Monoclonal antibodies that bind the renal Na+/glucose symport system. 1. Identification

Phlorizin is a specific, high-affinity ligand that binds the active site of the Na+/glucose symporter by a Na+-dependent mechanism but is not itself transported across the membrane. We have isolated a panel of monoclonal antibodies that influence high-affinity, Na+-dependent phlorizin binding to pig renal brush border membranes. Antibodies were derived after immunization of mice either with highly purified renal brush border membranes or with apical membranes purified from LLC-PK1, a cell line of pig renal proximal tubule origin. Antibody 11A3D6, an IgG2b, reproducibly stimulated Na+-dependent phlorizin binding whereas antibody 18H10B12, an IgM, strongly inhibited specific binding. These effects were maximal after 30-min incubation and exhibited saturation at increased antibody concentrations. Antibodies did not affect Na+-dependent sugar uptake in vesicles but significantly prevented transport inhibition by bound phlorizin. Antibodies recognized a 75-kDa antigen identified by Western blot analysis of brush border membranes, and a 75-kDa membrane protein could be immunoprecipitated by 18H10B12. These properties, taken together with results in the following paper [Wu, J.-S.R., & Lever, J.E. (1987) Biochemistry (following paper in this issue)], provide compelling evidence that the 75-kDa antigen recognized by these antibodies is a component of the renal Na+/glucose symporter.     

Purification of a putative Na+/D-glucose cotransporter from pig kidney brush border membranes on a phlorizin affinity column

Phlorizin, a potent inhibitor of the Na+/D-glucose cotransporter, was derivatised to 3-aminophlorizin and subsequently coupled to Affi-Gel 15. Affinity chromatography of pig kidney brush border membranes solubilised in Triton X-100 allowed the purification of a 60 kDa protein on this resin. We consider this protein to be the Na+/D-glucose cotransporter, or part of it, for the following reasons: (i) binding of this protein to Affi-Gel 15 specifically requires phlorizin covalently attached to the resin and is lowered when phlorizin is replaced by phloretin; (ii) binding of the 60 kDa protein to a phlorizin affinity column requires the presence of Na+; (iii) polyclonal as well as monoclonal antibodies against the 60 kDa protein inhibit binding of phlorizin to brush border membranes from rabbit and pig kidney.     

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.     

Partial purification of the sugar carrier of intestinal brush border membranes. Enrichment of the phlorizin-binding component by selective extractions

Summary The [³H] phlorizin-binding component of brush border vesicles was enrichedin situ by negative purification. Several procedures, known to effect selective solubilization of membrane components, were used separately or in combination to remove proteins unrelated to the binding. Deoxycholate ruptured the vesicles and released 67% of their protein, thereby increasing the specific [³H] phlorizin-binding activity of the pellet three-to fourfold. Extracting the deoxycholate-pellets with either NaI or alkaline solutions released up to 38% of the deoxycholate-insoluble protein without significantly affecting phlorizin binding. The polypeptide composition of the membranes at the different stages was analyzed by NaDodSO₄-polyacrylamide gel electrophoresis. A number of polypeptides present in the original vesicles could be ruled out as essential components of the [³H] phlorizin binding entity.Intact and deoxycholate-treated vesicles were subjected to proteolytic attack. Papain liberated sucrase and isomaltase from intact vesicles, but affected neither other Coomassiestained bands nor phlorizin binding. Neither the protein composition nor the binding properties of sealed vesicles were influenced by trypsin or chymotrypsin. However, all the proteolytic enzymes tested on deoxycholate-treated membranes substantially reduced [³H] phlorizin binding and produced concomitantly the disappearance of several bands from the electrophoretic profile.Pretreatment of vesicles with papain, followed by deoxycholate extraction and incubation in alkaline media, increased the specific binding activity of the membranes up to ninefold by removing close to 90% of the protein. A limited number of polypeptides are suggested as possible candidates for the glycoside-binding site of intestinal brush borders.     

Dexamethasone blocks adaptive increase of Na+-Pi cotransport in renal brush border membrane elicited by thyroid hormone

Dexamethasone administered to rats blocks and/or reverses adaptive increases in the rate of Na+-Pi cotransport, and also in the Na+-dependent binding of [14C]-phosphonoformic acid (PFA) by renal brush border membrane (BBM) vesicles elicited by thyroid hormone (T3). In contrast, dexamethasone had no effect on Na+-independent binding of [14C]-phosphonoformic acid, on Na+-dependent transport of D-glucose or on Na+-dependent binding of phlorizin by BBMV which indicates that its inhibitory effect is specific for Na+-Pi cotransport system of BBM. These findings suggest that glucocorticoids antagonize T3-elicited adaptive enhancement of Na+-Pi cotransport in renal proximal tubules by blocking the T3-stimulated de novo synthesis of Na+-Pi symporters and/or their insertion into BBM.     

Comparison of glucose transport mechanisms at opposing surfaces of the renal proximal tubular cell

This review contrasts the glucose transport mechanisms at opposing surfaces of the renal proximal convoluted tubule: the Na+-dependent D-glucose transporter localized at the brush border membrane and the Na+-independent transporter localized at the basolateral surface. The two sugar transport mechanisms are discussed from the point of view of their specificity, kinetic, and regulatory behaviors. Recent results focussing on molecular characterization of these different carrier proteins are also described, including some newer information on purification of the Na+-dependent glucose carrier from the brush border membrane.     

The Na+ gradient-dependent transport of D-glucose in renal brush border membranes.

The Na+-dependent transport of D-glucose was studied in brush border membrane vesicles isolated from the rabbit renal cortex. The presence of a Na+ gradient between the external incubation medium and the intravesicular medium induced a marked stimulation of D-glucose uptake. Accumulation of the sugar in the vesicles reached a maximum and then decreased, indicating efflux. The final level of uptake of the sugar in the presence of the Na+ gradient was identical with that attained in the absence of the gradient, suggesting that equilibrium was established. At the peak of the overshoot the uptake of D-glucose was more than 10-fold the equilibrium value. These results suggest that the imposition of a large extravesicular to intravesicular gradient of Na+ effects the transient movement of D-glucose into renal brush border membranes against its concentration gradient. The stimulation of D-glucose uptake into the membranes was specific for Na+. The rate of uptake was enhanced with increased concentration of Na+. Increasing Na+ in the external medium lowered the apparent Km for D-glucose. The Na+ gradient effect on D-glucose transport was dissected into a stimulatory effect when Na+ and sugar were on the same side of the membrane (cis stimulation) and an inhibitory effect when Na+ and sugar were on opposite sides of the membrane (trans inhibition). The uptake of D-glucose, at a given concentration of sugar, reflected the sum of the contributions from a Na+-dependent transport system and a Na+-independent system. The relative stimulation of D-glucose uptake by Na+ decreased as the sugar concentration increased. It is suggested, however, that at physiological concentrations of D-glucose the asymmetry of Na+ across the brush border membrane might fully account for uphill D-glucose transport. The physiological significance of the findings is enhanced additionally by observations that the Na+-dependent D-glucose transport system in the membranes in vitro possessed the sugar specificities and higg phlorizin sensitivity characteristic of more intact preparations. These results provide strong experimental evidence for the role of Na+ in transporting D-glucose across the renal proximal tubule luminal membrane.     

gamma-Glutamyltranspeptidase activity in newborn rat kidney brush border.

gamma-Glutamyltranspeptidase, known to be localized in the proximal tubule cell brush border in the rat, is a membrane-bound enzyme which transfers the gamma-glutamyl moiety of glutathione or its analogue gamma-glutamyl-p-nitroanilide to an amino acid or dipeptide acceptor. Brush borders were isolated from the kidneys of newborn and adult Sprague-Dawley rats and assayed for gamma-glutamyltranspeptidase activity. There is an increase in specific activity in the brush border with maturation. Newborn and adult brush border preparations exhibit similar pH optima, substrate affinities, apparent Km values, patterns of heat inactivation, inhibition by glutathione, and migration on polyacrylamide gels. Polyacrylamide gel electrophoresis of a deoxycholate extract of brush border proteins and subsequent reaction with substrate within the gel reveal the presence of two bands, suggesting the presence of two forms of gamma-glutamyltranspeptidase in the rat kidney brush border.     

Characterization of rat kidney proximal tubule brush border membrane-associated phosphatidylinositol phosphodiesterase

Isolated rat kidney proximal tubule brush border membrane vesicles exhibit an increase in diacylglycerol levels (20- to 30-fold) and a concomitant decrease in phosphatidylinositol when incubated with [3H]arachidonate-labeled lipids, Ca2+, and deoxycholate. Levels of free arachidonate, triglyceride, and noninositol phospholipids are not altered. These results suggest phosphatidylinositol phosphodiesterase activity is associated with rat proximal tubule brush border membrane. Presence of both deoxycholate and certain divalent cations was necessary to demonstrate enzyme activity. Optimum pH ranged from 7.0 to 8.5. Ca2+, Mg2+, and Mn2+ stimulated diglyceride production while Ba2+, Zn2+, Hg2+, and K+ were ineffective. HgCl2 inhibited Ca2+-stimulated phosphatidylinositol phosphodiesterase. Mg2+ and deoxycholate-dependent enzyme activity was shown to be phosphatidylinositol specific. Sodium lauryl sulfate, tetradecyltrimethylammonium bromide, and Triton X-100 did not activate phosphatidylinositol phosphodiesterase in the presence of Ca2+. In combination with deoxycholate, diglyceride formation was not affected by sodium lauryl sulfate, partially inhibited by Triton X-100, and completely abolished by tetradecyltrimethylammonium bromide. Diglyceride kinase activity was not found associated with brush border membrane phosphatidylinositol phosphodiesterase. ATP (1-5 mM) inhibited Ca2+- or Mg2+-stimulated, deoxycholate-dependent phosphatidylinositol hydrolysis by chelating the required divalent cation.     

Sugar uptake into a primary culture of dog kidney proximal tubular cells

We describe the functional characteristics of a new primary culture system derived from a suspension of dog proximal tubular cells. The culture system is maintained on alpha minimal essential medium with 15% fetal calf serum supplementation. At confluency the cultured cells demonstrate the following: (i) typical epithelial morphology using light microscopy, with multiple dome formation inhibited by ouabain; (ii) strong binding with a polyclonal antibody directed against dog proximal tubular brush border membrane antigens; (iii) high concentration of alkaline phosphatase activity by histochemical staining; and (iv) 25-hydroxycholecalciferol (25(OH)D3)-24-hydroxylase activity. Sugar transport was assessed using alpha-methyl-D-glucopyranoside (alpha MG), a nonmetabolizable analog of D-glucose, as well as L-glucose, and 3-O-methyl-D-glucose (3OMG). The transport of alpha MG was stereospecific; temperature sensitive; inhibited strongly by phlorizin but not by cytochalasin B, phloretin, or 3OMG; and Na dependent. The transport of 3OMG is stereospecific, temperature sensitive, and inhibited strongly by phloretin and cytochalasin B, but not by phlorizin. Despite the apparent heterogeneity of cell type, this primary culture system exhibits many features of normal dog proximal tubule function.     

Monoclonal antibodies that bind the renal Na+/glucose symport system. 2. Stabilization of an active conformation

Conformation-dependent fluorescein isothiocyanate (FITC) labeling of the pig renal Na+/glucose symporter was investigated with specific monoclonal antibodies (MAb's). When renal brush border membranes were pretreated with phenyl isothiocyanate (PITC), washed, and then treated at neutral pH with FITC in the presence of transporter substrates Na+ and glucose, most of the incorporated fluorescence was associated with a single peak after resolution by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The apparent molecular mass of the FITC-labeled species ranged from 79 to 92 kDa. Labeling of this peak was specifically reduced by 70% if Na+ and glucose were omitted. Na+ could not be replaced by K+, Rb+, or Li+. FITC labeling of this peak was also stimulated after incubation of membranes with MAb's known to influence high-affinity phlorizin binding, and stimulation was synergistically increased when MAb's were added in the presence of Na+ and glucose. Substrate-induced or MAb-induced labeling correlated with inactivation of Na+-dependent phlorizin binding. MAb's recognized an antigen of 75 kDa in the native membranes whereas substrate-induced FITC labeling was accompanied by loss of antigen recognition and protection from proteolysis. These findings are consistent with a model in which MAb's stabilize a Na+-induced active conformer of the Na+/glucose symport system.     

Partial purification and reconstitution of the Na+-D-glucose cotransport protein from pig renal proximal tubules

Brush border membranes from renal proximal tubules were solubilized with deoxycholate, and the proteins were incorporated into liposomes formed from cholesterol and phosphatidylserine by a freeze-thaw procedure. In the proteoliposomes Na+-D-glucose cotransport was demonstrated by showing that the D-glucose concentration in the liposomes increased far above the equilibrium value if a Na+ gradient was applied. The initial D-glucose uptake rate, stimulated by an inside directed gradient of 89 mM Na+, was 4 pmol/mg of protein-1 s-1. High affinity phlorizin binding could not be measured. After two precipitation steps with the solubilized membrane proteins, a protein fraction was obtained in which significantly high affinity phlorizin binding was detected. After reconstitution, proteoliposomes were formed in which more than 70% of the protein was represented by two polypeptides with molecular weights of 94,000 and 52,000. An initial Na+ gradient-dependent D-glucose uptake rate of 118 pmol/mg of protein-1 s-1 was obtained. In these liposomes, the D-glucose uptake rate could be inhibited by phlorizin (Ki = 0.3 microM), and 55-pmol phlorizin-binding sites per mg of protein (KD = 0.5 microM) were measured. In different liposomal preparations a correlation between Na+ gradient-dependent D-glucose uptake rate and the amount of 52,000 molecular weight polypeptide was observed.     

Purification and reconstitution of a 75-kilodalton protein identified as a component of the renal Na+/glucose symporter

A 75-kilodalton (kDa) protein was purified from solubilized renal brush border membranes by using high-pressure liquid chromatography (HPLC) and preparative sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Functional and immunological properties identified the 75-kDa protein as a component of the Na+/glucose symport system. The purified protein was specifically recognized by a monoclonal antibody that functionally interacts with the Na+/glucose symporter. Na+-dependent phlorizin binding activity was associated with fractions containing the 75-kDa protein during HPLC fractionation on the anion exchanger Mono-Q and was greatly increased after reconstitution into egg yolk phosphatidylcholine vesicles. The final purified preparation contained glucosamine and a blocked N-terminus.     

Insulin regulates Na+/glucose cotransporter activity in rat small intestine

In order to examine the involvement of insulin in the activity of Na+/glucose cotransporter in rat small intestine, we compared Na(+)-dependent uptake of D-glucose by brush-border membrane vesicles prepared from control, streptozotocin-induced diabetic, insulin-treated diabetic and starved diabetic rats. In four groups, the uptake of D-glucose showed a transient overshoot in the presence of Na+ gradient between medium and vesicles (medium greater than vesicles). The overshoot magnitude was increased (1.8-fold of controls) in diabetic brush border membrane vesicles and recovered to the control level by the treatment of diabetic rats with insulin. In contrast, increased uptake of D-glucose in diabetic rats was not recovered by the starvation of diabetic rats although the blood glucose level was the same as that of controls. Furthermore, we attempted to examine phlorizin binding activities among four groups. Scatchard analysis indicated that phlorizin binding to diabetic brush border membrane vesicles was increased (1.6-fold of controls) without a change of the affinity for phlorizin as compared with controls. Increased binding of phlorizin to diabetic brush border membrane vesicles was also recovered to the control level by the treatment of diabetic rats with insulin, but not by starvation. These results suggested that the increased activity of Na+/glucose cotransporter in diabetic rats was due to the increase of the number of cotransporter and that intestinal cotransporter was physiologically controlled by insulin, but not by blood glucose levels.     

Interaction of the sugar carrier of intestinal brush-border membranes with HgCl2

HgCl2 was used as an inhibitor and potential label for the glucose carrier of intestinal brush-border membranes. Half-maximal inhibition of Na+-dependent D-glucose uptake was reached with micromolar concentrations of HgCl2 when the protein concentration was 1.2 mg/ml. Similar concentrations were found to inhibit the binding of [3H]phlorizin, a reversible competitive inhibitor of sugar transport. Inhibition was reversed by dithioerythritol but only marginally by EDTA. The data support the involvement of a sulfhydryl group in the inhibitory process. Deoxycholate-extracted membranes, which are enriched in specific phlorizin binding activity, were used for labeling studies using 203HgCl2. The polypeptides were separated by gel electrophoresis and analyzed by protein staining and autoradiography. Non-specific 203HgCl2 labeling was minimized by pre-treatment with sulfhydryl reagents which do not inhibit phlorizin binding. Several bands, which are lost from the autoradiographic pattern during a negative purification of the phlorizin binding sites, could be ruled out as essential components of the sugar carrier. The polypeptide profile was also analyzed following proteolysis, which abolished phlorizin binding. Those radioactive bands of which apparent Mr values were alterd by the treatment were considered as possible candidates. Finally, samples in which inhibition was reversed by thiols were also studied. The possible identity of the polypeptide(s) involved in glucose translocation is disussed in the light of these observations.     

Further studies of proximal tubular brush border membraned-glucose transport heterogeneity

Summary The properties of two sodium-dependentd-glucose transporters previously identified in renal proximal tubule brush border membrane (BBM) vesicles are studied. The low-affinity system, found in BBM vesicles from the outer cortex (early proximal tubule), is shown to be associated with the high-affinity phlorizin binding site typically found in renal BBM preparations. The high-affinity system, found in BBM vesicles from the outer medulla (late proximal tubule), is almost two orders of magnitude less sensitive to inhibition by phlorizin and is apparently not associated with high-affinity phlorizin binding. The sodium/g;ucose stoichiometry of the outer medullary transporter is found to be 21 by two independent methods. Previous measurements have established that the stoichiometry of the outer cortical system is 11. It is suggested that this arrangement of transporters in series along the proximal tubule enables the kidney to reabsorb glucose from the urine in an energy-efficient fashion. The bulk of the glucose load is reabsorbed early in the proximal tubule at an energetic cost of one Na⁺ per glucose molecule. Then in the late proximal tubule a larger coupling ratio and hence a larger driving force is employed to reabsorb the last traces of glucose from the urine.     

Reconstitution of the partially purified renal phosphate (Pi) transporter.

Proteins from rabbit kidney brush border membranes were solubilized with 1% Nonidet P-40 (crude membrane proteins) and fractionated according to their isoelectric points (pI) by chromatofocusing. The eluate was pooled into three fractions according to the pI of the samples (1, greater than 6.8; 2, 6.8-5.4; 3, 5.4-4.0). The crude membrane proteins as well as the three fractions were reconstituted into liposomes and transport of Pi was measured by a rapid filtration technique in the presence of an inwardly directed K+ or Na+ gradient. Arsenate-inhibitable Na+-dependent transport of Pi was reconstituted into an osmotically active intravesicular space from both the crude membrane proteins and Fraction 1. In contrast, Fractions 2 and 3 were inactive. Treatment of the crude membrane proteins and the three fractions with the method for extracting phosphorin (a Pi-binding proteolipid found in brush border membranes) yielded Mn2+-dependent binding of Pi characteristic of phosphorin only in the extracts from crude membrane proteins and Fraction 1, the same fractions in which Na+-dependent transport of Pi was found in the reconstituted system. When reconstituted into liposomes, phosphorin was, however, unable to yield Na+-dependent transport of Pi. Moreover, we cannot eliminate the possibility that Na+-Pi transport can occur in the absence of phosphorin, since complete recovery of Na+-Pi transport was not achieved. However, the present data showing localization of the recovered binding and transport systems for Pi in the same protein fraction lend support to the hypothesis that phosphorin might be a constituent of the renal Pi transport system. Whether the presence of phosphorin is necessary or accessory for Na+-dependent Pi transport in intact brush border membrane vesicles or in liposomes reconstituted with crude or purified membrane proteins requires further investigation.     

Energetics of the Na+-dependent transport of D-glucose in renal brush border membrane vesicles.

The energetics of the Na+-dependent transport of D-glucose into osmotically active membrane vesicles, derived from the brush borders of the rabbit renal proximal tubule, was studied by determining how alterations in the electrochemical potential of the membrane induced by anions, ionophores, and a proton conductor affect the uptake of the sugar. The imposition of a large NaCl gradient (medium is greater than vesicle) resulted in the transient uptake of D-glucose into brush border membranes against its concentration gradient. In the presence of Na+ salts of isethionate or sulfate, both relatively impermeable anions, there was no accumulation of D-glucose above the equilibrium value. With Na+ salts of two highly permeable lipophilic anions, NO3- and SCN-, the transient overshoot was enhanced relative to that with Cl-. With Na+ salts whose mode of membrane translocation is electroneutral, i.e. acetate, bicarbonate, and phosphate, no overshoot was found. These findings suggest that only anions which penetrate the brush border membrane and generate an electrochemical potential, negative on the inside, permit the uphill Na+-dependent transport of D-glucose.     

High-affinity phlorizin binding in Mytilus gill.

The gill of the marine mussel, Mytilus, contains a high affinity, Na-dependent D-glucose transporter capable of accumulating glucose directly from sea water. We examined the ability of the beta-glucoside, phlorizin, to act as a high-affinity ligand of this process in intact gills and isolated brush border membrane vesicles (BBMV). The time course of association of nanomolar [3H]phlorizin to gills and BBMV was slow, with t50 values between 10 and 30 min, and a half-time for dissociation of approx. 30 min. 1 mM D-glucose reduced equilibrium binding of 1 nM phlorizin by 90-95%, indicating that there was little non-specific binding of this ligand to the gill. In addition, there was little, if any, hydrolysis by the gill of phlorizin to its constituents, glucose and phloretin. Phlorizin binding to gills and BBMV was significantly inhibited by the addition of 50 microM concentrations of D-glucose and alpha-methyl-D-glucose, and unaffected by the addition of L-glucose and fructose. Binding to gills and BBMV was reduced by greater than 90% when Na+ was replaced by K+. Replacement of Na+ by Li+ effectively blocked binding to the intact gill, although Li+ did support a limited amount of glucose-specific phlorizin binding in BBMV. The Kd values for glucose-specific phlorizin binding in intact gills and BBMV were 0.5 nM and 6 nM, respectively. We conclude that phlorizin binds with extremely high affinity to the Na-dependent glucose transporter of Mytilus gill, which may be useful in future efforts to isolate and purify the protein(s) involved in integumental glucose transport.     

Electron microscopic immunohistochemical localization of components of Na+-cotransporters along the rat nephron

The localization of Na+-cotransport proteins in cortex and outer medulla of rat kidney was investigated with five monoclonal antibodies. Recently, it was found that these antibodies altered Na+-D-glucose cotransport and/or Na+-dependent high affinity phlorizin binding in pig kidney cortex and that three of these antibodies interacted also with Na+-cotransporters for lactate, L-alanine and/or L-glutamate (Koepsell, H., K. Korn, A. Raszeja-Specht, S. Bernotat-Danielowski, D. Ollig, J. Biol. Chem. 263, 18,419-18,429 (1988]. In pig and rat the monoclonal antibodies bind to two brush-border membrane polypeptides with identical molecular weights and isoelectric points of 75,000 and pI 5.5, and 47,000 and pI 5.4. These polypeptides have been previously identified as components of the porcine renal Na+-D-glucose cotransporter (Neeb, M., U. Kunz, H. Koepsell, J. Biol. Chem. 262, 10,718-10,727 (1987] and may also be part of other Na+-cotransporters. The electron microscopic localization of antibody binding was demonstrated by protein A-gold labeling on ultrathin plastic sections. Three antibodies bound to brush-border membranes of proximal convoluted and straight tubules. In the proximal convoluted tubules all antibodies reacted with apical endocytic vacuoles, apical dense tubules and lysosomes. Since dense tubules are supposed to originate from endocytic vacuoles and to fuse with brush-border membranes the data suggest recycling of Na+-cotransporters in the proximal convoluted tubule. In the outer medulla two antibodies bound to apical membranes of descending thin limbs (DTL) of short loops of Henle and to apical and basal membranes of DTL of long loops of Henle. Three antibodies bound to apical membranes of collecting ducts. These data indicate that Na+-cotransporters or homologous proteins exist beyond the proximal tubule.