Abstract
Abstract. Aquaporin-4 (AQP4) is a member of the aquaporin water-channel family. AQP4 is expressed primarily in the brain, but it is also present in the collecting duct of the kidney, where it is located in the basolateral plasma membrane of principal cells and inner medullary collecting duct (IMCD) cells. Recent studies in the mouse also have reported the presence of AQP4 in the basolateral membrane of the proximal tubule. The purpose of this study was to establish the pattern of AQP4 expression during kidney development and in the adult kidney of both the mouse and the rat. Kidneys of adult and 3-, 7-, and 15-d-old mice and rats were preserved for immunohistochemistry and processed using a peroxidase pre-embedding technique. In both the mouse and the rat, strong basolateral immunostaining was observed in IMCD cells and principal cells in the medullary collecting duct at all ages examined. Labeling was weaker in the cortical collecting duct and the connecting tubule, and there was no labeling of connecting tubule cells in the mouse. In adult mouse kidney, strong AQP4 immunoreactivity was observed in the S3 segment of the proximal tubule. However, there was little or no labeling in the cortex or around the corticomedullary junction in 3- and 7-d-old mice. Between 7 and 15 d of age, distinct AQP4 immunoreactivity appeared in the S3 segment of the mouse proximal tubule concomitant with the differentiation of this segment of the nephron. Labeling of proximal tubules was never observed in the rat kidney. These results suggest that there are differences in transepithelial water transport between mouse and rat or that additional, not yet identified water channels exist in the rat proximal tubule.
Aquaporin-4 (AQP4) is a member of the aquaporin water-channel family that now includes at least 10 mammalian aquaporins, many of which are expressed in the kidney (1,2,3). The water-channel proteins that have been identified in the kidney include AQP1, AQP2, AQP3, and AQP4, which are plasma membrane water channels that are involved in transcellular water transport, and AQP6, which is located in intracellular vesicles. Recent studies also have reported the presence of AQP7 (4,5) and AQP8 in the kidney (6).
AQP1 is expressed constitutively in the proximal tubule, where most of the water in the glomerular filtrate is reabsorbed (7,8,9). Because AQP1 is present in both the apical and basolateral membrane of the cells, it provides a transcellular pathway for water movement across the epithelium of the proximal tubule. Until recently, there was no evidence that other members of the aquaporin family were expressed in the plasma membrane of the proximal tubule. AQP1 is also expressed in the apical and basolateral plasma membrane of the descending thin limb of Henle's loop (7,8) and in the nonfenestrated endothelium of the descending vasa recta (10), and it is believed to play an important role in the urinary concentrating mechanism. Recent studies demonstrated that AQP7 also is expressed in the proximal tubule of both the mouse (4) and the rat (4,5). It is located mainly in the S3 segment, where it is confined to the brush border.
AQP2 is found only in the renal collecting duct and connecting tubule (CNT), where it is located in the apical plasma membrane and in intracellular vesicles of principal cells, inner medullary collecting duct (IMCD) cells, and CNT cells (11). Its expression and intracellular location are regulated by the antidiuretic hormone vasopressin (12).
In the kidney, AQP3 and AQP4 are expressed in the basolateral plasma membrane of principal cells, IMCD cells, and CNT cells (13,14,15) and had not been detected outside of the collecting duct system until recently. However, a recent study by Van Hoek et al. (16) and preliminary studies from our laboratory (17) reported that AQP4 also is expressed in the S3 segment of the mouse proximal tubule. This contrasts with observations in the rat kidney, where AQP4 is found in the collecting duct only (15).
The purpose of this study was to explore further the expression of AQP4 in the kidney of both the mouse and the rat and to establish the time and pattern of expression during kidney development by the use of immunohistochemical techniques.
Materials and Methods
Animals and Tissue Preservation
The distribution of AQP4 in the kidneys of adult animals was examined in six mice (four transgenic wild-type mice, AQP1 +/+, produced by the intercrossing of CD1 AQP1 heterozygotes (18), and two ICR white mice) and five Sprague-Dawley rats. Kidney tissue from an AQP4 knockout mouse (19) was used as a control for the specificity of AQP4 immunostaining. Studies of AQP4 expression in the developing kidney were performed in 3-, 7-, and 15-d-old mice and rats. At each of the three ages, nine AQP1 +/+ mice that were derived from one litter and four to six Sprague-Dawley rats that were derived from two litters were examined.
The animals were anesthetized with an intraperitoneal injection of pentobarbital sodium (50 mg/kg body wt). Kidneys from mice and neonatal rats were preserved by in vivo perfusion through the heart, whereas kidneys from adult rats were perfused through the abdominal aorta. The kidneys were perfused briefly with phosphate-buffered saline (PBS) before perfusion with a periodate-lysine-2% paraformaldehyde solution for 10 min. The kidneys were removed and cut into 1- to 2-mm-thick slices that were fixed additionally by immersion in the same fixative for 2 h at room temperature and overnight at 4°C. Sections were cut transversely through the kidney on a Vibratome Pelco 101 (Sectioning Series 1000; Technical Products International, St. Louis, MO) at a thickness of 50 μm and processed for immunohistochemical studies by using a horseradish peroxidase pre-embedding technique.
Immunohistochemistry
Vibratome sections were washed with 50 mM NH4Cl in PBS three times for 15 min. Before incubation with the primary antibody, the sections were incubated for 3 h with PBS that contained 1% bovine serum albumin, 0.05% saponin, and 0.2% gelatin (solution A). The tissue sections then were incubated overnight at 4°C with polyclonal antibody against AQP4, diluted 1:250 in PBS containing 1% bovine serum albumin (solution B). The antibody against AQP4 labels the basolateral plasma membrane of principal cells and IMCD cells in rat and has been characterized in detail by Terris et al. (15). After several washes in solution A, the sections were incubated for 2 h in peroxidase-conjugated goat anti-rabbit immunoglobulin G Fab fragment (Jackson ImmunoResearch Laboratories, West Grove, PA) diluted 1:100 in solution B. The sections then were rinsed in solution A and subsequently in 0.05 M Tris buffer (pH 7.6). For detection of horseradish peroxidase, the sections were incubated in 0.1% 3,3′-diaminobenzidine in 0.05 M Tris buffer for 5 min. H2O2 then was added to a final concentration of 0.01%, and the incubation was continued for 10 min. After washing with 0.05 M Tris buffer three times, the sections were dehydrated in a graded series of ethanol and embedded in TAAB resin (Marivac Ltd., Nova Scotia, Canada). From all animals, 50-μm-thick Vibratome sections through the entire kidney were mounted in TAAB between polyethylene vinyl sheets. Sections from the cortex and the outer and inner medulla were excised and glued onto empty blocks of TAAB, and 1-μm sections were cut for light microscopy. The sections were photographed on an Olympus photomicroscope (Tokyo, Japan) that was equipped with differential interference contrast optics.
Results
Expression of AQP4 in the Adult Kidney
The pattern of AQP4 distribution in kidneys from mice and rats was examined at low magnification on 50-μm-thick sections (Figure 1) and at high magnification on 1.5-μm-thick sections (Figures 2,3,4). Sections from AQP4 knockout mice served as a control for the specificity of AQP4 labeling. There was no AQP4 immunostaining in the kidney of the AQP4 -/- animal (Figure 1a).
Light micrographs that illustrate immunostaining for aquaporin-4 (AQP4) in 50-μm-thick sections from adult mouse and rat kidneys. (a) Section from AQP4 knockout mouse that served as a control for the specificity of immunolabeling. There was no AQP4 immunoreactivity and particularly no labeling of tubules in the outer stripe of the outer medulla (OS) in the AQP4 knockout mouse. (b) In normal mouse, a broad band of intense AQP4 immunoreactivity was observed in the OS, representing the S3 segment of the proximal tubule, and in the medullary collecting duct. (c) In rat, there was strong AQP4 immunostaining in the medullary collecting duct and weaker labeling in the connecting tubule (CNT) and cortical collecting duct (CCD). There was no labeling of the proximal tubules in the OS. Co, cortex; IS, inner stripe of outer medulla. Magnifications: ×50 in a and b; ×33 in c.
Light micrographs of 1.5-μm-thick sections from renal cortex of adult mouse (left panels) and rat (right panels) illustrating immunostaining for AQP4. In the mouse, there was no immunoreactivity in the proximal convoluted tubule (PCT), distal convoluted tubule (DCT), or CNT (a). However, a few AQP4-positive cells (arrows) were observed in the CCD (c). In the rat, basolateral AQP4 immunostaining was observed in CNT cells (b) and principal cells in the CCD (d). There was no labeling of intercalated cells (arrowheads) in either mouse or rat. Magnification: ×1000 in a through d.
Light micrographs of 1.5-μm-thick sections from outer stripe of outer medulla of adult mouse (left panels) and rat kidney (right panels) illustrating immunostaining for AQP4. c, d, and e are higher magnifications of areas indicated by rectangles in a and b. In the mouse (a, c, and d), basolateral AQP4 immunoreactivity was observed in the S3 segment of the proximal tubule as well as in the outer medullary collecting duct (OMCDo). There was no labeling of intercalated cells (arrowhead). In the rat (b and e), AQP4 immunostaining was observed in principal cells in the OMCDo. There was no AQP4 immunoreactivity in the S3 segment of the proximal tubule or in intercalated cells (arrowheads). TAL, thick ascending limb. Magnifications: ×440 in a and b; ×1000 in c through e.
Light micrographs of 1.5-μm-thick sections from inner stripe of outer medulla (a and b) and inner medulla (c and d) of adult mouse (left panels) and rat kidney (right panels) illustrating immunostaining for AQP4. In the inner stripe portion of the outer medullary collecting duct (OMCDi) of the mouse, strong AQP4 immunoreactivity was observed in the basal membrane of principal cells, whereas labeling of the lateral membrane was very weak (a). As the collecting duct descended through the papilla, AQP4 immunoreactivity gradually increased, and intense labeling was present in both basal and lateral membranes of inner medullary collecting duct (IMCD) cells (c). In the rat, strong basolateral AQP4 immunoreactivity was present in both OMCDi (b) and IMCD (d). Arrowheads indicate AQP4-negative intercalated cells. Magnification, ×1000 in a through d.
In the 50-μm sections of normal mouse kidney, intense AQP4 immunostaining was observed in the outer and inner zones of the medulla, especially in the outer stripe of the outer medulla, where AQP4-positive tubular profiles formed a broad band (Figure 1b). In the cortex, however, there was very weak labeling for AQP4. In contrast, in the rat, tubules with strong AQP4 immunostaining were observed in all regions of the kidney, extending from the outer cortex through the inner medulla (Figure 1c). However, the broad band of AQP4-positive tubules that was observed in the outer stripe of the outer medulla of the mouse was not present in the rat kidney.
The cellular distribution of AQP4 immunoreactivity was determined at higher magnification on the 1.5-μm sections. In the cortex of the mouse kidney, there was no AQP4 immunostaining in the proximal convoluted tubule, distal convoluted tubule, or CNT (Figure 2a). However, in the cortical collecting duct (CCD) in the medullary ray, principal cells exhibited weak basolateral staining (Figure 2c). In contrast, in the rat, strong AQP4 immunostaining was observed in the basolateral plasma membrane of both CNT cells and principal cells in the CCD (Figure 2, b and d). In the outer stripe of the outer medulla of the mouse, strong AQP4 immunostaining was observed in the basolateral plasma membrane of S3 segments of the proximal tubule and principal cells of the outer medullary collecting duct (OMCD) (Figure 3, a, c, and d). In the outer medulla of the rat kidney, AQP4 immunostaining was observed in the OMCD only, and there was no AQP4 immunoreactivity in S3 segments of the proximal tubule (Figure 3, b and e). In both the mouse and the rat, strong AQP4 immunostaining was observed in principal cells of the OMCD in the inner stripe of the outer medulla and in principal cells and IMCD cells in the IMCD (Figure 4). In the mouse, principal cells in the OMCD exhibited mainly basal plasma membrane labeling and there was very weak labeling of the lateral membrane (Figure 4a). There was a gradual increase in AQP4 immunoreactivity as the collecting duct descended through the papilla, and intense labeling was observed in both the basal and lateral plasma membrane of principal cells and IMCD cells in the IMCD (Figure 4c). In the rat, however, intense immunolabeling was observed in both the basal and lateral plasma membrane of principal cells throughout the collecting duct (Figures 2d, 3e, and 4, b and d).
Expression of AQP4 in the Developing Kidney
The distribution of AQP4 immunoreactivity was examined on both 50-μm-thick and 1.5-μm-thick kidney sections from 3-, 7-, and 15-d-old mice (Figures 5 and 6) and rats (Figures 7 and 8). In the developing mouse kidney, strong AQP4 immunostaining was observed in the renal papilla at all ages examined (Figure 5). The intensity of staining was similar to what was observed in adult animals and represented labeling of the basolateral plasma membrane of principal cells and IMCD cells. However, there was little or no staining of the renal cortex or the area around the corticomedullary junction in 3-and 7-d-old pups (Figure 5, a and b). At 15 d of age, weak immunostaining was observed in the collecting ducts in the outer medulla and in a band that corresponded to the outer stripe of the outer medulla (Figure 5c). However, the intensity of labeling was less than in adult animals (Figure 5d). Higher magnification revealed that the staining in the outer stripe of the outer medulla represented labeling of the basolateral plasma membrane of the S3 segment of the proximal tubule and the principal cells of the collecting duct (Figure 6). In the developing rat kidney, intense AQP4 immunostaining was observed in the medullary collecting duct at all ages examined. Weak labeling also was present in the CCD (Figure 7). Higher magnification demonstrated strong labeling of the basolateral plasma membrane of principal cells and IMCD cells but weak basolateral labeling of CNT cells (Figure 8). AQP4 immuno-reactivity was not observed in the S3 segment of the rat proximal tubule at any of the ages examined.
Light micrographs of 50-μm-thick sections of kidneys from 3- (a), 7- (b), and 15-d-old (c) pups and adult mouse (d), illustrating immunostaining for AQP4. In 3- and 7-d-old pups (a and b), AQP4 was strongly expressed in the collecting duct, but there was little or no labeling of proximal tubules. In 15-d-old pup (c), distinct AQP4 immunostaining was observed in the S3 segment of proximal tubules in the outer stripe of the outer medulla (OS); however, the intensity of staining was less than in adult kidney (d). Magnifications: ×40 in a; ×25 in b through d.
Light micrographs of 50-μm-thick sections of renal outer medulla from 3- (a), 7- (b), and 15-d-old (c) pups and adult mouse (d) illustrating immunostaining for AQP4. Very faint immunostaining is present in the differentiating S3 segment of the proximal tubule (*) in 3-and 7-d-old pups (a and b). In 15-d-old pup (c), there was increased AQP4 immunoreactivity in the S3 segment of the proximal tubule (*), although the intensity of staining was less than in adult kidney (d). CD, collecting duct. Magnifications: ×500 in a through c; ×1000 in d.
(a) Light micrographs of 50-μm-thick sections of kidneys from 3- (a), 7- (b), and 15-d-old (c) rats illustrating immunostaining for AQP4. Immunolabeling is present along the entire collecting duct from cortex to inner medulla as well as in the CNT at all ages. Note that AQP4 immunoreactivity is weak in the cortex of 15-d-old pups (c). Magnification, ×16 in a through c.
Light micrographs of 1.5-μm-thick sections of cortex (a), outer medulla (b), and terminal part of inner medulla (c) from 3-d-old rat illustrating immunostaining for AQP4. There is weak basolateral AQP4 immunostaining of the CNT (*) (a) and intense labeling of principal cells in the OMCD (b) and IMCD cells (c). IMCDt, terminal IMCD. Magnification, ×400 in a through c.
Discussion
This study represents the first detailed description of AQP4 protein expression in the developing kidney and includes a detailed comparison between the distribution of AQP4 in mouse and rat kidney. In adult animals, the most striking difference in the distribution of AQP4 between the two species was observed in the S3 segment of the proximal tubule, which exhibited very strong basolateral AQP4 immunostaining in the mouse but no labeling in the rat. However, differences in the expression of AQP4 between the mouse and the rat also were observed in the CNT and the CCD. In the rat, AQP4 immunoreactivity was present in CNT cells and in principal cells throughout the collecting duct. In contrast, there was no labeling of CNT cells in the mouse and the intensity of AQP4 immunoreactivity in the CCD was much weaker than in the rat.
The demonstration of AQP4 immunoreactivity in the basolateral plasma membrane of the S3 segment of the mouse proximal tubule is in agreement with the results of a recent study by Van Hoek et al. (16), who reported similar findings in the proximal tubule. It is well established that another member of the aquaporin family, AQP1, is the major water-channel protein in the proximal tubule (7,8). AQP1 is expressed strongly in the brush-border membrane in all segments of the proximal tubule, and it also is present in the basolateral membrane. Thus the S3 segment of the mouse has two different water channels, AQP1 and AQP4, in the basolateral plasma membrane. To ensure that our antibody against AQP4 did not cross react with another protein in the mouse kidney, we also performed immunohistochemical staining in kidneys from a transgenic mouse that lacked AQP4. The lack of proximal tubule staining in the AQP4 knockout mouse indicates that the observed immunolabeling of the S3 segment of normal mouse proximal tubule does represent AQP4. Expression of AQP4 in the proximal tubule has been observed only in the mouse, and so far there is no evidence that it occurs in other species. The functional significance of having two different aquaporins in the basolateral membrane of the S3 segment of the proximal tubule is not known, and it is not clear why there would be a difference between the mouse and the rat. However, it is entirely possible that another as-yet-unknown aquaporin might be expressed in the basolateral plasma membrane of the S3 segment of the rat proximal tubule. It is interesting that there is now evidence that a second aquaporin, AQP7, also is present in the apical plasma membrane of the proximal tubule, especially in the S3 segment, of both the mouse (4) and the rat (4,5). In the mouse, apical AQP7 and basolateral AQP4 may provide another pathway for transepithelial water movement across the S3 segment of the proximal tubule in addition to the one created by the expression of AQP1.
A recent study of proximal tubule water reabsorption in transgenic AQP1 knockout mice demonstrated a 78% decrease in osmotic water permeability in microdissected S2 segments from AQP1 knockout mice compared with wild-type mice (20). Although there was a striking reduction in water permeability and reabsorption, there was no sign of increased fluid delivery to the distal tubule. This was attributed to a measured decrease in the single-nephron glomerular filtration rate that presumably was secondary to the activation of the tubuloglomerular feedback mechanism. A subsequent study demonstrated a markedly increased transtubular osmotic gradient in the proximal tubules of AQP1 knockout mice (21). On the basis of these observations, it was suggested that enhanced osmotic water flow through either the paracellular pathway or the lipid bilayer might provide partial compensation for the defect in water permeability that is caused by the lack of AQP1. Although this might account for the difference between the reduction in osmotic water permeability and the reduction in fluid reabsorption in the proximal tubule, it does not explain the lack of increased fluid delivery to the distal tubule. Other than a possible activation of the tubuloglomerular feedback mechanism, an alternative explanation for the lack of increased distal fluid delivery would be a compensatory increase in water reabsorption between the S2 segment of the proximal tubule and the distal tubule through other pathways. The expression of AQP7 in the brush-border membrane and AQP4 in the basolateral membrane of the S3 segment of the proximal tubule might constitute such an alternative pathway for water reabsorption in the AQP1 knockout mice. However, it should be pointed out that the function of AQP4 in the proximal tubule has not been established and that it is possible that AQP4 might have a role that is unrelated to water transport.
In both the mouse and the rat, there was very strong immunostaining for AQP4 in the IMCD, with labeling of the basolateral plasma membrane of both principal cells and IMCD cells. The importance of AQP4 for water reabsorption in the IMCD was substantiated by Chou et al. (22) in a recent study of transepithelial water permeability in isolated microperfused IMCD segments from transgenic AQP4 knockout mice. Those investigators demonstrated a fourfold reduction in osmotic water permeability in IMCD segments from AQP4 knockout mice compared with wild-type mice, indicating that AQP4 is the main water channel in the basolateral membrane of the IMCD. Whether AQP4 deletion also is associated with changes in the water permeability of the S3 segment of the mouse proximal tubule remains to be determined.
The distribution and intensity of AQP4 immunolabeling in the IMCD of the rat was similar to what was reported previously by Terris et al. (15). However, our study also demonstrates strong AQP4 immunoreactivity in the basolateral plasma membrane of CNT cells and principal cells in the CCD and particularly in the OMCD of the rat. This contrasts with the observations by Terris et al. (15), who reported little or no expression in the CCD and OMCD. The two studies were performed with the same polyclonal anti-AQP4 antibody that was derived from the same rabbit. Therefore, the most likely explanation for the discrepancy is the higher sensitivity of the pre-embedding immunoperoxidase method that was used in this study. It also is possible that the expression of AQP4 is different in the two strains of rats that were used in these studies. The experiments that are described in this study were performed in Sprague-Dawley rats, whereas Wistar rats were used in the immunohistochemical studies reported by Terris et al. (15).
In contrast to observations in the rat, AQP4 immunoreactivity was absent from the CNT of the mouse. Furthermore, the intensity of immunostaining in the CCD was lower in the mouse. The functional significance of this difference is not known. However, it should be pointed out that principal cells and CNT cells in both the mouse and the rat also express AQP3, another member of the aquaporin water-channel family, in the basolateral plasma membrane. Therefore, it is possible that CNT cells and possibly principal cells in the CCD do not need two different basolateral aquaporins in the mouse. Whether there are differences between the rat and the mouse in the expression of AQP3 in the CNT and the CCD remains to be established. It also is conceivable that there might be another, not-yet-identified water channel in the CNT and the CCD of the mouse.
The expression of AQP4 was similar in the neonatal kidney of the mouse and the rat; both species exhibited very intense immunoreactivity in the medullary collecting duct. However, weak immunoreactivity also was observed in the CCD of the rat, although this segment is not differentiated fully during the first 1 to 2 wk of life. There was little or no expression of AQP4 in the renal cortex of the neonatal mouse kidney. The expression of AQP4 in the S3 segment of the mouse proximal tubule coincided with the differentiation of this segment and the subsequent delineation of the outer stripe of the outer medulla. In 3- and 7-d-old pups, the descending limb of Henle's loop is not yet differentiated into the S3 segment of the proximal tubule and the descending thin limb. Therefore, there is no clear distinction between the outer and inner stripe of the outer medulla. With the differentiation of structures in the outer stripe, including the S3 segment of the proximal tubule, strong AQP4 immunoreactivity appeared in this region, creating a broad band of staining that corresponded to the outer stripe of the outer medulla. It is interesting that the expression of AQP4 in the S3 segment of the proximal tubule and the demarcation of the outer stripe of the outer medulla occur at the time that the animal develops the capacity to concentrate urine (23,24).
In the rat, the development of the outer stripe of the outer medulla, including the differentiation of the S3 segment of the proximal tubule, was similar to what was observed in the mouse kidney, indicating that AQP4 is not required for the normal development of the structures in this region.
In summary, the results of this study demonstrate significant differences between the expression of AQP4 in mouse and rat kidneys. The expression of AQP4 in the S3 segment of the proximal tubule in mouse but not rat kidney suggests that differences exist between the two species in transepithelial water transport in this segment or that additional, as-yet-unidentified water channels are present in the rat proximal tubule.
Footnotes
Juha Kokko, MD, served as guest editor and supervised the review and final disposition of this manuscript.
- © 2001 American Society of Nephrology
References
