Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The mammalian monoamine transporters for dopamine (DA), norepinephrine (NE) and serotonin (5HT) are the sites of action for two major classes of psychoactive drugs: psychomotor stimulants and antidepressant medications. Psychostimulants like cocaine and amphetamines are potent inhibitors of all three monoamine transporters; however, most of their reinforcing properties and abuse potential have been attributed to the blockade of the dopamine transporter (DAT) (Ritz et al., 1987). Antidepressant medications, including the selective serotonin reuptake inhibitors and the tricyclic antidepressants, predominantly block the serotonin and norepinephrine transporters (SERT and NET) (Barker and Blakely, 1995), but have decidedly lower potencies for the mammalian DATs.

Functional disruption of the monoamine transporter genes in mice result in profound behavioral (Giros et al., 1996) and neurochemical (Bengel et al., 1998; Xu et al., 2000) changes and demonstrate the crucial role of the transporters in maintaining low extracellular amine concentrations in vivo. Progress in our understanding of changes induced by cocaine and antidepressant drugs may also come from studies of simpler organisms. The fruit fly, Drosophila melanogaster, has emerged as a promising model organism to study the molecular genetics of responsiveness to volatilized free-base cocaine. D. melanogaster responds to cocaine exposure with many of the same stereotypic motor behaviors seen in vertebrates and also displays behavioral sensitization, as reflected in an increase in the locomotor response that develops after repeated cocaine exposures (McClung and Hirsh, 1998). DA and/or 5-HT mediate the locomotor responses to cocaine (Bainton et al., 2000; Li et al. 2000). The receptor pharmacology underlying the behavioral response to psychostimulants seems to be mediated through a quinpirole-sensitive, D2-like dopamine receptor at least in the nerve cord, but the molecular identity of the receptor remains unknown (Yellman et al., 1997). Additional studies have shown that the trace amine tyramine (TA) is required for cocaine sensitization (McClung and Hirsh, 1999) but does not stimulate nerve cord locomotor responses by itself. However, it is not clear whether TA in flies acts on postsynaptic receptors or on presynaptic transporters to produce a sensitized response. TA can activate 2-like adrenergic receptors in D. melanogaster (Saudou et al., 1990; Kutsukake et al., 2000), but in mammals, TA acts as an indirect sympathomimetic in that it stimulates carrier-mediated efflux of neurotransmitters by an exchange mechanism (Bönisch and Trendelenburg, 1988).

By analogy to the carriers identified in vertebrates, we reasoned that the D. melanogaster monoamine transporters would belong to the Na⁺/Cl-dependent symporter family (Amara and Kuhar, 1993) and would also be sensitive to cocaine. Moreover, a D. melanogaster cocaine-sensitive serotonin transporter, dSERT, has been identified (Corey et al., 1994; Demchyshyn et al., 1994), but the restricted expression of dSERT in serotonergic neurons implies the existence of additional high-affinity transporters to efficiently clear other monoamines such as DA, OA, and TA. All monoamine transporters cloned thus far contain a characteristic aspartate-residue in the first transmembrane domain (TMD1), which is critical for substrate transport activity (Kitayama et al., 1992). Using these criteria, we identified a putative monoamine carrier in the fly genome that displays ~65% sequence similarity with the mammalian DATs and NETs and a dopamine transporter from Caenorhabditis elegans (ceDAT; Jayanthi et al., 1998).

Here we report the identification, anatomical localization, and functional characterization of the D. melanogaster dopamine transporter, dDAT. Transport by dDAT has a pharmacological profile distinct from the DATs of mammalian species: although it retains the substrate selectivity of a dopamine carrier, dDAT is potently blocked by antidepressants and displays an inhibitory profile more similar to the mammalian NETs. The hybrid molecular features observed in dDAT and the phylogenetic analysis of vertebrate and invertebrate NTTs suggest that dDAT, along with ceDAT, represents a primordial monoamine transporter gene that existed before the emergence of the mammalian catecholamine carrier subtypes, DAT and NET. Our studies further demonstrate that dDAT is a cocaine target in D. melanogaster and that its functional interaction with TA in vitro could potentially link the essential role of TA in the cocaine-induced sensitization phenomenon with the modulation of dopaminergic signaling.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. All tritiated compounds used in this study were from NEN Life Science Products (Boston, MA), except [³H]DA, which was purchased from Amersham Pharmacia Biotech (Piscataway, NJ). CitalopramHBr was a kind gift from Lundbeck A/S (Copenhagen, Denmark); amphetamine isomers and RTI-55 were obtained from the NIDA Drug Supply Program. Other chemicals and drugs were purchased from Sigma-Aldrich-RBI (St. Louis, MO).

cDNA Cloning. We used the sequence information of the 1st transmembrane domain (TMD1) of hDAT to search for related sequences in the Berkeley Drosophila Genome Project database (http://www.fruitfly.org) and identified a 526-bp EST, GH22929. Translation of GH22929 revealed an essential aspartate residue (D79 in hDAT), suggesting that this EST belonged to a monoamine transporter cDNA. This EST was also part of the genomic clone AC005647, located on chromosome 2R, region 53C7-C14. Detailed analysis of the genomic sequence enabled us to identify the start and stop codons of an open reading frame of a potential monoamine transporter. We designed two oligonucleotides (5' cgc ggt acc g tcg cag atg tca cca acc gga c 3' and 5' ccg tct aga tca gac atc gac ggg ttc ctt ggc g 3') flanked by consensus sites for Asp718 and XbaI, respectively. A cDNA pool was generated by reverse transcription (RT-AMV, Roche Biochemicals, Nutley, NJ) of random hexamer primed total RNA isolated from D. melanogaster heads (Canton-S, wild type strain, a kind gift of Mike Forte, Vollum Institute) using the TRIzol reagent from Life Technologies (Gaithersburg, MD). The oligonucleotide pair was used to amplify a single 1.9 kb PCR fragment from this D. melanogaster head cDNA pool. The cycling program was 94°C for 3 min, 15 cycles touch down at 94°C for 45 s, 65°C (1°C/cycle) for 30 s, 72°C for 3 min, and 25 cycles 94°C for 45 s, 55°C for 15 s, 72°C for 3 min (TaqDNA polymerase, Roche Biochemicals).

In Situ Hybridization. Distribution of dDAT mRNA in D. melanogaster third instar larval brains (Oregon-R, wild-type strain) was analyzed by in situ hybridization as described in Lehmann and Tautz (1994), with modifications. Briefly, digoxigenin-labeled antisense and sense single-stranded DNA probes were prepared by two independent single primer polymerase chain reactions against 200 ng of linearized dDAT cDNA using the primers, 5'-GCC CGT AAA CCG TGA TGA AGA GCA G-3' for antisense and 5'-CTG GGT CAG CAC AAT CGT AAG GGT G-3' for sense probes. Hand-dissected larval brains were fixed with 4% paraformaldehyde, treated briefly with proteinase K, and incubated with digoxigenin-labeled antisense and sense probes in parallel. After extensive washing, digoxigenated probes were detected using alkaline phosphatase-conjugated antidigoxigenin antibody (Roche Biochemicals) and subsequent standard nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate color reaction.

Expression Profile Analysis. PCR-based Drosophila Rapid-Scan Gene Expression Panel (OriGene Technologies, Rockville, MD) was used to examine the level of dDAT mRNA in different developmental stages as well as in male and female head and body following the manufacturer's instruction. Using the primers (forward, 5'-CCGTC GATTTA ACAAA CGTCT GG-3'; reverse, 5'-ACTGC ATAGGG AAAGA GGGCA G-3') for dDAT, PCRs were carried out with AmpliTaq Gold (Perkin Elmer, Norwalk, CT) through 25 cycles of 94°C/30 s, 55°C/30 s and 72°C/1 min. The ethidium bromide-stained agarose gel was imaged on an Image Station 440CF (Kodak Digital Science, Rochester, NY) and the net intensity of bands were analyzed by 1D Image Analysis Software (Kodak Digital Science).

Tissue Cultures and Transfection Protocols. COS-7 cells (SV40-transformed African green monkey kidney cells; American Type Culture Collection, Manassas, VA), and MDCK cells (Madin-Darby canine kidney cells) were grown at 37°C in a 5% CO₂, humidified atmosphere in Dulbecco's modified Eagle's medium (Life Technologies) supplemented with 10% fetal calf serum, 10 U/ml penicillin, and 10 µM/ml streptomycin. The medium for stable-transfected MDCK cell lines was supplemented with 0.5 mg/ml G418 (Life Technologies).

Uptake Experiments. Uptake assays were performed at room temperature in Krebs-Ringer's-HEPES (KRH) buffer (125 mM NaCl, 4.8 mM KCl, 1.2 mM MgSO₄, 1.2 mM KH₂PO₄, 1.3 mM CaCl₂, and 25 mM HEPES, pH 7.4), supplemented with 0.1% D-glucose, 1 mM ascorbic acid, 1 mM tropolone [catechol-O-methyltransferase (EC 2.1.1.6)-inhibitor] and 10 µM pargyline (monoamine oxidase-B inhibitor). Before the assay, cells were washed once with KRH and equilibrated for 5 min. COS-7 cells were assayed in 24-well plates and incubated for 2 min with tritiated amines, whereas MDCK cells were incubated for 6 min in 48-well plates. Nontransported inhibitors were preincubated for 5 min, and substrates were applied together with the tritiated substrate. The uptake assay was terminated with two washes of ice-cold KRH, and the accumulated radioactivity was recovered by lysing the cells in 0.2% SDS and 0.1 N NaOH and counting on a Liquid Scintillation Analyzer 1900 TR (Packard, Meriden, CT). Nonspecific uptake was determined in the presence of 20 µM nisoxetine (for dDAT and hNET), 10 µM GBR12909 (for hDAT), 20 µM fluoxetine (for hSERT), or 20 µM mazindol (for dSERT and dDAT).

In Vitro Transcription, Oocyte Injection and Voltage Clamp Recording. Xenopus laevis oocytes were prepared as described in Sonders et al. (1997) and maintained at 17° or 21°C. cRNAs were transcribed and capped in vitro (mMessage mMachine, Ambion, Austin, TX) from pOTV plasmid containing the coding sequence of dDAT, and oocytes were tested 1 to 7 days after cRNA injection. Uptake and electrophysiological experiments were performed in a frog Ringer's buffer (100 mM NaCl, 2 mM KCl, 1 mM MgCl₂, 7.5 mM HEPES/3 mM Tris, pH 7.4) at room temperature as described previously (Sonders et al., 1997). Nonspecific uptake was defined using water-injected oocytes.

Efflux Experiments. Efflux assays were performed with dDAT- and dSERT-expressing COS-7 cells 2 days after transfection. The assays were done at room temperature with the same KRH-buffer that was used for uptake assays, except that it contained 2 µM Ro 41-0960 to inhibit catechol-O-methyltransferase-activity, instead of tropolone. Cells were washed once with KRH-buffer and then preloaded for 10 min with 20 to 50 nM [³H]amine. The preloading buffer was aspirated and cells were washed for 5 min in KRH-buffer. Efflux was initiated by replacing the wash buffer with KRH-buffer containing the cold test compounds. After 10 min, the supernatant was collected for scintillation counting (bath counts = efflux), cells were immediately washed twice with ice-cold KRH-buffer and subsequently lysed with 0.2% SDS, 0.1N NaOH, and the remaining radioactivity was counted (cell counts). Efflux (bath counts) was expressed as percent of the total recovered tritium (bath counts + cell counts) for each condition done in duplicates. Spontaneous (nonspecific) release of tritium was determined in parallel by incubating preloaded cells with KRH-buffer alone (vehicle) and was slightly lower than efflux measured in the presence of saturating concentrations of the high-affinity uptake inhibitor, mazindol (Fig. 7). The slight stimulating effect of mazindol reflects inhibition of transporter-mediated clearance of spontaneous released [³H]amine. Time-course experiments showed that DA-stimulated efflux reaches a plateau level after 5 to 10 min, whereas spontaneous efflux hardly changes after 5 min (data not shown). Therefore, an efflux time of 10 min was chosen to obtain robust efflux over vehicle control levels, even for poor substrates.

Data Analysis. The results of liquid scintillation counting were used to calculate the [³H]amine uptake into cells, expressed as femtomoles per milligram per well. In each experiment, the mean result from triplicate wells for each treatment was used. Specific uptake was calculated as the difference between uptake of [³H]amine in the absence (total uptake) and presence (nonspecific uptake) of 10 to 20 µM high-affinity transport inhibitor for each plate. K_T and IC₅₀ values were calculated from nonlinear regression analysis of the data for each individual experiment according to Michaelis-Menten kinetics. All uptake data were analyzed using GraphPad Prism 2c software (GraphPad Software, San Diego, CA).

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Identification of a D. melanogaster Monoamine Transporter cDNA. Using the sequence information of the first transmembrane domain (TMD1) of the human dopamine transporter (hDAT), we identified a D. melanogaster gene for a potential monoamine transporter on chromosome 2R in the Berkeley fly genome database (accession no. AC005647) (Fig. 1). PCR primers complementary to the predicted first and last exons of the open reading frame were used to amplify a single 1.9-kb DNA fragment from a D. melanogaster head cDNA pool. The isolated cDNA encoded a polypeptide of 631 amino acids with highest homology to the human monoamine transporters NET, DAT, and SERT and the C. elegans DAT (52, 49, 45, and 51% identity, respectively; Fig. 2), and was classified as the D. melanogaster dopamine transporter (dDAT), because of its restricted expression in dopaminergic cells and the functional properties outlined below. Sequence and hydrophobicity analyses of the carrier protein predicted 12 TMDs, intracellularly located N- and C-termini, and a large extracellular loop (EL2) between TMD3 and TMD4 (Fig. 1B). The highest degree of similarity to the mammalian catecholamine carriers is found in the 12 TMDs (~80% to both hNET and hDAT), whereas the amino- and carboxyl-terminal tails and the large EL2 show significantly less homology (50%).

View larger version (32K): [in this window] [in a new window]   Fig. 1.   The dDAT gene. A, exon-intron structure of the dDAT gene. Eight exons are shown as boxes, introns are the lines between the boxes, drawn approximately to scale. The first 5'-untranslated exon is indicated as gray box. Coding exons are shown as alternating white and black boxes that correspond to domains in the dDAT protein shown below in B. B, transmembrane topology of dDAT as predicted from hydrophobicity analysis. Domains that are encoded by sequential exons are drawn alternating white and black. Numbers at the exon borders indicate the respective amino acid residues. Location of intracellular consensus-sites for protein-phosphorylation are shown: T12, T14, S259, and S581 (*, R/KXXS/T for cAMP-dependent kinases); T14, S31, S261, T500, S585, and T624 (, S/TXR/K for PK-C); and S24, S434, S504, and T611 (+, S/TXE/D for casein kinase II). Y, potential N-glycosylation sites.

View larger version (117K): [in this window] [in a new window]   Fig. 2.   Amino acid alignment of monoamine transporters. Alignment was carried out with CLUSTAL W using MacVector software. Conserved and similar residues (present in 4 transporters) are shaded. Putative TMDs as predicted by hydrophobicity analysis are indicated by the black bars drawn above the sequences. The accession numbers of the aligned sequences are: ceDAT (AF79899), dDAT (AF260833), dSERT (U04809), fET (U72877), hDAT (NP_001035.1), hNET (NP_001034.1), and hSERT (NP_001036.1).

The Organization of the dDAT Gene. Comparison of the isolated cDNA fragments with the genomic sequences indicated that the dDAT gene is composed of eight exons spanning 5.9 kb on chromosome 2R (Fig. 1A). In contrast, the human DAT and NET genes are composed of 14 coding exons (Kawarai et al., 1997; Pörzgen et al., 1995). The locations of introns within the D. melanogaster and human monoamine transporter genes are conserved across species with the sole exception of an additional intron in the dDAT sequence encoding the large EL2 (at proline 159). Exon 5, which encodes most of this extracellular loop, contains five of the seven predicted N-glycosylation sites and displays only modest homology to other mammalian transporter exons. Interestingly, the algorithms used by the Drosophila Genome Project to predict coding sequences did not identify this particular region of the gene as an exon, highlighting the importance of comparing results from in silico and in vitro approaches. The dDAT gene contained only short introns (60-70 bp), except for the first two introns, which are 0.9 and 2.4 kb in size. One striking feature of the dDAT gene is that the large intron 2 shows islands of remarkable homology (up to 45% nucleic acid identity) to regions in the corresponding hDAT intron (data not shown), implying a strong evolutionary pressure to conserve the sequence and function of this intron. Interestingly, this intron has been implicated in cell-type specific expression of hDAT (Kouzmenko et al., 1997). In addition, we used two RT-PCR based strategies to investigate whether alternatively spliced dDAT cDNAs exist. Both approaches were successfully used to identify alternative splicing of the human norepinephrine transporter gene (Pörzgen et al. 1998). The first strategy used gene-specific primers designed to amplify the predicted ORF to detect alternative splicing of internal exons. The second approach used a gene-specific sense primer and an oligo-dT18-adapter primer to identify alternative C-terminal exons. After sequencing 35 dDAT cDNA clones, we did not find any evidence for alternative splicing of the dDAT gene.

Temporal and Spatial Expression of dDAT. Although the monoamines DA, OA, and 5-HT are widely distributed in the D. melanogaster CNS, immunocytochemical methods have identified specific neuronal cell groups, each synthesizing one of the three transmitters (Budnik and White, 1988; Valles and White, 1988; Monastirioti et al., 1995). We used in situ hybridization in whole-mount D. melanogaster third instar larval CNS to demonstrate that the distribution of the dDAT mRNA was expressed in a pattern consistent with that of DA-containing cell bodies. In the larval brain (Fig. 3B), the dDAT mRNA-probe reacts with the three DA cell groups that are found in each lobe. Figure 3C shows the characteristic staining in the ventral ganglion of the eight unpaired abdominal medial DA neurons along the midline of the nerve cord and the two rows of seven dorsal lateral DA neurons. This pattern of expression is consistent with the pattern of DA cells bodies, but is not consistent with the distribution of either serotonin or octopamine cell bodies (Valles and White, 1988; Monastirioti et al., 1996). The in situ data presented here strongly suggest that the DA system in D. melanogaster possesses a unique mechanism for DA clearance that is not shared by 5-HT- or OA-containing cells. We also determined the developmental profile of dDAT mRNA expression (Fig. 4). Semiquantitative RT-PCR analysis showed robust dDAT expression in late embryos (12-24 h) and during larval development. dDAT mRNA expression decreases during pupal development and increases again in adult flies, with heads showing a greatly enhanced signal over bodies. The low level of dDAT RNA expression in 0- to 4-h embryos may be attributable to residual, maternal RNA.

View larger version (128K): [in this window] [in a new window]   Fig. 3.   Restricted expression of dDAT mRNA in dopaminergic cell bodies in D. melanogaster third instar larvae. A, schematic map of DA () and 5-HT () cell bodies in fly larval brain. Ventral and dorsal image of dDAT expressing cells in the brain lobe (B) and in the nerve cord (C). Arrows indicate midline.

View larger version (40K): [in this window] [in a new window]   Fig. 4.   Developmental regulation of dDAT expression. A, agarose gel electrophoresis of a dDAT. B, a rp49 (ribosomal protein 49) PCR-product obtained after 25 cycles with a Drosophila Rapid-Scan Gene Expression Panel (details described under Experimental Procedures). C, densitometric analysis of A and B: Relative level of dDAT expression compared with rp49 (normalized to dDAT/rp49 of male head = 1).

Functional Expression of dDAT. To investigate the functional properties of the transporter encoded by the dDAT cDNA, we established a Madin-Darby canine kidney cell line stably expressing dDAT (dDAT-MDCK cells). In these cells, the uptake of monoamines obeyed Michaelis-Menten kinetics and was effectively blocked by nisoxetine (Fig. 5A). The rank order for the maximal uptake velocities (V_max) for the biogenic amines was DA > NE > TA  Epi. DA was clearly the preferred substrate of dDAT because it exhibits the highest V_max and the highest apparent transport affinity (K_T) of any substrate examined (Table 1). dDAT-expressing cells did not show specific, saturable transport of 5-HT, indicating that 5-HT is not a substrate for dDAT (data not shown). To estimate the transport efficacies (Bönisch, 1998) for the biogenic amines DA, NE, and TA, we determined the ratio V_max/K_T and compared these values with data obtained from MDCK cells stably transfected with hNET and hDAT (Nguyen and Amara, 1996; Daniels and Amara, 1999). The data, as summarized in Table 1, show that the transport kinetics and the estimated substrate efficacies of dDAT are more similar to hDAT than hNET.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Functional characterization of dDAT. A, saturation kinetics of specific amine uptake into stable dDAT-MDCK cells. Cells were assayed with 50 nM ([3H]DA), 100 nM ([3H]NE) or 200 nM ([3H]TA or [3H]Epi) and increasing concentrations of cold amine as described under Experimental Procedures. Nonspecific uptake was determined in the presence of 20 µM nisoxetine. Shown is a representative set of data from 2 to 5 independent experiments for each amine. The KT and Vmax values for [3H]DA, [3H]NE, [3H]TA, and [3H]Epi were 4.6 µM and 153 fmol/s · well; 53 µM and 107 fmol/s · well; 39 µM and 45 fmol/s · well; and 22 µM and 23 fmol/s · well, respectively. B, ion-dependence of DA uptake into dDAT-MDCK cells. Uptake assays were performed with 25 nM [3H]DA. DA transport was dependent on extracellular sodium. Equimolar replacement of sodium by choline (ChoCl) or lithium (LiCl) abolished >95% of the nisoxetine-sensitive uptake. Extracellular chloride only facilitated [3H]DA uptake, but was not required because complete replacement of chloride by gluconate/nitrate (NaGluc) maintained 25 to 30% of the control uptake. C, transport-elicited currents in voltage-clamped X. laevis oocytes, injected with dDAT-cRNA. Oocytes were clamped at 60 mV; amine concentrations are given in micromolar above each drug application and were 5- to 10-fold of their apparent affinities, as determined in dDAT-MDCK cells (see Table 2).

Transporter-Mediated Conductances. Because substrate uptake by monoamine neurotransmitter transporters is an electrogenic process (Mager et al., 1994; Galli et al., 1995; Sonders et al., 1997), we also evaluated the capability of various amines to elicit transport-associated currents. dDAT cRNA was injected into X. laevis oocytes and drug-mediated currents were recorded using a two-electrode voltage clamp protocol as described under Experimental Procedures. The amines were applied at concentrations approximately 5- to 10-fold higher than their observed IC₅₀ values (Table 2) to ensure sufficient occupancy of the transporter. Consistent with the results of the uptake experiments in dDAT-MDCK cells, DA, NE, and TA all generated transport-associated currents in oocytes clamped at 60 mV (Fig. 5C). However, the magnitude of the TA-elicited currents was larger than would be predicted by the results of radiotracer flux studies in dDAT-MDCK cells. This discrepancy implies either that the charge/substrate flux ratio for TA uptake exceeds that for DA uptake, or that the maximum uptake velocity for each substrate varies with cellular background.

View larger version (9K): [in this window] [in a new window]   Fig. 6.   Voltage-dependence of steady-state currents in dDAT- and hDAT-expressing X. leavis oocytes. Currents were recorded from voltage-clamped oocytes before, during, and after drug application and are represented as those elicited by drugs (Idrug  Icontrol). At negative potentials DA elicits robust inward currents in both hDAT- and dDAT-expressing oocytes. In contrast, dDAT does not mediate the pronounced ligand-sensitive proton leak conductance that is found in hDAT-expressing oocytes (Sonders et al., 1997). Block of the leak conductance is manifested in these substractions as regions of negative slope of the current-voltage curves.

Inhibitor and Substrate Pharmacology of dDAT. Initial radiotracer flux studies in dDAT expressing X. laevis oocytes and COS-7 cells suggested that transport of DA is most sensitive to the NET-selective ligands nisoxetine and desipramine but relatively insensitive to GBR 12909, a selective inhibitor of the mammalian DATs (Tatsumi et al., 1997). This similarity of dDAT to the mammalian NETs was surprising, because D. melanogaster do not have significant amounts of NE and because dDAT mRNA was not detected in cell bodies that contain OA, the arthropod equivalent of NE. To further explore the unexpected finding, we studied the pharmacological sensitivity (IC₅₀ values) of [³H]DA uptake in dDAT-MDCK cells and compared these values with data obtained for mammalian monoamine transporters (Table 2). Data for the mammalian carriers were generated using stably transfected MDCK cell lines or by using previously reported values.

Transporter-Mediated Efflux. The most salient function of the different monoamine carriers is the sodium-dependent, high-affinity uptake of their cognate substrates. However, monoamine transporters can also catalyze the outward transport (efflux) of intracellular amines (Bönisch and Trendelenburg, 1988). For instance, a variety of pharmacological agents that are carrier substrates, including amphetamines and TA, produce their psychostimulant and sympathomimetic effects by inducing transporter-mediated efflux of catecholamines. Because the potency of a drug to induce transporter-mediated efflux correlates with its apparent affinity to be transported, efflux studies can also provide indirect evidence whether a nonlabeled solute is a substrate for the transporter under investigation (Langeloh et al., 1987). In flies, TA has been shown to be involved in the development of sensitization to repeated cocaine exposures (McClung and Hirsh, 1999), but the mechanism for the increase in behavioral responses has not been established. A mechanism by which TA could act on dopaminergic or serotoninergic pathways and contribute to behavioral sensitization could be through the stimulation of transporter-mediated efflux of neurotransmitters. To address this possibility, we examined the ability of TA and other bioactive amines to stimulate outward transport of preloaded [³H]DA or [³H]5-HT through dDAT or dSERT, respectively. COS-7 cells transiently expressing dDAT or dSERT were loaded with their respective tritiated substrates and subsequently superfused with cold amine substrates. The amount of radiolabeled neurotransmitter released during uptake of test substrate was expressed as a fraction of the total recovered radioactivity. Consistent with the idea that DA, TA and (+)-amphetamine are substrates for the carrier, these compounds stimulated efflux of preloaded [³H]DA from dDAT expressing COS-7 cells, whereas 5-HT did not stimulate outward transport (Fig. 7). OA stimulated a small but significant amount of efflux, consistent with its limited efficiency in generating transport-associated currents and its low apparent affinity for dDAT. In similar experiments using dSERT-expressing COS-7 cells, 5-HT, TA, and (+)-amphetamine all induced significant efflux of preloaded [³H]5-HT. However, somewhat unexpectedly, DA and, to a lesser extent, OA, stimulated dSERT-mediated efflux, suggesting that substrate recognition by dSERT is less selective than that of dDAT. These results provide additional evidence that DA and TA, but not 5-HT and OA, are efficient substrates for dDAT and support the idea that transporter-mediated efflux through dDAT and dSERT could contribute to the neuromodulatory actions of TA on behavioral responses in flies.

View larger version (32K): [in this window] [in a new window]   Fig. 7.   Stimulation of transporter-mediated 3H-amine efflux. dDAT and dSERT-transfected COS-7 cells were loaded with [3H]DA or [3H]5HT as described under Experimental Procedures. Transporter-mediated efflux was stimulated by addition of cold amine to the extracellular medium. Amine/drug concentrations were: DA, 5-HT (at their cognate carrier), and (+)-amphetamine, 30 µM; TA, 300 µM; OA, 1 mM; mazindol, 20 µM; DA (at dSERT), 300 µM; 5-HT (at dDAT), 300 µM. The amount of efflux is expressed as percentage of the total recovered 3H tracer per well. Spontaneous efflux (vehicle) was determined in parallel in dDAT or dSERT-transfected cells incubated without any drug present. Data points are means ± S.E.M. from two pooled independent experiments, each performed in duplicate. Data were analyzed using one-way ANOVA analysis, followed by Tukey's multiple comparison tests, comparing all drug treatments for a given transporter with each other. P-values are shown only for the comparison of vehicle versus drug-stimulated efflux. **P <0.001; *, P <0.05.

Phylogenetic Relationships between dDAT and Other Neurotransmitter Carriers. Because the sequence homology and pharmacological profile of dDAT overlaps with both mammalian catecholamine transporters, NET and DAT, we considered the possibility that the dDAT gene could represent an ancestral precursor of the two mammalian carrier subtypes. Thus, we explored the evolutionary relationship of dDAT within the NTT family. Phylogenetic analysis by Lill and Nelson (1998) has shown that the mammalian monoamine transporters resemble a distinct subfamily within the NTT family and can be separated into three branches: SERTs, DATs, and NETs. By constructing an evolutionary tree using nucleic acid sequence alignments of the invertebrate carriers dDAT, dSERT, ceDAT, and the mammalian biogenic amine and amino acid carrier sequences, we found that dSERT was grouped in a branch together with its mammalian homologs, rSERT and hSERT. In contrast, dDAT and ceDAT formed a separate branch that originated from the main line before the branch point for the vertebrate catecholamine carriers, DAT and NET (including the frog epinephrine transporter, fET, Apparsundaram et al., 1997) (Fig. 8). Alignments were also performed with complete and partial amino acid sequences (composed of individual exons or TMDs); in the resulting phylogenetic trees, the branching point of dDAT was placed consistently before the divergence of the vertebrate catecholamine transporters. In contrast, dSERT was predominantly found grouped in the same branch with hSERT (data not shown). These results support the assumption that the invertebrate dopamine transporters (dDAT and ceDAT) represent a primordial catecholamine transporter gene, which is also consistent with the observed hybrid NET- and DAT-like pharmacology of the two transporters (this report; Jayanthi et al., 1998).

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Phylogenetic relationships of dDAT. Nucleic acid sequence alignment and bootstrapping of members of the Na/Cl-dependent NTT family was done with CLUSTAL W and CLUSTALTREE software, respectively (Thompson et al., 1994). Bootstrapping is a method to derive confidence values for the groupings of the tree and were calculated from n = 1000 independent trials with a generator seed = 333. The bootstrap values that were generated are noted for each branch. The unrooted phylogenetic tree was constructed by DRAWTREE (Felsenstein, 1989). The vertebrate catecholamine transporters (NET, DAT and fET) and the serotonin transporters (including dSERT) were grouped in two separate branches of the phylogenetic tree. The origin of a third branch containing dDAT and ceDAT is placed between these two branches, suggesting that dDAT and ceDAT originated from a gene that represents a common ancestor of both vertebrate DATs and NETs. The accession numbers of the aligned sequences are: bNET (X79015), ceDAT (AF79899), dDAT (AF260833), dSERT (U02296), fET (U72877), human betaine/-aminobutyric acid transporter subtype 1 (hBGT1) (L42300), human creatine transporter (hCRET; L31409), hDAT (L24178), human -aminobutyric acid transporter subtype 1(hGAT1; X54673), human glycine transporter c (hGlyT1c; S70612), hNET (M65105), hSERT (L05568), rat DAT (rDAT; M80233), rat NET (rNET; Y13223), rat SERT (rSERT; Y11024).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

We report here the isolation and characterization of the D. melanogaster dopamine transporter gene, which was identified in the fly genome based on its sequence similarity to the human DAT. Because of its comparable sequence homology to the mammalian catecholamine transporters (NET and DAT) and the dopamine carrier from C. elegans (Jayanthi et al., 1998), it seemed equally possible that the identified gene could encode either a fly octopamine or dopamine carrier or a nonselective transporter for both monoamines. We applied several functional and anatomical methods to classify the isolated transporter as the D. melanogaster dopamine transporter. Kinetic analysis of [³H]amine uptake in dDAT-MDCK cells demonstrated that DA is the preferred substrate for dDAT as reflected by the rank order of the biogenic amines for their apparent transport affinities (K_T) and transport efficacies (V_max/K_T), DA > NE  TA = Epi (Table 1). Low micromolar concentrations of DA elicit transport-associated currents in dDAT-expressing oocytes (Fig. 5c and 6) and also effectively stimulate dDAT-mediated efflux of preloaded [³H]amines from transfected cells (Fig. 7). Furthermore, dDAT is stereoselective for (+) over () amphetamines and discriminates against the NET substrates bretylium, guanethidine, and OA (Table 2). Overall, the substrate selectivity of dDAT is very similar to the selectivity observed for mammalian DATs. Compatible with the functional data, in situ hybridization analysis showed that dDAT mRNA expression is restricted to DA-producing cell bodies in the fly nervous system (Fig. 3).

Inhibitor Pharmacology. The pharmacology of dDAT is complex; however, a surprising aspect of the functional properties of dDAT is an inhibitory profile resembling that of the mammalian NETs (Table 2). This trend was also described for dDATs invertebrate homolog, ceDAT. Because nisoxetine and the tricyclic antidepressants (TCAs) desipramine, imipramine, and amitriptyline have such high affinities for dDAT, its pharmacological profile can be classified as NET-like and is clearly distinct from the mammalian DATs and SERTs as well as dSERT. We therefore reasoned that the amino acid sequences of dDAT and ceDAT could provide valuable sequence information to map the contact sites of high-affinity antidepressant binding. For example, amino acid residues conserved in dDAT, ceDAT, and the mammalian NETs, but different in the mammalian DATs, could potentially contribute to the formation of an antidepressant binding site. By comparing the various sequences, we identified several AA residues that matched these criteria and were located in TMD 4 to 8, a region that had been implicated in high-affinity antidepressant binding using chimeras of the mammalian DATs and NETs (Giros et al., 1994; Buck and Amara, 1995). Another region has also been implicated in tricyclic antidepressant binding to the serotonin carriers: studies with cross-species chimeras of human and rat SERTs (Barker et al., 1994; Barker and Blakely, 1996) have identified a phenylalanine in TMD12 that confers the higher affinity for imipramine observed for hSERT compared with its rat homolog. Interestingly, the analogous AA exchange introduced in hNET does not affect antidepressant affinity. These studies suggest that multiple domains and contact sites contribute to high-affinity antidepressant binding in a transporter-specific manner. Definition of the different amino acid residues that interact with antidepressant compounds may also guide modeling studies of transporter helix packing in the absence of higher resolution structural data.

Electrophysiologial Properties of the dDAT. Both dDAT and hDAT mediate transport-associated inward currents when expressed in X. laevis oocytes (Fig. 5C and 6). However, in dDAT-expressing oocytes, we were unable to detect the pronounced, substrate- and inhibitor-sensitive constitutive leak conductance, which is a characteristic feature of the mammalian monoamine carriers (Mager et al., 1994; Galli et al., 1995; Sonders et al., 1997). This could mean either that dDAT does not share the proton-selective leak conductance of the human DAT or that ligand-binding does not impede a leak associated with the dDAT. In any case, dDAT seems to be a unique template to dissect mechanistic and structural aspects of these transporter-mediated conductances.

dDAT Is a Target for Cocaine. Along with dSERT (Corey et al., 1994; Demchyshyn et al., 1994), dDAT is one of two cocaine-sensitive targets identified in flies thus far. Although, in the same expression system (transiently transfected COS-7 cells), cocaine displayed a 10-fold lower apparent affinity for dDAT than for dSERT (data not shown), cocaine's low micromolar affinity for dDAT (IC₅₀ = 2.6 µM) is potentially sufficient to effectively block the transporter in vivo under the experimental conditions of exposure to volatilized free base cocaine (McClung and Hirsh, 1998). Besides, cocaine could have a higher affinity for dDAT in vivo than for the cloned transporter in vitro, a phenomenon observed for the rat dopamine transporter (Richelson and Pfenning, 1984; Kilty et al., 1991).

Significance of dDAT to Neurotransmitter Actions in Flies. Although DA is the preferred substrate for dDAT, our studies also show that the biogenic amine TA is an efficient substrate for the carrier and thus dDAT could modulate the actions of TA in the fly. TA serves as the metabolic precursor of the neurotransmitter OA, and there also seems to be a subset of neurons that selectively contains TA but not OA (J. Hirsh, unpublished observations), consistent with a possible function of TA as a neurotransmitter/modulator. Furthermore, TA acts on 2-like adrenergic receptors in insects (Arakawa et al., 1990; Saudou et al., 1990; Kutsukake et al., 2000) and, in flies, mediates physiological effects that are clearly distinct from the actions of DA and OA. For example, TA is essential for the sensitization of D. melanogaster toward cocaine, and a single cocaine exposure induces TA biosynthesis in flies (McClung and Hirsh, 1999). However, TA does not directly stimulate the locomotor activity of decapitated fly nerve cord preparations as observed after DA, 5-HT, or OA treatment (Yellman et al., 1997). Thus, TA could act on either or both presynaptic transporters or postsynaptic receptors to modulate the behavioral sensitization observed after repeated cocaine administration.

High-Affinity Octopamine Clearance Does not Seem to Be Mediated by dDAT. In contrast to TA and NE, OA did not seem to be a good substrate for dDAT, based on its low apparent affinity (IC₅₀ = 281 µM) for inhibiting DA uptake, the small transport-associated currents in dDAT-expressing oocytes (Fig. 5C), and its poor ability to stimulate efflux of [³H]DA from dDAT-expressing COS-7 cells (Fig. 7). Moreover, we did not detect dDAT expression in octopaminergic cells, suggesting the existence of an alternative clearance mechanism for OA. Even though the genome of D. melanogaster is nearly completely sequenced, we were not able to identify any additional, putative monoamine transporter candidate in the GADFLY database (http://hedgehog.lbl.gov:8000/cgi-bin/annot/query), raising the possibility that OA clearance in insects is accomplished by a different transporter or metabolizing enzyme. Removal of OA by less-selective transporters for cationic amino acids (Sloan and Mager, 1999) or organic cations (Gründemann et al., 1994) could provide a plausible alternative mechanism for inactivation.

dDAT May Be a Primordial Catecholamine Transporter. Both dDAT and ceDAT share with the mammalian SERTs their high affinity for TCAs, the capability to discriminate between (+)- and ()-amphetamines, and a moderate sensitivity for cocaine (Table 2), suggesting that these features are mediated by the same structural correlates and that these structures may represent old motifs of the monoamine transporter family. It seems unlikely that these complex features have emerged independently within the NTT gene family. Curiously, the mammalian NETs lack the selectivity to discriminate between the stereoisomers of amphetamine but are sensitive to TCAs, whereas the opposite situation is found for the mammalian DATs. Because the functional profile of dDAT has characteristics resembling those of both mammalian NETs and DATs, and the fly genome does not seem to contain an additional monoamine transporter, it seems conceivable that the dDAT gene represents a common ancestral gene for the vertebrate catecholamine transporters. Consistent with this idea, phylogenetic analysis of members of the Na⁺/Cl-dependent NTT family show the invertebrate carriers dDAT and ceDAT grouped together in a branch that emerges from the main line before the vertebrate catecholamine carriers branch into DAT and NET families, whereas dSERT is clearly grouped together with its mammalian SERT homologs. Consequently, we propose that the genomes of such invertebrates as D. melanogaster and C. elegans contain only two distinct monoamine transporter genes, a SERT and a catecholamine transporter. The emergence of the distinct, vertebrate DAT and NET families may be the consequence of a gene duplication event of a primordial catecholamine transporter gene that is still unique in the invertebrate genomes of D. melanogaster and C. elegans.