Conserved Glutamate Residues Are Critically Involved in Na+/Nucleoside Cotransport by Human Concentrative Nucleoside Transporter 1 (hCNT1)*

Human concentrative nucleoside transporter 1 (hCNT1), the first discovered of three human members of the SLC28 (CNT) protein family, is a Na+/nucleoside cotransporter with 650 amino acids. The potential functional roles of 10 conserved aspartate and glutamate residues in hCNT1 were investigated by site-directed mutagenesis and heterologous expression in Xenopus oocytes. Initially, each of the 10 residues was replaced by the corresponding neutral amino acid (asparagine or glutamine). Five of the resulting mutants showed unchanged Na+-dependent uridine transport activity (D172N, E338Q, E389Q, E413Q, and D565N) and were not investigated further. Three were retained in intracellular membranes (D482N, E498Q, and E532Q) and thus could not be assessed functionally. The remaining two (E308Q and E322Q) were present in normal quantities at cell surfaces but exhibited low intrinsic transport activities. Charge replacement with the alternate acidic amino acid enabled correct processing of D482E and E498D, but not of E532D, to cell surfaces and also yielded partially functional E308D and E322D. Relative to wild-type hCNT1, only D482E exhibited normal transport kinetics, whereas E308D, E308Q, E322D, E322Q, and E498D displayed increased K50Na+ and/or Kmuridine values and diminished VmaxNa+ and Vmaxuridine values. E322Q additionally exhibited uridine-gated uncoupled Na+ transport. Together, these findings demonstrate roles for Glu-308, Glu-322, and Glu-498 in Na+/nucleoside cotransport and suggest locations within a common cation/nucleoside translocation pore. Glu-322, the residue having the greatest influence on hCNT1 transport function, exhibited uridine-protected inhibition by p-chloromercuriphenyl sulfonate and 2-aminoethyl methanethiosulfonate when converted to cysteine.

Most nucleosides, including nucleoside analogs with antineoplastic and/or antiviral activity, are hydrophilic molecules that require specialized plasma membrane nucleoside transporter (NT) 4 proteins for uptake into or release from cells (1)(2)(3). NTmediated transport is a critical determinant of nucleoside and nucleotide metabolism and, for nucleoside drugs, their pharmacologic actions (3)(4)(5). By regulating adenosine availability to cell-surface purinoreceptors, NTs also profoundly affect neurotransmission, vascular tone, and other physiological processes (5,6). Two structurally unrelated families of integral membrane proteins exist in human and other mammalian cells as follows: the SLC28 concentrative nucleoside transporter (CNT) family, and the SLC29 equilibrative nucleoside transporter (ENT) family (3, 6 -8). ENTs are normally present in most, possibly all, cell types (8). CNTs, in contrast, are found predominantly in intestinal and renal epithelia and other specialized cells, suggesting important roles in absorption, secretion, distribution, and elimination of nucleosides and nucleoside drugs (1-4, 6, 7).
Although considerable progress has been made in molecular studies of ENT proteins (6,8), studies of structurally and functionally important residues within the CNT protein family are still at an early stage. Topological investigations suggest that hCNT1-3 and other eukaryote CNT family members have a 13 (or possibly 15)-transmembrane helix (TM) architecture, and multiple alignments reveal strong sequence similarities within the C-terminal half of the proteins (20). Prokaryote CNTs lack the first three TMs of their eukaryote counterparts, and functional expression of N-terminally truncated human/rat CNT1 in Xenopus oocytes has established that the first three TMs are not required for Na ϩ -dependent uridine transport activity (20). Consistent with these findings, chimeric studies between hCNT1 and hfCNT (12) and between hCNT1 and hCNT3 (17) have demonstrated that residues involved in Na ϩ -and H ϩ -coupling reside in the C-terminal half of the protein.
In hCNT1, two sets of adjacent residues in TM 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 of the transporter from CNT1-type to CNT2-type (21). Mutation of Ser-319 in TM 7 of hCNT1 to glycine was sufficient to enable transport of purine nucleosides, whereas mutation of the adjacent residue Gln-320 to methionine (which had no effect on its own) augmented this transport. The additional mutation of Ser-353 in TM 8 of hCNT1 to threonine converted S319G/Q320M from broadly selective (CNT3-type) to purine nucleoside-selective (CNT2type) but with relatively low adenosine transport activity. Further mutation of Leu-354 to valine increased the adenosine transport capability of S319G/Q320M/S353T, producing a full CNT2-type phenotype. Residues in both TMs 7 and 8 therefore play key roles in determining hCNT1/2 nucleoside selectivities. Confirming this, the double TM 8 mutant (S353T/L354V) was recently shown to exhibit a unique uridine-preferring transport phenotype (22). Mutation of Leu-354 alone markedly increased the affinity of the transporter for Na ϩ and Li ϩ , demonstrating that TM 8 also has a role in cation coupling (22).

EXPERIMENTAL PROCEDURES
Site-directed Mutagenesis and DNA Sequencing-hCNT1 cDNA (GenBank TM accession number U62968) in the Xenopus expression vector pGEM-HE (34) provided the template for construction of hCNT1 mutants by the oligonucleotide-directed technique (35), using reagents from the QuikChange site-directed mutagenesis kit (Stratagene) according to the manufacturer's directions. Constructs were sequenced in both directions by Taq dye deoxy terminator cycle sequencing to ensure that only the correct mutation had been introduced.
Production of Wild-type and Mutant hCNT1 Proteins in Xenopus Oocytes-hCNT1 cDNAs were transcribed with T7 polymerase and expressed in oocytes of Xenopus laevis by standard procedures (36). Healthy defolliculated stage VI oocytes were microinjected with 10 nl of water or 10 nl of water containing RNA transcript (1 ng/nl) and incubated in modified Barth's medium at 18°C for 72 h prior to the assay of transport activity.
Flux Assays-Transport was traced using 14 C/ 3 H-labeled nucleosides at 1 Ci/ml. Flux measurements were performed at room temperature (20°C) as described previously (36,37). Briefly, groups of 12 oocytes were incubated in 200 l of transport medium containing either 100 mM NaCl or choline chloride and 2 mM KCl, 1 mM CaCl 2 , 1 mM MgCl 2 , and 10 mM HEPES, pH 7.5. Unless otherwise indicated, the uridine concentration was 10 M. At the end of the incubation period, extracellular label was removed by six rapid washes in ice-cold Na ϩ -free (choline chloride) transport medium, and individual oocytes were dissolved in 1% (w/v) SDS for quantitation of oocyte-associated radioactivity by liquid scintillation counting (LS 6000 IC; Beckman). Initial rates of transport (influx) were determined using an incubation period of 1 min, except for mutants E322Q and E322D, which had low transport activity and required a longer incubation time (5 min) to achieve cellular uptake comparable with that of wild-type hCNT1 and the other mutants. In PCMBS inhibition studies, oocytes were pretreated with PCMBS (0.1 mM) on ice for 30 min and then washed five times with ice-cold transport medium to remove excess organomercurial before the assay of transport activity. Corresponding pretreatment with the MTS reagents MTSEA, MTSES, and MTSET (2.5, 10, and 1 mM, respectively) was performed at room temperature for 5 min. To demonstrate substrate protection, unlabeled uridine (20 mM) was included along with inhibitor during the preincubation step (38). 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. Kinetic (K m , K 50 , V max , and Hill coefficient) parameters (ϮS.E.) were determined using SigmaPlot software (Jandel Scientific). Statistical significance of the reported data sets was evaluated using t tests.
Electrophysiology-Steady-state and presteady-state currents were measured using the two-microelectrode voltage clamp as described previously (16).
Isolation of Membranes and Immunoblotting-Total membranes (plasma ϩ intracellular membranes) and purified plasma membranes were isolated by centrifugation from groups of 100 oocytes at 4°C in the presence of protease inhibitors as described previously (39,40). Colloidal silica (Sigma) was used to increase the gravitational density of plasma membranes and enhance their yield and purity (40). Protein was determined by the bicinchoninic acid protein assay (Pierce) using bovine serum albumin as standard.

RESULTS
Residues Identified for Mutagenesis-In this study, we employed site-directed mutagenesis and heterologous expression in Xenopus oocytes to analyze the roles of hCNT1 acidic amino acid residues. The locations of the residues selected for study are shown in Fig. 1. hCNT1 contains 51 aspartate and glutamate residues. Of these, 10 are conserved in other mammalian CNT family members and were included in the present analysis (hCNT1 residues Asp-172, Glu-308, Glu-322, Glu-338, Glu-389, Glu-413, Asp-482, Glu-498, Glu-532, and Asp-565). All but one (Asp-172) were located in the C-terminal half of the protein. In initial mutagenesis experiments, each of the 10 hCNT1 aspartate and glutamate residues were individually replaced by the corresponding neutral amino acids (asparagine or glutamine, respectively). All mutations were verified by sequencing the entire coding region of the double-stranded plasmid DNA in both directions. Except for the desired base changes, all sequences were identical to wildtype hCNT1.
Uridine Uptake and Cell-surface Expression of Wild-type hCNT1 and Mutants-The hCNT1 mutant transporters were expressed in Xenopus oocytes and assayed for uridine transport activity (10 M uridine influx, 1 min fluxes) in the presence and absence of Na ϩ as described under "Experimental Procedures." Representative values for mediated uridine uptake, corrected for basal nonmediated uptake in control water-injected oocytes, are presented in Table 1. Five of the mutants (D172N, E338Q, E389Q, E413Q, and D565N) exhibited Ͼ75% of wild-type Na ϩdependent transport activity and were not investigated further. In contrast, substitution of Glu-308 reduced transport activity by almost 90%, whereas mutation of E322Q, D482N, E498Q, and E532Q resulted in Ͼ99% loss of uridine transport activity. The time course of uridine uptake by wild-type hCNT1 shown in Fig. 2A demonstrates that the measured fluxes corresponded to initial rates of transport. Fig. 2A also demonstrates the absence of uridine uptake in water-injected oocytes.
Cell-surface expression of mutants E308Q, E322Q, D482N, E498Q, and E532Q was investigated by immunoblotting of purified oocyte plasma membranes using polyclonal antibodies (22) directed against amino acid residues 31-55 at the N terminus of the protein (Fig. 3A). Wild-type hCNT1 and transporters with mutations at positions 308 and 322 were present in similar amounts, indicating that these single amino acid substitutions resulted in loss of intrinsic hCNT1 transport activity without altering surface quantities in the oocyte plasma membrane. Antibody specificity was confirmed by lack of immunoreactivity in membranes prepared from control water-injected oocytes. Unlike E308Q and E322Q, very little plasma membrane immunoreactivity was detected for the transporters having mutations at positions 482, 498, and 532, indicating that the lack of transport activity was associated with reduced cell-surface expression. Membrane-spanning ␣-helices predicted from bioinformatic analyses of currently identified CNT family members are numbered 1-13 (strongly predicted) and 5A and 11A (weakly predicted). Highly conserved acidic amino acid residues are indicated with filled circles.

TABLE 1 Uridine uptake by wild-type hCNT1 and mutants expressed in Xenopus oocytes
Oocytes producing recombinant hCNT1 and hCNT1 mutants were incubated in transport medium with or without Na ϩ at 20°C as described under "Experimental Procedures." Each value is the mean Ϯ S.E. from 10 to 12 oocytes. For influx in the presence of Na ϩ , significant differences in mediated uridine uptake (p Ͻ 0.05) compared with wild-type hCNT1 are indicated by *.

Mediated uridine uptake
Protein expression a Na ؉ medium Na ؉ -free medium Relative expression in the plasma membrane is from Fig. 3. b ND indicates not determined.
hCNT1 Glutamate Residues OCTOBER 19, 2007 • VOLUME 282 • NUMBER 42 In a second round of mutagenesis experiments, Glu-308, Glu-322, Asp-482, Glu-498, and Glu-532 of hCNT1 were replaced by the alternative negatively charged amino acid (i.e. glutamate to aspartate or aspartate to glutamate). Time courses of uridine accumulation in the presence of Na ϩ and the initial rate of uridine uptake in the presence and absence of Na ϩ by E308D, E322D, D482E, E498D, and E532D produced in Xenopus oocytes are shown in Fig. 2A. Relative to the corresponding asparagine and glutamine mutations (Table 1), oocytes expressing E308D, D482E, or E498D exhibited higher uridine transport activities with fluxes that were broadly similar to those of wild-type hCNT1. E322D, in contrast, displayed only partial transport activity. In all cases, uridine uptake was Na ϩ -dependent (Fig. 2B). Transport was also concentrative because the 15and 30-min uptake values exceeded the initial extracellular uridine concentration of 10 M ( Fig. 2A), assuming an intracellular water volume of 1 l (36, 37). Immunological analysis revealed that low activity E322D was present in normal quantities at the cell surfaces (Fig.  3B), demonstrating that its low transport activity was not because of defective insertion and/or stability of the transporter in oocyte plasma membranes. Unlike the other positions, hCNT1 with Glu-532 substituted by aspartate (E532D) lacked functional activity (Fig. 2, A and B) and remained undetectable at the cell surface by immunoblotting (Fig. 3B). The immunoblots shown in Fig. 3C demonstrated that oocyte total membranes (plasma ϩ intracellular membranes) contained mutants D482N, E498Q, E532Q, and E532D, indicating that the "no activity" mutants were present within cells, confirming improper targeting to the plasma membrane.
Kinetic Characterization of Wild-type hCNT1 and Mutants-To further elucidate the effects of individual replacement of amino acids Glu-308, Glu-322, Asp-482, and Glu-498 on hCNT1 transport function, mutants E308D, E308Q, E322D, E322Q, D482E, E498D, and wild-type hCNT1 were characterized for Na ϩ activation and uridine concentration dependence kinetics. Na ϩ activation was investigated by measuring 10 M uridine influx as a function of Na ϩ concentration (Fig. 4), whereas the concentration dependence of uridine uptake was determined at a saturating concentration (100 mM) of Na ϩ (Fig.  5). Kinetic parameters derived from these data, including measures of apparent Na ϩ affinities (K 50 Naϩ ), apparent uridine affinities (K m uridine ), and maximal rates of transport (V max Naϩ and V max uridine ) are summarized in Table 2. In agreement with previous studies (16,18), the rate of uridine transport by oocytes expressing wild-type hCNT1 increased markedly as the Na ϩ concentration was increased from 0 to 100 mM and was essentially saturated at Na ϩ concentrations above 40 mM. In contrast, rates of uridine transport by oocytes producing E308Q, E322D, or E322Q did not reach saturation, even at the highest concentration of Na ϩ (100 mM) (Fig. 4). Hill-type analysis of the data yielded an apparent K 50 Naϩ value of 8.2 mM for wild-type hCNT1 compared with 20 mM for E308D and Ͼ40 mM for E308Q, E322D, and E322Q. E308Q, E322D, and E322Q also showed major reductions in V max Naϩ values (Ͻ1 pmol/oocyte⅐min Ϫ1 , compared with 5.6 pmol/oocyte⅐min Ϫ1 for wild-type hCNT1). V max Naϩ :K 50 Naϩ ratios (a measure of transport efficiency) were 0.68 for wildtype hCNT1, 0.17 for E308D, and Ͻ0.03 for E308Q, E322D, and  hCNT1 Glutamate Residues E322Q. In the absence of Na ϩ , uridine uptake by all mutants was close to zero, indicating an absence of measurable slippage (uncoupled, Na ϩ -independent uridine uptake). As also shown in Fig. 4 and summarized in Table 2, D482E and E498D exhibited Na ϩ activation kinetics broadly similar to those of wildtype hCNT1. In the case of E498D, however, there was a noticeable reduction in the V max Naϩ values (2.5 pmol/oocyte⅐min Ϫ1 ), leading to a corresponding decrease in the V max Naϩ :K 50 Naϩ ratio (0.26). Where measurable (hCNT1, E308D, D482E, and E498D), Hill coefficients were consistent with a Na ϩ :uridine coupling ratio of 1:1 (16,18,19).
To minimize the potential effects of altered Na ϩ apparent affinity on uridine kinetic parameters, experiments to investigate uridine transport kinetics were undertaken at the maximum possible Na ϩ concentration of 100 mM. As shown in Fig. 5 and summarized in Table 2 E308Q, E322D, and E322Q (3.8, 1.8 ,and 1.2 pmol/ oocyte⅐min Ϫ1 , respectively). Compared with a value of 0.66 for wild-type hCNT1, V max uridine :K m uridine ratios ranged from 0.18 and 0.25 for E308D and E498D, respectively, to 0.08, 0.06, and 0.02 for E308Q, E322D, and E322Q, respectively. The nucleoside selectivities of E308D, E322D, D482E, and E498D were identical to wild-type hCNT1 (Table 3).
Electrophysiological Characterization of Wild-type hCNT1 and Mutants-In steady-state electrophysiological experiments (16), all mutants (E308D, E308Q, E322D, E322Q, D482E, and E498D) were confirmed to mediate uridine-induced Na ϩ inward currents (Fig. 6A). No currents were detected in the absence of Na ϩ or in control water-injected oocytes. Measured in the same batch of oocytes on the same day, there was excellent correlation between the magnitudes of the currents recorded and the corresponding Na ϩ -dependent fluxes of radiolabeled uridine (Fig. 6B). The one exception was mutant E322Q, which exhibited an elevated uridine-induced Na ϩ current disproportionate to its very low Na ϩ -dependent uridine transport activity (compare, for example, E322D and E322Q in Fig. 6, A and B).
In parallel presteady-state electrophysiological experiments performed in the absence of uridine, oocytes producing wildtype hCNT1 and mutants E308D, E308Q, E322D, and E322Q . Sodium activation kinetics of wild-type hCNT1 and mutants. Initial rates of transporter-mediated radiolabeled uridine uptake (10 M, 20°C) were measured in transport media containing 0 -100 mM NaCl, using choline chloride to maintain isosmolality. Each value is the mean Ϯ S.E. from 10 to 12 oocytes. Mediated transport was calculated as uptake in RNA-injected oocytes minus uptake in oocytes injected with water alone. Kinetic parameters derived from the data are presented in Table 2.
hCNT1 Glutamate Residues OCTOBER 19, 2007 • VOLUME 282 • NUMBER 42 were voltage-clamped at a holding potential (V h ) of Ϫ50 mV, and presteady-state currents were activated by voltage steps to a series of test potentials (V t ) (Fig. 7, A and B). In agreement with previous studies (16), current relaxations that were largely eliminated upon removal of external Na ϩ were observed for wild-type hCNT1. Presteady-state currents were greatly reduced in mutant E322D and absent from mutants E308Q, E308D, E322Q, and control water-injected oocytes.
Na ϩ :Nucleoside Coupling Ratios of Wild-type hCNT1 and Mutants-The Na ϩ /uridine coupling stoichiometries of wildtype hCNT1 and mutants E308Q and E322Q were determined by simultaneously measuring Na ϩ currents and radiolabeled uridine uptake under voltage-clamp conditions. The linear cor-relations between integrated uridine-dependent charge and radiolabeled uridine accumulation measured in Na ϩ -containing transport medium (100 mM) gave calculated Na ϩ /nucleoside coupling ratios of 0.88 Ϯ 0.06 and 0.82 Ϯ 0.03 for hCNT1 and E308Q, respectively (Fig. 8, A and B). Consistent with E322Q functioning as a partially uncoupled uridine-gated Na ϩ channel, there was no correlation between Na ϩ current and radiolabeled uridine uptake, charge:flux ratios for individual E322Q-producing oocytes ranging from 5 to 74 (Fig. 8C).
PCMBS and MTS Inhibition of Uridine Transport by Wildtype hCNT1 and Mutants-Residues lining the translocation pore can be identified through the use of hydrophilic thiolreactive reagents such as p-chloromercuribenzene sulfonate FIGURE 5. Uridine kinetics of wild-type hCNT1 and mutants. Initial rates of transporter-mediated radiolabeled uridine uptake (20°C) were measured in transport medium containing 100 mM NaCl. Each value is the mean Ϯ S.E. from 10 to 12 oocytes. Mediated transport was calculated as uptake in RNA-injected oocytes minus uptake in oocytes injected with water alone. Kinetic parameters derived from the data are presented in Table 2.
Although wild-type hCNT1 contains 20 endogenous cysteine residues, and consistent with results of previous studies (38), there was no change in hCNT1-mediated uridine uptake following incubation with membrane-impermeant PCMBS at a concentration of 0.1 mM (Fig. 9A). PCMBS exposure also had no measurable effect on E308C transport activity. In contrast, uridine uptake by E322C was strongly inhibited by PCMBS, and the presence of extracellular uridine (20 mM) protected the transporter against this inhibition (Fig. 9A). Reflecting their different reactivities with thiol groups, MTSEA, MTSES, and MTSET were tested on wild-type hCNT1 and mutants E308C and E322C at concentrations of 2.5, 10, and 1 mM, respectively. Only MTSEA gave significant inhibition, and only E322C was affected (Fig. 10A). Like PCMBS, addition of uridine (20 mM) to the extracellular medium protected E322C against MTSEA inhibition (Fig. 10B).

DISCUSSION
As shown in Fig. 1 for hCNT1, current models of hCNT topology have 13 putative TMs (9 -11, 20). Computer algorithms also weakly predict two additional potential transmembrane regions, designated in Fig. 1 as TMs 5A and 11A (20). Location of the N and C termini as intracellular and extracellular, respectively, derives from immunocytochemical experiments with site-specific antibodies, and from studies of native and introduced glycosylation sites (20). Cytoplasmic exposure of the loop linking TMs 4 and 5 has been similarly confirmed (20). Both a 13 TM and 15 TM membrane architecture are consistent with these landmarks. Initial substituted cysteine accessibility method analyses of TMs 11, 12, and 13 of a functional cysteine-free version of a human CNT3 (hCNT3CϪ) using MTS reagents (46), as well as other published structure/ function studies (e.g. 21,22), are also consistent with both models. TMs 1-3 of human and rat CNT1 are not required for Na ϩ -dependent uridine transport activity (20).
The present study identified five conserved acidic amino acid residues of Na ϩ /nucleoside cotransporter hCNT1 (Glu-308, FIGURE 6. Uridine-induced steady-state currents of wild-type hCNT1 and mutants. Oocytes were injected with 10 ng of RNA transcripts or water alone and incubated for 3 days. A, averaged inward currents in hCNT1-and mutant-producing oocytes perfused with 1 mM uridine in the presence or absence of 100 mM Na ϩ in the incubation medium. Each value is the mean Ϯ S.E. from five oocytes. No currents were observed in control water-injected oocytes. B, initial rates of radiolabeled uridine uptake (10 M, 20°C) measured in the presence or absence of 100 mM Na ϩ in the incubation medium. H 2 O, control water-injected oocytes. Each value is the mean Ϯ S.E. from 10 to 12 oocytes.

TABLE 3 Nucleoside uptake by wild-type hCNT1 and mutants expressed in Xenopus oocytes
Oocytes producing recombinant hCNT1 and hCNT1 mutants were incubated with different radiolabeled nucleosides (10 M) in NaCl transport medium at 20°C as described under "Experimental Procedures." Each value is corrected for basal uptake in control water-injected oocytes and is the mean Ϯ S.E. from 10 to 12 oocytes. Significant differences in mediated nucleoside uptake (p Ͻ 0.05) compared with wild-type hCNT1 are indicated by *. To permit a side-by-side comparison of pyrimidine and purine nucleoside uptake, all fluxes were determined using a 5-min incubation interval. Glu-322, Asp-482, Glu-498, and Glu-532) whose mutation to the corresponding neutral amino acid (glutamine or asparagine) dramatically reduced transport activity of the recombinant protein produced in Xenopus oocytes. All are situated in the C-terminal half of the protein, a region previously implicated in both nucleoside and cation translocation (12,17,21). Glu-308 and Glu-322 are located in predicted TM 7; Asp-482 is adjacent to the extracellular face of TM 11; Glu-498 is located in the middle of TM11A; and Glu-532 is on the extracellular boundary of TM 12 (Fig. 1). Although TM 7 mutants E308Q and E322Q targeted normally to cell surfaces, their impaired intrinsic transport activities suggested key roles for both glutamate residues in hCNT1-mediated transport. D482N, E498Q, and E532Q, in contrast, were retained in intracellular membranes. Mutant E532D, in which the negative charge of Glu-532 was replaced with aspartate, was also retained in intracellular membranes, implying an absolute requirement of this glutamate residue to achieve normal cell-surface levels. The predicted location of Glu-532 on the immediate extramembranous surface of TM 12 (Fig. 1) suggests that it may be necessary for the anchoring and/or correct membrane configuration of this TM. Replacement of Asp-482 with glutamate, and Glu-498 with aspartate, in contrast, yielded normal processing to cell surfaces and resulted in uridine fluxes broadly similar to those of wild-type hCNT1. The functional activities of TM 7 mutants E308Q and E322Q were also partially restored by replacement  , indicating that this residue is unlikely to have a mechanistic role in hCNT1 Na ϩ /nucleoside cotransport. Mutation of Glu-498 to aspartate (mutant E498D), however, led to a Ͼ50% reduction in V max Naϩ and an almost 2-fold increase in the K m uridine value such that both the V max Naϩ :K 50 Naϩ and V max uridine :K m uridine ratios were reduced relative to those of wild-type hCNT1. Although substitution of Glu-498 with glutamine would be predicted to exhibit more marked changes in transport function, mutant E498Q was not processed to cell surfaces and thus could not be characterized. Consistent with a key role for Glu-498, it is centrally positioned in the most highly conserved sequence motif in the entire CNT family (G/A)XKX 3 NEFVA(Y/M/F). Mutation of Glu-519, the residue in hCNT3 that corresponds to hCNT1 Glu-498, impaired both the Na ϩ /nucleoside and H ϩ /nucleoside transport activities of the transporter. 5 As illustrated in Fig. 1, Glu-498 and the conserved motif of which it is a part are located in a region of the protein that is potentially exofacial (in the loop linking TMs 11 and 12) or membrane-associated (TM 11A). The functional significance of Glu-498 revealed by this study favors the latter possibility. If TM 11A is transmembrane (as opposed to a re-entrant loop), there is the likelihood that the TM 5A region is also transmembrane (to preserve the experimentally determined endofacial and exofacial locations of the N and C termini of the protein, respectively). A consequence of this is that the central TM 6 -11 region of the protein may be in an orientation opposite to that shown in Fig. 1  (please see below).

Mediated
Residues Glu-308 and Glu-322-Effects on hCNT1 transport kinetics were also apparent by mutation of Glu-308 and Glu-322. In the presence of a saturating concentration of Na ϩ , substitution of Glu-308 by aspartate (mutant E308D) resulted in an almost 4-fold increase in K m uridine with no change in V max uridine . A 3-fold increase in K 50 Naϩ combined with a modest reduction in V max Naϩ was also observed. Therefore, both V max Naϩ : K 50 Naϩ and V max uridine :K m uridine ratios of the transporter were affected. More pronounced kinetic effects were apparent for E308Q and E322D/E322Q, a consistent feature of these mutants being a dramatic reduction in apparent affinities for Na ϩ (increase in K 50 Naϩ ) as well as V max Naϩ and V max uridine values. Two of the mutants (E308Q and E322Q) also exhibited reductions in apparent affinities for uridine (increase in K m uridine ). As a result, V max Naϩ :K 50 Naϩ and V max uridine :K m uridine ratios were severely compromised, indicating critical roles for Glu-308 and Glu-322 in cation and nucleoside binding and/or translocation. Because removal of the carboxylate groups at these positions (mutants E308Q and E322Q) still allowed residual Na ϩ -dependent uridine transport activity, albeit with very low kinetic efficiency, electrostatic interactions involving Glu-308 and Glu-322 must not in themselves be obligatory for function.
Similar to the proposed roles of Asp-187 in the E. coli PutP Na ϩ /proline transporter (27), Glu-269 and Glu-325 of E. coli LacY H ϩ -coupled lactose permease (24,25,47), and Asp-369 and Asp-404 of the Aquifex aeolicus LeuT Aa Na ϩ /Cl Ϫ -dependent leucine transporter (48), the profound and complementary effects of hCNT1 Glu-308 and Glu-322 mutations on both V max Naϩ and V max uridine values suggest that these residues may facilitate conformational transitions within the Na ϩ /nucleoside transport cycle. As indicated by the observed effects of Glu-308 and Glu-322 mutations on binding affinities for Na ϩ and uridine, secondary roles in cation and nucleoside binding are also possible. Examples where this occurs in other transporters include Asp-55 of PutP (26), and glutamate and aspartate residues in the E. coli MelB Na ϩ /melibiose and GlpT glycerol 3-phosphate transporters (23,49), the mammalian NaDC-1 Na ϩ /dicarboxylate transporter (28), the Na ϩ /H ϩcoupled EAAC1 glutamate transporter (30), the NHE1 Na ϩ /H ϩ exchanger (31), and the Na ϩ /Cl Ϫ -dependent dopamine transporter (32). As hypothesized for Glu-325 of LacY  hCNT1 Glutamate Residues OCTOBER 19, 2007 • VOLUME 282 • NUMBER 42 transport V max :K m ratio with the major effect on V max (50), a kinetic outcome similar to that seen here for mutation of hCNT1 residues Glu-308 and Glu-322. Mutant E308Q retained the wild-type Na ϩ /nucleoside coupling ratio of 1:1. In addition to possessing low level Na ϩ -dependent uridine transport activity, however, and different from E308Q, mutant E322Q also exhibited features consistent with uncoupled Na ϩ transport. This was manifest by disproportionately high uridine-induced Na ϩ currents causing variable Na ϩ / nucleoside charge:flux ratios in excess of the expected wildtype value of 1:1. Residues of other transporters where mutation causes channel-like behavior include Asp-204 of the human SGLT1 and Asn-177 of the rat 5-hydroxytryptamine transporter (51)(52).
Glu-308 and Glu-322 are both located on the putative hydrophilic surface of TM 7 (21). This surface also includes the residues Asn-315 and Ser-319 (previously shown to determine hCNT1/2 nucleoside selectivities) (21). The present results further strengthen the functional importance of TM 7 and suggest that the helix face containing these four residues lines a common Na ϩ /nucleoside translocation pore. TM 8 is also likely to be pore-lining (22). In TM 7, Glu-322 is located close to its extracellular aspect according to the putative 13-TM topology model shown in Fig. 1, but it would be close to its intracellular aspect if TMs 6 -11 were in the opposite orientation as predicted by the alternate 15-TM models of hCNT3 topology. When converted to cysteine (mutant E322C), this residue was accessible to membrane-impermeant PCMBS added to the extracellular medium, which resulted in marked transport inhibition that was prevented by externally applied uridine. Supporting the pore-lining location of Glu-322, the corresponding residue in a cysteine-free version of hCNT3 (hCNT3CϪ) (46), when converted to cysteine, also resulted in PCMBS inhibition of uridine uptake. 5 The smallest of the MTS reagents tested (MTSEA) was also inhibitory against hCNT1 E322C. As with PCMBS, externally applied uridine protected the transporter against MTSEA inhibition. Therefore, Glu-322 evidently lies within the permeant translocation channel in a position with restricted access that is occluded by uridine. Potentially located deep inside the translocation channel in a position within or in close proximity below the uridine binding pocket, this is more consistent with the TM 7 orientation shown in Fig. 11 (15-TM model) than with that in Fig. 1 (13-TM model). Lack of a corresponding effect of PCMBS and MTS reagents on uridine transport by E308C may reflect the more external location of this residue within the translocation vestibule (Fig. 11). Possibly, therefore, Glu-308 and Glu-322 may include parts of the extracellular and internal gates of the transporter, respectively, FIGURE 10. MTS reagent inhibition of wild-type hCNT1, E308C, and E322C. A, oocytes producing wild-type hCNT1 or mutants E308C or E322C were incubated with or without MTSEA (2.5 mM), MTSES (10 mM), or MTSET (1 mM) in NaCl transport medium at 20°C. After 5 min, excess MTS reagents were removed by washing in ice-cold medium. Initial rates of uridine uptake were then determined (10 M) at 20°C. B, protection from MTSEA inhibition by uridine. Oocytes producing wild-type hCNT1 or mutants E308C or E322C were incubated with or without MTSEA (2.5 mM) in NaCl transport medium at 20°C in the absence or presence of unlabeled uridine (20 mM). After 5 min, excess MTSEA and uridine were removed by washing in ice-cold medium. Initial rates of uridine uptake were then determined (10 M) at 20°C. Each value is the mean Ϯ S.E. from 10 to 12 oocytes and was corrected by subtraction of the corresponding basal uptake value in control water-injected oocytes.
FIGURE 11. Topological model of TM7 of hCNT1. Residues Glu-308 and Glu-322 are indicated with filled circles. Helix orientation is that predicted by a 15-TM membrane architecture. hCNT1 Glutamate Residues a function supported by the channel-like behavior of mutant E322Q revealed by steady-state currents, and by the potential gating function of negatively charged residues within the common cation/solute translocation pore of the recently solved three-dimensional crystal structure of A. aeolicus LeuT Aa (48). In the latter protein, negatively charged residues stabilize the transporter in a closed conformation that occludes closely associated Na ϩ and leucine-binding sites halfway across the membrane bilayer. Similar to the mammalian GAT1 Na ϩ /Cl Ϫ -dependent ␥-aminobutyric acid transporter (53,54), a member of the same protein family as LeuT Aa , hCNT1 (and hCNT3) presteady-state currents largely reflect binding and potential occlusion of extracellular Na ϩ . 5 Consistent with a gating function for residues Glu-308 and Glu-322, their mutation markedly impaired hCNT1 presteady-state currents.
Conclusions-The present results for hCNT1 suggest close proximity integration of cation/solute binding and transport within a common cation/permeant translocation pore, and reveal important roles for three postulated intramembranous glutamate residues (Glu-308, Glu-322, and Glu-498) in cation/ nucleoside translocation. Setting the stage for additional future substituted cysteine accessibility method and other analyses of hCNT structure and function, the findings favor a revised 15-TM model of hCNT1 membrane architecture.