Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The catecholamines dopamine (DA) and norepinephrine (NE) are essential modulatory neurotransmitters in the nervous system of vertebrates. DA and NE neurotransmission have many common features. DA and NE derive from the aromatic amino acid tyrosine by the same metabolic pathway (Nagatsu and Stjarne, 1998). All their receptors belong to the same group of G-protein-coupled receptors (Valdenaire and Vernier, 1997; Bockaert and Pin, 1999). Termination of the catecholaminergic transmission occurs mainly through a fast and active reuptake by membrane transporters into the presynaptic terminals. In mammals, the neuronal dopamine transporter (DAT) and the norepinephrine transporter (NET) are both closely related members of the Na⁺/Cl-coupled neurotransmitter transporter family (Amara and Kuhar, 1993; Giros and Caron, 1993). They are intrinsic membrane proteins containing 12 putative transmembrane domains (TMD), with both the N and C termini residing within the cytoplasm (Giros and Caron, 1993; Brüss et al., 1995).

DAT and NET share many pharmacological features, including the high-efficiency transport of both DA and NE. The uptake specificity of these carriers is not solely a consequence of their pharmacological characteristics. The contribution of each transporter to the synaptic regulation of DA and NE essentially depends on the transporter localization in the corresponding nerve terminals. NET transcript is present only in noradrenergic cells of the nervous system, which are found posterior to isthmus, such as the locus ceruleus, and in peripheral sympathetic nerves (Ordway et al., 1997; Comer et al., 1998; Nishimura et al., 1999). In contrast, DAT mRNA is found anterior to the isthmus, mainly in the mesencephalic dopaminergic nuclei substantia nigra and ventral tegmental area. The corresponding protein is transported to the terminals of these neurons, mostly to the dorsal and ventral striatum (Freed et al., 1995; Nirenberg et al., 1996). The catecholaminergic nuclei found on both sides of the mid-hindbrain junction are well conserved in all vertebrates but are specified by different developmental mechanisms (Smeets and Reiner, 1994; Ye et al., 1998; Goridis and Brunet, 1999).

DAT and NET have been isolated from several mammalian species, including human (Pacholczyk et al., 1990; Giros et al., 1992), rat (Giros et al., 1991; Kilty et al., 1991; Shimada et al., 1991; Brüss et al., 1997), mouse (Donovan et al., 1995; Fritz et al., 1998), and bovine (Usdin et al., 1991; Lingen et al., 1994). Recently, a peripheral epinephrine transporter (fET) was characterized in the bullfrog Rana catesbiana (Apparsundaram et al., 1997), and a single catecholamine transporter (ceDAT) was found in the genome of the nematode Cænorhabditis elegans (Jayanthi et al., 1998) and the arthropod Drosophila melanogaster (Pörzgen et al., 2001). Thus, it is possible that the simultaneous presence of DAT and NET is not a conserved character in all Bilateria (including vertebrates). In particular, DAT and NET expression may depend on specific gene duplications and the differentiating mechanisms of the corresponding neurons. The presence or absence of one of these transporters may have strong influence on the effects of catecholamines in the nervous system of a particular species.

To examine in more detail the role of catecholamine transporters in the vertebrate brain, we chose to search and characterize them in the teleost fish medaka (Oryzias latipes). The medaka is becoming a popular vertebrate model (Ishikawa, 2000), in addition to the widely used zebrafish (Danio rerio). It is an easy-to-breed, small, aquarium fish with transparent embryos. It is suitable for large-scale mutagenesis, and its genome is certainly less redundant than that of zebrafish. An important reason to study a ray-finned fish is that this vertebrate group has evolved independently from the sacropterygians, the lineage that led to tetrapods and includes mammals. Thus, the characteristics shared by ray-finned fishes and tetrapods are likely to be ancestral and will serve as a basis for analyzing changes in the roles of catecholamine transporters in vertebrates.

The brain anatomy and catecholaminergic systems of several teleost species have been studied extensively (Ekström et al., 1986; Roberts et al., 1989; Ekström et al., 1990; Sas et al., 1990; Corio et al., 1991; Meek, 1994), including medaka (Kapsimali et al., 2001). These studies show that despite the everted telencephalic development of the teleost brain, the anatomy of catecholaminergic nuclei and pathways are highly conserved compared with mammals (Smeets and Reiner, 1994; Reiner et al., 1998). Consequently, it seems very attractive to examine, in a comparative perspective, the nature, characteristics, and localization of the catecholamine transporters in fishes.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. [³H]Dopamine (42 Ci/mmol) and [³H]norepinephrine (39 Ci/mmol) were purchased from Amersham Pharmacia Biotech UK, Ltd. (Little Chalfont, Buckinghamshire, UK). [¹²⁵I]RTI-55 (2200 Ci/mmol) was purchased from PerkinElmer Life Science (Boston, MA). Dopamine, norepinephrine, nortriptyline, nomifensine, desipramine, fluoxetine, and citalopram were obtained from Sigma/RBI (Natick, MA). Cocaine was kindly provided by Dr. M.-H. Thiebot (Paris, France), and d- and l-amphetamine were kindly provided by Dr. C. Pifl (Vienna, Austria).

Cloning of meNET. A cDNA probe from the rat DAT coding sequence was obtained by PCR amplification of the plasmid TS3-pCMV5 (Giros et al., 1991) with the primers 5'-CCTGCTATCAGTCATCGGCTTTGC and 5'- AGCAGAACAATGACCAGCACCAGG, which correspond to nucleotides 207 to 230 and 749 to 726 of the rat DAT sequence and flank the large extracellular loop. The PCR amplification was run at 94°, 54°, and 72°C for 30 cycles (1 min each). Plaques (700,000) from a cDNA ZAPII library prepared from total brain of medaka (Joly et al., 1997) were transferred onto nitrocellulose filters (BAS85; Schleicher & Schuell, Dassel, Germany) and prehybridized at 42°C for 3 h in 30% formamide, 1× Denhardt's solution (0.1% polyvinylpyrrolidone, 0.1% bovine serum albumin, and 0.1% ficoll), 0.01% SDS, 20 mg/ml salmon sperm DNA, 20 mg/ml yeast tRNA, 20 mM Tris-HCl, pH 7.4, and 4× SSC (60 mM, Na₃ citrate, 0.6 M NaCl, pH 7.0). Hybridization was carried out overnight at 42°C in the same buffer containing also 10% dextran sulfate and 5 × 10⁵ cpm/ml of the nick-translated PCR probe (labeled with [³²P]dATP and [³²P]dCTP at a specific activity of 1 to 2 × 10⁹ cpm/mg of template). The filters were washed twice for 20 min in 2× SSC, 0.1% SDS at 42°C, and twice for 20 min in 0.2× SSC, 0.2% SDS at 42°C. Positive clones were plaque-purified after a second round of enrichment, and the pBluescript-containing fragments were excised from the phages and sequenced in both orientations by the dideoxynucleotide chain-termination method using an automated sequencer (ABI Prism 377; Applied Biosystems, Foster City, CA).

Transfection of COS-7 Cells and Uptake Experiments. A full-length EcoRI fragment from a positive clone was subcloned in the EcoRI site of the pRc/CMV expression vector (Invitrogen, Carlsbad, CA) and sequenced to confirm the proper orientation. COS-7 cells were grown in Dulbecco's modified Eagle's medium (Invitrogen), 10% fetal calf serum (Valbiotech, Paris, France), and 100 Units/ml penicillin-streptomycin (Invitrogen) at 37°C and 7% CO_2. The cells were transfected by the phosphate-calcium method using 5 to 10 µg of a cesium chloride-prepared plasmid. One day after transfection, cells were plated in 24-well dishes, and uptake experiments were performed 72 h after transfection, as described previously (Giros et al., 1992). For the determination of IC₅₀ values, uptake assays were performed for 10 min in the uptake buffer (5 mM Tris base, 7.5 mM HEPES, 120 mM NaCl, 5.4 mM KCl, 1.2 mM CaCl_2, 1.2 mM MgSO₄, 1 mM ascorbic acid, and 5 mM D-glucose, final pH, 7.4) at 37°C using 20 nM [³H]DA with competitors added 5 min before. For determination of K_M and V_max values, uptake assays were performed with 20 nM [³H]DA diluted with increasing concentrations of unlabeled DA (20-30 mM). Nonspecific [³H]DA uptake was determined in the presence of 10 mM nomifensine (Giros et al., 1992). Assays were terminated by rapid removal of the supernatant followed by two successive washes with ice-cold uptake buffer. Cells were lysed in 0.5 ml of 0.1 M NaOH, and the radioactivity was quantified by direct liquid-scintillation counting. Calculations of V_max, K_M, and IC₅₀ values were performed as described previously (Giros et al., 1992). All experiments were carried out in triplicate.

In Situ Hybridization cDNA Probes. A meNET fragment of 680 base pairs was amplified by PCR (HiTaq; Bioprobe, Montreuil sous Bois, France) on meNET cDNA using two specific primers (sense 5'-AAAGGTGTGGGCTACGCTGT and antisense 5'-TTTTGAGCCATGTATCCCAG) and subcloned into the pCRII vector (Invitrogen). A medaka tyrosine hydroxylase fragment of 388 base pairs was amplified by PCR (Promega, Charbonnières, France) on medaka adult brain cDNA by using two oligonucleotides (antisense 5'-CGTGCCTTCCGYGTGTTCCAGTG and sense 5'-CTGGTAGKTCTGGTCYTGGTAGGGCT) and subcloned in the pCRII vector. In both cases, plasmid DNA were digested by appropriate restriction enzymes and transcribed with T7 RNA polymerase from the corresponding promoter. Probes were labeled by adding digoxigenin-UTP to the RNA synthesis reaction medium, and all other nucleotides were unlabeled and present in excess, according to the manufacturer's protocol (Roche Molecular Biochemicals, Meylan, France). Sense probes transcribed from the SP6 promoter were used as negative controls.

Tissue Preparation and In Situ Hybridization. Thirty medaka fish were killed by immersion in ice-cold water and fixed overnight in 4% paraformaldehyde in phosphate buffer. The brains were dissected, postfixed for 1 h in the same fixative, and kept in methanol at 20°C at least for 2 h before the in situ hybridization experiments. Whole medaka brains were processed for in situ hybridization according to the methods used by Joly et al. (1997), except that the duration of the proteinase K treatment was extended from 30 to 45 min to ensure efficient permeabilization of the tissue. After the final step (alkaline phosphatase reaction), the brains were postfixed for 15 min in PAF 4%, washed in PBS, dehydrated, and wax-embedded. Serial sections (8 µm) were prepared in the transverse, sagittal, and horizontal planes. They were counterstained with nuclear-fast red and photographed using a DMRD microscope (Leica, Wetzlar, Germany).

Reverse Transcriptase-Polymerase Chain Reaction Experiments. Total RNA was extracted from the medaka telencephalon and rat brain using the acid/guanidium method (Chomczynski and Sacchi, 1987) and treated with DNase, recovered by phenol-chloroform extraction, and ethanol. Reverse transcriptase (RT)-PCR was performed according to the manufacturer's instructions using the Access RT-PCR system (Promega). The primers used (sense 5'-TGCTACAARAAYGGHGGHGGTGCC and antisense 5'-CCYTTCCAKAGGCTRAARTA) flanked the large second extra-cytoplasmic loop. The RT-PCR products were inserted into PCRII.1 vector (Invitrogen). Competent cells were transformed according to the manufacturer's instructions. The clones were screened by restriction analysis, and those bearing the estimated 530-base-pair product were sequenced.

In Vitro Autoradiographic Binding of the Cocaine Analog [¹²⁵I]RTI-55. Medaka brains were extracted, fixed overnight with 0.1 M phosphate buffer, pH 7.4 (1× PBS) containing 4% paraformaldehyde, rinsed twice with 1× PBS, immersed in 30% sucrose for 48 h, frozen in cold isopentane (40°C), and kept at 80°C until used. Slide-mounted tissue sections (15 µm) were thawed and brought to equilibrium in sucrose buffer (320 mM sucrose, 10 mM sodium phosphate, 10 mM sodium iodide, pH 7.4) for 10 min at room temperature. The sections were then incubated in the absence or presence of cocaine (50 µM), desipramine (100 nM), citalopram (100 nM), or fluoxetine (10 µM) for 20 min before the addition of [¹²⁵I]RTI-55 (50 pM) for 60 min at room temperature. Free and nonspecifically bound [¹²⁵I]RTI-55 was removed by washing the sections twice for 20 min in ice-cold sucrose buffer, twice in water for 5 s, and once in 20% ethanol for 10 s. The sections were dried under a cool stream of air and exposed to a -max film (Amersham Pharmacia Biotech).

Synaptosome Preparation and Uptake Experiments. Synaptosomes were prepared according to the method described by Javitch et al. (1985) with minor modifications. Ten medaka fish brains were dissected at 4°C and homogenized in 1 ml of ice-cold sucrose buffer (sucrose 0.32 M, 1 mM EDTA, and 10 mM Tris-HCl) in a tapered grinder with Teflon pestle (Kontes Glass, Vineland, NJ). The homogenate was diluted (1:3) in sucrose buffer and centrifuged at 1000g for 5 min. The supernatant was collected and the pellet was resuspended in 3 ml of sucrose buffer and centrifuged again at 1000g for 5 min. The supernatants were pooled and centrifuged at 12,000g for 25 min. This pellet was resuspended in 1 ml of sucrose buffer and used for uptake experiments. The uptake buffer was the same as the one used for uptake into cells (except when 120 mM LiCl was used to substitute NaCl). The experiments were performed in 500 µl, comprising 25 µl of the synaptosome suspension, 4 nM [³H]DA (or alternatively 5 nM [³H]5-hydroxytryptamine when specified) (Amersham Pharmacia Biotech), and various uptake inhibitors as indicated and then preincubated for 5 min before the radiolabeled substrates. The dose-response curve for DA uptake was obtained by isotopic dilution with unlabeled DA mixed with [³H]DA. After an incubation of 10 min at 37°C, the uptake was stopped by the addition of 3 ml of ice-cold buffer and immediate filtration through GF/B glass-fiber filters presoaked in 0.05% polyethylenimine (Whatman, Clifton, NJ). The filters were washed twice with 3 ml of ice-cold buffer, and the radioactivity was quantified by direct liquid-scintillation counting. Nonspecific [³H]DA binding on filters (5-6% total) was systematically subtracted. Total uptake (100%) in the presence of Na⁺Cl buffer was 3153 ± 258 cpm of [³H]DA for 10 min.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Low stringency screening of 700,000 plaque-forming units of a medaka brain cDNA library with a rat DAT cDNA probe led to the isolation of 19 clones that were purified and sequenced. Among them, six overlapping clones corresponded to the same partial or complete coding region that strongly resembles the mammalian DAT and NET. The other isolated clones all shared similarities with various subtypes of the GABA transporter family; 5, 2, 3, and 3 clones appeared orthologous to the rat GAT1, GAT2, GAT3, and BGT, respectively.

The longest DAT- and NET-related sequence was 3.3 kilobase pairs long and contained an open reading frame of 1875 base pairs encoding a 625-amino-acid protein, thereafter called meNET (Fig. 1). The meNET sequence displays 12 hydrophobic segments that correspond to the putative TMD characterizing the members of the Na⁺/Cl-dependent transporter family. The sequence of meNET bears five putative N-linked glycosylation sites (N42, N192, N200, N207, and N304). The consensus site of the N-terminal tail (N42) is unlikely to be used, because it is anticipated to be located in the cytoplasmic compartment. Among the three consensus glycosylation sites located in the large extracellular loop between TMD3 and TMD4, two sites (N192 and N200) are highly conserved in other vertebrate catecholamine transporters (Melikian et al., 1996), suggesting that they will also be used in meNET (Fig. 1). In addition, an N-glycosylation site in EL3 is unique to meNET. Three possible phosphorylation motives for protein kinase C (S267, S587, and T593) are found in the meNET sequence as well as one consensus site for phosphorylation by cAMP-dependent protein kinase A (S13). Like the already known NETs, meNET also possesses a leucine zipper motif within TMD2 (L103-L122; Fig. 1). This leucine zipper also exists in DATs (Giros and Caron, 1993), in which the second leucine of the motif is substituted with another hydrophobic residue (methionine). This motif has been implicated in protein-protein interactions, but no experimental evidence is currently available concerning its role in the function of Na⁺/Cl-dependent neurotransmitter transporters. The regions encompassing TMD5 to TMD8 correspond to the tricyclic antidepressant binding domain of catecholamine transporters (Giros et al., 1994; Buck and Amara, 1995). Three positions, F331, S355, and I369, are conserved among all tricyclic-sensitive transporters (meNET, NETs, fET, SERTs, dDAT, and ceDAT) but differ only in mammalian DATs and are substituted by Cys, Thr, and Val, respectively. The mutation from Phe to Cys in this position of the hNET has been shown recently to decrease the affinity of tricyclic antidepressants (Roubert et al., 2001).

View larger version (118K): [in this window] [in a new window]   Fig. 1.   Alignment of deduced amino acid sequence of meNET with other members of the catecholamine transporter gene family. Alignment was performed by using amino acid sequences of the cloned meNET (this study), frog ET (fET; Apparsundaram et al., 1997), human NET (hNET; Pacholczyk et al., 1990), rat NET (rNET; Brüss et al., 1997), bovine NET (bNET; Lingen et al., 1994), bovine DAT (bDAT; Usdin et al., 1991), rat DAT (rDAT; Giros et al., 1991), human DAT (hDAT; Giros et al., 1992), D. melanogaster DAT (dDAT; Pörzgen et al., 2001), and C. elegans DAT (ceDAT; Jayanthi et al., 1998). Consensus sequence for PKC (), leucine zipper motif () and N-linked glycosylation sites (*) are shown. A continuous bar () is located on top of the putative transmembrane domains. White-lettered residues on a black background are strictly conserved for at least 6 of 10 sequences. The PKC site in IL2 (S267) as well as two glycosylation sites in EL3 (N192 and N200) are highly conserved among all the cloned catecholamine transporters.

Alignment of the meNET amino acid sequence with those of the other known monoamine transporters (Fig. 2A) indicates that meNET mostly resembles the fET sequence (72% identity) cloned from the frog R. catesbiana (Apparsundaram et al., 1997). In fact, meNET is related more closely to the NET subfamily than to the mammalian DAT, as illustrated in the phylogenetic tree shown in Fig. 2B. Bootstrap values also support this contention (100 for the branching separating DAT subfamily from the NET group, which contains the medaka NET). Although the number of species studied is still small, it is possible that DAT sequences exhibit a larger apparent sequence divergence than NET sequences (note the difference in phylogenetic distances between human and bovine for the NET and the DAT sequences). This accounts for the deeper branching of DAT, indicating that vertebrate DAT sequences did not diverge before NET in vertebrate evolution; instead, they diverged faster. Incidentally, the three monoamine transporter subfamilies (DAT, NET, and SERT) are clearly individualized from each other and from the GABA transporters. In this respect, it should be noted that the so-called DAT from C. elegans and D. melanogaster branched at the basis of the catecholamine transporter group and cannot be assigned either to the NET or the DAT groups.

View larger version (43K): [in this window] [in a new window]   Fig. 2.   Phylogenetic relationships of meNET with other members of the catecholamine transporter family. b, bovine; h, human; r, rat; ce, C. elegans; f, bullfrog R. Catesbiana; d, arthropod D. melanogaster; and t, Torpedo marmorata. A, identities shared among catecholamine transporters cloned in different species. The values are expressed as the percentage of identical amino acids. B, tree of the phylogenetic distances calculated by the Neighbor Joining Method from an optimized alignment of the amino acid sequences of representative Na+/Cl-dependent neurotransmitter transporters. Numbers correspond to the bootstrap values (occurrence of the presented branching after 100 iterations). Unless indicated, these values are equal to 100. Sequences for DAT and NET are those described in the legend to Fig. 1. Other sequences are from cloned rat betaine/GABA transporter (rBET/GAT; Burnham et al., 1996), rat GAT-1 (rGAT1; Guastella et al., 1990), rat GAT-2 (rGAT2; Borden et al., 1992), rat GAT-3 (rGAT3; Clark et al., 1992), T. marmorata GAT (tGAT; Guimbal and Kilimann, 1994), and C. elegans GAT (CAA98519).

To assess the functional characteristics of the cloned meNET cDNA, we directed its expression into COS-7 cells. Cells transfected with meNET cDNA were able to mediate the uptake of [³H]DA and [³H]NE with high affinity in a saturable manner and with a first-order kinetics, whereas in mock transfected cells, no specific accumulation of catecholamines could be detected. These experiments indicate that meNET was able to transport DA (K_M = 290 nM) with a higher affinity than NE (K_M = 640 nM). The meNET displayed similar capacities for DA (V_max = 1354 ± 308 fmol/min/10⁵cells) and NE (V_max = 860 ± 150 fmol/min/10⁵ cells; Fig. 3). These properties are comparable with those displayed by human or rat NETs, which share a higher affinity for DA but a lower capacity than DAT itself (Table 1).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Functional characteristics of meNET. Eadie-Hofstee representation of concentration dependence of NE and DA transport into COS-7 cells transiently transfected with meNET, hNET, or rDAT cDNA. One representative experiment is shown here. V is expressed in fmol/min/105 cells1.

The ability of various inhibitors of catecholamine transport to block the uptake of [³H]DA was assessed in COS-7 cells transfected with either rDAT, hNET, or meNET (Table 2). We found that meNET exhibited a pharmacological pattern similar to that displayed by DAT and NET for the drugs that block the two catecholamine transporters: cocaine (IC₅₀ = 169 ± 42 nM), d-amphetamine (IC₅₀ = 150 ± 18 nM), benztropine (IC₅₀ = 85 ± 23 nM), and nomifensine (IC₅₀ = 54 ± 18 nM). In contrast to DAT, meNET was inhibited by nanomolar concentrations of the tricyclic antidepressants desipramine (IC₅₀ = 0.92 ± 0.36 nM), nortriptyline (IC₅₀ = 16 ± 4 nM), and by the NET-specific compound nisoxetine (IC₅₀ = 2.6 ± 1.25 nM). Furthermore, we found no significant differences for meNET inhibition by the d- and l-stereoisomers of amphetamine as observed with hNET, whereas they displayed stereospecificity toward rDAT. In addition, both the serotonin transporter inhibitor fluoxetine (IC₅₀ = 1070 ± 90 nM) and serotonin (IC₅₀ > 12,000 nM) showed weak inhibition of meNET-mediated DA uptake (Table 2). Therefore, these pharmacological experiments undoubtedly classified the medaka catecholamine transporter-like sequence as a NET and not a DAT, further justifying its naming.

In mammals, the cellular localization of the transporter transcripts is certainly a better criterion than pharmacology to assign a sequence to the DAT or NET family. The distribution of meNET mRNA was analyzed by in situ hybridization in the medaka brain and compared with the localization of the medaka tyrosine hydroxylase (TH) transcripts. The localization of meNET mRNA was strikingly restricted to a few distinct nuclei of the isthmus and hindbrain (Fig. 4). The midbrain and forebrain were completely devoid of meNET labeling, whereas TH transcripts were detected in numerous nuclei all along the anteroposterior axis of the brain (Fig. 4, A-C). Labeling of the meNET transcripts was observed exclusively in a limited subset of TH-positive cells (Fig. 4). The more anterior meNET-positive nucleus was the locus ceruleus, where all the cells seemed to be labeled by both the meNET and the TH cRNA probes (Fig. 4, D and E). More posteriorly, meNET labeling was observed in a few cells of the nucleus of the solitary tract and in the nucleus of the vagus nerve (Fig. 4F). A strong labeling by both probes was also present in a group of dispersed cells (Fig. 4, F and G) found among the medial longitudinal fascicle, the more lateral descending trigeminal root, and the lateral longitudinal fascicle. Ma (1997) has referred to this group in zebrafish as the interfascicular catecholaminergic neurons. A strong labeling was also observed in the area postrema, but in this area, the number of cells labeled with the meNET probe was approximately half the number of cells labeled with the TH probe (Fig. 4, H and I). Finally, more ventrally, the meNET mRNA was detected in a few cells of the reticular formation.

View larger version (80K): [in this window] [in a new window]   Fig. 4.   Distribution of meNET mRNA in the medaka brain and comparison with TH transcript detected with digoxigenin-labeled probes. A, sagittal section (anterior on the right) showing TH transcripts concentrated in discrete areas along the anteroposterior axis of the brain, such as the internal cellular layer of the olfactory bulb (ICL), posterior preoptic parvocellular nucleus (PPp), ventrolateral thalamic nucleus (VL), nucleus of the periventricular posterior tuberculum (TP), dorsal (Hd), and ventral periventricular hypothalamus (Hv). Black lines correspond to the sections presented in B (anterior) and C (posterior), in which no meNET transcripts are detected, as opposed to TH mRNA. meNET (D) and TH (E) transcripts are detected in all the cells of the locus ceruleus (lc). More posteriorly, cells of the interfascicular catecholamine group (arrows) are labeled with meNET (F) and TH (G) probes, respectively. In the area postrema (ap), the number of cells labeled with the meNET probe (H) represents approximately 50% of those labeled with the TH probe (I). I, TH-positive cells (arrow) are located in the posterior part of the interfascicular catecholamine group.

The striking discrepancy between the distribution of the TH and meNET mRNAs could reflect the presence of an additional catecholamine transporter, which would be present at least in some of the TH-positive and meNET-negative neurons. In particular, the mammalian DAT has been found in the mesencephalic dopaminergic areas, but no labeling with the meNET probe was detected in the midbrain and forebrain in medaka fish. Therefore, to determine whether a DAT-like transporter exists in the medaka brain, we performed a series of additional experiments to specifically address this issue.

We first carried out RT-PCR with degenerate primers designed to recognize all the vertebrate DAT or NET sequences, including meNET. These primers were located in the most conserved regions of DAT and NET (in TMD3 and TMD5), but the resulting PCR product can be easily characterized by the nonconserved second extracellular loop between TMD3 and TMD4. These primers were able to amplify sequences corresponding to rDAT and GAT3 using total rat telencephalon mRNA. They were used to amplify related sequences from mRNA extracted from the whole medaka brain, but also from dissected areas of the forebrain, midbrain, and hindbrain. We reasoned that if a DAT-related transporter exists in the medaka brain, it would be present in the dopaminergic areas anterior to the isthmus. Forty-seven clones were isolated from these experiments. Nine of them were identical to meNET, but no other sequence resembling an additional catecholamine transporter was detected (data not shown).

Furthermore, we looked for specific ligand binding on tissue sections. We used as radioligand [¹²⁵I]RTI-55, an iodinated cocaine analog that exhibits similar affinity for DAT and SERT but has a 10-fold lower affinity for NET in mammals (Eshleman et al., 1999). In coronal sections of the medaka brain, a strong binding of [¹²⁵I]RTI-55 was observed in preoptic, thalamic regions, and other diencephalic areas, as well as in the dorsal midbrain and more posterior regions of the medaka brain (Fig. 5). However, no labeling was detected in the telencephalon. The observed labeling was totally displaced by 10 µM cocaine, an amine transporter inhibitor (data not shown). Fluoxetine 10 µM (Fig. 5, C and G) and desipramine 100 nM (data not shown) completely displaced [¹²⁵I]RTI-55 binding, and no remaining labeling could be observed, even after a long exposure time. These high doses of desipramine and fluoxetine should have been able to compete with SERT and NET but not with DAT. After incubation with 100 nM citalopram, a dose that should compete only with SERT but not with meNET or a putative DAT, a significant labeling remained only in subcortical structures (Fig. 5, B and F), which should represent specific meNET labeling. However, we had to use an overnight incubation of the medaka brain with 4% paraformaldehyde to preserve the tissue structures. Thus, we cannot totally exclude that such a treatment may differentially affect DAT binding compared with NET and SERT binding. The results of this experiment showed that apart from SERT and NET, no other cocaine-sensitive binding site could be found in the medaka brain, and thus, no DAT-like binding site was detected.

View larger version (41K): [in this window] [in a new window]   Fig. 5.   Detection of [125I]RTI-55 binding sites on medaka brain slices. A and E, [125I]RTI-55; B and F, [125I]RTI-55 + 100 nM citalopram; C and G, [125I]RTI-55 + 10 µM fluoxetine; D and H, cresyl violet staining of the sections shown in A, B, C and E, F, G, respectively. A, strong labeling in the optic tract and in the diencephalic preoptic area is observed. E, strong labeling in some diencephalic areas, such as the posterior preglomerular complex, and in mesencephalic areas, such as the torus semicircularis. Scale bar represents 0.5 mm.

To determine the pharmacological profile of the endogenous uptake of catecholamine in the medaka brain, synaptosomes were prepared to analyze [³H]DA transport. The dose-response for DA indicates a saturable uptake (Fig. 6A). Eadie-Hofstee transformation (Fig. 6A, inset) gives a K_M value of 355 ± 80 nM. The [³H]DA was mostly Na⁺-dependent (80% of total uptake in the presence of LiCl) and inhibited by classic NET blockers (Fig. 6B). In fact, nisoxetine (1 µM), benztropine (10 µM), and desipramine (10 nM) were equally potent (80% inhibition), and cocaine (1 mM) was less potent (70% inhibition) in blocking the uptake of [³H]DA detected in medaka synaptosome preparations (Fig. 6B). The dose-response inhibition with cocaine and desipramine gave IC₅₀ values of 350 nM and 1.6 nM, respectively (data not shown). Citalopram (1 µM), a specific SERT inhibitor, could not displace the DA uptake by meNET. A high concentration of DA (1 mM) was able to decrease the [³H]DA uptake by up to 90%, which was significantly lower than that observed in the absence of Na⁺ (LiCl). The uptake of [³H]5-HT was blocked (79%) by the addition of citalopram (data not shown), therefore confirming the presence of a SERT-like transporter as seen with [¹²⁵I]RTI-55 labeling. These data show that 80% of the total DA uptake was achieved in an Na⁺-dependent and desipramine-sensitive manner, whereas the remaining 20% uptake did not fulfill the specific criteria of uptake by an Na⁺-dependent transporter of the NET/GAT family.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Assessment of native [3H]DA uptake from freshly prepared medaka brain synaptosomes. A, dopamine saturation curve and Eadie-Hofstee analysis. Mean value ± S.E.M. (n = 3) for KM (nM) and Vmax (pmol/min/mg protein) are 355 ± 80 and 2.1 ± 0.7, respectively. B, [3H]DA uptake in the presence of various inhibitors and with substitution of NaCl by LiCl.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Molecular and Pharmacological Characteristics of meNET. A cDNA encoding a unique putative catecholamine transporter was isolated from a medaka brain cDNA library. The corresponding protein structure, its pharmacological and functional properties, and the transcript distribution suggest it to be a transporter for NE. The protein was thus named meNET, according to the accepted rules of transporter naming.

Brain Distribution of meNET. The localization of meNET mRNA in the medaka brain also strongly suggests that it is mainly a noradrenaline carrier. The in situ hybridization demonstrated that meNET mRNA is restricted to well-known catecholaminergic nuclei of the isthmus and hindbrain. The meNET mRNA is beyond detection levels in the numerous well-characterized dopaminergic nuclei of the olfactory bulb, hypothalamus, and in the nucleus of the posterior periventricular tuberculum, considered the diencephalic homolog in fish of the substantia nigra in tetrapod (Reiner and Northcutt, 1992; Kapsimali et al., 2001). In mammals, this latter nucleus presents the highest expression of DAT mRNA and a strong TH staining but is completely devoid of NET transcript as found in the medaka brain, using in situ hybridization. The fact that meNET transcripts were detected only by RT-PCR in the anterior brain further suggests that meNET mRNA expression is very weak in the anterior brain. Given the expression and pharmacological profile of meNET, there is little doubt that it transports NE in the noradrenergic synapses of the medaka brain. However, a role of this transporter in DA uptake is also probable in nuclei known to contain both DA-positive and NA-positive cell bodies, such as the area postrema and the interfascicular catecholaminergic neurons, as well as in all the anterior brain regions, which are innervated by noradrenergic fibers originating from the locus ceruleus. This possibility has already been proposed in the prefrontal cortex of the rat (Tanda et al., 1997; Yamamoto and Novotney, 1998). In addition, other high-capacity, low-affinity transport systems for monoamines exist in the fish brain. In particular, it has been shown that the paraventricular organ of the hypothalamus, the neurons of which are in contact with the cerebrospinal fluid, is able to concentrate very high levels of DA and other monoamines (Vigh and Vigh-Teichmann, 1998). Thus, the precise role of the several monoamine clearance systems in the fish brain needs to be more precisely investigated.

Is meNET the Unique Catecholamine Transporter in the Medaka Brain? The pharmacological characterization and anatomical data defined the medaka catecholamine transporter only as a bona fide NET. If the regulation of extracellular catecholamine levels in fish is comparable with that of mammals, a DAT should be present in the medaka brain. However, we were unable to provide any evidence of this.