THE JOURNAL OF BIOLOGICAL CHEMISTRY (0 1993 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 268, No. 3, Issue of January 25, pp. 2106-2112,1993 Printed in U.S.A.

Molecular Characterization of Four Pharmacologically Distinct a-Aminobutyric Acid Transporters in Mouse Brain*

(Received for publication, August 5, 1992)

Qing-Rong Liu, Beatriz Lopez-Corcueral, Sreekala Mandiyan, Hannah Nelson, and Nathan Nelsons From the Roche Institute of Molecular Biology, Roche Research Center, Nutley, New Jersey 07220

Two novel y-aminobutyric acid (GABA) transport- ers, GAT3 and GAT4, were cloned from the mouse neonatal brain cDNA library and expressed in Xenopus oocytes. Sequence analysis indicated they were mem- bers of the Na+-dependent neurotransmitter trans- porter family. The GABA uptake activities were meas- ured in cRNA injected Xenopus oocytes. The K,,, for GABA uptake by GAT3 was 18 MM and by GAT4 was 0.8 PM. GAT3 also transports &alanine and taurine with K , of 28 and 540 MM, respectively. Similarly, GAT4 transports ,&alanine with K , of 99 PM and tau- rine with a K , of 1.4 mM. The newly cloned GABA transporters were compared with two previously cloned GABA transporters, GATl and GATB, in terms of molecular and pharmacological properties. While GATl and GAT4 gene expression were neural specific, GAT2 and GAT3 mRNAs were detected in other tissues such as liver and kidney, in which GAT3 mRNA was especially abundant. The expression of GAT3 mRNA in mouse brain is developmentally regulated, and its mRNA is abundant in neonatal brain but not in adult brain. High affinity GABA transporters GATl and GAT4 were more sensitive to inhibition by nipecotic acid. Low affinity GABA transporters GAT2 and GAT3 were inhibited most effectively by betaine and 0-alanine, respectively. The differential tissue distri- bution and distinct pharmacological properties of those four GABA transporters suggest functional speciali- zation in the mechanisms of GABA transmission ter- mination.

The amino acid y-aminobutyric acid (GABA)’ is the pre- dominant inhibitory neurotransmitter in brain and is widely distributed throughout the nervous system (1,Z). The uptake of GABA was extensively studied in various brain tissue preparations such as slices, homogenates, synaptosomal-en- riched fractions, and reconstituted transporters (3-10). A multitude of sodium-dependent GABA-uptake systems with K , ranging from 1 p~ to 4 mM were reported. In general, the uptake systems were divided into two groups of neuronal and

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) LO4663 and L04662.

$, Recipient of a North Atlantic Treaty Organization fellowship. 3 To whom all correspondence should be addressed. Tel.: 201-235-

:3790; Fax: 201-235-5848. ‘ The abbreviations used are: GABA, y-aminobutyric acid; DABA,

1.-2,4-diamino-n-butyric acid PGPA, (3-guanidinopropionic acid; DAPA, ~-2,3-diaminopropionic acid PEG, polyethylene glycol; MOPS, (3-[N-morpholino]propanesulfonic acid).

glial transporters according to their sensitivity to nipecotic acid or P-alanine (11, 12). However, a closer look at the distribution of one of the GABA transporters (GAT1) revealed a much more complicated picture suggesting the presence of more GABA transporters localized in specific parts of the brain (10-13). Moreover, different types of GABA transport- ers could be present not only in different brain areas but also in different stages during nervous system development. Such a versatility has already been proven for GABA receptors (14), and it would not be surprising if the same phenomenon would be observed for GABA transporters, whose activity is physiologically coupled with that of the receptors in the synapses.

The cDNA encoding two different brain GABA transporters (GAT1 and GAT2) were cloned, sequenced, and expressed in Xenopus oocytes (15, 16). These genes are part of a family of neurotransmitter transporters sharing similar structures and amino acid sequences but different pharmacological proper- ties (15-24). In this paper we report on the cloning and expression of two new cDNAs encoding GABA transporters (GAT3 and GAT4). They also encode highly hydrophobic proteins of about 70 kDa with 12 potential transmembrane segments. All four GABA transporters differ in their expres- sion, substrate specificity, and sensitivity to inhibitors.

EXPERIMENTAL PROCEDURES

Expression of the GABA Transporters in Xenopus Oocytes-Follow- ing linearization of the plasmid by XhoI, RNA was synthesized using T3-RNA polymerase with RNA synthesis and capping kit from Stratagene. The synthetic RNA was injected into Xenopus oocytes, and the GABA uptake was assayed as previously described (15). Briefly, oocytes were surgically removed from frogs and defolliculated by collagenase treatment. After recovery for 24 h, the oocytes were injected with 50 nl containing 2-10 ng of synthetic RNA. After 2-3 days they were assayed for GABA transport. Prior to the uptake assay, the oocytes were washed three times with 1 ml of a solution containing 102 mM KCI, 1 mM CaC12, 1 mM MgC12, and 10 mM HEPES (pH 7.5) and preincubated in 0.5 ml of this solution for 15 min. After removal of the preincubation medium the transport reac- tion was started by the addition of 0.5 ml of solution containing 100 mM NaCl, 2 mM KCI, 1 mM CaCl,, 1 mM MgCL, 10 mM HEPES (pH 7.5), about 0.1 pCi of [3H]GABA and 1 FM or the specified amounts of cold GABA. After incubation for 30 min at room temperature, the oocytes were washed three times with 1 ml of the incubation solution in which the substrate was omitted. Individual oocytes were solubi- lized in 500 bl of 1% SDS and the radioactivity measured by scintil- lation counting. Six to 12 oocytes were used for each experimental point, and the data are expressed as the average uptake/hour. The effect of various chemicals on GABA uptake was assayed by including them at the indicated concentrations during the preincubation and assay periods.

cDNA library, cloned into the EcoRI-XhoI site of the Uni-ZAP X R Screening, Cloning, and Sequencing-A neonatal mouse brain

cloning vector, was obtained from Stratagene. This library was screened with a 32P-labeled cDNA encoding the taurine transporter (28). The clone NTTl2 that encodes a GABA transporter GAT3 was isolated from this library. This library also yielded the clone NTT13

2106

Four Different GABA Transporters 2107

that encodes about two-thirds of the GABA transporter GAT4. To obtain the full-length clone of GAT4, an adult mouse-brain 5'- STRETCH cDNA library (XgtlO, Clontech) was screened with NTT13. A cDNA of 4 kilobases was isolated and identified as a full- length GABA transporter by expression in Xenopus oocytes. The screening was performed as follows. The host Escherichia coli strain BB4 was streaked on an LB plate containing 12.5 pg/ml tetracycline and grown a t 37 "C overnight. A single colony was picked and inoc- ulated into 50 ml of LB medium containing 0.2% maltose and 12.5 pg/ml tetracycline. The bacterial culture was incubated overnight a t 37 "C with vigorous shaking, then spun down a t 6,000 X g for 5 min. The pellet was resuspended in 25 ml of medium for dilution and storage of bacteriophage X (SM). About 1 million plaque-forming units were taken from the diluted cDNA library (100 times dilution with SM) and added to 8 ml of the bacterial suspension. The mixture was incubated a t 37 "C for 20 min and then divided into 12 aliquots of 0.6 ml. At the same time, 0.6% agarose in LB medium was melted by microwave and was left to cool down to about 55 "C and set on a 65 "C heat block. The 150 X 15 mm LB-agar plates were dried a t 42 "C for 1 h before plating. Seven ml of 0.6% agarose-LB medium was added to each of the 12 culture tubes and immediately placed on the LB-agar plates. After solidification the plates were incubated at 37 "C in the inverted position for 6-8 h until clear plaques appeared. The plates were stored a t 4 "C for a t least 2 h, after which a nitrocel- lulose filter (137 mm, 40-pm pore size) was laid on the plates for about 1-2 min. After lifting, the filters were dried for about 10 min, and the denaturation and neutralization were carried out by soaking the filters in 0.5 M NaOH, 1.5 M NaCl for 2 min, transferring the filter in order to 0.5 M Tris-HC1 (pH 8.0), 1.5 M NaCl for 5 min and then 2 X SSC (0.15 M NaCl in 15 mM sodium citrate (pH 7.0) for 5 min. The DNA was cross-linked to the filters with a UV Stratalinker (Stratagene) for two cycles. The filters were wet by 5 X SSC for 5 min and then put one by one into a plastic bag containing 30 ml of hybridization solution (50% formamide, 5 X SSC, 14 mM Tris-HC1 (pH 7.5), 1 X Denhardt's and 10% dextran sulfate) and prehybridized at, 42 "C for 2 h. After prehybridization, the filters were removed and a denatured 32P-labeled DNA probe was added to the prehybridization solution. The filters were put back into the hybridization solution one by one and hybridized a t 42 "C overnight. Washing of the filters was carried out a t low stringency. The filters were first washed with 2 x SSC, 0.1% SDS at room temperature 3 times for 10 min each, then with 1 X SSC, 0.1% SDS at 42 "C twice for 30 min each. The washed filters were placed between two sheets of Saran Wrap sup- ported by a used x-ray film for autoradiography. The positive plaques were picked up by using a cut-off 1-ml pipette tip. The agar plug was added to 0.5 ml of SM with one drop of chloroform in an Eppendorf tube which was then stored a t 4 "C overnight to allow phages to be released. One p1 of the phage suspension from the stock of the primary screening was diluted into 1 ml of SM, from which 2 pl was used to infect the E. coli host cells. The secondary screening was performed similar to the primary screening except 100 X 15-mm LB-agar plates and 82-mm nitrocellulose filters (45-pm pore size) were used. A single positive plaque was picked up from the secondary screening using a sterile Pasteur pipette and added to 500 pl of SM buffer with one drop of chloroform. The phage plug was stored a t 4 "C overnight to release the phage.

The excision of Bluescript plasmid with a positive cDNA insert was performed as follows. A single colony of XL1-blue cells was grown in 2 X 8 g/liter bactotryptone, 5 g/liter bactoyeast extract in 5 g/liter NaCl (YT), 0.2% maltose/lO mM MgClt overnight. Two-hundred p1 of the XL1-blue overnight cell culture was mixed with 200 p1 of phage stock and 1 pl of R408 helper phage (106-7 pfu/ml) and incubated a t 37 "C for 15 min. Five ml of 2 X YT was added to the mixture and incubated at 37 "C with shaking until lysis occurred (about 3-5 h). The phage lysate was heated a t 70 "C for 20 min, and the cell debris were spun down a t 6,000 X g for 10 min. One p1 of supernatant was diluted with 2 X YT to 100 &I, from which 20 pl was added to 200 p1 of XL1-blue cell culture. The mixture was incubated a t 37 "C for 15 min, and 50 pl of the mixture were spread on LB-agar plates contain- ing 100 pg/ml ampicillin and incubated a t 37 "C overnight. The minipreparation and dot blot were performed as described previously (25). The isolation of XgtlO phage DNA with positive cDNA inserts was performed as follows. A single positive plaque was picked up with a sterile Pasteur pipette and added to a sterile Eppendorf tube containing 300 p1 of SM buffer with one drop of chloroform. The phages were allowed to release a t 4 "C overnight. The host cell C600 H f l was inoculated to LB medium containing 0.2% maltose and grown a t 37 "C overnight with shaking. About 300 pl of the phage

stock was combined with 500 pl of overnight culture of C600 Hfl cells and incubated a t 37 "C for 20 min. The infected C600 Hfl cells were added to 100 ml LB, 10 mM MgSO, medium and incubated a t 37 "C with vigorous shaking until cell lysis occurred (when the turbid bacteria suspension became clear after 4-8 h). 0.1 ml of chloroform was added and the incubation was continued for another 10 min a t 37 "C with shaking. The cell debris were spun down a t 6,000 X g for 10 min. The supernatant was transferred to a 250-ml centrifuge tube and half volume (50 ml) of 15% PEG (8,000), 1.5 M NaCl was added. The X g t l O phage particles were precipitated by storing a t 4 "C over- night and then centrifuged at 12,000 X g for 30 min at 4 "C. The pellet was dispensed in 3 ml of SM by pipetting up and down. The suspension was transferred to a 15-ml polypropylene tube containing an equal volume of chloroform. The mixture was vortexed for 30 s and centrifuged a t 6,000 X g for 10 min a t 4 "C. The aqueous phase was transferred to a fresh 15-ml tube. DNase I and RNase A digestion was carried out by adding 15 p1 of 1 mg/ml DNase I and 30 p1 of 1 mg/ml RNase A and incubated at 37 "C for 30 min. Proteinase K digestion was carried out by adding 30 p1 of 0.5 M EDTA (pH 8.0), 8 pl of 20 mg/ml proteinase K, and 30 p1 of 10% SDS sequentially and incubated at 68 "C for 15 min. The reaction mixture was extracted twice with an equal volume of phenol-chloroform. Sodium acetate (pH 6.0) was added to give a final concentration of 0.2 M and the DNA was precipitated with 2 volumes of ethanol. The pellet was dissolved in 480 p1 of 10 mM Tris-HC1 (pH 8.01, 0.1 mM EDTA (TE) buffer containing 20 pg/ml RNase A and incubated a t 37 "C for 30 min. The solution was mixed with 120 p1 of 5 M NaCl and 600 p1 of 13% PEG (8,000). The mixture was set on ice for 30 min and centrifuged a t 10,000 X g for 15 min a t 4 "C. The pellet was rinsed with 70% ethanol, lyophilized, and dissolved in 50 pl of sterile deionized water. The final yield of X y t l O DNA was about 100 pg. The DNA inserts were released with EcoRI digestion and subcloned into the EcoRI site of Bluescript vector (25).

Sequencing of the cDNA in Bluescript plasmid was performed by overlapping deletions generated by exonuclease 111 (26) and verified by sequencing with synthetic oligonucleotides as previously described (15). Preliminary sequencing was performed by an Applied Biosystem 373A Sequencer and verified by radioactive sequencing using ATP. The sequences were assembled using DNAstar software and analyzed by a GCG program in the VAX computer.

RNA Isolation and Northern Blot-Mouse brain was dissected into its various parts from decapitated mice and immediately frozen in liquid nitrogen. One g of the tissue was ground in liquid nitrogen with a pestle and mortar. The tissue powder was added to 14 ml of guanidinium thiocyanate, 0.07% P-mercaptoethanol and homogenized with three 30-s bursts of a Polytron, according to the manufacturer's recommendation supplied with the RNA isolation kit (Stratagene). Following acid-phenol extraction and isopropyl alcohol precipitation, the samples were lyophilized, and the total cellular RNA was dissolved in RNase-free water and the concentration was determined by UV absorption at ODZm nm. The aliquots of RNA could be stored a t -80 "C before running the gel. Equal amounts of RNA (40 pg) were mixed 1:l with RNA denaturing solution (50% formamide, 2 X MOPS buffer and 6% formaldehyde), incubated a t 65 "C for 5 min, then chilled on ice. After adding 0.1 volume of 10 X dye solution (50% glycerol, 0.25% bromphenol blue, 0.25% xylene cyano], and 1 mM EDTA), the RNA samples were loaded onto 1% agarose-formaldehyde gels (1 x MOPS buffer (pH 7.0),6.6% formaldehyde) and the electro- phoresis was run a t 100 V in 1 X MOPS buffer (pH 7.0). The RNA gel was stained with ethidium bromide in 10 mM sodium phosphate buffer (pH 7.2), for 30 min. The gel was destained in the same buffer and then soaked in 10 X SSC for 30 min. RNA was transferred to a Duron-UV membrane (Stratagene) and cross-linked to the membrane by UV. The labeled DNA probe for GAT3 was the full-length cDNA inserts, and GAT4 was the insert ofthe partial cDNA clone of NTTl3. Prehybridization for 2 h and hybridization overnight with specific probes were carried out in the same hybridization buffer (1% bovine serum albumin, 7% SDS, 0.55 M sodium phosphate buffer (pH 7.2), and 1 mM EDTA) a t 65 "C. The filters were washed a t 65 "C first with washing solution 1 (1% bovine serum albumin, 5% SDS, 40 mM sodium phosphate buffer (pH 7.2), and 1 mM EDTA) twice for 15 min each, then with washing solution 2 (0.05% bovine serum albumin, 1% SDS, 40 mM sodium phosphate buffer, and 1 mM EDTA) three times at 10 minluiash (27). The exposure of autoradiography for GAT1 and GAT4 was overnight, GAT3 for 1 week, and GAT2 for 2 weeks.

2 108 Four Different GABA Transporters

RESULTS

The tissue-specific localization of the two GABA-trans- porters (GAT1 and GAT2) prompted us to look for additional transporters active in sodium-dependent GABA uptake. Sev- eral positive clones isolated as described under “Experimental Procedures” were assayed for GABA uptake following expres- sion in Xenopus oocytes. Two different cDNA clones exhib- ited GABA uptake activity and were named GAT3 and GAT4. Fig. 1 shows the nucleotide and predicted amino acid se- quences of GAT3. The predicted open reading frame encodes a protein of 602 amino acids with a molecular mass of 68,293 Da. An unusual low isoelectric point was calculated to be PI = 6.1. Most neurotransmitter transporters have PI > 8. The hydropathy plot suggests 12 transmembrane segments, a sim- ilar structure to all the other neurotransmitter transporters sequenced so far (12-23).

Alignment of GAT3 with predicted amino acid sequences of the other transporters showed 69% identity with GAT2 (16), 64% identity with the taurine transporter (28), 51% identity with GATl (15), 42% identity with the glycine trans- porter (24), and 42-45% identity with the catecholamine transporters (17-21). Therefore, GAT3 is a novel neurotrans- mitter transporter related to GAT2.

The substrate specificity of GAT3 was determined by

1

61

121

181

241

301

361

421

481

541

601

661

121

181

841

901

961

1021

1081

GMTTCGGCACGAGGAMCTTCACCMGGTGGGATGGAGMCAGGGCCTCGGGAACMCC U E N R A S G T T

TGCTACAGGGACTGCATCGCCCTTTGCATCCTCMCAGCAGCACCAGCTTCATGGCCGGG C Y R D C I A L C I L N S S T S F W A G TTTGCTATCTTCTCCATCCTGGGCTTCATGTCTCffiGAGCAGGGTGTGCCCATATCTGAG F A I F S I L G F U S P E Q G V P I S E CTTGCTGMTCAGGCCCTGGCCTGGCATTCATCGCCTACCCTCGAGCTGTGGTGATGTTA

60

120

180

240

300

360

420

480

540

600

660

720

180

840

900

960

1020

1080

1140

1111 CCCTTCTCACCGTTGTGGGCCTGCTGTTTCTTCTTCATGGTCGTTCTCCTGGGACTAGAT 1200

1201 AGCCAGTTTGTGTGTGTAGAMGCCTTGTGACAGCATTAGTGGACATGTATCCCCGGGTG 1260

1261 RCCGTM~MTCffiAGGAGGTCCTCATCCTTATTGTGTCCGTCAT~CGTTCTTC 1320

V A E S G P G L A F I A Y P R A V V U L

P F S P L W A C C F F F U V V L L G L D

S P F V C V E S L V T A L V D ~ Y P R V

1321 ATCGGGCTCATCATGCTCACAG~GGCGGCAT~AC~GTTCCAGCTCTTTGACTACTAT 1380

1381 GCffiCCAGTGGCATGTGTCTTCTCTTCGTGGCCATCTTTGAGTCCCTCTGTGTGGCTTGG 1440

1411 GTTTATffiffiCCGGCCGCTTCTACGACMCATTGAGGATATGATTGffiTATMGCCATGG 1500

1501 CCTCTTATCAMTATTGTTGGCTCTTTTTCAC~CAGCTGTGTGCCTGGCGACCTTCCTG 1560

1561 TTCTCCCTGATCAMTACACGCCACTGACCTACMCMGMGTACACGTACCCATGGTGG 1620

1621 GGGGATGCCCTAGGGTGGCTCCTA~TCTGTCCTCCATGATCTGCATTCCTGCCTGGAGC 1680

1681 ATCTACMGCTGAGGACTCTCMGGGCCCACTCAGAGAGAGACTTCGCCAGCTCGTGTGC 1140

1141 CCGGCTGMGACCTTCCCCAGMCCMCCAGAGCCMCTGCTCCTGCGACACCGATG 1800

1801 ACATCCCTTCTCffiGCTCACAGMCTGGAGTCTMCTGCTAGGCCAGGGCCTTTGACACA 1860

1861 CCTGTGAGCCTGGCTGTGGGGCAGCTATAGA~CAGAGGGGCGGMCTACCCC3CTGTG 1924 1921 GTGGGCCAG~GACGILMGffiGTA~TGffiACTCACCTTCTGMCTGGGTACCTCCTA 1980 1981 TCTTGATGCCCCTATTCCAGGGTGGTCCACACATTGTGACGTGCCCGMTCMTMCATC 2040 2041 CCATTGAGGCAGCTTGGGMGCCCAGGGAGGGCTGGGGAGGCMAAGCCTGGCTTGTTGC 2100 2101 TGGGCCCTCTTCCTTCCATGTCATTAMTCCMCTTCCTCTCAC- 2160 2161 A M 2163

~~

F R ~ K K ~ N R ~ R E ~ V L ~ I L I V S V I S F F

I G L I U L T E G G M Y V F Q L P D Y Y

A A S G U C L L F V A I F E S L C V A W

V Y G A G R F Y D N I E D W I G Y K P W

P L I K Y C W L F F T P A V C L A T F L

F S L I K Y T P L T Y N K K Y T Y P W W

G D A L G W L L A L S S U I C I P A W S

I Y K L R T L K G P L R E R L R Q L V C

P A E D L P Q K N Q P E P T A P A T P U

T S L L R L T E L E S N C ‘

FIG. 1. Nucleotide and deduced amino acid sequences of cDNA encoding the GABA-transporter GAT3. The cDNA was cloned and sequenced as described under “Experimental Procedures.” Both DNA strands were sequenced by the dideoxy termination method (36).

expression of its cRNA in Xenopus oocytes. The injected oocytes accumulated up to 350 times as much [3H]GABA as uninjected oocytes or oocytes injected with mRNA of the glycine transporter. Cold GABA competed with the radioac- tive GABA uptake. The anion and cation specificity for GABA uptake by the expressed GAT3 was similar to that of GATl and GAT2. Fig. 2 shows the effect of GABA concentration on its uptake into GAT3 cRNA-injected oocytes. The Eadie- Hofstee plot revealed a Michaelis constant K,,, = 18 @M. The expressed GAT3 also transports @alanine and to a lesser extent taurine. As shown in Fig. 2, relatively high affinity ( K , = 28 PM) for p-alanine uptake was observed. Taurine was transported by GAT3 at a relatively low affinity of K, = 540 PM. All of these substances were taken up at relatively high rates. This suggests that GAT3 is the transporter of GABA, p-alanine, and taurine.

Screening of the neonatal and adult mouse brain libraries

z 4 6 a l o vm

D I

0.1 0.2 0.3 0.4 0.5

GABA (NM)

I 0.1 0.2 0.3 0.4 0.5

palanins (pM)

I C

1 0.5 1 1.5 2 2.5

Taurine (mM)

FIG. 2. Kinetics of GABA, &alanine, and taurine uptake into Xenopus oocytes injected with synthetic mRNA of GAT3. The uptake of 3H-labeled GABA, 0-alanine, and taurine was assayed as described for GABA uptake under “Experimental Procedures.” A, GABA uptake. B, 0-alanine uptake. C, taurine uptake. The values obtained with uninjected oocytes were subtracted from the corre- sponding injected samples. The values obtained with the injected oocytes were 350- (GABA), 330- @-alanine), and 72- (taurine) fold greater than with the uninjected oocytes. Eadie-Hofstee analysis is depicted in the insets.

Four Different GABA Transporters 2 109

yielded another GABA transporter (GAT4). Fig. 3 shows the nucleotide and predicted amino acid sequences of GAT4. The open reading frame encodes a protein of 627 amino acids with a calculated molecular mass of 69,868 Da. Alignment of the amino acids of GAT4 with the other GABA transporters revealed that it is highly related to GAT2 and GAT3 and more remotely related to GATl (Fig. 4). The alignment of GAT4 with the known neurotransmitter transporters showed 70% identity with GAT3, 65% identity with GAT2 (16), 53% identity with GATl (15), 66% identity with the taurine trans- porter (28), 42% identity with the glycine transporter (24), and 42-44% identity with the catecholamine transporters (17- 21). The affinity of GATl to GABA was much higher than tha t of GAT2 or GAT3; therefore, it was surprising that the affinity of GAT4 to GABA was even greater than that of GATl (Fig. 5). The Eadie-Hofstee plot revealed a Michaelis constant (K,,,) of 0.65 p ~ , and two other experiments gave K, values of 0.86 and 0.94 p ~ . The mean value of about 0.8 p~ suggests that GAT4 is a high affinity GABA transporter with

1

61

121

1 8 1

241

301

361

421

4 8 1

541

601

661

121

781

841

901

9 6 1

1021

1081

1141

1201

1261

1321

1381

1441

1501

1561

1621

1611

1141

1801

1861

1981 1921

2041 2101 2161 2221 2281 2341 2401

2521 2461

2581 26al 2101 2161 2821

2941 3001 3061 3121

3241 3301 3361 3421

3601 3541

3721 3661

3841 3111

3901 3961 4021 4081 41.1

2881

3181

3481

60

120

180

240

300

360

420

480

540

600

660

120

n o 840

900

960

1020

lono

1140

1200

1260

1320

1380

1440

1500

1560

1620

1680

1140

I800

1860

1920

1980 2040 2100 2160 2220

2340

2460 2400

2520

2640 2100

2820 2760

2880 2940 3000 3060

3180 3120

3240 3300 3360 3420

3540

3660 3600

3180 3720

3840 3900 3960 4020 4080 4140

2280

2580

3480

FIG. 3. Nucleotide and deduced amino acid sequences of cDNA encoding the GABA-transporter GATI. The experimen- tal conditions were as described in Fig. 1.

the highest affinity for GABA than all transporters cloned so far. Like GAT3, GABA uptake by expressed GAT4 was also inhibited by @-alanine and taurine. Moreover, @-alanine up- take with K, = 99 PM and taurine uptake with K, = 1.4 mM were measured in the GAT4 cRNA injected oocytes (not shown).

The pharmacology of the four GABA transporters is de- picted in Table I. b-Alanine inhibited the GABA uptake by the expressed GAT3 and GAT4, and to much lesser extent by GAT2. Betaine inhibited GABA transport only by GAT2. GABA transport by GATl and GAT4 was more sensitive to DABA, guvacine, and nipecotic acid than that of GAT2 and GATS. While GAT2 was the most sensitive to phloretin and quinidine, GAT3 and GAT4 were the most sensitive to @GPA, DAPA, and hypotaurine. It is apparent that even though GAT2, GAT3, and GAT4 exhibit high sequence homology they have different substrate and inhibitor specificities. More- over, GATl and GAT4 have the highest affinity for GABA and have greater sensitivity to nipecotic acid.

The specificity of the various GABA transporters suggests specialization for distinct functions. Northern hybridization was performed with RNA isolated from different parts of the mouse brain as well as from kidney and liver. The GAT4- cDNA hybridized exclusively with the brain tissues and hy- bridization with RNA from kidney and liver could not be detected (Fig. 6). The 5-kilobase GAT4 mRNA was enriched in the brain stem and the rest of the brain at least five times more than the cerebellum and cerebral cortex. In contrast, the GATl mRNA was distributed evenly in different parts of the brain (13). GAT3 cDNA hybridized weakly with the RNA isolated from adult mouse brain, however, it gave a strong positive 2.5-kilobase hybridization band with liver and kidney RNA. Since colonies containing the cDNA encoding GAT3 were abundant in the neonatal brain library, we looked for the presence of mRNA in neonatal and young mice. As shown in Fig. 7 neonatal brains contain substantial amounts of GAT3 mRNA. Its amounts increase in the first few days after birth and decreases in the adult brains. No significant pattern of mRNA change for GAT1, GATS, and GAT4 was observed in the neonatal mouse brain in comparison with adult mouse brain (not shown). Therefore, GAT3 may have a role in brain development, and the amounts of its mRNA are develop- mentally regulated.

DISCUSSION

Studies on GABA uptake into brain slices and isolated cells revealed the presence of more than one sodium-dependent GABA transport systems (1-7, 29). Recently, the presence of two pharmacologically distinct GABA transporters was dem- onstrated (10). It was shown that while one of the transport- ers, presumably GAT1, is not inhibited by p-alanine, the second GABA transporter is sensitive to @-alanine and may also function in @-alanine transport. An apparent K, of about 44 pM was observed for the @-alanine uptake activity (10). Out of the four GABA transporters analyzed in this work only GAT3 and GAT4 transported P-alanine at appreciable rates and with relatively high affinity of K, = 28 and 99 p ~ , respectively. These values are close enough to the reported K, = 44 PM for @-alanine uptake in the reconstituted prepa- ration from rat synaptosomes (10). Recently, we cloned a taurine transporter that also transports @-alanine with a K, = 56 pM (28). Therefore, the observed p-alanine transport in preparations of brain synaptosomes may be the result of uptake by more than a single transporter. The taurine trans- porter, GAT3, and GAT4, can be distinguished by the sensi- tivity of their @-alanine transport to inhibition by GABA.

2110

FIG. 4. Alignment of the amino

GATS, GAT3, and GAT4. Deduced acid sequences of mouse GAT1,

GABA-transporters GATl (15), GAT2 amino acid sequences of the mouse

(161, GAT3 and GAT4, were aligned using a DNAstar program. Identical amino acids are indicated by an asterisk (*I. Putative transmembrane domains are underlined.

GAT4

GAT3

GAT2

GAT 1

GAT 4

GAT3

GAT2

GAT 1

GAT4

GAT3

GAT2

GATl

GAT4

GAT3

GAT2

GATl

GAT 4

GAT3

GhTZ

GATl

GAT4

GAT3

GAT2

GATl

Four Different GABA Transporters ~ ~ P W " ~ 1 I G " P ~ I ~ I P w T m ~

) O l l l 8 0 R s n c r m m m - ~ x x G - ~ I p n r - I - w

~ A ~ W W I I W ~ ~ M R R ~ I P Y ? I ~ S = G I W ? ~ ~

u ~ x ~ I ~ ~ ~ L ~ ~ ~ ~ ~ ~ ~ P Y ? L n x ~ ~ ~

. . . ........................................................ ........................................................... . . . ..................................... 1 2

T S . f f i I T ~ Y L E I G l ~ I I ~ ~ I ~ ~ ~ ~ ~ ~ S ~ - S ~ ~ l E f f i l ~ I ~ ......................................................................... Z * P G I T M R I B I I M i I ~ M 9 1 N ~ I ~ E ~ ~ ~ ~ - D ~ - ~ ~ T ~ I ~ ~ I ~ ............................................. ..................... 8 " l l I B L I 4 0 1 P I C ~ I ~ I I I ~ S ~ ~ S - - ~ ~ S S ~ ~ I T ~ I ~ ~ . . . .......................................... r s 1 ~ ~ Y I l I I ~ I v l ~ ~ I m n r ~ ~ - - r s ~ s " " - ~ - ~ ~ f f i w ~ w

~ ~ I c ~ ~ G ~ ~ ~ A ~ I ~ I ~ - ~ G I ~ ~ D ~ ~ ~ ~ o I ~ ~ A I ~ L T ~

3 ........................................................................................... ~ ~ I I C I I C I ~ ~ ~ ~ ~ ~ ~ ~ I ~ ~ N I ~ ~ O I ~ ~ A I ~ ~ ................................................................................................. ~ I I C I F C I ~ ~ ~ Y ~ I I ~ ~ I P U ~ I V D ~ ~ I ~ S ? N ~ T ~ ~ ...................................................................... 1 I O O U ~ ~ ~ ~ S A ~ r x ~ x x ~ ~ x ~ x ~ " ~ ~ I ~ s r ~ ~ ~ ~ ~ s

4 5 6 I ~ U ~ ~ I I S I ~ ~ U ~ M ~ ~ ~ T - I ~ ~ S ~ ~ L ~ ~ ~ ~ I ~ .................................................................................... I ~ I I l I ~ ~ A I F E I L C . Y S O L a O W I ? a V 1 Y I ( P O ............................................................................... I ~ U S 1 r ~ M G m E I M l P I ~ ~ ~ M ~ ~ ~ l ~ w ~ ~ ~ ~ ~ ~ I ~ s ................ ................................... I N O C I M ~ ~ I I S I H I ~ ~ ~ ~ ~ ~ ~ I ~ ~ ~ ~ I D ~ I ~ ~ ~ ~ U ~ 7 8 I I E ~ ~ ~ ~ ~ I ~ ~ ~ I ~ I ~ ~ ~ I ~ ~ ~ ~ L I ~ G I ~ I ~ I ~ ~ ~ ~ ~ G I G .......................................................................... V I E ~ I G L I ~ ~ p L m l ~ N ~ ~ D U I ~ ~ I ~ L I ~ ~ A ~ ~ Y S L I ~ L ~ ~ ......................................................................... ~ I I U . L L L Y ~ ~ ~ I I ~ ~ M ~ I C ~ C I M W M D S I G ................................................ V - E I L I G L S ~ I ~ I Y ~ ~ ~ S ~ ~ ~ I ~ l ~ ~ I l ~ ~ ~ L ~ ~ ~ ~ 7 10 11

1 2 3 4 5 6 7

(rM) FIG. 5. Effect of GABA concentration on GABA uptake into

Xenopus oocytes injected with synthetic mRNA of GAT4. The experimental conditions were as described in Fig. 2.

While the transport of /?-alanine by GAT3 and GAT4 is strongly inhibited by GABA, the taurine transporter is insen- sitive to GABA. On the other hand taurine is a potent inhib- itor of /?-alanine transport by the taurine transporter, but relatively high concentrations of taurine are required to in- hibit /?-alanine transport by GAT3 and GAT4.

The pharmacology of the four GABA transporters is influ- enced by their substrate specificity. Taurine and its analogs BGPA, DAPA, and hypotaurine specifically inhibited GAT3 and GAT4 that transport taurine. On the other hand GAT2 that transports betaine was the most sensitive to betaine and its analogs phloretin and quinidine. The two high affinity GABA transporters GATl and GAT4 were the most sensitive t o guvacine and nipecotic acid. Therefore, GABA transporters might not simply be divided into neuronal and glial subtypes according to their pharmacological properties, and sensitivity of GABA uptake to &alanine is only first approximation for glial GABA transporters. GATl is considered to be neuronal because of its relatively high affinity for GABA and its sen- sitivity to nipecotic acid (12). Recent studies on the localiza-

TABLE I Inhibitor sensitivity of the four different GABA transporters

expressed in Xenopus oocytes The transport assay with the various GABA transporters was

performed as described under "Experimental Procedures." The dif- ferent chemicals were present at the indicated concentrations during the preincubation and assay periods.

Addition Conc. GABA transport activity (7% of control)

GATl GAT2 GAT3 GAT4

m None 100% 2 1 O O f 11 100 f 4 1 O O f 7 j3-Alanine 10 1 0 8 f 6 106 f 15 5 9 + 3 7 4 f 11

100 9 8 2 3 100 f 8 38 + 3 22 f 2 500 9 2 f 5 5 8 f 5 1 5 2 2 S f 1

Betaine 200 101 f 9 5 5 f 10 102 f 6 1 1 O f 10 500 107 f 2 25 + 2 104 f 5 103 f 5

DABA 100 92 f 5 1 0 2 f 10 97 f 7 7 6 f 7 500 49, 3 61 f 10 82 + 5 5 7 k 5

Guvacine 100 34 f 3 101 + 6 62 f 5 37 f 4 500 1 0 2 1 83 f 11 60 + 6 1 6 f 3

Nipecotic acid 100 2 3 f 2 104 f 9 69 f 3 46 f 4 500 S f 1 92 f 15 61 f 10 14 f 2

Phloretin 100 6 5 f 4 27 f 4 60 f 7 7 0 f 10

j3GPA Quinidine 100 99 f 13 33 + 5 6 4 f 3 94 + 15

DAPA 100 8 0 f 14 7 9 f 16 18f 2 9 f 1 100 9 3 f 6 8 9 f 9 4 6 f 7 S f 1

Hypotaurine 100 91 f 11 9 0 + 6 47 + 9 15 f 2

tion of its mRNA by in situ hybridization and immunocyto- chemistry with antibodies raised against this transporter con- firmed its function in neurons (30, 31).' However, the other GABA transporters may also contribute to the termination of neurotransmission by GABA. The affinity of GAT4 for GABA is about %fold greater than that of GAT1, and therefore it may remove GABA into much lower concentrations. On the other hand the rate of reuptake in the synaptic clefts is

* Q.-R. Liu, unpublished results.

Four Diffcrpnt GARA Transpor t~rs 21 11

- 5 k b

FIG. 6. Localization of GAT4 mRNA in various tissues. Northern hybridization was performed as described under "Experi- mental Procedures." About 40 g of total RNA were applied to each well. The equal amount of RNA in each well was verified hy staining with ethidium bromide. The size of the transcripts was determined I)y RNA standards. The source of RNA was as follows: I ) liver, 2) kidney, 3 ) cerehellum, 4 ) cerebral cortex, T i ) brain stem, and 6 ) the rest of the hrain.

1 2 3 4 5 6 7 8 9

- 2.2 kb

FIG. 7. The presence of GAT3 mRNA in various adult tis- sues and developing neonatal brain. The experimental conditions were as descrihed in Fig. 6. The source of RNA was as follows: I ) adult liver, 2) adult kidney, 3 ) adult cerebellum, 4 ) adult cerebral cortex, 5 ) adult brain stem, 6 ) the rest of the adult brain, 7) whole neonatal brain, 8) whole brain 2 days after hirth, and 9) whole brain 4 days after birth.

dependent on the amounts of the transporters and their turnover. So far we have no precise information on these two parameters for the various GARA transporters. In addition to t.heir function as GARA transporters, GAT3 and GAT4 can transport &alanine at relatively high rates (Fig. 2). Therefore, the transport. of GARA may be modulated by @-alanine and the transport of 8-alanine may be controlled by the presence of GARA. Taurine that is present at high concentrations in the brain (32, 33) , may also play a role in the regulation of

GARA and @-alanine transport in parts of the brain that GAT.? and GAT4 prevail. Since taurine is present in the brain at. millimolar concentrations it may be transported by GAT3 and GAT4 regardless of their high K , for taurine. The inter- play of the four different GARA and taurine transporters may elucidate the fine mechanisms of neurotransmission by GABA, /3-alanine, and taurine.

The expression of the various GARA transporters was studied by Northern hybridization of the transporters cDNA with RNA isolated from different parts of the brain. As was shown before, GATl is specific to the central nervous system, and it could not be detected in other tissues such as kidney and liver (1.7, 15). Transcripts of GATl could he found in all brain parts in recent studies with in situ hybridization' and immunocytochemistry revealed that it is primarily a neuronal transporter (30, 31). Transcripts of GAT2 are also present in all brain parts, kidney and liver (16). Similar but not identical transporter of canine kidney was found to be regulated by hypertonicity and suggested to be functional in betaine trans- port in the kidney (34). The pharmacolom of GAT2 is con- sistent with its being a glial GARA transporter (16). Like GATl the transcript . of GAT4 is specific for the central nervous system, and it is not present in kidney or liver. However, GAT4 mRNA was more concentrated in brain stem in comparison with cerebellum and cerebral cortex. Its high affinity to GARA suggests an important role for GAT4 in terminat,ion of GARA transmission in the central nervous system. Initially, we detected transcripts of GAT.? only in kidney and liver. This was surprising because the cDNA encoding GAT3 was very abundant in the neonatal mouse library. Indeed northern hybridization (Fig. 7 ) revealed that its expression is developmentally regulated, and its transcript is much more abundant in the brains of newly born mice than in the adult brain. The interchange of GARA transporters expression may play an important role in the development of the GARAergic neuronal functions in the brains of young mice.

The emerging family of neurotransmitter transporters can now be divided into subfamilies that share not only similar substrates but also extensive amino acid sequence identity. The subfamily of GARA transporters contains the transport- ers G A T l (12, 13, 151, GAT2 ( I6) , GAT.7, GAT4, choline (3.5), and taurine (28). Sequence identity among all of these trans- porters suggests that they evolved from a common ancestor highly related to the taurine transporter (28). Another subfamily of catecholamine transporters contains the norep- inephrine, serotonin, and dopamine transporters (17-21). The glycine (23, 24) and proline (22) transporters may belong to a subfamily of amino acid transporters. The evolutionary dis- tance revealed hy the percentage of sequence identity among the subfamilies is ahout equal. Moreover, a common exon- intron junction was identified at the most conserved region between the first and second transmembrane domains in all t h e genes encoding neurotransmitter transporters (1.5). Therefore, it, seems likely that the subfamilies evolved from a single ancestral gene, and they diverged at about the same time. The challenge now is to learn about the mechanism of action and regulation of the various neurotransmitter trans- porters as well as their specific function in the central nervous system and in other tissues.

Acknoulkrd~mrnls-We thank Dr. John Reev~s nnd Dr. Nnt firr~t for critical reading of the manuscript.

REFERESCES 1 . Ivernen. L. L.. and Kelly. -1. S. I19751 Hiochrvn f'hormorol 24, 933-93R 2. Martin. I). (1976) in ,Vcrc,our .Yytcrn Funrfwn ( ( 'hnw. 7'. Si.. nnd 'lower.

I ) . H.. edn) pp. 357-334, Ibvrn. S r w l ' o r k

2112 Four Different GABA Transporters 3. 4. 5. 6. 7. 8. 9.

10.

12. 11.

13.

15. 14.

16.

17.

Iversen, L. L., and Neal, M. J. (1968) J. Neurochem. 15,1141-1149

Bond, P. A. (1973) J. Neurochem. 20,511-517 Iversen, L. L., and Johnston, G . A. R. (1971) J. Neurochem. 18,1939-1950

Levi, G., and Raiteri, M. (1973) Life Sei. 12, 81-88 Martin, D. L. (1973) J. Neurochem. 21 , 345-356 Wood, J. D., and Sidhu, H. S. (1986) J. Neurochem. 46,739-744 Radian, R., Bendahan, A,, and Kanner, B. I. (1986) J. B i d . Chem. 261,

Kanner, B. I., and Bendahan, A. (1990) Proc. Natl. Acad. Sci. U. S . A. 87, 15437-15441

2550-2554 Kanner, B. I. (1989) Curr. Opin. Cell Biol. 1 , 735-738 Gaustella, J., Nelson, N., Nelson, H., Czyzyk, L., Keynan, S., Miedel, M.

C., Davidson, N., Lester, H. A,, and Kanner, B. I. (1990) Science 249, 1 ?,nx-lmG

Nelson, H., Mandiyan, S., and Nelson, N. (1990) FEBS Lett. 269, 181-

Olsen, R. W., and Tobin, A. J. (1990) FASEB J. 4, 1469-1480 Liu. Q.-R.. Mandivan. S.. Nelson. H.. and Nelson. N. (1992) Proc. Natl.

- I - - - - - -

184

Acid. Sci. U. S. A. &, 6639-6643 '

(1992) J. Biol. Chem.. 267. 17491-17493

, . ,

Lopez-Corcuera, B., Liu, Q.-R., Mandiyan, S., Nelson, H., and Nelson, N.

Pacholczyk, " . T., Blakely, R. D., and Amara, S. G . (1991) Nature 350,350-

22. Fremeau, R. T., Jr., Caron, M. G., and Blakely, R. D. (1992) Neuron 8 , M., Prince, H. K., and Bradley, C. C. (1991) Nature 354,66-69

23. Smith, K. E., Borden, L. A., Hartig, P. R., Branchek, T., and Weinshank, 915-926

24. Liu, Q. -R., Nelson, H., Mandiyan, S., L6pez-Corcuera, B., and Nelson, N. R. L. (1992) Neuron 8,927-935

25. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: a (1992) FEBS Lett. 305,110-114

Laboratory Manual, Second ed., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

~~

26. Henikoff, S. (1984) Gene (Amst.) 28,351-359 27. Church, G. M., and Gilbert, W. (1984) Proc. Natl. Acad. Sei. U. S. A. 81,

1qq1-1 qq.5 28.

29.

30.

31.

32. 33.

Liu, Q.-R., Lopez-Corcuera, B., Nelson, H., Mandiyan, S., and Nelson, N.

Larsson, O. M., Griffths, R., Allen, I. C., and Schousboe, A. (1986) J.

Radian, R., Ottersen, 0. P., Storm-Mathisen, J., Castel, M., and Kanner,

Mabjeesh, N. J., Frese, M., Rauen, T., Jeserich, G., and Kanner, B. I. (1992)

Jacobsen, J. G., and Smith, L. N. (1968) Physiol. Reu. 48,424-511 Wright. C. E.. Tallan. H. H.. and Lin. Y. Y. (1986) Annu. Reu. Biochem.

"" "I-

(1992) Proc. Natl. Acad. Sci. U. S. A. , in press

Neurochem. 47,426-432

B. I. (1990) J. Neurosci. 10,1319-1330

FEES Lett. 299,99-102

18. Shimada. s., Kitayama, s., Lin, c.-L., Patel, A., Nanthakumar, E., G r e w , 34. Yamauchi, A,, (Jchida, s., Kwon, H. M., Preston, A. S., Robey, R. B.,

19. Kilty, J. E., Lorang, D., and Amara, S. G . (1991) Science 264,578-579 P., Kuhar, M., and Uhl, G. (1991) Science 254,576-578 Garcia-Perez, A,, Burg, M. B., and Handler, J. S. (1992) J. Biol. Chem.

20. Hoffman, B. J., M e w , E., and Brownstein, M. J. (1991) Science 254,579- 35. Mayser, W., Schloss, P., and Betz, H. (1992) FEBS Lett. 305, 31-36 267,649-652

21. Blakely, R. D., Berson, H. E., Fremeau, R. T., Jr., Caron, M. G., Peek, M. $5'. A. 74,5463-5467

234 56,427-453 . .

580 36. San er, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sei.