A Conformationally Mobile Cysteine Residue (Cys-561) Modulates Na+ and H+ Activation of Human CNT3*

In humans, the SLC28 concentrative nucleoside transporter (CNT) protein family is represented by three Na+-coupled members; human CNT1 (hCNT1) and hCNT2 are pyrimidine and purine nucleoside-selective, respectively, whereas hCNT3 transports both purine and pyrimidine nucleosides and nucleoside drugs. Belonging to a phylogenetic CNT subfamily distinct from hCNT1/2, hCNT3 also mediates H+/nucleoside cotransport. Using heterologous expression in Xenopus oocytes, we have characterized a cysteineless version of hCNT3 (hCNT3C-). Processed normally to the cell surface, hCNT3C-exhibited hCNT3-like transport properties, but displayed a decrease in apparent affinity specific for Na+ and not H+. Site-directed mutagenesis experiments in wild-type and hCNT3C-backgrounds identified intramembranous Cys-561 as the residue responsible for this altered Na+-binding phenotype. Alanine at this position restored Na+ binding affinity, whereas substitution with larger neutral amino acids (threonine, valine, and isoleucine) abolished hCNT3 H+-dependent nucleoside transport activity. Independent of these findings, we have established that Cys-561 is located in a mobile region of the hCNT3 translocation pore adjacent to the nucleoside binding pocket and that access of p-chloromercuribenzene sulfonate to this residue reports a specific H+-induced conformational state of the protein ( Slugoski, M. D., Ng, A. M. L., Yao, S. Y. M., Smith, K. M., Lin, C. C., Zhang, J., Karpinski, E., Cass, C. E., Baldwin, S. A., and Young, J. D. (2008) J. Biol. Chem. 283, 8496-8507 ). The present investigation validates hCNT3C- as a template for substituted cysteine accessibility method studies of CNTs and reveals a pivotal functional role for Cys-561 in Na+- as well as H+-coupled modes of hCNT3 nucleoside transport.

Nucleosides and nucleoside analogs with antineoplastic and/or antiviral activity require specialized nucleoside transporter (NT) 4 proteins for uptake into or release from cells (1)(2)(3). NT-mediated transport is therefore a critical determinant of nucleoside and nucleotide metabolism and the pharmacologic actions of nucleoside drugs (3)(4)(5). NTs also regulate adenosine availability to cell surface purinoreceptors and profoundly affect neurotransmission, vascular tone, and other physiological processes (5,6). NTs in human and other mammalian cells belong to two structurally unrelated families of integral membrane proteins: the SLC28 concentrative nucleoside transporter (CNT) family and the SLC29 equilibrative nucleoside transporter (ENT) family (3, 6 -8). ENTs are ubiquitously expressed in most, possibly all, cell types, whereas CNTs are found predominantly in intestinal and renal epithelial and other specialized cells (1-4, 6 -11).
In humans, the SLC28 protein family is represented by three members. hCNT1 and hCNT2 are pyrimidine nucleoside-selective and purine nucleoside-selective, respectively, whereas hCNT3 transports both pyrimidine and purine nucleosides (12)(13)(14)(15). hCNT1 and hCNT2 are predominantly Na ϩ -coupled cotransporters and share a common Na ϩ /nucleoside stoichiometry of 1:1 (16 -18). hCNT3, in contrast, can utilize electrochemical gradients of Na ϩ and/or H ϩ to accumulate nucleosides within cells with Na ϩ /nucleoside and H ϩ /nucleoside coupling stochiometries of 2:1 and 1:1, respectively (16 -19). hCNT3 has a broader tissue and cellular distribution than hCNT1/2 (14) and is especially prominent in kidney (10,11). Nonmammalian members of the CNT protein family that have been characterized include hfCNT from the ancient marine prevertebrate the Pacific hagfish Eptatretus stouti (19), CaCNT from the pathogenic yeast Candida albicans (20,21), CeCNT3 from the nematode Caenorhabditis elegans (22), and NupC from the bacterium Escherichia coli (23,24). hfCNT is predominantly Na ϩ -coupled and, similar to hCNT3, exhibits a Na ϩ /nucleoside stoichiometry of 2:1 (19). CaCNT, CeCNT3, and NupC function exclusively as H ϩ /nucleoside * This work was supported in part by the National Cancer Institute of Canada, with funds from the Canadian Cancer Society and the Alberta Cancer Board. 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 U.S.C. Section 1734 solely to indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. S1-S3. 1 Supported by a Studentship from the Alberta Heritage Foundation for Medical Research. 2 Canada Research Chair in Oncology. 3  cotransporters (20,22,24). The H ϩ /nucleoside coupling ratio for CaCNT is 1:1 (20). Human and other eukaryote CNTs have 13 (or possibly 15) transmembrane helices (TMs), with the C-terminal halves of the proteins showing greatest sequence similarity (14,25). Prokaryote CNTs lack the first three TMs of their eukaryote counterparts, and functional expression of N-terminally truncated hCNT1/rCNT1 in Xenopus oocytes has established that these TMs are not necessary for Na ϩ -dependent uridine transport activity (25). Consistent with this, chimeric studies involving hCNT1/2 (26), hCNT1/3 (17), and hCNT1/hfCNT (19) have established that residues involved in nucleoside selectivity and Na ϩ and H ϩ coupling reside in the C-terminal half of the proteins. Mutagenesis studies also reveal a key role for the C-terminal residues. In hCNT1, for example, two sets of adjacent residues in TMs 7 and 8 have been identified (Ser 319 /Gln 320 and Ser 353 /Leu 354 ) that, when converted to the corresponding residues in hCNT2, change the nucleoside specificity from CNT1 type to CNT2 type (26). Confirming the roles of these residues in hCNT1/2 nucleoside selectivity, the double TM 8 mutant (S353T/L354V) exhibits a unique uridine-preferring transport phenotype (27). Mutation of Leu 354 alone markedly increased the affinity of the transporter for Na ϩ and Li ϩ (27). Also in hCNT1, Glu 308 and Glu 322 in TM 7 and Glu 498 in the region between TMs 11 and 12 (TM 11A) have been shown to have key roles in Na ϩ coupling and translocation (28).
To avoid confounding background reactivity with endogenous cysteine residues, SCAM analysis requires construction of a functional cysteineless version of the transporter to serve as a template for subsequent cysteine reinsertion at defined positions. In the case of hCNT3, the protein was engineered by mutation of all 14 endogenous cysteine residues to serine, resulting in the cysteineless construct hCNT3CϪ (37). Expressed in yeast, hCNT3CϪ was used to assess residues in TMs 11-13 for accessibility to methanethiosulfonate reagents (37). More recently, hCNT3CϪ, expressed in Xenopus oocytes, was used to identify a H ϩ -mediated conformational shift that allows access of p-chloromercuribenzene sulfonate (PCMBS) to specific residues in TM 12 under acidified conditions (38). One of these conformationally mobile amino acids, native cysteine residue 561, was fully protected against PCMBS inhibition by micromolar concentrations of extracellular uridine, suggesting likely proximity to the nucleoside binding pocket. In the present study, we have utilized heterologous expression in Xenopus oocytes, cell surface biotinylation, and site-directed mutagenesis in combination with radioisotope flux and presteady-and steady-state electrophysiological kinetic experiments to undertake an in depth functional comparison between wild-type and cysteineless hCNT3. The present investigation, which validates hCNT3CϪ as a template for SCAM analyses of CNTs, reveals that Cys-561 modulates Na ϩ -as well as H ϩ -coupled modes of hCNT3 nucleoside transport.

EXPERIMENTAL PROCEDURES
Site-directed Mutagenesis and DNA Sequencing-hCNT3 DNA (GenBank TM accession number AF305210) provided the template for the construction of a cysteineless version of hCNT3 (hCNT3CϪ) with all 14 endogenous cysteine residues converted to serine (37). hCNT3CϪ was transferred from the yeast E. coli shuttle vector pYPGE15 (37) into the Xenopus oocyte expression vector pGEM-HE. The latter provides additional 5Ј-and 3Ј-untranslated regions from the Xenopus ␤-globin gene, resulting in greater protein expression and functional activity when produced in Xenopus laevis oocytes (39). Wildtype hCNT3C and hCNT3CϪ, both in pGEM-HE, provided the templates for construction of corresponding mutants by the oligonucleotide-directed technique (40), using reagents from the QuikChange site-directed mutagenesis kit (Stratagene) according to the manufacturer's directions. All constructs were sequenced in both directions by Taq DyeDeoxy terminator cycle sequencing to ensure that only the correct mutation(s) had been introduced.
Production of Recombinant Transport Proteins in Xenopus Oocytes-hCNT3 and hCNT3CϪ DNAs were transcribed with T7 polymerase using the mMESSAGE mMACHINE TM (Ambion) transcription system and produced in oocytes of X. laevis by standard procedures (41). Healthy defolliculated stage VI oocytes were microinjected with 20 nl of water or 20 nl of water containing RNA transcript (20 ng) and incubated in modified Barth's medium (changed daily) at 18°C for 72 h prior to the assay of transport activity.
Radioisotope Flux Assays-Transport assays were performed as described previously (27,41,42). Groups of 12 oocytes were incubated at room temperature (20°C) in 200 l of transport medium containing 100 mM NaCl, 2 mM KCl, 1 mM CaCl 2 , 1 mM MgCl 2 , and 10 mM HEPES (pH Ն 7.0) or 10 mM MES (pH Յ 6.5). In experiments examining the H ϩ dependence of transport, 100 mM NaCl was replaced with 100 mM ChCl. ChCl also substituted for NaCl in Na ϩ activation experiments. Except where otherwise indicated, the nucleoside concentration was 10 M. Uptake was traced with 14 C-or 3 H-radiolabeled nucleosides (1 or 2 Ci/ml, respectively) (GE Healthcare) using a 1-min uptake interval to measure initial rates of transport (influx). In experiments with adenosine, 1 M deoxycoformycin was added to inhibit breakdown by adenosine deaminase. At the end of the incubation period, extracellular label was removed by six rapid washes in ice-cold Na ϩ -free, ChCl transport medium (pH 7.5), and individual oocytes were dissolved in 1% (w/v) SDS for quantitation of cell-associated radioactivity by liquid scintillation counting (LS 6000 IC; Beckman). The flux values shown are means Ϯ S.E. of 10 -12 oocytes, and each experiment was performed at least twice on different batches of cells. Values for mediated transport were calculated as uptake in RNA-injected oocytes minus uptake in control oocytes injected with water alone. Kinetic parameters (K m , K 50 , V max , Hill coefficient) Ϯ S.E. were calculated from data for mediated transport using SigmaPlot software (Jandel Scientific Software).
Electrophysiology-Steady-state and presteady-state hCNT3 and hCNT3CϪ membrane currents were measured at room temperature (20°C) using the whole cell, two-electrode voltage clamp technique (GeneClamp 500B; Molecular Devices Corp.), as described previously (16 -18). The GeneClamp 500B was interfaced to an IBM-compatible PC via a Digidata 1322A A/D converter and controlled by pCLAMP software (version 9.0; Molecular Devices Corp.). The microelectrodes were filled with 3 M KCl and had resistances ranging from 0.5 to 1.5 megaohms. Following microelectrode penetration, resting membrane potential was measured over a 10-min period prior to the start of the experiment. Oocytes exhibiting an unstable membrane potential or a potential more positive than Ϫ30 mV were discarded. Individual oocytes with good resting membrane potentials were clamped at Ϫ50 mV. Steady-state currents were measured using a uridine concentration of 100 M. In Na ϩ activation experiments, uridine-evoked currents from six or more oocytes from the same batch of oocytes used on the same day were individually normalized and averaged to produce mean values Ϯ S.E.
Presteady-state (transient) currents were studied using a voltage pulse protocol, as described previously (16). Membrane voltage was stepped from the holding potential (V h ) of Ϫ50 mV to a range of test potentials (V t ) from Ϫ150 to ϩ75 mV in 25-mV increments. The voltage rise time of the clamp was adjusted by use of an oscilloscope such that it varied between 200 and 500 s. Current signals were measured in medium of the same composition used in radioisotope flux assays and filtered at 2 kHz (four-pole Bessel filter) at a sampling interval of 200 s/point (corresponding to a sampling frequency of 5 kHz). For data presentation, representative traces were averaged from two pulses for which the current at each test potential was averaged from five sweeps and further filtered at 0.75 kHz by pCLAMP 9.0 software (Molecular Devices Corp.).
Charge/Nucleoside Stoichiometry-Na ϩ /nucleoside and H ϩ /nucleoside coupling ratios for hCNT3CϪ were determined by simultaneously measuring Na ϩ or H ϩ currents and 100 M [ 3 H]uridine (2 Ci/ml) uptake under voltage clamp conditions, as described above. Individual hCNT3CϪ-producing oocytes were placed in a perfusion chamber and voltage-clamped at Ϫ90 mV in the appropriate nucleoside-free medium for a 10-min period to monitor base-line currents. When the base line was stable, the nucleoside-free medium was exchanged with medium of the same composition containing radiolabeled uridine. Current was measured for 30 s, and uridine uptake was terminated by washing the oocyte with nucleoside-free medium until the current returned to base line. The oocyte was then transferred to a scintillation vial and solubilized with 1% (w/v) SDS for quantitation of oocyte-associated radioactivity. Nucleoside-induced current was obtained as the difference between base-line current and the inward uridine current. The total charge translocated into the oocyte during the uptake period was calculated from the current-time integral and correlated with the measured radiolabeled flux for each oocyte to determine the charge/uptake ratio. Basal [ 3 H]uridine uptake was determined in control water-injected oocytes (from the same donor frog) under equivalent conditions and used to correct for endogenous nonmediated uridine uptake over the same incubation period. Coupling ratios Ϯ S.E. were calculated from slopes of least squares fits of uridine-dependent charge versus uridine accumulation for eight or more oocytes.
Cell Surface Expression and Glycosylation-Production of recombinant hCNT3 and hCNT3CϪ proteins at the oocyte cell surface was determined by labeling of intact oocytes with EZ-Link sulfo-NHS-LC-biotin (Pierce), followed by isolation of the resultant biotinylated plasma membrane proteins using immobilized streptavidin resin (Pierce) according to the manufacturer's instructions. Glycosylation status was established by digestion with N-glycosidase F (Roche Applied Science). Identically treated samples omitting enzyme were used as controls.

RESULTS
As depicted in Fig. S1, hCNT3 contains 14 endogenous cysteine residues. Nine of these are located within either the extramembranous N-terminal (Cys-82, Cys-91, and Cys-94) or C-terminal (Cys-621, Cys-643, Cys-644, Cys-673, Cys-674, and Cys-684) regions of the transporter. Five are positioned in predicted TMs (Cys-121 in TM 1, Cys-486 in TM 11, Cys-561 in TM 12, and Cys-602 and Cys-607 in TM 13). To enable transmembrane protein topology mapping and structure/function analysis of hCNT3 by SCAM, all 14 cysteine residues of hCNT3 were mutated to serine. The resulting cysteineless version of hCNT3 (hCNT3CϪ) was capable of Na ϩ -dependent nucleoside transport when produced in yeast and has been used as template in an initial series of SCAM analyses of hCNT3 TMs 11-13 (37,38). The current work extended these studies by using the dual radioisotope flux and electrophysiological capabilities of the Xenopus oocyte heterologous expression system to undertake a detailed functional characterization of hCNT3CϪ.
hCNT3CϪ Nucleoside Selectivity and Cation Dependence-The experiment of Fig. 1A investigated the ability of hCNT3CϪ to transport a panel of radiolabeled physiological pyrimidine and purine nucleosides. Since wild-type hCNT3 is both Na ϩand H ϩ -coupled (14,17,18) and since Na ϩ -and H ϩ -coupled hCNT3 exhibit different nucleoside selectivity profiles (17), the experiment was performed both in the presence of Na ϩ and H ϩ (100 mM NaCl, pH 7.5, and 100 mM ChCl, pH 5.5, respectively). As described under "Experimental Procedures," the flux values shown in this and subsequent experiments depict mediated transport activity, defined as the difference in uptake between RNA transcript-injected and control water-injected oocytes using a 1-min uptake interval to determine initial rates of transport (Fig. S2). Nucleoside uptake in water-injected oocytes was Ͻ0.1 pmol/oocyte⅐min Ϫ1 under all conditions tested (data not shown).

hCNT3 Residue Cys-561
Similar to wild-type hCNT3 (14,23), hCNT3CϪ in Na ϩcontaining medium transported uridine, adenosine, thymidine, cytidine, guanosine, and inosine at equivalent rates (4.8 -7.0 pmol/oocyte⅐min Ϫ1 ). Also similar to the wild-type protein (17), this broad selectivity profile was markedly altered in H ϩ -containing medium (uridine Ͼ Ͼ adenosine, thymidine Ͼ cytidine Ͼ guanosine, inosine) with only uridine showing similar rates of transport in the two media. The different nucleoside selectivity profiles of Na ϩ -and H ϩ -coupled hCNT3 and hCNT3CϪ suggest that Na ϩ -and H ϩ -bound versions of the transporters have significantly different conformations of the nucleoside binding pocket and/or translocation channel (17).
Only the H ϩ -coupled form of hCNT3 is susceptible to inhibition by the hydrophilic thiol-reactive agent PCMBS (38). Fig. 1B compares the cation dependence of hCNT3 and hCNT3CϪ influx of 10 M 14 C-labeled uridine in the presence of Na ϩ and/or H ϩ (100 mM NaCl or ChCl, pH 7.5 or 5.5, as indicated). Like hCNT3, hCNT3CϪ displayed high levels of uridine uptake in the presence of Na ϩ (NaCl, pH 7.5), H ϩ (ChCl, pH 5.5), and both Na ϩ and H ϩ (NaCl, pH 5.5). Relative to hCNT3, hCNT3CϪ exhibited an equivalent rate of uridine transport in the presence of H ϩ (ChCl, pH 5.5) but did not show the wild-type increase in uridine influx in the presence of Na ϩ (NaCl, pH 7.5) or Na ϩ plus H ϩ (NaCl, pH 5.5). As a result, the ratio of uridine influx in the presence of Na ϩ (NaCl, pH 7.5) to that in the presence of H ϩ (ChCl, pH 5.5) decreased from 1.60 for wild-type protein to 0.98 for hCNT3CϪ. These uridine uptake ratios are in good agreement with that reported previously for hCNT3 (17) and with the other results for hCNT3CϪ shown in Fig. 1A. The residual uridine fluxes evident in the absence of Na ϩ and at a low H ϩ concentration (ChCl, pH 7.5) can be attributed to the small but significant amount of H ϩ activation that occurs under these conditions (17).
Cell Surface Expression and Glycosylation of hCNT3C-The equivalent levels of transport activity shown in Fig. 1B suggest that hCNT3 and hCNT3CϪ were present in oocyte plasma membranes in similar quantities. This was investigated directly by cell surface labeling with sulfo-NHS-LC-biotin using immobilized streptavidin resin to separate cell surface protein ( Fig. 2A) from that associated with total (plasma ϩ intracellular) membranes (Fig. 2B). Immunoblots of the fractions were probed with hCNT3 polyclonal antibodies (10) directed against amino acid residues 45-69 of the extramembranous N-terminal region of the protein (Fig. S1). To evaluate the glycosylation status of hCNT3 and hCNT3CϪ, the immunoblots in Fig. 2, A and B, also include aglyco-hCNT3, a construct produced by mutating all four potential N-terminal glycosylation acceptor sites, asparagine residues 630, 636, 664, and 678, to aspartate (Fig. S1). For comparison with the electrophoretic mobility of aglyco-hCNT3, Fig. 2B also shows the same streptavidin fractions of total hCNT3 and hCNT3CϪ  Labeling of oocytes with sulfo-NHS-LC-biotin was used in conjunction with immobilized streptavidin resin to separate recombinant hCNT3, hCNT3CϪ, and aglyco-hCNT3 cell surface immunoreactivity (A) from that associated with total (plasma ϩ intracellular) membranes (B). Immunoblots of the fractions were probed with anti-hCNT3 antibodies. In B, plus and minus signs refer to digestion with N-glycosidase F. The positions of molecular weight standards are shown on the left. Water, control water-injected oocytes. Blots shown in A and B are from different gels. membranes before and after digestion with N-glycosidase F. Antibody specificity was evaluated in membrane fractions prepared from control water-injected oocytes.
Consistent with the presence of multiple possible sites of N-linked glycosylation, cell surface hCNT3 and hCNT3CϪ exhibited three discreet immunobands at 100, 86, and 75 kDa ( Fig. 2A). Total cell surface immunoreactivity was similar for the two transporters, the majority of hCNT3 and hCNT3CϪ staining being associated with the highest and intermediate molecular weight forms, respectively ( Fig. 2A). Corresponding patterns of immunoreactivity were apparent in total membranes (Fig. 2B). Indicative of proper protein folding; therefore, hCNT3CϪ was processed to the oocyte plasma membrane in amounts similar to wild-type hCNT3. Digestion with N-glycosidase F shifted hCNT3 and hCNT3CϪ immunoreactivity to the lower molecular mass band at 75 kDa, an electrophoretic mobility identical to that of aglyco-hCNT3 and in good agreement with the molecular mass of hCNT3 calculated from its amino acid sequence (77 kDa). Confirming antibody specificity for hCNT3/hCNT3CϪ, no immunoreactivity was detected in blots from control water-injected oocytes.
Kinetic Characterization of hCNT3C-To further investigate hCNT3CϪ function, the concentration dependence of radiolabeled uridine (0 -1 mM) influx by hCNT3 and hCNT3CϪ was measured either in the presence of Na ϩ (100 mM NaCl, pH 7.5) or in the presence of H ϩ (100 mM ChCl, pH 5.5) (Fig. S3). The corresponding kinetic parameters are given in Table 1. The results demonstrate robust high affinity transport of uridine by hCNT3CϪ. Similar to hCNT3 and in agreement with previous kinetic data for the wild-type transporter (17), the apparent K m value of hCNT3CϪ for uridine in Na ϩ -containing transport medium was lower than that in the presence of H ϩ (15 and 62 M, respectively, for hCNT3 compared with 18 and 63 M, respectively, for hCNT3CϪ). Calculated uridine V max /K m ratios, an indicator of transporter efficiency and a kinetic predictor of relative transport rates at permeant concentrations less than K m values, were consistent with the 10 M influx data shown in Fig. 1B (hCNT3(Na ϩ ) Ͼ hCNT3(H ϩ ), hCNT3CϪ(Na ϩ ), and hCNT3CϪ(H ϩ )).
Cation activation curves were also determined for both transporters using a radiolabeled uridine concentration of 10 M. Na ϩ activation was measured over the concentration range 0 -100 mM NaCl in transport medium at pH 7.5 and 8.5 (Fig. 3,  A and B, respectively), the latter to eliminate the small amount of H ϩ activation that occurs at neutral pH (17). Isosmolarity was maintained with ChCl. H ϩ activation was measured in 100 mM ChCl transport medium at pH values ranging from 4.5 to 8.5 (Fig. 3C). Kinetic parameters derived from these data are presented in Tables 2 and 3 (Na ϩ and H ϩ , respectively). Reflecting previously reported differences in Na ϩ and H ϩ cou- Radiolabeled 10 M uridine influx was measured in 100 mM NaCl transport medium, pH 7.5, and 100 mM ChCl transport medium, pH 5.5, under initial rate conditions (1-min) in oocytes producing hCNT3 and hCNT3CϪ. Kinetic parameters were calculated from data shown in Fig. S3 corrected for basal (nonmediated) uptake in control water-injected oocytes.   Tables 2 and 3.

hCNT3 Residue Cys-561
pling by hCNT3 (17,18), the wild-type protein exhibited saturable Na ϩ and H ϩ activation curves that were sigmoidal and hyperbolic in nature, respectively, with apparent K 50 values of 12 (pH 7.5) and 11 mM (pH 8.5) for Na ϩ and 690 nM for H ϩ . Calculated Hill coefficients were consistent with 2:1 (Na ϩ ) and 1:1 (H ϩ ) cation/nucleoside coupling ratios. Validating the subsequent use of pH 8.5 to study Na ϩ coupling of hCNT3/ hCNT3CϪ without interference from H ϩ , there was no difference in hCNT3 Na ϩ activation kinetics at pH 7.5 and 8.5.
In marked contrast to wild-type hCNT3 and consistent with the relative decrease in Na ϩversus H ϩ -coupled uridine transport activity noted for the cysteineless transporter in Fig. 1B and Table 1, the corresponding cation activation curves for hCNT3CϪ revealed a marked and specific decrease in apparent affinity for Na ϩ (apparent K 50 Ͼ 40 mM). This shift in Na ϩ affinity did not extend to H ϩ , since kinetic parameters for hCNT3CϪ H ϩ activation were comparable with those of hCNT3 (apparent K 50 value of 550 nM). Similarly, the corresponding Hill coefficient was consistent with a H ϩ /nucleoside coupling ratio of 1:1. That for Na ϩ could not be determined, although the curve appeared sigmoidal (Fig. 3, A and B) and therefore potentially consistent with the wild-type Na ϩ /nucleoside coupling ratio of 2:1. Similar to the wild-type protein, there was no difference in hCNT3CϪ Na ϩ activation kinetics at pH 7.5 and 8.5. All subsequent hCNT3/ hCNT3CϪ Na ϩ activation and Na ϩ coupling experiments were performed at pH 8.5.
Electrophysiological Determination of hCNT3CϪ Na ϩ Activation Kinetics-Since hCNT3 is electrogenic, the Na ϩ activation kinetic parameters of hCNT3CϪ were also determined by electrophysiology. Fig. 4 depicts the relationship between Na ϩ concentration and 100 M uridine-evoked current for oocytes expressing wild-type hCNT3 or hCNT3CϪ clamped at a membrane holding potential of Ϫ90 mV. Oocytes were individually normalized to their predicted I max values and subsequently averaged to produce mean kinetic parameters. In agreement with previous studies (17,18), the apparent K 50 value for Na ϩcoupled uridine uptake by hCNT3 under these conditions was 2.2 Ϯ 0.1 mM. Similar to the radioisotope flux data in Fig. 3, A and B, hCNT3CϪ exhibited a marked decrease (ϳ11-fold) in the apparent affinity for Na ϩ in comparison with that of the wild-type protein, with an increased apparent K 50 value of 24.7 Ϯ 0.8 mM. The corresponding Hill coefficients for hCNT3CϪ and hCNT3 were Ն1.5, suggesting that the wildtype Na ϩ /uridine stoichiometry of 2:1 was maintained by hCNT3CϪ. The apparent K 50 values for hCNT3 and hCNT3CϪ as determined by electrophysiology were lower than those derived from radioisotope flux data because of (i) the different uridine concentrations used in the two studies (100 and 10 M, respectively) and (ii) the difference in membrane potential in the two situations (Ϫ90 mV for the voltageclamped oocytes versus ϳϪ40 mV under radioisotope flux conditions) (17).
Presteady-state Electrophysiology of hCNT3C-Presteadystate electrophysiological experiments were performed on oocytes producing hCNT3CϪ voltage-clamped at a holding potential (V h ) of Ϫ50 mV. Presteady-state currents were activated by voltage steps to a series of test potentials (V t ) in the presence of varying concentrations of Na ϩ (0 -100 mM NaCl, pH 8.5; Fig. 5, A-E) and in the presence of both Na ϩ and extra-  From Fig. 3, A and B. b ND, could not be determined; in 100 mM transport medium with 0 -100 mM NaCl (pH 7.5 or 8.5, as indicated, isosmolarity maintained by ChCl). c From Fig. 7 Fig. 3C. b Data not shown. c From Fig. 8B; in 100 mM ChCl transport medium with pH values ranging from 4.5 to 8.5.   5F). hCNT3CϪ exhibited presteady-state currents similar to those previously described for wild-type hCNT3 (17) and wildtype hCNT1 (16), which increased in magnitude upon exposure to extracellular Na ϩ and were largely eliminated upon the addition of uridine. However, consistent with a reduced binding affinity for Na ϩ and different from the wild-type protein, presteadystate currents approached maximum values only at Na ϩ concentrations of Ͼ50 mM, a behavior also observed for hfCNT, another CNT with low apparent affinity for Na ϩ . 5 Presteadystate currents were absent from control water-injected oocytes (data not shown) (16,28). Na ϩ /Uridine and H ϩ /Uridine Stoichiometry-The Na ϩ /uridine and H ϩ /uridine stoichiometries of hCNT3CϪ were determined directly by simultaneous measurement of uridineevoked currents and 100 M radiolabeled uridine uptake under voltage clamp conditions, as described previously (16,17). A representative current trace is depicted in Fig. 6A for an oocyte expressing hCNT3CϪ exposed to 100 M radiolabeled uridine in 100 mM NaCl transport medium, pH 8.5. Presented in Fig. 6, B and C, each data point represents a single oocyte, and the coupling ratio is given by the slope of the linear fit of charge (pmol) versus uptake (pmol). In 100 mM NaCl transport medium (pH 8.5) and at a holding potential of Ϫ90 mV, the linear correlation between uridine-dependent charge and uridine accumulation gave an Na ϩ /uridine coupling ratio of 2.08 Ϯ 0.11 (Fig. 6B). The corresponding H ϩ /uridine stoichiometry was determined in acidified 100 mM ChCl transport medium (pH 5.5) and resulted in a coupling ratio of 1.11 Ϯ 0.06 (Fig. 6C). Both hCNT3CϪ stoichiometries agree with previously published values for hCNT3 (17,18) and with the Hill coefficients derived from hCNT3 and hCNT3CϪ from Figs. 3 and 4 (Tables 2 and 3).
Cation Activation of Aglyco-hCNT3-The experiment of Fig.  2 demonstrated that wild-type hCNT3 and hCNT3CϪ have different patterns of N-linked glycosylation. To exclude the possibility that the shift in hCNT3CϪ Na ϩ affinity was secondary to the altered glycosylation status of the protein, the apparent K 50 and V max values for Na ϩ and H ϩ activation of aglyco-hCNT3 were determined under conditions identical to those used in Fig. 3B for hCNT3CϪ. Kinetic parameters for both cations were not significantly different from those of the wildtype protein measured in parallel in the same experiment (aglyco-hCNT3 and hCNT3 apparent K 50 and V max values of 10.5 Ϯ 1.1 and 8.8 Ϯ 1.2 mM and 6.8 Ϯ 0.4 and 7.4 Ϯ 0.5 pmol/ oocyte⅐min Ϫ1 , respectively, for Na ϩ and 399 Ϯ 98 and 506 Ϯ 182 nM and 4.7 Ϯ 0.2 and 5.3 Ϯ 0.4 pmol/oocyte⅐min Ϫ1 , respectively, for H ϩ ) (cation activation curves not shown).
Cation Activation of hCNT3CϪ Mutants-The hCNT3CϪ construct differs from wild-type hCNT3 by the replacement of all 14 endogenous cysteine residues with serine. Five of these substitutions lie within predicted TM regions of the protein (supplemental Fig. S1). Of these, four reside in the C-terminal half of the protein, which corresponds to the region that chimeric studies between hCNT1 and hCNT3 (17) and between hCNT1 and hfCNT (19) have demonstrated to be responsible for Na ϩ coupling. Site-directed mutagenesis experiments to identify the residue(s) responsible for the shift in hCNT3CϪ Na ϩ affinity therefore focused on the four C-terminal intramembranous cysteine residues: Cys-486 (TM 11), Cys-561 (TM 12), Cys-602 (TM 13), and Cys-607 (TM 13). In the hCNT3CϪ background, cysteine residues were individually reintroduced at each of these positions, yielding constructs S486C(CϪ), S561C(CϪ), S602C(CϪ), and S607C(CϪ). Measured under conditions identical to those used in Fig. 3B, Fig. 7 shows oocyte Na ϩ activation curves for each mutant. Corresponding kinetic parameters are given in Table 2. S486C(CϪ), S602C(CϪ), and S607C(CϪ) (Fig. 7, A, C, and D) each retained Na ϩ activation curves similar to that of hCNT3CϪ (Fig. 3B), with apparent K 50 values of Ͼ40 mM. In contrast, the Na ϩ activation curve for S561C(CϪ) (Fig. 7B) closely resembled that of wild-type hCNT3 (Fig. 3B) with an apparent K 50 value of 7.2 Ϯ 0.6 mM. H ϩ activation kinetics were also determined for S561C(CϪ), yielding apparent K 50 and V max values that were similar to those of both hCNT3CϪ and hCNT3 (cation activation curve not shown; Table 3).
Cation Activation of hCNT3 Mutant C561S-To confirm the role of Cys-561 in the shift in Na ϩ affinity of hCNT3CϪ, serine was substituted for cysteine in wild-type hCNT3, creating the mutant protein C561S. Na ϩ and H ϩ activation kinetics for oocytes producing C561S were determined under experimental conditions identical to those described above for other hCNT3/ hCNT3CϪ constructs and are depicted in Fig. 8. Kinetic parameters derived from the curves are presented in Tables 2  and 3. Similar to hCNT3CϪ (Fig. 3, B and C), C561S exhibited hCNT3 Residue Cys-561 a decreased apparent affinity for Na ϩ (Ͼ40 mM) in the absence of a parallel shift in H ϩ apparent binding affinity.
Cation Activation of hCNT1 and hCNT1 Mutant C540S-In contrast to hCNT3, which is both H ϩ -and Na ϩ -coupled and exhibits H ϩ /nucleoside and Na ϩ /nucleoside stoichiometries of 1:1 and 2:1, respectively, hCNT1 and hCNT2 are Na ϩ -specific and have an Na ϩ /nucleoside stoichiometry of 1:1 (12, 16 -18). All three proteins share similar high affinities for Na ϩ binding with apparent K 50 values of ϳ10 mM (12, 16 -18). The residue corresponding to hCNT3 Cys-561 in hCNT1 is Cys-540. To investigate the effects of a cysteine to serine mutation at this position in another CNT family member, the same change was made in hCNT1 to create the mutant protein C540S. Mutant C540S and wild-type hCNT1 were then produced in oocytes, and their Na ϩ activation kinetics were investigated under experimental conditions identical to those described above for hCNT3/hCNT3CϪ and mutants (Fig. 9). In agreement with previous studies (12,27), hCNT1 displayed a hyperbolic Na ϩ activation curve with an apparent affinity for Na ϩ of 6.6 Ϯ 1.0 mM and a V max of 4.5 Ϯ 0.2 pmol/oocyte⅐min Ϫ1 (Fig. 9B). Similarly, Na ϩ activation of C540S was also hyperbolic with apparent K m and V max values of 8.4 Ϯ 1.1 mM and 4.3 Ϯ 0.1 pmol/ oocyte⅐min Ϫ1 , respectively (Fig. 9A). Corresponding Hill coefficients were both 1.0 Ϯ 0.1.
Other Amino Acid Substitutions of hCNT3 Residue Cys-561-Sequence comparisons of 126 eukaryotic and prokaryotic CNT family members revealed that the residue corresponding to hCNT3 Cys-561 is highly conserved. However, in addition to cysteine (90 family members), the amino acids alanine, valine, threonine, and isoleucine were also represented at this position (four, 21, six, and five family members, respectively). Therefore, to further elucidate the role of Cys-561 in hCNT3 cation coupling, hCNT3 mutants C561A, C561V, C561T, and C561I were constructed and produced in oocytes. Substitution of hCNT3 Cys-561 with glycine, the amino acid with the smallest side chain, was also included in the series (mutant C561G). Functional activity and cation selectivity were compared with wild-type hCNT3 and mutant C561S by measuring 10 M radiolabeled uridine influx in both Na ϩ -and H ϩ -containing transport medium (100 mM NaCl, pH 7.5 and 100 mM ChCl, pH 5.5, respectively) (Fig. 10). Similar to these transporters, C561G and C561A exhibited both Na ϩ -and H ϩ -coupled influx of radiolabeled uridine. However, relative to transport in the presence of Na ϩ and in contrast to C561A, C561G exhibited reduced influx of uridine in H ϩ -containing medium. Extending this trend, substitution of Cys-561 with larger neutral amino acids (mutants C561V, C561T, and C561I) led to almost total abolition of uridine influx in H ϩ -containing medium, a finding that was confirmed for C561I by corresponding measurements of 100 M uridine-induced steady-state currents (Fig. 11).
Kinetic parameters for Na ϩ activation of C561G, C561A, C561S, C561T, C561V, and C561I, measured as described in Fig. 3B, are given in Table 2. Replacement of Cys-561 with residues of a similar size (C561A and C561V) allowed full (C561A) FIGURE 6. Na ؉ /uridine and H ؉ /uridine stoichiometry of hCNT3C؊. A, a representative current trace for an oocyte expressing hCNT3CϪ clamped at a membrane potential of Ϫ90 mV and exposed to 100 M radiolabeled uridine in 100 mM NaCl, pH 8.5. Integration of the uridine-evoked current over the uptake period yielded the charge moved; this charge, representing net cation influx, was correlated to the net uptake (Flux) of radiolabeled uridine by the oocyte during the same time interval. Following the same protocol (V h of Ϫ90 mV; 100 M uridine), B and C show charge/flux ratio plots generated for oocytes producing hCNT3CϪ in either 100 mM NaCl transport medium at pH 8.5 (B) or ChCl transport medium at pH 5.5 (C). Each point represents a single oocyte. Linear regression fits of both data sets passed through the origin and are indicated by solid lines. The dashed lines represent theoretical 2:1 (B) and 1:1 (C) charge/uptake ratios.

hCNT3 Residue Cys-561
or partial (C561V) recovery of Na ϩ apparent affinity (K 50 values of 10.7 Ϯ 0.7 and 31.3 Ϯ 1.9 mM, respectively). In contrast, C561G, C561T, and C561I exhibited reduced affinities for Na ϩ similar to that of hCNT3CϪ (apparent K 50 values of Ͼ40 mM). The ability of alanine to mimic cysteine at residue position 561 extended to the reciprocal hCNT3CϪ mutant S561A(CϪ), which exhibited a partially restored apparent K 50 value of 20.5 Ϯ 0.9 mM (Na ϩ activation curve not shown; Table 2).
Corresponding kinetic parameters for H ϩ activation of mutants C561G and C561A were also determined. Measured as described in Fig. 3C and similar to wild-type hCNT3, C561G and C561A displayed apparent K 50 values for H ϩ of 900 Ϯ 90 and 490 Ϯ 100 nM, respectively (H ϩ activation curves not shown; Table 3).
Complementary to these cation activation analyses, we additionally determined the uridine kinetic parameters of hCNT3 mutants. Values were determined as described in Table 1 and are presented in Table 4 (concentration dependence curves not shown). In Na ϩ -containing transport medium (100 mM NaCl, pH 8.5), C561A and C561S exhibited high affinity for uridine similar to wild-type hCNT3 (Fig. S3 and Table 1 (Table 1 and Fig. S3), uridine apparent K m values increased when determined in H ϩ -containing transport medium (100 mM ChCl, pH 5.5). C561A (65.0 Ϯ 6.3 M) and C561S (91.0 Ϯ 11.0 M) were again comparable with the wild-type transporter, whereas C561G was higher (234 Ϯ 32 M).
Presteady-state Electrophysiology of hCNT3 Cys-561 and hCNT3CϪ S561(CϪ) Mutants-In a final series of experiments, presteadystate electrophysiological experiments were performed on oocytes producing the hCNT3 mutants C561A and C561S and the hCNT3CϪ mutant S561C(CϪ). As undertaken for hCNT3CϪ in Fig. 5, oocytes were voltage-clamped at a holding potential (V h ) of Ϫ50 mV and stepped to a series of test potentials (V t ) in the presence of varying concentrations of Na ϩ (0 -100 mM NaCl, pH 8.5). C561A, C561S, and S561C(CϪ) exhibited presteadystate currents, which increased in magnitude upon exposure to extracellular Na ϩ and were eliminated following the addition of uridine (500 M) (data not shown). Consistent with the flux demonstration that both C561A and S561C(CϪ) displayed high affinity for Na ϩ ( Table  2 and Fig. 7B), the profiles of the presteady-state currents as a function of Na ϩ concentration resembled that of wild-type hCNT3 (17) and approached maximum values at low Na ϩ concentrations consistent with their respective apparent K 50 values for Na ϩ activation ( Table 2). In agreement with the decreased apparent affinity of C561S for Na ϩ ( Table 2 and Fig. 8A), however, C561S presteadystate currents were similar in profile to hCNT3(CϪ) (Fig. 5) and hfCNT, 5 and maximal only at Na ϩ concentrations of Ͼ50 mM.

DISCUSSION
hCNT3, the most recently discovered and functionally versatile of three human members of the SLC28 (CNT) protein family, is 691 amino acids in length, has a 13-TM membrane (or possibly 15-TM membrane) architecture, and utilizes electrochemical gradients of both Na ϩ and H ϩ to accumulate a broad range of pyrimidine and purine nucleosides and nucleoside drugs within cells (14,15,17,18). Although paralogs hCNT1 and hCNT2 have a similar predicted membrane topology, they function predominantly as Na ϩ -coupled transporters and are pyrimidine nucleoside-preferring and purine nucleoside-preferring, respectively (12,13,16,18). More widely distributed in cells and tissues than hCNT1 or hCNT2 (14) and with a central role in renal transepithelial nucleoside and nucleoside drug transport (9 -11), the multifunctional capability of hCNT3 makes it the protein of choice for systematic in depth molecular characterization by SCAM. A prerequisite of this approach is the availability of a functional cysteineless version of the transporter.
Cysteineless hCNT3CϪ, in which all 14 endogenous cysteine residues of hCNT3 have been converted to serine, is shown  Table 2.
hCNT3 Residue Cys-561 here to have a robust functional phenotype when produced in Xenopus oocytes. Similar to wild-type hCNT3 and consistent with correct folding of the cysteineless transporter, hCNT3CϪ exhibited broad selectivity for both pyrimidine and purine nucleosides in Na ϩ -containing medium (Fig. 1A), displayed the characteristic narrowing of this permeant selectivity in H ϩ -containing medium (Fig. 1A), was processed to the cell surface in amounts equivalent to hCNT3 (Fig. 2), exhibited apparent binding affinities for uridine in both Na ϩ -and H ϩ -containing medium similar to those of the wild-type transporter ( Table  1 and Fig. S3), and retained the wild-type Na ϩ /uridine and H ϩ /uridine coupling stoichiometries of 2:1 and 1:1, respectively (Fig. 6). Recombinant hCNT3CϪ is also functional in yeast and was used in initial SCAM analyses to screen residues in TMs 11-13 for reactivity to methanethiosulfonate reagents (37). In oocytes, hCNT3CϪ was previously used in SCAM analysis of TM 12 to test residues for inhibition by PCMBS (38). Consistent with correct folding of the nucleoside binding pocket of hCNT3CϪ, methanethiosulfonate and PCMBS accessibility to some residues was blocked by exofacial uridine (37,38). Taken together, these analyses validate use of   hCNT3 Residue Cys-561 SEPTEMBER 5, 2008 • VOLUME 283 • NUMBER 36 hCNT3CϪ as a template for SCAM analyses of CNT structure and function.
Characterization of hCNT3CϪ revealed an altered apparent affinity for Na ϩ ; hCNT3CϪ and wild-type hCNT3 exhibited apparent K 50 values for Na ϩ activation of 10 M radiolabeled uridine influx of Ͼ40 mM and 11-12 mM, respectively (Fig. 3, A  and B). Under voltage clamp conditions of Ϫ90 mV with a uridine concentration of 100 M, Na ϩ activation curves for hCNT3 and hCNT3CϪ were sufficiently shifted to the left to establish that the change in binding affinity for Na ϩ between hCNT3 and hCNT3CϪ was ϳ11-fold (apparent K 50 values of 2.2 and 25 mM, respectively) (Fig. 4). This difference, which was also apparent in presteady-state current measurements (Fig. 5), was specific to Na ϩ and did not extend to H ϩ (Fig. 3C). The change was also specific to one particular cysteine residue, Cys-561, in TM 12. Reintroduction of cysteine at position 561 (hCNT3CϪ mutant S561C(CϪ)) restored wild-type Na ϩ binding (Fig. 7), whereas the reciprocal single cysteine-to-serine substitution in hCNT3 (mutant C561S) produced an altered Na ϩ binding affinity corresponding to that found for hCNT3CϪ (Fig. 8). Presteady-state current measurements confirmed the Na ϩ -binding characteristics of S561C(CϪ) and C561S.
Recently, a naturally occurring hCNT3 variant has been described in the Spanish population in which a T/C transition leads to the substitution of cysteine at position 602 by arginine (44). This single amino acid replacement in TM 13 led to a shift in Hill coefficient consistent with a change in Na ϩ /nucleoside stoichiometry from 2:1 to 1:1. In agreement with our findings here, the C602R phenotype was a consequence of the insertion of arginine at this position rather than the loss of cysteine.
Fully processed to the cell surface, hCNT3CϪ exhibited an altered pattern of N-linked glycosylation relative to that of the wild-type protein. As shown in Fig. 2, two glycosylated forms of hCNT3/hCNT3CϪ were apparent on SDS-polyacrylamide gels (100 and 86 kDa). Most hCNT3CϪ immunoreactivity was associated with the 86 kDa band, in contrast to hCNT3, for which most immunoreactivity was found in the higher molecular mass band of 100 kDa. Since no fewer than six of the 14 endogenous cysteine residues of hCNT3 are located alongside the four potential N-terminal glycosylation acceptor sites in the extracellular N-terminal tail of the protein (supplemental Fig. S1) and since these cysteines probably form intramolecular disulfide bridges within the protein, it is likely that the modified glycosylation status of the cysteineless transporter is secondary to changes in the secondary/tertiary structure of this extramembranous region of the protein. The demonstration that aglyco-hCNT3 exhibited normal Na ϩ activation kinetics verified that altered glycosylation was not responsible for the shift in hCNT3CϪ Na ϩ affinity.
As well as replacement of hCNT3 Cys-561 with serine, each of the alternative amino acids found at position 561 in other CNT family members (alanine, valine, threonine, and isoleucine) were introduced into hCNT3. To provide a more complete spectrum of neutral amino acid side chain structures in the analysis, cysteine was also mutated to glycine. In addition to cysteine, only alanine at this position (hCNT3 mutant C561A) elicited wild-type cation binding (Tables 2 and 3 and Fig. 10). The rank order of Na ϩ binding affinities (cysteine and alanine Ͼ valine Ͼ glycine, serine, threonine, and isoleucine) demonstrated that normal hCNT3 Na ϩ activation was critically dependent upon a combination of residue 561 side chain bulk and polarity. Tolerance of alanine at position 561 was con- FIGURE 11. Steady-state currents of hCNT3 mutants. Steady-state currents were recorded for wild-type hCNT3 and mutants C561G, C561A, C561S, and C561I in oocytes that were voltage-clamped at Ϫ90 mV and exposed to 100 M uridine in transport medium containing 100 mM NaCl, pH 7.5 (A), or 100 mM ChCl, pH 5.5 (B). No uridine-induced currents were evident in control water-injected oocytes (data not shown).

TABLE 4 Uridine kinetic parameters for hCNT3 mutants
Radiolabeled 10 M uridine influx was measured in 100 mM NaCl transport medium, pH 8.5, and 100 mM ChCl transport medium, pH 5.5, under initial rate conditions (1 min) in oocytes producing C561G, C561A, C561S, C561V, C561T, and C561I. Kinetic parameters were calculated from data (not shown) corrected for basal (nonmediated) uptake in control water-injected oocytes. hCNT3 Residue Cys-561 firmed in the hCNT3CϪ background, in that mutant S561A(CϪ) exhibited partially recovered Na ϩ binding affinity (Table 2). Additionally, only C561A displayed wild-type hCNT3-like apparent affinities for uridine in both Na ϩ -and H ϩ -containing transport medium (Table 4). Therefore, mutation of Cys-561 to alanine instead of serine provides an alternative to hCNT3CϪ as template for future SCAM analyses of hCNT3 structure and function. Despite retaining Na ϩ -coupled transport capability, albeit with reduced Na ϩ binding affinity, substitution of Cys-561 by valine, threonine, and isoleucine had the additional and unanticipated consequence of dramatically decreasing H ϩ -coupled uridine transport activity (Figs. 10 and 11). Relative to Na ϩ , substitution of cysteine with the smaller amino acid glycine also led to reduced H ϩ -dependent uridine transport activity. Therefore, the nature of the amino acid side chain at position Cys-561 in hCNT3 influences both Na ϩ -and H ϩ -coupled nucleoside transport but with seemingly dissimilar structure-function profiles. For example, whereas the apparent affinity of mutant C561V for Na ϩ was only moderately decreased (apparent K 50 of 31 mM compared with 11 mM for wild-type hCNT3; Table 2), H ϩ -coupled uridine transport was essentially eliminated (Figs. 10 and 11). In the opposite direction, the apparent affinity of mutant C561G for Na ϩ was not measurable (Table 2), whereas H ϩ activation curves revealed wild-type hCNT3-like high affinity binding of H ϩ (apparent K 50 of 900 nM; Table 3). The structural requirements for Na ϩ and H ϩ binding and/or translocation at this single residue position are not, therefore, the same. At the level of the whole protein, other indications that Na ϩ and H ϩ binding and/or translocation have different structural requirements include the demonstration that hCNT3 exhibits different Na ϩ and H ϩ binding stoichiometries (17,18). Na ϩand H ϩ -coupled hCNT3 also have markedly different nucleoside and nucleoside drug selectivities, a finding that provides evidence for two distinct cation-dependent conformational states of the protein (17). Independent of changes in cation interactions and consistent with cation/nucleoside cotransport through a common translocation pore, amino acid substitutions at position Cys-561 were also found to have a marked influence on uridine transport kinetics, increasing the uridine apparent K m value up to 15-fold in Na ϩ -containing transport medium and up to 4-fold in H ϩ -containing transport medium (Tables 1 and 4).
In a separate study, we have independently identified Cys-561 as the cysteine residue responsible for inhibition of wildtype hCNT3 by PCMBS (38). Access of this membrane-impermeant probe to Cys-561 required H ϩ but not Na ϩ and was blocked by micromolar concentrations of extracellular uridine. Although this cysteine residue is conserved in Na ϩ -specific hCNT1 and hCNT2, neither transporter is affected by PCMBS. When converted to cysteine, two other residues in hCNT3 adjacent to Cys-561 (Ile-554 and Tyr-558) also led to H ϩ -activated inhibition by PCMBS (38). These findings suggest that Cys-561 is located in the translocation pore in a mobile region within or closely adjacent to the nucleoside binding pocket and that accessibility of PCMBS to this residue reports a specific H ϩ -induced conformational state of the protein (38). Positioned in the middle of TM 12 (Fig. S1), this residue position is now also revealed to be capable of modifying the functionality of the protein, with marked influences on both Na ϩ /nucleoside and H ϩ /nucleoside cotransport. Within the plane of the membrane, Cys-561 is located at the interface between those residues sensitive to inhibition by PCMBS in H ϩ -containing medium only and those where inhibition occurs in the presence of both Na ϩ and H ϩ (38). Remarkably intolerant to substitution by other amino acids, only cysteine replacement by alanine approached full wild-type kinetic functionality. Matching the previous observation that H ϩ -induced inhibition by PCMBS is specific to hCNT3 and not found in hCNT1/2 (38), we have similarly found that mutation of hCNT1 Cys-540 to serine (the cysteine substitution corresponding to hCNT3 C561S) had no influence on Na ϩ binding affinity (Fig. 9). hCNT3 has two Na ϩbinding sites, one of which may be Na ϩ -specific and the other of which may be shared functionally with H ϩ (17,38). hCNT3 Cys-561 seems to be primarily associated with the site interacting with both cations.
In conclusion, two independent lines of investigation have converged to identify Cys-561 as a key residue that resides in a conformationally mobile region of hCNT3 and is intimately involved in both Na ϩ /nucleoside and H ϩ /nucleoside cotransport. With actions seemingly specific to one of two hCNT3 cation binding sites, future investigations of Cys-561 and adjacent residues will be central to understanding the molecular intricacies of CNT cation/nucleoside cotransport and, in particular, to functionally separating and structurally identifying the two cation binding domains of hCNT3. Without access to a CNT crystal structure, SCAM-based approaches employing hCNT3CϪ as a template will be essential to these endeavors.