Abstract

Urea transport in the kidney is mediated by a family of transporter proteins that include the renal urea transporter (UT-A) and the erythrocyte urea transporter (UT-B). The purpose of this study was to determine the location of the urea transporter isoforms in the mouse kidney and to examine the effects of prolonged potassium depletion on the expression and distribution of these transporters by ultrastructural immunocytochemistry. C57BL6 mice were fed a low-potassium diet for 2 wk, and control animals received normal chow. After 2 wk on a low-potassium diet, urinary volume increased and urinary osmolality decreased (833 ± 30 vs. 1,919 ± 174 mosmol/kgH2O), as previously demonstrated. Kidneys were processed for immunocytochemistry with antibodies against UT-A1 (L446), UT-A1 and UT-A2 (L194), UT-A3 (Q2), and UT-B. In normal mice, UT-A1 and UT-A3 were expressed mainly in the cytoplasm of the terminal inner medullary collecting duct (IMCD). UT-A2 immunoreactivity was observed mainly on the basolateral membrane of the type 1 epithelium of the descending thin limb (DTL) of short-looped nephrons. The intensity of UT-A1 and UT-A3 immunoreactivity in the IMCD was markedly reduced in potassium-depleted mice. In contrast, there was a significant increase in UT-A2 immunoreactivity in the DTL. The intensity of UT-B immunoreactivity in the descending vasa recta (DVR) was reduced in potassium-depleted animals compared with controls. In control animals, UT-B immunoreactivity was predominantly observed in the plasma membrane, whereas in potassium-depleted mice, it was mainly observed in cytoplasmic granules in endothelial cells of the DVR. In summary, potassium depletion is associated with reduced expression of UT-A1, UT-A3, and UT-B but increased expression of UT-A2. We conclude that reduced expression of urea transporters may play a role in the impaired urine-concentrating ability associated with potassium deprivation.

  • immunogold labeling
  • descending thin limb of loop of Henle
  • urine concentration

prolonged hypokalemia causes profound alterations in renal structure (22) and is associated with renal functional abnormalities, including an impairment of the urinary concentrating ability in humans (31, 32) and experimental animals, including the rat (3, 21) and rabbit (29). The mechanisms by which hypokalemia causes a urine-concentrating defect are incompletely understood. Several mechanisms have been proposed to explain the defect, including reduced expression of inner medullary aquaporin (AQP)-2 (1, 21), blunted responsiveness of the collecting duct epithelium to anti-diuretic hormone (ADH; vasopressin) due to intrarenal prostaglandin overproduction (12, 30) or impairment of ADH-sensitive adenylate cyclase (3), altered ADH release by the pituitary gland (33), and down-regulation of major renal Na+ transporters (8).

Potassium depletion is also accompanied by severe structural changes, such as hyperplasia, hypertrophy, and the accumulation of cytoplasmic granules in medullary collecting duct cells and in medullary endothelial cells (36, 39). The droplets are believed to constitute lysosomal structures and are positive for AQP2 in hypokalemia (21). Therefore, they may represent sites of AQP2 degradation and may be involved in the decrease in AQP2 levels observed in potassium depletion.

Urea is an important contributor to the processes of urinary concentration. Physiological data provide evidence that urea transport in the kidney and red blood cells is mediated by specific urea transporter proteins, the renal urea transporter (UT-A) and the erythrocyte urea transporter (UT-B) (34). These are encoded by two genes, Slc14a2 (UT-A) and Slc14a1 (UT-B), which occur in tandem on chromosome 18 (10). The UT-A gene Slc14a2 was initially cloned from the rat (24) and subsequently from humans (2). The UT-B gene Slc14a1 was cloned from a human erythropoietic cell line (26) and from the rat (6). The cDNAs of four isoforms of rat UT-A, i.e., UT-A1, UT-A2, UT-A3, and UT-A4, have been cloned as has a fifth isoform from the mouse, UT-A5. They are thought to be mRNA variants produced by alternative splicing from a single gene (23). The cDNA of UT-B has been cloned from a human bone marrow library (26) and subsequently isolated from a rat kidney inner medullary library by homology screening (6).

In the rat, UT-A1–4 is predominantly expressed in the kidney. UT-A1 and UT-A3 are localized in the inner medullary collecting duct (IMCD; 25, 37), and UT-A2 is localized in the descending thin limb (DTL) of Henle's loop (37, 45). The precise segmental distribution of UT-A4 has not been elucidated, and indeed the existence of UT-A4 protein has not been unequivocally demonstrated in kidney. UT-B mRNA and protein are expressed in the endothelial cells of the vasa recta (26, 42, 48). In a study of the subcellular localization of UT-A in the IMCD of the rat, Nielsen and colleagues (25) demonstrated labeling of both the apical plasma membrane and intracellular vesicles of IMCD cells by immunoelectron microscopy, suggesting the possibility that UT-A1 may undergo intracellular trafficking. However, Inoue and colleagues (14) showed that regulated trafficking of UT-A1 does not occur in the rat IMCD. Studies by Kim et al. (17) using a horseradish peroxidase technique provided evidence that UT-A and UT-B are expressed in both the apical and basolateral plasma membrane of the DTL and descending vasa recta (DVR), respectively. However, it is not known whether UT-A2 and UT-B are located in intracellular vesicles.

In a recent study of the mouse kidney, Fenton and colleagues (11), using an antibody raised to the 19 COOH-terminal amino acids of rat UT-A1 (L194), the COOH terminus of which is identical to that of UT-A2 (11, 25), demonstrated strong staining in the type 1 and type 3 epithelium of the DTL and in the middle and terminal IMCD. However, there are no reports of the subcellular localization of either UT-A or UT-B in the mouse kidney. Moreover, it remains to be established whether potassium depletion is associated with changes in the expression of these transporters in the kidney.

The purpose of this study was to determine the exact location of urea transporter isoforms in the mouse kidney and to examine the effects of prolonged potassium depletion on the expression and distribution of these transporters at the subcellular level using ultrastructural immunocytochemistry with antibodies against UT-A, L446 (UT-A1), L194 (UT-A1, UT-A2, and UT-A4), Q2 (UT-A3), and antiserum directed against UT-B.

MATERIALS AND METHODS

Animals and Tissue Preservation

A total of 36 male C57BL6 mice (Samtako, Seoul, Korea) were used in these experiments, and they all had free access to food and water. The animals were divided into two groups: the first group was fed a normal diet (control) and the second group a potassium-depleted diet (Harlan Teklad, Madison, WI) for 2 wk. The two diets were the same except for the potassium content. Animals were anesthetized with an intraperitoneal injection of pentobarbital sodium (50 mg/kg body wt). The kidneys were first perfused briefly through the abdominal aorta with PBS (pH 7.4) to rinse out all blood. For Western blot analysis, the right kidneys of animals were excised after the renal artery was clamped. For immunohistochemical studies, the left kidneys were then perfused with fixative solution, periodate-lysine-paraformaldehyde for light microscopy or 8% paraformaldehyde for ultrastructural studies. The kidneys were removed and cut transversely into 1- to 2-mm-thick slices, which were immersed in the same fixative overnight at 4°C.

Antibodies

The following published and characterized antibodies were used for these studies: rabbit L194 (UT-A1 and UT-A2; 25); rabbit L446 (UT-A1; 45); rabbit Q2695–2 (UT-A3; 41); and rabbit UT-B (42). The DTL of the long loop of Henle was identified using a rabbit polyclonal antibody to AQP1 (Chemicon, Temecula, CA). The thick ascending limb (TAL) was identified using a rabbit polyclonal antibody directed against Na-K-ATPase-α1 (Chemicon). The antibodies have been characterized in detail in previous studies (47).

Immunoperoxidase Preembedding Method

Vibratome sections (50 μm thick) were washed three times for 15 min with 50 mM NH4Cl in PBS. Before incubation with the primary antibodies, all tissue sections were incubated for 3 h with PBS containing 1% BSA, 0.05% saponin, and 0.2% gelatin (solution A). The tissue sections were then incubated overnight at 4°C in rabbit antibodies against UT-A (L194, L446, Q2) and UT-B in PBS containing 1% BSA (solution B). After several washes with solution A, the tissue sections were incubated for 2 h with peroxidase-conjugated donkey anti-rabbit IgG Fab fragments (Jackson ImmunoResearch Laboratories, West Grove, PA) diluted 1:100 in solution B. The tissues were then rinsed, first in solution A, and then in 0.05 M Tris buffer (pH 7.6). To detect horseradish peroxidase, the sections were incubated in 0.1% 3,3′-diaminobenzidine (DAB) in 0.05 M Tris buffer for 5 min, after which H2O2 was added to a final concentration of 0.01% and the incubation was continued for 10 min. The sections were washed three times with 0.05 M Tris buffer, dehydrated in a graded series of ethanol, and embedded in Epon 812 resin (Polysciences, Warrington, CA). For UT-A and AQP1 or UT-A and Na-K-ATPase-α1 double immunostaining, vibratome sections were labeled with L194 (1:1,000) using DAB as the chromogen (brown), as described above. The sections were then rinsed with PBS, incubated for 1 h in solution B, and then incubated overnight at 4°C in rabbit antibodies against AQP1 (1:500) or Na-K-ATPase-α1 in solution B. After several washes with PBS, the tissue sections were incubated for 2 h in horseradish peroxidase-conjugated donkey anti-rabbit IgG Fab fragments (Jackson ImmunoResearch Laboratories) diluted 1:100 in solution B. The tissues were then rinsed in PBS (pH 7.4), and, to detect horseradish peroxidase, the sections were incubated in vector SG (blue; Vector Laboratories, Burlingame, CA) in PBS for 5 min, followed by incubation with H2O2 for 10 min. Sections were washed three times with PBS, dehydrated in a graded series of ethanol, and embedded in PolyBed 812 resin. The sections were examined with an Olympus photomicroscope.

Immunogold Labeling for Ultrastructural Studies

Fifty-micrometer vibratome sections were processed as described above, using the preembedding method and primary antibodies. Tissue sections were incubated overnight at 4°C with L194 (1:1,000), Q2 (1:300), or UT-B (1:1,000) antibody diluted in 1% BSA-PBS. Sections were incubated in gold buffer (2% gelatin, 0.5% BSA, 0.2% sodium azide) at room temperature. Sections were then incubated overnight with 1-nm gold-conjugated donkey anti-rabbit IgG (Amersham Pharmacia Biotech, Buckinghamshire, UK) diluted 1:50 in gold buffer. The sections were then rinsed in gold buffer and washed in distilled water. To detect 1-nm gold particles, sections were treated with an Intense M Silver Enhancement Kit (Amersham Pharmacia Biotech) for 8–10 min. Sections were washed with distilled water and then fixed for 1 h in 1% glutaraldehyde at 4°C. Tissue sections were washed with 0.1 M phosphate buffer and then postfixed in 1% osmium tetroxide in 0.1 M phosphate buffer for 1 h at 4°C. After a final wash with 0.1 M phosphate buffer, the sections were dehydrated in a graded series of ethanol and embedded in Epon 812. Vibratome sections (50 μm) through the entire kidney, from each animal, were mounted in Epon 812 between polyethylene vinyl sheets. Sections from the cortex, outer medulla (OM), and inner medulla (IM) were excised and glued onto empty blocks of Epon 812. After examination of 1-μm-semithin sections by light microscopy, ultrathin sections were cut and photographed with a JEOL 1200EX (Tokyo, Japan) transmission electron microscope.

Semiquantitative Immunoblotting

For immunoblotting analysis, a kidney from each mouse was dissected into cortex, OM, and IM. The tissues were homogenized in lysis buffer containing 0.3 M sucrose, 25 mM imidazole, and 1 mM EDTA, using a homogenizer (Biospec Products, Bartlesville, UK). The homogenates were centrifuged at 4,000 g for 20 min at 4°C to remove cell debris. The supernatant was centrifuged at 200,000 g for 1 h (Beckman ultracentrifuge) to separate a membrane fraction. Protein concentrations were determined with a BCA Protein Assay Reagent Kit (Pierce, Rockford, IL), and samples were dissolved in Laemmli buffer and then heated to 60°C for 15 min. Samples were separated by SDS-polyacrylamide gel electrophoresis (Bio-Rad, Hercules, CA). Proteins were then transferred to nitrocellulose membranes by electroblotting. To reduce nonspecific antibody binding, the membranes were blocked with 5% nonfat dried milk for 30 min at room temperature and then incubated for 24 h at 4°C with affinity-purified antibodies, as described above. The membranes were then washed in several changes of blotting buffer [0.01 M PBS (pH 7.4) and 0.1% Tween 20] and incubated for 1 h with peroxidase-labeled donkey anti-rabbit IgG (1:1,000). Samples were visualized after a 5- to 30-min exposure at room temperature to enhanced chemiluminescence (Amersham Life Sciences, Buckinghamshire, UK). Densitometric analysis was performed using the Zero-Dscan software of the Eagle Eye II Still Video System (Stratagene, La Jolla, CA). Optical density was calculated as means ± SD of three readings for each band.

Functional Data and Statistical Analysis

Mice were kept in metabolic cages (Techniplast, Buguggiate, Italy) for at least 48 h before starting the low-potassium diet. Urine samples were measured over 24 h during this period. Blood samples were also collected before the mice were killed. Serum and urine potassium, sodium, and chloride were analyzed on an Ektachem 400 device (Eastman Kodak, Rochester, NY).

Results are presented as means ± SE, and all statistical analyses were performed with Microsoft Excel for Windows 98. Comparisons between groups were made using unpaired t-tests. P values <0.05 were considered significant.

RESULTS

Functional Data

Table 1 shows the functional parameters of the control and experimental groups. The potassium-depleted groups showed a significant decrease in serum potassium levels (2.9 ± 0.2 vs. 4.6 ± 1.1 mM) and urinary osmolality (834 ± 30 vs. 1,919 ± 174 mosmol/kgH2O) compared with the control groups, respectively. There were no significant differences in serum sodium or serum chloride between the control and experimental groups.

View this table:
Table 1.

Balance, blood, and urinary data

Light Microscopic Immunohistochemistry

Expression of UT-A. control mouse kidney. Light microscopy of 50-μm sections demonstrated UT-A immunolocalization with antibody L194 in the inner stripe of the outer medulla (ISOM) and IM. There was no labeling in the cortex, outer stripe of the outer medulla (OSOM), or initial part of the IM (Fig. 1A). To test whether each subtype of UT-A is expressed in mouse kidney, we used immunohistochemistry with antibodies L194 (UT-A1 and UT-A2), L446 (UT-A1), and Q2 (UT-A3). UT-A1 (L446) immunoreactivity was observed in the IMCD of the middle and terminal IM. There was no labeling in the cortex, OM, or the initial part of the IM (Fig. 1B). UT-A2 (L194) immunoreactivity was observed exclusively in the ISOM (Fig. 1A). UT-A3 (Q2) immunoreactivity was observed only in the middle and terminal parts of the IM (Fig. 1C).

Fig. 1.

Light micrographs of 50-μm-thick vibratome sections of the mouse kidneys from control animals illustrating single immunostaining for L194 (A and D), L446 (B), and Q2 (C) and double immunostaining for Na-K-ATPase-α1 (blue) and L194 (brown; EG). A: urea transporter (UT)-A1 and UT-A2, detected as immunoreactivity to L194, were present in the inner stripe of the outer medulla (ISOM) and the middle and terminal part of inner medulla (IM). No immunoreactivity to L194 was observed in the cortex (C). B and C: UT-A1 and UT-A3, detected as immunoreactivity to L446 and Q2, respectively, appeared only in the IMCD in the renal papilla. DG: higher magnification of ISOM (DF), the initial part of IM (IMi), and the border between the outer medulla (OM) and IM (G), respectively. D: single immunostaining for L194. Note that L194 immunolabeling in the ISOM is localized in the middle part of the ISOM. F: higher magnification of a part of middle region of ISOM in E. Arrows indicate the abrupt transition from strong UT-A2-positive descending thin limbs (DTLs; brown) to Na-K-ATPase-α1-positive thick ascending limbs (TAL; blue). In the rectangle, the open arrow indicates the junction of the DTL and TAL. G: UT-A2 is also expressed in the some DTLs (arrowheads) located at the innermost part of ISOM. Note that these UT-A2-positive DTLs were connected to UT-A2-negative DTLs in the IMi. Magnification: AC, ×15; D and E, ×60; F and G, ×140.

At higher magnification of the ISOM in 50-μm-thick vibratome sections with immunostaining for UT-A2, most of the strongly labeled tubular profiles were located in the middle part of the ISOM, and a few labeled tubular profiles were sparsely distributed in the innermost part of the ISOM (Fig. 1D). However, there were no immunolabeled thin limbs of Henle in the IM. To determine where UT-A2 is expressed in the ISOM, double-labeling experiments were performed using antibodies directed against UT-A (L194) and Na-K-ATPase-α1. Na-K-ATPase-α1 immunoreactivity is observed in the cells of the TAL and the distal convoluted tubules, as demonstrated previously (47). The strongly labeled tubular profiles located in the middle portion of the ISOM were identified as the terminal portion of the short-loop DTL by the abrupt transition to the Na-K-ATPase-α1-positive TALs, which usually bend at this level (Fig. 1, E and F). In the innermost part of the ISOM, the weakly labeled thin limbs were not connected to Na-K-ATPase-α1-positive TALs but to UT-A-negative thin limbs at various levels proximal to the border between the OM and IM (Fig. 1G).

potassium-depleted mouse kidney. Figure 2, A and A, shows representative immunolabeling of the medulla from a control mouse and a potassium-depleted mouse, respectively, using antibody L194 (for UT-A1 and UT-A2). In potassium-depleted mice, the intensity of UT-A2 immunoreactivity in the ISOM and the initial part of the IM was significantly increased (Fig. 2A). In contrast, there was a marked decrease in UT-A1 immunoreactivity in the IM (Fig. 2A). In both the innermost part of the ISOM and initial part of the IM, the number of UT-A-positive thin limbs was increased (Fig. 2B). The UT-A-positive thin limbs did not show AQP1 immunoreactivity, a marker of DTLs of long loop nephrons, whereas UT-A-negative thin limbs expressed AQP1 (Fig. 2, C and C). Similar results were obtained in immunohistochemical studies with antibodies specific for the different UT-A subtypes. Immunoreactivity for L446 (UT-A1) was reduced in the IMCD (Fig. 2D) compared with the control (Fig. 2D). A striking increase in the number of immunolabeled DTLs and in the intensity of immunostaining with L194 (UT-A2) was observed both in the innermost part of the ISOM and the initial part of the IM (Fig. 2E).

Fig. 2.

Light micrographs of 50-μm-thick vibratome sections of the mouse kidney from control (A, B, DF) and potassium-depleted animals (A′, B′, C, C′, D′, E′, and F′). A and A′: single immunostaining for L194. In potassium-depleted animals, UT-A2 was upregulated in the DTLs of the ISOM and IMi, whereas UT-A1 was downregulated in the inner medullary collecting ducts (IMCDs). BC′: higher magnifications of the ISOM and IMi. B and B′: double immunostaining for Na-K-ATPase-α1 (blue) and L194 (brown). B′: in the potassium-depleted kidney, UT-A2 immunoreactivity was markedly increased in both the ISOM and IMi. C and C′: double immunostaining for aquaporin-1 (blue) and L194 (brown) in potassium-depleted kidney. C′: higher magnification of the area indicated by the square in C. Note that UT-A2-positive tubules cells (brown) did not colocalize with AQP1-positive DTLs of the long-loop nephron (blue, stars) located at the border between the OM and IM and the IMi. DF′: single immunostaining for L446 (D and D′), L194 (E and E′), and Q2 (F and F′). UT-A1 and UT-A3 immunoreactivity in the IMCD was reduced in the potassium-depleted kidney (D′); UT-A3 especially is markedly reduced compared with the control animals (F), whereas UT-A2 immunoreactivity was upregulated in the ISOM (E′). Magnifications: A, D, and F: ×20; B, E, and C: ×200; C′, ×400.

Immunoreactivity for Q2 (UT-A3) was significantly reduced in the IMCD in potassium-depleted animals (Fig. 2F).

Expression of UT-B. control mouse kidney. Strong UT-B immunoreactivity was observed in the endothelial cells of the DVR in the medulla (Fig. 3, A and E), as previously reported in human and rat kidneys (42). In the mouse kidney, UT-B protein was also identified in the proximal convoluted tubules in the cortical labyrinth (Fig. 3, A and C) and the papillary surface epithelium (PSE; Fig. 3G).

Fig. 3.

Light micrographs of 50-μm-thick vibratome sections of mouse kidneys from control (A, C, E, and G) and potassium-depleted animals (B, D, F, and H) illustrating UT-B immunoreactivity in proximal tubule (C and D), descending vasa recta (DVR; E and F), and papillary surface epithelium (PSE; G and H). In potassium-depleted kidney, UT-B immunoreactivity was reduced in the proximal tubule (D) and distributed diffusely in the cytoplasm of hypertrophied endothelial cells (arrows) of DVR (F). The number of UT-B-positive cells was decreased in PSE (H). Note that UT-B immunoreactivity in the PSE cells was localized only in basolateral plasma membrane (arrows) of both control and potassium-depleted animals.

potassium-depleted mouse kidney. UT-B immunoreactivity was markedly reduced in intensity not only in the DVR (Fig. 3F) and PSE (Fig. 3H) but also in the proximal tubules (Fig. 3D) compared with control animals (Fig. 3A). In the DVR, immunoreactivity for UT-B was translocated from the plasma membrane to the cytoplasm (Fig. 3F). In the PSE, the number of UT-B-positive cells was markedly reduced in the potassium-depleted group (Fig. 3H).

Electron Microscopic Immunocytochemistry

To identify the type of epithelium present in the UT-A2-positive tubules in the ISOM, electron microscopy was used with the preembedding immunoperoxidase method. In the upper part of the ISOM, no UT-A immunoreactivity was observed in either type 1 short-loop DTL or type 2 long-loop DTL (Fig. 4). In the middle part of the ISOM, strong UT-A2 labeling was found in the type 1 epithelium of the short-loop DTL, which is composed of very flat and noninterdigitating cells without microvilli (Fig. 5A). There was no immunolabeling for UT-A2 in the type 2 epithelium in the middle part of the ISOM (Fig. 5B) or in the type 3 epithelium of the long-loop DTL in the innermost part of the ISOM (Fig. 5C). Surprisingly, in the innermost part of the ISOM and the initial part of the IM, immunoreactivity for UT-A2 was detected in a very flat epithelium with the ultrastructural characteristics of type 1 epithelium. This epithelium was found to continue directly into a UT-A2-negative epithelium that also has the features of type 1 epithelium (Fig. 6, A and B). There was no immunolabeling for UT-A2 in the type 4 epithelium of the long-loop ascending thin limb (ATL) or in the type 3 epithelium of the long-loop DTL in the IM (Fig. 6, B and C).

Fig. 4.

Transmission electron microscopic localization of L194 in the outer part of ISOM of the control mouse kidney. In this area, there was no UT-A immunoreactivity in type 1 (A) and type 2 (B) epithelium. Type 1 epithelium has very simple cytoplasm, and the apical cell membrane is generally smooth. In contrast, type 2 epithelium has tall cell bodies and irregular microvilli. Inset: higher magnification of the area indicated by rectangle in B. Bars = 1 μm.

Fig. 5.

Transmission electron microscopic localization of L194 in the middle part (A and B) and in the innermost part (C) of ISOM of mouse kidney. A: strong UT-A2 was observed in the type 1 epithelium of the short-loop DTL. B: there was no UT-A2 immunoreactivity in the type 2 epithelium of the long-loop DTL in the middle part of ISOM. C: also, UT-A2 was not labeled in the type 3 epithelium in the innermost stripe of ISOM. Bars = 1 μm.

Fig. 6.

Transmission electron microscopic localization of L194 of the IMi of mouse kidney. A: strong UT-A2 immunoreactivity was observed in the type 1-like epithelium. Arrow indicates the abrupt transition from type 1-like epithelium, labeled with L194, to unlabeled type 1-like epithelium. B: note that type 1-like epithelium with UT-A2 immunoreactivity was coexisting with type 4 epithelium in IM. Type 4 epithelium is characterized by an absence of surface microvilli but has an abundance of lateral cell processes and interdigitations. C: note UT-A2-negative DTL (type 3 epithelium) and ascending thin limb (type 4 epithelium) of loop of Henle in the IM. Bars = 1 μm. AVR, ascending vasa recta.

Immunogold electron microscopy was used to determine the changes in subcellular localization of UT-A and UT-B in the potassium-depleted mouse kidney compared with the control kidney. In the ISOM, UT-A2 was located mainly in the basal plasma membrane and sparsely in the cytoplasm of the type 1 epithelium of the short-loop DTL in the control animals (Fig. 7A). In the potassium-depleted mouse kidney, however, UT-A2 was located not only in the basal plasma membrane but also in the apical plasma membrane of the type 1 epithelium (Fig. 7B). In the innermost part of the ISOM and the initial part of the IM in potassium-depleted animals, UT-A2-positive DTLs with type 1 epithelium were increased in number, and UT-A2 was detected diffusely in the cytoplasm, as well as in the apical and basal plasma membranes (Fig. 7, C and D). UT-A1 was located throughout the cytoplasm of the IMCD cells in the middle and terminal parts of the IMCD cells in control animals (Fig. 8A). In the potassium-depleted animals, cytoplasmic droplets were accumulated in the cytoplasm of IMCD cells in which UT-A1 immunolabeling was markedly decreased (Fig. 8B). The cytoplasmic droplets, which characteristically form during chronic potassium depletion, were not labeled for UT-A1 (Fig. 8B).

Fig. 7.

Transmission electron microscopic localization of L194 in the middle part of ISOM (A and B) and IMi (C and D) using the immunogold method in control (A) and potassium-depleted (BD) animals. A: in control kidney, numerous gold particles present mainly in the basal plasma membrane of type 1 epithelium. B: however, in potassium-depleted kidney, immunoreactivity was observed in the apical plasma membrane (double arrows) as well as in the basal plasma membrane of type 1 epithelium. C and D: note that type 1-like epithelium with L194 immunoreactivity coexisted with type 4 epithelium in IM. Labeling of UT-A2 was observed in both apical and basal plasma membranes of type 1-like epithelium. Arrows indicate cytoplasmic droplets in the endothelium of DVR in potassium-depleted groups.

Fig. 8.

Transmission electron microscopic localization of L194 in the IMCD cells of the middle part of IMCD using the immunogold method. A: in control animals, UT-A1 was distributed mainly in the cytoplasm of IMCD cells. B: in potassium-depleted animals, UT-A1 labeling (arrows) was markedly decreased in the cytoplasm. Note that there was no gold particles in the cytoplasmic droplet. Bar = 1 μm.

Electron microscopic studies revealed strong UT-B immunolabeling in the apical and basolateral plasma membranes and little labeling of the cytoplasm in the continuous endothelial cells of the DVR (Fig. 9A). There was no immunoreactivity in the pericytes embedded in the basal lamina of the DVR or in the fenestrated endothelial cells of the ascending vasa recta (Fig. 9A). In potassium-depleted animals, UT-B-positive endothelial cells were reduced in number (Fig. 9C). Cytoplasmic droplets were accumulated in the endothelial cells of the DVR, and these cytoplasmic droplets were labeled for UT-B (Fig. 9D). In the PSE, UT-B-positive and -negative cells were intermingled, and UT-B was detected mainly in the basolateral plasma membrane but not in the apical membrane (Fig. 10A). In potassium-depleted kidneys, UT-B immunolabeling of the basolateral plasma membrane of the PSE was significantly reduced. Cytoplasmic droplets also accumulated in the PSE cells but did not contain UT-B (Fig. 10B).

Fig. 9.

Transmission electron microscopic localization of UT-B in the DVR in the IM of mouse kidney from control (A and B) and potassium-depleted animals (C and D). A and B: UT-B immunostaining was observed in both apical and basolateral plasma membranes of the continuous endothelial cells of the DVR. However, there was no immunoreactivity in the fenestrated endothelial cells of the AVR. Small arrowheads indicate the fenestrae. A′: areas outlined in A. Note a process (arrow) of pericyte embedded in the basal lamina of UT-B-labeled endothelial cell of DVR. C and D: labeling of UT-B immunostaining was reduced and observed in the cytoplasm as well as apical and basolateral plasma membranes (C). Arrows indicate a UT-B-negative endothelial cell of DVR. In the cytoplasm, UT-B immunostaining was found in the vesicular membrane of cytoplasmic droplets (*). Bars = 1 μm.

Fig. 10.

Transmission electron microscopic localization of UT-B in the PSE of mouse kidney from control (A) and potassium-depleted animals (B). A: UT-B immunostaining was observed mainly in the basolateral plasma membrane but not in the apical plasma membrane of PSE. Note the UT-B-negative cells have an electron-dense cytoplasm (*). B: however, UT-B immunoreactivity was decreased in the potassium-depleted kidney. Arrows indicate gold particles. Bar = 1 μm.

Semiquantitative Analysis

UT-A. An immunoblot with relative optical densities of UT-A is shown in Fig. 11. The relative optical densities of the UT-A bands in each lane were compared using control values as the 100% reference. Western blot analysis of UT-A1 showed two bands at 97 and 117 kDa in the IM (Fig. 11A), which were slightly less intense in the potassium-depleted mice (79 ± 18% of the control value). In contrast, there was a significant increase in the band density of UT-A2 (195 ± 51% of the control, P < 0.01) in the OM of the potassium-depleted mice. However, UT-A3 expression revealed bands at 44 and 67 kDa, which were less intense in the IM of the potassium-depleted mice (80 ± 17% of the control, P < 0.05). In particular, the 44-kDa band was markedly reduced in the potassium-depleted mice. These data suggest that potassium depletion increases the abundance of UT-A2, whereas it decreases that of UT-A1 and UT-A3.

Fig. 11.

Immunoblot of mouse outer (B) and inner medullary (A and C) proteins probed with antibodies to UT-A in homogenates from control and potassium-depleted animals. Each lane was loaded with a sample from a different animal. Values are means ± SD normalized to the mean value for the control group. We used L194 for UT-A1 and -A2 and Q2 for UT-A3. A: densitometric analysis showed a decrease in UT-A1 abundance in the IM of the potassium-depleted mice. B: there are prominent bands at 55 kDa, which were significantly increased in the potassium-depleted mice. C: immunoblot for UT-A3 has 67- and 44-kDa bands. Note that the 44-kDa band was markedly reduced in the potassium-depleted mice. *Significantly different from control, P < 0.05.

UT-B. Semiquantitative immunoblotting analysis demonstrated that UT-B abundance was markedly reduced in membrane fractions from the cortex, OM, and IM of the potassium-depleted kidney (Fig. 12). This downregulation was in agreement with the results of the immunohistochemical studies, which showed reduced labeling in the proximal tubule, DVR, and PSE from potassium-depleted kidneys compared with control kidneys.

Fig. 12.

Immunoblot of mouse cortex (A), outer (B), and inner medullary (C) proteins probed with antibodies to UT-B in homogenates from control and potassium-depleted animals. Each lane was loaded with a sample from a different animal. Values are means ± SD normalized to the mean value for the control group. Western blot shows bands at 31, 44, and 98 kDa. The 31- and 44-kDa bands were analyzed together. A-C: densitometric analysis showed a decrease in UT-B abundance in the cortex (Co) and outer (OM) and inner medulla (IM) of the potassium-depleted mice. *Significantly different from control, P < 0.05.

DISCUSSION

The present study represents the first detailed description of the subcellular localization of urea transporters in the mouse kidney and the first report of the effect of hypokalemia on the expression of these transporters in the kidney. Both UT-A1 and UT-A3 immunostaining was expressed mainly in the cytoplasm of the cells in the middle and terminal IMCD, whereas UT-A2 was mainly located on the basolateral plasma membrane of type 1 epithelium of DTL (see schematic drawing in Fig. 13). UT-B was expressed on both the apical and basolateral plasma membrane of endothelial cells of DVR and on the basolateral plasma membrane of PSE. Potassium depletion had different effects on the expression of the UT subtypes in the mouse kidney. Expression of UT-A1 and UT-A3 was markedly reduced in the IM compared with that in control kidneys. UT-B expression was also reduced in the DVR and PSE. However, UT-A2 was significantly increased in the OM and expressed de novo in the initial part of the IM in potassium-depleted animals. These results are in agreement with the results of Western blot analysis. Thus the urinary concentrating defect in prolonged potassium depletion is associated with decreased expression of UT-A1, UT-A3, and UT-B, and with increased expression of UT-A2 in the mouse kidney.

Fig. 13.

Diagram summarizing the distribution of various subtypes of urea transporters in control and potassium-depleted mouse kidneys. VR, vasa recta; CD, collecting duct; 1, type 1 epithelium; 2, type 2 epithelium; 3, type 3 epithelium; 4, type 4 epithelium.

Expression of UT-A1 and UT-A3 in Mouse Kidney

By Western blot analysis of the IM of mouse kidney, L194 detected 97- and 117-kDa bands, which is similar to what was reported in the rat IM in previous studies (5). Recently, Terris et al. (41) succeeded in making a UT-A3-specific antibody (Q2) that detects bands at both 44 and 67 kDa. In this study, these two bands were clearly identified in the IM of the mouse kidney. These results indicate that L194 and Q2 were recognizing mouse UT-A1 and UT-A3, respectively.

Our immunolocalization studies using L194 for UT-A1 and Q2 for UT-A3 demonstrated that both UT-A1 and UT-A3 localize exclusively to the middle and terminal IMCD in a pattern very similar to that observed in rat kidney (41, 46). Surprisingly, electron microscopic studies, employing the immunogold method, demonstrated that both UT-A1 and UT-A3 were localized mainly in the cytosol of the IMCD cells with little or no labeling of the plasma membrane. This is different from what was reported in previous studies in the rat in which UT-A1 was observed in the apical plasma membrane of IMCD cells (25). Whether UT-A1 is inserted into the plasma membrane of mouse IMCD in conditions with increased urea absorption remains to be established.

Both light microscopic immunoperoxidase and ultrastructural immunogold studies demonstrated a marked decrease in UT-A1 and UT-A3 labeling in IMCD cells of potassium-depleted animals. For UT-A3, this was confirmed by semiquantitative immunoblotting demonstraing a decrease in its abundance. For UT-A1, the decrease did not reach statistical significance. These observations suggest that urea absorption in the IMCD by UT-A1 and UT-A3 may be impaired during potassium depletion, leading to a decrease in the concentration of urea in the IM and subsequently causing an impairment of urea recycling and a decreased corticomedullary osmotic gradient. These changes may play a role in the defective concentration of urine seen in the potassium-depleted kidney.

It has previously been shown that UT-A1 protein abundance is increased in several conditions with impaired urinary concentrating ability such as water diuresis, which would be expected to suppress endogenous vasopressin release (15, 16). In contrast, an increase in plasma vasopressin by exogenous administration (38), or by water restriction, is associated with a decrease in the abundance of UT-A1 proteins in the rat inner medulla (40). These observations suggest that there is long-term modulation of UT-A1 expression in response to circulating vasopressin levels. The mechanisms by which hypokalemia causes a concentrating defect are incompletely understood. It is known that potassium depletion is associated with a reduction in the production of cAMP in response to vasopressin either via increased intrarenal prostaglandin synthesis (30) or by direct inhibition of adenylate cyclase (4). However, the effect of vasopressin in increasing urea transport occurs rapidly and is not associated with an increase in urea transporter abundance. In addition, the short-term effect of vasopressin on urea transport in the IMCD is not believed to be due to stimulation of trafficking of UT-A1 (14). However, there is evidence that the short-term effect is associated with direct phosphorylation of UT-A1 (50). Studies under chronic conditions (long-term regulation) have recently provided evidence that expression levels of UTs are regulated by a transcriptional mechanism (24). However, the decrease in UT-A1 protein abundance and basal urea permeability in response to an increase in vasopressin levels is not a result of a decrease in UT-A1 mRNA, because most studies showed no changes in UT-A1 mRNA abundance in response to either water loading or restriction (34). Thus transcriptional regulation does not appear to be the mechanism for changes in UT-A1 protein abundance in response to changes in hydration and/or vasopressin level.

Recent studies by Wang and colleagues (46) demonstrated that aldosterone-induced volume expansion is associated with a decrease in the abundance of UT-A1 and UT-A3 in the IM and a dramatic decrease in serum urea concentrations in association with a transition from low-NaCl intake to high-NaCl intake. A similar decrease in the expression of UT-A1 and UT-A3 was observed after treatment with candesartan, an inhibitor of the ANG II AT1 receptor. Because sodium retention of sufficient degree to quickly expand the extracellular fluid accompanies even modest potassium depletion, it is conceivable that extracellular volume expansion and the associated suppression of the reninangiotensin system may play a role in the decrease in UT-A1 and UT-A3 expression demonstrated in the present study.

Expression of UT-A2 in Mouse Kidney

By Western blot analysis of the OM of mouse kidney, L194 labeled a 55-kDa protein band indicating that this antibody recognized mouse UT-A2. Immunohistochemical localization of UT-A2 using antibody L194 in mouse kidney revealed labeling of type 1 short-loop DTL in the middle part of ISOM in a pattern very similar to that reported previously (11, 45). In addition, L194-labeled tubular profiles were observed at the border between outer and inner medulla in control animals predominantly in the innermost part of ISOM. Previous studies using RT-PCR (37), in situ hybridization (43), and immunohistochemistry (11, 17, 45) demonstrated that UT-A2 is also expressed in the type 3 epithelium of long-loop DTL in the IM. Interestingly, it has been reported that in mouse kidney, type 3 epithelium is found in the innermost part of the ISOM and there was no type 1 epithelium in this part of ISOM or in the IM (7). However, our results demonstrate that UT-A2-positive tubular profiles in the innermost part of ISOM have ultrastructural characteristics of type 1 epithelium. Moreover, in potassium-depleted animals, tubules in which UT-A2 was expressed de novo in the initial part of the IM also had type 1 epithelium. To determine whether UT-A2 and AQP1 were expressed in the same DTL segments, double-labeling experiments were performed. In the middle part of ISOM, the DTL of short-loop nephrons expressed UT-A2 but not AQP1, whereas the DTL of the long-loop nephrons expressed AQP1 and not UT-A2. However, only the most distal portion of the short-loop DTL expressed UT-A2, as previously demonstrated in rat kidney (25). In the rat kidney, UT-A2-positive DTLs at the border between OM and IM are AQP1-positive long-loop DTL (17). In contrast, we found no overlap between UT-A2-positive DTL and AQP1-positive DTL in the innermost part of ISOM and initial part of IM, indicating that in mouse kidney UT-A2-positive DTLs with type 1 epithelium in this region are not typical long-looped DTLs of juxtamedullary nephrons but belong to long loops of midcortical nephrons that bend in the outer part of the IM (Fig. 13).

In contrast to the effect on UT-A1 and UT-A3, prolonged potassium depletion was associated with increased expression of UT-A2 protein in the DTL of the ISOM and the initial part of the IM. Moreover, in potassium-depleted animals, UT-A2 was expressed in both the apical and basolateral plasma membrane, whereas in control animals UT-A2 immunolabeling occurred predominantly in the basolateral membrane of the DTL suggesting that urea secretion into the DTL is increased in the OM during potassium depletion. Furthermore, electron microscopy revealed that UT-A2 was expressed de novo in the DTL with type 1 epithelium in the initial part of the IM. It is known that UT-A2 protein and mRNA abundances are increased in rodents with increased vasopressin levels, due to either vasopressin administration or water restriction (28, 38). However, Rutecki and colleagues (33) demonstrated a blunted response of vasopressin secretion in response to hypertonic saline during potassium depletion. Whereas baseline vasopressin values were normal, the slope of the vasopressin level vs. plasma sodium concentrations decreased greatly. The mechanism underlying the defect in vasopressin release during potassium depletion is unclear. Although the functional role of UT-A2 upregulation in ISOM and initial part of IM has not been clearly defined, it is likely that this serves as a mechanism to maintain high levels of urea and increased hypertonicity in the IM during potassium depletion.

Expression of UT-B in Mouse Kidney

Western blot analysis of mouse kidney using rabbit polyclonal antibody to the human erythrocyte urea transporter (hUT-B1) showed strong bands at 31, 44, and 98 kDa in both OM and IM, as previously observed in rats and humans (42). In the mouse kidney, the labeling of UT-B in the continuous endothelium of the DVR is consistent with that previously reported in human (48) and rat (42) kidney. Our demonstration by immunogold cytochemistry that UT-B is localized on both the apical and basolateral plasma membrane of the continuous endothelium of DVR indicates that UT-B is responsible for the high urea permeability demonstrated in microperfused DVR (27). The presence of UT-B in the plasma membrane of the DVR allows urea in the medullary interstitium to enter the DVR. When blood leaves the IM via ascending vasa recta, urea can exit these vessels through the UT-B-negative fenestrated endothelium. Studies in UT-B knockout mice demonstrated decreased urinary osmolality and increased blood urea nitrogen, which was due to decreased countercurrent exchange of urea (49). Thus the production of maximally concentrated urine appears to require UT-B protein expression in DVR and/or red blood cells (20).

Interestingly, in the mouse kidney, UT-B was also observed in the PSE where it was located in some of the light cells. Immunoelectron microscopy revealed that UT-B labeling was present only in the basolateral plasma membrane not in the apical membrane. There was no labeling of UT-A1, UT-A2, and UT-A3 in the PSE. Direct measurements of urea transport in isolated PSE of rabbits demonstrated that the urea permeability is very low (35), consistent with the absence of a urea transporter in the apical plasma membrane. The functional significance of the expression of UT-B in the PSE is not clear. However, the presence of UT-B in the basolateral membrane suggests that it may be involved in regulation of cell volume during changes in interstitial urea concentration.

In the potassium-depleted mouse kidney, expression of UT-B in the DVR and PSE was decreased. The downregulation of UT-B protein abundance has been studied in several conditions associated with reduced urinary concentrating ability such as dDAVP and furosemide administration (44), nephrectomy (13), and lithium administration (18). Promeneur et al. (28) reported that chronic dDAVP infusion causes a fall in UT-B protein abundance. Decreased V1 or V2 receptor-dependent vasopressin effects in renal medulla may be implicated in the reduction in UT-B expression induced by potassium depletion. Another possible explanation for a decrease of UT-B expression is that such a decrease may be an adaptive process to the reduced need for urea transport resulting from decreased production of urea within endothelial cells of DVR. However, future studies are needed to address these questions. In our study, UT-B immunolabeling was observed in the membranes of cytoplasmic droplets in the DVR. Hypokalemia is associated with structural changes, such as renal hypertrophy and accumulation of lipid droplets and lysosome-like bodies in the kidneys of rats, mice, hamsters, gerbils, and humans (39). Our study shows that UT-B immunolabeling was located in the membranes of cytoplasmic droplets in the DVR in potassium-depleted animals. Interestingly, the droplets that accumulated in the IMCD and PSE during potassium depletion were not UT immunopositive. This is in contrast to the results of previous studies that demonstrated AQP2 immunoreactivity in cytoplasmic droplets in IMCD cells during hypokalemia.

In summary, this study demonstrates that prolonged potassium depletion is associated with decreased expression of UT-A1 and UT-A3 in the IMCD and decreased expression of UT-B in the DVR and PSE. In contrast, the expression of UT-A2 was increased in DTL in ISOM and the initial part of IM. These observations suggest that changes in UT abundance in the renal medulla may play a role in the impaired urinary concentration ability associated with potassium deprivation.

DISCLOSURES

This work was supported by the Korea Science and Engineering Foundation (R01-2000-000-00117-0) and presented at American Society of Nephrology, Philadelphia, PA (2002).

Acknowledgments

We gratefully acknowledge the technical assistance of H.-D. Roh, I.-S. Lee, H.-L. Kim, and K.-A. Ryu.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

References

View Abstract