Abstract
“Ecstasy” [3,4-methylenedioxymetamphetamine (MDMA)] is of considerable interest in light of its prosocial properties and risks associated with widespread recreational use. Recently, it was found to bind trace amine-1 receptors (TA1Rs), which modulate dopaminergic transmission. Accordingly, using mice genetically deprived of TA1R (TA1-KO), we explored their significance to the actions of MDMA, which robustly activated human adenylyl cyclase-coupled TA1R transfected into HeLa cells. In wild-type (WT) mice, MDMA elicited a time-, dose-, and ambient temperature-dependent hypothermia and hyperthermia, whereas TA1-KO mice displayed hyperthermia only. MDMA-induced increases in dialysate levels of dopamine (DA) in dorsal striatum were amplified in TA1-KO mice, despite identical levels of MDMA itself. A similar facilitation of the influence of MDMA upon dopaminergic transmission was acquired in frontal cortex and nucleus accumbens, and induction of locomotion by MDMA was haloperidol-reversibly potentiated in TA1-KO versus WT mice. Conversely, genetic deletion of TA1R did not affect increases in DA levels evoked by para-chloroamphetamine (PCA), which was inactive at hTA1 sites. The TA1R agonist o-phenyl-3-iodotyramine (o-PIT) blunted the DA-releasing actions of PCA both in vivo (dialysis) and in vitro (synaptosomes) in WT but not TA1-KO animals. MDMA-elicited increases in dialysis levels of serotonin (5-HT) were likewise greater in TA1-KO versus WT mice, and 5-HT-releasing actions of PCA were blunted in vivo and in vitro by o-PIT in WT mice only. In conclusion, TA1Rs exert an inhibitory influence on both dopaminergic and serotonergic transmission, and MDMA auto-inhibits its neurochemical and functional actions by recruitment of TA1R. These observations have important implications for the effects of MDMA in humans.
Introduction
The monoamine releaser and reuptake suppressor, “ecstasy” [3,4-methylenedioxymetamphetamine (MDMA)], facilitates prosocial behaviors in animals and humans (Bedi et al., 2009; Johansen and Krebs, 2009), actions related to enhanced release of oxytocin in centers controlling social behavior, like amygdala and septum (McGregor et al., 2008). However, greater attention has been devoted to the psychostimulant properties of MDMA and its widespread recreational use (Vollenweider et al., 1999, 2002). In addition, MDMA can trigger a serotonergic syndrome and disrupt thermoregulatory mechanisms, neurotoxic effects especially prominent in young adults under conditions of dehydration (Schifano, 2004; Baumann et al., 2007; Karlsen et al., 2008). It is thus important to elucidate the precise mechanisms of action of MDMA.
MDMA displaces neuronal serotonin (5-HT) via reversal of vesicular and plasma membrane transporters (Rudnick and Wall, 1992; Mlinar and Corradetti, 2003). Further, it prevents 5-HT reuptake (Iravani et al., 2000) and inhibits the activity of the catabolytic enzyme monoamine oxidase A (Leonardi and Azmitia, 1994). Concomitant elevations in dopamine (DA) and noradrenaline (NA) levels occur either via a similar pattern of direct actions upon catecholaminergic pathways (Gough et al., 2002; Baumann et al., 2005, 2008) and/or events downstream of serotonergic transmission, such as activation of 5-HT2A receptors (Gobert et al., 2000). Moreover, at high concentrations, MDMA binds to 5-HT2A, histamine H1, and muscarinic M1/M2 receptors, though their putative roles in its functional actions remain poorly characterized (Green et al., 2003).
A potentially novel dimension to the neurobiology of MDMA was unveiled by the suggestion that it recognizes Gs-coupled TA1 receptors [TA1Rs (sometimes called TAAR1s)] (Bunzow et al., 2001; Lindemann and Hoener, 2005; Maguire et al., 2009). While most subtypes of trace amine (TA) receptor are preferentially expressed in olfactory epithelium (Liberles and Buck, 2006), TA1Rs are mainly distributed in limbic structures and regions containing monoaminergic perykarias (Borowsky et al., 2001; Wolinsky et al., 2007; Lindemann et al., 2008). The TAs, tyramine, tryptamine, and β-phenethylamine are implicated in psychiatric and neurological disorders associated with monoaminergic dysfunction (Branchek and Blackburn, 2003; Berry, 2004; Burchett and Hicks, 2006; Sotnikova et al., 2008), yet it remains unproven that they act via TA1Rs in vivo. Conversely, the thyroxine derivatives 3-iodothyronamine (T1AM) and o-phenyl-3-iodotyramine (o-PIT) behave as agonists at TA1Rs, and their hypothermic actions are blunted in TA1 knock-out (TA1-KO) mice (Scanlan et al., 2004; Hart et al., 2006; Doyle et al., 2007) (M.J. Millan, unpublished observations).
Likewise of pertinence to MDMA, TA1Rs inhibit the electrical and synthetic activity of nigrostriatal dopaminergic neurons (Geracitano et al., 2004; Lindemann et al., 2008; Bradaia et al., 2009; Ledonne et al., 2010). Moreover, functional interactions have been reported between colocalized TA1R and DA transporters (Xie and Miller, 2008; Xie et al., 2008). Interestingly, T1AM and o-PIT suppress locomotor activity in mice (Scanlan et al., 2004), actions opposite to those of MDMA (Baumann et al., 2008).
Collectively, these observations raise the intriguing possibility that TA1R may modulate the actions of MDMA. Though EPPTB (N-(3-ethoxy-phenyl)-4-pyrrolidin-1-yl-3-trifluoromethyl-benzamide) was described as a selective TA1R antagonist (Bradaia et al., 2009; Stalder et al., 2011), it is not centrally active upon systemic administration. The present study addressed the role of TA1R in the physiological, neurochemical, and behavioral actions of MDMA by exploiting TA1-KO mice.
Materials and Methods
Construction of the targeting vector.
Targeting vector construction and knockout strategy have been designed and performed by genOway. Genomic clones containing the murine Taar1 locus were isolated from a 129S6/SvEvTacRPCI-22 bacterial artificial chromosome (BAC) genomic library using a probe corresponding to the murine Taar1 (nucleotides 45–950, GenBank accession #AF380187). Two BAC clones (37K19 and 393O10) containing Taar1 locus have been isolated. The genomic organization of targeted locus was determined by subcloning XbaI (6 kb) and SpeI (5 kb) genomic fragments into the pZEro-2 vector (Invitrogen). The XbaI and SpeI genomic inserts were sequenced and used to construct the targeting vector. Briefly, a 4.4 kb XbaI-NsiI fragment comprising Taar1 exons 1 (4334 bp 5′UTR + 120 bp of coding region) and a 1.1 kb EcoRI-BglII fragment located downstream of Taar1 exon 1 (122 bases of CDS + 1046 bp 3′ UTR), were used to flank an IRES LacZ-NEO-tk cassette (IRES LacZ-LoxP site-PGK promoter-NEO-tk fusion cDNA-LoxP) (Fig. 1A). The insertion of an IRES-LacZ sequence into the coding region of Taar1 allowed the expression of a truncated TA1 protein of 40 aa instead of 332 aa for the entire protein. The insertion of the selection cassette at 3′ of the EcoRI-BglII fragment left 122 bp of the 3′ Taar1 coding sequence. A negative diphtheria toxin selection cassette was introduced at the 5′ of the short arm of homology.
Screening of Taar1-targeted ES cell clones.
SmaI-linearized targeting vector was transfected by electroporation into 129SvPas ES cells (108 ES cells in the presence of 100 μg of linearized plasmid, 260 V, 500 μF). Positive selection was achieved 48 h later by addition of 200 μg/ml G418 (150 μg/ml active component, Life Technologies). Ninety-nine resistant clones were isolated and amplified in 96-well plates (duplicates). The set of plates containing ES cell clones amplified on gelatin were screened by PCR and further confirmed by Southern blot. The 3′ PCR screening conditions were as follows: GW224 primer specific for the Neo-Tk selection cassette (5′-GGCTGCTAAAGCGCATGCTCCAGAC-3′); and GW475 primer that hybridizes the Taar1 gene (5′-CCGCGGACAGTTTTATGGTGTAACCCTGCCTGACCTG-3′). PCR conditions are 94°C/5 min; 35 cycles of 94°C/30 s and 68°C/2 min 30 s; and then 68°C/7 min, which resulted in a 1505 bp band for the mutated allele. PCR reactions were performed using Long Expand High Fidelity polymerase (Roche) and reaction buffer 3. The 5′ PCR screening conditions were as follows: GW114 primer specific for the IRES-LacZ cassette (5′-CAGTCACGACGTTGTAAAACGAC-3′), and GW587 primer specific for the Taar1 gene (5′-GGAGTCCATCATGCTAAGGTGTCAGG-3′). PCR conditions were 94°C/5 min; 35 cycles of 92°C/30 s and 58°C/30 s; and then 68°C/7 min, which resulted in a 5374 bp band for the mutated allele. For Southern blot analysis, genomic DNA was digested with XbaI and then hybridized with a 0.8 kb internal probe; Taar1+/− clones gave rise to a 5.4 kb wild-type signal and a 10.5 kb targeted signal. Two clones (1A2 and 1D2) were identified both by PCR and Southern blot as targeting the Taar1 locus (Fig. 1B–D).
Generation of germ line chimera mice and homozygous breeding.
Two floxed mutated Taar1 ES cell clones (namely, 1A2 and 1D2) were microinjected into C57BL/6 blastocysts. They gave rise to male chimeras with a significant ES cell contribution (Agouti coat color). After mating with C57BL/6 females, germ line transmission was confirmed by the genotyping of tail DNA offspring (PCR and Southern blot analyses). Floxed heterozygous F1 animals were discriminated from wild type by PCR using couples primers specific for targeted (GW224/GW475) and wild-type alleles (GW264, 5′-CGACTGGTCAAGAGAAGTCC-3′/GW265, 5′-AGGAGAACCATCTTCAAGGC-3′) and by Southern blot on XbaI-digested genomic DNA using an XbaI/HindIII probe specific for the 5′-targeted region (Fig. 1F,G). Thirty-two heterozygous F1 mice for mutation were generated. Mating of the six heterozygous F1 mice yielded the generation of eight heterozygous and eight homozygous F2 mice for the mutation. Heterozygous and homozygous animals were screened by Southern blot analysis as described above.
Animal.
Homozygous TA1-KO mice were backcrossed on pure C57BL/6J genetic background (WT) for eight generations (Charles River). In all experiments, 5–8-week-old male mice were used. They were housed four to five per cage under a 12 h light/dark cycle and had free access to food and water. Housing (four to five per cage) and experimental procedures (12 h light/dark cycle) were fully compliant with the principles of the Care and Use of Laboratory Animals (European Economic Community directive/86/609).
qRT-PCR.
Brains were dissected and frozen at −80°C, and cerebral structures were removed using a 1.8 mm diameter punch. Poly A+ RNA was extracted using the MagnaPure LC Isolation station and the MagnaPure LC mRNA Isolation Kit I (Roche Molecular Biochemicals). The RNA samples were subjected to a reverse transcription step using the high-capacity cDNA archive Kit (Applied Biosystems). Single-stranded cDNA products were then analyzed by PCR using the ABI PRISM 7900HT Sequence Detection System (Applied Biosystems). References of the assays (Applied Biosystems) were as follows: Mm00447557_m1 [tyrosine hydroxylase (TH)]; Mm00460472_m1 (DβH); Mm00558009_g1 (MAO-A); Mm00555412_m1 (MAO-B); Mm00514377_m1 (COMT); Mm00553058_m1 (VMAT2); Mm01353211_m1 (D1); Mm00438541_m1 (D2); Mm00432887_m1 (D3); Mm00432893_m1 (D4); Mm00658653_s1 (D5); and Mm99999915_g1 (GADPH). The TA1R assay (Applied Biosystems) was a Custom TaqMan Gene Expression assay (Forward primer TTGAGCGCTGTTGGTATTTTGG, Reverse primer GAGGCGGAGCTCAGCAT, Taqman probe CACCAGCACCGATATC). Triplicate threshold cycle (Ct) values were averaged and relative gene expression was analyzed using the ΔCt method (Livak and Schmittgen, 2001) with GADPH as an endogenous control/reference assay.
Histoenzymology.
Brains were dissected, frozen at −25°C/−30°C, and embedded in O.C.T. Compound (Tissue-Tek). Sagittal brain sections (30 μm) were placed on superfrost-plus glass slides (LABOnord) at −20°C. Staining was made with the β-Galactosidase Reporter Gene Staining Kit (GALS; Sigma-Aldrich) according to the manufacturer's instructions. Counterstained (Nuclear Fast Red, 5 min) and dehydrated sections were then mounted with hydrophobic media then digitalized with NanoZoomer (Hamamatsu).
Cell culture and cAMP accumulation.
HeLa cells were incubated at 37°C (95% O2/5% CO2) in DMEM with fetal bovine serum 5%, penicillin 100 Us/ml, and streptomycin 100 μg/ml. Cells were transiently transfected with the 3HA_B2_hTARR1 receptor (Barak et al., 2008) and adenylyl cyclase V plasmids (2 μg each) using the lipofectamine method. Two days later, HeLa cells were plated into 96-well (5.104 cells/well) with partially serum-free medium (0.5% FBS) for 24 h. Accumulation of cAMP was determined using the Alphascreen cAMP detection kit (PerkinElmer). Cells were incubated with MDMA, o-PIT, or PCA for 15 min at 37°C. The buffer was then removed, and the reaction terminated by lysing and freezing the cells (−70°C). Samples were transferred to a 384-well white Optiplate (PerkinElmer) and anti-cAMP acceptor beads, then donor beads with biotinylated cAMP were added. Plates were incubated then read using a Fusion-α microplate reader (PerkinElmer). Experiments were run in triplicate. Data were calculated against a standard curve of cAMP concentrations using Origin Pro 7.5 software (OriginLab).
TH phosphorylation in brain slices.
Striatal slices (300 μm; Vibratome, Leica) were preincubated in Krebs buffer (118 mm NaCl, 4.7 mm KCl, 1.5 mm Mg2SO4, 1.2 mm KH2PO4, 25 mm NaHCO3, 11.7 mm glucose, 1.3 mm CaCl2) at 30°C under constant oxygenation (95% O2/5% CO2) for 60 min, with a change of buffer after 30 min. They were then treated with MDMA for 5 min. The buffer was then removed, and the slices were rapidly frozen on dry ice, sonicated in 1% SDS, and boiled for 10 min. The amount of protein in homogenates was determined using the bicinchoninic acid protein assay method (Pierce). Equal amounts of protein were processed by using 10% acrylamide gels. Immunoblotting was performed with phosphorylation state-specific antibodies against p-Ser19-TH (1:500; AB5425, Millipore), p-Ser31-TH (1:500; AB5423, Millipore, and p-Ser40-TH (1:500; AB5935; Millipore), or antibodies that are not phosphorylation state specific against total TH (1:1000; AB152, Millipore) and actin (1:1000; A3853; Sigma). All antibodies stained bands of correct size. The peptide sequences and procedures to generate the antibodies toward TH have been published previously (Haycock et al., 1998; Lew et al., 1999). The site- and phosphorylation state-specific antibodies show selective reactivity of the phospho versus nonphospho forms of TH (Haycock et al., 1998; Lew et al., 1999). Antibody binding was detected by enhanced chemiluminescence (GE Healthcare) and quantified by densitometry (NIH IMAGE 1.61 software).
Core temperature.
Experiments were performed in a randomized sequence between 10.00 and 12.00:00 A.M. under standard conditions (ambient temperature 21°C). Core temperature was determined by a procedure adapted for mice and using a rectal thermoprobe (Millan et al., 1994). Core temperature was measured under basal conditions before treatment and 30, 60, and 120 min after vehicle or MDMA administration. Under conditions of high ambient temperature, mice were exposed to a temperature of 27°C during the 4 h before the basal core temperature measurement.
Microdialysis in freely moving mice.
Mice were anesthetized with pentobarbital (6 mg/kg, i.p.) before the guide cannula implantation in the dorsal striatum, frontal cortex, or nucleus accumbens. The stereotaxic coordinates were as follows (in millimeters from bregma): dorsal striatum: anteroposterior (AP) +1.0, lateral (L) ±1.9, dorsoventral (DV) −2.2; frontal cortex: AP −2.2, L ±0.3, DV −1.0; and nucleus accumbens: AP −1.3, L ±0.6, DV −4.5. One week later, a 2.0 or 1.0 mm (nucleus accumbens) length CMA/7 probe was inserted and perfused at 1.0 μl/min with PBS (NaCl 147 mm, KCl 4 mm, CaCl2 2.3 mm, pH 7.2). After a 150 min elimination period, 20 min samples were collected at 4°C on 20 μl of ascorbic acid 60 μm, for 240 min. They were stored at −80°C pending injection (35 μl). The mobile phase was composed of sodium dihydrogen citrate 25 mm, sodium decanesulfonate 1.8 mm, and methanol 24% (v/v), pH 5.7, and delivered at 0.3 ml/min. NA, DA, and 5-HT were separated on a BETASIL C18 column (Thermo Fisher Scientific; 150 × 2.1 mm, 3 μm) at 40°C. A coulometric detector (Coulochem III, ESA) with analytical cell settled at E1 = −70 mV and E2 = +300 mV (ESA5041, ESA) was used. The limit of sensitivity was 0.2 pg. Dialysis contents of exogenous MDMA were determined using mass spectrometry (Applied Biosystems). The detection limit was 0.2 ng/ml. Area under the curve (AUC) values were calculated using the standard trapezoidal method and expressed as arbitrary units (percentage × minutes × 10−3).
Synaptosomal preparation.
Dorsal striata were dissected at 4°C, and synaptosomal pellet (P2) fractions prepared by suspending the pellet in glucose 0.32 m (8 mg of protein/ml). Fractions were diluted (1:10) with Krebs–Ringer medium (130 mm NaCl, 3 mm KCl, 1.2 mm MgSO4, 2.5 mm Na2HPO4, 1 mm ascorbic acid, and 20 mm Tris buffer, pH 7.5) and preincubated for 10 min at 37°C in a Dubnoff water bath. Synaptosomes were labeled with 0.1 μm [3H]DA or [3H]5-HT (specific activity 36 and 30 Ci/mmol, respectively; PerkinElmer) for 10 min at 37°C. Suspensions were diluted (1:10) with Krebs–Ringer medium (basal medium, glucose 10 mm, pH 7.5). Aliquots were placed on Millipore filters lying at the bottom of parallel superfusion chambers (37°C). Under a moderate vacuum, basal medium was used to wash synaptosomes (10 ml), which were then superfused at 0.45 ml/min. After 20 min, synaptosomes were superfused with basal medium alone or a medium containing PCA and/or o-PIT. One-minute fractions were collected on 100 μl of a protective solution (ascorbic acid 1%, EDTA 1.5%, and unlabeled neurotransmitter 0.001%). [3H]DA and [3H]5-HT in fractions versus synaptosomes were separated from [3H]metabolites on Biorex-70 columns and counted using a Packard liquid-scintillation β-counter. Experiments were run in quadruplicate. Values are percentage of total [3H]DA or [3H]5-HT recovered.
Locomotor activity.
Testing started at 09:00 A.M. and lasted 180 min. The test cage was made of Plexiglas (27 × 20 × 20 cm) and was illuminated by a 6 W light. Two opposite sets of six photocells were located 6 cm apart, 2 cm above the floor (Hesperid). Every beam interruption corresponded to a locomotion count.
Statistical analysis.
Statistical comparisons of the effect of MDMA in WT versus TA1-KO mice were performed using two-way ANOVA with genotype (gen) and drug (drug) between factors and followed by Tukey's post hoc test. The following three-way ANOVAs were used: plus time (time) for within-factor and repeated measurement in kinetic studies of the effect of MDMA; plus pretreatment (pret) for between-factors analysis in drug interaction studies; and plus ambient temperature (amb) for between-factors analysis in studies for the effect of ambient temperature on thermoregulation. A value of p < 0.05 was considered significant.
Drugs.
S(+)-3–4-methylenedioxymetamphetamine, dl-para-chloroamphetamine, and haloperidol were provided by Sigma. o-Phenyl-3-iodotyramine hydrochloride and ketanserin were synthesized by Servier chemists. Compounds were dissolved in sterile water plus lactic acid when necessary, then the pH was adjusted to near neutrality. Drugs were injected in a volume of 10 ml/kg, and doses are expressed in terms of free base. All drugs, with the exception of ketanserin (administered subcutaneously), were administered intraperitoneally. In neurochemical and behavioral studies, a full dose range of MDMA was tested (0.63–20.0 mg/kg), corresponding to doses shown to exert robust and specific actions in previous studies of this agent (Green et al., 2003; Colado et al., 2004), and the actions of MDMA were tested over a range of times (20–180 min).
Results
Targeting strategy and generation of the TA1-KO mouse line
A line of homozygous mice lacking TA1 receptors (TA1-KO) was generated in which the coding sequence of Taar1 was replaced by the LacZ reporter gene, resulting in the production of a radically truncated (42 vs 322 aa) and inactive TA1R in vivo (Fig. 1A). PCR of the targeted and WT TA1R gene locus was used to diagnose homologous recombination in embryonic stem cells and in homozygous mutants (Fig. 1B). The mouse line was generated using targeted derived ES cells and was maintained on a pure C57BL/6J genetic background (WT mice). In TA1-KO mice, qRT-PCR analysis revealed an absence of mRNA coding TA1R in structures known to be enriched in TA1Rs (Table 1) (Wolinsky et al., 2007; Lindemann et al., 2008). By contrast, in WT mice, higher levels of mRNA coding TA1R were detected in the substantia nigra/ventral tegmental area, and in nucleus accumbens, striatum, and frontal cortex (Table 1). Conversely, in TA1-KO mice, these and other regions known to express TA1Rs revealed an expression of β-galactosidase (product of the LacZ reporter gene), as observed by histoenzymology (Fig. 1H). Mice were healthy, and ate and grew normally (body weights as adults were 21.6 ± 0.1 and 21.4 ± 0.2 g, respectively, in WT and TA1-KO mice). They showed no gross perturbation of behavior and displayed no difference in basal motor activity (see Fig. 7).
Agonist actions of MDMA at recombinant HeLa-expressed hTA1R
Since trace amine family receptors are notoriously difficult to functionally express in cell systems (Lindemann and Hoener, 2005; Grandy, 2007), we designed a TA1R bearing an asparagine-linked glycosylation site at its N terminus to facilitate plasma membrane expression and pharmacological characterization: in addition, we boosted coupling sensitivity by cotransfecting adenylyl cyclase V, which may also act as a chaperone to promote TA1R cell surface expression (Dupré et al., 2007). Using this system, MDMA was found to concentration-dependently stimulate the production of cAMP with an efficacy similar to that of o-PIT (Scanlan et al., 2004; Hart et al., 2006). However, MDMA was less potent than o-PIT, as indicated by a significant difference in EC50 values (F(1,4) = 29.2, p < 0.01), which were 167.5 ± 38.8 nm for MDMA and 9.5 ± 0.6 nm for o-PIT. In contrast to MDMA, PCA only weakly stimulated cAMP production even at very high concentrations up to 10 μm (Fig. 2).
Influence of MDMA upon core temperature in TA1-KO mice
Under standard conditions of 21°C ambient temperature, basal core temperature did not differ between WT and TA1-KO mice (38.1 ± 0.1°C in each case; F(1,73) = 0.9, NS). Administration of MDMA to WT mice induced a dose-dependent (0.63–20 mg/kg, i.p.) and biphasic thermoregulatory response with a rapidly occurring hypothermia (30 min postinjection; T30) (Fig. 3A) followed by a gradual hyperthermia from 60 (T60) to 120 min (T120) (Fig. 3B,C). In contrast, TA1-KO mice displayed only a long-lasting hyperthermia. The maximal genotype difference was detected at T30, with −1.5 ± 0.2°C and +1.5 ± 0.1°C, respectively, in WT and TA1-KO mice. ANOVA revealed a significant gen × drug interaction (F(5,63) = 16.6, p < 0.01). Differences in genotype were also detected at T60 and T120, at which time point a more pronounced hyperthermia was measured in TA1-KO mice (gen × drug interactions: F(5,63) = 6.0, p < 0.01 and F(5,63) = 2.3, p < 0.05 at T60 and T120, respectively).
Increasing ambient temperature to 27°C (elevated temperature) did not reveal genotype difference in basal core temperature (gen × amb interaction: F(1,48) = 0.7, NS), which was significantly lower compared with standard condition (F(1,50) = 57.3, p < 0.01): 37.2 ± 0.1°C and 37.3 ± 0.2°C, respectively, in WT and TA1-KO mice. The effect of genotype upon the MDMA response at T30 (10 mg/kg) (Fig. 3D) was found to not be influenced by ambient temperature (gen × drug × amb interaction: F(1,44) = 0.2, NS). The effects of MDMA under increased ambient temperature were magnified compared with standard temperature (drug × amb interaction: F(1,48) = 10.2, p < 0.01) and were still significantly different between TA1-KO (+2.8 ± 0.3°C) and WT (+1.2 ± 0.5°C) mice (gen × drug interaction: F(1,48) = 10.9, p < 0.01).
The effect of the 5-HT2A antagonist ketanserin (0.63 mg/kg, s.c.) on the hyperthermia elicited by MDMA (20 mg/kg, i.p.) was evaluated under standard conditions at 90 min after injection (data not shown). Ketanserin significantly reduced MDMA-induced hyperthermia (pret × drug interaction: F(1,58) = 5.1, p < 0.05), and its action was similar in both genotypes as indicated by a nonsignificant gen × pret × drug interaction (F(1,58) = 3.7, NS), lowering the MDMA-induced hyperthermia from +0.5 ± 0.1°C to −0.7 ± 0.4°C and from +1.3 ± 0.1°C to +0.5 ± 0.1°C, respectively, in WT and TA1-KO mice. Ketanserin, by itself, slightly decreased core temperature (−0.2 ± 0.2°C in WT mice; −0.6 ± 0.2°C in TA1-KO mice) when compared with vehicle (0.0 ± 0.1°C change in both cases). By contrast with ketanserin, core temperature was unaffected by antagonists at D1, D2, α2, and 5-HT1A sites SCH23390, raclopride, RX821002, and WAY100635, respectively (data not shown).
Dialysis levels of exogenous MDMA in the dorsal striatum of TA1-KO mice
A dedicated set of microdialysis experiments was performed to verify the exposure to MDMA in the dorsal striatum of WT and TA1-KO mice. To this end, dialysate levels of (exogenous) MDMA were determined by mass spectrometry following its intraperitoneal administration (10 mg/kg). Exposure to MDMA was similar as determined by both (1) AUC (165 ± 19 in WT mice and 183 ± 33 in TA1-KO mice) and (2) maximal concentrations (2085 ± 378 ng/ml in WT mice and 1997 ± 252 ng/ml TA1-KO mice). No statistical difference was observed throughout the 180 min of observation, and ANOVA indicated a nonsignificant gen × time interaction (F(8,72) = 0.5, NS). Hence, differences in the functional response to MDMA between TA1-KO and WT mice cannot be attributed to an alteration in its penetration of the blood–brain barrier and differential accumulation of MDMA in the brain.
Effect of MDMA on dialysis levels of DA in the dorsal striatum and nucleus accumbens of TA1-KO mice
Under basal conditions, extracellular levels of DA in dorsal striatum and nucleus accumbens of freely moving mice did not significantly differ between WT and TA1-KO mice (Table 2).
In the dorsal striatum, MDMA (10 mg/kg, i.p.) elicited a sustained and robust increase in DA levels (Fig. 4A). Its maximal magnitude was significantly higher in TA1-KO mice compared with WT mice (1756 ± 361 and 909 ± 133% of baseline, respectively) and was revealed by a significant gen × drug × time interaction (F(11,198) = 7.8, p < 0.01). The overall influence of genotype upon the effect of MDMA was dependent upon the dose of MDMA (5–10 mg/kg) (Fig. 4B), as indicated by a significant gen × drug interaction (F(3,38) = 5.0, p < 0.01).
In the nucleus accumbens, MDMA (10 mg/kg) (Fig. 4C) also provoked a pronounced increase in DA levels, which reached a higher maximal value in TA1-KO mice than in WT mice (1316 ± 145 and 630 ± 73% of baseline). The hypersensitivity of TA1-KO mice was revealed by a significant gen × drug × time interaction (F(11,220) = 16.8, p < 0.01). The overall effect of MDMA was also significant with a gen × drug interaction (F(1,20) = 13.0, p < 0.01) (Fig. 4C).
Influence of o-PIT upon PCA-induced release of DA in the dorsal striatum of TA1-KO mice
Administration of PCA (5 mg/kg, i.p.) elevated DA levels in the dorsal striatum with a similar magnitude in WT and TA1-KO mice (571 ± 63 and 655 ± 50%, respectively). Pretreatment with the TA1 agonist, o-PIT (10 mg/kg, i.p.), which did not affect DA levels by itself, significantly blunted the action of PCA in WT mice, lowering its maximal effect to 301 ± 81% (Fig. 5A) and was, however, inactive in TA1-KO mice (Fig. 5B). ANOVA indicated a significant pret × drug × time interaction in WT mice (F(11,187) = 3.6, p < 0.01) but not TA1-KO (F(11,176) = 2.1, NS) mice. The influence of the genotype was revealed by the analysis of the AUC values and the significant gen × pret × drug interaction (F(1,34) = 6.1, p < 0.05) (Fig. 5C).
The influence of o-PIT upon PCA-induced [3H]-DA overflow was also evaluated in striatal and accumbal synaptosomes from TA1-KO mice. Basal release of [3H]-DA was not influenced by genotype and was 15.4 ± 0.6 and 15.8 ± 1.7% of total release, respectively, in WT and TA1-KO mice. Perfusion of o-PIT (100 nm) did not affect the basal release of [3H]-DA (Fig. 5D), whereas PCA (1 μm) significantly increased [3H]-DA overflow in preparation from WT and TA1-KO mice (33.9 ± 0.9 and 34.3 ± 1.2%, respectively). In WT mice, the effect of PCA upon [3H]-DA was significantly reduced by coperfusion of o-PIT (gen × drug interaction: F(1,6) = 18.1, p < 0.01), lowering its overflow to 27.9 ± 1.2% of total release. In contrast, o-PIT was inactive in preparations from TA1-KO mice (34.1 ± 1.4%).
Levels of p-Ser-TH and TH activity in dorsal striatum of TA1-KO mice, and influence of MDMA upon p-Ser40-TH
The basal phosphorylation state of TH at Ser19, Ser31, and Ser40 residues was significantly (p < 0.05) increased in striatal slices from TA1-KO mice when compared with WT mice (Fig. 6A). No corresponding difference was found in total levels of TH, which correlates with the absence of significant difference in the expression of mRNA coding TH (Table 1). Basal TH activity, measured by DOPA accumulation in striatal slices, was significantly higher in TA1-KO mice when compared with WT mice (p < 0.05) (Fig. 6B). In response to MDMA, levels of p-Ser40-TH were decreased in TA1-KO mice, but not in WT mice (Fig. 6C). ANOVA indicated a significant gen × drug interaction (F(1,32) = 6.6, p < 0.05).
By analogy to TH, as depicted in Table 1, there were no significant differences between WT and TA1-KO mice as regards the levels of mRNA encoding various substrates controlling the activity of dopaminergic pathways, including catabolic enzymes, DA transporters, and both D2/D3 autoreceptors. Hence, putative changes in these mechanisms cannot account for alterations in the influence of MDMA upon extracellular levels of DA in TA1-KO versus WT mice.
Influence of MDMA in interaction with haloperidol upon locomotor activity in TA1-KO mice
Spontaneous locomotor activity was not significantly different between WT and TA1-KO mice (392 ± 20 and 406 ± 19 counts, respectively). MDMA (10 mg/kg, i.p.) increased locomotion more robustly in TA1-KO than WT mice (gen × drug × time interaction: F(20,520) = 2.2, p < 0.01) (Fig. 7A). The locomotor-activating action of MDMA was blocked by haloperidol (Fig. 7B). This interaction was not influenced by genotype (gen × pret × drug interaction: F(3,98) = 1.4, NS). In both genotypes, the effect of MDMA was dose-dependently prevented by haloperidol (0.01–0.16 mg/kg, i.p.; pret × drug interaction: F(3,49) = 5.1, p < 0.01 and F(3,49) = 9.1, p < 0.01, respectively, in WT and TA1-KO mice), and was totally abolished by 0.16 mg/kg haloperidol. Furthermore, haloperidol was more potent in reducing the effect of MDMA in TA1-KO than in WT mice (minimal effective doses were 0.01 and 0.16 mg/kg, respectively).
Effect of MDMA upon DA transmission in the frontal cortex of TA1-KO mice
Basal extracellular levels of DA in the frontal cortex of TA1-KO mice were not significantly different from WT mice (Table 2). In the frontal cortex, MDMA (10 mg/kg, i.p.) increased levels of DA (Fig. 8A), an effect that was amplified in TA1-KO compared with WT mice (3280 ± 650 and 2096 ± 425% of baseline, respectively; gen × drug × time interaction: F(11,209) = 3.4, p < 0.05).
In synaptosomal preparations from frontal cortex, basal release of cortical [3H]-DA was not statistically influenced by genotype and was 15.4 ± 0.7 and 17.8 ± 1.0% of total release, respectively, in WT and TA1-KO mice. The release of [3H]-DA was not altered by the perfusion of o-PIT (100 nm), but was significantly increased by PCA (1 μm) in WT and TA1-KO mice (27.1 ± 1.1 and 24.8 ± 1.2% of total release, respectively) (Fig. 8B). Coperfusion of o-PIT with PCA significantly reduced the effect of PCA to 21.0 ± 1.1% of total release in synaptosomes from WT mice (gen × drug interaction: F(1,8) = 5.8, p < 0.05) but not TA1-KO mice (24.9 ± 1.4%).
Effect of MDMA on 5-HT transmission in TA1-KO mice
Extracellular levels of 5-HT in the dorsal striatum, nucleus accumbens, or frontal cortex of freely moving mice did not significantly differ between WT and TA1-KO mice (Table 2).
In the dorsal striatum, MDMA (10 mg/kg, i.p.) elicited a sustained and robust increase in 5-HT levels (Fig. 9A). Its maximal magnitude was significantly higher in TA1-KO mice compared with WT mice (2139 ± 175 and 938 ± 174% of baseline, respectively), as revealed by a significant gen × drug × time interaction (F(11,198) = 12.2, p < 0.01). The overall influence of genotype upon the effect of MDMA was dependent upon the dose of MDMA (5–10 mg/kg) (Fig. 9B), as indicated by a significant gen × drug interaction (F(3,37) = 10.9, p < 0.01).
In the nucleus accumbens, MDMA (10 mg/kg) (Fig. 9C) increased 5-HT levels to a greater extent in TA1-KO than in WT mice (2937 ± 389 and 1931 ± 163% of baseline). ANOVA indicated a significant gen × drug × time interaction (F(11,220) = 4.2, p < 0.01).
In frontal cortex, MDMA (10 mg/kg) increased levels of 5-HT (Fig. 9D), an effect that was not significantly influenced by genotype (2074 ± 130 and 1797 ± 355% of baseline, respectively, in WT and TA1-KO mice; F(11,220) = 0.2, NS).
Influence of o-PIT upon PCA-induced release of 5-HT in the dorsal striatum of TA1-KO mice
Administration of PCA (5 mg/kg, i.p.) elevated 5-HT levels in the dorsal striatum with a comparable magnitude in WT and TA1-KO mice (1180 ± 105 and 1250 ± 111%, respectively). Pretreatment with the TA1 agonist o-PIT (10 mg/kg, i.p.), which did not affect 5-HT levels by itself, significantly blunted the action of PCA in WT mice—lowering its maximal effect to 663 ± 100% (Fig. 10A)—but not in TA1-KO mice (Fig. 10B). ANOVA indicated a significant pret × drug × time interaction in WT mice (F(11,187) = 11.9, p < 0.01), but not in TA1-KO mice (F(11,176) = 2.0, NS). The influence of the genotype was revealed by analysis of the AUC values and the significant gen × pret × drug interaction (F(1,34) = 3.4, p < 0.05) (Fig. 10C).
In synaptosomal preparation from dorsal striatum, basal release of [3H]-5-HT was not influenced by genotype and was 7.0 ± 0.6 and 8.7 ± 0.8% of total release, respectively, in WT and TA1-KO mice. Perfusion of o-PIT (100 nm) did not affect the release of [3H]-5-HT (Fig. 10D), whereas PCA (1 μm) significantly increased [3H]-5-HT overflow in preparations from WT and TA1-KO mice (27.2 ± 1.3% and 27.7 ± 2.2%, respectively). In WT mice, the effect of PCA upon [3H]5-HT was significantly reduced by coperfusion of o-PIT (gen × drug interaction: F(1,6) = 6.8, p < 0.05), lowering its overflow to 22.6 ± 1.2% of total release. In contrast, o-PIT was inactive in TA1-KO mice (26.9 ± 1.4%).
Discussion
Agonist properties of MDMA at hTA1R
The prototypical agonist, o-PIT (Scanlan et al., 2004; Barak et al., 2008) stimulated TA1Rs, and less potent but equi-effective activation of TA1R was seen with MDMA. Like o-PIT, MDMA also binds murine TA1Rs (Bunzow et al., 2001; Hu et al., 2009). In contrast to MDMA, PCA was essentially inactive at hTA1R. The interaction of o-PIT with TA1R (Scanlan et al., 2004) was specific in that it showed ∼100-fold lower potency (M. J. Millan, unpublished observation) in G-protein coupling studies of sites controlling monoaminergic transmission, core temperature, and locomotor activity: notably, dopamine D2/D3, 5-HT1A, and α2A-adrenergic receptors, which act as autoreceptors on dopaminergic, serotonergic, and adrenergic pathways, respectively (Millan et al., 2000a).
Reinforcement of the hyperthermic actions of MDMA in TA1-KO mice
MDMA exerts a complex influence upon core temperature (Green et al., 2003; Rodsiri et al., 2011) and elicited a biphasic time- and dose-dependent hypothermia-hyperthermia in WT mice, which was transformed into a monophasic, rapid, and dose-dependent hyperthermia in TA1-KO counterparts. Increasing ambient temperature favors hyperthermia (Green et al., 2003; Feduccia et al., 2010) and exaggerated the hyperthermic response to MDMA in TA1-KO mice, consistent with the hypothermic impact of TA1R activation (Scanlan et al., 2004; Hart et al., 2006). These observations amplify the suggestion of Panas et al. (2010) that recruitment of TA1R by MDMA restrains its hyperthermic properties. The hypothermia-mediating populations of TA1R that modulate the influence of MDMA on core temperature are probably localized in thermoregulatory centers like medial optic area and arcuate hypothalamus (Borowsky et al., 2001; Hargreaves et al., 2007; Bratincsák et al., 2008). Alternatively, they might be situated in monoaminergic neurons themselves (Mechan et al., 2002; Herin et al., 2005; Bexis and Docherty, 2008). Nonetheless, D2/D3, 5-HT1A, and α2-adrenergic autoreceptors are not involved in MDMA-induced hypothermia since antagonists did not modify its influence on core temperature (see Results) (Fig. 3). Conversely, the 5-HT2A receptor blocker ketanserin blunted the hyperthermic effects of MDMA both in TA1-KO and WT mice (Liechti et al., 2000; Docherty and Green, 2010), consistent with clinical studies of the hyperthermia provoked by MDMA in humans (Vollenweider et al., 2002; Docherty and Green, 2010).
Amplification of MDMA-induced increases in dialysis levels of DA in TA1-KO mice
TA1-KO mice are modestly hypersensitive to increases in extracellular levels of DA provoked by amphetamine in dorsal striatum (Wolinsky et al., 2007; Lindemann et al., 2008). The present data expand these studies in demonstrating that genetic deletion of TA1R markedly enhances MDMA-induced extracellular DA not only in dorsal striatum but also nucleus accumbens and frontal cortex. This robust potentiation reflects the potent and high-efficacy actions of MDMA at TA1R, compared with the weak partial agonist effects of amphetamine (Borowsky et al., 2001; Barak et al., 2008; Lindemann et al., 2008). The phasic role of TA1R in auto-inhibiting actions of MDMA at striatal and frontocortical dopaminergic terminals was further exemplified using o-PIT (Scanlan et al., 2004; Hart et al., 2006), which blunted elevations in dialysis levels of DA elicited by PCA in WT but not TA1-KO mice. This action of o-PIT was reproduced in striatal-accumbal and frontocortical synaptosomes, demonstrating that it occurs at dopaminergic terminals. This demonstration that TA1R modulate mesolimbic and frontocortical as well as nigrostriatal dopaminergic projections is important in view of their distinctive functional roles (Hauber, 2010). The precise mechanisms involved require additional study, but activation of TA1R probably interferes with the operation of dopamine transporters (DATs) and/or D2/D3 autoreceptors both at the level of perikarya (Table 1; Lindemann et al., 2008; Bradaia et al., 2009; Ledonne et al., 2010) and terminals (Xie and Miller, 2007, 2008; Xie et al., 2008).
Amplification of the influence of MDMA upon locomotor activity
Mirroring data from dialysis, induction of locomotor activity by MDMA was more pronounced in TA1-KO versus WT mice, and its actions were dose-dependently abrogated by haloperidol (Ball et al., 2003; Jaworski et al., 2003). Similarly, an enhanced locomotor response to amphetamine was seen in mice genetically deprived of TA1R (Wolinsky et al., 2007; Lindemann et al., 2008). Future studies should examine MDMA-responsive procedures monitoring reward mechanisms and drug-seeking behavior, such as place preference and self-administration. It would also be interesting to examine the influence of MDMA on prepulse inhibition (PPI) since MDMA disrupts sensorimotor gating in mice and humans (Vollenweider et al., 1999; Braff et al., 2001; Liechti et al., 2000), and PPI is blunted in TA1-KO versus WT mice (Wolinsky et al., 2007).
Alterations in MDMA-induced striatal TH phosphorylation in TA1-KO mice
TH phosphorylation at Ser19, Ser31, and Ser40, was increased in TA1-KO mice. These sites are regulated by distinct mechanisms: increases in intracellular Ca2+ stimulate phosphorylation of Ser19; extracellular signal-regulated protein kinases promote phosphorylation at Ser31; and phosphorylation of Ser40 is catalyzed by protein kinase A (Haycock, 1993). Since the phosphorylation state of all three sites is increased by stimulation of dopaminergic fibers (Haycock and Haycock, 1991), it is interesting that the firing rate of dopaminergic neurons is accelerated in TA1-KO versus WT littermates (Lindemann et al., 2008), and (in vitro) in WT mice treated with the TA1R antagonist EPPTB (Bradaia et al., 2009). Correspondingly, increased basal phosphorylation of TH in TA1-KO mice may reflect elevated activity of dopaminergic neurons. Furthermore, enhanced phosphorylation of TH stimulates enzymatic activity (Haycock and Haycock, 1991), and increased TH activity was also seen in TA1-KO mice.
Intriguingly, MDMA did not affect TH phosphorylation at Ser40 in WT versus TA1-KO mice. DA, acting via inhibitory D2 autoreceptors on dopaminergic terminals, inhibits Ser40-TH phosphorylation (Haycock and Haycock, 1991; Lindgren et al., 2001), and an increased proportion of striatal high-affinity D2 receptors was seen in TA1-KO versus WT mice (Wolinsky et al., 2007). Likewise, quinpirole more potently stimulated D2 receptor-mediated functional responses in TA1-KO mice (Bradaia et al., 2009). Hence, MDMA-released DA may more potently trigger D2 autoreceptor-mediated inhibition of TH phosphorylation in TA1-KO compared with WT mice. Furthermore, o-PIT increases phosphorylation at Ser40-TH independently of dopaminergic transmission (P. Svenningsson, unpublished observations), so loss of this TA1R-mediated stimulatory influence of MDMA on TH phosphorylation likely contributes to lower TH phosphorylation following MDMA treatment of TA1-KO versus WT mice.
Amplification of MDMA-induced increases in dialysis 5-HT of TA1-KO mice
Despite localization of TA1R in raphe nuclei (Wolinsky et al., 2007), their potential influence on serotonergic pathways remains undocumented. Suggesting a phasic, inhibitory influence of TA1R upon serotonergic transmission and the effects of MDMA, its elevation of extracellular 5-HT was magnified in dorsal striatum and nucleus accumbens of TA1-KO mice, whereas the influence of PCA was unaffected. Together with the observation that o-PIT reduced PCA-elicited increases in extracellular 5-HT in WT but not TA1-KO mice, these observations suggest that intrinsic agonist actions of MDMA at TA1R blunt its influence upon serotonergic transmission. Moreover, a direct influence of TA1R on serotonergic nerve endings is probably involved since o-PIT attenuated PCA-elicited increases in superfusate concentrations of 5-HT in synaptosomal preparations derived from WT but not TA1-KO mice. By analogy to interactions between TA1R and DATs in dopaminergic neurons (Sotnikova et al., 2008; Xie and Miller, 2007), TA1Rs may alter extracellular 5-HT levels through functional interactions with serotonin transporters (SERTs): activation of TA1R by MDMA may decrease the operation by SERTs via kinase-mediated phosphorylation (Ramamoorthy et al., 2007; Millan et al., 2008). Alternatively, mimicking interactions between TA1R and D2/D3 autoreceptors (see Introduction), TA1R may interact with 5-HT1B autoreceptors (Millan et al., 2000b) and/or their protein partner, p11, to inhibit liberation of 5-HT (Svenningsson et al., 2006; Millan et al., 2008).
The frontal cortex and dorsal striatum are predominantly innervated by the raphe nuclei, and PCA-induced elevations in 5-HT levels were blunted by o-PIT in frontocortical and striatal synaptosomes of WT but not TA1-KO mice. Surprisingly, then, there was no phenotypic variation as regards the influence of MDMA upon 5-HT levels in dialysates of frontal cortex. Such a regional difference in the impact of MDMA on serotonergic transmission resembles observations with neurokinin1 receptor antagonists (Guiard et al., 2007; Gobert et al., 2009).
MDMA-induced increases in dialysis levels of NA in TA1-KO mice
Finally, by analogy to striatal levels of amphetamine (Lindemann et al., 2008), MDMA-elicited increases in NA levels were slightly amplified in frontal cortex and nucleus accumbens of TA1-KO mice. In frontal cortex, NA reuptake is controlled by DATs (Millan et al., 2000b), so a mechanism similar to that modulating levels of DA may be involved. As regards nucleus accumbens, NA is primarily derived from adrenergic neurons (Gobert et al., 2004), suggesting that NA overflow may be under the control of TA1R. However, mRNA encoding TA1R has not yet been observed in the locus ceruleus, so these observations await mechanistic clarification.
Pathophysiological relevance and concluding comments
The present studies reveal that TA1Rs exert a widespread, phasic, and inhibitory influence upon cerebral monoaminergic transmission expressed not only in dorsal striatum but also in nucleus accumbens and frontal cortex, and not only against DA but also 5-HT and, albeit weakly, NA. In contrast to Gi-coupled monoaminergic autoreceptors (Millan et al., 2000a), TA1Rs couple positively via Gαi to adenylyl cyclase. Hence, cellular cascades intervening in its influence upon monoamine release likely differ to those engaged by classical G-protein-coupled receptors (Xie and Miller, 2007; Xie et al., 2008).
The present data also demonstrate that MDMA auto-restrains its influence upon extracellular levels of monoamines by recruiting colocalized TA1R. Since the present work was undertaken in mice with constitutively deleted TA1Rs, it would be interesting to perform comparable studies of adult-onset and/or regionally specific deletion of TA1R. In addition, because young adults frequently take multiple doses of MDMA over long periods, it would be interesting to examine the effects of chronic exposure to MDMA in TA1-KO mice.
Finally, from a clinical perspective, the present data raise the possibility that genetic factors, epigenetic programming, and/or environmental modulation of TA1R gene expression may modify individual risk upon exposure to MDMA.
Footnotes
The authors declare no competing financial interests.
- Correspondence should be addressed to Mark J. Millan, Institut de Recherche Servier, Department of Neurobiology, 125, Chemin de Ronde, 78290 Croissy-sur-Seine, France. mark.millan{at}fr.netgrs.com