|Home | About | Journals | Submit | Contact Us | Français|
Glutamate transporters in the brain remove the neurotransmitter from the synapse by cotransport with three sodium ions into the surrounding cells. Recent structural work on an archaeal homolog suggests that, during substrate translocation, the transport domain, including the peripheral transmembrane helix 3 (TM3), moves relative to the trimerization domain in an elevator-like process. Moreover, two TM3 residues have been proposed to form part of a transient Na3′ site, and another, Tyr-124, appears close to both Na3′ and Na1. To obtain independent evidence for the role of TM3 in glutamate transport, each of its 31 amino acid residues from the glial GLT-1 transporter was individually mutated to cysteine. Except for six mutants, substantial transport activity was detected. Aqueous accessibility of the introduced cysteines was probed with membrane-permeant and membrane-impermeant sulfhydryl reagents under a variety of conditions. Transport of six single cysteine mutants, all located on the intracellular side of TM3, was affected by membrane-permeant sulfhydryl reagents. However, only at two positions could ligands modulate the reactivity. A120C reactivity was diminished under conditions expected to favor the outward-facing conformation of the transporter. Sulfhydryl modification of Y124C by 2-aminoethyl methanethiosulfonate, but not by N-ethylmaleimide, was fully protected in the presence of sodium. Our data are consistent with the idea that TM3 moves during transport. Moreover, computational modeling indicated that electrostatic repulsion between the positive charge introduced at position 124 and the sodium ions bound at Na3′ and Na1 underlies the protection by sodium.
In the brain, the signaling by glutamate is terminated by transporters that remove this excitatory neurotransmitter from the cleft into the cells surrounding the synapse. Glutamate transport is an electrogenic process (1–3), which consists of two half-cycles (4–6): (i) cotransport of the neurotransmitter with sodium and hydrogen ions (1, 7) and (ii) countertransport of potassium (1, 5) (see Fig. 1A). The stoichiometry is three sodium ions, one proton, and one potassium ion per transported glutamate molecule (8, 9).
Several crystal structures of a glutamate transporter homolog, GltPh, from the archaeon Pyrococcus horikoshii are now available (10–13). The structure reveals a trimer with a permeation pathway through each of the protomers, indicating that the protomer is the functional unit. This is also the case for the eukaryotic glutamate transporters (14–17). The protomer contains eight transmembrane domains and two oppositely oriented reentrant loops, one between domains 6 and 7 (HP1) and the other between domains 7 and 8 (HP2) (10). This unusual topology is in excellent agreement with that inferred from biochemical studies on the brain transporters (18–20). Moreover, many of the amino acid residues of the brain transporters that have been inferred to be important in the interaction with sodium (21, 22), potassium (4, 23), and glutamate (24, 25) are facing toward the binding pocket. Thus, the GltPh structures represent excellent models for the brain transporters.
Because of the limited resolution of the GltPh structure, Tl+ ions, which exhibit a robust anomalous scattering signal, have been used in an attempt to visualize the sodium sites in this homolog (11), which also uses three Na+ ions per transported substrate molecule (26). Two Tl+ sites were identified. However, in contrast to Na+, Tl+ could not support transport (11). Nevertheless, functional evidence supports the role of one of the Tl+ sites (Na1) as a sodium-binding site (27). Suggestions for additional sodium-binding sites have been searched using a combination of computational and functional studies (28–34). Based on computational studies, the Na3 site was proposed. The side chains of conserved threonine and asparagine residues, from transmembrane helix (TM)2 7 and TM8, respectively, as well as the carboxyl oxygens of the acidic amino acid substrate participate in this site (Fig. 1, B and C) (29). Recently, further experimental evidence for this site was reported (35). There is also experimental evidence for a Na3′ site (28, 30), which was predicted to be a transient site (29). The side chains of four conserved residues participate in this site, namely Asn-366 and Asp-368 from TM7 and Tyr-123 and Thr-127 from TM3, using the numbering of the glial transporter GLT-1 (Fig. 1, B and C).
Recent studies indicate that substrate translocation by GltPh, as well as by the neuronal transporter EAAC1, occurs by an “elevator-like” mechanism (12, 36), where the transport domain, which includes HP1, HP2, TM3, and TM6–8, moves relative to the fixed trimerization domain (37). Because TM3 is located at the periphery of the transport domain and suggested to participate in sodium binding, we set out to obtain independent evidence for the role of TM3 in glutamate transport. For this purpose, we introduced cysteine residues at each of the 31 TM3 positions of a cysteine-less version of GLT-1. Aqueous accessibility of the introduced cysteines was probed with membrane-permeant and membrane-impermeant sulfhydryl reagents under conditions favoring either inward- or outward-facing conformations. Two conformationally sensitive positions were observed: 120 and 124. The reactivity of A120C was diminished under conditions favoring the outward-facing conformation of the transporter. On the other hand, modification by 2-aminoethyl methanethiosulfonate of Y124C, which is located close to the Na1 and Na3′ sites, was fully protected in the presence of sodium.
Mutations were made by site-directed mutagenesis of cysteine-less WT GLT-1 (18) in the vector pBluescript SK− (Stratagene) using uracil-containing single-stranded DNA as described previously (38, 39). Briefly, the GLT-1-containing plasmid was used to transform Escherichia coli CJ236 (dut−,ung−). Uracil-containing single-stranded DNA was isolated from one of the transformants upon growth in uridine-containing medium, according to the standard Stratagene protocol, using R408 helper phage. This yields the sense strand, and consequently, mutagenic primers were designed to be antisense. The resulting mutant DNA plasmid was then sequenced in its entirety to verify the mutation and to ensure that no other mutations were introduced.
HeLa cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 200 units/ml penicillin, 200 μg/ml streptomycin, and 2 mm glutamine. Infection with recombinant vaccinia/T7 virus vTF7-3 (40), subsequent transfection with plasmid DNA, and d-[3H]aspartate transport were performed as described previously (41). In all of the experiments described, the expression vector used was pBluescript SK−, and all transport data were corrected for those obtained with cells transfected with the vector alone. In the case of the wild type, transport values were typically 25-fold over this background, and even for the low-activity Y124C mutant, the signal-to-noise ratio was at least 4-fold. Kinetic constants were determined using the standard transport solution (200 μl/well) containing d-aspartate concentrations ranging from 0 to 20 μm and supplemented with 0.8 μm d-[3H]aspartate (10 Ci/mmol) per well. Transport was done for 3 min for the determination of initial rates. Km and Vmax were calculated from linear regression using the Lineweaver-Burk plot.
Prior to transport assay, cells adhering to 24-well plates were washed once with 1 ml of transport medium containing 150 mm choline chloride instead of NaCl. Each well was then incubated at room temperature with 200 μl of preincubation medium (the different compositions are indicated in the figure legends) with the indicated concentrations of sulfhydryl reagents, (2-trimethylammonium)ethyl methanethiosulfonate (MTSET; Anatrace), N-ethylmaleimide (NEM; Sigma-Aldrich), or 2-aminoethyl methanethiosulfonate (MTSEA; Anatrace). After 5 min, the medium was aspirated, and the cells were washed twice with 1 ml of transport medium, followed by d-[3H]aspartate transport. For each mutant, the concentration of the different sulfhydryl reagents (see figure legends) chosen was determined after preliminary titration experiments in choline-containing medium for each mutant separately. An example of such a titration is shown in Fig. 5B. The aim of these titrations was to find a concentration at which the reagent caused a partial inhibition so that the effects of the different preincubation conditions could be seen most clearly. Results were plotted using normalized data for each mutant, where the untreated activity levels are defined as 100%.
Labeling of wild-type and mutant transporters at the cell surface using Sulfo-NHS-SS-Biotin (Pierce), quenching the reaction, cell lysis, and isolation of the biotinylated proteins by streptavidin-agarose beads (Pierce) were done as described previously (42). After SDS-PAGE (10% gel) and transfer to nitrocellulose, the GLT-1 protein was detected with an affinity-purified antibody directed against a polyclonal antibody raised against GLT-1 purified from rat brain (43) at a 1:5000 dilution, with horseradish peroxidase-conjugated secondary antibody at a 1:40,000 dilution, and with ECL. 1% goat serum was present in all antibody, blocking, and washing solutions to minimize the appearance of nonspecific bands.
Models of rat trimeric GLT-1 were constructed using a 2.96 Å resolution structure of GltPh (Protein Data Bank code 2NWL) as a template with MODELLER 9v6 (44). To create the alignment, we first identified 250 sequence homologs each of rat GLT-1, GltPh, and EAAC1 with PSI-BLAST using five iterations with an E-value cutoff of 0.005 for inclusion in the next round. The two sets of 250 sequences were then clustered to remove redundant sequence homologs with >70% identical residues using CD-HIT (45) and aligned using T-Coffee version 5.31 (46). After extracting the sequences of rat GLT-1 and GltPh and removing 52 residues between helices 4b and 4c, as well as 44 N-terminal and 73 C-terminal residues, the final alignment (supplemental Fig. 1) contained 34.6% identical residues. Guided by this alignment, at least 300 models were built; one with a reasonable molpdf score and optimal Ramachandran plot distributions was selected for further modeling. The resulting model of the outward-open conformation of rat GLT-1 has 13 and 5 residues in the generously allowed and disallowed regions of the Ramachandran plot, respectively, similar to the template structure (i.e. 18 and 5 residues, respectively, for Protein Data Bank code 2NWW). For the outward-occluded conformation of rat GLT-1, there are 16 generously allowed and 5 disallowed positions, compared with 10 and 7, respectively, for the template structure (Protein Data Bank code 2NWL). None of these residues interact directly with the ions, substrates, or cysteine-modified side chains. In each of the protomers, an aspartate was modeled in the substrate-binding site by analogy with the substrate-bound structure (code 2NWL), and sodium ions were added at the positions of all four proposed binding sites because it was not yet clear which would be bound simultaneously. Hydrogen atoms were added according to Ref. 47, and the entire model was then energy-minimized with CHARMM v34a2 (48) using 500 steps of steepest descents and 2000 steps of Powell optimization in each of two stages, first allowing side chains only and then the entire protein to be optimized. The protein was represented using the CHARMM27 force field, from which parameters were adapted, with the aid of PRODRG2 (49), ParaChem, and Gaussian 09 (50), for the cysteine-modified NEM (1-ethyl-1H-pyrrole-2,5-dione) or MTSEA (1-thialysine) residues (parameters are provided in the supplemental data). In one protomer of each model, Tyr-124 was converted to either MTSEA- or NEM-modified cysteine. Side chain rotamers of both MTSEA- and NEM-modified Cys-124 were sampled systematically in steps of 30° or 20°, respectively, for all five side chain χ angles. Those rotamers without clashes (i.e. whose side chain atoms are >1.4 Å from any protein atom) were included in the calculations of the electrostatic contribution to the binding free energy.
The electrostatic contribution to the binding free energy of the four putative sodium ions in the presence of each modifier was calculated under the Poisson-Boltzmann continuum electrostatics framework as implemented in CHARMM. Models were first aligned to the xy plane by fitting to the coordinates of GltPh (Protein Data Bank code 2NWL) obtained by OPM (51). The surface of the protein, ligand, and ions was determined using a 1.4-Å radius spherical probe with atomic radii (52). The region within this surface was assigned a dielectric constant of 2, as was a 25-Å-thick slab centered at z = −0.9 Å that represents the hydrophobic core of the membrane. To account for the large aqueous cavity in the center of the trimer, however, a 25-Å radius spherical region centered at (0, 0, 15) Å was assigned a dielectric constant of 80, as was the remainder of the 109 × 108 × 89 Å system. Electrostatic energies of the protein bound to the ion (Epi), the ion alone (Ei), or the protein alone (Ep) were calculated using a cubic grid whose cell dimension was 1 Å and then repeated using a finer (0.5 Å) grid whose boundary potentials were derived from the former calculation. The electrostatic contribution to the binding free energy for each ion (i) to the protein (p), containing a cysteine modification in a given rotamer, was then calculated as Ebind = Epi − Ep − Ei.
Each of the amino acid residues in TM3 of a cysteine-less version of the glial GLT-1 transporter was mutated to cysteine, one at a time. After expression of the mutants in HeLa cells, their ability to transport d-[3H]aspartate was determined and compared with that of the cysteine-less WT transporter (Fig. 2). Except for six mutants, substantial transport activity was detected. The inactive single cysteine mutants included replacements of Tyr-123 and Thr-127 (Fig. 2), both of which are proposed to participate in the Na3′ site (28). The other four inactive mutants were T128C, A131C, G135C, and I142C (Fig. 2). To determine whether these inactive mutant transporters reached the membrane at all, surface biotinylation was performed. Only T128C was biotinylated to levels comparable to the cysteine-less parent (Fig. 3A), indicating that this mutant has an intrinsic transport defect. The loss of function of the other five inactive mutants can be explained by the inability of the transporters to reach the plasma membrane, as evidenced by their almost complete absence from the biotinylated fraction (Fig. 3A). However, the biosynthesis of these five mutants was not affected, as judged from the “total” samples containing also transporters present in intracellular membranes (Fig. 3B). Moreover, in these five mutants, we observed only one band, reflecting immature transporters residing in internal membranes (Fig. 3B). On the other hand, an additional slower moving band was observed in the total samples of WT and T128C (Fig. 3B). This band probably represents the fully processed N-glycosylated transporters, which reside in the plasma membrane. We also analyzed the surface expression of five other mutants, where the reduction of the activity was less severe. For G114C, L134C, and L138C, the fully processed transporter band was the predominant one (supplemental Fig. 2, A and B). Their intensities in the linear exposure range were 51.7 ± 0.3, 37.3 ± 4.8, and 77.4 ± 1.2% of that of WT (n = 3), respectively (supplemental Fig. 2). For L134C, this is roughly similar to the percentage of their transport activity relative to that of WT (Fig. 2), indicating that the reduced transport activity of this mutant is predominantly due to impaired surface expression. Based on the same considerations, the reduced activity of G114C and L138C seems to be due to a combination of impaired membrane trafficking and intrinsic functionality. In the case of R115C, R119C, and Y124C, the intensity of the biotinylated bands was very low, especially in the case of the former two (supplemental Fig. 2). Because the transport activity of R115C and R119C is comparable with that of G114C and L134C, the low intensity of the bands seen with the former mutants is presumably due to proteolysis. The same is probably true for Y124C, which has an activity similar to that of L138C (Fig. 2). In support of this idea is the observation that the mobility of the bands for Y124C was variable from experiment to experiment. Two bands with increased mobility were seen in the experiment shown in supplemental Fig. 2, but in other experiments, the mobility of the band was similar to that of WT (data not shown). Consistent with this, the Vmax values of R115C, R119C, and Y124C were close to that of L138C (Table 1). The reduced Vmax of these mutants was in part offset by a decreased Km relative to that of WT (Table 1). In contrast to Y124C, the kinetic parameters of A120C, for which the accessibility of the introduced cysteine was also studied in detail (see below), were similar to those of WT (Table 1).
In sodium-containing media, the addition of non-transportable substrate analogs such as dl-threo-β-benzyloxyaspartate (TBOA) is expected to increase the proportion of outward-facing transporters (Fig. 1A). To determine sulfhydryl modification from the extracellular medium, HeLa cells expressing each of the 25 single cysteine mutants, which have measurable transport activity, were preincubated with 20 μm TBOA in the presence or absence of the membrane-impermeant sulfhydryl reagent MTSET at 1 mm. After washing away the reagents, transport was measured under standard conditions. None of the mutants was impacted by MTSET (Fig. 4A). This indicates either that the cysteine residues were not accessible to the extracellular side or that their modification did not impact the transport activity of the mutants.
To maximize the chances of sulfhydryl modification of the mutants from the cytoplasm, the cells were exposed to membrane-permeant NEM (1 mm) in the presence of potassium. It is expected that the proportion of inward-facing transporters will be increased in the presence of this cation (Fig. 1A). Under these conditions, a pronounced inhibition of the transport activity of A120C was observed, and a smaller inhibition was seen with S113C, L116C, and Y124C (Fig. 4B). Mutants R115C and G117C were also modified (p < 0.01), as evidenced by the fact that their activity was increased by the sulfhydryl reagent (Fig. 4B).
One implication of the proposed elevator-like movement of the transport domain is that the observed reactivity of the intracellular part of TM3 to the membrane-permeant NEM should be dependent on the conformation of the transporters. Therefore, thiol modification of six NEM-sensitive mutants was compared under conditions favoring outward- and inward-facing conformations. Because these six introduced cysteine residues are located on the intracellular side of TM3, they are most likely modified from the cytoplasmic side of the membrane. Only in the case of A120C was there a marked difference between the effects of NEM under the two conditions. Compared with the result in the presence of potassium, a marked protection was observed with sodium and TBOA in the preincubation medium (Fig. 5A). As described under “Experimental Procedures,” the NEM concentration used was based on preliminary titration experiments. Clearly, the protection observed in the presence of sodium and TBOA was also seen at other concentrations of NEM (Fig. 5B). This protection is consistent with the idea that the cysteine at position 120 becomes inaccessible to NEM when the transporters become outward-facing. There was also some protection by TBOA against sulfhydryl modification of Y124C, whereas there was little difference between the two conditions for the other mutants (Fig. 5A).
Due to the presence of high intracellular glutamate concentrations (~10 mm), a predominantly outward-facing conformation is expected when either sodium or choline is in the external medium (Fig. 1A). However, comparing the activities measured in the presence of sodium with or without TBOA, there is a very marked protection by the TBOA against inactivation of A120C by NEM. This suggests that, compared with WT, a larger proportion of the mutant transporters are inward-facing despite the presence of only sodium in the extracellular medium (Fig. 6). This is further supported by the fact that the addition of l-glutamate did not lead to any significant potentiation of the inhibition (Fig. 6). Such a potentiation was observed, however, in the presence of potassium (Fig. 6), perhaps due to an effect additional to that on the proportion of inward-facing transporters. The effect of TBOA was sodium-dependent; much less protection by TBOA was seen in the presence of choline (Fig. 6).
As mentioned above, there was only a small difference in the inhibition of Y124C by NEM in the presence of either sodium plus TBOA or potassium (Fig. 5A). However, a dramatic difference was seen under these two conditions when the positively charged membrane-permeant MTSEA (2.5 mm) was used rather than NEM (Fig. 7). In contrast, the other five NEM-sensitive mutants were hardly impacted by this reagent either in the presence of sodium plus TBOA or in the presence of potassium (Fig. 8). Moreover, none of the cysteine residues located on the extracellular side of position 124 were impacted by MTSEA, not even in the presence of choline (data not shown). The activity of Y124C was very sensitive to this reagent in the presence of choline, whereas a complete protection by sodium was observed regardless of the presence of l-glutamate or TBOA (Fig. 7). Significant protection was also seen in the presence of lithium, which was almost as efficient as sodium. In contrast, in the presence of potassium, which is thought to bind at a different location and which is less similar to sodium than lithium, a lower degree of protection was observed (Fig. 7).
The protection of Y124C against inactivation by MTSEA was observed even at very low concentrations of sodium, with a half-maximal effect seen at ~6 mm (Fig. 9A). Therefore, it is likely that at 150 mm sodium no effect of l-glutamate or TBOA was observed (Fig. 7) because full protection was already observed in the presence of sodium alone. To increase the possibility of detecting modulatory effects by either the substrate or the blocker, the experiments were repeated at 3 mm sodium, a concentration at which the protection by sodium was only partial (Fig. 9A). Under these conditions, both l-glutamate and TBOA afforded marked protection (Fig. 9B). This is in contrast to the different effects of the substrate and blocker on the reactivity of A120C to NEM (Fig. 6). The specificity of the protection of Y124C against MTSEA in the presence of 3 mm sodium is illustrated by the fact that GABA, which is not a substrate of the glutamate transporters, had basically no effect (Fig. 9B). The effect of l-glutamate and TBOA was much smaller in the absence of sodium than in its presence presumably because of the sodium dependence of their binding to the transporter (Fig. 9B). As shown side-by-side in Fig. 10, the dramatic effects of different cations on the accessibility of Y124C to MTSEA were not seen with NEM; this issue will be addressed under “Discussion.”
The comparison between outward- and inward-facing GltPh structures indicates that TM3, located at the periphery of the transport domain, is vertically translocated by ~12 Å (12). This α-helix is in contact with TM6–8 as well as with the membrane bilayer (Fig. 1B). During the translocation, the spatial interactions between TM3 and its neighboring TMs do not change, but part of its contact area with the bilayer is now predicted to become accessible to the internal aqueous space, assuming a minimal adaptation of the membrane. Indeed, reactivity to membrane-permeant NEM is observed only with positions located in the intracellular part of TM3 (Fig. 4B), suggesting that NEM is approaching those cysteine residues from the cytoplasm. For several of these introduced cysteines, this reactivity is not conformationally sensitive (Fig. 5A), suggesting that they are always accessible to the cytoplasm.
The reactivity to NEM of only one engineered cysteine (position 120) is markedly protected by the blocker TBOA. We cannot entirely exclude a steric effect of the blocker, but because it is highly likely that the modification by NEM is from the cytoplasmic side, this protection by TBOA appears to be due to its ability to increase the proportion of the outward-facing transporters (Fig. 1A). This protection is sodium-dependent (Fig. 6) apparently because binding of the analog to the transporter requires sodium. The substrate, expected to increase the proportion of inward-facing wild-type transporters (Fig. 1A), does not potentiate the inhibition of transport of A120C by NEM (Fig. 6). Therefore, it appears that, already in the presence of sodium alone, almost all of the A120C transporters are inward-facing. This could be due to a reduced apparent affinity for internal sodium and/or glutamate, which would limit the transition from the inward to the outward conformation.
Because the reactivity to NEM of only one cysteine (position 120) is conformationally sensitive, one might conclude that the movement of TM3 is much smaller than the predicted 12 Å. However, it should be kept in mind that the effects of the sulfhydryl reagents on transport activity are used as the criterion for accessibility. Thus, it is possible that modification of some of the introduced cysteines does not have any effect on transport. For instance, modification can take place when exposed to the cytoplasm, but the adduct could be accommodated by the bilayer during the “upward” movement of the transporter. In contrast to the expectation for an adduct with MTSET, which has a permanent positive charge, the adducts with the membrane-permeant reagents are much more hydrophobic, and accommodation by the bilayer is a real possibility. Thus, our data are nevertheless consistent with the proposed vertical movement of the transport domain during substrate translocation (12, 36) but do not rule out the possibility that the movement of TM3 is smaller than 12 Å. Another possibility that should be considered along this line is adaptation of the lipid bilayer; it is not yet clear whether or to what extent the membrane is able to adapt to the large conformational changes of the transport domain in glutamate transporters to maintain hydrophobic matching. However, if such a large adaptation were to occur, it could also result in smaller apparent changes in accessibility.
The characteristics of the inhibition of Y124C by sulfhydryl reagents were completely different from those of A120C. Even though the activity of Y124C was somewhat inhibited by NEM (Fig. 4B), this inhibition was hardly sensitive to the conformational state of the transporter (Fig. 5A). Interestingly, a potent inhibition of transport of this mutant by MTSEA in the absence of sodium was observed. Specifically, in the absence of sodium, Y124C was very sensitive to MTSEA, whereas physiological sodium concentrations fully protected against this reaction (Fig. 7). When the sodium concentration was reduced, the effects of l-glutamate and TBOA on the inhibition of Y124C by MTSEA could be monitored, showing that the substrate and blocker had the same protective effect (Fig. 9B). Thus, the reactivity of the cysteine inserted at position 124 is the same regardless of whether the transporters are outward- or inward-facing. Instead, the reactivity of this cysteine is dependent on whether sodium is or is not bound. Lithium is the closest sodium congener and is able to support transport by EAAC1 (22), and in the case of GLT-1, this cation appears to be able to replace sodium at some but not all sites (53). Consistent with these observations, lithium is quite effective in protecting Y124C against MTSEA (Fig. 7). On the other hand, the larger potassium is only slightly more effective than choline, which basically is an inert cation and is thought not to bind to the excitatory amino acid transporters.
The dramatic effects of cations on the accessibility of Y124C to MTSEA were not seen with NEM (Fig. 10). A possible explanation for this is that whereas NEM reacts from the intracellular side of the membrane, MTSEA approaches the cysteine at position 124 from the extracellular side. Consistent with this explanation, the MTSEA reactivity of the more internally located cysteines was much lower than that of the cysteine at position 124 (Fig. 8). We have tried to obtain further clues to the possible external accessibility of Y124C to MTSEA. However, for the single cysteines located extracellularly relative to position 124, MTSEA did not affect transport, even if choline was present (data not shown). The positions that are most likely impacted by MTSEA are those participating in the Na3′ site, namely Tyr-123 and Thr-127 (28, 30). However, the introduction of a cysteine residue at or near these positions rendered the transporter inactive (Fig. 2), and therefore, these mutants were not available to be assayed.
One possibility is that the occupation of the Na3′ site physically protects the access to the cysteine at position 124 (Fig. 1C). On the other hand, the Na1 site (11, 27) is also nearby, and thus, it is possible that the protection of Y124C is due to occupation of the Na1 and/or Na3′ site. Computational modeling of NEM- or MTSEA-modified GLT-1(Y124C) suggested that the sodium ions would not sterically compete directly with bound reagent, as all atoms of NEM or MTSEA are >3 Å from either ion (see “Experimental Procedures”). However, it is possible that the approach of and/or reaction with the positively charged MTSEA is electrostatically unfavorable in the presence of an ion at one or more of these sites. To test this hypothesis, we calculated the electrostatic contribution to the binding free energy of sodium ions to the NEM- or MTSEA-modified GLT-1(Y124C) model (Table 2). This is equivalent to the protection of reagent binding by bound ions. As shown in Table 2, the binding energies of ions at the Na1′ and Na3 sites were both significantly (12.6–13.3 kcal/mol) less favorable in the presence of MTSEA than in the presence of NEM. This result was independent of whether the system included three or four ions in total and independent of which combination of ions was considered. Thus, these calculations support the possibility that addition of a cationic moiety at position 124 destabilizes binding at Na1 and/or Na3′. Destabilization by MTSEA was also observed for binding to the more distal Na2 and Na3 sites, albeit to a lesser extent (~5 kcal/mol), suggesting that ion binding to those sites may also offer some protection against MTSEA modification of Y124C. We note that the electrostatic contribution to the binding energy of an ion to Na2 was unfavorable even in the presence of NEM, except in the case that no ion was bound at Na3, presumably reflecting the direct repulsion of the ions at these two positions.
The overall transport process requires three sodium ions, rather than the two predicted to be involved in the protection of Y124C. The apparent affinity for sodium as measured by the protection of Y124C (Fig. 9A) appears to be significantly higher than that for transport (22). Therefore, it is possible that the apparent affinity of the transporter for the “third” sodium is very low, and this would result in a lower apparent sodium affinity when transport is measured. The further protection by either substrate or analog suggests that their binding affects the liganding of the sodium ion(s) involved in the protection, consistent with earlier proposals (29, 54). The sodium concentration dependence of the protection (Fig. 9A) suggests the involvement of more than one site. Together, these observations suggest that the protection is due to binding of this cation to both Na1 and Na3′ sites. It appears that the combination of computational and functional studies used here could give further insights into the role of other structural elements of the transporters.
*This work was supported, in whole or in part, by National Institutes of Health Grant NS16708 from NINDS (to B. I. K). This work was also supported by United States-Israel Binational Science Foundation Grant 2007051 (to B. I. K.) and by German Research Council (DFG) Collaborative Research Group 807 “Transport and Communication across Membranes” (to L. R. F.).
This article contains supplemental data and Figs. 1 and 2.
2The abbreviations used are: