A 40-kDa polypeptide from papain digestion of the rabbit intestinal Na+/phosphate cotransporter retains Na+ and phosphate cotransport

We searched for new fluorescent probes of catalytic-site nucleotide binding in F(1)F(0)-ATP synthase by introducing Cys mutations at positions in or close to catalytic sites and then reacting Cys-mutant F(1) with thiol-reactive fluorescent probes. Four suitable mutant/probe combinations were identified. beta F410C labeled by 7-fluorobenz-2-oxa-1,3-diazole-4-sulfonamide (ABD-F) gave very large signal changes in response to nucleotide, allowing facile measurement of fluorescence and nucleotide-binding parameters, not only in F(1) but also in F(1)F(0). The results are consistent with the presence of three asymmetric catalytic sites of widely different affinities, with similar properties in both enzymes, and revealed a unique probe environment at the high-affinity site 1. beta Y331C F(1) labeled by ABD-F gave a large signal which monitored catalytic site polarity changes that occur along the ATP hydrolysis pathway. Two other mutant/probe combinations with significant nucleotide-responsive signals were beta Y331C labeled by 5-((((2-iodoacetyl)amino)ethyl)amino)naphthaline-1-sulfonic acid and alpha F291C labeled by 2-4'-(iodoacetamido)anilino)naphthalene-6-sulfonic acid. The signal of the latter responds differentially to nucleoside diphosphate versus triphosphate bound in catalytic sites.     

Establishing a definitive stoichiometry for the Na+/monocarboxylate cotransporter SMCT1

Several different stoichiometries have been proposed for the Na(+)/monocarboxylate cotransporter SMCT1, including variable Na(+)/substrate stoichiometry. In this work, we have definitively established an invariant 2:1 cotransport stoichiometry for SMCT1. By using two independent means of assay, we first showed that SMCT1 exhibits a 2:1 stoichiometry for Na(+)/lactate cotransport. Radiolabel uptake experiments proved that, unlike lactate, propionic acid diffuses passively through oocyte membranes and, consequently, propionate is a poor candidate for stoichiometric determination by these methods. Although we previously determined SMCT1 stoichiometry by measuring reversal potentials, this technique produced erroneous values, because SMCT1 simultaneously mediates both an inwardly rectifying cotransport current and an outwardly rectifying anionic leak current; the leak current predominates in the range where reversal potentials are observed. We therefore employed a method that compared the effect of halving the external Na(+) concentration to the effect of halving the external substrate concentration on zero-current potentials. Both lactate and propionate were cotransported through SMCT1 using 2:1 stoichiometries. The leak current passing through the protein has a 1 osmolyte/charge stoichiometry. Identification of cotransporter stoichiometry is not always a trivial task and it can lead to a much better understanding of the transport activity mediated by the protein in question.     

Isolation and reconstitution of the intestinal Na+/glucose cotransporter

The intestinal Na+/glucose cotransporter was isolated from brush border membrane vesicles using a three-step procedure and Na(+)-dependent phlorizin binding as the measure of cotransporter enrichment. The initial step was to treat the Ca2(+)-precipitated brush border membrane vesicles with 0.02% sodium dodecyl sulfate (SDS) followed by sucrose gradient centrifugation which resulted in a 5-fold enrichment of the Na+/glucose cotransporter. The second step was chromatofocusing chromatography over the pH range from pH 7.4 to pH 4.0. This step resulted in an additional 20-fold purification as compared with the SDS-brush border membrane vesicle protein which served as the starting material. The final step was affinity chromatography on con A-Sepharose which resulted in a 5-fold enrichment of the chromatofocused protein. The glycoprotein fraction from the concanavalin A column reconstituted into phosphatidyl choline: cholesterol liposomes demonstrated Na(+)-dependent, phlorizin-sensitive, and osmotic strength-sensitive glucose uptake. This fraction consisted of a single 75-kDa polypeptide on SDS-polyacrylamide gel electrophoresis upon staining with silver. On the basis of these criteria it appears that a protocol for the isolation of the Na+/glucose cotransporter has been developed.     

Examination of substrate-induced conformational changes in the Na+/glucose cotransporter

Conformations of the Na+/glucose cotransporter were examined using tryptophan fluorescence and substrates to induce cotransporter conformational changes. Addition of Na+ but not K+ or TMA+ resulted in a saturable quenching of tryptophan fluorescence with a K0.5 for Na+ of 28 mM. In the presence of saturating Na+ concentrations, d-glucose but not l-glucose, fructose, or phlorizin resulted in a partial return of tryptophan fluorescence to approximately 70% of the substrate-free levels. This return of tryptophan fluorescence was a saturable function of d-glucose concentration with a K0.5 of 43 microM. The three conformations were compared with respect to their sensitivity to tryptophan quench reagents. Acrylamide quenching was unaffected by substrates. In contrast, I- quenching decreased 40% in the presence of Na+, while Cs+ quenching increased 64%. Addition of saturating d-glucose concentrations resulted in the return of I- quenching to 90% of the substrate-free values and reduced Cs+ quenching to substrate-free levels. In contrast, phlorizin did not mimic the effect of d-glucose on tryptophan fluorescence. These results are interpreted in terms of a second substrate-induced cotransporter conformational change which based on similar substrate specificities appears directly related to cotransporter-mediated Na+ and d-glucose transport.     

Two substrate sites in the renal Na+-d-glucose cotransporter studied by model analysis of phlorizin binding and-d-glucose transport measurements

Summary Time courses of phlorizin binding to the outside of membrane vesicles from porcine renal outer cortex and outer medulla were measured and the obtained families of binding curves were fitted to different binding models. To fit the experimental data a model with two binding sites was required. Optimal fits were obtained if a ratio of low and high affinity phlorizin binding sites of 1:1 was assumed. Na⁺ increased the affinity of both binding sites. By an inside-negative membrane potential the affinity of the high affinity binding site (measured in the presence of 3 mM Na⁺) and of the low affinity binding site (measured in the presence of 3 or 90 mM Na⁺) was increased. Optimal fits were obtained when the rate constants of dissociation were not changed by the membrane potential. In the presence of 90 mM Na⁺ on both membrane sides and with a clamped membrane potential,K _D values of 0.4 and 7.9 M were calculated for the low and high affinity phlorizin binding sites which were observed in outer cortex and in outer medulla. Apparent low and high affinity transport sites were detected by measuring the substrate dependence ofd-glucose uptake in membrane vesicles from outer cortex and outer medulla which is stimulated by an initial gradient of 90 mM Na⁺(out>in). Low and high affinity transport could be fitted with identicalK _m values in outer cortex and outer medulla. An inside-negative membrane potential decreased the apparentK _m ofhigh affinity transport whereas the apparentK _m of low affinity transport was not changed. The data show that in outer cortex and outer medulla of pighigh and low affinity Na⁺-d-glucose cotransporters are present which containlow and high affinity phlorizin binding sites, respectively. It has to be elucidated from future experiments whether equal amounts of low and high affinity transporters are expressed in both kidney regions or whether the low and high affinity transporter are parts of the same glucose transport moleculc.     

Effects of membrane potential on Na cotransports in eel intestinal brush-border membrane vesicles: Studies with a fluorescent dye

Summary The Na-dependent transport of a number of organic molecules (d-glucose,l-proline,l-alanine,l-phenylalanine) in brush-border membrane vesicles isolated from the intestine of the eel (Anguilla anguilla) was monitored by recording the fluorescence quenching of the voltage-sensitive cyanine dye 3,3-diethylthiacarbocyanine iodide (DiS-C₂(5)). The experimental approach consisted of: a) generating an inside-negative membrane potential mimicking in vivo conditions: b) measuring the rate of membrane potential decay (i.e., the rate of fluorescence quenching decay) due to Na-neutral substrate cotransport. Rates of membrane potential decay showed saturation on substrate concentration andK _app values (the substrate concentration giving 50% of the maximal rate) were estimated for Na-dependent transport ofd-glucose (0,099mm),l-alanine (0.516mm),l-proline (0.118mm) andl-phenylalanine (2.04mm). The influence of an inside-negative membrane potential on the affinity of the transporter for glucose and for sodium is discussed.     

Determinants of substrate and cation affinities in the Na+/dicarboxylate cotransporter.

The Na+/dicarboxylate cotransporter (NaDC-1) couples the transport of sodium and tricarboxylic acid cycle intermediates, such as succinate and citrate. The rabbit and human homologues (rbNaDC-1 and hNaDC-1, respectively) are 78% identical in amino acid sequence but exhibit several differences in their functional properties. rbNaDC-1 has a greater apparent affinity for citrate and sodium than hNaDC-1. Furthermore, unlike hNaDC-1, rbNaDC-1 is inhibited by low concentrations of lithium. In this study, chimeric transporters were constructed to identify the protein domains responsible for the functional differences between rbNaDC-1 and hNaDC-1. Individual substitutions of transmembrane domain (TMD) 7, 10 or 11 produced transporters with intermediate properties. However, substitution of TMD 7, 10, and 11 together resulted in a transporter with the citrate Km of the donor, suggesting that interactions between these domains determine the differences in apparent citrate affinities. TMDs 10 and 11 are most important in determining the differences in apparent sodium affinities, and TMD 11 determines the sensitivity to lithium inhibition. We conclude that transmembrane domains 7, 10, and 11 in NaDC-1 may contain at least one of the cation binding sites in close proximity to the substrate binding domain.     

Evidence for tyrosyl residues at the Na+ site on the intestinal Na+/glucose cotransporter

A tyrosine group has been identified at, or near, the Na+-binding site of the Na+/glucose and Na+/proline cotransporters of rabbit intestinal brush-borders. Three tyrosine group-specific reagents, n-acetylimidazole, tetranitromethane, and p-nitrobenzene sulfonyl fluoride, were used to evaluate the role of tyrosyl groups in Na+-dependent glucose transport, Na+-dependent phlorizin binding, and the Na+-induced fluorescence quenching of fluorescein isothiocyanate bound to the glucose site of the carrier. All three reagents inhibited glucose transport, phlorizin binding, and fluorescein isothiocyanate quenching by 50-85% with Ki values in the range 7-50 microM. The presence of Na+ during the exposure of membranes to the reagents completely protected against inhibition, the Na+ concentration required to produce 50% protection was 14-36 mM. Fluorescent derivatives of n-acetylimidazole were synthesized to identify the tyrosyl residues on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. A total of five polypeptide bands were labeled with eosin or fluorescein n-acetylimidazole in a Na+-sensitive manner. Two of these bands, previously identified as the glucose (75,000-dalton) and proline (100,000-dalton) binding sites of the glucose and proline carriers, account for 50% of the Na+-sensitive tyrosyl residues. On the basis of these studies, we believe that the Na+/glucose cotransporter contains both the Na+ and glucose active sites on the same polypeptide or that the cotransporter consists of two similar polypeptides, each containing one substrate binding site.     

Stimulation of intestinal Na+/d-glucose cotransport by monoclonal antibodies

Summary The small intestinal brush border membrane is endowed with a number of transport systems. Monoclonal antibodies were produced against integral membrane proteins and tested for their ability to bind to such membranes. For this purpose papain-digested, deoxycholate-extracted BBMVs from rabbit small intestine were used to immunize mice. Of the 765 hybridoma supernatants tested, 119 gave a significantly higher extent of binding to the crude antigen preparation as compared with the background. The monoclonal antibodies were also tested for their ability to influence the sodium-dependent uptake of solutes into intact BBMVs. Two monoclonal antibodies clearly showed stimulation of secondary actived-glucose transport, whereas sodium-dependent uptake ofl-alanine andl-proline was not affected. Hydrophobically labeled, i.e. intrinsic, membrane proteins of 175, 78 and 65 kilodaltons could be immunoprecipitated by both monoclonal antibodies, the 78 kDa band corresponding in all likelihood to the Na⁺/glucose cotransporter.     

Perturbation analysis of the voltage-sensitive conformational changes of the Na+/glucose cotransporter

Conformational changes of the human Na(+)/glucose cotransporter (hSGLT1) were studied using voltage-jump methods. The cotransporter was expressed in Xenopus laevis oocytes, and SGLT1 charge movements were measured in the micro- to millisecond time scale using the cut-open oocyte preparation and in the millisecond to second time scale using the two-electrode voltage clamp method. Simultaneous charge and fluorescence changes were studied using tetramethylrhodamine-6-maleimide-labeled hSGLT1 Q457C. In 100 mM external [Na(+)], depolarizing voltage steps evoked a charge movement that rose initially to a peak (with time constant tau = 0.17 ms) before decaying to steady state with two time constants (tau = 2-30 and 25-150 ms). The time to peak (0.9 ms) decreased with [Na(+)], and was not observed in 0 mM [Na(+)]. In absence of Na(+), charge movement decayed monotonically to steady state with three time constants (0.2, 2, and 150 ms). Charge movement was accompanied by fluorescence changes with similar time courses, indicating that global conformational changes monitored by charge movement are reflected by local environmental changes at or near Q457C. Our results indicate that the major voltage-dependent step of the Na(+)/glucose transport cycle is the return of the empty carrier from inward to outward facing conformations. Finally, we observed subtle differences between time constants for charge movement and for optical changes, suggesting that optical recordings can be used to monitor local conformational changes that underlie the global conformational changes of cotransporters.     

Expression and characterization of the intestinal Na+/glucose cotransporter in COS-7 cells

Cells derived from the simian kidney, COS-7 cells, were transfected with a eucaryotic expression vector (pEUK-C1) containing the clone for the rabbit intestinal Na+/glucose cotransporter. Expression was monitored after transfection with lipofectin by measuring the initial rate of alpha-methylglucopyranoside (MeGlc) uptake. Cells transfected with vector containing the cDNA for the Na+/glucose cotransporter expressed Na(+)-dependent MeGlc transport. Neither control cells nor cells transfected with vector lacking cloned cDNA expressed the cotransporter. Na(+)-dependent MeGlc uptake into transfected cells was saturable (Km 150 microM), phlorizin-sensitive (Ki 11 microM), and inhibited by sugar analogs (D-glucose greater than MeGlc greater than D-galactose greater than 3-O-methyl-D-glucoside greater than D-allose much greater than L-glucose). Europium was able to mimic Na+ in driving MeGIC uptake. Finally, tunicamycin, an inhibitor of asparagine-linked glycosylation, inhibited the expression of Na(+)-dependent MeGlc transport 80%. We conclude that the rabbit intestinal Na+/glucose cotransporter expressed in COS-7 cell exhibits very similar kinetic properties to that in the native brush border and to that expressed in Xenopus oocytes. In addition, N-linked glycosylation appears to be important for functional expression of this membrane protein.     

Disaccharidase-deficient animals have normal ultrastructure of intestinal brush border membranes

Summary The intestinal disaccharidases, lactase, sucrase-isomaltase complex, and glucoamylase are proteins intimately associated with the brush-border membrane of the epithelial cell. These three enzyme activities are found in the intestine of the adult rat; lactase and glucoamylase activities are primarily associated with the intestine of the infant rat. Only glucoamylase and isomaltase activities are detected in the intestine of the California sea lion, Zalophus californianus. The activities of these enzymes are detected only in villus cells, and not in crypt cells.We have carried out electron microscopic studies of negatively stained brush-border preparations of intestinal crypt and villus cells; from the intestine of the 10-day-old rat and from that of the California sea lion. The density of the knob-like structures protruding from the brush-border membranes was not significantly different in any of these preparations. The diameter of the knobs on the preparations from crypt cells was smaller than the diameters of the knobs found on membranes prepared from the other sources. These data are discussed in terms of the relationship between the presence of knob structures and disaccharidase activities associated with the brush-border membranes.     

Mechanism of interaction of the cyanine dye DiS−C3-(5) with renal brush-border vesicles

Summary The equilibrium binding mechanism and kinetics of binding of diS–C₃-(5) (3,3-dipropylthiodicarbocyanine iodide) to rabbit renal brush-border membrane vesicles (BBMV) were examined using steady-state and time-resolved fluorescence, and fluorescence stopped-flow methods. In aqueous solution, diS–C₃-(5) exists as a monomer at concentrations <5 m with fluorescence emission peak at 670 nm (excitation 622 nm), anisotropyr=0.102, and lifetime =1.2 nsec (23°C). Upon addition of increasing BBMV (voltage clamped to 0 mV using K⁺/valinomycin), the 670 nm emission peak decreases, corresponding to formation of a nonfluorescent membrane dimer, and subsequently a new emission peak at 695 nm increases, corresponding to membrane monomer. Dynamic depolarization studies show that aqueous diS–C₃-(5) rotation is unhindered with a rotational rateR=0.57 nsec^–1 while membrane monomer is hindered with steady-state anisotropyr=0.190, lifetime =2.1 nsec,R=0.58 nsec^–1 and limiting anisotropyr =0.11. Based on equilibrium fluorescence titrations, the membrane monomer-dimer (M-D) dissociation constant,K _d=[M]²/[D][BBMV], is 0.0013, where BBMV is expressed as membrane phospholipid concentration. Three distinct kinetic processes are identified by stopped-flow experiments in which BBMV are mixed with diS–C₃-(5). There is rapid binding of diS–C₃-(5) to the membrane to form bound monomer with a 6-msec exponential time constant. The membrane monomer at the membrane outer surface then aggregates to form bound dimer at the outer surface with a concentration independent time constant of 30 msec. The overall dimerization reaction probably consists of a rate-limiting reorientation process (30 msec) followed by a rapid dimerization which occurs on a nanosecond time scale. Finally, there is a 0.8 to 1 sec translocation of membrane dimer between symmetric sites at the inner and outer membrane surfaces. The translocation reaction is the step which is probably sensitive to changes in transmembrane electrical potential.     

Characterization of a Na+/glucose cotransporter cloned from rabbit small intestine

Summary The Na⁺/glucose cotransporter from rabbit intestinal brush border membranes has been cloned, sequenced, and expressed inXenopus oocytes. Injection of cloned RNA into oocytes increased Na⁺/sugar cotransport by three orders of magnitude. In this study, we have compared and contrasted the transport properties of this cloned protein expressed inXenopus oocytes with the native transporter present in rabbit intestinal brush borders. Initial rates of¹⁴C--methyl-d-glucopyranoside uptake into brush border membrane vesicles andXenopus oocytes were measured as a function of the external sodium, sugar, and phlorizin concentrations. Sugar uptake into oocytes and brush borders was Na⁺-dependent (Hill coefficient 1.5 and 1.7), phlorizin inhibitable (K _i 6 and 9 m), and saturable (-methyl-d-glucopyranosideK _m 110 and 570 m). The sugar specificity was examined by competition experiments, and in both cases the selectivity wasd-glucose>-methyl-d-glucopyranoside>d-galactose>3-O-methyl-d-glucoside. In view of the close similarity between the properties of the cloned protein expressed in oocytes and the native brush border transporter, we conclude that we have cloned the classical Na⁺/glucose cotransporter.     

Induction of Na+/K+/2Cl- cotransporter expression mediates chronic potentiation of intestinal epithelial Cl- secretion by EGF

Alterations in EGF receptor (EGFR) signaling occur in intestinal disorders associated with dysregulated epithelial transport. In the present study, we investigated a role for the EGFR in the chronic regulation of intestinal epithelial secretory function. Epithelial Cl(-) secretion was measured as changes in short-circuit current (Isc) across voltage-clamped monolayers of T84 cells in Ussing chambers. Acute treatment of T84 cells with EGF (100 ng/ml, 15 min) chronically enhanced Isc responses to a broad range of secretagogues. This effect was apparent within 3 h, maximal by 6 h, and sustained for 24 h after treatment with EGF. The Na+/K+/2Cl(-) cotransporter (NKCC1) inhibitor bumetanide (100 microM) abolished the effect of EGF, indicating increased responses are due to potentiated Cl(-) secretion. Neither basal nor agonist-stimulated levels of intracellular Ca2+ or PKA activity were altered by EGF, implying that the effects of the growth factor are not due to chronic alterations in levels of second messengers. EGF increased the expression of NKCC1 with a time course similar to that of its effects on Cl(-) secretion. This effect of EGF was maximal after 6 h, at which time NKCC1 expression in EGF-treated cells was 199.9 +/- 21.9% of that in control cells (n = 21, P < 0.005). EGF-induced NKCC1 expression was abolished by actinomycin D, and RT-PCR analysis demonstrated EGF increased expression of NKCC1 mRNA. These data increase our understanding of mechanisms regulating intestinal fluid and electrolyte transport and reveal a novel role for the EGFR in the chronic regulation of epithelial secretory capacity through upregulation of NKCC1 expression.     

cAMP-mediated regulation of murine intestinal/pancreatic Na+/HCO3- cotransporter subtype pNBC1

Basolateral Na(+)-HCO(3)(-) cotransport is essential for intestinal anion secretion, and indirect evidence suggests that it may be stimulated by a rise of intracellular cAMP. We therefore investigated the expression, activity, and regulation by cAMP of the Na(+)-HCO(3)(-) cotransporter isoforms NBC1 and NBCn1 in isolated murine colonic crypts. Na(+)-HCO(3)(-) transport rates were measured fluorometrically in BCECF-loaded crypts, and mRNA expression levels and localization were determined by semiquantitative PCR and in situ hybridization. Acid-activated Na(+)-HCO(3)(-) cotransport rates were 5.07 +/- 0.7 mM/min and increased by 62% after forskolin stimulation. NBC1 mRNA was more abundant in colonic crypts than in surface cells, and crypts expressed far more NBC1 than NBCn1. To investigate whether the cAMP-induced Na(+)-HCO(3)(-) cotransport activation was secondary to secretion-associated changes in HCO(3)(-) or cell volume, we measured potential forskolin-induced changes in intracellular pH and assessed Na(+)-HCO(3)(-) transport activity in CFTR -/- crypts (in which no forskolin-induced cell shrinkage occurs). We found 30% reduced Na(+)-HCO(3)(-) transport rates in CFTR -/- compared with CFTR +/+ crypts but similar Na(+)-HCO(3)(-) cotransport activation by forskolin. These studies establish the existence of an intracellular HCO(3)(-) concentration- and cell volume-independent activation of colonic NBC by an increase in intracellular cAMP.     

Investigating the conformational states of the rabbit Na+/glucose cotransporter

The Na(+) and voltage-dependence of transient rabbit Na(+)/glucose cotransporter (rSGLT1) kinetics was studied with the two-electrode voltage-clamp technique and Xenopus laevis oocytes. Using step changes in membrane potential, in the absence of glucose but with 100 or 10 mM Na(+), transient currents were measured corresponding to binding/debinding of Na(+) and conformational changes of the protein. Previously, only a single time constant has been published for rSGLT1. We, however, observed three decay components; a fast (tau(f), 0.5-1 ms) voltage- and Na(+)-independent decay, and medium (tau(m), 0.5-4 ms) and slow (tau(s), 8-50 ms) voltage- and Na(+)-dependent decays. Transient currents were simulated and fit using a four-state model to obtain kinetic parameters for the system. The four-state model was able to reconstitute an assortment of experimental data.     

On the stoichiometry of intercellular Na+/K(+)-exchange

Literature data on the atomic and ionic sizes of sodium (Na+) and potassium (K+), and the thermodynamic, and physicochemical characteristics of solutions containing Na+ and K+ have been compared. It was shown that in solutions containing Na+ only the activity of H+ is 1.31 +/- 0.038 times higher than in solutions containing only K+. It was concluded that there is reason to believe that the stoichiometry of Na+/K(+)-exchange is close to 3:4. The different degree of hydratation of Na+ and K+ caused by different density of their charges is considered to be the main reason for the result obtained.     

Acidic residues involved in cation and substrate interactions in the Na+/dicarboxylate cotransporter, NaDC-1.

The role of acidic amino acid residues in cation recognition and selectivity by the Na+/dicarboxylate cotransporter, NaDC-1, was investigated by site-directed mutagenesis and expression in Xenopus oocytes. Four of the residues tested, Asp-52, Glu-74, Glu-101, and Glu-332, were found to be unimportant for transport activity. However, substitutions of Asp-373 and Glu-475, conserved residues found in transmembrane domains M8 and M9, respectively, altered transport kinetics. Replacements of Asp-373 with Ala, Glu, Asn, and Gln resulted in changes in sodium affinity and cation selectivity in NaDC-1, indicating that the carbonyl oxygen at this position may play a role in the topological organization of the cation-binding site. In contrast, substitutions of Glu-475 led to dramatic reductions in transport activity and changes in transport kinetics. Substitution with Gln led to a transporter with increased substrate and sodium affinity, while the E475D mutant was inactive. The E475A mutant appeared to have poor sodium binding. Substrate-induced currents in the E475A mutant exhibited a strong voltage dependence, and a reversal of the current was seen at -30 mV. The results suggest that Glu-475 may play a role in cation binding and possibly also in mediating anion channel activity. Remarkably, mutations of both Asp-373 and Glu-475 affected the Km for succinate in NaDC-1, suggesting dual roles for these residues in determining the affinity for substrate and cations. We propose that at least one of the cation-binding sites and the substrate-binding site are close together in the carboxy-terminal portion of NaDC-1, and thus transmembrane domains M8 and M9 are candidate structures for the formation of the translocation pathway.