 100, and 650 ng/ml.

Sample Preparation. Plasma samples were analyzed as 4-fold dilutions (0.25 ml of sample plus 0.75 ml of Milli-Q water), 10-fold dilutions (0.10 ml of sample plus 0.90 ml of Milli-Q water), or 50-fold dilutions (0.02 ml of sample plus 0.98 ml of Milli-Q water). The samples were transferred to separate silanized glass culture tubes that contained 50 μl of 250 mM ascorbic acid, 50 μl of 250 mM sodium EDTA, 50 μl of 250 mM sodium bisulfite, and 50 μl of 4-methyl catechol.

Extraction Preparation. For the internal standard addition, 50 μl containing 1 ng/μl MDMA-d₅ and MDA-d₅ was transferred to each plasma sample tube. The tubes were then vortexed. For hydrolysis, a 1-ml volume of 0.5 N hydrochloric acid was added to each tube. Teflon-lined screw caps were placed on each tube, and the tubes were vortexed. The tubes were then placed in a 100°C oven and allowed to incubate for 30 min. The tubes were then removed from the oven and allowed to cool to room temperature. One hundred microliters of 70% perchloric acid was added to each tube to precipitate proteins. The tubes were immediately vortexed and then centrifuged for 10 min at 2000 rpm. The supernatants were collected into clean silanized 16 × 100-mm culture tubes. The pH of the sample preparation supernatants was adjusted to pH 6 (as assessed using pH paper) by adding 1 ml of 1 M potassium phosphate, followed by 5 drops of 10 N sodium hydroxide. The tubes were centrifuged as described above and reserved for the SPE.

Solid-Phase Extraction. Bond Elute LRC SCX (Varian, Palo Alto, CA) SPE columns (1 ml, 100 mg) were used for the extraction procedure. A separate SPE column was used for each submitted sample, calibration standard, or quality control sample. The SPE columns were conditioned by the successive addition of 2 ml of methanol, 2 ml of Milli-Q water, and 2 ml of 1 M potassium phosphate. Vacuum was not used in the conditioning steps. The sample preparation supernatants were added to their respective conditioned SPE columns. After the addition of the samples, the columns were washed by successive additions of 1 ml of Milli-Q water and 4 ml of methanol. After the methanol passed through the SPE columns, vacuum was applied (>10 in mercury) for 2 min to dry the columns. After drying, the SPE columns were eluted by adding 2 ml of methanol/12 M hydrochloric acid (99:1). The eluted samples were collected into clean, separate 5-ml conical glass tubes. A Turbovap was used to evaporate the extracts under a stream of nitrogen in 40°C water. Following evaporation, 200 μl of methanol was added to reconstitute the extracts. To remove particulate material in the extracts, the conical tubes were centrifuged for 10 min at 2000 rpm. The supernatants were transferred to clean separate 5-ml conical glass tubes. The cleaned extracts were evaporated as described above and then reconstituted with 100 μl of 0.1% formic acid/methanol (97:3). The extracts were transferred to clean separate plastic 11-mm conical autosampler vials.

Liquid Chromatography-Mass Spectrometry Analysis. A Thermo-Finnigan TSQ 7000 MS-MS (Thermo Fisher Scientific, Waltham, MA) coupled with an Agilent 1100 HPLC system (Agilent Technologies, Palo Alto, CA) was used for the analysis. Chromatographic separation used a Metasil Basic, 100 × 3.0-mm, 3-μ HPLC column (Varian). The mobile phase consisted of a 0.1% formic acid and methanol gradient that varied from 97:3 to 88:12 at a flow rate of 0.2 ml/min. The analytes were ionized by electrospray ionization. The capillary column temperature was set at 250°C. Selected reaction monitoring was used. The following transitions were monitored: MDMA (m/z 194 → 163), MDMA-d₅ (m/z 199 → 165), MDA (m/z 180 → 163), MDA-d₅ (m/z 185 → 168), HHMA (m/z 182 → 151), and HHA (m/z 168 → 151). MDA-d₅ was also used as the internal standard for HHMA and HHA.

Statistical Analysis. Core temperature data were analyzed with repeated measures mixed linear regression using SAS Proc Mixed (version 8.2; SAS, Cary, NC) with ambient temperature and time as factors. Tests of the main effects of ambient temperature and time and their interaction were performed using F-tests. In the presence of a significant interaction, post hoc least-squares means difference tests were performed for planned comparisons between: 1) core temperature at time 0 min and core temperature at each time point after MDMA within an ambient temperature, and 2) each ambient temperature within time 0.

Maximum concentration (C_max), time of maximum concentration (T_max), volume of distribution at steady state (Vd_ss), clearance (CL), area under the concentration-time curve (AUC), and elimination half-life (t_½) were determined using a noncompartmental intravenous bolus model (WinNonlin Professional 5.1; Pharsight Co., Mountain View, CA). Pharmacokinetic data were analyzed with repeated measures mixed linear regression. Ambient temperature was the main factor, and a significant main effect was followed by a post hoc least-squares means difference test. All results were considered significant at the 95% level of confidence (p < 0.05).

Results

Effects of Ambient Temperature on MDMA-Induced Hyperthermia. Saline administration at the cool, room, and warm ambient temperatures did not significantly change core temperature over the 240-min monitoring period (data not shown). When 1.5 mg/kg MDMA was administered intravenously, a significant ambient temperature and time interaction was detected (F_14,70 = 7.44, p < 0.05). Post hoc analysis revealed that when administered at room temperature (24°C), MDMA did not significantly change core temperature over the 240-min sampling period. In contrast, at 18°C, this dose of MDMA induced a significant decrease (p < 0.05) in core temperature starting at 60 min and lasting through 240 min (Fig. 2). At 31°C, MDMA induced a significant increase (p < 0.05) in core body temperature starting at 30 min and lasted through 240 min. Between ambient temperatures, there was a significant difference (p < 0.05) only between baseline temperatures at 31°C and 24°C.

Effects of Ambient Temperature on MDMA Disposition. At room temperature (24°C), MDMA C_max reached T_max at 5 min and was not significantly different when tested at different ambient temperatures (Fig. 3 and Table 1). HHMA C_max reached T_max between 60 and 120 min after MDMA administration regardless of ambient temperature. HHA C_max reached T_max at ∼90 min regardless of ambient temperature. MDA C_max also was not significantly affected by ambient temperature. In contrast to the C_max data, there was a significant main effect of ambient temperature on MDA AUC_0-240min (F_2,3 = 11.1, p < 0.05). Post hoc analysis revealed that the AUC at 18°C was significantly greater (p < 0.05) than either 24°C or 31°C (Table 1).

Discussion

The aim of the present study was to examine the relationships between ambient temperature, MDMA-induced changes in core temperature and its metabolism. MDMA (1.5 mg/kg i.v.) did not affect core temperature when administered at room temperature (24°C) but induced a significant hypothermic response at 18°C and a significant hyperthermic response at 31°C. However, the metabolic disposition of MDMA could not account for the effects of ambient temperature on thermoregulation. Similar to humans, a major route of MDMA metabolism in nonhuman primates involves O-demethylenation to HHMA. After HHMA, HHA had the next highest metabolite/MDMA ratio, followed by MDA. At 24°C, the t_½ of MDMA was approximately 90 min. Both increasing and decreasing the ambient temperature significantly increased this approximate t_½. In addition, lowering the ambient temperature altered MDMA disposition such that a significantly greater MDA AUC was observed, with a trend for a decreased HHMA AUC compared with both 24°C and 31°C. These findings indicate a more complex interaction between MDMA and central nervous system function than simply ambient temperature induced changes in metabolism.

Previous nonhuman primate MDMA pharmacokinetic studies (Bowyer et al., 2003; Mechan et al., 2006) have reported only MDMA and MDA plasma concentrations. Thus, data are lacking in nonhuman primates on plasma concentrations of HHMA, which is a major metabolite in humans. Limited in vitro evidence (Bogaards et al., 2000) suggests that the rats and monkeys have cytochrome P450 activity similar to humans for the specific isozymes involved in MDMA metabolism. In contrast to the in vitro findings, however, in vivo studies suggest that species differences exist. MDA is a major metabolite in rats (Chu et al., 1996), whereas HHMA is a major metabolite in monkeys (present study) and humans (Segura et al., 2001, 2002). The results of the present study confirm that MDMA metabolism in monkeys is similar to humans, providing additional support for the use of monkeys to model human MDMA abuse.

A direct comparison between the present study and previous pharmacokinetic studies in monkeys (Bowyer et al., 2003; Mechan et al., 2006) is complicated by the use of different doses and routes of administration. Nevertheless, several general comparisons can be drawn. Although a lower dose was used, the MDMA C_max in the present study is at least 5-fold higher than reported by Mechan et al. (2006) or Bowyer et al. (2003). Furthermore, MDMA t_½ (∼1.5 h at 24°C) in the present study is decreased compared with 2.8 and ∼3.5 h in the Bowyer et al. (2003) and Mechan et al. (2006) studies, respectively. These differences in C_max and t_½ can be explained in part by differences in routes of administration. Future studies in nonhuman primates should extend these characterizations of ambient temperature to oral MDMA administration, which is the primary route used by humans.

Previous research in rodents (Gordon et al., 1991) and monkeys (Von Huben et al., 2007) has demonstrated that ambient temperature can alter the thermogenic effects of MDMA, with MDMA inducing a hypothermic effect at ambient temperatures lower than 20°C and a hyperthermic effect at ambient temperatures greater than 24°C. In contrast to Von Huben et al. (2007), in the present study, MDMA did not significantly change core temperature at 24°C. One potential reason for this discrepancy could be monkeys were seated in restraint chairs in the present study compared with freely moving in the Von Huben et al. (2007) study. However, MDMA-induced hyperthermia in monkeys (and rodents for that matter) does not seem to be related to motor activity (Crean et al., 2006; Taffe et al., 2006; Von Huben et al., 2007). A more likely explanation is that core body temperature was measured in the present study, whereas Von Huben et al. (2007) measured temperature using subcutaneous telemetry. This conclusion implies that at least two different processes are operating following MDMA administration that influence thermoregulation. Rodent models have demonstrated that the hyperthermia induced by MDMA results from the activation of the sympathetic nervous system and the hypothalamic-pituitary-thyroid/adrenal axis. The release of norepinephrine mediated by MDMA generates heat through regulation of uncoupling proteins by α₁- and β₃-adrenoreceptors and loss of heat dissipation through norepinephrine-mediated vasoconstriction (for a review, see Mills et al., 2004).

Curiously, the t_½ of MDMA was increased both by lowering and elevating the ambient temperature. These data should be considered as approximate estimations of t_½ because the 240-min sampling period was probably too short to reliably estimate the t_½. Although both CL and Vd_ss are factors affecting t_½, there was no statistically significant difference in CL or Vd_ss at the different ambient temperatures. There was a trend, however, for a smaller Vd_ss at 18°C, which could increase t_½. Consistent with this possibility, Bansinath et al. (1988) examined the effects of hypo- and hyperthermia on morphine disposition and found increases in t_½ and decreases in Vd_ss in hypothermic, but not hyperthermic, dogs compared with normothermic controls. Another potential reason for the increased t_½ at 18°C could be hypothermia-induced impairment of biliary secretion and microsomal function in the liver (Kalser et al., 1968). MDMA can also induce cutaneous vasoconstriction regardless of ambient temperature (Pedersen and Blessing, 2001), which would presumably decrease the Vd_ss. In the present study, however, Vd_ss was similar at both 24°C and 31°C, despite different thermogenic responses. Thus, the reasons for similar effects on t_½ of both increasing and decreasing the ambient temperature are unclear but may involve changes in peripheral circulation.

In conclusion, the results of the present study indicate that a major metabolite of MDMA in nonhuman primates is HHMA, the major MDMA metabolite in humans, confirming that nonhuman primates are a more analogous model than rodents for the pharmacokinetic characterization of MDMA. These results also indicate that either lowering or elevating the ambient temperature increases the t_½ of MDMA. Furthermore, lowering the ambient temperature significantly altered MDMA disposition, shunting metabolism toward the N-demethylation pathway and away from the O-demethylenation pathway. In contrast, elevating the ambient temperature did not alter the metabolic pathway of MDMA. If changes in ambient temperature and/or core temperature differentially alter MDMA disposition, these results may have implications for the study of MDMA-induced serotonergic neurochemical consequences.