Cysteine 144 in the third transmembrane domain of the creatine transporter is located close to a substrate-binding site.

All creatine transporters contain a cysteine residue (Cys(144)) in the third transmembrane domain that is not present in other members of the Na+,Cl(-)-dependent family of neurotransmitter transporters. Site-directed mutagenesis and reaction with methane thiosulfonates were used to investigate the importance of Cys(144) for transporter function. Replacement of Cys(144) with Ser did not significantly affect the kinetics or activity of the transporter, whereas a C144A mutant had a higher K(m) (0.33 compared with 0.18 mm). Substitution of Cys(144) with Leu gave a mutant with a 5-fold higher K(m) and a reduced specificity for substrate. Low concentrations of 2-aminoethyl methanethiosulfonate (MTSEA) resulted in rapid inactivation of the creatine transporter. The C144S mutant was resistant to inactivation, indicating that modification of Cys(144) was responsible for the loss of transport activity. Creatine and analogues that function as substrates of the creatine transporter were able to protect from MTSEA inactivation. Na+ and Cl(-) ions were not necessary for MTSEA inactivation, but Na+ was found to be important for creatine protection from inactivation. Our results indicate that cysteine 144 is close to the binding site or part of a permeation channel for creatine.

Creatine and creatine phosphate are essential for the maintenance of ATP levels in tissues with high and fluctuating energy demands such as skeletal muscle and brain (1)(2)(3). Creatine kinase catalyzes the reversible transfer of a phosphate from creatine phosphate to ADP regenerating ATP. In mammals, creatine is either synthesized by sequential reactions occurring in the kidney and liver or obtained from the diet. A specific uptake system for creatine has been demonstrated in skeletal muscle (4), some cultured cell preparations (5), human monocytes and macrophages (6), and astroglial-rich cultures from neonatal rats and mice (7).
Molecular cloning studies have identified muscle and brain cDNAs encoding high affinity sodium-and chloride-dependent creatine transporters (8). The deduced sequence of the rabbit CreaT 1 exhibits significant homology to the Na ϩ and Cl Ϫ -de-pendent GABA/norepinephrine (GAT-1/norepinephrine transporter) gene family of neurotransmitter transporters (9,10). All members of this family are predicted to contain 12 membranespanning domains, to contain a large extracellular loop containing sites for N-linked glycosylation between the third and fourth transmembrane domains, and to have the amino and carboxyl termini facing the cytoplasmic side of the membrane.
In the absence of a three-dimensional structure, indirect approaches have been used to gain insight into the structure and function of transporter proteins. The identification of residues required for substrate selectivity and specificity is particularly important. Chimeras were prepared by exchanging domains between the norepinephrine and dopamine transporters, two transporters that have high sequence similarity but differ in their inhibitor sensitivity. One group suggested that regions spanning TMs 1-3, 11, and 12 were required for substrate affinity (11), whereas another found that TMs 8 -12 were dominant for substrate selectivity and specificity (12). Sitedirected mutagenesis has also been useful in identifying amino acid residues involved in substrate binding. Tyr 140 in TM3 of the GABA transporter (GAT-1) was found to be critical for substrate recognition and transport (13). Replacement of the equivalent residue in the SerT (Tyr 176 ) with cysteine resulted in a reduced affinity for serotonin and cocaine (14). The tyrosine at a position equivalent to 140 of GAT-1 is conserved in all transporters, including the CreaT, and it has been suggested that this residue may be required for a common function (13).
Methanethiosulfonate (MTS) reagents are useful tools to identify residues that may be important for substrate binding and the specificity of neurotransmitter transporters. If cysteine residues are exposed to the external medium or part of a water-accessible channel or binding site, then they will react with polar MTS reagents (15). Two MTS derivatives, 2-aminoethyl methanethiosulfonate (MTSEA) and 2-(trimethylammonium)ethyl methanethiosulfonate (MTSET), are positively charged, whereas 2-sulfonatoethyl methanethiosulfonate (MTSES) is negatively charged. Mutants of the SerT in which the Ile at positions 172 and 179 in TM3 was replaced with cysteine were readily inactivated by MTSET (14). The I172C mutant was protected from MTSET inactivation by substrate, suggesting that it was located in a permeation pathway close to the substrate-binding site (16). This protection was not dependent on temperature or Na ϩ , suggesting that it results from direct occlusion of I172C by serotonin and not secondary to a substrate-induced conformational change. In contrast, Ile 179 of the SerT and the equivalent residue of the norepinephrine transporter, Ile 155 , were found to be in a conformationally sensitive region of TM3.
A comparison of the protein sequence of transporters separated by a large phylogenetic distance may also help identify candidate amino acid residues required for substrate specificity. Guimbal and Kilimann (17) found that the CreaT from the electric ray, Torpedo marmorata, was 64% identical to the rabbit CreaT. Only three residues in predicted TMs were found to be better conserved between CreaTs than other members of the neurotransmitter transporter family. One of these was a cysteine residue in TM3 that is also conserved in rabbit, rat, human, and bovine CreaTs (8, 18 -21). We have investigated the importance of this residue (cysteine 144 in mammalian CreaTs) for the function of the creatine transporter.
Transient Expression of Wild-type and Mutant CreaT in HEK293 Cells-HEK293 cells were grown in minimum essential medium containing 10% fetal bovine serum, streptomycin, and penicillin and plated out at a density of 6 ϫ 10 4 cells/well in 12-well (3.8-cm 2 area) plastic culture dishes (Falcon) 42 h prior to transfection. The cells were transfected with 0.4 g of plasmid and 3.2 l of LipofectAMINE in 0.4 ml of OptiMEM (Life Technologies, Inc./Invitrogen) per well as recommended by the manufacturer. After 5 h the DNA/LipofectAMINE suspension was removed from the cells and replaced with minimum essential medium, 10% fetal bovine serum without antibiotics. After another 19 h, this medium was removed and replaced with normal growth medium. For biotinylation experiments, the cells were plated out at a density of 1.5 ϫ 10 5 cells/well in 6-well (9.6-cm 2 area) plastic culture dishes. For immunofluorescence, 1.5 ϫ 10 5 cells were plated onto polylysine-coated glass coverslips. The cells were transfected with 1 g of plasmid and 8 l of LipofectAMINE in 1 ml of OptiMEM/well.
Creatine Uptake Assay-Cells expressing CreaT or its mutants were assayed for creatine uptake activity 48 h after transfection. The medium was removed from wells, and the cells were washed with 1 ml of prewarmed KRH uptake buffer. The KRH buffer contained 120 mM NaCl, 4.7 mM KCl, 2.2 mM CaCl 2 , 1.2 mM MgSO 4 , 1.2 mM KH 2 PO 4 , 25 mM HEPES, 10 mM glucose, adjusted to pH 7.4 with NaOH. The cells were incubated for 3 min at 37°C with 1 ml of KRH containing various concentrations of [4-14 C]creatine (American Radiolabeled Chemicals, ARC-176, 53 mCi/mmol, diluted to a specific activity of 6 mCi/mmol and a concentration of 2 mM). After the incubation the uptake medium was removed by aspiration, and the cells were washed three times with 1 ml of ice-cold KRH buffer. The cells were solubilized in 0.5% Triton X-100, and aliquots were taken for scintillation counting. Samples were also taken to determine protein concentration using the detergent-compatible protein assay (Bio-Rad) with bovine serum albumin as standard. All assays were carried out in triplicate.
The assay was modified to determine the effect of MTS reagents on creatine uptake. Cells prepared as above were washed once with KRH before incubation for 5 min at 37°C with various concentrations of MTSEA, MTSES, or MTSET (Toronto Research Chemicals). Creatine uptake was then assessed as described above. In protection experiments, the cells were exposed to MTSEA in the presence of creatine or its analogues. The analogues guanidinopropionate, GABA, guanidi-nobutyric acid, guanidinoacetate, cyclocreatine, and arginine were used at a concentration of 1 mM.
In experiments performed without Na ϩ , KRH was prepared with N-methylglucamine in place of NaCl. To prepare KRH without Cl Ϫ , NaCl, KCl, and CaCl 2 were replaced with sodium glucuronate, K 2 HPO 4 , and CaSO 4 , respectively.
Preparation of Antibodies against the CreaT-A 417-bp fragment encoding a 56-residue segment of the predicted carboxyl-terminal region of the bovine CreaT (GenBank TM accession number AF027197) was synthesized by the polymerase chain reaction using the bovine CreaT cDNA as a template. The forward 5Ј-CGGGATCCCTCCT-CAGGGCCAAGGGGACC-3Ј and the reverse 5Ј-GGAATTCCTAAAG-CAGGGATGCTATGGC-3Ј primers encoded sites for BamHI and EcoRI (underlined), respectively, to facilitate subcloning into the vector pGEX-2T (Amersham Pharmacia Biotech). This positioned the CreaT sequence downstream and in-frame with the sequence of glutathione S-transferase. The fusion protein (GST-CreaTC) was expressed in E. coli (DH5␣ or BL21) and purified on glutathione-agarose. Material corresponding to the intact fusion protein was purified on a 12% SDSpolyacrylamide gel and recovered by electroelution (23). Rabbits were injected subcutaneously with 100 g of the fusion protein in Freund's complete adjuvant and boosted three times at monthly intervals with 100 g of protein in Freund's incomplete adjuvant. One week after the final immunization the rabbits were deeply anesthetized and bled out by cardiac puncture. The serum was collected and stored at Ϫ20°C. Prior to use, the antiserum was depleted of antibodies against GST. The serum (10 ml) was diluted with an equal volume of PBS and applied twice to a column containing 8 ml of GST linked to CNBr-activated Sepharose 4BCL (5.7 mg of GST/ml). Preliminary testing of the antiserum depleted of GST-reactive antibodies indicated that most of the specific antibodies were directed toward the carboxyl terminus of the fusion protein. Specific antibodies (␣CreaTC) were purified using a second GST fusion protein containing the carboxyl-terminal 21 amino acids of the CreaT. This was prepared and purified in an identical fashion to that described above except for the use of a new forward primer, 5Ј-CGGGATCCGGCCTGACCACCCTGACCCCA-3Ј, which also contained a BamHI site (underlined). This fusion protein showed no signs of degradation and could be isolated in high yield from E. coli and was therefore more suitable for affinity purification of the antibodies.
Cell Surface Biotinylation-Cell surface expression of the transporters was determined using a membrane-impermeant biotinylation reagent (Pierce) and a method based on that in Ref. 24. Cells expressing CreaT or its mutants were labeled with sulfo-NHS-SS-biotin 48 h after transfection. The medium was removed from wells, and the cells were washed three times with 2 ml of ice-cold PBS. The PBS buffer contained 137 mM NaCl, 2.7 mM KCl, 1.5 mM KH 2 PO 4 , 1 mM Na 2 HPO 4 , 0.1 mM CaCl 2 , 1.2 mM MgCl 2 , adjusted to pH 7.2 with HCl. The cells were incubated for 30 min on ice with 0.5 ml of 1.5 mg/ml sulfo-NHS-SSbiotin (Pierce) in biotinylation buffer (150 mM NaCl, 2 mM CaCl 2 , 10 mM triethanolamine, pH 7.5). The biotinylation step was repeated with a fresh aliquot of sulfo-NHS-SS-biotin. This reagent was removed by aspiration, and the cells were washed twice with 2 ml of ice-cold PBS containing 100 mM glycine. The cells were quenched for 30 min on ice with this same buffer before washing again with PBS. Lysis buffer (0.25 ml) was added to each well, and the plates were incubated for 30 min on ice with gentle rocking. Lysis buffer consisted of 137 mM NaCl, 20 mM Tris, 1 mM EDTA, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, protease inhibitors (Complete TM Mini protease inhibitor mixture; Roche Molecular Biochemicals) and was adjusted to pH 7.5 with HCl. Lysates were transferred to 1.5-ml microcentrifuge tubes and centrifuged at 20,000 ϫ g for 20 min at 4°C. NeutrAvidin TM beads (Pierce) were prepared as a 50% suspension in lysis buffer, and 50 l were added to 0.19 ml of each supernatant from the lysate samples. Biotinylated proteins were allowed to bind to the beads for 1 h at 4°C with occasional mixing. The NeutrAvidin TM beads were pelleted by centrifugation at 5000 ϫ g at 4°C for 15 min. After removing the unbound supernatant fraction, the pellet was washed three times with 0.5 ml of lysis buffer, twice with 0.5 ml of a high salt lysis buffer (as above but with 0.5 M NaCl), and twice with 0.5 ml of a no-salt lysis buffer. The residual wash was completely removed, and the biotinylated proteins were extracted from the beads by adding 50 l of 2ϫ SDS reducing buffer (125 mM Tris, 4% SDS, 20% glycerol, and 10% ␤-mercaptoethanol) and incubating for 30 min at 37°C followed by centrifugation at 10,000 ϫ g for 5 min. Total, unbound, and biotinylated proteins were analyzed by SDS-polyacrylamide gel electrophoresis and Western blotting as described previously (25). Affinity-purified ␣CreaTC antibodies were used at a concentration of 0.5 g/ml.
Immunofluorescence Microscopy-The expression of CreaT mutants in HEK293 cells was examined by immunofluorescence microscopy. The culture medium was removed, and the cells were washed twice with 2 ml of PBS. The cells were fixed for 15 min in 3% paraformaldehyde in PBS, washed twice, and then quenched for 15 min with 15 mM glycine in PBS. The cells were washed again in PBS and permeabilized by treatment with 0.5% Triton X-100 and 10% fetal bovine serum in PBS for 10 min. The cells were incubated with affinity-purified ␣CreaTC antibodies (10 g/ml) in PBS containing 10% fetal bovine serum for 1 h at room temperature. After incubation the cells were washed four times with PBS for 3 min each. The cells were then incubated with Cy3 TM labeled goat anti-rabbit IgG (Amersham Pharmacia Biotech) diluted 1:500 in PBS containing 10% fetal bovine serum for 1 h at room temperature. The cells were washed as described above, and the coverslips were rinsed briefly in water and mounted onto microscope slides with 15 l of Vectashield® (Vector Laboratories). Images were collected on a confocal microscope at the Biomedical Imaging Research Unit in the Department of Anatomy, University of Auckland.

RESULTS
Creatine transporters contain a uniquely conserved cysteine (Cys 144 ) in the third transmembrane domain, whereas other members of the Na ϩ , Cl Ϫ -neurotransmitter transporter family have more bulky and hydrophobic side chains at this position ( Fig. 1). In this study we have investigated the importance of Cys 144 on the activity of the CreaT by both site-directed mutagenesis and reaction with MTS derivatives.
Cys 144 Is Important for the Activity of the CreaT-Variants of the CreaT were prepared in which Cys 144 was replaced by Ser, Ala, or Leu. The latter residue is found in an identical position in the homologous GABA transporter (GAT-1). The activity of the C144S, C144A, and C144L mutants were 75, 55, and 15% of the wt CreaT (Fig. 2). The activity of all mutants was significantly above the level of endogenous creatine uptake activity of HEK293 cells. Cells transfected with empty pcDNA3.1 ϩ vector showed ϳ3% of the activity obtained by transfection with wt CreaT (0.75 nmol/min/mg protein).
We compared the expression of CreaT proteins by Western blotting and immunofluorescence. Similar levels and molecular forms of CreaT immunoreactivity were seen in cells expressing C144S, C144A, or C144L or wt CreaT. Treatment of transfected cells with sulfo-NHS-SS-biotin followed by purification of biotinylated proteins on NeutrAvidin TM beads showed expression of the wt and all mutant transporters at the cell surface. The biotinylated fraction from the wt CreaT and all three Cys 144 mutants contained a 90-kDa form (Fig. 3A) that corresponds to the glycosylated form of the CreaT (data not shown). The re-sults of immunofluorescence microscopy were also very similar for the Cys 144 mutants and wt CreaT (Fig. 3B). It appears that differences in expression or altered trafficking to the cell surface are not the reason for the changes seen in the activity of Cys 144 CreaT variants.
The saturation kinetics of creatine transport in cells expressing C144S, C144A, or C144L mutants or wt CreaT were compared (Fig. 4). The K m of the C144S mutant (0.21 mM) was very similar to the wt CreaT (0.20 mM). However, the K m values of the C144A (0.33 mM) and C144L (0.82 mM) variants were significantly higher than wt CreaT. The large increase in the K m seen for the C144L variant suggested that this mutant may have lost some specificity for substrate recognition. We compared the ability of GABA to compete with creatine uptake in cells expressing C144L and wt CreaT. In the presence of 1 and 10 mM GABA the activity of the C144L mutant was reduced by 26.3 and 62.8%, respectively. GABA at 1 mM had no effect on the wt CreaT, whereas 10 mM GABA decreased uptake activity by 6.2% (data not shown). It appears that the amino acid side chain at position 144 is an important determinant of the specificity of the CreaT for creatine.
The CreaT Is Inactivated by MTSEA-We tested the ability of three charged methanethiosulfonates, MTSEA (2.5 mM), MTSES (10 mM), and MESET (1 mM) to inhibit the CreaT (Fig. 5). These concentrations were chosen because they allow for the intrinsic reactivity of these MTS derivatives in solution (26). No effects were seen with MTSES and MTSET, but MTSEA inhibited creatine transport activity by Ն95%. It appears that the CreaT is very sensitive to MTSEA. Exposure of HEK293 cells expressing wt CreaT to 250 or 50 M MTSEA for 1-2 min resulted in 92-95% and 85-95% inhibition, respectively (data not shown).
MTSEA Modification of Cys 144 Results in CreaT Inactivation-To determine whether Cys 144 was the target of MTSEA modification, cells expressing C144S and C144A mutants or wt CreaT were treated with MTSEA prior to determining creatine transport activity. The activity of the C144S and C144A variants was unaffected by MTSEA under conditions (incubation with 50 M MTSEA for 5 min) that resulted in the loss of the majority of wt CreaT activity (Fig. 6). This identifies Cys 144 as the site modified by MTSEA. The C144S variant retains ϳ70% of wt CreaT activity and is resistant to MTSEA. Thus, the C144S variant will provide a valuable background to study the structure and function of CreaT by cysteine-scanning mutagenesis.
Creatine and Its Analogues Protect the CreaT from Inactivation by MTSEA-Cells expressing wt CreaT were treated with  (21), human (20), rat (18,19,31), and rabbit (8). The T. marmorata CreaT sequence is from Ref. 17. CreaT sequences are compared with the rat GABA transporter, GAT-1 (32), rat taurine transporter, TauT (33), and the rat serotonin transporter, SerT (34,35). MTSEA in both the absence and presence of various concentrations of creatine. The presence of creatine was found to prevent inactivation by MTSEA in a dose-dependent manner (Fig. 7). The concentration giving ϳ50% protection was similar to the known K m (ϳ0.2 mM) of the transporter. We also compared creatine and various creatine analogues for their ability to protect from MTSEA inactivation ( Table I). Analogues that are known to be substrates of CreaT, guanidinopropionate, cyclocreatine, guanidinobutyric acid, and guanidinoacetate all reduced the inactivation by MTSEA. The degree of protection correlated with their known potency to competitively inhibit creatine transport. GABA and arginine, two molecules that are not substrates, did not protect the CreaT from inactivation by MTSEA.
Ion Dependence of MTSEA Inactivation and Creatine Protection-Cells expressing the CreaT were treated with MTSEA in media lacking Na ϩ and Cl Ϫ ions. The absence of Na ϩ or Cl Ϫ had no effect on inactivation by MTSEA (Table II). By comparison the ability of creatine to protect from MTSEA inactivation was much reduced in the absence of Na ϩ . DISCUSSION A cysteine residue (Cys 144 ) is found in the third transmembrane domain of all CreaTs but not other members of the Na ϩ ,Cl Ϫ -dependent neurotransmitter transporter family. We have found that modification of this residue with MTSEA leads to rapid inactivation of the transporter and that substrates protect from this inactivation, whereas a C144S mutant retains ϳ70% of transporter activity. Our results are consistent with the view that Cys 144 is located in a substrate-binding site and that it is an important but not an essential determinant of creatine transport.
There are parallels between the results we have obtained for Cys 144 of the creatine transporter and those obtained by others for Cys 148 of the E. coli lactose transporter (see Ref. 27 for a review). Neither of these cysteine residues are required for creatine/Na ϩ or lactose/H ϩ symport. However, in each case the transporter is protected from inactivation from sulfhydryl modifying reagents by specific substrates. Replacement mutants of lac permease were used to distinguish between the possibilities that Cys 148 was either (i) a component of the substrate-binding site or (ii) far removed from the binding site with long range conformational effects providing substrate protection from in- Samples of total cell lysates and biotinylated (surface) proteins were run on a 10% SDS-PAGE gel, blotted onto nitrocellulose, and probed for creatine transporter using the ␣CreaTC antibody. The relative molecular masses (kDa) of protein standards are shown on the left. B, cells expressing wt CreaT and C144S, C144A, or C144L CreaT mutants were fixed, permeabilized, and probed with the antibody ␣CreaTC followed by Cy3-labeled anti-rabbit IgG before imaging by confocal microscopy. activation. Evidence that the size and polarity of the side chains affected the activity and specificity of the lactose transporter led to the conclusion that Cys 148 was part of a substratebinding site (28). Similarly, we have found differences in both the initial rates and saturation kinetics of mutants in which the Cys 144 of the CreaT was replaced with Ser, Ala, or Leu. Although C144S had a K m (0.21 mM) close to that of wt CreaT (0.20 mM), the Ala mutant had a higher K m (0.33 mM), indicating a lower affinity for substrate. Both Ser and Ala are smaller than Cys and are often used interchangeably to replace cysteine in scanning mutagenesis studies. However, Ser is more hydrophilic than Ala, and this may be important if the side chain of residue 144 is in a water-accessible site. The rapid inactivation of CreaT by low concentrations of MTSEA (50 M) suggests that Cys 144 may be present in an ionized, thiolate (RS Ϫ ) form. A thiolate is more likely to be present in a wateraccessible site and is reported to react a billion times faster with methanethiosulfonates than un-ionized thiols (15). It also appears that size and hydrophobicity may decrease the affinity of the CreaT for substrate because the K m of the C144L mutant was approximately four times higher than that of the wt CreaT. This mutation also resulted in a decreased specificity of the transporter because high concentrations of GABA inhibited creatine uptake by the C144L mutant but not the wt CreaT.  MTSES are also larger than MTSEA and contain bulky charged groups that may prevent them entering a substrate translocation pathway for creatine. The rapid reaction with low concentrations of MTSEA and the ability of creatine to protect from inactivation suggest Cys 144 to be located in a wateraccessible site. However, it is difficult to determine whether MTSEA enters this site from the extracellular or cytoplasmic side of the membrane. Creatine appears to protect from MTSEA inactivation by being transported, because protection is significantly reduced by the absence of Na ϩ or Cl Ϫ . We thus favor a view in which Cys 144 is accessible from the extracellular side of the membrane. Na ϩ or Cl Ϫ are not required for MTSEA inactivation, so it appears unlikely that a conformational change is induced by these ions. A close correlation was found for the concentration of creatine required for 50% protection and the experimentally determined K m for creatine transport (ϳ0.2 mM). Also, the rank order of potency of creatine analogues that protect from MTSEA (guanidinopropionate Ͼ creatine Ͼ cyclocreatine Ͼ guanidinobutyric acid Ͼ guanidinoacetate) is in accordance with known affinities of these substrates for the transporter. Furthermore, only substrates of the CreaT could protect from MTSEA inactivation. No protection was obtained with GABA or arginine, derivatives that contain functional groups present in creatine but are not competitive substrates of CreaT. These data are consistent with the view that Cys 144 is close to or part of the creatine-binding site of the CreaT.
The identification of a residue within the third transmembrane domain of the CreaT as part of a substrate binding site is consistent with research on other members of the Na ϩ ,Cl Ϫ -dependent neurotransmitter transporter family. Tyr 140 in TM3 of GAT-1 has been found to be critical for substrate recognition (13). All members of this transporter superfamily contain a Tyr in an equivalent position, and Bismuth et al. (13) suggested a role for this residue in binding an amino-group. A mutant form of the SerT in which this Tyr was replaced with a cysteine had reduced serotonin and cocaine binding (14). It was suggested that Ile 172 and Ile 179 of SerT are part of a serotonin-and cocaine-binding site. Further studies from this group led to the conclusion that TM3 constitutes part of a permeation pathway for both SerT and norepinephrine transporter (16). Cys 144 of the CreaT and Ile 172 of the SerT both occupy equivalent positions within the predicted helices of TM3 ( Fig. 1 and Refs. 16 and 17). The cysteine residues in wild-type CreaT and the I172C SerT mutant react with MTS reagents, and in each case the substrate protects from inactivation. The present work with the naturally occurring cysteine at position 144 of the CreaT strengthens the importance of this region of TM3 for substrate binding by Na ϩ ,Cl Ϫ -dependent neurotransmitter transporters.
Cysteine-scanning mutagenesis is a powerful technique when combined with hydrophilic MTS reagents to study structure-function relationships and the topology of membrane proteins. The ideal starting point is a cysteine-less protein. This has been possible for only a few membrane proteins, the lactose permease, the Nha-Na ϩ /H ϩ transporter, and the glutamate transporter (17,29,30). The CreaT contains 22 cysteine residues, several of which are conserved in the related GABA and SerT and are likely to be involved in disulfide bonds. It appears that production of a cysteine-less CreaT that retains function would be very unlikely. We have found the C144S mutant of the creatine transporter to retain ϳ70% of the activity of wt CreaT and be resistant to MTS reagents. This mutant will provide an excellent background for probing residues in TM3 and other interesting domains of the CreaT by cysteine-scanning mutagenesis.