Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2004 Apr 23;557(Pt 3):747–759. doi: 10.1113/jphysiol.2004.062521

New inhibitors for the neutral amino acid transporter ASCT2 reveal its Na+-dependent anion leak

Christof Grewer 1, Eva Grabsch 2
PMCID: PMC1665628  PMID: 15107471

Abstract

The neutral amino acid transporter ASCT2 catalyses uncoupled anion flux across the cell membrane in the presence of transported substrates, such as alanine. Here, we report that ASCT2 conducts anions already in the absence of transported substrates through a leak anion-conducting pathway. The properties of this leak anion conductance were studied by electrophysiological recording from ASCT2-expressing HEK293 cells. We found that the leak anion conductance was inhibited by the binding of the newly characterized inhibitors benzylserine and benzylcysteine to ASCT2. These inhibitors competitively prevent binding of transported substrates to ASCT2, suggesting that they bind to the ASCT2 binding site for neutral amino acid substrates. The leak anion conductance exhibits permeation properties that are similar to the substrate-activated anion conductance of ASCT2, preferring hydrophobic anions such as thiocyanate. Inhibition of the leak anion conductance by benzylserine requires the presence of extracellular, but not intracellular Na+. The apparent affinity of ASCT2 for extracellular Na+ was determined as 0.3 mm. Interestingly, a Na+-dependent leak anion conductance with similar properties was previously reported for the related excitatory amino acid transporters (EAATs), suggesting that this leak anion conductance is highly conserved within the EAAT protein family.

The transport of neutral amino acids across membranes of mammalian cells is catalysed by a variety of different transport systems (reviewed in Kilberg et al. 1993; Christensen et al. 1994; Bode, 2001). The alanine–serine–cysteine transporter (ASCT), which belongs to the superfamily of excitatory amino acid transporters (EAATs; Arriza et al. 1993; Utsunomiya-Tate et al. 1996; Broer et al. 1999), is one of these neutral amino acid transport systems. ASCT is specific for small, neutral amino acids, including glutamine in the case of ASCT2 (Arriza et al. 1993; Shafqat et al. 1993). In addition to sequence homology, EAATs and ASCTs share many functional features, most importantly their specificity for Na+ as the major cotransported ion. However, some functional differences were also observed for the two systems. Whereas EAATs counter-transport potassium ions, ASCT function is independent of the intracellular K+ concentration (Zerangue & Kavanaugh, 1996). Furthermore, ASCTs are not able to complete a full transport cycle and are therefore assumed to be locked in the exchange mode (Zerangue & Kavanaugh, 1996; Broer et al. 2000). In this mode, amino acids can only be transported by homo- or heteroexchange with the same or other neutral amino acids.

A characteristic functional feature of excitatory amino acid transporters is their glutamate-gated anion conductance (Wadiche et al. 1995). The magnitude of this anion conductance varies with the subtype of the glutamate transporter. Recently, it was observed that ASCT1 and ASCT2 share this feature with their EAAT counterparts (Zerangue & Kavanaugh, 1996; Broer et al. 2000). Although the characteristics of the anion conductance may be different for ASCT1 and ASCT2 with regard to permeation properties, the anion conductance is activated by the binding of neutral instead of acidic amino acids in both ASCT subtypes. In addition to the anion conductance activated by the transported substrate, EAATs catalyse a leak anion flux (Otis & Jahr, 1998). This leak anion flux is observed as a tonic current that can be inhibited by applying competitive inhibitors of EAATs, such as kainic acid, to the transporter. Both the glutamate-activated anion conductance and the leak anion conductance require the presence of Na+ in the extracellular solution. It is not known whether ASCTs also catalyse a leak anion conductance.

Here, we report the characterization of two new inhibitors for ASCT2. Although these inhibitors bind to ASCT2 only with high micromolar affinity, they reveal new information about the functional properties of ASCT2. Application of the inhibitors to ASCT2-expressing cells in the absence of a neutral amino acid inhibits a tonic leak current that is carried by anions. This leak conductance is sensitive to the extracellular Na+ concentration. Thus, our results indicate that the functional features of the substrate-induced and leak anion conductance are highly conserved within the EAAT and ASCT transporter families. Furthermore, the new inhibitors provide a useful structural scaffold for the design of compounds that bind to ASCT2 with higher affinity.

Methods

The cDNA coding for the rat ASCT2 was kindly provided by S. Bröer (Bröer et al. 1999, 2000) and was subcloned into the EcoR I site of the pBK-CMV vector (Stratagene) for mammalian expression. EAAC1 cloned from rat retina was subcloned into pBK-CMV (Stratagene) as previously described (Rauen et al. 1996; Grewer et al. 2000). The ASCT2 and EAAC1 cDNA constructs were used for transient transfection of subconfluent human embryonic kidney cells (HEK293, ATCC No. CGL 1573) using the calcium phosphate-mediated transfection method as described (Chen & Okayama, 1987). Electrophysiological recordings were performed between days 1 and 3 post-transfection.

ASCT2- and EAAC1-mediated currents were recorded with an Adams & List EPC7 amplifier (HEKA, Lambrecht, Germany) under voltage-clamp conditions in the whole-cell current-recording configuration (Hamill et al. 1981). The typical resistance of the recording electrode was 2–3 MΩ, the series resistance was 5–8 MΩ. Because of the low membrane conductance changes associated with ASCT2 and EAAC1 activation (typically < 5 nS), series resistance (RS) compensation had no effect on the magnitude of the observed currents and therefore was not used. The extracellular bath buffer solution contained (mm): 140 NaCl or NaSCN, 2 MgCl2, 2 CaCl2, and 10 Hepes (pH 7.4/NaOH). For testing the [Na+] dependence of the currents, Na+ in the extracellular solution was replaced with choline. The pipette solution used for back-filling the recording electrode contained (mm): 130 NaSCN or NaCl, 2 MgCl2, 10 EGTA, 10 Hepes, and 10 l-alanine (pH 7.4/NaOH). The high intracellular alanine concentration used served the purpose of saturating the alanine binding site of ASCT2 when it was exposed to the cytoplasm. For recordings with EAAC1, alanine was replaced by l-glutamate. Using this intracellular solution, the transporters are locked in the exchange mode. Thiocyanate was used because it enhances ASCT-associated currents and allows the detection of the anion-conducting mode (Zerangue & Kavanaugh, 1996; Broer et al. 2000). For the investigation of the dependence of currents on the intracellular cation composition the pipette solution contained (mm): 130 KSCN, 2 MgCl2, 10 EGTA, and 10 Hepes (pH 7.4/KOH). For some experiments intracellular Na+ was replaced by N-methylglucamine+. The time course of equilibration of the cell solution with the recording pipette solution after establishing the whole-cell configuration was determined by measuring how fast the new resting potential, which is now mainly determined by the dominant SCN− permeability, was reached. For cells typically used in the current recordings this equilibration took place within 1 min of breaking into the whole-cell configuration. Current recordings were started at 3 min after establishing the whole-cell mode. The currents were amplified with an Adams & List EPC-7 amplifier, low pass filtered at 1–10 kHz (model 3200, Krohn-Hite, Brockton, MA, USA) and digitized with a digitizer board (Digidata 1200, Axon Instruments, Foster City, CA, USA) at a sampling rate of 10–50 kHz which was controlled by software (Axon pCLAMP7). All the experiments were performed at room temperature.

HEK293 cells are reported to have an intrinsic ASCT-like transport activity for neutral amino acids (Matthews et al. 1997), but we were unable to detect substantial alanine-activated anion currents in non-transfected cells. Application of 1 mm alanine to non-transfected cells resulted in anion currents no larger than 3 pA, which is negligible compared to the average of 110 pA in cells expressing recombinant ASCT2 (all at 0 mV and with 140 mm internal SCN−). ASCT2-expressing cells were selected by fluorescence microscopy after coexpression of green fluorescent protein (GFP). The GFP cDNA concentration used for the transfection was kept at 1/3 of the ASCT2 cDNA concentration. Under these conditions, more than 95% of GFP-expressing cells also expressed ASCT2.

Rapid solution exchange was performed as described (Grewer et al. 2000; Watzke et al. 2001). Briefly, substrates and inhibitors were applied to the ASCT2- and EAAC1-expressing cells by means of a quartz tube (opening diameter 350 μm) positioned at a distance of ∼0.5 mm to the cell. The linear flow rate of the solutions emerging from the opening of the tube was approximately 5–10 cm s−1, resulting in typical rise times of the whole-cell current of 30–100 ms (10–90%).

Data evaluation and fitting were performed using Axon pCLAMP8 and Origin software (OriginLab, Northampton, MA, USA). Each experiment was repeated at least 3 times with at least two different cells. The error bars represent the error of the single measurement (mean ± s.d.), unless stated otherwise. Dose–response curves were fitted to a Michaelis–Menten-like relationship. For analysis of the dose dependence of the inhibition of substrate-induced currents by the inhibitors the following equation was used:

graphic file with name tjp0557-0747-m1.jpg (1)

In this equation, I is the current amplitude in the presence, and I0 that in the absence, of inhibitor. Ki is the apparent inhibition constant.

The compounds tested for ASCT2 inhibition, benzylserine and benzylcysteine, were obtained from Bachem Bioscience (Torrance, CA, USA). dl-threo-β-oxybenzylaspartic acid (TBOA) was purchased from Tocris (Ellisville, MO, USA). Glutamate, glutamine, alanine, phenylalanine, salts and other reagents were purchased from Sigma-Aldrich (St Louis, MO, USA).

Results

Application of alanine to ASCT2-expressing HEK293 cells resulted in inwardly directed currents in the presence of alanine, Na+ and SCN− at the intracellular side of the membrane (Fig. 1A and 0 mV transmembrane potential). Under these conditions, ASCT2 operates in the exchange mode. In the absence of intracellular SCN−, which permeates the anion conductance of ASCT2, no alanine-induced currents were observed (data not shown), indicating that the inward current shown in Fig. 1A is carried by SCN−. The alanine-induced currents were dependent on the alanine concentration (Fig. 1B). This concentration dependence followed Michaelis–Menten-like behaviour with a Km of 420 ± 40 μm (n = 4) and an Imax of 190 ± 45 pA (n = 20). In contrast to ASCT1, ASCT2 also transports glutamine (Utsunomiya-Tate et al. 1996). Therefore, we tested whether glutamine generates exchange currents in ASCT2-expressing cells. As shown in Fig. 1B, glutamine elicited currents with the same amplitude as those induced by alanine. Glutamine-induced currents saturated with a Km of 61 ± 2 μm(n = 5). The Km values for glutamine and alanine are in agreement with previously published values of 24 μm (Utsunomiya-Tate et al. 1996), 70 μm (Bröer et al. 1999) and 90 μm (Bröer et al. 2000) for glutamine in ASCT2-expressing Xenopus oocytes, and 100–380 μm for alanine determined for native system ASC (Bass et al. 1981; Franchi-Gazzola et al. 1982; Kimmich et al. 1994). However, our Km of alanine for rat ASCT2 is significantly higher than that reported for recombinant ASCT2 from the mouse of 18 μm (Utsunomiya-Tate et al. 1996). In the latter study, the Km was determined by radiotracer flux methods in Xenopus oocytes. In contrast to our experiments, which were carried out in the presence of 140 mm intracellular Na+ and 10 mm intracellular alanine, oocytes have much lower intracellular Na+ (5 mm) and neutral amino acid concentrations (1 mm for glutamine) (Bröer et al. 2002). We therefore determined the Km for alanine under ionic conditions and at a transmembrane potential (–40 mV) matching those of oocytes and found a value of 250 ± 20 μm. This result shows that at least part of the difference in the observed Km values is caused by different experimental conditions.

Figure 1. Characterization of ASCT2 substrates and inhibitors.

Figure 1

Open in a new tab

All currents shown in this figure were recorded at 0 mV transmembrane potential in the presence of 140 mm intracellular NaSCN and 10 mm intracellular alanine (exchange mode). A, currents induced in HEK293 cells transiently expressing ASCT2 after application of 1 mm alanine, 1 mm benzylserine, 1 mm phenylalanine, and 0.1 mmdl-TBOA. B, dependence of the substrate-induced anion current on the glutamine (•) and alanine (○) concentration (c). The currents were normalized to the current at saturating substrate or inhibitor concentration. The continuous lines represent fits to a Michaelis–Menten relationship. The Km values obtained from the fit are shown in the main text. C, currents at saturating concentrations of the respective compounds, normalized to the current induced by a saturating concentration of glutamate (EAAC1, left) and a saturating concentration of alanine (ASCT2, right). Data were averaged from at least 3 individual experiments from at least 2 different cells.

Next, we tested potential inhibitors of neutral amino acid transport with respect to their activity towards ASCT2. We first tested benzylserine which is a structural analogue of the glutamate transporter inhibitor TBOA, but lacks the β-carboxy function (Fig. 2). Application of 1 mm benzylserine to an ASCT2-expressing cell resulted in an outwardly directed current in the presence of intracellular alanine, Na+ and SCN− (Fig. 1A and 0 mV transmembrane potential), suggesting that the compound is not a transported substrate of ASCT2. The outward current was dependent on the benzylserine concentration, as shown in Fig. 3, and it saturated with a Km value of 0.9 ± 0.4 mm (n = 5). In the absence of intracellular SCN−, application of benzylserine to ASCT2 did not generate any detectable currents (data not shown), indicating that the outward current may be specifically carried by anions (see below). Similar outward currents were found for the other structural TBOA analogue tested, benzylcysteine (Figs 1C and 3). The results obtained with benzylserine and benzylcysteine are summarized in Table 1. To test whether the benzylserine-induced outward current is specific for ASCT2, we performed additional experiments with phenylalanine, which is known to not interact with ASCT2 (Utsunomiya-Tate et al. 1996). Consistent with this fact, application of 1 mm phenylalanine to ASCT2 did not result in detectable currents (Fig. 1A and C). We further tested the effects of the glutamate transporter blocker TBOA on ASCT2. As shown in Fig. 1A and C, no current was generated by the application of 0.1 mm TBOA to ASCT2. The concentration of 0.1 mm is about 500 times the Km value of TBOA at glutamate transporters (Shimamoto et al. 1998; Grewer et al. 2000).

Figure 2. Structural comparison of transported substrates, non-transported amino acids, and inhibitors for EAATs and ASCT2.

Figure 2

Open in a new tab

Phenylalanine, although being neither a transported substrate nor an inhibitor, is included because of its structural similarity to the ASCT2 inhibitors.

Figure 3. Determination of the mechanism of ASCT2 inhibition by benzylserine and benzylcysteine.

Figure 3

Open in a new tab

A, typical currents activated by the application of 0.4 mm alanine (grey bar) in the absence and presence of 0.5 mm and 5 mm benzylcysteine (open bar). The transmembrane potential was 0 mV and the pipette contained 140 mm NaSCN and 10 mm alanine. B, dose–response relationships of benzylserine activation of outward currents in the absence of alanine (•) and inhibition of inward currents activated by 0.5 mm alanine (○). Currents are normalised to the current recorded after application of 0.5 mm alanine in the absence of inhibitor. The lines represent best fits according to a Michaelis–Menten-type relationship (•) and eqn (1) (see Methods, ○). C, inhibition constant Ki for benzylcysteine at different concentrations of alanine. The continuous line represents the result from a linear regression analysis with a slope of 2.1 ± 0.1 and an intercept of 0.74 ± 0.14 mm. The dashed line shows the relationship expected for competitive inhibition according to eqn (2).

Table 1.

Relative maximum currents and inhibition constants for ASCT2 inhibitors determined in the absence and presence of 0.5 mm alanine at Vm= 0 mV under exchange conditions

Benzylserine Benzylcysteine Phe TBOA
Ki (0 mm Ala) (mm)  0.9 ± 0.4 0.78 ± 0.50 – –
Ki(0.5 mm Ala) (mm)  1.9 ± 0.3 1.44 ± 0.3  – –
Imax/Imax(Ala) 0.26 ± 0.10 0.25 ± 0.05 0.02 ± 0.05 0.01 ± 0.001
Open in a new tab

TBOA, dl-threo-β-oxybenzylaspartic acid.

Additional experiments were designed to test whether benzylserine and benzylcysteine are competitive blockers of ASCT2, by determining their effects on alanine-induced currents. As shown in Fig. 3A, benzylcysteine inhibited inward currents elicited by 0.5 mm alanine in a concentration-dependent manner. At the highest concentration tested (5 mm), benzylcysteine generated an outward current even in the presence of alanine, indicating that it competitively inhibited alanine binding to ASCT2. To confirm competitive inhibition, we determined the inhibition constant Ki(S), where S represents transported substrate for benzylcysteine and benzylserine inhibition of alanine-induced currents at different concentrations of alanine. As an example, the concentration dependence of benzylserine inhibition in the presence of 0.5 mm alanine is shown in Fig. 3B. The results for the analysis of benzylcysteine inhibition at four different alanine concentrations are shown in Fig. 3C. Increasing concentrations of alanine linearly shifted the Ki(S) for benzylcysteine inhibition towards higher values, as expected for a competitive mechanism according to the following equation (dashed line in Fig. 3C):

graphic file with name tjp0557-0747-m2.jpg (2)

In this equation, [S] is the concentration of the substrate (alanine), and Ki(0) is the inhibition constant for benzylcysteine inhibition in the absence of alanine. From the slope of the linear relationship of 2.1 ± 0.1 and the known Ki(0) (Table 1), we calculated the Km for alanine to be 370 μm. This value is in good agreement with the Km for alanine determined experimentally (420 μm, see above).

Next, we tested whether the benzylserine-induced ASCT2 conductance is selective for anions. Figure 4 shows the voltage dependence of ASCT2-mediated currents in the presence of the transported substrate alanine (Fig. 4A), and the inhibitor benzylcysteine (Fig. 4B). In the presence of 140 mm SCN− only on the intracellular side and within the voltage range tested (–90 to +60 mV), alanine-induced currents were always inwardly directed, whereas benzylserine-induced currents were always outwardly directed. When the SCN− concentration gradient across the membrane was reversed, currents induced by the application of both alanine and benzylserine reversed their direction at all voltages. Finally, when 140 mm SCN− was present on both sides of the membrane, the current–voltage relationship was linear and showed a reversal potential near 0 mV, which was expected from a predominantly SCN−-conducting system. These results show that alanine activates an anion conductance, whereas benzylserine inhibits a tonic leak anion conductance.

Figure 4. ASCT2-mediated currents are carried by anions.

Figure 4

Open in a new tab

A, voltage dependence of currents generated by the application of 1 mm benzylserine to ASCT2-expressing cells in the presence of 140 mm SCN− only in the intracellular solution (▪), only in the extracellular solution (□), and in both solutions (▴). In all experiments the SCN−-free solution contained Cl− as the main anion. B, similar experiment to that in A, but currents were evoked by the application of 1 mm alanine to the cells.

In additional experiments, we tested the anion selectivity of the ASCT2 substrate-induced conductance and the leak conductance. Figure 5A shows the voltage dependence of the alanine-induced current in the presence of the anions SCN−, NO3− and Cl− in the extracellular solution. Within the range of membrane potentials studied (–90 to +60 mV), the currents for SCN− and NO3− were outwardly directed and did not reverse. At all voltages NO3− carried smaller currents than SCN−. Within experimental error, no current was detected in the presence of Cl− as the sole extracellular anion, confirming results from a previous report that showed that the permeability of ASCT2 for Cl− is very low (Bröer et al. 2000). Inward currents elicited by the application of 2 mm benzylserine to the same cell showed a similar dependence on the nature of the extracellular anion with I(SCN−) > I(NO3−) > I(Cl−) (Fig. 5B), indicating similar permeation properties of the substrate-induced and the leak anion conductance. In order to determine the permeability ratios between SCN− and NO3−, we measured the reversal potential in the presence of NO3− in the extracellular and SCN− in the intracellular solution, as shown in Fig. 5C. When the anion conductance was activated by alanine, the reversal potential was Vrev= 50 ± 7 mV (n = 5). A similar value of Vrev (52 ± 10 mV) was found for glutamine as a substrate. Inhibiting the leak anion conductance by the application of benzylserine resulted in a Vrev of 53 ± 5 mV (n = 6). Using the Goldman–Hodgkin–Katz zero current equation (Hille, 2001), we calculated the permeability ratio P(SCN−)/P(NO3−) to be 7.0 for the substrate-induced anion conductance, and 7.9 for the leak anion conductance. When the same experiment was repeated in the presence of extracellular Cl− instead of NO3−, the current did not reverse within the accessible range of membrane potentials (up to –90 mV). Therefore, we can only estimate a lower limit of P(SCN−)/P(Cl−) of 34 for both the substrate-induced and the leak anion conductance. The permeability ratios for the substrate-induced anion conductance agree well with values previously published for the substrate-dependent ASCT1 anion conductance (Zerangue & Kavanaugh, 1996) and the glutamate transporter anion conductance (Wadiche & Kavanaugh, 1998).

Figure 5. Determination of the ion selectivity of the ASCT2 anion conductance.

Figure 5

Open in a new tab

A, voltage dependence of current generated by the application of 1 mm benzylserine to ASCT2-expressing cells in the presence of SCN− (•), NO3− (○), and Cl− (▵) at concentrations of 140 mm at the extracellular side of the membrane. The intracellular solution contained 130 mm Cl−. B, similar experiment to that in A, but currents were evoked by the application of alanine to the cells. C, alanine-induced (•) and benzylserine-induced (○) currents in the presence of 130 mm intracellular SCN− and 140 mm extracellular NO3−.

Activation of the substrate-dependent anion conductance of ASCTs requires the presence of extracellular Na+ (Zerangue & Kavanaugh, 1996; Bröer et al. 2000). Consistently, alanine-induced currents were dependent on the extracellular Na+ concentration and were abolished in the absence of extracellular Na+ (data not shown). The Km for Na+ was determined to 18.6 ± 1.8 mm (n = 3). In addition, we tested whether the presence of sodium ions is also necessary to activate the ASCT2 leak anion conductance. Figure 6 shows the dependence of the current induced by the application of 1 mm benzylserine on the extracellular sodium concentration. In the absence of extracellular Na+, no detectable current was induced by benzylserine. In the presence of increasing [Na+], currents increased to reach a maximum at about 5 mm Na+ (Fig. 6A). From the data in Fig. 5, an apparent Km for Na+ activation of the current of 0.3 ± 0.1 mm (n = 4) was obtained. The voltage dependence of the anion current as a function of Na+ concentration is shown in Fig. 6B. At all concentrations of Na+, the current is outwardly directed, further demonstrating that it is carried by anions and not by Na+.

Figure 6. The ASCT2 leak anion conductance depends on the extracellular Na+ concentration.

Figure 6

Open in a new tab

A, the alanine-induced current is plotted as a function of the extracellular Na+ concentration. The line represents a fit to a Michaelis–Menten-like relationship with a Km of 0.25 ± 0.06 mm.B, voltage dependence of the alanine-induced current at different concentrations of extracellular Na+: 140 mm (○), 1 mm (▴) and 0 mm (▵). The intracellular solution contained 130 mm NaSCN and 10 mm alanine.

So far, anion currents were measured in the presence of Na+ as the sole intracellular cation (Na+-exchange mode). Next, we determined whether intracellular Na+ is required to generate ASCT2 anion current. When Na+ was replaced by intracellular K+, application of 1 mm benzylcysteine to ASCT2-expressing cells still inhibited the leak anion current, as shown in Fig. 7A. The benzylcysteine-dependent outward current was of about the same amplitude as that determined in the presence of intracellular Na+ (Fig. 7C, SCN−-containing pipette solution). In contrast to the results obtained in the homoexchange mode, application of alanine generates an outwardly directed current in the presence of internal K+ and SCN− (Fig. 7A). This result indicates that under these ionic conditions alanine inhibits the leak anion conductance, but does not activate the substrate-induced anion conductance. To test whether the alanine-induced outward current is carried by SCN−, we determined the voltage dependence of this current (Fig. 6B). Up to 60 mV the current is outwardly directed and does not reverse direction, consistent with SCN− outflow under zero-trans conditions. In additional experiments we substituted intracellular Na+ with NMG+ (N-methylglucamine) which is inert with respect to ASCT2. As shown in Fig. 7C, benzylcysteine was still able to inhibit the leak anion current under these conditions. Together, these results suggest that the nature of the intracellular cation is not important for the inhibitory effect of benzylcysteine.

Figure 7. Requirements of intracellular ions for activation of the anion conductance.

Figure 7

Open in a new tab

A, typical currents activated by the application of 1 mm alanine (left panel) and 1 mm benzylcysteine (right panel) in the presence of 130 mm KSCN in the intracellular solution (Vm = 0 mV). B, voltage dependence of currents as shown in A for alanine (○) and benzylcysteine (•). C, currents at 1 mm alanine (open columns) and 1 mm benzylcysteine (grey columns) in the presence of 140 mm of the intracellular cation indicated at the bottom of the graph (Vm= 0 mV). The currents were normalized to the current measured at 1 mm alanine. The extracellular solution contained 140 mm SCN−. Data were averaged from 3 individual experiments from 2 different cells.

Finally, we characterized the inhibitors with respect to their effect on the glutamate transporter EAAC1. As shown in Fig. 8A, application of 100 μm glutamate to EAAC1-expressing cells induced large inward currents in the presence of intracellular SCN−. In contrast, application of 100 μm TBOA to EAAC1 elicited outward currents (Watzke & Grewer, 2001). No currents were observed in the same cell after application of 1 mm benzylserine (Imax/Imax(Glu) = 0.008 ± 0.001, n = 3, Fig. 8C), demonstrating that benzylserine does not interact with EAAC1 at concentrations up to 1 mm. At 5 mm a small outward current was induced by benzylserine (Imax/Imax(Glu) = 0.05 ± 0.01, n = 3), indicating that the compound does bind to EAAC1, but with very low affinity. To further test the effects of benzylserine on EAAC1, we determined whether benzylserine inhibits glutamate-induced currents. As shown in Fig. 8D, inhibition of anion currents induced by the application of 10 μm glutamate to EAAC1-expressing cells was observed at benzylserine concentrations of 2 mm and 5 mm. From these data, the Ki for inhibition was estimated to be about 11 mm.

Figure 8. Characterization of the effect of benzylserine on the glutamate transporter EAAC1.

Figure 8

Open in a new tab

Typical currents activated by the application of 100 μm glutamate (A), 100 μm TBOA (B), and 1 mm benzylserine (C) to an EAAC1-expressing HEK293 cell. The time of application of the respective compound is indicated by the bar. The currents were recorded in the presence of 130 mm intracellular NaSCN and 10 mm intracellular glutamate at Vm= 0 mV. D, ratio of the current in the presence of benzylserine (I) and the current in the absence of benzylserine (I0) as a function of the benzylserine concentration. The glutamate concentration was 10 μm.

Discussion

This article describes the characterization of two new inhibitors for the neutral amino acid transporter ASCT2. Benzylserine and benzylcysteine inhibit ASCT2 function based on a competitive mechanism, indicating that they bind to the substrate-binding site of ASCT2.The compounds inhibit the anion conductance that is associated with translocation of transported substrates. Therefore, our data suggest that translocation of transported substrates is competitively blocked by these compounds. Benzylserine application to ASCT2 in the absence of permeant anions does not result in detectable currents, suggesting that the compound is either not transported or that transport is electroneutral. Since substrate transport by ASCT2 is based on an electroneutral exchange mechanism, this absence of benzylserine-induced transport current does not allow us to rule out the possibility that the compound is a transported substrate of ASCT2. In glutamate transporters transported substrates generate anion current (Wadiche et al. 1995), whereas non-transportable blockers inhibit the tonic leak anion conductance (Wadiche & Kavanaugh, 1998). The latter behaviour is found for the two ASCT2 inhibitors characterized here which suggests that they are non-transportable blockers of ASCT2-mediated transport. However, a direct test of whether benzylserine is transported would involve the determination of the radiotracer flux of the radiolabelled inhibitor.

The selection of the two new inhibitors was based on their structural analogy with known blockers of the excitatory amino acid transporter family that compete with the transported substrate for binding to the substrate-binding site. They contain the bulky aromatic benzylether moiety that was found to be an important structural element in EAAT blockers (Shimamoto et al. 1998; Koch et al. 1999; Campiani et al. 2001), but they lack the β-carboxy group. It was previously reported that one of the differences between the substrate binding sites of EAATs and ASCTs is based on the presence of the basic amino acid side chain of arginine in EAATs that associates with the γ-carboxy function of glutamate (Bendahan et al. 2000). In ASCTs, this basic amino acid residue is replaced with an amino acid with a neutral side chain (cysteine in ASCT2). In fact, mutation of R447 in EAAC1 to cysteine converts EAAC1 to a transporter for neutral amino acids (Bendahan et al. 2000). Consistent with this interpretation, we found that the EAAT blocker TBOA does not bind to ASCT2. TBOA has two carboxy groups and one amino group, and is, like glutamate, negatively charged at physiological pH. This negative charge, most likely, prevents TBOA from binding to ASCTs. By simply removing the β-carboxy group, and therefore the negative charge from TBOA, the compound is converted into the ASCT2 inhibitor benzylserine. This finding has two implications: (1) TBOA interacts with EAATs such that its β-carboxy group associates with the conserved arginine residue in the substrate-binding site, and (2) the substrate binding sites of ASCTs and EAATs appear to be relatively conserved because the acceptor for the bulky, hydrophobic benzyl moiety of glutamate transporter blockers is also present in ASCT2. Interestingly, phenylalanine, although having a bulky aromatic side chain, is not an inhibitor of ASCT2, indicating that the length of the linker between the α-carbon and the aromatic ring is critical for the inhibitory effect of the compound. For glutamate transporters, inhibitors are known that do not have a bulky aromatic side chain, such as THA (threo-hydroxyaspartic acid). THA is not a blocker of EAATs, but a transportable inhibitor that is, however, transported at a low rate compared to glutamate (Arriza et al. 1994). Interestingly, the structural analogue of THA for ASCT2 that lacks the β-carboxy group is serine. Serine is a fully transported substrate of ASCT2, being transported at the same rate as alanine (Utsunomiya-Tate et al. 1996).

Application of the ASCT2 inhibitors to the transporter revealed the existence of a tonic leak conductance associated with ASCT2. This leak conductance is already present even in the total absence of transported substrates, such as alanine or glutamine. A similar leak conductance was previously described for the glutamate transporters of the EAAT family (Otis & Jahr, 1998; Watzke et al. 2001; Bergles et al. 2002). In analogy to EAATs the ASCT2 leak conductance is selective for anions and is inhibited by competitive inhibitors of transport. The ASCT2 leak conductance exhibits a high permeability for hydrophobic anions, displaying the permeability sequence P(SCN−) > P(NO3−) > P(Cl−). This permeability sequence is identical to that found for the substrate-induced ASCT2 anion conductance and very similar to that of the glutamate transporter anion conductance (Wadiche & Kavanaugh, 1998). Therefore, our results suggest that the permeation properties of the ASCT2 anion conductance are independent of whether it is activated in the absence or presence of transported substrate. This finding is in contrast to recent results obtained for the glutamate transporters EAAT2 and EAAT4 which showed that the permeability sequence for anions is different for the leak anion conductance and the glutamate-induced anion conductance (Melzer et al. 2003). This difference could be based on the different strategies employed in the study by Melzer et al. as compared to ours for investigating the anion conductance. Here, we observed only the part of the anion current that is blockable by competitive inhibitors. Therefore, the total ASCT2 leak anion current may be underestimated since it may contain a current that is not blockable by the inhibitors. In the study by Melzer et al. (2003) anion current was induced by application of SCN− to EAAT-expressing cells, possibly leading to an overestimation of anion current due to non-specific background anion currents. Like in most of the EAATs, ASCT2 currents carried by Cl− are insignificant due to the low Cl− permeability of the anion conductance, even at physiological transmembrane potentials. This result suggests that the Cl− leak conductance of ASCT2 is unlikely to play a significant physiological role.

Leak anion currents of ASCT2 blocked by benzylserine and benzylcysteine are about 24% of the maximum anion currents induced by saturating concentrations of alanine and glutamine (Fig. 1), suggesting that either the relative population of the leak anion conducting state (‘channel open probability’) is lower than that of the substrate-bound anion conducting state, or that the unitary anion currents catalysed by the two states are different from each other. Thus it appears that either the ‘open probability’ or the single transporter conductance is modulated along the transport pathway, in agreement with recent suggestions regarding the anion conductance of EAATs (Melzer et al. 2003; Ryan et al. 2004). The maximum amplitude of the anion current at saturating inhibitor concentrations was independent of the inhibitor used. This amplitude was also similar for the TBOA-inhibited leak current in EAAC1 (I(TBOA)/I(Glu) = 0.22, Fig. 1C. This inhibitor-independent nature of the leak conductance indicates that it is an intrinsic feature of the transport protein and that this feature is highly conserved within the ASC and glutamate transporter family. This interpretation is in line with a recent study on EAAT1 (Ryan et al. 2004) in which the authors propose that the EAAT1 anion conductance is an intrinsic feature of the transport protein and that transmembrane helix 2 (TM2) forms part of the permeation pathway for anions. Consistently, the amino acid sequence of TM2, and specifically the important residues S103 and D112 (EAAT1 numbering; Ryan et al. 2004), are conserved between the EAAT and ASCT families.

Activation of the ASCT2 leak anion conductance requires the presence of extracellular Na+. This finding is consistent with results obtained for the leak conductance of glutamate transporters which also depends on the external Na+ concentration (Watzke et al. 2001). In the case of EAAC1, the Km for the Na+ ion which activates the leak anion conductance is about 80 mm (Watzke et al. 2001). Interestingly, Na+ interacts with ASCT2 much more strongly, showing a Km of 0.3 mm. This is consistent with previous observations of glutamine uptake by ASCT2 which saturates with a Km for Na+ of 2 mm (Bröer et al. 2000). The results suggest that the Na+ binding sites of EAATs and ASCTs may have very different properties. On the other hand, leak anion current carried by ASCT2 is not affected by changes in the intracellular cation composition. Replacing intracellular Na+ by NMG+ or K+, ions that are not thought to be transported by ASCT2, has no effect on the magnitude of the leak anion current. This result suggests that the resting transporter resides in a state that is (1) anion permeable and (2) ready to accept extracellular substrate or inhibitor, independent of the nature of the intracellular cation. Therefore, we propose that exposure of the transporter binding sites to the extracellular side of the membrane is strongly favoured in the absence of transported substrate. It can be speculated that the reason for this behaviour is the exceptionally high affinity of the transporter for extracellular sodium, which tends to pull the transporter into the state that has extracellular Na+ bound and which is available for binding the neutral amino acid substrate. Surprisingly, activation of the substrate-dependent anion conductance by Na+ occurs with a much higher Km(19 mm). Two interpretations are possible to explain this result: (1) two binding sites for Na+ exist on ASCT2, one with high affinity which has to be occupied for activation of the leak conductance, and one with low affinity which has to be occupied for the activation of the alanine-dependent conductance; and (2) the binding of benzylserine to ASCT2 creates a high affinity Na+ binding site, whereas the binding of alanine creates a low-affinity Na+ binding site (see also the model shown in Fig. 9). Since there is no other evidence for the involvement of more than one Na+ ion in ASCT-catalysed amino acid transport, we prefer the second possibility, which is consistent with previous proposals (Bröer et al. 2000).

Figure 9. Two possible mechanisms of ASCT2 inhibition by benzylserine.

Figure 9

Open in a new tab

T is the ASC transporter, S the transported substrate, and I the non-transported inhibitor. The bars designate the anion-conducting state.

The diagrams in Fig. 9 show two possible mechanisms for the ASCT2 leak anion conductance that are compatible with our results. In the first mechanism (model 1, Bröer et al. 2000) either the inhibitor (I) or the amino acid substrate (S) can bind to the empty transporter (T) in the absence of Na+. The sodium ion then binds to the amino acid-loaded transporter to form a complex that is anion conducting, as indicated by the bar. This complex can also translocate the substrate across the membrane. In contrast, the binding of Na+ to the inhibitor-bound form of ASCT2 leads to a complex which is not anion conducting and which cannot undergo translocation. This model is also compatible with previous reports stating that the rate of Na+ exchange is about five times higher than that of amino acid exchange and that this ratio of exchange rates is variable and depends on the transported substrate (Koser & Christensen, 1971; Bröer et al. 2000). Such a variable stoichiometry can only be explained if Na+ unloading on both sides of the membrane is preferred over amino acid unloading, meaning that Na+ can dissociate from ASCT2 in the presence of bound amino acid, but amino acid cannot dissociate from ASCT2 in the presence of bound Na+. This mechanism could also explain why the affinity for Na+ is different depending on whether transported substrate or non-transported inhibitor is bound to ASCT2. However, model 1 cannot explain previous results showing that some amino acids, such as proline, exhibit a higher exchange rate than Na+ (Koser & Christensen, 1971). Therefore, we show here a second possible mechanism (model 2), which is analogous to that of the glutamate transporter leak anion conductance (Watzke et al. 2001). In model 2, Na+ binds to the empty transporter to form a Na+–transporter complex that is permeant to anions. Competitive inhibitors bind to this Na+–transporter complex and convert it to a non-conducting form, whereas the binding of substrate converts it to an anion-conducting form. The inhibitor-bound transporter is locked in this non-conducting form and cannot undergo any other state transitions, such as substrate translocation. According to this mechanism, ASCT2 is a Na+-gated anion channel which is analogous to EAATs (Wadiche et al. 1995; Watzke et al. 2001). Based on our data, we cannot differentiate between models 1 and 2. Further experiments will be necessary to determine which mechanism is the correct one.

The inhibitor with the highest affinity tested here, benzylcysteine, binds to ASCT2 with an apparent Ki of 780 μm. This affinity is relatively low, meaning that the compound would have to be used at high concentrations in possible physiological studies. Using high inhibitor concentrations may result in unspecific interactions with other proteins. Therefore, benzylserine and benzylcysteine are intended to serve as a proof of principle that it is possible to develop inhibitors for neutral amino acid transporters. We suggest that the two compounds should be used as lead structures for the development of future inhibitors that bind to ASCT2 with higher affinities. One possibility for improving the affinity of the compounds would be to add a hydrophilic OH or amide group to the β-carbon of the serine or cysteine residue. It will be also important to test the inhibitors for their activity towards other neutral amino acid transporters, such as system A and system N. For example, glutamine shuttling between cellular compartments of the brain is thought to be mediated by a variety of neutral amino acid transporters, including ASCT2 (Bode, 2001; Deitmer et al. 2003). Having selective inhibitors for neutral amino acid transporters available would facilitate the determination of which individual amino acid transport systems contribute to total glutamine uptake and release in brain tissue preparations or in vivo.

Acknowledgments

C.G. is grateful to the Max-Planck-Institute for Biophysics in Frankfurt, Germany, and E. Bamberg for support during initial experiments of this work. We thank S. Bröer for providing the ASCT2 cDNA and H.-G. Breitinger for critical reading of the manuscript.

References

  1. Arriza JL, Fairman WA, Wadiche JI, Murdoch GH, Kavanaugh MP, Amara SG. Functional comparisons of three glutamate transporter subtypes cloned from human motor cortex. J Neurosci. 1994;14:5559–5569. doi: 10.1523/JNEUROSCI.14-09-05559.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Arriza J, Kavanaugh M, Fairman W, Wu Y, Murdoch G, North R, et al. Cloning and expression of a human neutral amino acid transporter with structural similarity to the glutamate transporter gene family. J Biol Chem. 1993;268:15329–15332. [PubMed] [Google Scholar]
  3. Bass R, Hedegaard HB, Dillehay L, Moffett J, Englesberg E. The A, ASC, and L systems for the transport of amino acids in Chinese hamster ovary cells (CHO-K1) J Biol Chem. 1981;256:10259–10266. [PubMed] [Google Scholar]
  4. Bendahan A, Armon A, Madani N, Kavanaugh MP, Kanner BI. Arginine 447 plays a pivotal role in substrate interactions in a neuronal glutamate transporter. J Biol Chem. 2000;275:37436–37442. doi: 10.1074/jbc.M006536200. [DOI] [PubMed] [Google Scholar]
  5. Bergles DE, Tzingounis AV, Jahr CE. Comparison of coupled and uncoupled currents during glutamate uptake by GLT-1 transporters. J Neurosci. 2002;22:10153–10162. doi: 10.1523/JNEUROSCI.22-23-10153.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bode BP. Recent molecular advances in mammalian glutamine transport. J Nutr. 2001;131:2475S–2485S. doi: 10.1093/jn/131.9.2475S. [DOI] [PubMed] [Google Scholar]
  7. Bröer A, Albers A, Setiawan I, Edwards RH, Chaudhry FA, Lang F, et al. Regulation of the glutamine transporter SN1 by extracellular pH and intracellular sodium ions. J Physiol. 2002;539:3–14. doi: 10.1113/jphysiol.2001.013303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bröer A, Brookes N, Ganapathy V, Dimmer KS, Wagner CA, Lang F, et al. The astroglial ASCT2 amino acid transporter as a mediator of glutamine efflux. J Neurochem. 1999;73:2184–2194. [PubMed] [Google Scholar]
  9. Broer A, Wagner C, Lang F, Broer S. Neutral amino acid transporter ASCT2 displays substrate-induced Na+ exchange and a substrate-gated anion conductance. Biochem J. 2000;346:705–710. [PMC free article] [PubMed] [Google Scholar]
  10. Campiani G, De Angelis M, Armaroli S, Fattorusso C, Catalanotti B, Ramunno A, et al. A rational approach to the design of selective substrates and potent non-transportable inhibitors of excitatory amino acid transporter EAAC1 (EAAT3). New glutamate and aspartate analogues as potential neuroprotective agents. J Med Chem. 2001;44:2507–2510. doi: 10.1021/jm015509z. [DOI] [PubMed] [Google Scholar]
  11. Chen C, Okayama H. High-efficiency transformation of mammalian cells by plasmid DNA. Mol Cell Biol. 1987;7:2745–2752. doi: 10.1128/mcb.7.8.2745. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Christensen HN, Albritton LM, Kakuda DK, Macleod CL. Gene-product designations for amino acid transporters. J Exp Biol. 1994;196:51–57. doi: 10.1242/jeb.196.1.51. [DOI] [PubMed] [Google Scholar]
  13. Deitmer JW, Broer A, Broer S. Glutamine efflux from astrocytes is mediated by multiple pathways. J Neurochem. 2003;87:127–135. doi: 10.1046/j.1471-4159.2003.01981.x. [DOI] [PubMed] [Google Scholar]
  14. Franchi-Gazzola R, Gazzola GC, Dall'asta V, Guidotti GG. The transport of alanine, serine, and cysteine in cultured human fibroblasts. J Biol Chem. 1982;257:9582–9587. [PubMed] [Google Scholar]
  15. Grewer C, Watzke N, Wiessner M, Rauen T. Glutamate translocation of the neuronal glutamate transporter EAAC1 occurs within milliseconds. Proc Natl Acad Sci USA. 2000;97:9706–9711. doi: 10.1073/pnas.160170397. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hamill OP, Marty A, Neher E, Sakmann B, Sigworth FJ. Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers Arch. 1981;391:85–100. doi: 10.1007/BF00656997. [DOI] [PubMed] [Google Scholar]
  17. Hille B. Ion Channels of Excitable Membranes. Sunderland MA USA: Sinauer Associates, Inc.; 2001. [Google Scholar]
  18. Kilberg MS, Stevens BR, Novak DA. Recent advances in mammalian amino acid transport. Annu Rev Nutr. 1993;13:137–165. doi: 10.1146/annurev.nu.13.070193.001033. [DOI] [PubMed] [Google Scholar]
  19. Kimmich GA, Randles J, Wilson J. Na+-coupled alanine transport in LLC-PK1 cells. Am J Physiol. 1994;267:C1119–C1129. doi: 10.1152/ajpcell.1994.267.4.C1119. [DOI] [PubMed] [Google Scholar]
  20. Koch HP, Kavanaugh MP, Esslinger CS, Zerangue N, Humphrey JM, Amara SG, et al. Differentiation of substrate and nonsubstrate inhibitors of the high-affinity, sodium-dependent glutamate transporters. Mol Pharmacol. 1999;56:1095–1104. doi: 10.1124/mol.56.6.1095. [DOI] [PubMed] [Google Scholar]
  21. Koser BH, Christensen HN. Effect of substrate structure on coupling ratio for Na+-dependent transport of amino acids. Biochim Biophys Acta. 1971;241:9–19. doi: 10.1016/0005-2736(71)90298-7. [DOI] [PubMed] [Google Scholar]
  22. Matthews JC, Aslanian AM, McDonald KK, Yang W, Malandro MS, Novak DA, et al. An expression system for mammalian amino acid transporters using a stably maintained episomal vector. Anal Biochem. 1997;254:208–214. doi: 10.1006/abio.1997.2432. [DOI] [PubMed] [Google Scholar]
  23. Melzer N, Biela A, Fahlke C. Glutamate modifies ion conduction and voltage-dependent gating of excitatory amino acid transporter-associated anion channels. J Biol Chem. 2003;278:50112–50119. doi: 10.1074/jbc.M307990200. [DOI] [PubMed] [Google Scholar]
  24. Otis TS, Jahr CE. Anion currents and predicted glutamate flux through a neuronal glutamate transporter. J Neurosci. 1998;18:7099–7110. doi: 10.1523/JNEUROSCI.18-18-07099.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Rauen T, Rothstein JD, Waessle H. Differential expression of three glutamate transporter subtypes in the rat retina. Cell Tiss Res. 1996;286:325–336. doi: 10.1007/s004410050702. [DOI] [PubMed] [Google Scholar]
  26. Ryan RM, Mitrovic AD, Vandenberg RJ. The chloride permeation pathway of a glutamate transporter and its proximity to the glutamate translocation pathway. J Biol Chem. 2004 doi: 10.1074/jbc.M304433200. 10.1074/jbc.M304433200. [DOI] [PubMed] [Google Scholar]
  27. Shafqat S, Tamarappoo BK, Kilberg MS, Puranam RS, McNamara JO, Guadano-Ferraz A, et al. Cloning and expression of a novel Na+-dependent neutral amino acid transporter structurally related to mammalian Na+/glutamate cotransporters. J Biol Chem. 1993;268:15351–15355. [PubMed] [Google Scholar]
  28. Shimamoto K, Lebrun B, Yasuda-Kamatani Y, Sakaitani M, Shigeri Y, Yumoto N, et al. dl-threo-beta-Benzyloxyaspartate, a potent blocker of excitatory amino acid transporters. Mol Pharmacol. 1998;53:195–201. doi: 10.1124/mol.53.2.195. [DOI] [PubMed] [Google Scholar]
  29. Utsunomiya-Tate N, Endou H, Kanai Y. Cloning and functional characterization of a system ASC-like Na+-dependent neutral amino acid transporter. J Biol Chem. 1996;271:14883–14890. doi: 10.1074/jbc.271.25.14883. [DOI] [PubMed] [Google Scholar]
  30. Wadiche JI, Amara SG, Kavanaugh MP. Ion fluxes associated with excitatory amino acid transport. Neuron. 1995;15:721–728. doi: 10.1016/0896-6273(95)90159-0. [DOI] [PubMed] [Google Scholar]
  31. Wadiche JI, Kavanaugh MP. Macroscopic and microscopic properties of a cloned glutamate transporter/chloride channel. J Neurosci. 1998;18:7650–7661. doi: 10.1523/JNEUROSCI.18-19-07650.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Watzke N, Bamberg E, Grewer C. Early intermediates in the transport cycle of the neuronal excitatory amino acid carrier EAAC1. J General Physiol. 2001;117:547–562. doi: 10.1085/jgp.117.6.547. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Watzke N, Grewer C. The anion conductance of the glutamate transporter EAAC1 depends on the direction of glutamate transport. FEBS Lett. 2001;503:121–125. doi: 10.1016/s0014-5793(01)02715-6. [DOI] [PubMed] [Google Scholar]
  34. Zerangue N, Kavanaugh MP. ASCT-1 is a neutral amino acid exchanger with chloride channel activity. J Biol Chem. 1996;271:27991–27994. doi: 10.1074/jbc.271.45.27991. [DOI] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES