Electrogenic L-glutamate uptake in Xenopus laevis oocytes expressing a cloned rat brain L-glutamate/L-aspartate transporter (GLAST-1).

The transport of L-glutamate into Xenopus laevis oocytes expressing the cloned L-glutamate/L-aspartate transporter (GLAST-1) from rat brain was studied using the voltage clamp technique. At a holding potential of -90 mV, a bath application of 100 microM L-glutamate induced an inward current (IGLAST) with an amplitude ranging from -5 to -30 nA. IGLAST did not require extracellular Ca2+, Mg2+, or Cl-, was larger at negative potentials, and did not reverse up to +80 mV. The current was dependent on external L-glutamate and Na+ with half-maximal amplitudes at 11 microM L-glutamate and 41 mM Na+. IGLAST saturated at 100 microM L-glutamate and 80 mM Na+. The Hill coefficient for Na+ and L-glutamate was 3.3 and 1.3, respectively, suggesting that 3 Na+ accompany the transport of 1 L-glutamate molecule. At low [Na+]o, IGLAST was enhanced by reducing [K+]o, an indication for the countertransport of K+. Reducing external pH from 7.4 to 6.0 did not change the amplitude of IGLAST. This argues against a glutamate/proton cotransport. The results provide evidence for GLAST-1 carrying out a high affinity, sodium-dependent L-glutamate transport with a proposed stoichiometry of 3 Na+, 1 L-glutamate-/1 K+.

t-Glutamate mediates neurotransmission at the vast majority of excitatory synapses in mammalian central nervous system (for review see Collingridge and Lester (1988)). The excitatory action of this amino acid is involved in complex physiological processes like learning and the establishment of memory (Monaghan et al., 1989). On the other hand, L-glutamate has potent excitatoxic properties leading to neurodegeneration and brain damage (for review see Choi (1988)).
To terminate the excitatory signal, L-glutamate is rapidly removed from the synaptic clefts. Sodium-dependent L-glutamate transporters residing in the plasma membranes of the presynaptic nerve ending and surrounding glia cells are thought to be responsible for this transport process (Flott and Seifert, 1991). The structural specificity and ion dependence of this transport system have been investigated by * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. [14Clglutamate uptake studies using brain slices (Balcar and Johnston, 19721, synaptosomes (Bennett et al., 1973), isolated cells (Gordon and Balazs, 1983;Flott and Seifert, 1991), and cell lines (Waniewski and Martin, 1984) for more than two decades. Further information about the mechanism of transport is obtained from electrophysiological measurements. It has been shown in Muller cells from salamander retina that L-glutamate evokes an inward current, which is most probably associated with L-glutamate uptake into the cytosol (Brew and Attwell, 1987;Barbour et al., 1988Barbour et al., , 1991Schwartz and Tachibana, 1990). This electrogenic uptake of L-glutamate was dependent on external Na+ and internal K+ (Barbour et al. (1988(Barbour et al. ( , 1991 but see Schwartz and Tachibana (1990)).
The recent cloning of cDNAs coding for three rodent L-glutamate transporters offers for the first time the opportunity to investigate independently single members of this family upon expression in heterologous systems (Storck et al., 1992;Pines et al., 1992;Kanai and Hediger, 1992). Using electrophysiological techniques we describe important functional properties of the cloned L-glutamateh-aspartate transporter (GUST-1)' from rat brain expressed in Xenopus laevis oocytes. We demonstrate that the cloned transporter is electrogenic and has a high affinity for L-glutamate. The transport of L-glutamate depends on the transmembrane gradients of sodium and potassium ions. There is no evidence for the cotransport of protons.

MATERIALS AND METHODS
20-72 h after injection of GUST-1 cRNA (Storck et al., 1992) the oocytes were superfused by gravity flow with a solution composed of (mM): NaCl(1201, CaCI2 (1.81, KC1 (2), HEPES-Tris (101, pH 7.4. Oocytes were voltage-clamped using a two-electrode voltage clamp amplifier (Warner Instruments Corp., Hamden, CT). Microelectrodes filled with 3 M KC1 had resistances ranging from 1 to 2 megaohms. The software and hardware package IS01 (MFK, Frankfurt, Germany) was used to generate the voltage commands to acquire and evaluate the data. Data were filtered at 10 Hz (-3 db) and digitized at 100 Hz. Assuming that L-glutamate has no other electrical effects apart from activating the transporter, IGwT can be obtained as the difference of the membrane current before and during application of L-glutamate.
In experiments in which the sodium or chloride concentrations of the bath solution were vaned, these were equimolarly substituted by choline or methanesulfonate, respectively. All measurements were done at room temperature (20-22 "C). Data are expressed if not otherwise noted as mean * S.D. All chemicals were purchased from Sigma.

The Glutamate-induced Inward Current Is Generated by the
Activity of GUST-1-Bath application of 100 PM L-glutamate induced an inward current into Xenopus oocytes (stage V-m) that peaked within 2 s to a plateau of -7 nA and remained stable up to 2 h as long as L-glutamate was present in the bath with less than a 20% decline in amplitude (Fig. lA). Upon washout of L-glutamate the inward current disappeared. NO inward currents were observed upon L-glutamate application in oocytes injected with water ( n = 8, per batch) or in non-injected oocytes ( n = 5, per batch).
The L-glutamate-induced current did not depend on extracellular Ca2+, M e , or C1-, suggesting that inward current did not flow through t-glutamate-gated channels. Under control conditions 100 VM L-glutamate induced a current with a mean am- plitude of -7.2 * 1.7 nA(n = 6). Reducing [Ca2+l, (substituted by 1.8 mM Mg2+ + 1 mM EGTA) resulted in an inward current of similar amplitude (-7.6 * 2.5 nA), and likewise changing [Cl-I, from 119.6 to 4.6 mM left the inward current with an amplitude of -7.2 f 1.8 nA unimpaired.

Electrogenic Cloned Rat
Using ramp depolarizations (within 10 s from -100 to +80 mV) the L-glutamate-induced current is larger at negative potentials and did not reverse up to +80 mV (Fig. 1B). For unselective cation channels opened by L-glutamate the reversal potential should be around 0 mV (for review see Monaghan (1989)). Further evidence for the inward current reflecting the L-glutamate uptake into the oocyte (ZG-T) came from the application of glutamate agonists (trans-l-amino-1,3-cyclopentane dicarboxylate, N-methybaspartate, and kainate). They completely failed to induce currents in oocytes injected with The amplitude of Z C~~T depended on the extracellular Lglutamate concentration. Fig. 1C shows the concentration response curve obtained from 12 different oocytes at a holding potential of -90 mV. In order to compare the different amplitudes OfzGLAST for each oocyte the amplitudes were normalized to the maximal current induced with 316 p~ t-glutamate. When plotted as a function of the L-glutamate concentration the data fit with a curve obeying the equation, GLAST-1 cRNA (n = 7).
is the normalized maximum current amplitude, glutamate] the L-glutamate concentration in the bath, K, the apparent affinity constant, and n the Hill coefficient. The best fit was obtained with a K , of 11 J~M and n = 1.3. This result is compatible with the hypothesis that L-glutamate interacts with one high affinity binding site per transporter.
Dependence of Z G~~T on External Zons-Exchanging extracellular Na' with choline' completely blocked induced by bath application of 15 VM L-glutamate. At [Na'], larger than 10 mM the current rose steeply and saturated at [Na'J, greater than 90 mM (Fig. 2 A ) . The relation between the amplitude of ZCLAsT and [Na+l, could be best described by the following equation. Electrogenic Cloned Rat Brain GUST-1 amplitude of -1.7 nA (Fig. 2B, left). Under these conditions, however, the omission of K+ from the bath solution produced dramatic changes in ZGLMT. Applying L-glutamate induced a current with an amplitude of -7.2 nA. In five similar experiments at reduced [K' I, Z G u s T was enhanced by a factor of 3.5 * 0.8. Changing back to control conditions (120 m M Na+, 2 m M K+) resulted in an inward current with an amplitude of -8.5 nA ( Fig. 2 B , right).
Reducing extracellular pH has no effect on the amplitude of ZG-T.
At pH 7.4 100 VM L-glutamate induced a current with an amplitude of -21.2 nA (Fig. 2C, left). Changing pH to 6.0 (10 m~ MES buffer) has no appreciable effect on the amplitude of Z C~~T (-22.3 nA, Fig. 2C, right). This finding suggests that GLAST-1 does not transport protons together with L-glutamate into the cell.

DISCUSSION
Our results demonstrate for the first time that the GLAST-1-induced transport of L-glutamate in X. laevis oocytes is electrogenic. Application of 100 PM t-glutamate induced an inward that is independent of extracellular Ca2+, Mg2+, or C1-. This suggests that the current did not result from the opening of L-glutamate-gated channels but reflects the L-glutamate uptake into the oocytes. The current flowing through L-glutamategated unselective cation channels has a reversal potential around 0 mV (for review see Monaghan et al. (1989)). In contrast, the current studied here did not reverse up to +80 mV. A similar voltage dependence has been shown for the glutamate uptake into glial cells of the salamander retina (Barbour et al., 1988(Barbour et al., , 1991Schwartz and Tachibana, 1990).
At a holding potential of -90 mV the half-maximal current was obtained at a L-glutamate concentration of 11 PM. This value is lower than the value estimated from tracer flux measurements (77 * 27 p~ (Storck et d., 1992)) but similar to the 6.9 PM reported for a cloned intestinal L-glutamate transporter (Kanai and Hediger, 1992) or the 5-20 PM reported for salamander glial cells (Barbour et al., 1991). The low K,,, of 11 VM indicated that the GLAST-1 from rat brain has a high affinity for L-glutamate.
There is evidence that the transport of L-glutamate depends on the presence of extracellular Na' (for review see Nicholls and Attwell(1990)). Also the electrogenic uptake of L-glutamate into glial cells depends on the presence of extracellular Na' with half-maximal inhibition at 43-50 m M (Barbour et al., 1991;Schwartz and Tachibana, 19901, with Hill coefficients between 2 and 3. Our results fit best with a curve showing half-maximal inhibition at 41 m M [Na+], with a Hill coefficient of 3.3, suggesting that the transport of L-glutamate might go along with the binding of 3 Na'. These results cannot exclude the possibility that only 2 Na+ are transported in countertransport of pH changing anions as proposed by Bouvier et al. (1992).
From tracer flux studies it is evident that the uptake of L-glutamate depends on [K' ], (Kanner and Sharon, 1978;Nicholls and Attwell, 1990). However, voltage clamp experiments gave conflicting results. Barbour et al. (1991) reported that K+ is necessary for the L-glutamate-induced current, while in the same preparation Schwartz and Tachibana (1990) did not find any influence of K' . Our finding that ZGLMT is unim-paired by high [K+l, at saturating "physiological" [Na' ], does not exclude K+ as a countertransported ion. With 120 m~ Na+ and 100 p~ L-glutamate the transporter is maximally stimulated so that increasing [K+], is not sufficient to decrease the high transport rate. The situation is different when the transport rate is not maximally stimulated, e.g. at reduced ma+],. Now a reduction of [K+l, dramatically increases ZGWT suggesting that K+ is countertransported. Since the inward current associated with the uptake of L-glutamate requires the translocation of 1 netto positive charge, at least 3 Na+ per molecule of L-glutamate have to be transported into the oocyte 1 in countertransport of 1 K+. Erecinska et al. (1983) have reported that protons were transported together with aspartate whereas according to Gazzola et al. (1981) L-glutamate is transported as an anion without a proton. Using a pK, for the y-carboxyl group of the Lglutamate group of 4.31, one can calculate that at a pH of 7.4 the ratio of protonated to deprotonated L-glutamate is approximately 1:1230, i.e. when adding 100 p~ L-glutamate to the bath 0.08 PM is protonated. At pH 6.0 this relation is 1:49. Assuming that L-glutamate is transported as a zwitterion the amplitude Z G u s T should dramatically increase at acid pH because the amount of the protonated form has increased by a factor of 25. If L-glutamate is transported as an anion, there should be no appreciable influence on the amplitude of the current, because the amount of the protonated L-glutamate decreases only from 100 to 98 PM. Since acid pH has no detectable influence on the amplitude of ZGLAsT it is likely that t-glutamate is transported as a negatively charged molecule. This hypothesis is in accordance with the result reported by Schwartz and Tachibana (1990) in salamander glial cells.
In this study we have characterized basic functional properties of the GUST-1 cloned from rat brain and expressed in X.
laevis oocytes. Our results provide evidence that GLAST-1 is a high affinity sodium-dependent L-glutamate transporter with a proposed stoichiometry of 3 Na+, 1 L-glutamate71 K' .