|Home | About | Journals | Submit | Contact Us | Français|
The Na+/dicarboxylate symporter (SdcS) from Staphylococcus aureus is a homolog of the mammalian Na+/dicarboxylate cotransporters (NaDC1) from solute carrier family 13 (SLC 13). The present study examined succinate transport by SdcS heterologously expressed in Escherichia coli, using right-side-out (RSO) and inside-out (ISO) membrane vesicles. The Km values for succinate in RSO and ISO vesicles were similar, about 30 μM. The single cysteine of SdcS was replaced to produce the cysteineless transporter, C457S, which demonstrated similar functional characteristics as the wild-type. Single cysteine mutants were made in SdcS-C457S at positions that are functionally important in the mammalian NaDC1. Mutant N108C of SdcS was sensitive to chemical labeling by MTSET ([2-(trimethylammonium)ethyl]-methanethiosulfonate) from both the cytoplasmic and extracellular side, depending on the conformational state of the transporter, suggesting that Asn-108 may be found in the translocation pore of the protein. Mutant D329C was sensitive to MTSET in the presence of Na+ but only from the extracellular side. Finally, mutant L436C was insensitive to MTSET although changes in its kinetic properties indicate that this residue may be important in substrate binding. In conclusion, this work identifies Asn-108 as a key residue in the translocation pathway of the protein, accessible in different states from both sides of the membrane. Functional characterization of SdcS should provide useful structural as well as functional details about mammalian transporters from the SLC 13 family.
The Na+/dicarboxylate symporter (SdcS) from Staphylococcus aureus is a member of the divalent anion sodium symporter (DASS) family that also includes the mammalian solute carrier 13 (SLC 13) family (1). SLC13 family members are plasma membrane transporters that use energy from the movement of Na+ down its electrochemical gradient to transport dicarboxylates or inorganic anions across the membrane (2). There are three different SLC13 transporters for citric acid cycle intermediates in humans: the low affinity Na+/dicarboxylate cotransporter 1 (NaDC1), the high affinity Na+/dicarboxylate cotransporter 3 (NaDC3) and the Na+/citrate cotransporter (NaCT). The transport properties of SdcS are very similar to those of the mammalian NaDC transporters. SdcS is a Na+-coupled transporter that carries four-carbon terminal dicarboxylates, including succinate, malate and fumarate, with Km values between 5 and 15 μM (3).
Although a considerable amount of structure-function information has been determined using the mammalian NaDC transporters (4–7), there are technical difficulties with mammalian expression systems that limit the studies to the extracellular surface of the membrane. In contrast, prokaryotic homologs of mammalian proteins can be expressed in large quantities in bacteria such as Escherichia coli, which allows detailed structural and functional characterization (8;9). Established methods for preparing E. coli membrane vesicles of known orientation permits the analysis of transporters from both the inside and outside of the cell (10–12). The sequence of SdcS is about 35–40% identical to that of the mammalian NaDC1 transporters, and several key residues are conserved or are similar. Therefore, studies of SdcS should provide valuable information that is applicable to the other members of the DASS family.
In the present study, we examined the functional properties of SdcS and mutants expressed in E. coli, using right-side-out (RSO) and inside-out (ISO) membrane vesicles to examine the transport reactions in both directions. SdcS exhibited similar succinate Km values in RSO compared with ISO vesicles. The cysteineless transporter C457S had similar kinetic properties as the wild type, although there were some differences in the effects of sodium on transport. We also found that amino acids previously found to be important in mammalian NaDC1 are also important in SdcS. The N108C mutant of SdcS was found to be accessible to membrane impermeant methanethiosulfonate reagents from both sides of the membrane, indicating that this residue may be located within the translocation pathway of the transporter, possibly close to the substrate binding site. The D329C mutant was accessible to MTSET in RSO vesicles and only in the conformation seen in the presence of Na+. Mutant L436C was not functionally affected by MTSET treatment but this mutant exhibited a decrease in substrate affinity in the forward direction. In conclusion, this study establishes SdcS as a model system for the eukaryotic NaDC1 transporters and identifies key residues that may be important during the transport cycle from both sides of the membrane.
Site-directed mutagenesis was performed using the QuikChange site-directed mutagenesis kit (Stratagene) according to the manufacturer’s instructions. Plasmid pQE-80L/SdcS encoding SdcS with a hexa-histidine tag at the N-terminal (3) was used as a template to construct the cysteineless mutant, C457S. The cysteine-substituted mutants were then made using the C457S mutant as the template. Mutants were verified by sequencing at the University of Texas Medical Branch (Galveston, TX) DNA Sequencing facility.
Expression of recombinant SdcS in E. coli BL21 [F− ompT hsdSB(rB− mB−) gal dcm] was as described (3). Overnight cultures of E. coli transformed with plasmids (SdcS and mutants C457S, D329C, L436C and N108C in vector pQE-80L or pQE-80L alone) were used to inoculate 500 ml of LB-Lennox broth (1:10 dilution) containing 50 μg/ml of carbenecillin. Cells were grown at 37°C to an optical density at 660 nm of 0.4 to 0.6. To induce protein expression, 150 μM isopropyl-β-D-galactopyranoside (IPTG) was added and cells were harvested 2 hours later by centrifugation. Cell pellets were washed and resuspended with 100 mM potassium phosphate buffer (KPi), pH 7.
Right-side-out vesicles (RSO) were prepared by a modification of Kaback’s method developed by Quick and Jung ((12) and personal communication, M. Quick). Briefly, washed centrifuged cell pellets were resuspended in 300 ml per liter of original culture of 30% sucrose, 30 mM Tris-HCl, pH 8. RNAse A (1 μg/ml) and K2-EDTA, pH 7 (10 mM) were added and the mixture was incubated at room temperature for 2–3 min. Lysozyme (6.4 mg to 300 ml resuspended pellets) was added and flasks were incubated with shaking 15 min at room temperature. The cell suspension was centrifuged for 30 min at 17,000 × g (rav) and pellets were resuspended in 7.5 ml of 30% sucrose, 20 mM MgSO4, 100 mM potassium phosphate buffer (KPi), pH 6.6. RNAse A and DNAse (1.7 mg each) were added and homogenates were poured into 1 L flasks containing 600 ml of 50 mM KPi, pH 6.6. Flasks were incubated 10 min at 30°C with slow shaking at 75 rpm. 1M K2-EDTA, pH 7 (10 mM final concentration) was added and the incubation continued for 15 min, followed by 1 M MgSO4 (15 mM final concentration) and an additional 15 min incubation. The cells were centrifuged at 27,500 × g for 70 min. The pellets were resuspended in a total volume of approximately 100 ml of 100 mM KPi, 10 mM EDTA, pH 6.6. Low speed centrifugation was performed at ~ 500 × g for 30 min to remove unbroken cells. The supernatants were then centrifuged at 43,400 × g for 30 min. RSO vesicles were resuspended by homogenization in 100 mM KPi, pH 7, separated into aliquots, quick-frozen in liquid N2 and stored at −80°C.
Inside-out membrane vesicles (ISO) were prepared according to Quick and Jung ((12;13) and personal communication, M. Quick), with an additional step to purify the membranes by sucrose density gradient centrifugation (14). The cells were washed and suspended in KPi buffer, pH 7.4 to approximately 50 ml per liter of culture. Phenylmethylsulfonyl fluoride (PMSF), 0.5 mM, was added to all steps to reduce proteolysis. The suspensions were incubated on a platform shaker at 4°C with 0.5 mg/ml each of DNAse and RNAse A. The cell suspensions were then passed through a French Pressure cell (Thermo Scientific) at 4000 psi. Unbroken cells and organelles were removed by low speed centrifugation for 15 min at 10,000 × g, 4°C. Vesicles were then pelleted from the supernatants by ultracentrifugation at ~184,000 × g, 45 min at 4°C. The pellets were resuspended by homogenization in 2 ml 100 mM KPi buffer, pH 7.4 and further purified by sucrose density gradient centrifugation. The sucrose density step gradients were prepared with 0.77 M, 1.44 M and 2.02 M sucrose in 10 mM HEPES, pH 7.4. Crude ISO vesicles were added to the top of the gradient and samples were centrifuged at 27,000 rpm (SW28 rotor, Beckman) for 15–18 hrs. at 4°C. The upper band of proteins at the interface between 0.77 M and 1.44 M sucrose was collected by gradient collector. Preliminary studies showed that this membrane fraction contains most of the Na+-dependent succinate transport activity in vesicles prepared from E. coli expressing SdcS. The lower band between 1.44 M and 2.02 M sucrose contains cell wall fragments and other cellular debris (14). The upper band from the step gradient containing ISO vesicles was diluted approximately 5-fold with 100 mM KPi, pH 7.4 and centrifuged at ~184,000 × g, 45 min at 4°C. ISO membrane vesicles were resuspended by homogenization in 100 mM KPi, pH 7.4. The vesicles were separated into aliquots, quick-frozen in liquid N2 and stored at −80°C. For each experiment, at least three independent right-side-out and inside-out membrane vesicle preparations were made.
Uptake of 14C-succinate by RSO and ISO membrane vesicles was measured using a rapid filtration assay (15). The sodium transport buffer contained (unless otherwise specified) 10 mM NaCl, 90 mM choline chloride, 50 mM MOPS, pH adjusted to 7 with 1M Tris. The assays were done at room temperature. For the assay, 40 μl of transport buffer containing ~30 μM 14C-succinate (44 mCi/mmol, PerkinElmer Life Sciences) was placed in the bottom of a 5 ml polystyrene Falcon tube. The reaction was initiated by the addition of 10 μl vesicles with vortexing. The reaction was terminated after the appropriate time, usually 10 sec, with 1 ml ice-cold choline buffer (100 mM choline chloride, 50 mM MOPS, pH 7), filtered immediately through a Millipore filter (0.45 μm pore size, type HAWP) with suction and washed with 4 ml cold choline buffer. The radioactivity retained by the filters was measured by liquid scintillation counting.
Sodium activation of succinate uptake was determined by replacing Na+ with choline up to a final concentration in the transport mixture of 80 mM. For experiments to determine kinetic constants, succinate concentrations between 1 μM to 200 μM (or 400 μM for RSO vesicles containing L436C mutant) were used. Kinetic constants (Km and Vmax) were determined by fitting initial transport rates (10 sec) to the Michaelis-Menten equation [v = (Vmax[S])/(Km + [S])] using nonlinear regression analysis (SigmaPlot 9, Systat Software Inc.).
The MTSET labeling reaction was done by combining 40 μl of the vesicles with 10 μl of 5 X concentrated solutions: 5 mM MTSET in either 25 mM Na+/75 mM choline buffer, 100 mM choline buffer or 25 mM Na+/75 mM choline buffer with 50 mM succinate for 10 min at room temperature. MTSET was weighed into tubes, kept dark and on ice, and the buffer added just before use. A final concentration of 1 mM MTSET was used for all the preincubations. Control groups of vesicles were preincubated with the same buffers but without MTSET. After the preincubations, the transport activity remaining in the vesicles was measured as described above.
RSO and ISO vesicles (25 μg protein) were diluted in sample buffer (50 mM Tris-HCl pH 7, 10% glycerol, 4% SDS, 2 % β-mercaptoethanol, 0.1 mg/ml Coomassie blue R-250) and heated in a boiling water bath for 2 min. Proteins were separated by Tricine SDS-PAGE with 10% (w/v) acrylamide and transferred to nitrocellulose membranes (16). Blots were incubated with 1:2000 dilutions of mouse monoclonal antibody reactive to the SdcS N-terminal RGS(H)4 epitope tag (RGS-His antibody, QIAGEN) followed by 1:5000 dilution of horseradish peroxidase-conjugated anti-mouse immunoglobulin G antibody (Jackson ImmunoResearch Laboratories, Inc.). The Supersignal West Pico chemiluminescent substrate kit (Pierce) was used to detect antibody binding. Images were captured with a Kodak Image Station 440CF and Image 1D analysis software (Eastman Kodak Co.) was used to quantitate the protein expression.
The protein content of membrane vesicles was measured using the Bio-Rad protein assay with γ–globulin as a standard.
Duplicate or triplicate measurements were made for each data point. The experiments were repeated with at least three different membrane preparations. Significant differences between groups were identified by Student’s t-test or ANOVA with P<0.05.
Figure 1 shows time courses of succinate transport by right-side-out (RSO) and inside-out (ISO) membrane vesicles of E. coli BL21 cells harboring pQE-80L vector only, pQE-80L/SdcS or pQE-80L/C457S, encoding the cysteineless mutant. Vesicles expressing SdcS and C457S showed rapid accumulation of succinate followed by a slow decline. The succinate uptake was linear until ~15 sec and therefore 10 sec was used as an initial time point in subsequent experiments. The peak of the overshoot was seen at ~1 min for both SdcS and C457S in RSO and ISO vesicles. The succinate content of ISO vesicles expressing SdcS and C457S reached equilibrium by 60 min, whereas the RSO vesicles still retained some of their succinate content by that time point, possibly due to differences in membrane permeability or vesicle volume. The succinate content of vesicles prepared from E. coli carrying the control plasmid, pQE-80L, showed a steady increase toward equilibrium with no overshoot in succinate concentration. The time course of succinate transport by SdcS in choline was similar to that of vesicles containing the control plasmid (not shown).
To verify the orientation of RSO and ISO membrane vesicles, succinate transport activity was measured after pretreatment with membrane-impermeant cysteine-specific methanethiosulfonate (MTS) reagents. RSO and ISO membrane vesicles were preincubated 10 min with 1 mM each [2-(trimethylammonium)ethyl]-methanethiosulfonate (MTSET) or (2-sulfonatoethyl) methanethiosulfonate (MTSES). There was no effect of the MTS reagents on succinate uptake activity in RSO vesicles expressing SdcS (Figure 2). However, succinate transport activity in ISO vesicles expressing SdcS was inhibited after pretreatment with the MTS reagents; preincubation with MTSET decreased the succinate uptake by ~ 85 % and preincubation with MTSES decreased the uptake by ~ 90 % (Figure 2). We also tested the membrane permeant reagent MTSEA, which inhibited ISO vesicles completely but showed variable effects in RSO vesicles (results not shown). In two experiments there was no effect of 1 mM MTSEA in RSO vesicles and in two experiments there was almost 100% inhibition. The membrane permeant reagent, N-ethyl maleimide, inhibited SdcS in RSO vesicles by 73% (results not shown). RSO and ISO vesicles expressing the cysteineless mutant, C457S, were insensitive to MTS reagents (results not shown) verifying that the single cysteine at position 457 in SdcS mediates the effects of these reagents. The results show that the single cysteine at position 457 in SdcS is accessible to membrane-impermeant reagents only from the inside of the cell but not from the outside. In a single dose-response experiment, the K0.5 for MTSET was 99 μM (result not shown).
It should be noted that membrane vesicle preparations often contain a mixture of orientations. For example, one study showed that RSO vesicle preparations from E. coli could contain between 5–25% ISO when counting numbers of vesicles (17). However, the ISO vesicles are much smaller in size and make up only 2–3% of the membrane surface area. Our time course experiment is consistent with this observation because the equilibrium vesicle volumes of ISO vesicles are lower than those of RSO (Fig. 1) The same study found that ISO membrane vesicles prepared using a French Pressure Cell were approximately 60–80% pure (17), although the sucrose density gradient step in our study might increase the purity of the ISO vesicles. One might expect to see some inhibition of “RSO membranes” by MTS reagents due to inhibition of contaminating ISO vesicles. It is possible that the smaller size and relatively lower transport activity of ISO vesicles could explain the lack of apparent inhibition.
SdcS is a sodium-coupled transporter. Therefore, the relationship between Na+ concentration and succinate transport activity was examined in membrane vesicles expressing SdcS and C457S (Figure 3). Our previous transport studies of SdcS showed strong cation-mediated inhibition of succinate uptake at higher Na+ concentrations in whole cell assays (3) but not in proteoliposomes (18). In preliminary studies with membrane vesicles, we found that succinate transport by SdcS expressed in RSO membrane vesicles was inhibited by sodium concentrations greater than 20 mM; therefore, the assay was done at lower sodium concentrations. The C457S mutant, in contrast, was not inhibited by sodium concentrations up to 80 mM (not shown). In the experiment shown in Figure 3A, activation of succinate transport by sodium was sigmoidal with Hill coefficients of 2.48 (SdcS) and 1.97 (C457S). The KNa values for the experiment shown in Figure 3A were 4.5 mM (SdcS) and 7.7 mM (C457S). The average KNa values in SdcS were 3.3 ± 0.6 mM (n=3 experiments) and in C476S, 8.2 ± 0.4 mM (n=5 experiments), significantly different from one another at p<0.05. In ISO vesicles, both SdcS and C457S were much more sensitive to inhibition by sodium, with decreased activity above 7 mM Na+ and almost complete inhibition of activity at 80 mM Na+ (Figure 3B).
Succinate kinetic analysis was performed for SdcS and C457S in RSO and ISO membrane vesicles. The Km for succinate in SdcS was similar in RSO and ISO vesicles, 26 and 36 μM, respectively (Table 1). This value was higher than the previous Km measurements of 7 μM in whole cells and 12 μM in proteoliposomes (3;18). The Vmax for succinate transport by SdcS in RSO vesicles was significantly higher than in ISO vesicles. The Km values for succinate in C457S in RSO and ISO vesicles, 42 and 51 μM, respectively, were not significantly different from the values measured for SdcS. Similar to SdcS, C457S had a significantly higher succinate Vmax in RSO compared with ISO vesicles. However, it should be noted that the volume differences between RSO and ISO vesicles could contribute to the apparent differences in Vmax values.
Previous studies with the rabbit NaDC1 have shown that cysteines substituted for residues Lys-84, Asp-373 and Met-493 are accessible to membrane-impermeant methanethiosulfonate (MTS) reagents (6;7;19). D373C is accessible to MTSET in both the presence and absence of Na+, and MTSET labeling is prevented by substrate. M493C and K84C require sodium for inhibition by MTS reagents, and they also show substrate protection. K84 is not sensitive to inhibition by MTSET, only by MTSES. These results show that the accessibility of these residues to the outside of the cell depends on the conformational state of the transporter. SdcS is approximately 40% identical in sequence to rbNaDC1, and Asn-108, Asp-329, and Leu-436 of SdcS correspond with Lys-84, Asp-373, and Met-493 of rbNaDC1 (Figure 4). Therefore, the single cysteine-substituted mutants N108C, D329C, and L436C were constructed in the cysteineless SdcS mutant, C457S. The accessibility of these substituted cysteines to MTS reagents from both sides of the membrane was then tested using RSO and ISO vesicles.
The protein abundance of SdcS and mutants in RSO and ISO membrane vesicles was determined by Western blotting (Figures 5 and and6).6). The summary of succinate transport activity and protein expression of single cysteine mutants is shown in Figure 6. In RSO vesicles there was similar protein expression for SdcS, C457S and the cysteine-substituted mutants. In SdcS, C457S and L436C, the transport activity was similar to the relative amount of protein expression. However, N108C and D329C in RSO vesicles had much lower relative transport activity than protein expression. In the ISO vesicles, the protein expression of SdcS was about half that of the C457S mutant whereas the transport activity of SdcS was higher than in C457S. The L436C mutant had lower activity compared with protein expression, whereas the other mutants had similar relative activity and expression.
The kinetics of succinate transport was measured in mutants N108C, D329C and L436C in RSO and ISO membrane vesicles. As shown in Table 1, the Km for succinate of mutant L436C in RSO vesicles was approximately 3.5-fold higher than that of the parental C457S. However, the Km for succinate of L436C mutant in ISO vesicles was similar to that of C457S. The other mutants had similar succinate affinity in RSO and ISO compared with C457S. The maximum velocity (Vmax) of L436C and D329C mutants in right-side-out vesicles were significantly lower than C457S. The cysteine-substituted mutants in ISO vesicles showed similar succinate Km and Vmax values as C457S.
To determine whether the substituted cysteines are accessible to MTSET in different conformational states of the transporter, the vesicles were pretreated with MTSET in different buffers: Na+, choline (or Na+-free) and 10 mM succinate in Na+ buffer. Previous experiments suggest that SdcS follows an ordered binding mechanism in which Na+ binds first followed by substrate (3). Therefore, different conformational states are likely to predominate in the presence and absence of Na+ and substrate. SdcS in ISO vesicles was affected similarly by chemical labeling with MTSET in the presence or absence of Na+ and substrate (Figure 7), indicating that the cysteine at position 457 is exposed to the cytoplasm in all conformational states. There was no effect of MTSET on SdcS in RSO vesicles or on the cysteineless mutant, C457S, and L436C in RSO and ISO vesicles (Figure 7). Mutant N108C was accessible to MTSET in both RSO and ISO vesicles, and the inhibition by MTSET was greatest in the presence of sodium. There was substrate protection in both RSO and ISO vesicles containing N108C. Mutant D329C was only accessible to MTSET from the outside of the cell; in the RSO but not the ISO vesicles. MTSET labeling of D329C in RSO required sodium; there was no inhibition in choline buffer, and there was no substrate protection.
The kinetics and functional properties of the Na+/dicarboxylate symporter SdcS from Staphylococcus aureus were investigated by heterologous expression in Escherichia coli followed by the preparation of right-side out (RSO) and inside out (ISO) membrane vesicles. The vesicles enabled us to examine the forward and reverse transport reactions as well as the accessibility of introduced cysteines from both sides of the membrane. Because SdcS is related in sequence to the mammalian Na+/dicarboxylate cotransporters, studies of SdcS should provide information about the transport mechanism of other members of the SLC13 family. A major finding of this study was the identification of amino acid Asn-108 that is accessible from both the cytoplasmic and periplasmic sides of the membrane, indicating that it is found in the translocation pore of the protein.
Succinate transport by SdcS in both RSO and ISO vesicles exhibited an overshoot in concentration above equilibrium, verifying that it is an active transport process. The substrate affinity from the cytoplasmic and extracellular side appeared to be very similar with Km values for succinate between 26 and 36 μM. However, the Vmax of the forward reaction measured in RSO vesicles was higher than in ISO vesicles. It should be noted, though, that the size of ISO vesicles tends to be much smaller than that of RSO vesicles (17), which could contribute to a lower Vmax. Similar to SdcS, the rabbit Na+/succinate cotransporter from renal brush-border membranes (most likely NaDC1) exhibits similar Km values for influx and efflux, but has a Vmax for succinate influx three times higher than that for efflux (20). The succinate kinetic properties of the cysteineless mutant, C457S, were very similar to those of the wild type SdcS.
The C457S mutant exhibited some differences from SdcS in sodium kinetics. Succinate transport by SdcS in RSO vesicles had a mean KNa value of 3.3 mM and was inhibited by sodium concentrations above ~15 mM. In previous whole cell assays, the KNa for sodium activation in SdcS was about 1.5 mM, and there was complete inhibition at high sodium concentrations (3). The cysteineless mutant in RSO vesicles was relatively unaffected by high concentrations of sodium, with a higher mean KNa of 8.2 mM, suggesting that Cys-457 may mediate the inhibition by sodium at particular conformational states of the transporter. In contrast, both the wild-type SdcS and C457S in ISO vesicles were inhibited by sodium concentrations above 7 mM. The transport inhibition at high sodium concentrations has also been reported for another bacterial transporter, Tyt1, assayed in whole cells (21). Interestingly, partially purified SdcS assayed in proteoliposomes no longer exhibits inhibition by high concentrations of sodium, which suggests that a protein in the E. coli membranes mediates the effect. One possible candidate could be the E. coli Na+/H+ exchanger, NhaA, which would transport Na+ into the vesicle in exchange for protons. The activity of NhaA could result in succinate transport inhibition by collapsing the Na+ gradient, trans-inhibition by increased intracellular Na+, or inhibition due to a change in intravesicular pH. SdcS has a pH optimum of about 7–7.4 and alkalinization would result in decreased activity (3). It is possible that the effects are more pronounced in ISO because of their smaller size.
The transport mechanism of SdcS involves ordered binding of two sodium ions followed by binding of substrate (3). A simplified kinetic model with 6 states is shown in Figure 8A. The results of the present study show that residue Asn-108 from SdcS is accessible from the extracellular as well as from the cytoplasmic side of the cell. Asn-108 of SdcS corresponds to Lys-84 of rbNaDC1, which has different accessibility to the outside of the cell during the transport cycle and it is likely located within the substrate binding pocket (7;22). The N108C mutant of SdcS was sensitive to inhibition by MTSET in the conformation adopted in the presence of Na+. In RSO vesicles in Na+ buffer, SdcS should be predominantly in State 2, an outward-facing conformation with Na+ ions bound to the transporter (Figure 8A). This state has the highest affinity for substrate. In ISO vesicles in the presence of sodium, the transporter should be predominantly State 5, the Na+-bound conformation facing the cytoplasmic side of the membrane (Figure 8A). Note that the cytoplasmic side of the membrane faces out in inverted vesicles.
The N108C mutant also showed substrate protection of MTSET labeling, which was due to steric hindrance rather than a conformational change in the protein. Because the mutation of Asn-108 did not change the kinetic properties of the transporter, it is likely that this residue is located near the substrate binding site but is not part of it. Figure 8B shows the secondary structure model of SdcS, based on the rbNaDC1 model, with Asn-108 located approximately midway through TM 3. Asn-108 could be exposed alternately to the inside and outside of the cell by tilting or twisting of TM3. Similar results were seen previously for the UhpT transporter in which cysteine mutants located in the central portion of transmembrane helix 7 are accessible to both the outside and inside of the cell (23;24). These residues also exhibited substrate protection (24). The endogenous cysteine at position 176 in TM5 of the glycerol 3-phosphate transporter, GlpT, is also located in the transport pathway (25). Cys-176 is accessible to pCMBS from both the outside and inside of the cell, and the accessibility is blocked by the presence of substrate. However, an alternate explanation for our results is that Asn-108 of SdcS resides in a reentrant loop that changes accessibility during the transport cycle. The Na+-citrate symporter CitS contains two endogenous cysteines, accessible from both sides of the membrane, that are found in a reentrant loop lining part of the substrate permeability pathway (26).
Mutant D329C of SdcS was accessible to MTSET in the presence but not absence of Na+ and only in RSO vesicles, indicating that its accessibility also changes in the different conformational states of the transporter. The accessibility of D329C would be greatest in states 2 and 3 (Figure 8). This aspartic acid residue is conserved among all members of the SLC13 family. The secondary structure model places this residue at the the extracellular surface of TM 7, explaining its accessibility from the outside (Figure 10). The L436C mutant of SdcS was not affected by MTSET, which could mean that Leu-436 is buried in the helix and not accessible to the outside or that chemical labeling by MTSET does not have any functional effect. This residue appears to determine substrate affinity in the forward direction (from outside to inside) because L436C in RSO membrane vesicles showed a 3.5 fold higher Km compared with ISO vesicles. The other members of the SLC13 family contain either leucine or methionine in this position. By comparison, the M493C mutant (at the equivalent position) of rbNaDC1 is very sensitive to inhibition by MTSET (6).
The results of this study, particularly with the N108C mutant, are consistent with the alternating access model of transport in which the transporter protein contains a single substrate binding site that is alternately accessible from both sides of the membrane. The lactose permease (LacY) and glycerol-3-phosphate transporter (GlpT), both secondary active transporters from the major facilitator superfamily, follow an alternating access mechanism for transport of substrates (23;27–29). The structure of the lactose permease shows a large water-filled cavity containing the amino acid side chains involved in binding protons and substrate (27). Experiments involving cross-linking of residues in LacY show that transport activity involves opening and closing of the hydrophilic cavity (30). The structure of GlpT has a funnel-shaped vestibule leading to the substrate binding site (31). Similar to LacY, the GlpT structure suggests a rocker-switch type movement of two halves of the protein during the transport cycle, involving salt-bridge formation and breakage.
The present work represents the characterization of SdcS in rightside out and inside out membrane vesicles. The kinetic properties of succinate transport in SdcS appear asymmetrical, which may indicate differences in transport velocity (Vmax) rather than substrate affinity. The N108C mutant of SdcS was found to be accessible to membrane impermeant methanethiosulfonate reagents from both sides of the membrane, indicating that this residue may be located within the translocation pathway of the transporter, possibly close to the substrate binding site. The sensitivity of mutant D329C to MTSET inhibition was greatest in RSO vesicles in presence of sodium, whereas residue Leu-436 may be a determinant of substrate affinity in the forward direction, measured in RSO vesicles. In conclusion, this study establishes SdcS as a model system for the eukaryotic NaDC1 transporters and identifies key residues, particularly Asn-108, that may be important during the transport cycle from both sides of the membrane.
We thank Dr. Matthias Quick for the use of his membrane vesicle protocols and discussions about E. coli membrane preparations. Thanks also to Kathleen Randolph for preparation of bacterial media and solutions.
†This work was supported by National Institutes of Health Grant DK46269 (AMP).
1Portions of this work were done at the Department of Biochemistry and Molecular Biology, University of Texas Medical Branch, Galveston, TX 77555.