The Family of Na + / Cl Neurotransmitter Transporters
Nathan Nelson
Department of Biochemistry, Tel Aviv University, Tel Aviv, Israel
Abstract: The termination of neurotransmission isachieved by rapid uptake of the released neurotransmitterby specific high-affinity neurotransmitter transporters.Most of these transporters are encoded by a family ofgenes (Na~ICI transporters) having a similar membranetopography of 12 transmembrane helices. An evolutionarytree revealed fivedistinct subfamilies: y-aminobutyric acidtransporters, monoamine transporters, amino acid trans-porters, “orphan” transporters, and the recently discov-ered bacterial transporters. The bacterial transporters thatbelong to this family may help to develop heterologousexpression systems with the aim of solving the three-dimensional structure of these membrane proteins. Someof the neurotransmitter transporters have been implicatedas important sites for drug action. Monoamine transport-ers, for example, are targeted by major classes of antide-pressants, psychostimulants, and antihypertensive drugs.Localization of individual transporters in specific cells andbrain areas is pertinent to understanding their contributionto neurotransmission and their potential as targets fordrugs. The most important questions in the field includeresolving the mechanism of neurotransmitter transport,the structure of the transporters, and the interactionof each transporter in complex neurological activities.Key Words: Neurotransmission— y-Aminobutyric acidtransporters—Monoamine transporters—Amino acidtransporters—Orphan transporters— Bacterial trans-porters.J. Neurochem. 71, 1785—1803 (1998).
Communication between separate cells requires therelease of extracellular messengers and specific recep-tor mechanisms on or in the target cells. Signalingsubstances include hormones, neurotransmitter sub-stances, trophic factors, and other diffusible sub-stances. In higher organisms, chemical transmissionis the principal way of communication between cells,especially in the nervous system. Nerve cells mediatefast signaling between sensory systems and the CNSand between the CNS and effector systems. Within theCNS, nerve cells form complex circuits that serve theintegration of inputs and the generation of specific ac-tivity patterns. In many cases, synaptic transmissionis chemical and involves the secretion of signaling
substances. In most systems, the termination of thechemical transmission is achieved by rapid uptake ofthe released neurotransmitter, by specific high-affinityneurotransmitter transporters, into the synaptic termi-nal or the surrounding glial cells (Kuhar, 1973; Iversenand Kelly, 1975; Kanner, 1983, 1989; Kanner andSchuldiner, 1987). It has been known for many yearsthat neurons and glia canaccumulate neurotransmittersby a Nat-dependent transport process. Neurotransmit-ters are cotransported with Na+ using the energy storedin transmembrane electrochemical gradients generatedby primary ion pumps (Kanner, 1983).
Studies on neurotransmitter uptake indicated the ex-istence of multiple uptake systems, each relatively se-lective for a specific neurotransmitter (Iversen andJohnston, 1971). Neurotransmitters are transportedacross membranes by at least four distinct families oftransporters: (a) vesicular transporters that function inthe uptake of neurotransmitters into synaptic vesiclesand granules (Schuldiner, 1994); (b) Nat- and Cl-dependent (Nat/Cl) transporters that operate on theplasma membrane of neuronal and glia cells (Schlosset al., 1992; UhI, 1992; Amara and Kuhar, 1993); (c)Na~/K~-dependenttransporters that function on theplasma membranes, especially in glutamate transport(Kanner, 1993); and (d) general amino acid transportsystems that participate in controlling the availabilityof neurotransmitters outside the cells (McGivan andPastor-Anglada, 1994).
A defining moment in the advancement of ourknowledge of neurotransmitter transporters came in1990 when the first cDNAs encoding the y-aminobu-tyric acid (GABA) transporters were cloned and se-quenced (Guastella et al., 1990; Nelson et a!., 1990).
Address correspondence and reprint requests to Dr. N. Nelson atDepartment of Biochemistry, Tel Aviv University, Tel Aviv 69978,Israel.
Abbreviations used: ACHC, cis-3-aminocyclohexanecarboxylate;GABA, y-aminobutyric acid; GATI —4, GABA transporters 1—4;GLYTI and GLYT2, glycine transporters I and 2; Na~/C1 , Na -
Subsequently, several cDNAs encoding these plasmamembrane transporters have been cloned, either by in-dependent expression cloning or by using sequenceconservation between GABA and nonadrenaline trans-porters, with the resulting promise of rapid new ad-vances in understanding the mechanism of neurotrans-mitter reuptake (Blakely et al., 1991; Hoffman et al.,1991; Kilty et al., 1991; Pacholczyk et al., 1991; Shi-mada et al., 1991; Usdin et al., 1991). This reviewwill dea! exclusively with the family of Na + /Cl trans-porters.
STRUCTURE OF Na~IC!NEUROTRANSMITTER TRANSPORTERS
The gene products of the Na~/Cl transporter fam-ily are highly conserved. In mammals, they can begrouped into four subfamilies: GABA transporters,monoamine transporters, amino acid transporters, and“orphan” [typified by neurotransmitter transporter 4(NTT4)] transporters (Nelson and Lill, 1994; Uhl andJohnson, 1994). The transporters of the first three sub-families show a common structure of presumably 12transmembrane helices/domains (TM) with a single!arge loop in the external face of the membrane withpotential glycosylation sites (Fig. 1). The orphan sub-family (NTT4), somewhat remote from the other fam-ily members, was identified by sequencing eDNA frommammalian sources (Uhl et al., 1992; Liu et al.,!993c). Their structure deviates from the norm byhaving two potential glycosylated loops outside themembrane. Knowing the limitations of hydropathyplots as predictors for the existence of transmembranehelices and the topography of membrane proteins(Eisenberg, 1984), it was proposed that the GABAtransporter GATI contains 12 transmembrane helices(Guastella et al., 1990; Nelson et al., 1990). This as-sumption was accepted as a likely working hypothesisby the groups that cloned and sequenced eDNA encod-ing members of this family of transporters (Amara andKuhar, 1993). Similar structures have been reportedfor several other families of transporters, includingbacterial, plant, and mammalian sugar transporters(Hediger et al., 1987; Kaback, 1992; Caspari et al.,1994). The mechanistic significance of this structureis not understood, and there are no apparent commonmotifs in all transporters with the 12-transmembranetopology. However, a dimeric organization of six heli-ces was proposed as a common structure for membranetransporters (Maloney, 1990). Although the number12 has been maintained for transmembrane helices,there is no apparent reason for the position of short andextended loops between the helices. Some transporterscontain charged residues inside their hydrophobicmembrane spanning segments, but several have nocharges in their transmembrane helices. Therefore, ifthere is a common mechanism for transport of substrateacross the membrane, it is hiding behind motifs thatare as yet unrecognized.
The proposed general structure of the Na~!Cltransporters was verified by experimental evidence andlogical arguments (Nelson, 1993; Pantanowitz et a!.,1993). The evidence includes the observations thatthere is no cleavable N-terminal sequence, suggestingthat the N-terminal part of the transporter is situatedin the cytoplasmic face of the plasma membrane (seeFig. 1). Studies using peptide-specific antibodies toseveral regions of the GABA transporter (GAT1 ) com-bined with limited proteolysis with pronase suggestedthat the N- and C-terminal regions are present insidethe membrane (Mabjeesh et al., 1992; Mabjeesh andKanner, 1993). Three transmembrane segments are re-quired to position the large glycosylated !oop at theexternal surface of the plasma membrane. In addition,in the orphan (NTT4) transporter subfamily, an ex-tended hydrophilic loop with potential glycosylationsites exists between the proposed TM 7 and 8. Takingthis together with the experimental evidence on the E.coli lactose permease, which withstood the scrutiny ofnumerous molecular studies (Kaback et al., 1997), itis likely that the structure of Na + /Cl — transporterscontains 12 transmembrane helices.
A direct experimental approach to identify the posi-tion and boundaries of the various transmembrane heli-ces was performed by the use of site-directed antibod-ies and selective cleavage of representative transport-ers. An immunofluorescent study with antibodiesraised against peptide sequences from the N- and C-termini and the large extracellular loop of the noradren-aline transporter, provided data to support the proposedorientation of these domains (Melikian et al., 1994;Bruss et al., 1995). An electron microscopic immuno-chemical study provided support for the cytoplasmiclocation of the dopamine transporter N-terminus (Nir-enberg et al., 1996). Fragments with apparent differentmolecular masses were generated by digestion ofGAT1 in native or reconstituted vesicles with pronase(Mabjeesh and Kanner, 1993). The fragments wererecognized by site-specific antibodies, and the resultingfragments were in good fit with the proposed mem-brane structure of GAT 1.
Although the general structure of the Na/Cltransporters was verified by experimental evidence, theposition of the first three transmembrane helices wasdebated recently (Bennett and Kanner, 1997; Olivareset al., 1997). One of the hallmarks of Nat/Cl trans-porters is the conserved amino acid sequence that in-cludes the residue WRFPXXXYXNGGGAF. Withvery few exceptions, this sequence is highly conservednot only for the amino acid residues, but also in thenucleotide sequences of the mRNA. Moreover, in allthe transporter genes, there is an intron within the cod-ing sequence of the three glycine residues (Liu et a!.,l992a). The position of this intron is conserved notonly in the vertebrate genes, but also in the genesencoding Nat/Cl - transporters in insects. Accordingto the original proposal for the membrane organizationof Nat/Cl transporters, the vicinity of these three
J. Neurochern., Vol. 71, No. 5, 1998
Na ~/(7 NEUROTRANSMITTER TRANSPORTERS 1787
FIG. 1. Schematic representation of themembrane topology of mammalian andbacterial transporters that belong to thefamily of Nat/CL neurotransmitter trans-porters. Potential sizes of the varioustransmembrane helices are depicted, andsome highly conserved amino acids po-tentially situated in the membrane are indi-cated. The net charge of each loop inSTRYP1 (gene of the Symbiobacteriumthermophilum genome) is indicated.
glycine residues is situated between TM 1 and 2. Thisassumption was challenged by two recent studies onthe membrane topology of GAT 1 and the glycine trans-porter GLYTI (Bennett and Kanner, 1997; Olivareset al., 1997). N-Glycosylation scanning mutagenesiswas performed with GAT1 in which the glycosylationsites in the large external 1oop were inactivated by Nto D substitutions. Functional transporters with glyco-sylation at the predicted external loops 3 and 6 essen-tially demonstrated the validity of the membrane to-pography from the large external 1oop (no. 2) to theend of the transporter (Bennett and Kanner, 1997).Surprisingly, insertion of the sequence NNSST to thepredicted internal 1oop 1 resulted in a glycosylation atthis site. However, the mutated transporter was notactive in GABA transport. Similar glycosylation scan-ning was performed with GLYT1, leading to the sameconclusion that the predicted membrane topographyfrom the large glycosylated loop to the C-terminus of
the transporter is as originally predicted by the hydrop-athy plot. This work further suggested that the predi-cated internal loop 1 is essentially facing the outersurface of the membrane (Olivares et al., 1997). Toreconcile this with the fact that the N-terminus is atthe cytoplasmic side of the membrane, both groupsproposed that the rather long TM 3 essentially traversesthe membrane twice, generating TM 2 and 3 (see Fig.1). The original amino acid residues proposed to beTM 2 are the first to cross the membrane. Indeed, insome, but not all, Na~/C1 transporters, the aminoacid sequence that was proposed to be comprised ofthe third transmembrane segment contains up to 38uncharged amino acids (Nelson, 1993). However, inone of the bacterial transporters that are homologousto the mammalian Nat/Cl transporters, thereare only24 uncharged amino acids in the same position (N.Nelson et al., unpublished observations). Moreover, arecent study on the topography of the serotonin trans-
J. Neurochem., Vol. 71, No. 5, 1998
1788 N. NELSON
porter using cysteine-modifying reagents and biotinyl-ated probe scanning endorses the original proposal forthe topography of Na f/Cl — transporters (Chen et a!.,1997a; Barker and Blakely, 1998). In any case, thelong stretch of uncharged amino acids (124—161 inGATI) may have an important role in the mechanismof energy coupling and/or substrate translocationacross the membrane (Chen et al., 1997b). High-reso-lution structural information on even one of the familymembers may shed light not only on their membranetopography, but also on their function at a molecularlevel.
THE FIVE SUBFAMILIES OF Na k/C! -
NEUROTRANSMITTER TRANSPORTERS
Genomic cloning of genes encoding neurotransmit-ter transporters and the search for sequences withoutknown function in the GenBank revealed that insectsand C. elegans contain genes encoding potential neuro-transmitter transporters (Liu et al., I 992a). Therefore,it was assumed that this family of transporters evolvedfrom a common ancestor about 1 billion years ago(Nelson, 1993). Recently, we identified and cloned abacterial gene with relatively high homology to neuro-transmitter transporters (N. Nelson et al., unpublishedobservations). This gene is present in a potential tryp-tophanase operon of Symbiobacterium thermophilum(Hirahara et al., 1992), which contains the genes en-coding the enzyme tryptophanase and a potential tryp-tophan transporter. Recently, the entire genomic se-quences of Haemophilus influenzae and Methanococ-cus jannaschii were determined (Fleischmann et al.,1995; Bult et al., 1996). Each genome contains a singlegene homologous to the mammalian neurotransmittertransporters. These findings indicate that the family ofneurotransmitter transporters is related to the onset oflife and is present not only in eukaryotes, but also ineubacteria and archaebacteria. So far, genes encodingtransporters of this family have not been identified inplants and fungi. It is also noteworthy that homologoustransporters were not found in bacteria, such as E. coli,Bacillus subtilis, and cyanobacteria, where most of thegenomic sequence is known. Therefore, it appears thatplants, fungi, and those bacteria that primarily use pro-tonmotive force for their transport systems gave up theuse of these transporters.
An evolutionary tree constructed with different partsof the various transporters suggests that the bacterialtransporters are loosely grouped with the orphan(NTT4) transporters or with the amino acid transport-ers (Fig. 2). The separation into five subfamilies oftransporters is quite apparent. The subfamilies as wesee them now are GABA transporters, the subfamilyof monoamine transporters, the subfamily of aminoacid transporters, the subfamily of orphan (NTT4)transporters, and the subfamily of bacterial transport-ers. The evolutionary tree reveals some interesting re-lationships between the various transporters. In the
subfamily of GABA transporters, GATI appears to bequite remote from the other subfamily members. Thisis in line with its substrate specificity that is exclusivefor GABA. Similarly, the creatine transporter is alsoseparated from the other subfamily members. All theother subfamily members transport /3-alanine, and theyare grouped together in the evolutionary tree. It is inter-esting that /9-alanine transport requires a higher degreeof amino acid sequence conservation than does GABAtransport. The three subfamilies with known substratesare distinguished not only by their sequence homology,but also by their substrate specificity and pharma-cology.
The GABA transporter as a prototypeGABA is the predominant inhibitory neurotransmit-
ter and is widely distributed in the mammalian brain.Its uptake is catalyzed by transporters ubiquitously dis-tributed in the various brain parts. Biochemical andphysiological studies revealed two major subtypes ofGABA transport systems. The GABAA uptake system,which is inhibited by diaminobutyric acid or cis-3-aminocyclohexanecarboxylate (ACHC), is thought tobe neuronal (Bowery et al., 1976), and the GABABtransport system, which is inhibited by /3-alanine, isthought to be of glial origin (Schon and Kelly, 1975).One of the GABA transporters (GAT1) was the firstto have been purified to apparent homogeneity (Radianet al., 1986). A reconstitution assay was used in whichdetergent-solubilized membrane protein fractions wereincorporated into liposomes of asolectin and brain lip-ids, and then radiolabeled GABA uptake was deter-mined (Radian and Kanner, 1985). A glycosylatedprotein with an apparent molecular mass of 80 kDawas isolated and, following reconstitution, exhibitedACHC-sensitive GABA uptake into the liposomes.The GABA uptake was not sensitive to /3-alanine, andit was concluded that GAT1 is the neuronal GABAtransporter (Kanner et al., 1989). This conclusion wasendorsed by immunocytochemical localization ofGAT1 in rat brain sections using an antibody that wasraised against the purified protein (Radian et al.,1990). However, some of the results indicated thatGATI may also be present in glial cells.
This purified preparation of GAT1 was used for ob-taining cyanogen bromide fragments, and the aminoacid sequences of the polypeptides were used to designoligonucleotide probes to isolate the eDNA encodingthe rat transporter (Guastella et al., 1990). The openreading frame in the eDNA encodes a hydrophobicprotein containing 599 amino acids and having a calcu-lated molecular mass of 67 kDa. The human GABAtransporter was found to be almost identical to the ratprotein, with just 17 amino acid differences, mainlynear the N- (four residues) and C- (11 residues) ter-mini (Nelson et al., 1990). Expression of the GATImRNA in Xenopus oocytes generated GABA uptakeactivity with pharmacology similar to the neuronalhigh-affinity GABA transporter (Guastella et al.,
.1. Neurochem., Vol. 71, No. 5, 1998
Na k/Cl NEUROTRANSMITTER TRANSPORTERS 1789
FIG. 2. Evolutionary tree constructed from the amino acid sequences of representative neurotransmitter transporters. The tree wasconstructed using the amino acid sequences of the part corresponding to TM 4—7 (see [ill and Nelson, 1998). The source of sometransporters is indicated by the last letter: -D, dog; -H, human; -R, rat; -M, mouse. HAEM1 is a gene of the Haemophilus influenzaegenome, METHAN1 is a gene of the Methanococcus jannaschii genome, and STRYP1 is a gene of the Symbiobacterium thermophilumgenome. BETAT, betaine transporter; CRETR, creatine transporter; DOPAT, dopamine transporter; NORAT, noradrenaline transporter;PROT, proline transporter; SEROT, serotonin transporter; TAUT, taurine transporter.
References: (1) Yamauchi et al., 1992; (2) Guimbal and Kilimann, 1993; (3) Shimada et al., 1991; (4) Liu et al., 1992a; (5) LOpez-Corcuera et al. 1992; (6) Liu et al., 1993a; (7) Liu et al., 1992c; (8) Liu et al., 1993b; (9) Fleischmann et al., 1995; (10) Bult et al.,1996; (11) Pacholczyk et al., 1991; (12) Liu et al., 1993c; (13) Smith et al., 1995; (14) Fremeau et al., 1992; (15) Wasserman et al.,1994; (16) Blakely et al., 1991; (17) N. Nelson et al., unpublished observations; (18) Liu et al., 1992b.
J. Neuroche,n., Vol. 7/, No. 5, 1998
1790 N. NELSON
1990). The rat brain eDNA encoding GATI has alsobeen expressed in mammalian cells (Keynan et al.,1992). Tunicamycin sensitivity of the expression ofthe transport activity suggested that the N-linked gly-cosylation is important for the activity and/or properassembly of the transporter. The eDNA of GABAtransporter GAT! is the only neurotransmitter trans-porter that was cloned based on its amino acid se-quence. Most of the other eDNAs encoding membersof this transporter family were cloned based on theinitial sequence information obtained for the GABAand noradrenaline transporters.
A multitude of Na + -dependent GABA uptake sys-tems with K~values ranging from 1 p~Mto 4 mM werereported (Krogsgaard-Larsen et al., 1987). A closerlook at the distribution of one of the GABA transport-ers (GATI) revealed a much more complicated pic-ture, suggesting the presence of more GABA transport-ers localized in specific parts of the brain (Kanner andBendahan, 1990). Moreover, different types of GABAtransporters could be present not only in different brainareas, but also in different stages during nervous sys-tem development (Cherubini et al., 1991). Therefore,it was not surprising that four pharmacologically dis-tinct GABA transporters were cloned from the mousebrain library (Liu et al., l993a). Homologous cDNAswere cloned from rat brain and other mammaliansources (Borden et al., 1992; Clark et al., 1992). Wetermed the transports GAT1 to GAT4, and we will usethis nomenclature in this review (mouse GAT3 = ratGAT2 and mouse GAT4 = rat GAT3 or GAT-B; seeBorden et al., 1992; Clark et al., 1992; Liu et al.,1 993a). The GABA uptake activities of the four trans-porters were measured in mRNA-injected Xenopu.s’ 00-
cytes (Liu et al., l993a). The K~,values for GABAuptake by the expressed GATI to GAT4 were 6, 79,18, and 0.8 ~M, respectively. GAT2 is also an effectivebetaine transporter with a Km of ~-~200,uM. GAT3 andGAT4 also transport /9-alanine with Ku, values of 28and 99 tiM, respectively. They also transport taurinewith K
0 values of 99 ,uM and 1.4 mM. Whereas GATIand GAT4 gene expression is brain-specific, GAT2and GAT3 mRNAs were detected in other tissues, suchas liver and kidney, in which GAT3 mRNA was espe-cially abundant (López-Corcuera et al., 1992; Liu etal., 1993a; Jursky et al., 1994). The expression ofGAT3 mRNA in mouse brain is developmentally regu-lated, and its mRNA is abundant in neonatal brain, butnot in adult brain.
Pharmacological studies of the four GABA trans-porters revealed that /9-alanine inhibited GABA uptakeby the expressed GAT3 and GAT4, and to much lesserextent by GAT2 (López-Corcuera et al., 1992; Liu etal., !993a). Betaine inhibited GABA transport onlyby GAT2. GABA transport by GATI and GAT4 wasmore sensitive to diaminobutyric acid, guvacine, andnipecotic acid than that by GAT2 and GAT3. GAT2was the most sensitive to ph!oretin and quinidine,whereas GAT3 and GAT4 were the most sensitive to
guanidinopropionic acid, diaminopropionic acid, andhypotaurine. It is apparent that even though GAT2,GAT3, and GAT4 exhibit high sequence homology,they have different substrate and inhibitor specificities.Moreover, GATI and GAT4 have the highest affinityfor GABA and are more sensitive to nipecotic acid.
The subfamily of GABA transporters contains sev-eral transporters exhibiting high sequence homologyto each other (Liu et al., l993a). It includes the trans-porters GAT1 (Guastella et al., 1990; Liu et a!.,1992a), GAT2 (López-Corcuera et al., 1992), GAT3(Liu et al., 1993a), GAT4 (Liu et al., 1993a), creatine(Guimbal and Kilimann, 1993), and taurine (Liu etal., 1992b; Smith et al., l992h; Uchida et al., 1992).Although GAT2, GAT3, GAT4, and the taurine trans-porters exhibit >60% identity in their amino acid se-quences, they are only =50% identical to GAT1 and=40% identical to members of the other subfamiliesof Na~/C1 transporters. Despite the high sequenceidentity between the GABA and taurine transporters,the latter is not capable of GABA transport and GABAdoes not inhibit its taurine uptake activity (Liu et al.,l992b; Uchida et al., 1992). In contrast, the taurinetransporter is also an efficient /9-alanine transporter.Xenopus oocytes injected with mRNA of the clonedtaurine transporter expressed uptake activities with Kmvalues of 4.5 p.M for taurine and 56 p.M for /9-alanine(Liu et al., 1992b). This value is close to the reportedaffinity for /9-alanine uptake into neurons or astrocytesand synaptosomal preparations (Kanner and Benda-han, 1990). Competition between GABA and /3-ala-nine, as well as between taurine and /9-alanine, wasreported. It was concluded that one of the GABA up-take systems, presumably the glial transporter, trans-ports /9-alanine (Kanner and Bendahan, 1990). Thecloned high-affinity GABA transporter (GAT1) doesnot transport /9-alanine, but GAT2 transports /9-alanineat very low rates (Borden et a!., 1992; Liu et al.,1993a). Therefore, it can be concluded that /3-alaninetransport is shared by both GABA and taurine transportsystems.
The precise function of each individual GABAtransporter is not entirely clear. It is apparent thatGAT1 and GAT4 are the principal players not only inthe termination of GABA neurotransmission, but alsoin maintaining the homeostasis of GABA concentra-tions in the various brain parts. The principle of cou-pled transporters, one neuronal and one glial, that func-tion in the termination of GABA neurotransmissionalso holds for the termination of glycine neurotrans-mission (Jursky and Nelson, !996a,h). The expressionof the various transporters during the development ofthe various brain parts is modulated to fulfill the neuro-logical function required for each developmental stage.For example, it was demonstrated that GABA acts asan excitatory transmitter in the early postnatal stage ofthe developing hippocampus (Ben-An et al., 1997).At about postnatal day 5, the GABA switches to itsadult role as an inhibitory neurotransmitter. Although
I. Neurochem., Vol. 71, No. 5, 1998
Na k/Cl NEUROTRANSMITTER TRANSPORTERS 1791
this phenomenon was attributed to functional changesin the GABAA, N-methyl-D-aspartic acid (NMDA),and a-amino-3-hydroxy-5-methylisoxazole-4-propio-nate (AMPA) ionotropic receptors, it may be meaning-ful for the process that at postnatal day 0, GABA trans-porters are almost absent from the hippocampus, andat postnatal day 5 GAT! is present at relatively highamounts in this brain region (Jursky and Nelson,1996b). Therefore, the presence of specific transport-ers in the various brain parts is not only vital for thecorrect neurotransmission, but also for the proper de-velopment of the CNS.
Monoamine transportersThis subfamily of neurotransmitter transporters con-
tains the transporters of noradrenaline, serotonin, anddopamine (Rudnick and Clark, 1993). An importantbreakthrough in the study of these transporters wasmade by cloning a human eDNA encoding the nor-adrenaline transporter (Pacho!czyk et a!., 1991). Notonly did it give an exciting opportunity to study thetransporter in depth, but it also helped to define thefamily of Nat/Cl neurotransmitter transporters. Thesequence of this transporter was found to be very simi-lar to that of the GABA transporter, particularly in thetransmembrane regions. Expression of the noradrena-line transporter eDNA in HeLa cells allowed noradren-aline uptake that was saturable, Na + -dependent, andsensitive to inhibition by tricyclic antidepressants, withspecificity identical to those of the native membrane-bound transporter (Pacholczyk et al., 1991). A!! thefundamental properties of noradrenaline uptake in thebrain were found to be coded by this single eDNAspecies. The transporter is widely distributed through-out the brain and is inhibited by cocaine, tricyclic anti-depressants, and amphetamines. Therefore, the cloningof eDNA encoding noradrenaline transporter circum-vented the necessity to purify the protein that hadproved difficult due to its low abundance in mamma-lian brain.
The sequence identity between the GABA and nor-adrenaline transporter led a number of workers to as-sume that other transporters might belong to this genefamily. By using a degenerative oligonucleotide probecorresponding to the region of greatest identity be-tween the GABA and noradrenaline transporters, acDNA clone encoding the dopamine transporter wasisolated from rat brain (Kilty et al., 1991; Shimada etal., 1991), and the cloning and sequencing of the bo-vine brain eDNA was reported subsequently (Usdin etal., 1991). Expression of the rat brain eDNA in bothCOS and HeLa cells was achieved, generating cocaine-sensitive dopamine transport activity. It was shown intransfeeted COS cells that two cocaine binding siteswere conferred from a single eDNA, suggesting thata single transporter species codes for both low- andhigh-affinity binding sites (Kilty et al., 1991; Shimadaet al., 1991; Usdin et al., 1991; Vandenbergh et al.,1992). The cloned human dopamine transporter shows
>95% amino acid identity to the rat protein and hasone less N-linked glycosylation site.
In both platelets and neurons, serotonin is trans-ported by a high-affinity uptake system that cotrans-ports Na~and Cl and countertransports K~(Rudnickand Clark, 1993). By using affinity chromatographywith serotonin or 6-fluorotryptamine as ligands, thetransporter was purified to an apparent homogeneity(Launay et al., 1992). However, here too, in the clon-ing attempts for this neurotransmitter transporter, theidentity between the GABA and noradrenaline trans-porters was used for designing degenerative oligonu-eleotide probes corresponding to the regions of greatestidentity. The attempt resulted in the isolation of aeDNA clone encoding the serotonin transporter in ratbrain (Blakely et al., 1991). Simultaneously, by usingexpression cloning, the serotonin transporter from ratbasophilic leukemia cells was also cloned (Hoffmanet al., 1991). Expression of each eDNA in nonneuronalcells generated Nat-dependent serotonin uptake abil-ity that was sensitive to antidepressants. Comparisonof the predicted amino acid sequences with those ofthe GABA and noradrenaline transporters showed 38%and 47% identity, respectively. However, several re-gions in the protein shared relatively high homologywith the noradrenaline and dopamine transporters, de-fining these three transporters as a tight subfamily inthe family of Na /Cl -- neurotransmitter transporters(Fig. 2). Monoamine transporters, as well as otherfamily members, are regulated by phosphorylation anddephosphorylation by protein kinases and phospha-tases. This subject was discussed recently in an excel-lent review and will not be covered in this review (seeBlakely et a!., 1997).
Because of its relevance to drug addiction and highlyadvanced pharmacology, the subfamily of monoaminetransporters attracted significant attention. Numerousexcellent reviews were written on this subject (Rud-nick and Clark, 1993; Blakely eta!., 1997). Therefore.it is not surprising that the first knockout mice lackinggenes encoding neurotransmitter transporters were ob-tained for members of this subfamily (Giros et al.,1996; Bengel et al., 1998). Disruption of the mousedopamine transporter gene resulted in a mouse pheno-type that is reminiscent of the wild-type mouse underthe influence of cocaine (Giros et al., 1996). Disrup-tion of the mouse serotonin transporter gene resultedonly in minor locomotor changes (Bengel et al., 1998;Sora et al., 1998). These observations stress the mainobstacle of using global knockout of genes and theunexpected effects of such intervention. It was ex-pected that the serotonin gene knockout would resultin a wide range of physiological and behavioral alter-ations, yet the mutant mice were able to cope withsuch a drastic intervention with no major handicap.The dopamine gene knockout resulted in one of themost definite phenotypes, yet it causes major adaptivechanges, such as decreases in neurotransmitter and re-
I. Nczoochenc, Vol. 71, No, 5, /998
1792 N. NELSON
ceptor levels (Bosse et a!., 1997; Gainetdinov et a!.,1997; Jaber et a!., 1997).
There are several obstacles to analyzing the effect ofgene interruption through conventional gene knockouttechniques. Unrestricted genetic deletion may lead tosevere developmental defects or premature death(Joyner, 1994). This approach can preclude analysisof postdevelopmental gene functions. Furthermore,such an approach may sometimes cause one to over-look many important effects or to obtain misleadingresults due to developmental circumvention of the ac-tivity that is missing due to gene knockout. If mutantmice complete development, interpretation of the ex-perimental results often runs into two types of diffi-culties. First, global gene knockout makes it difficultto attribute abnormal phenotypes to a particular systemin the CNS. Second, it is often difficult to exclude thepossibility that the abnormal phenotype observed inadult animals is caused directly by a developmentaldefect. Recently, a system of Cre/loxP was developedto circumvent these obstacles (Lakso et al., 1992;Tsien eta!., 1996). This systemis based on a phage P1 -
derived site-specific recombination in which the Crcrecombinase catalyzes recombination between two 34-bp IoxP recognition sequences (Lakso et al., 1992).The loxP recognition sequences are introduced in in-trons of genes encoding neurotransmitter transporters.Transgenic mice are generated by introducing theseengineered sequences into the original genes by homol-ogous recombination (Tsien et al., 1996). This genemanipulation should result in a modified gene that iscompetent to transcribe and translate intact neurotrans-mitter transporter. The homozygous mutant mice arethen crossed with homozygous transgenie mice thatexpress the Crc gene in specific brain regions. This isachieved by cloning the Crc gene after a promoter thatexpresses the downstream gene only in the desiredparts of the brain. Several types of mutant mice con-taining region- and cell-specific Crc expression arealready available, and it is likely that in a short periodof time the repertoire of mutant mice will increasetremendously. Therefore, it is anticipated that, in thefuture, mutant mice with neurotransmitter transporterscontaining loxP will be able to mate with several mu-tant mice containing the Crc expression system, andwill therefore allow gene deletion in many cell- andsystem-specific parts of the CNS.
The subfamily of Na ~IC1 -dependent amino acidtransporters
The subfamily of Nat/Cl-dependent amino acidtransporters contains genes encoding transporters forglycine and proline (see Fig. 2). The amino acid gly-cine is a classical inhibitory neurotransmitter localizedin the spinal cord, brainstem, and retina (Daly, 1990).Its effects are mediated by the glycine receptor, a Ii-gand-gated Cl - channel that is competitively antago-nized by strychnine (Grenning!oh et al., 1987). Gly-cine also modulates excitatory neurotransmission as an
obligatory coagonist with glutamate at NMDA-aeti-vated glutamate receptors (Johnson and Aseher, 1987;Kemp and Leeson, 1993). Although a specific prolinereceptor has not yet been identified, it was argued thatthis amino acid may function as a neurotransmitter(Fremeau et al., 1992). It was shown that radiolabeledL-proline is released from brain slices and synapto-somes in a Ca2~-dependent manner following K~-induced depolarization (Nickolson, 1982). These andother experimental findings supported a synaptic rolefor L-proline in specific excitatory pathways in thebrain. Proline may also be involved in sustaining eyto-plasmic glutamate levels and thus modulate gluta-matergie neurotransmission (Miller et al., 1997).
By using low-stringency screening and PCR,cDNAs encoding the amino acid transporters of gly-cine (GLYT1) and proline were cloned and sequenced(Fremeau et al., 1992; Guastella et al., 1992; Liu eta!., l992c; Smith et a!., 1992a). During the course ofcloning the first mouse glycine transporter (GLYTI),we obtained a clone that was different from the one thatwas published previously (Liu et al., 1992c, 1993b) inits 5’ untranslated region and the first 15 amino acids.From this point to the 3’ end of the eDNA, the twoclones were identical. The kinetics and pharmacologyof the two expressed glycine transporters namedGLYT1a and GLYT1b were indistinguishable. Geno-mie cloning and sequencing revealedthat GLYT1a andGLYTIb are coded by a single gene (Liu et a!.,1993b). The first exon encodes the 5’ untranslatedsequence and the N-terminal 10 amino acids ofGLYTIa. The second exon encodes the 5’ untranslatedsequence and the N-terminal 15 amino acids ofGLYTIb. Therefore, the GLYTI variants may be gen-erated by alternative splicing. It was also pointed outthat GLYT1b may be transcribed from a promoter thatis localized in the intron between the first and secondexons. This gene organization and differential splicingmay also exist in the rat genome. The two reportedg!yeine transporters cloned from eDNA rat brain librar-ies are homologous to the mouse GLYTIa (Guastellaet al., 1992) and GLYT1b (Smith et al., 1992a), re-spectively. Recently, a third splicing variant of GLYT1(GLYT!e) was found in human brain (Kim et al.,1994).
Cloning and expression of eDNAs encoding GLYT1left several unanswered questions regarding glycinereuptake in the CNS (Guastella et a!., 1992; Liu etal., !992c; Smith et al., 1992a). The distribution ofGLYTI mRNA did not correlate with the glycine re-ceptor, and its predicted amino acid sequence was con-siderably shorter than expected from the determinedmolecular weight of the purified brainstem transporter(López-Corcuera et al., 1991). Therefore, we soughta novel glycine transporter that would be associatedwith the glycine receptor in the glycinergic neurons.A novel glycine transporter (GLYT2) was cloned fromarat brain eDNA library (Liu et a!., !993b). GLYT2 is=48% and 50% homologous to the previously cloned
1. Neurochcm., Vol. 71, No. 5, 1998
Na k/Cl- NEUROTRANSMITTER TRANSPORTERS 1793
mouse glycine transporter (GLYT1) and rat prolinetransporter (PROT), respectively. GLYT2 differs fromGLYTI in molecular structure, tissue specificity, andpharmacological properties. The eDNA of GLYT2 en-codes for 799 amino acid residues with an extendedN-terminal peptide containing 200 amino acids beforethe first transmembrane domain. The calculated molec-ular weight of GLYT2 was in agreement with that ofthe isolated glycine transporter from spinal cord (Ló-pez-Coreuera et al., 1991). Potential phosphorylationsites for protein kinase C, cyclic AMP-dependent ki-nase, and calmodulin-dependent kinase were identifiedin the extended N-terminal region. Xenopus ooeytesinjected with GLYT2 eRNA transport glycine had aKm of 17 pM, and the uptake of glycine is resistant toinhibition by sareosine. These data are also in linewith those obtained with the purified and reconstitutedglycine transporter (López-Coreuera et al., 1991). Theexperimental data suggested that GLYT2 may play amajor role in the termination of the inhibitory effectof glyeine in the brainstem and spinal cord of verte-brates. This assumption was substantiated by in situhybridization and immunocytoehemical studies (Boro-wsky et al., 1993; Jursky et al., 1994; Jursky and Nel-son, 1995; Luque et al., 1995). The distribution ofGLYT2 correlated very well with that of the strych-nine-sensitive glycine receptors in most CNS regions,except cerebellum (Jursky and Nelson, 1995; Luqueet a!., 1995). It was concluded that GLYT2 is theprincipal transporter that mediates the termination ofglycine neurotransmission.
The mRNAs of GLYTI and the proline transporterwere more abundant near excitatory pathways in theCNS, suggesting a neuromodulatory role for prolineand glycine in glutamatergie cells. It was proposed thatone of the functions of both transporters is to removeglycine and proline from the environment of glutamatereceptors (Liu et al., 1992c; Smith et al., 1992a). Thispossibility has important pharmacological implicationsthat will be discussed later.
The subfamily of amino acid neurotransmitter trans-porters contains only glycine and proline transporters.The extensive screening of brain eDNA libraries inseveral laboratories left almost no stones unturned. Yetlow abundant transporters are emerging (N. Nelson etal., unpublished observations).
The orphan (NTT4) subfamily of transporterswith no known substrate
This subfamily of transporters consists of four geneproducts that appear to encode transporters that differstructurally from the three other subfamilies. They dis-play large second and fourth extracellular domainswith sites for N-linked glycosylation in both large do-mains (Fig. 2). Substrates for all of these gene prod-ucts, which are highly expressed in the brain or kidney,have never been identified. During the cloning of sev-eral neurotransmitter transporters, we encountered aeDNA clone structurally related to the other members
of the family of Na + /Cl -dependent transporters (Liuet al., 1993c). We denoted this alleged transporter asNTT4 and observed that it exhibits =35% identity withthe other members of this family of transporters. Twoother closely related members of this subfamily werecloned and their predicted amino acid sequences are=65% identical withNTT4 (UhI et al., 1992;El Mesti-kawy et a!., 1994). They have been termed orphantransporters because their substrates have not yet beenidentified. Recently, eDNA encoding another memberof this subfamily was cloned from a kidney cell line(Wasserman et al., 1994). It expresses quite specifi-cally in the kidney, and once again the substrate forthis transporter could not be identified. Northern blotanalysis of peripheral and neural tissues demonstratedthat the expression of NTT4 mRNA is restricted essen-tially to the nervous system (Liu et a!., 1993c). Insitu hybridization demonstrated a broad, but discrete,localization of the NTT4 message in the CNS, particu-larly in the cerebellum (Purkinje and granule cell lay-ers), hippoeampus (pyramidal and granular cell lay-ers), and thalamus and throughout the cerebral cortex(El Mestikawy et al., 1994; Luque et al., 1996). Thisdistribution parallels that of the neurotransmitters glu-tamate and aspartate. One noticeable exception to thedistribution of mRNA for NTT4 and these excitatoryneurotransmitters is the cerebellar Purkinje cell layer,in which GABAergic neurons are localized. NTT4 wasfound to codistribute widely with the subset of specificvariants of the NMDA receptor subunit I, namely,NMDARI I —4b. Polyclonal antibodies obtained byimmunizing guinea pigs with fusion proteins were usedto identify and characterize NTT4 in the rat CNS.Overall, the immunochemical localization of NTT4correlates with the distribution of NTT4 mRNA. Forexample, although prominent hybridization signals arefound in the CA3 hippocampal neurons and Purkinjecell layer, correspondingly high levels of immunostain-ing are present in the mossy fibers of the hippocampalregion, as well as in the cerebellar molecular layer,into which bothcell populations extend their respectiveprocesses.
Identification of the substrate for NTT4 is difficult,because in situ hybridization and immunohistoehemis-try indicate that this transporter is synthesized by phe-notypically different neuronal populations, e.g., gluta-minergie, GABAergic, cateeholaminergic (locus ceru-!eus), histaminergie (mammillary nuclei), andserotoninergie neurons (raphe nuclei). The failure toobtain any uptake with >100 different substancestested in our and other laboratories may be the resultof special properties of this transporter: (a) this sub-family of transporters may require a /3 subunit thatmight be necessary for their uptake activity; (b) un-known or unappreciated cofactors or heterooligomersmay be involved in the functional expression of thistransporter; (c) they may require other facilitative ionsbesides Na ~, Cl -, and K ~ (d) they may be expressedin the vacuolar system; and (e) this subfamily of trans-
J. Neurochem., Vol. 71, No.5, 1998
1794 N. NELSON
porters may transport an unknown neurotransmitter.As the NTT4 transporter appears to be present in CNSregions with specific glutaminergie innervation, it islikely to be involved in excitatory pathways. It is note-worthy that the glutamate transporters, discovered sofar, fail to explain the termination of glutamatergicneurotransmission (Luque et al., 1996).
Bacterial transporters of unknown functionGenomic cloning of genes encoding neurotransmit-
ter transporters and a search for sequences withoutknown function in the GenBank revealed that insectsand C. elegans contain genes encoding potential neuro-transmitter transporters (Liu et al., l992a). Recently,we identified and cloned a bacterial gene with rela-tively high homology to neurotransmitter transporters(S. Mandiyan et al., unpublished observations). Thisgene (STRYP!) is present in a potential tryptophanoperon of Svmbiobacterium thermophilum that con-tains the gene encoding the enzyme tryptophanase(Hirahara et al., 1992). In many other bacteria, a simi-lar operon exists and includes a gene for tryptophanasefollowed by the gene encoding a tryptophan transporter(Deeley and Yanofsky, 1981; Kamath and Yanofsky,1992). However, the tryptophan transporter in thesebacteria has no homology whatsoever to neurotrans-mitter transporters. Therefore, we suspect that, in thosethermophilie bacteria, the gene that follows the trypto-phanase and is homologous to neuronal transportersfunctions as tryptophan permease. Recently, the entiregenomic sequences of Haemophilus influenzae andMethanococcus jannaschii were determined (Fleisch-mann et al., 1995; Bult et al., 1996). Each genomecontains a single gene homologous to the mammalianfamily of neurotransmitter transporters. It is highlylikely that, in light of the high sequence identity be-tween the bacterial protein and neurotransmitter trans-porters, their mechanism of action may be similar, ifnot identical.
The complete sequence of the yeast genome and theextensive sequencing of plant genomes revealed nogenes with homology to neurotransmitter transporters.This raises an interesting evolutionary question,namely, why have fungi and plants avoided the use ofthese transporters? One possible cause is that Na~’/Cl — transporters strictly require Na + gradients for theiraction and they could not be adjusted to protonmotiveforce, which is the main driving force for transportin fungi and plants. Bacterial proteins are usually notg!ycosylated, and indeed the bacterial homologues ofneurotransmitter transporters are missing the large ex-ternal loop (no. 2) (N. Nelson et a!., unpublished ob-servations). This configuration may provide importantstructural information not only for the bacterial, butalso for the mammalian, transporters. Hydropathy plotssuggest 12 transmembrane helices for the membranetopography of these transporters. Figure 1 depicts theproposed arrangement of the transporters in the bacte-rial membrane. Like the mammalian ones, both the N-
and C-termini may be situated at the cytoplasmic faceof the membrane. Figure 1 also indicates the chargebalances of the proposed internal and external loops.Like most other membrane proteins, the rule of netpositive charges in internal loops is maintained. It isinteresting to note that the third transmembrane helixcontains a long stretch of uncharged amino acids, 24in Symbiobacterium thermophilum and 38 in Methano-coccus jannaschii. However, it is not apparent howthese stretches of amino acids can fold into two trans-membrane helices. Thus, the original proposal of themembrane topography for the Nat/Cl ‘ transporters issupported by the sequence analysis of the bacterialtransporters. The function of these transporters is notknown, but their existence opens up new and interest-ing research avenues.
TRANSPORTERS AS PHARMACOLOGICALTARGETS
In the normal brain, transporters maintain low intra-synaptic and extracellular neurotransmitter levels, andby this action they (a) regulate synaptic efficacy, (b)ensure synaptic fidelity, and (e) reduce the potentialneurotoxicity of excessive excitatory neurotransmit-ters. Deviation from a well controlled state might causeneurological disease, and modulation of an aberrantactivity may remedy malfunctioning systems.
Many of the neurotransmitter transporters have beenimplicated as important sites for drug action. The func-tion of transporters in terminating the synaptic activi-ties of released neurotransmitters makes them an im-portant target for several drugs. It was demonstratedthat transporters may be involved in the action of nu-merous drugs and specific toxins that affect the nervoussystem (Snyder and D’Amato, 1986; Bergman et al.,1989; Kuhar et al., 1991). Monoamine transporters,for example, are targeted by major classes of antide-pressants, psychostimulants, and antihypertensivedrugs (Blakely et al., 1997). They are also necessaryfor the actions of several classes of neurotoxins thatselectively poison specific neuronal targets (Snyderand D’Amato, 1986). As a result of these activities, themembers of these four gene subfamilies have becomecandidate genes for involvement in several neuropsy-chiatrie disorders. The subfamily of monoamine trans-porters is a target for important addictive drugs, suchas cocaine. Several lines of evidence suggest that co-caine’s rewarding and reinforcing actions in the brainare due to activities of dopamine transporters. The drugalso possesses a reasonable affinity to the other Na + -
and Cl -dependent plasma membrane monoaminetransporters, as well as to some Na~channels (Berg-man et al., 1989; Kuhar et al., 1991). Effective antide-pressants share the ability to inhibit the serotonin trans-porter (Blakely et al., 1997). Transporter-mediated ac-cumulation and/or sequestration of neurotoxins playsa key role in several current experimental models ofthe specific dopaminergic neurodegenerative processes
J. Neurochem., Vol. 71, No.5, 1998
Na k/Cl - NEUROTRANSMITTER TRANSPORTERS 1795
found in Parkinson’s disease (Snyder and D’Amato,1986). For example, accumulation of MPP~can berendered toxic in cells expressing the dopamine trans-porter (Kitayama et a!., 1992). Conversely, cells be-come resistant to this poison if they overexpress thevesicular monoamine transporter. The presence of hightransporter levels results in the accumulation of MPP +
into their vacuolar system (Schuldiner, 1994). Thereis a good fit between the onset of Parkinson’s diseaseand the expression levels of transporter mRNA in ratand human brains (Uhl et a!., 1994). Moreover, the“designer” drug of the I 990s, Prozac, interacts withthe serotonin transporter that is a member of the familyof Na f/Cl ‘ neurotransmitter transporters (Vanden-bergh et al., 1992; Ramamoorthy et al., 1993).
Human studies indicated that GABA transporterfunction is reduced in epileptic hippocampi (Duringet al., 1995; Williamson et al., 1995). This may resultfrom a reversed GABA transport due to membranedepolarization. This action reduces the risk of hyperac-tive glutamatergic neurons that, when they pass a cer-tain threshold, cause seizures. Therefore, there is muchinterest in agents that block GABA uptake. Theseagents may be used as anticonvulsive drugs by actingas selective inhibitors of either glial or neuronal uptake.In theory, GABA uptake inhibitors would be“cleaner” therapeutic agents than GABA receptor ag-onists, because inhibitors of GABA uptake would bepresent only when GABA was released physiologi-cally. The way by which certain drugs exhibit uniqueselectivity is not apparent. There are many receptorsubtypes for every neurotransmitter, most of whichhave a distinctive distribution. If the duration of neuro-transmitter action of all of them is regulated by reup-take, a given drug would be expected to affect all ofthem. The action of Prozac, for example, speaksagainst such a sweeping effect. Prozac is quite selectivein blocking serotonin transporters. It is now widelyprescribed not only for depression, but also for lessserious prevalent behavioral symptoms. The questionwhy one blockerof a given transport system is a betterdrug than another with similar inhibitory properties isnot easy to answer. The selectivity of reuptake blockersmay result from the different accessibility of differentbrain parts to the drug. Therefore, the drug augmentsthe action of those brain synapses where the neuro-transmitter is already being released. Thus, the elevatedneurotransmitter concentration in the various synapsesmay affect differently various brain cells or develop-mental stages (Ben-An et a!., 1997). The therapeuticeffect of a transporter blocker may take weeks to de-velop. It is logical to assume that their primary actionin prolonging neurotransmitter effects is to set up aseries of molecular changes in the brain that changethe homeostasis of several metabolic processes. Thispossibility is supported by recent studies with knockoutmice lacking the dopamine transporter (Giros et a!.,1996). The lack of dopamine transporter affected notonly the duration in which the neurotransmitter was
present in the synapse, but also its biosynthesis andseveral other metabolic processes. Understanding thesechanges should help in the development of rationalnew drugs that are based on the modulation of neuro-transmitter transporter activity.
Due to the special structure of glycine, which is thesmallest amino acid and has no L- and D-isomers, thereis no pharmacology for glycine transporters. This phar-macology can now be developed using expression sys-tems for the various glycine transporters. This searchshould consider the finding that even alanine, whichis structurally related to glycine, fails to inhibit glycinetransport, as well as the observation that sareosine in-hibits GLYTI isoforms, but not GLYT2 (Liu et al.,l993b). These properties suggest a compact structureof the substrate binding site that would not allow evenslight alteration of the glycine structure. Nevertheless,a hydrophobic pocket may exist near the binding siteof glycine transporters. This site may enable the designof inhibitory drugs for glycine uptake. An alternativeavenue may reveal an allosteric site for antagonists ofglycine uptake. The importance of such an antagonistis apparent from the important function of glycine re-ceptors in the CNS and the possibility of modulatingthe glutamate (NMDA) receptors. The glycine site ofNMDA receptors has generated enormous interestsince it was first described 10 years ago (Johnson andAseher, 1987). Potent pharmacological agents that in-teract selectively with this site were discovered, andthey include both agonists and antagonists that modu-late the activity of the receptor. Glycine is a coagonistof this receptor, and the receptor is inactive in theabsence of glycine (Kleckner and Dingledine, 1988).It was demonstrated that both the glutamate and theglycine recognition sites reside on the same protein(Moriyoshi et al., 1991). The affinity of the NMDAreceptor to glycine is in the low mieromolar concentra-tion. The glycine concentration in brain fluids is ordersof magnitude higher. This raised some doubts aboutthe possible physiological role of glycine. because pre-sumably the glycine binding site is saturated at alltimes. In this respect, glycine transporters may play acrucial role in the modulation of the synaptic responsemediated by NMDA receptors (Liu et al., !992c;Smith et al., 1992a). The presence of glycine trans-porters in the genera! area of the NMDA receptorsmay lower the glycine concentration in these areas.Moreover, placement of glycine transporters in the vi-cinity of an NMDA receptor may completely alter itsresponse to glutamate. There is evidence suggestingthat the glycine site on NMDA receptors is not satu-rated (Wood and Sidhu, 1986). The modulatory roleof the glycine site makes it an attractive target fortherapeutic anticonvulsant and neuroproteetive agents(Priestley et al., 1990; Singh et a!., 1990). It is appar-ent that modulating the activity of the glycine trans-porters in the vicinity of the NMDA receptor mayachieve similar pharmacological effects with less side
J. Neurochein., Vol. 71, No. 5, 1998
1796 N. NELSON
effects than the use of agonists and antagonists thatinteract with the glycine site in the NMDA receptor.
Recently, it was suggested that the primary distur-bance in glutamatergie function may be etiological inschizophrenia (Tsai et al., 1995). Dopaminergicagents, such as amphetamines, induce a psychotic stateby elevating dopamine in the synapses. In contrast,pheneyclidine (PCP) induces a schizophrenia-like psy-chotic state by blocking the NMDA receptor (Javitt etal., 1994, 1996). Experimental evidence suggests thatelevation of glycine concentrations in the brain re-verses PCP-indueed behaviors in rodents (Toth andLajtha, 1986). Moreover, a significant improvementin negative symptoms was observed when a g!yeinetransporter inhibitor glycyldodecylamide was given toPCP-treated animals (Javitt et al., 1997). These initialfindings stress the importance of improving the phar-macology of glycine transporters. The availability ofeDNAs encoding glycine transporters may facilitatethe discovery of efficient agonists and antagonists oftheir activity.
The few examples described above are only part ofa long list of the potential actions of neurotransmittertransporters in regulating the activity of neurotransmis-sion. New drugs can now be synthesized and assayedin systems that express specific neurotransmitter trans-porters. By following the activity of the new drugs,one can predict their possible effects on certain neuro-transmission pathways in the brain. It is apparent thatexperimental animal models in which specific neuro-transmitter transporters are knocked out or geneticallyaltered will be very helpful in assessing the resultsobtained with the newly developed drugs.
MECHANISM OF TRANSPORT
At the end of his life, King Solomon admitted thatsome things were unknown to him, such as the wayof a man with a woman (in spite of being married toa thousand wives). There are about a thousand trans-porters, yet their mechanism of action is still anenigma. The lactose permease of E. coli is the closestto being understood (Kaback et al., 1997). The knownpart of the mechanism of lactose transport includes anelusive substrate binding site and a conformational-driven twisting helix. To reach a similar position inNa + /Cl - neurotransmitter transporter mechanisms,much more work is still needed. There are a fewcardinal questions that have to be answered. Does the12 transmembrane helix structure indicate a similarmechanism of action in all such transporters? Is therea general transporter fold? Are twisting helices a com-mon denominator in all transporters? How is the dee-troehemical gradient coupled with the vectorial trans-port of substrates? To answer these questions, distinctsteps in substrate transport have to be addressed: (a)What is the nature of the substrate-binding site andwhich part(s) of the transporters are involved in sub-strate binding? (b) Where do the different inhibitors
FIG. 3. Schematic representation of the minimal steps in theuptake cycle and the possibility of channel formation with andwithout the involvement of additional protein. The scheme de-scribes the kinetic coupling and binding order of GABA trans-porters. Serotonin transport is stimulated by internal K but theion is not absolutely required. The thermodynamic coupling fordopamine transport is not clear. Two Na’~ions may be involved,but one of them may not be transported across the membrane.5, substrate.
bind and how much do their binding sites overlap withthe substrate binding site? (e) Which are the aminoacid residues involved in the Na~and C1 binding?(d) How does energy, expressed as electroehemicalgradient of Na drive the substrate transport acrossthe membrane? (e) How does the translocation ma-chinery operate? (f) What is the significance of thetransporter channel activities?
Having stated the main open questions, we will ex-amine some of the available information that is perti-nent to answering them. Like the other transporters,the catalytic cycle of neurotransmitter transporters canbe described by a conventional representation of sub-strate- and ion-induced conformational changes (Rud-nick, 1997). Figure 3 depicts a schematic representa-tion of the minimal requirements for the transport ofneurotransmitters across the membrane. The drivingforce for transport is a Na~eleetroehemical gradientthat is used through cotransport of 2Na~,1C1, and
J. Neurochem., Vol. 71, No. 5, /998
Na k/Cl - NEUROTRANSMITTER TRANSPORTERS 1797
!S (substrate). These ions bind first, and the conforma-tional change induced by this binding enables the sub-strate binding. Following the transport step, the Na +
ions are released first, and then the substrate and C1are also released. The transporter is rendered compe-tent for the next cycle by returning to the originalconformation spontaneously, or through facilitation byK” or H’ gradients (see Cao et al., 1997), or bya membrane potential (Au’). Each transporter has avariation on the above theme. The serotonin transporterrepresents one of the extreme deviations, because ithas an eleetroneutral transport that is dependent onan outward K~gradient (Rudnick, 1997). We are alllooking forward to the surprises that the orphan trans-porters have in store for us.
Substrate binding and trans!ocationThe sequence homology of all the neurotransmitter
transporters in this family suggests a common mecha-nism for transport. However, even in the subfamily ofmonoamine transporters, significant differences in themode of ion transport and stoichiometry were ob-served. Dopamine may be cotransported with 2Naand lCl, noradrenaline is cotransported with lNaand 1C1, and serotonin is also cotransported withINa’ and lCl, but its transport is also dependenton the presence of K in the cytoplasmie side of thetransporter (Rudniek and Nelson, 1978; Gu et al.,1996; Rudniek, 1997). Strategies based on proteinmodification have provided some information on someprobable residues involved in transport activity. Unlikeconventional methods, such as analyses of site-directedor deletion mutants, chimeras can provide phenotypesthat allow the delineation of functions associated withparticular protein domains. Chimerie proteins havebeen constructed from domains of the dopamine andnoradrenaline transporters (Buck and Amara, 1994,1995; Giros et al., 1994) and human or rat serotonintransporter (Barker et al., 1994). The ehimerie proteinswere constructed using two experimental strategies:(a) exchanging domains by exploiting unique restric-tion sites that are already present in the sequence orengineered into the sequence by PCR; and (b) using anin vivo method that generates chimeras within bacteriatransformed with linear plasmid DNA containing a sin-gle copy of each parental eDNA in a tail-to-head con-figuration (Buck and Amara, 1994, 1995; Moor andBlakely, 1994; Barker and Blakely, 1998). The resultssuggest that a region spanning TM 5—8 in the nor-adrenaline and dopamine transporters is important forconferring inhibitor sensitivity to a variety of inhibi-tors, including imipramine (Buck and Amara, 1995).The domain in the noradrenaline transporter spanningTM 1 —3 also contributes to the affinity for antidepres-sants. Determinants responsible for the substrates’ af-finity were localized in the region spanning TM 1 —3,as well as TM 11 and 12 (Buck and Amara, 1994).In another study, it was reported that the region of theTM 8—12 has a predominant influence on the affinity
and selectivity for different substrates (Giros et al.,1994). Similarly, ehimeras constructed between the ratand human serotonin transporter indicated that deter-minants in the region of TM 11 and 12 are responsiblefor the species-specific differences in the binding ofthe imipramine and amphetamine to the transporter(Barker et al., 1994). Mutations in F586, in TM 12of the human serotonin transporter, revealed this resi-due as responsible for high-affinity interaction of triey-die antidepressants with the transporter (Barker andBlakely, 1996). Chimeric serotonin transporters fromhuman and Drosophila were used recently for elucidat-ing their antagonist binding sites (Barker et al., 1998).Domains in TM 1 and 2 were identified as being im-portant for mazindol and citalopram binding. Eightamino acids that differ between the two transporterswere substituted. A single mutation (Y95F) in TM I ofthe human serotonin transporter shifted the antagonistbinding to the Drosophila transporter (Barker et a!.,1998).
The observations using ehimerie transporters are inline with conclusions reached from site-directed muta-genesis experiments. Substitution of D79 in TM 1 ofthe dopamine transporter resulted in some alteration tothe binding of a cocaine analogue and a reduction ofthe apparent affinity to dopamine (Kitayama et at.,1992). Recently, it was shown that the correspondentamino acid residue in TM I of the noradrenaline trans-porter is critical for its function (Barker and Blakely,1998). Further studies with mutations in TM 7 andTM 11 indicated that both domains contain determi-nants that influence substrate translocation and the af-finity for MPP~,respectively (Kitayama eta!., 1993).The large external 1oop between TM 3 and TM 4 isunlikely to have a direct role in the transport mecha-nism. However, TM 3 and the beginning of the largeloop are quite likely to be involved in transport. Thethree cysteines in the external loops of the rat serotonintransporter were substituted with alanine or serine andthe mutants were analyzed for sensitivity to methane-thiosulfonate reagents (Chen et al., !997a,b; Stephanet al., 1997). Although the mutant CIO9A located be-tween TM I and 2 had no effect on transport activity,substitution of C200 or C209 inflicted severe effectson the expression and activity of the transporter. Theresults support the possibility that C200 and C209 maybe linked by a disulfide bond in the second externalloop of the serotonin transporter and that C 109 is situ-ated on the external face of the transporter. Assumingthe original model proposed for the membrane archi-tecture, TM 3 is unique in its length (Fig. I). It con-tains from 24 up to 37 uncharged amino acids. Thisstructure may be important for the transport activityof the family members. One of the possible mecha-nisms for solute transport across membranes is a slid-ing a-helix that changes its position in the membraneby substrate binding and/or eleetroehemica! potential(Racker, 1976). If such a mechanism exists, TM 3 isa very good candidate to fulfil! this function. TM 3
J, Neurochem., Vol. 7/, No. 5, /998
1798 N. NELSON
contains a tyrosine residue that is conserved in all thefamily members, including the bacterial transporters.Recently, it was shown that this tyrosine residue(Yl40) is critical for GABA recognition and transportby GATI (Bismuth eta!., 1997). Even substitution tothe other aromatic amino acids, phenylalanine(Y!40F) and tryptophan (Y14OW), results in com-pletely inactive transporters. Electrophysio!ogicalcharacterization reveals that both mutants exhibit theNa’’ -dependent transient currents associated with Na”binding, as well as the Cl -dependent lithium leakcurrents characteristic of GAT1. Both mutants are in-active in GABA-induced steady-state transport cur-rents. It was concluded that Y140 may be involved inthe ligand binding of the amino group of GABA andother neurotransmitters (Bismuth et al., 1997). Flee-trophysiological studies also indicate that the con-served residue W68 in the first putative TM I ofGATImay be involved in the binding of Na’’ and/or CL’(Mager et al., 1996).
The availability of four cDNAs encoding homolo-gous GABA transporters presented some opportunitiesto analyze the substrate binding sites of these transport-ers. The importance of short external loops for theactivity of GAT 1 was demonstrated by their high sensi-tivity to changes. Insertion of glycosylation sites intosome of the hydrophilic loops of GAT I resulted in itsinactivation, suggesting high sensitivity to amino acidreplacements in those sites (Bennett and Kanner,1997). This is in line with the observations showingthe importance of hydrophilie loops for transport andsubstrate specificity (Kanner et al., 1994; Tamura etal., 1995). Inspection of amino acid sequences re-vealed that the putative short external loops of GAT2,GAT3, and GAT4 are nearly identical, but are signifi-cantly different from that of GAT I. Site-directed muta-genesis was used to generate GATI transporters inwhich the amino acid sequences of these 1oops wereidentical to those of GAT3 (Tamura et al., 1995).When the loop structure of GAT3 was introduced intoloop 4 of GATI, the apparent K,0 was lowered to 2.0pM, which was similar to the Km value reported forthe high-affinity GABA transporter GAT4 (Liu et at.,1993a; Tamura et a!., 1995). This K11, value was aboutfourfold lower than that reported for GABA uptake byGAT1 (K11, = 8.7 pM). When the amino acid sequenceof 1oop 6 of GAT I was substituted with that of GAT2,an apparent K11, of 35 p.M was observed, which is nine-fold higher than the value for GATI. When aminoacids in loop 5 of GATI were substituted to give anamino acid sequence identical to that in the corre-sponding loop of GAT3, /3-alanine inhibited theGABA uptake activity of the mutated GATI trans-porter. To determine whether the three amino acids inloop 5 of GAT3 were important for /3-alanine binding,the three residues of GAT3 were replacedby the corre-sponding amino acids in GATI. This reverse mutationof GAT3 exhibited less sensitivity in its GABA uptaketo /3-alanine. This result suggested that three amino
acids in loop 5 participated in the /3-alanine bindingdomain of GAT3 (Tamura et al., 1995). The pharma-cology of the mutated transporters followed thechanges in the substrate binding specificity and affin-ity. These reciprocal effects suggest that the short ex-ternal loops 4, 5, and 6 take part in the substrate bind-ing of /3-alanine and GABA. It was suggested thatthese three external loops form a pocket on the trans-porter into which the substrate binds.
The mechanism of GABA transport was analyzedby tracer fluxes and eleetrophysiological approaches(Kanner, 1983; Mager et al., 1993). It was demon-strated that the transporter cotransports the neurotrans-mitter with Na’’ and Cl - in an electrogenic fashion.Although the transport is absolutely dependent on thepresence of Na’’, the Cl - dependency is not absolute.It is not clear whether this phenomenon is a result ofa loose coupling for Cl- ions or the replacement ofthe Cl by another anion, such as hydroxyl. The func-tion of the substrate in the electrogenicity of the trans-porter was analyzed by following transient currentsgenerated across the plasma membrane by expressingGAT1 in Xenopus oocytes (Kavanaugh et a!., 1992;Mager et al., 1993) or HeLa cells (Risso eta!., 1996).Although the transport of positive charge across themembrane was GAB A-dependent, the transient cur-rents generated by Na’ ions were independent of thesubstrate. This phenomenon probably reflects a confor-mationa! change of the transporter that takes place fol-lowing the binding of Na’’ to the protein. Mutationalanalysis of negatively charged amino acids in GAT!revealed that the conserved residue ElOl is critical forthe function of the transporter (Keshet et al., 1995).Even its replacement by aspartate left only =1% ofthe transport activity. It was concluded that ElOl iscritical for one or more of the obligatory conforma-tional changes during the transport cycle. The mecha-nism of substrate translocation is not clear, and it mayinvolve movements and twists of a-heliees or a novelmechanism of channel activity.
Transporters as ion channelsRecently, it became apparent that several transport-
ers are involved in additional activities, such as waterand ion transport, that may have significant effects onthe osmolarity and the membrane potential of certaincells and membranes (Mager et a!., 1994; Loo et a!.,1996; Panayotova-Heiermann et al., 1997; Sonders etal., 1997). This bulk transport of water and/or ions ischaracteristic of channels, and one of the main openquestions is whether the apparent channel activity isalso involved in the mechanism of substrate transport.This subject was discussed recently in several excellentand thoughtful reviews (DeFelice and Blakely, 1996;Lester et al., 1996; Sonders and Amara, 1996; Su etal., 1996). Several of the Na”/Cl neurotransmittertransporters were shown to have channel activity asso-ciated with their transport activity (Sonders et al.,1997; see Fig. 3). Two major implications are to be
.1. Nco’ochei,,., Vol. 71, No. 5, /998
Na‘/Cl - NEUROTRANSMITTER TRANSPORTERS 1799
drawn from these observations. It was argued thattransporters preceded channels as sensors of the exter-nal environment (Nelson, 1993; Rudnick, 1997). Ac-cordingly, it is logical to assume that channel activitywas evolved for sensing the changes in the compositionof the external milieu. The second implication is mech-anistic and may contribute to the understanding oftransport activity. From the work performed with sev-eral transporters, it appears that not all of them exhibitchannel activity. Even those transporters that wereshown to have channel activity manifested differentproperties of this extrinsic activity. The uncoupled ionflux is likely to result from events in which the trans-porter transiently operates as an ion channel insteadof like a carrier. The physiological and mechanisticsignificance of this phenomenon depends on the proba-bility and the duration of this event. The turnover ofthe carrier mode is =10/s, and the ion flux throughchannels is >10,000/s. Hence, the existence of a chan-nel activity at 0. I % of the time would have high mech-anistic and physiological implications. It was possibleto demonstrate that serotonin transporter channels areopen very rarely relative to the number of catalyticcycles, and it is likely that this uncoupled current isprobably not part of the transport process. However,it is important to note that even though serotonin trans-port is eleetroneutral, inward currents were recordedin ooeytes expressing the serotonin transporter (Mageret at., 1994). As the uncoupled current is influencedby substrate binding or transport, it is difficult to esti-mate the contribution of each of these processes tothe steady-state current during transport. Fluctuationanalysis with Drosophila serotonin transporter, ex-pressed in Xenopus oocytes, showed that =500 seroto-nm molecules are transported per channel opening(Galli et al., 1996, 1997). At the same time, at —2()mV, 10,000 electronic charges were translocated.This observation can be explained by assuming thatone channel opening takes place every 500 transportcycles or that 500 serotonin molecules and 10,000 ionspass through a common channel. Even though seroto-nm transporters from Drosophila and mammals arehighly homologous, the recorded channel events in thetwo transporters are notentirely compatible. The detec-tion of channel activities in transporter molecules arevery interesting, but more work is required to revealtheir significance for the transport mechanism.
Another unsettled issue is the possible involvementof additional protein(s) in the uncoupled ion transport.The results of site-directed mutagenesis (Lin et al.,1996) and the observed correlation between their ef-fects on transport and channel activities favor the pos-sibility that the neurotransmitter transporter gene prod-ucts by themselves can fulfill a dual function. Thisassumption is also supported by the detection of chan-nel activity in two different heterologous expressionsystems. However, there is no direct evidence to provethis point.
EPILOGUE
The field of neurotransmitter transporters has justentered a turning point. Most of the cDNAs encodingthese transporters have been cloned and expressed inheterologous systems to reveal the transport and phar-macological properties of individual transporters. Thefew that are missing are likely to be cloned in the nearfuture, and the function of the orphan transporters willbe solved sooner or later. A three-dimensional struc-ture of even one of the neurotransmitter transportersis highly desired. The main tools to achieve this goalare being developed. These include the expression ofthe bacterial homologue in proper bacterial expressionsystems and/or the expression of the mammalian trans-porters in yeast. Achieving these goals may help largeamounts of crystallizable transporter to be obtainedand may provide a system in which second-site sup-pressors of inactive transporters may be studied. Re-cent structural studies revealed a pattern of commonfolds for many proteins without sequence similarity.The membrane transporter fold is one of the most im-portant folds that is still completely unknown. Wehopethat this fold will be understood soon.
Acknowledgment: I thank Dr. Holger Lil! for con-structing the evolutionary tree. This work was supported bygrants from the Israel Science Foundation and the UnitedStates—Israel Binational Scientific Foundation.
REFERENCES
Amara S. and Kuhar M. (1993) Neurotransmitter transporters—recent progress. A,wu. Rev. Neurosci. 16, 73—93.
Barker E. L. and Blakely R. D. (1996) Identification of a singleamino acid. phenylalanine 586. that is responsible for high af-unity interactions of tricyclic antidepressants with the humanserotonin transporter. Mo!. Pharniacol. 50, 957—965.
Barker E. L. and Blakely R. D. (1998) Insight into structural deter-minants of neurotransrnitter transport using cross-species chi-meras: studies in the sei’otonin transporter. Methods En~vmo1.(in press).
Barker E. L., Kimmel H. L., and Blakely R. D. (1994) Chimerichuman and rat serotonin transporters reveal domains involvedin recognition of transporter ligands. Mo!. Pharmaco!. 46, 799—807.
Barker E. L.. Perlman M. A.. Adkins P. E.. Houlihan W. J., PristupaZ. B., Niznik H. B., and Blakely R. D. (1998) Antagonist bind-ing sites on thc serotonin transporter revealed by species-scan-ning mutagenesis. J. Biol. Chem. (in press).
Ben-An Y., Khazipov R.. Leinekugel X.. Caillard 0.. and GaiarsaJ. L. (1997) GABA
5. NMDA and AMPA receptors: a develop-mentally regulated ménage a trois. Trend.s Neurosci. 20, 523—529.
Bengel I)., Murphy D. L., Andrews A. M.. Wichems C. H.. Feltner
D.. Heils A., Mossner R., Westphal H.. and Lesch K. P. (1998)Altered brain serotonin homeostasis and locomotor insensitivityto 3,4-methylenedioxymethamphetamine C Ecstasy’) in scm-tonin transporter-deficient mice. Mo!. Pharmacol. 53, 649—655.
Bennett E. R. and Kanner B. 1. (1997) The membrane topologyof CAT-I, a (Nat + CV)-coupled gamma-aminobutyric acidtransporter from rat brain. I. Iliol. Chem. 272, 1203—1211).
Bergman J.. Madras B. K.. Johnson S. F., and Spealman R. D.(1989) Et’fects of cocaine and related drugs in nonhuman pri-mates. III. Self-administration by squirrel monkeys. J. Pharma-col. Exp. Ther. 251, 150—155.
.1. Ncu,’ochen,,. Vol. 71. No. 5, / 998
1800 N. NELSON
Bismuth Y., Kavanaugh M. P., and Kanner B. 1. (1997) Tyrosine140 of the gamma-aminobutyric acid transporter GAT- I playsa critical role in neurotransmitter recognition. J. Biol. Chem.272, 16096—16102.
Blakely R. D., Berson H. E., Fremeau R. T. Jr., Canon M. C., PeekM. M., Prince H. K., and Bradley C. C. (1991) Cloning andexpression of a functional serotonin transporter from rat brain.Nature 354, 66—69.
Blakely R. D., Ramamoorthy S., Qian Y., Schroeter S., and BradleyC. C. (1 997) Regulation of antidepressant-sensitive serotonintransporter, in Neurotransmitter Transporters—Structure,Function, and Regulation (Reith E. A., ed), pp. 29—72. Hu-mana Press, Totowa, New Jersey.
Borden L. A., Smith K. E., Hartig P. R., Branchek T. A., and Wein-shank R. L. (1992) Molecular heterogeneity ofthe y-aminobu-tyric acid (GABA) transport system. J. Biol. Chem. 267,21098—21104.
Borowsky B., Mezey E., and Hoffman B. J. (1993) Two glycinetransporter variants with distinct localization in the CNS andperipheral tissues are encoded by a common gene. Neuron 10,85 1—863.
Bosse R., Fumagalli F., Jaber M., Giros B., Gainetdinov R. R., Wet-sd W. C., Missale C., and Canon M. G. (1997) Anterior pitu-itary hypoplasia and dwarfism in mice lacking the dopaminetransporter. Neuron 19, 127—138.
Bowery N. G., Jones G. P., and Neal M. J. (1976) Selective inhibi-tion of neuronal GABA uptake by cis-1,3-aminocyclohexanecarboxylic acid. Nature 264, 281—284.
Bruss M., Hammermann R., Bnimijoin S., and Bonisch H. (1995)Antipeptide antibodies confirm the topology of the human nor-epinephnine transporter. J. Biol. Chem. 270, 9197—9201.
Buck K. J. and Amara S.C. (1994) Chimenic dopamine—norepi-nephnine transporters delineate structural domains influencingselectivity for catecholamines and I -methyl-4-phenylpynidin-ium. Proc. Nat!. Acad. Sc!. USA 91, 12584—12588.
Buck K. J. and Amara S. G. (1995) Structural domains of catechol-amine transporter chimeras involved in selective inhibition byantidepressants and psychomotor stimulants. Mol. Pharmacol.48, 1030—1037.
Bult C. J., White 0., Olsen G. J., Zhou L., Fleischmann R. D., SuttonG. C., Blake J. A., FitzGerald L. M., Clayton R. A., GocayneJ. D., Kerlavage A. R., Dougherty B. A., Tomb J.-F., AdamsM. D., Reich C. 1., Ovenbeek R., Kirkness E. F., Weidman J. F.,Fuhrmann J. L., Nguyen D., Utterback T. R., Kelley J. M., Pe-terson J. D., Sadow P. W., Hanna M. C., Cotton M. D., RobertsK. M., Hurst M. A., Kaine B. P., Borodovsky M., Klenk H-P.,Fraser C. M., Smith H. 0., Woese C. R., and Venter J. C.(1996) Complete genome sequence of the methanogenic Arch-aeon Methanococcus jannaschii. Science 273, 1058—1073.
Cao Y., Mager S., and Lester H. A. (1997) H~permeation and pHregulation at a mammalian serotonin transporter. J. Neurosci.17, 2257—2266.
Caspani T., Will A., Opekarová M., Sauer N., and Tanner W. (1994)Hexose/H * symporters in lower and higher plants. J. Exp. Biol.196, 483—491.
Chen J.-G., Liu-Chen S., and Rudnick G. (1997a) External cysteineresidues in the serotonin transporter. Biochemistry 36, 1479—1486.
Chen J.-G., Sachpatzidis A., and Rudnick G. (1997b) The thirdtransmembrane domain of the senotonin transporter containsresidues associated with substrate and cocaine binding. J. Biol.Chem. 272, 28321—28327.
Cherubini F., Gaiarsa J. L., and Ben-An Y. (1991) GABA: an excit-atory transmitter in early postnatal life. Trends Neurosci. 14,515—519.
Clark J. A., Deutch A. Y., Gallipoli P. Z., and Amara S. G. (1992)Functional expression and CNS distribution of a /3-alanine-sensitive neuronal GABA transporter. Neuron 9, 337—348.
Daly E. C. (1990) The biochemistry of glycinergic neurons, in Gly-cine Transmission (Ottersen 0. P. and Storm-Mathisen J., eds),pp. 25—66. John Wiley, New York.
Deeley M. C. and Yanofsky C. (1981) Nucleotide sequence of the
structural gene for tryptophanase of Escherichia coli K- 12. J.Bacteriol. 147, 787—796.
DeFelice L. J. and Blakely R. D. (1996) Pore models for transport-ers? Biophys. J. 70, 579—580.
During M. J., Ryder K. M., and Spencer D. D. (1995) HippocampalGABA transporter function in temporal-lobe epilepsy. Nature376, 174—177.
Eisenberg D. (1984) Three-dimensional structure of membrane andsurface proteins. Annu. Rev. Biochem. 53, 595—623.
El Mestikawy, S., Cmos B., PohI M., Hamon M., Kingsmone S. F.,Seldin M. F., and Canon M. C. (1994) Characterization of anatypical member of the Na’ /Cl -dependent transporter family:chromosomal localization and distribution of GABAergic andglutamatergic neurons in the rat brain. J. Neurochem. 62, 445—455.
Fleischmann R. D., Adams M. D., White 0., Clayton R. A., KirknessE. F., Kerlavage A. R., Bult C. J., Tomb J.-F., Dougherty B. A.,Mernick J. M., McKenney K., Sutton G., FitzHugh W., FieldsC. A., Gocayne J. D., Scott J. D., Shirley R., Liu L.-I., GlodekA., Kelley J. M., Weidman J. F., Phillips C. A., Spniggs T.,Hedblom E., Cotton M. D., Uttenback T. R., Hanna M. C., Ngu-yen D. T., Saudek D. M., Brandon R. C., Fine L. D., FnitchmanJ. L., Fuhrmann J. L., Geoghagen N. S. M., Gnehm C. L., Mc-Donald L. A., Small K. V., Fraser C. M., Smith H. 0., and Ven-ter J. C. (1995) Whole-genome random sequencing and assem-bly of Haemophiius influenzae RD. Science 269, 496—512.
Fremeau R. T. Jr., Canon M. G., and Blakely R. D. (1992) Molecularcloning and expression of a high affinity L-prOline transporterexpressed in putative glutamatergic pathways of rat brain. Neu-ron 8, 9 15—926.
Gainetdinov R. R., Fumagalli F., Jones S. R., and Caron M. C.(1997) Dopamine transporter is required for in vivo MPTPneunotoxicity: evidence from mice lacking the transporter. J.Neurochem. 69, 1322—1325.
Galli A., Blakely R. D., and DeFelice L. J. (1996) Norepinephrinetransporters have channel modes of conduction. Proc. Nat!.Acad. Sci. USA 93, 8671—8676.
Galli A., Petersen C. I., deBlaquiere M., Blakely R. 0., and DeFeliceL. J. (1997) Drosophila senotonin transporters have voltage-dependent uptake coupled to a serotonin-gated ion channel. J.Neurosci. 17, 3401—3411.
Cmos B., Wang Y. M., Suter S., McLeskey S. B., Pill C., and CaronM. C. (1994) Delineation of discrete domains for substrate,cocaine, and tnicyclic antidepressant interactions using chimenicdopamine—norepinephrine transporters. J. Biol. Chem. 269,15985— 15988.
Cmos B., Jaber M., Jones S. R., Wightman R. M., and Canon M. C.(1996) Hypenlocomotion and indifference to cocaine and am-phetamine in mice lacking the dopamine transporter. Nature379, 606—612.
Grenningloh C., Rienitz A., Schmitt B.. Methfessel C., Zensen M.,Beyneuthen K., Gundelfingen E. D., and Betz H. (1987) Thestrychnine-binding subunit of the glycine receptor shows ho-mology with nicotinic acetylcholine receptors. Nature 328,215—220.
Cu H. H., Wall S., and Rudnick C. (1996) Ion coupling stoichiome-try for the norepinephrine transporter in membrane vesiclesfrom stably transfected cells. J. Biol. Chem. 271, 6911—6916.
Guastella J., Nelson N., Nelson H., Czyzyk L., Keynan S., MiedelM. C., Davidson N., Lester H. A., and Kanner B. I. (1990)Cloning and expression of a rat brain GABA transporter. Sci-ence 249, 1303—1306.
Guastella J., Brecha N., Weigmann C., Lester H. A., and DavidsonN. (1992) Cloning, expression, and localization of a rat brainhigh-affinity glycine transporter. Proc. Nat!. Acad. Sc!. USA 89,7189—7193.
Guimbal C. and Kilimann M. W. (1993) A Na*~dependentcreatinetransporter in rabbit brain, muscle, heart, and kidney. J. Biol.Chem. 268, 8418—8421.
Hediger M. A., Coady M. J., Ikeda T. S., and Wright E. M. (1987)Expression cloning and eDNA sequencing of the Na~/gIucoseco-transporter. Nature 330, 379—381.
J. Neurochem., Vol. 71, No. 5, 1998
Na ~/Cl NEUROTRANSMJTTER TRANSPORTERS 1801
Himahana T., Suzuki S., Honinouchi S., and BeppuT. (1992) Cloning.nucleotide sequences, and ovenexpression in E,ccherichia coilof tandem copies of a tryptophanase gene in an obligately sym-biotic thermophile, Symhiobocterium thermophilum. App!. En-viron. Microbiol. 58, 2633—2642.
Hoffman B. J., Mezey F., and Bnownstein M. J. (1991) Cloning ofa semotonin transporter affected by antidepressants. Science 254,579—580.
Humphneys C. J., Wall S. C., and Rudnick C. (1994) Ligand bindingto the serotonin transporter: equilibria, kinetics, and ion depen-dence. Biochemistry 33, 9118—9125.
Ivensen L. L. and Johnston C. A. R. (1971) GABA uptake in ratcentral nervous system: comparison of uptake in slices andhomogenates and the effects of some inhibitors. J. Neurochem.18, 1939—1950.
Iversen L. L. and Kelly J. 5. (1975) Uptake and metabolism of y-aminobutynic acid by neurons and glial cells. Biochem. Pharma-(‘01. 24, 933—938.
Jaber M., Jones S., Cmos B., and Camon M. G. (1997) The dopaminetransporter: a crucial component regulating dopamine transmis-sion. Mov. Disord. 12, 629—633.
Javitt D. C., Zylbenman 1., Zukin S. R., Heresco-Levy U., and Lin-denmayen J. P. (1994) Amelioration of negative symptoms inschizophrenia by glycine. Am. J. Psychiatry 151, 1234—1236.
Javitt D. C., Steinschneider M., Schroeder C. F., and Anezzo J. C.(1996) Role of cortical N-methyl-o-aspartate receptors in audi-tory sensory memory and mismatch negativity generation: im-plications for schizophrenia. Proc. Nat!. Acad. Sc!. USA 93,11962—11967.
Javitt D.C., Senshen H., Hashim A., and Lajtha A. (1997) Reversalof phencyclidmne-inducedhyperactivity by glycine and the gly-cine uptake inhibitor glycyldodecylamide. Neuropsychophar-macology 17, 202—204.
Johnson J. W. and Aschen P. (1987) Glycine potentiates the NMDAresponse in cultured mouse brain neurons. Trend.v Pharmaco!.Sc!. 14, 20—25.
Joynen A. L. (1994) Gene targeting and development of the nervoussystem. Curr. Op/n. Neurohio!. 4, 37—42.
Junsky F. and Nelson N. (1995) Localization of glycine neurotrans-mitten transporter (CLYT2) reveals correlation with the distri-bution of glycine receptor. J. Neurochem. 64, 1026—1033.
Jansky F. and Nelson N. (1996a) Developmental expression of theglycine transporters GLYTI and GLYT2 in mouse brain. J.Neurochem. 67, 336—344.
Jursky F. and Nelson N. (1996b) Developmental expression ofGABA transporters GAT1 and CAT4 suggests involvement inbrain maturation. J. Neurochem. 67, 857—867.
Jursky F., Tamura S., Tamuna A., Mandiyan S., Nelson H.. andNelson N. (1994) Structure, function and brain localization ofneurotnansmitter transporters. J. Exp. Bioi. 196, 283—295.
Kaback H. R. (1992) The lactose penmease of Escherichia coli: aparadigm for membrane transport proteins. Biochim. Biophy.c.Acta 1101, 210—213.
Kaback H. R., Voss J., and Wu J. (1997) Helix packing in polytopicmembrane proteins: the lactose permease of Escherichia coli.Curr. Opin. Struct. Biol. 7, 537—542.
Kamath A. V. and Yanofsky C. (1992) Characterization of the tryp-tophanase openon of Proteus vuigoris. Cloning, nucleotide se-quence, amino acid homology, and in vitro synthesis of theleader peptide and regulatory analysis. J. Bioi. Chem. 267,19978— 19985.
Kanner B. I. (1983) Bioenergetics of neurotnansmitten transport.Biochim. Biophv.c. Acta 726, 293—316.
Kannen B. (1993) Glutamate transporters from brain—a novel neu-notnansmitten transporter family. FEBS Lett. 325, 95—99.
Kannen B. I. and Bendahan A. (1990) Two pharmacologically dis-tinct sodium- and chloride-coupled high-affinity y-aminohu-tyric acid transporters are present in plasma membrane vesiclesand reconstituted preparations from rat brain. Proc. Nat!. Acad.Sc!. USA 87, 2550—2554.
Kannen B. I. and Schuldinen 5. (1987) Mechanism of transport andstorage of neunotransmitters. Crit. Rev. Bioche,n. 22, 1—38.
Kannen B. I., Keynan S., and Radian R. (1989) Structural and func-tional studies on the sodium- and chloride-coupled gammaaminobutynic acid transporter: deglycosylation and limited pro-teolysis. Biochemistr 28, 3722—3728.
Kanner B. 1., Bendahan A., Pantanowitz S.. and Su H. (1994) Thenumber of amino acid residues in hydnophilic loops connectingtransmembnane domains of the CABA transporter CAT-I iscritical for its function. FEBS Lett. 356, 191— 194.
Kavanaugh M. P., Arriza J. L.. North R. A., and Amana S. C. (1992)Electrogenic uptake of gamma-aminohutyric acid by a clonedtransporter expressed in Xenopus oocytes. I. Bio!. Chem. 267,22007—22009.
Kemp J. A. and Leeson P. D. (1993) The glycine site of the NMDAreceptor—five years on. Trend,c Pharmaco!. Sci. 14, 20—25.
Keshet C. I., Bendahan A., Su H., Mager S.. Lester H. A., and Kan-ncr B. I. (1995) Clutamate-lOl is critical for the function ofthe sodium and chloride-coupled GABA transporter CAT-I.FEBS Lett. 371, 39—42.
Keynan S., Sub Y. S.. Kanner B. I., and Rudnick C. (1992) Expres-sion of a cloned y-aminohutynic acid transporter in mammaliancells. Biochemistry 31, 1974—1979.
Kilty J. F., Lonang D., and Amara S. C. (1991) Cloning and expres-sion of a cocaine-sensitive rat dopamine transporter. Science254, 578—579.
Kim K. M., Kingsmore S. F., Han H.. Yang-Feng T. L., Codinot N.,Seldin M. F., Canon M. C., and Cmos B. (1994) Cloning of thehuman glycine transporter type I: molecular and pharmacologi-cal characterization of novel isoform variants and chnomosomallocalization of the gene in the human and mouse genomes. Mo!.Phormaco!. 45, 608—617.
Kitayama S., Shimada S., Xu H., Markham L., Donovan D. M., andUhI C. R. (1992) Dopamine transporter site-directed mutationsdifferentially alter substrate transport and cocainebinding. Proc.Nat!. Acad. Sc!. USA 89, 7782—7785.
Kitayama S., Wang J. B., and UhI C. R. (1993) Dopamine trans-porter mutants selectively enhance MPP transport. Synapse15, 58—62.
Kleckner N. W. and Dingledine R. (1988) Requirement for glycinein activation of NMDA-receptor expressed in Xenopus oocytes.Science 241, 835—837.
Knogsgaard-Lansen P., Falch E., Lansson 0. M., and Schousboe A.(1987) CABA uptake inhibitors: relevance to antiepileptic drugresearch. Epilepsy Res. 1, 77—93.
Kuhar M. J. (1973) Neunotnansmitten uptake: a tool in identifyingneunotransmitten-specific pathways. LiJb Sci. 13, 1623—1634.
Kuhar M. J., Ritz M. C., and Boja J. W. (1991) The dopamine hy-pothesis of the reinforcing properties of cocaine. Trend.s Neu-rosci. 14, 299—302.
Lakso M., Sauer B., Nosinger B. Jr., Lee E. J., Manning R. W., YuS-H., Mulder K. L., and Westphal H. (1992) Targeted oncogneactivation by site-specific necombination in transgenic mice.Proc. Nat!. Acad. Sc!. USA 89, 6232—6236.
Launay J.-M., Geoffroy C., Mutel V., Buckle M., Cesura A., AbutJ. E., and Da Prada M. (1992) One-step purification of thesenotonin transporter located at the human platelet plasma mem-brane. J. Bio!. Chem. 267, 11344—11351.
Lester H. A., Cao Y., and Magen S. (1996) Listening to neurotnans-mitten transporters. Neuron 17, 807—8 10.
Lill H. and Nelson N. (1998) Homologies and family relationshipsamong Na*/Cl neunotnansmitter transporters. Methods Enzy-ow!. (in press).
Lin F., Lester H. A., and Mager 5. (1996) Single-channel currentsproduced by the serotonin transporter and analysis of a mutationaffecting ion permeation. Biophys. J. 71, 3126—3135.
Liu Q.-R., Mandiyan S., Nelson H., and Nelson N. ( 1992a) A familyof genes encoding neurotransmitter transporters. Proc. Nat!.Acod. Sc!. USA 89, 6639—6643.
Liu Q.-R., López-Corcuena B., Nelson H.. Mandiyan S., and NelsonN. (l992h) Cloning and expression of a eDNA encoding the
J. Neurorheni., Vol. 7/, No. 5, 1998
1802 N. NELSON
transporter of taunine and /3-alanine in mouse brain. Proc. Nat!.Acad. Sci. USA 89, 12145—12149.
Liu Q.-R., Nelson H., Mandiyan S., López-Corcuera B., and NelsonN. (1992 c) Cloning and expression of a glycine transporterfrom mouse brain. FEBS Lett. 305, 110—114.
Liu Q.-R., López-Corcuena B., Mandiyan S., Nelson H., and NelsonN. (1993a) Molecular characterization of four pharmacologi-cally distinct y-aminobutynic acid transporters in mouse brain.J. Bio!. Chem. 268, 2106—21 12.
Liu Q.-R., Ldpez-CorcueraB., Mandiyan S., Nelson H., and NelsonN. (l993b) Cloning and expression of a spinal cord and brain-specific glycmne transporter with novel structural features. J.Biol. Chem. 268, 22802—22808.
Liu Q.-R., Mandiyan S., López-Corcuera B., Nelson H., and NelsonN. (1993c) A rat brain eDNA encoding the neurotransmittertransporter with an unusual structure. FEBS Lett. 315, 114—uS.
Loo D. D., Zeuthen T., Chandy G., and Wright F. M. (1996) Co-transport of water by the Na * / glucosecotransporter. Proc. Nat!.Acad. Sci. USA 93, 13367—13370.
López-Corcuena B., Vázquez J., and Aragón C. (1991) Purificationof the sodium- and chloride-coupled glycine transporter fromcentral nervous system. J. Bio!. Chem. 266, 24809—248 14.
Lopez-Corcuera B., Liu Q.-R., Mandiyan S., Nelson H., and NelsonN. (1992) Expression of a mouse brain eDNA encodmng novely-aminobutynic acid transporter. J. Bio!. Chem. 267, 17491—17493.
Luque J. M., Nelson N., and Richards C. (1995) Cellular expressionof the glycine transporter GLYT2 mRNA exclusively in rathindbrain and spinal cord. Neuroscience 64, 525—535.
Luque J. M., Jursky F., Nelson N., and Richards J. C. (1996) Distri-bution and sites of synthesis of NTT4, an orphan member ofthe Na/Cl-dependent neurotransmitter transporter family, in therat CNS. Eur. J. Neurosci. 8, 127—137.
Mabjeesh N. J., Frese M., Rauen T., Jesenich C., and Kanner B. I.(1992) Neuronal and glial ‘y-aminobutynic acid transporters aredistinct proteins. FEBS Lett. 299, 99—102.
Mabjeesh N.J. and Kanner B. I. (1993) The substrates ofa sodium-and chloride-coupled gamma-ammnobutynic acid transporter pro-tect multiple sites throughout the protein against proteolyticcleavage. Biochemistry 32, 8540—8546.
Magen S., Naeve J., Quick M., Labanca C., Davidson N., and LesterH. A. (1993) Steady states, charge movements, and rates for acloned GABA transporter expressed in Xenopus oocytes. Neu-ron 10, 177—188.
Mager S., Mm C., Henry D. J., Chavkin C., Hoffman B. J., DavidsonN., and Lester H. A. (1994) Conducting states of a mammaliansenotonin transporter. Neuron 12, 845—859.
Mager S., Klemnbengen-Doron N., Keshet G. I., Davidson N., KannerB. I., and Lester H. A. (1996) Ion binding and permeation atthe GABA transporter GATI. J. Neurosci. 16, 5405—5414.
Maloney P. C. (1990) A consensus structure for membrane trans-port. Res. Microbioi. 141, 374—383.
McGivan J. D. and Pastor-Anglada M. (1994) Regulatory and mo-lecular aspects of mammalian amino acid transport. Biochem.J. 299, 321—334.
Melikian H. E., McDonald J. K., Cu H., Rudnick G., Moore K. R.,and Blakely R. D. (1994) Human norepinephnine transporter.Biosynthetic studies using a site-directed polyclonal antibody.J. Biol. Chem. 269, 12290—12297.
Miller J. W., Kleven D. T., Domin B. A., and Fremeau R. T. (1997)Cloned sodium- (and chloride-) dependent high-affinity trans-porters for GABA, glycine, proline, betaine, taurine, and cre-atine, in Neurotransmitter Transporters—Structure, Function,and Regulation (Reith E. A., ed), pp. 101—150. Humana Press,Totowa, New Jersey.
Moore K. R. and Blakely R. D. (1994) Restriction site-independentformation ofchimeras from homologous neurotransmitter-trans-porter cDNAs. Biotechniques 17, 130—136.
Moniyoshi K., Masu M., Ishii T., Shigemoto R., Mizuno N., andNakanishi S. (1991) Molecular cloning and characterization ofthe rat NMDA receptor. Nature 354, 31—37.
Nelson H., Mandiyan S., and Nelson N. (1990) Cloning of thehuman brain GABA transporter. FEBS Lett. 269, 181—184.
Nelson N. (1993) Presynaptic events involved in neurotransmission.J. Physiol. (Paris) 87, 171—178.
Nelson N. and Lill H. (1994) Porters and neurotransmitter transport-ers. J. Exp. Bio!. 196, 213—228.
Nickolson V. J. (1982)’ ‘On” and “off” responses of K * -inducedsynaptosomal proline release: involvement of the sodium pump.J. Neurochem. 38, 289—292.
Nirenbeng M. J., Vaughen R. A., UhI C. R., Kuhar M. J., and PickelV. M. (1996) The dopamine transporter is localized to dendniticand axonal plasma membranes of nigrostniatal dopaminergicneurons. J. Neurosci. 16, 436—447.
Olivares L., Aragon C., Gimenez C., and Zafra F. (1997) Analysisof the transmembrane topology of the glycine transporterGLYTI. J. Bio!. Chem. 272, 1211—1217.
Pacholczyk T., Blakely R. D., and Amara S. G. (1991) Expressioncloning of a cocaine- and antidepressant-sensitive human nor-adrenaline transporter. Nature 350, 350—354.
Panayotova-Heiermann M., Eskandari S., Turk E., Zampighi C. A.,and Wright E. M. (1997) Five transmembrane helices form thesugar pathway through the Na * / glucose cotransporter. J. Biol.Chem. 272, 20324—20327.
Pantanowitz S., Bendahan A., and Kannen B. I. (1993) Only one ofthe changed amino acids located in the transmembrane alpha-helices of the gamma-aminobutynic acid transporter (subtypeA) is essential for its activity. J. Bio!. Chem. 268, 3222—3225.
Pniestley T., Home A. L., McKernan R. M., and Kemp J. A. (1990)The effect of NMDA receptor glycmne site antagonists on hy-poxia-induced neurodegeneration of rat cortical cell cultures.Brain Res. 531, 183—188.
Racker E. (1976) A New Look at Mechanisms in Bioenergetics.Academic Press, New York.
Radian R. andKanner B. I. (1985) Reconstitution and purification ofsodium- and chloride-coupled y-aminobutyric acid transporterfrom rat brain. J. Bio!. Chem. 260, 11859—11865.
Radian R., Bendahan A., and Kannen B. I. (1986) Purification andidentification of the functional sodium- and chloride-coupledgamma-ammnobutynic acid transport glycoprotein from rat brain.J. Bio!. Chem. 261, 15437—15441.
Radian R., Ottensen 0. P., Storm-Mathisen J., Castel M., and KannerB. I. (1990) Immunocytochemical localization of the GABAtransporter in rat brain. J. Neuro.sci. 10, 1319—1330.
Ramamoorthy S., Bauman A. L., Moore K. R., Han H., Yang-FengT., Chang A. S., Ganapathy V., and Blakely R. D. (1993) Anti-depressant- and cocaine-sensitive human serotonin transporter:molecular cloning, expression, and chromosomal localization.Proc. Nat!. Acad. Sc!. USA 90, 2542—2546.
Risso S., DeFelice L. J., and Blakely R. D. (1996) Sodium-depen-dent CABA-induced currents in CAT 1 -transfected HeLa cells.J. Physio!. (Lond.) 490, 691 —702.
Rudnick R. (1997) Mechanisms of biogenic amine neurotransmlttertransporters, in Neurotransmitter Transporters— Structure,Function, and Regulation (Reith E. A., ed), pp. 73—100. Hu-mana Press, Totowa, New Jersey.
Rudnick C. and Clark J. (1993) From synapse to vesicle: the reup-take and storage of biogenic amine neurotransmitters. Biochim.Biophys. Acta 1144, 249—263.
Rudnick C. and Nelson P. J. (1978) Platelet 5-hydroxytryptaminetransport, an electroneutral mechanism coupled to potassium.Biochemistry 17, 4739—4742.
Schloss P., Mayser W., and Betz H. (1992) Neurotransmitter trans-porters. A novel family of integral plasma membrane proteins.FEBS Lett. 307, 76—78.
Schon F. and Kelly J. S. (1975) Selective uptake of [3H)beta-alanine
by glia: association with glial uptake system for CABA. BrainRes. 86, 243—257.
Schuldiner S. (1994) A molecular glimpse of vesicular monoaminetransporters. J. Neurochem. 62, 2067—2078.
Shimada S., Kitayama S., Lin C.-L., Patel A., Nanthakumar E.,Cregon P., Kuhar M.. and UhI C. (1991) Cloning and expres-
Singh L., Donald A. E., Foster A. C., Hutson P. H., Iversen L. L.,Ivensen S. D., Kemp J. A., Leeson P. D., Marshall C. R., OlesR. J., Pniestley T., Thorn L., Tnicklebank M. D., Vass C. A.,and Williams B. J. (1990) Enantiomens of HA-966 (3-ammno-l-hydroxypynnolid-2-one) exhibit distinct central nervous systemeffects: (+ )-HA-966 is a selective glycine/N-methyl-D-aspan-tate receptor antagonist, but (— )-HA-966 is a potent y-butyro-lactone-like sedative. Proc. Nat!. Acad. Sc!. USA 87, 347—351.
Smith K. E., Borden L. A., Hartig P. R., Branchek T., and Wein-shank R. L. (1992a) Cloning and expression of a glycmne trans-porter reveal colocalization with NMDA receptors. Neuron 8,927 —935.
Smith K. F., Borden L. A., Wang C-H., Hantig P. R., BranchekT. A., and Weinshank R. L. (1992b) Cloning and expressionof a high affinity taurine transporter from rat brain. Mo!. Phar-maco!. 45, 563—569.
Smith K. E., Fried S. C., Durkin M. M., Custafson E. L., BordenL. A., Branchek T. A., and Weinshank R. L. (1995) Molecularcloning of an orphan transporter. A new member of the neuno-transmitter transporter family. FEBS Lett. 357, 86—92.
Snyder S. H. and D’Amato R. J. (1986) MPTP: a neunotoxin rele-vant to the pathophysiology of Parkinson’s disease. Neurology36, 250—258.
Sonders M. S. and Amara S. C. (1996) Channels in transporters.Curr. Opin. Neurob!o!. 6, 294—302.
Sonders M. S., Zhu S. J., Zahnisen N. R., Kavanaugh M. P., andAmara S. C. (1997) Multiple ionic conductances of the humandopammne transporter: the actions of dopamine and psychostim-ulants. J. Neurosci. 17, 960—974.
Sora I., Wichems C., Takahashi N., Li X. F., Zeng Z., Revay R.,Lesch K. P., Murphy D. L., and UhI C. R. (1998) Cocaine re-ward models: conditioned place preference can be establishedin dopamine- and in serotonin-transporten knockout mice. Proc.Nat!. Acad. Sc!. USA 95, 7699—7704.
Stephan M. M., Chen M. A., Penado K. M., and Rudnick C. (1997)An extracellulan loop region of the senotonin transporter maybe involved in the translocation mechanism. Biochemistry 36,1322— 1328.
Su A., Magen S., Mayo S. L., and Lester H. A. (1996) A multi-substrate single-file model for ion-coupled transporters. B!o-phys. J. 70, 762—777.
Tamura S., Nelson H., Tamura A., and Nelson N. (1995) Shortexternal loops as potential substrate binding site of CABA trans-porters. J. Bio!. Chem. 270, 28712—28715.
Toth E. and Lajtha A. (1986) Antagonism of phencyclidmne-mnduced
hyperactivity by glycine in mice. Neurochem. Res. 11, 393—400.
Tsai C., Passani L. A., Slushen B. S., Carter R., Baer L., KleinmanJ. F., and Coyle J. T. (1995) Abnormal excitatory neunotnans-mitten metabolism in schizophrenic brains. Arch. Gen. Psychia-try 52, 829—836.
Tsien J. Z., Chen D. F., Gerber D., Tom C., Mercer E. H., AndersonD. J., Mayford M., Kandel E. R., and Tonegawa 5. (1996) Sub-region- and cell type-restricted gene knockout in mouse brain.Ce!! 87, 1317—1326.
Uchida S., Kwon H. M., Yamauchi A., Preston A. S., Marumo F.,and Handler J. 5. (1992) Molecular cloning of the eDNA foran MDCK cell Na * - and Cl -dependent taunine transporter thatis regulated by hypertonicity. Proc. Nat!. Acod. Sci. USA 89,8230— 8234.
UhI C. R. (1992) Neurotransmitten transporters (plus): a promisingnew gene family. Trends Neurosci. 15, 265—268.
Uh1G. and JohnsonP.S. ( 1994)Neunotnansmitter transporters: threeimportant gene families for neunonal function. J. Exp. Bin!.196, 229—236.
UhI C. R., Kitayama S., Gregon P., Nanthakumar F., Pensico A., andShimada 5. (1992) Neurotransmitter transporter family cDNAsin a rat midbrain library: “orphan transporters” suggest sizablestructural variations. Mo!. Brain Re.s. 16, 353—359.
UhI C. R., Walthen D., Mash D., Faucheux B., and Javoy-Agid F.(1994) Dopamine transporter messenger RNA in Parkinson’sdisease and control substantia nigra neurons. Ann. Neuro!. 35,494—498.
Usdin T. B., Mezey F., Chen C., Bnownstein M. J., and HoffmanB. J. (1991) Cloning of the cocaine-sensitive bovine dopaminetransporter. Proc. Nat!. Acad. Sc!. USA 88, 11168—11171.
Vandenhergh D. J., Persico A. M., and UhI C. R. (1992) A humandopamine transporter eDNA predicts reduced glycosylation,displays a novel repetitive element and provides racially-dimon-phic Taq I RFLPs. Mo!. Brain Res. 15, 161—b 66.
Wasserman J. C., Delpire E., Tonidandel W., Kojima R., and CullansS. R. (1994) Molecular characterization of ROSIT, a renal os-motic stress-induced Na * -Cl -organic solute cotnansporten.Am. J. Physio!. 267, F688—F694.
Williamson A., Telfeian A. E., and Spencer D. D. (1995) ProlongedGABA responses in dentate granule cells in slices isolated frompatients with temporal lobe sclerosis. I. Neurophysio!. 74, 378—387.
Wood J. D. and Sidhu H. 5. (1986) Uptake of y-aminobutynic acidby brain tissue preparations: a reevaluation. J. Neurochem. 46,739—744.
Yamauchi A., Uchida S., Kwon H. M., Preston A. S., Robey R. B.,Garcia-Perez A., Burg M. B., and Handler J. S. (1992) Cloningof a Na(+)- and Cl(—)-dependent betaine transporter that isregulated by hypertonicity. J. Bin!. Chem. 267, 649—652.
