Sites important for Na+ and substrate binding in the Na+/proline transporter of Escherichia coli, a member of the Na+/solute symporter family.

To elucidate the functional importance of transmembrane domain II in the Na(+)/proline transporter (PutP) of Escherichia coli we analyzed the effect of replacing Ser-54 through Gly-58. Substitution of Asp-55 or Met-56 dramatically reduces the apparent affinity for Na(+) and Li(+) in a cation-dependent manner. Conversely, Cys in place of Gly-58 significantly reduces only the apparent proline affinity while substitution of Ser-57 results in a dramatic reduction of the apparent proline and cation affinities. Interestingly, upon increasing the proline concentration the apparent Na(+) affinity of Ser-57 replacement mutants converges toward the wild-type value, indicating a close cooperativity between cation and substrate site(s). This notion is supported by the fact that Na(+)-stimulated site-specific fluorescence labeling of a single Cys at position 57 is completely reversed by the addition of proline. Similar results are obtained upon labeling of a Cys at position 54 or 58. Taken together, these results indicate that Asp-55 and Met-56 are located at or close to the ion-binding site while Ser-54, Ser-57, and Gly-58 may be close to the proline translocation pathway. In addition, the data prod at an involvement of the latter residues in ligand-induced conformational dynamics that are crucial for cation-coupled transport.

The Na ϩ /solute symporter family (SSF, 1 TC 2.A.21) comprises more than 80 similar proteins of pro-and eukaryotic origin (1)(2)(3). These integral membrane proteins utilize the Na ϩ electrochemical gradient to drive the coupled uphill transport of a variety of substrates (sugars, amino acids, vitamins, osmolytes, ions, myo-inositol, urea, and water). Among the eukaryotic members of the SSF the Na ϩ /glucose transporter (SGLT1), the Na ϩ /iodide transporter, and the Na ϩ /multivita-min transporter are directly implicated in human disorders (4 -6). It has been proposed that all members of this family possess a common topological motif consisting of 13 ␣-helical transmembrane domains (2,(7)(8)(9). One of the current major challenges in the investigation of these transporters is the identification of regions involved in the binding and translocation of Na ϩ and substrate. Studies on SGLT1 implicate the C-terminal part of the transporter in sugar binding and transport (10 -12). However, little is known about the role of residues in the N-terminal domain of SSF members, and the coupling mechanism of this class of membrane proteins is well far from understood. To further identify functional important regions we are utilizing PutP of Escherichia coli as a model system.
PutP is an integral protein of the cytoplasmic membrane of E. coli. The transporter has been purified, reconstituted into proteoliposomes, and shown to be solely responsible for the coupled translocation of Na ϩ and proline (13,14). Labeling experiments, random and site-directed mutagenesis have been employed to identify functionally important sites in PutP of E. coli and Salmonella typhimurium (see Ref. 15, for a recent review). Recent results from our laboratory indicate the functional importance of putative transmembrane domain (TM) II in cation-coupled proline transport (16 -18). In particular, Asp-55 located in the center of TM II is the only residue known to be essential for PutP activity (16). While a neutral amino acid completely impairs all types of transport, Glu at this position reduces the apparent Na ϩ affinity dramatically but has only little effect on proline binding, implying a direct involvement of Asp-55 in Na ϩ binding. Another residue in TM II, Ser-57, is crucial for high-affinity proline uptake (18). Replacement with Ala, Cys, Gly, or Thr decreases the apparent affinity for proline by up to 2 orders of magnitude with little influence on the maximum rate of transport. Furthermore, substitution analysis and site-directed Cys labeling studies of Arg-40, located at the cytoplasmic interphase of TM II suggest that this conserved residue is important for the efficient coupling of ion and proline transport (17). In this vein, data obtained upon substitution of conserved Asp-187 (located in the putative cytoplasmic loop between TM V and VI) indicate that electrostatic interactions of the amino acid side chain at position 187 in PutP with other parts of the transporter and/or the coupling ion are important for active proline transport (19).
Based on the initial findings on the role of Asp-55 and Ser-57 in Na ϩ -coupled proline uptake we analyzed the functional importance of amino acids in the vicinity of both residues, namely Ser-54, Met-56, and Gly-58. In addition, former work on Asp-55 and Ser-57 was extended. To this end, we systematically studied the effect of individual amino acid substitution on (i) proline uptake activity, (ii) ion selectivity, and (iii) the influence of * This work was supported by Deutsche Forschungsgemeinschaft Grant SFB431/D4. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  ligand binding on the accessibility of Cys individually placed at positions 54 to 58. The results indicate that the region comprising Ser-54 to Gly-58 is particularly crucial for binding and translocation of the coupling ion (Na ϩ or Li ϩ ) and the substrate. Furthermore, it is shown that this part of the protein undergoes functional relevant conformational alterations that are essential for cation-coupled transport.

EXPERIMENTAL PROCEDURES
Materials-L-[ 14 C]Proline (247 Ci/mol) was obtained from ICN. Mouse anti-FLAG-M2-antibody was from Integra Biosciences and sheep anti-(mouse IgG)-horseradish peroxidase conjugate was from Amersham Bioscience, Inc. Restriction endonucleases, Taq-DNA polymerase, T4-DNA ligase, and alkaline phosphatase were purchased from New England Biolabs. Synthetic oligonucleotides were from Eurogentec or Invitrogen. E. coli DH5␣ competent cells were obtained from Invitrogen. Fluorescein 5-maleimide was from Molecular Probes. The SDS-PAGE low-range molecular mass standard was purchased from Bio-Rad. n-Dodecyl-␤,D-maltoside was from Anatrace and Ni 2ϩ -NTA spin columns were from Qiagen. All other chemicals were of analytical grade and obtained from commercial sources.
Site-directed Mutagenesis-Amino acid substitutions were created using a two-step polymerase chain reaction (PCR) protocol with plasmids pT7-5/putP or pT7-5/putP(⌬Cys) as templates. Mutagenic primers encoding the desired amino acid substitution, and suitable sense and antisense primers binding up-and downstream, respectively, of the site of substitution were used similarly as described (16,18). PCR fragments were digested with restriction endonucleases ApaI and XbaI and ligated to similarly treated plasmids pT7-5/putP and pT7-5/putP(⌬Cys) incubated with alkaline phosphatase to avoid religation of the vector. For overexpression of PutP, the mutated putP genes were cut with NcoI and HindIII and ligated to similarly treated vector pTrc99a incubated with alkaline phosphatase. The resulting constructs were verified by sequencing double-stranded plasmid DNA using dideoxynucleotide chain termination (23) after alkaline denaturation (24).
Transport Assay-Active transport was measured in E. coli WG170 (PutP Ϫ A Ϫ ) harboring derivatives of plasmids pT7-5/putP or pT7-5/ putP(⌬Cys) encoding PutP with given amino acid replacements. The cells were grown aerobically in Luria-Bertani medium (25) containing 100 g/ml ampicillin at 37°C. Overnight cultures were diluted 25-fold and were allowed to grow to an optical density at 420 nm (A 420 ) of 1.0, followed by induction with 0.5 mM isopropyl-1-thio-␤-D-galactopyranoside for 2 h. Cells were harvested by centrifugation at 13,200 ϫ g for 10 min and washed up to 6 times with 100 mM Tris/Mes, pH 6.0, at 4°C to reduce the Na ϩ contamination below 5 M. For transport assays, cells were resuspended in the same buffer, and adjusted to a total protein concentration of 0.35 mg/ml. Transport of 10 M L-[U-14 C]proline (if not otherwise indicated) was assayed under standard test conditions in the presence of 20 mM D-lactate (Na ϩ salt) and 50 mM NaCl or, for determination of the ion dependence on proline uptake (apparent affinities for Na ϩ and Li ϩ ), in the presence of 0.005 to 250 mM NaCl or LiCl at 25°C using the rapid filtration method as described (16). Initial rates of transport were calculated from the initial linear portion of the time course and data were plotted according to Eadie-Hofstee. Standard deviations were determined from at least three independent experiments.
Cysteine Accessibility Analysis-E. coli WG170 harboring derivatives of plasmid pTrc99a/putP(⌬Cys) encoding PutP(⌬Cys) with given amino acid replacements was grown aerobically (500 ml of culture) to an A 420 of 1.5 before induction with 0.5 mM isopropyl-1-thio-␤-D-galactopyranoside and further growth for 2 h. Cells were harvested, washed in 100 mM Tris/Mes, pH 7.0, and sonicated for 4 ϫ 30 s using a Branson Sonifier. Cell debris was removed by centrifugation at 13,200 ϫ g for 10 min at 4°C and the supernatant was subjected to ultracentrifugation at 230,000 ϫ g for 45 min at 4°C. The resulting membrane pellet was washed in 100 mM Tris/Mes, pH 7.0, resuspended in the same buffer, and stored in liquid nitrogen until use.
After thawing, 150-l aliquots of the membrane suspension (protein concentration adjusted to 10 mg/ml) were preincubated in the presence or absence of ligand at 25°C for 10 min. Subsequently, 200 M fluorescein 5-maleimide was added and incubation was continued for further 10 min. The reactions were stopped by addition of 5 mM ␤-mercaptoethanol. PutP was solubilized from the membrane and purified by Ni 2ϩ -NTA affinity chromatography as recently described (8). Equal amounts of protein were subjected to 10% SDS-PAGE. Fluorescein 5-maleimide fluorescence of PutP was visualized using the Multi-Imager TM (Bio-Rad, Munich) with an excitation wavelength of 360 nm and quantified by means of the Quantity One TM or MultiAnalyst TM software. After staining the gel with Coomassie Blue the density of the stained bands served to correct the fluorescence values for differences in the amount of protein using the ImageQuant TM software. The corrected fluorescence value of each PutP derivative obtained upon labeling in the absence of ligand was set to 100. Standard deviations were determined from at least three independent experiments.
Immunological Analysis-Relative amounts of PutP wild-type and PutP with given amino acid replacements in membranes of E. coli WG170 were estimated by Western blot analysis. Immunoblotting was performed with mouse anti-FLAG IgG against the Flag epitope at the C terminus of each PutP variant followed by incubation with horseradish peroxidase-linked sheep anti-(mouse IgG) antibody by the enhanced chemiluminescence method as described (18).
Protein Determination-The protein concentration of membrane suspensions was determined by the method of Peterson (26). Solubilized protein was quantified according to Bradford (27). Bovine serum albumin was used as standard.

Generation of Mutants
Asp-55 and Ser-57 in TM II of PutP were demonstrated to be important for ligand binding and transport (16, 18). To elucidate the function of amino acids located in the vicinity of these crucial residues, Ser-54, Met-56, and Gly-58 were individually replaced with Cys by oligonucleotide-directed, site-specific mutagenesis. In addition, Ser-54 was substituted by Ala and Thr. PutP molecules with verified amino acid replacements were subjected to detailed kinetic analyses and site-directed thiollabeling studies. Other mutants used in this study (PutP-D55C, 2 -D55E, -D55N, -S57A, -S57C, -S57G, and -S57T) were generated and initially characterized previously (16, 18).

Time Course of Proline Uptake by PutP Containing
Substitutions for Ser-54, Met-56, or Gly-58 To determine the effect of the newly generated amino acid substitutions on PutP activity, active transport of proline was assayed under standard test conditions (50 mM Na ϩ , 10 M proline) by using E. coli WG170 which lacks PutP and PutA (proline dehydrogenase), and therefore, cannot metabolize proline. Cells transformed with plasmids encoding PutP-S54A, -S54C, or -S54T exhibited initial rates and steady-state levels of proline accumulation comparable to those observed for PutP wild-type (Fig. 1). Severe effects on Na ϩ -coupled proline uptake were observed upon replacement of Met-56 and Gly-58. In particular the initial rate of proline uptake was reduced below 10% compared with PutP-wild-type (PutP-M56C, 6.7 Ϯ 1.4%, -G58C, 1.9 Ϯ 0.8%), while the steady-state level of proline accumulation was less affected.

Immunological Analysis
Relative concentrations of PutP molecules bearing given replacements in membranes of E. coli WG170 were approximated by Western blot analysis using an anti-FLAG-antibody as de-scribed (18). PutP-wild-type, -S54A, -S54C, -S54T, -M56C, and -G58C were present in the membrane in comparable amounts (Fig. 2). Therefore, the differences in proline uptake kinetics described between wild-type transporter and transporter molecules with given replacements cannot be attributed to defective insertion of the transporter into the membrane or to enhanced proteolytic degradation of the modified proteins subsequent to insertion. Similar observations were made previously for PutP containing substitutions for Glu-55 or Ser-57 (16, 18).

Kinetic Analyses
To further characterize the effect of the individual replacement of amino acids Ser-54 through Gly-58 on PutP function, the apparent proline affinity (K m Pro ) was determined under Na ϩ saturation (50 mM Na ϩ ). Furthermore, to obtain information on a possible participation of the described region in cation binding and/or coupling of cation and substrate transport, proline uptake was analyzed with varying Na ϩ or Li ϩ concentrations. The proline concentration in the latter experiments was 10 M (5-fold above the apparent K m Pro of PutP-wild-type) if not otherwise indicated. The cation concentration causing half-maximum stimulation of proline uptake (K 0.5 Na ϩ or K 0.5 Li ϩ ) was used as a measure of the apparent cation affinity, while V max Na ϩ and V max Li ϩ correspond to the maximum rates of proline uptake under saturating Na ϩ or Li ϩ concentrations. The obtained kinetic data led to a classification of the mutants into three groups and the results are presented accordingly (Table I).

Mutants with an Altered Ion Dependence (Group I Mutants)-Group I comprises PutP bearing substitutions for
Asp-55 or Met-56. The mutants are characterized by their low apparent affinity for the coupling ion. So, similar as previously reported for PutP-D55E (16), also substitution of Met-56 by Cys caused severe effects on K 0. 5 Na ϩ (20-fold increase) and V max Na ϩ (reduction to 14% of the wild-type value) (Table I). Furthermore, replacement of Asp-55 with Glu or Met-56 with Cys had a much more pronounced effect on K 0.5 Li ϩ than on K 0.5 Na ϩ (K 0.5 Li ϩ Ͼ 100 mM) yielding a dramatic change in the K 0.5 Li ϩ to K 0.5 Na ϩ ratio. While this ratio for PutP wild-type was about 5, PutP-D55E and -M56C exhibited a ratio of K 0.5 Li ϩ to K 0.5 Na ϩ of above 100, thereby suggesting a more distinct discrimination between the two coupling cations. Contrary to the apparent Li ϩ affinity, the maximum rate of Li ϩ -coupled proline uptake (V max Li ϩ ) of both mutants was in the same order of magnitude as the corresponding Na ϩ value. Compared with the apparent ion affinities, the apparent proline affinity was only moderately altered upon substitution of Asp-55 by Glu or Met-56 by Cys (about 8-fold increase of K m Pro ) (Table I). Finally, as shown for Na ϩ (16), Li ϩ also proved unable to drive proline transport by PutP-D55C and -D55N.
Mutants with Altered Affinities for Coupling Ion and Proline (Group II Mutants)-Group II encompasses PutP containing substitutions for Ser-57 and Gly-58. Group II mutants differ from group I mutants by their highly reduced apparent affinities for proline in addition to defects in ion binding (Table I).
Most severe effects were observed upon replacement of Ser-57. Substitution of this residue by Ala, Cys, Gly, or Thr increased K m Pro by up to 2 orders of magnitude (18). In addition, this study revealed severe defects in ion binding. Depending on the nature of the amino acid side chain at position 57, K 0.5 Na ϩ was increased by factors of 12 (PutP-S57T), 42 (-S57C), 45 (-S57G), and 79 (-S57A). PutP-S57A, -S57C, and -S57T exhibited a V max Na ϩ Յ 4% of the wild-type activity. A higher but still reduced V max Na ϩ was observed for PutP-S57G (40% compared with PutP-wild-type). Since the proline concentration used for the latter studies (10 M) was far below the apparent proline affinity of PutP containing replacements for Ser-57, the Na ϩ dependence of proline uptake of PutP-S57A and -S57C was also analyzed at higher proline concentrations. Saturating proline concentrations could not be used since the ion stimulatory fraction of the transport signal was to low to allow a precise calculation of kinetic parameters. However, in the presence of 100 M proline K 0.5 Na ϩ for PutP-S57A and -S57C was about 8-fold reduced compared with K 0.5 Na ϩ determined at 10 M proline. Furthermore, the V max Na ϩ obtained at 100 M proline was ϳ5-fold increased compared with that determined in the presence of 10 M proline. The 10-fold increase of the proline concentration had no significant effect on K 0.5 Na ϩ and V max Na ϩ for PutP-wild-type (data not shown). Substitution of Ser-57 with Cys, Gly, or Thr affected V max Li ϩ and K 0.5 Li ϩ to a similar extend as observed for the corresponding Na ϩ parameters (Table I). However, Li ϩ did not stimulate proline uptake by PutP-S57A (Li ϩ concentrations of up to 250 mM were used) which was independent on the proline concentration (10 or 100 M).
Substitution of Gly-58 by Cys resulted in a highly reduced apparent affinity for proline (40-fold increase of K m Pro ), similar as described above for PutP containing substitutions for Ser-57. The apparent Na ϩ affinity of PutP-G58C was less dramatically affected (5-fold increase of K 0.5 Na ϩ ) while the apparent Li ϩ affinity was more severely altered (19-fold increase of K 0.5 Li ϩ ). The changes in the apparent ligand affinities of the mutant were accompanied by reduced maximum rates of proline uptake (30-fold reduction of V max Na ϩ ; 8-fold reduction of V max Li ϩ ).

FIG. 2. Immunological detection of PutP containing replacements of amino acids in TM II.
Cells were cultivated as described under "Experimental Procedures." Ten micrograms of total cells protein from each culture were separated by 10% SDS-PAGE. Proteins were then transferred onto a nitrocellulose membrane (0.45-m pore size) and hybrid proteins were probed with mouse anti-FLAG M2 monoclonal antibody followed by incubation with horseradish peroxidase-linked sheep anti-mouse IgG antibody.

Mutants with Minor Effects on Kinetic Parameters (Group III
Mutants)-Group III contains PutP with substitutions for Ser-54. Substitution of Ser-54 by Ala, Cys, or Thr had the least effect on all kinetic parameters tested. The apparent affinity of the transporter for proline (K m Pro ) was not significantly affected by the replacements (Table I). Furthermore, K 0.5 Na ϩ and K 0.5 Li ϩ were increased by factors of maximum 6 and 8 (PutP-S54C). V max Na ϩ and V max Li ϩ were in the same order of magnitude as the corresponding wild-type values.

Influence of Ligands on Cys Accessibility
To obtain further information on the role of this region in ligand binding and ligand-induced conformational alterations, the accessibility of Cys residues placed individually at positions 54 to 58 was analyzed in the absence and presence of different ligands (Na ϩ , Li ϩ , and proline). For this purpose Cys was individually introduced into a functional PutP derivative devoid of all five native Cys residues (PutP(⌬Cys)) (8). The effects of the substitutions on transport activity of Cys-free PutP were similar to that observed in the wild-type background (data not shown). All mutants exhibited Na ϩ -coupled proline uptake except for PutP(⌬Cys)-D55C which was already reported to be completely inactive in an earlier study (16). Labeling experiments were performed with randomly oriented, de-energized membrane vesicles using the fluorescent dye fluorescein 5-maleimide as sulfhydryl reagent.
Contrary to the Cys-free transporter all five single-Cys PutP derivatives showed significant fluorescence labeling and the degree of labeling varied with the position of the Cys and the absence or presence of ligand (Fig. 3). In the absence of ligand (the Na ϩ concentration of the buffer was below 5 M) highest relative fluorescence was observed upon incubation of PutP(⌬Cys)-S54C (group III mutant), -D55C, and -M56C (group I mutants) with the fluorescence probe while PutP(⌬Cys)-S57C and -G58C (group II mutants) were only poorly labeled under these conditions (Fig. 3, lanes 1). Addition of Na ϩ or Li ϩ had little or no effect on the fluorescence labeling of Cys at positions 54, 55, and 56 (Fig. 3, lanes 2 and 4). However, Na ϩ ions caused a substantial stimulation of the labeling reaction of PutP(⌬Cys)-S57C and -G58C. In contrast to Na ϩ , Li ϩ ions did not significantly affect labeling at position 57 and had only a small stimulatory effect on the reaction of Cys at position 58 with the fluorescent maleimide. In the absence of either cation no significant change of labeling was observed for any single Cys-PutP derivative upon the addition of proline (not shown). However, proline caused efficient protection of Cys at position 54 in the presence of Na ϩ or Li ϩ (Fig. 3, lanes 3 and  5). Furthermore, in the presence of Na ϩ , proline inhibited labeling of PutP(⌬Cys)-S57C and G58C thereby reversing the stimulatory Na ϩ effect. Labeling of PutP(⌬Cys)-M56C was slightly reduced by proline in the presence of Na ϩ . Finally, reaction of PutP(⌬Cys)-D55C with the fluorescent probe remained unaffected under all conditions tested. DISCUSSION The present report investigates the role of amino acids Ser-54 to Gly-58 in TM II of PutP in ligand binding and the coupling of cation and substrate transport. The studies are based on earlier observations demonstrating that Asp-55 and Ser-57 are crucial for coupled transport (16, 18). The kinetic analysis of the effect of the individual substitution of amino acids Ser-54 through Gly-58 reveals three groups of mutations with the following properties (Fig. 4). Group I comprises PutP-D55E and -M56C, both of which have an altered ion dependence. For these constructs Li ϩ -dependent transport is much more affected than Na ϩ -coupled transport. The apparent affinity of the transporter for proline is only slightly changed. Group II encompasses PutP molecules with replacements at position Ser-57 or Gly-58 exhibiting a highly reduced affinity for proline and for the coupling cation. Group III (PutP with substitution of Ser-54) shows transport kinetics similar to PutP-wild-type. In more detail, PutP containing neutral replacements for Asp-55 is not only unable to catalyze Na ϩ -driven proline uptake as previously reported (16) but also cannot utilize the energy stored in the electrochemical Li ϩ gradient to drive substrate transport. However, PutP with Glu in place of Asp-55 (as well as PutP-M56C) catalyzes Na ϩ -and Li ϩ -coupled proline uptake with reduced maximum rates and reduced apparent ion TABLE I Proline uptake kinetics of PutP bearing replacements of given amino acid residues To determine K m pro , initial rates of L-[U-14 C]proline uptake by E. coli WG170 producing either PutP-wild-type or PutP with given replacements were measured in the presence of 50 mM NaCl and 20 mM D-lactate (Na ϩ salt) at proline concentrations from 0.2 to 1000 M. For determination of the ion specific parameters (K 0.5 Na ϩ , K 0.5 Li ϩ , V max Na ϩ , V max Li ϩ ) transport of 10 M (or 100 M if indicated) L-[U-14 C]proline was measured in the presence of 0.005 to 250 mM NaCl or LiCl at 25°C. The resulting data were plotted according to Eadie-Hofstee. Standard deviations were determined from at least three independent experiments. affinities. A striking feature of the latter mutants is the finding that the apparent Li ϩ affinity of the transporter is more than 100-fold lower than the apparent affinity for Na ϩ . In PutPwild-type both parameters differ only by a factor of 4.8. This change of ion discrimination may be due to the fact that the larger Na ϩ ion can undergo more interactions with the protein than the smaller Li ϩ ion, and thus alteration of the binding site is therefore less dramatic for Na ϩ than for Li ϩ . In addition to previous findings, this study identifies also Met-56 as a residue which is crucial for high-affinity ion binding. Since the side chain of Met is nonpolar and cannot interact with monovalent cations, e.g. via electrostatic interactions or hydrogen bond formation, a direct participation of Met-56 in ion binding is highly unlikely. Nonetheless, the observed strong effects on the apparent affinities for Na ϩ and Li ϩ may reflect an altered geometry of the ion-binding site caused by the relatively modest alteration of the side chain (replacement of Met-56 by Cys). The latter idea is in perfect agreement with the proposed role of Asp-55 in ion binding (16).
However, if Asp-55 and Met-56 are indeed located at or close to the ion-binding site one might expect a competition of ion and sulfhydryl reagents for Cys placed at either one of these positions. But this is clearly not the case even though both positions are accessible to sulfhydryl reagents. These results can be interpreted in either of two ways: ion binding takes place at a site different from this part of TM II or ion binding is prevented by the alteration of the cation site caused by the mutation itself. The kinetic data of the mutants strongly support the latter idea. In conclusion, we suggest a location of Asp-55 and Met-56 at (or in the vicinity of) the ion-binding site(s) in which Asp-55 directly interacts with the coupling cation(s).
To date there is no direct evidence for an involvement of amino acid residues in ion binding in other members of the SSF (28). Recently, evidence has been presented suggesting a participation of Ala-166 of the human Na ϩ /glucose transporter (SGLT1) in coupling Na ϩ to sugar transport (29). In the well studied melibiose permease (MelB) of E. coli, which is not homologous to PutP, amino acids also located in the N-terminal part of the protein (Asp-55, Asp-58, and Asp-124) are implicated in Na ϩ binding (30 -32). However, PutP clearly differs from the latter sugar transporter by the fact that in PutP the carboxylate at only one position (Asp-55) is essential for substrate transport while other acidic residues in the N-terminal domain prove to be dispensable for function (16).
In addition, members of group II mutants are characterized by substantially reduced apparent affinities for Na ϩ and Li ϩ suggesting that the amino acids at the corresponding  Bradford (27), and equal amounts were subjected to 10% SDS-PAGE. Fluorescent bands of the PutP derivatives were visualized with a MultiImager TM (lane A). Afterward, the same gel was stained with Coomassie Blue (lane B). For a quantitative analysis (lane C) the relative fluorescence was analyzed by using the software Quantity One TM or MultiAnalyst TM and normalized to the amount of protein based on the density of the Coomassie-stained protein bands. The fluorescence value of each PutP derivative obtained upon labeling in the absence of ligand was set to 100. Standard deviations were determined from at least three independent experiments. positions (Ser-57 and Gly-58) participate in the formation of an ion-binding site. However, there are several arguments contradicting this view and suggesting that Ser-57 and Gly-58 are rather important for proline than for ion binding. (i) The decrease of the apparent ion affinities is accompanied by highly reduced proline affinities and the apparent ion affinities are strongly dependent on the proline concentration. So, in the cases of PutP-S57A and -S57C a 10-fold increase in the proline concentration causes an about 8-fold decrease of K 0.5 Na ϩ of both transporters (if the concentration of proline is below K m Pro ). In contrast, the apparent proline affinity of PutP containing substitutions for Ser-57 remains decreased by more than 2 orders of magnitude at saturating Na ϩ concentrations (18). These findings suggest that substitution of Ser-57 primarily affects proline binding while the influence on ion binding is a secondary event due to cooperative interactions between the ligand-binding sites. (ii) Replacement of Gly-58 by Cys induces a strong decrease of the apparent affinity of the transporter for proline (40-fold increase of K m Pro ) while the apparent affinity for Na ϩ is only about 5-fold reduced. The more pronounced effect on the apparent Li ϩ affinity (19-fold decrease) may be explained by the close proximity of Gly-58 in a helix wheel arrangement to amino acid residues that are implicated in cation binding (Asp-55 and Met-56, see Fig. 4). (iii) Na ϩ (or Li ϩ ) does not inhibit the alkylation of a single Cys placed at position 57 or 58 by sulfhydryl reagents. Instead, Na ϩ stimulates the reaction of Cys at both positions with fluorescein maleimide. This stimulatory effect can only be explained by a conformational alteration induced by ion binding most likely at a site differing from Ser-57 and Gly-58. Strikingly, in the presence of Na ϩ , addition of proline to PutP with a single Cys at position 57 or 58 reverses the stimulatory ion effect completely. This protective effect of proline can either be explained by direct steric hindering or by a proline-induced structural change of this region in the protein that reduces the accessibility of the sulfhydryl group at both positions. Clearly, the former possibility strengthens the idea of Ser-57 being part of the prolinebinding site(s) in PutP and the kinetic data support this notion.
Substitution of Ser-54 (group III mutants) has little or no effect on transport kinetics. These results suggest that this residue does not play a primary role in ligand binding by PutP. Nonetheless, the reaction of Cys in place of Ser-54 with fluorescein maleimide is inhibited in the presence of both, Na ϩ (or Li ϩ ) and proline, while the cation alone has no influence on the labeling reaction. This result points out the possibility that the accessibility of Ser-54, which is located on the same ␣-helical phase of TM II as Ser-57, from the aqueous environment has changed due to conformational changes caused by high-affinity binding of proline to Ser-57 (Fig. 4). In fact, a proline-induced conformational change has previously been demonstrated by side-directed spin labeling of Cys individually placed at position 37 in the cytoplasmic loop preceding TM II or at position 45 on the same putative helix (TM II) as Ser-54 (9).
Taken together, the results emphasize the crucial role of N-terminal regions (in particular TM II) in high-affinity ligand binding and coupling of ion and solute transport for PutP, a member of the SSF. Results obtained with SGLT1 indicate that the C-terminal domain forms the major part of the substrate (sugar) pathway (10 -12, 33). Since a truncated version of SGLT1 consisting of the last five C-terminal transmembrane domains catalyzes Na ϩ -independent facilitated diffusion of sugar with a highly reduced sugar affinity, these results directly support the findings of the present report. In conclusion, the consensus idea might be that multiple regions in individual members of the SSF contribute to efficient substrate binding. For PutP the idea is also supported by the fact that deletion mapping of putP mutants with an altered substrate specificity cluster in three distinct regions of the putP gene (34). Furthermore, the proposed close proximity of ion and organic solute binding sites in the tertiary structure of this class of membrane proteins, thereby ensuring strong cooperativity, might be the key for the molecular mechanism of cation/substrate co-transport. FIG. 4. Sites important for Na ؉ and substrate binding in transmembrane domain II of PutP. A helical wheel representation of TM II (residues Ala-48 to Pro-65) is shown to summarize the main results of the current study. Substitution of residues shown in large open letters resulted in complete inhibition of Na ϩ -and Li ϩ -coupled proline uptake (neutral replacements for Asp-55) or dramatically reduced the apparent affinity for Na ϩ and Li ϩ in a cation-dependent manner (Glu in place of Asp-55 or Cys in place of Met-56). Residues highlighted in large bold letters proved particularly crucial for high-affinity proline binding. The effects of ligand on the accessibility of Cys individually placed at the corresponding amino acid position are represented as follows: Na ϩ 1, increase of Cys accessibility by Na ϩ ; Pro2, decrease of Cys accessibility by proline in the presence of Na ϩ (or Li ϩ in the case of Cys at position 54); Na ϩ /Pro3, no effect of ligand on Cys accessibility.