Expression of claudin-7 and -8 along the mouse nephron

【摘要】  Claudins are integral membrane proteins of the tight junction that determine the magnitude and selectivity of paracellular permeability in epithelial tissues. The mammalian renal tubule exhibits considerable heterogeneity in the permeability properties of its different segments. To determine the nephron segment localization of claudin-7 and -8, immunofluorescence staining of mouse kidney sections was performed using isoform-specific antibodies. Claudin-8 was found to be expressed primarily at the tight junction along the entire aldosterone-sensitive distal nephron and in the late segments of the thin descending limbs of long-looped nephrons. This pattern of expression is consistent with the putative role of claudin-8 as a paracellular cation barrier. By contrast, claudin-7 was found in the same nephron segments as claudin-8, but it was expressed primarily at the basolateral membrane.

【关键词】  tight junction paracellular transport immunofluorescence renal tubule

THE MAJOR ROLE OF THE RENAL tubule is to alter the volume and composition of the glomerular filtrate in accordance with the homeostatic needs of the body, thereby transforming tubular fluid into urine. The final steps in this adjustment of tubular fluid occur in the distal nephron and primarily involve active, transcellular solute reabsorption and secretion. For example, in the aldosterone-sensitive distal nephron (ASDN), which constitutes the segment of nephron from the distal convoluted tubule (DCT) to the inner medullary collecting duct (IMCD) ( 29 ), aldosterone stimulates active Na + reabsorption and active K + and H + secretion, generating transtubular gradients as high as 1:30 (lumen:peritubular) for Na + ( 27 ), 15:1 for K + ( 11 ), and 1,000:1 for H + ( 19 ). The role of the paracellular pathway in the ASDN is to act as a barrier to prevent dissipation of such gradients by passive backleak of these cations. Not surprisingly, the entire ASDN constitutes a tight epithelium ( 21, 38, 39 ). The only other part of the nephron that has such a low paracellular permeability is the late thin descending limb of Henle ( 22 ).

The tight junction constitutes the rate-limiting step in the paracellular pathway ( 36 ). Claudins are tetraspan proteins that are located in the tight junction ( 40 ). Recent evidence suggests that the presence of claudins is necessary and sufficient for the formation of tight junction strands ( 14 ) and that claudins directly affect the magnitude and selectivity of paracellular permeability in epithelia (reviewed in Ref. 47 ). This suggests that claudins form both the barrier and the pore of the paracellular pathway ( 47 ). More than 20 claudin isoforms have been identified in mammalian species, and each is expressed in a tissue-specific manner ( 47 ). Different claudin isoforms, when overexpressed or knocked out in cell lines and in genetic mouse models, alter paracellular permeability differently in terms of direction, magnitude, and selectivity. It is therefore likely that the particular combination of claudin isoforms expressed in each nephron segment is unique and determines the paracellular permeability properties unique to that segment.

We recently reported that claudin-8 decreases paracellular permeability to cations when overexpressed in Madin-Darby canine kidney (MDCK)-II cells, suggesting that it might function primarily as a paracellular cation barrier ( 48 ). In a comprehensive renal immunolocalization study of multiple claudin isoforms, Kiuchi-Saishin et al. ( 25 ) found claudin-8 expression in several mouse distal nephron segments. Together with our findings, this suggests that claudin-8 may contribute to the paracellular barrier that protects transtubular Na +, K +, and H + gradients in the distal nephron. Surprisingly, Kiuchi-Saishin also found claudin-8 in the thin ascending limb of Henle, a segment of the nephron that is known to be leaky to Na + ( 20 ).

We have now generated our own antibody against claudin-8 and performed a complete characterization of the nephron segments in mouse kidney in which claudin-8 is expressed. We find that claudin-8 is expressed continuously in tight junctions along the entire ASDN and also in the late thin descending limb of Henle, consistent with its putative role as a paracellular cation barrier protein, and that it is absent from the thin ascending limb. During the course of this project, a commercial antibody to claudin-7 also became available. Using this antibody for immunolocalization, we found that claudin-7 is expressed in the very same nephron segments as claudin-8 but localizes primarily to the basolateral membrane.

METHODS

Antibodies. Antibody 933 is a custom affinity-purified polyclonal antibody (Invitrogen Custom Antibody Service, Carlsbad, CA) raised in a rabbit against a synthetic 21-residue peptide (CQRSFHAEKRSPSIYSKSQYV) corresponding to the COOH terminus of claudin-8 ( 31 ). This sequence is unique among the known claudin isoforms. A BLAST search revealed no other close matches in the National Center for Biotechnology Information nonredundant protein database, although the possibility that it may resemble an as yet unidentified claudin cannot be totally excluded. Antibody 933 (0.4 mg/ml) was used at a dilution of 1:100 for unamplified immunofluorescence, 1:5,000 for immunofluorescence with tyramide signal amplification (TSA), and 1:1,000 for immunoblotting. Anti-claudin-7 antibody purchased from Zymed (San Francisco, CA) is an affinity-purified rabbit polyclonal antiserum raised against a synthetic peptide corresponding to the COOH terminus of claudin-7. It was used at a dilution of 1:100 for unamplified immunofluorescence and for immunoblotting, and 1:1,000 for immunofluorescence with TSA. Other antibodies used for double-label immunofluorescence colocalization studies are listed in Table 1.

Table 1. List of antibody markers used for colocalization studies

Protein samples and immunoblotting. To prepare crude kidney membranes, mouse kidney cortex was dissected, homogenized, and centrifuged at 47,000 g. To generate a glutathione- S -transferase (GST) fusion protein with the last 40 amino acids (residues 186-225) of mouse claudin-8, the corresponding nucleotide sequence of claudin-8 was amplified by PCR, cloned into the vector, pGEX-4T1 (Amersham Biosciences, Piscataway, NJ), and transformed into Escherichia coli DH10B cells. Protein was induced by bacterial growth in the presence of isopropyl - D -thiogalactoside, the cells were lysed by sonication in the presence of 1% Triton X-100, and the fusion protein was affinity-purified on a glutathione-Sepharose column. A full-length claudin-8 protein preparation was generated by immunoprecipitation. MDCK-II cells expressing NH 2 -terminal FLAG epitope-tagged mouse claudin-8 under the transcriptional control of the TetOff system ( 48 ) were grown for 5 days in the absence of doxycycline to induce claudin-8 expression. Cells were lysed in 1% Triton X-100 and immunoprecipitated with mouse monoclonal anti-FLAG antibody (M2; 5 µg/ml; Sigma-Aldrich) and protein G-agarose. To generate a claudin-7 protein preparation, the coding sequence of mouse claudin-7 (IMAGE cDNA clone 903882) was subcloned into the vector pcDNA3.1 (Invitrogen). Recombinant plasmid DNA, or vector alone, was transiently transfected into human embryonic kidney (HEK-293) cells, which were then harvested, and 100,000- g membranes were isolated. Reducing SDS-PAGE and immunoblotting were performed by standard methods, as described previously ( 9 ).

Immunofluorescence. For in situ fixation of mouse kidneys, we adapted a method developed for rats by McDonough and colleagues ( 45 ). Male CD1 mice (30-45 g) were anesthetized with pentobarbital sodium. The left kidney, still attached to its vascular stalk, was mobilized, decapsulated, placed in a small plastic cup, and partially immersed in fixative (2% paraformaldehyde in phosphate-buffered saline) at 37°C for 15 min. The kidney was then excised, bisected in the coronal plane, and the half that had been fully bathed in fixative was retrieved and postfixed in 2% paraformaldehyde for another 4 h. The fixed tissue was cryoprotected by overnight incubation in 30% sucrose in PBS, embedded in Tissue Tek OCT Compound (Sakura Finetek, Torrance, CA), frozen, and 5-µm cryosections were cut. In some studies in situ fixation was omitted, with similar appearance of claudin antibody staining.

Immunofluorescence staining was performed essentially as described previously ( 10 ). Secondary antibodies used were goat anti-rabbit, -mouse, or -guinea pig IgG conjugated to Alexa Fluor 488 or 555. For double-staining experiments in which both primary antibodies were of rabbit origin, the rabbit anti-claudin antibody was applied first at a lower concentration and amplified with the TSA kit using Alexa Fluor 488 tyramide (Molecular Probes, Eugene, OR). The second primary antibody was subsequently added and then was detected using goat anti-rabbit Alexa Fluor 555. For the peptide-block control, 933 serum (1:100, 4 µg/ml) was preincubated overnight at 4°C with its cognate peptide (1.25 µg/ml) and then applied to kidney sections.

Images were acquired at the USC Center for Liver Diseases using a Nikon PCM confocal microscopy system with argon and helium-neon lasers. For double-labeled slides with a large disparity in fluorescence brightness between the two channels, images were acquired sequentially for each fluorophore in single-label mode to minimize spectral overlap between channels. Each pair of images was then imported into Adobe Photoshop 7.0, pseudocolored, and merged to generate dual-color images.

RESULTS

Immunolocalization of claudin-8. To determine the intrarenal localization of claudin-8, we first generated a rabbit polyclonal antibody (933) against a unique 20-amino acid sequence at the COOH terminus of the mouse claudin-8 protein. This peptide sequence had been used successfully by Morita et al. ( 31 ) previously to generate an isoform-specific antibody.

The specificity of the 933 antibody for claudin-8 was first assessed by Western blotting. The 933 antibody was able to recognize a band in mouse kidney cortical membranes of the expected size (25 kDa) for claudin-8 ( Fig. 1 ). Specificity for claudin-8 was confirmed by demonstrating that the 933 antibody recognizes a 35-kDa GST fusion protein with the COOH-terminal 40 amino acids of claudin-8, as well as a 25-kDa FLAG epitope-tagged claudin-8 protein in lysates from transiently transfected HEK cells and in immunoprecipitates from a stably transfected MDCK cell line ( 48 ) ( Fig. 1 ). A second band of 40 kDa, which we have observed in immunoprecipitates of a variety of claudin isoforms (unpublished observations), was also seen in the claudin-8 immunoprecipitate.

Fig. 1. Western blot analysis of claudin antibodies. A : anti-claudin-8 antibody (933) blots. Samples ( left to right ) are mouse kidney cortex crude membranes, a fusion protein of Escherichia coli glutathione- S -transferase (GST) with a COOH-terminal polypeptide of claudin-8 (35 kDa), membranes from HEK cells transiently transfected with vector, claudin-7 or claudin-8, and Madin-Darby canine kidney (MDCK)-II cells stably transfected with a FLAG epitope-tagged claudin-8 construct and immunoprecipitated (IP) with the FLAG antibody. IB, immunoblot. B : anti-claudin-7 antibody blot. Samples indicated are the same as in A. Arrowheads denote the position of the 25-kDa native claudin band on each blot.

By immunofluorescence staining of mouse kidney cryosections with the 933 antibody, claudin-8 protein was found in numerous scattered tubules throughout the cortical labyrinth and medullary rays, in a small subset of the tubules in the outer stripe of the outer medulla, and in most of the tubules in the inner stripe and in the inner medulla ( Fig. 2 ). Within the tubules, staining was found primarily as either dots or reticular networks of lines near the apical surface, consistent with tight junction staining ( Fig. 3 ). This was confirmed by demonstrating that claudin-8 colocalized with zona occludens-1, a tight junction marker, but was just apical in location to -catenin, a marker of the adherens junction and basolateral membrane ( Fig. 4 ). This pattern of staining was absent with nonimmune rabbit serum or when the 933 antibody was blocked by preincubation with the claudin-8 peptide ( Fig. 3 ).

Fig. 2. Low-magnification overview of a mouse kidney stained with 933 anti-claudin-8 serum.

Fig. 3. Immunofluorescence staining of mouse kidney cryosections with 933 anti-claudin-8 serum. A : 933 antiserum (green) stains the tight junctions of tubule cells. Tight junction strands appear as lines (arrow) or reticular networks in tangential section and as bright dots in cross section (arrowheads). No 933 staining was observed with nonimmune serum ( B ) or peptide-blocked serum ( C ). Cell nuclei were counterstained with propidium iodide (red). Asterisks, tubule lumens. Scale bars = 25 µm.

Fig. 4. Subcellular localization of claudin-8 (Cldn8) in tubule epithelial cells. Dual-color immunofluorescence was performed using 933 antibody (green) and antibodies to the following markers (red): zona occludens-1 (ZO-1; tight junction); -catenin (adherens junction and basolateral membrane). Scale bars = 25 µm.

Dual-antibody immunofluorescence with a panel of nephron segment markers ( Table 1 ) was used to identify the sites of claudin-8 expression along the nephron ( Figs. 5 and 6 ). Claudin-8 was found in the early and late DCT (DCT1 and DCT2), connecting tubule, and the entire cortical and medullary collecting duct. Claudin-8 was also expressed in the late segment of the thin descending limbs of long-looped nephrons (TDL2), where it colocalized with aquaporin-1 (AQP1; Fig. 6 ). Claudin-8 was notably absent from thin ascending limbs, which were identified in the inner medulla by staining with anti-ClC-K antibody ( Fig. 6 ). In occasional distal nephron segments in the cortex, claudin-8 was observed not only at the tight junction but exhibited faint basal membrane staining as well ( Fig. 6 ).

Fig. 5. Nephron segment localization of claudin-8 in the renal cortex. Dual-color immunofluorescence was performed using 933 antibody (green) and antibodies to the following tubule markers (red): apical Na-K-2Cl cotransporter [NKCC2; thick ascending limb of Henle (TALH); A ]; thiazide-sensitive NaCl cotransporter [NCC; early and late distal collecting duct (DCT1/2); B ]; calbindin-D 28k [CBP; connecting tubule (CNT) and cortical collecting duct (CCD; C )]; and H + -ATPase (CCD intercalated cells; D ). Scale bars = 25 µm.

Fig. 6. Nephron segment localization of claudin-8 in the renal outer (OM) and inner (IM) medulla. Dual-color immunofluorescence was performed using 933 antibody (green) and antibodies to the following tubule markers (red): aquaporin-1 (AQP1; proximal tubule and thin descending limb); voltage-dependent chloride channel (ClC-K1 in the inner medulla, thin ascending limb); aquaporin-2 (AQP; collecting duct). Images of the OM demonstrate the transition (t) between the S3 segment of the proximal tubule, where AQP1 is weakly expressed at apical (a) and basolateral (b) membranes and the early segment of the thin descending limbs (TDL1), where AQP1 is strongly expressed. Claudin-8 is absent from both of these tubule segments. In the IM, claudin-8 colocalizes with AQP1 to the late segment of thin descending limbs (TDL2). Claudin-8 is also strongly expressed (arrowheads) in OM and IM collecting ducts (CD) and is absent from thin ascending limbs (asterisks). Scale bars = 25 µm.

Immunolocalization of claudin-7. During the course of this project, an antibody to claudin-7 became commercially available. We therefore used it to compare the intrarenal distribution of claudin-7 to that of claudin-8. Claudin-7 was expressed primarily at the basolateral membrane of distal nephron segments ( Figs. 7 and 8 ), including the connecting tubule and throughout the collecting duct (more strongly in intercalated cells than principal cells). Claudin-7 was expressed at the tight junction only in DCT1 and DCT2, but even in these segments it was also found at low abundance in the basolateral membrane ( Fig. 7 ). Claudin-7 was also found in the TDL2 ( Fig. 8 ). Our findings indicate that claudin-7 and claudin-8 are expressed in the exact same nephron segments but localize largely to different subcellular sites. This was confirmed by costaining the same kidney section with antibodies to both claudin-7 and -8 ( Fig. 9 ).

Fig. 7. Nephron segment localization of claudin-7 in the renal cortex. Dual-color immunofluorescence was performed using the claudin-7 antibody (green) and antibodies to the following tubule markers (red): NCC, (DCT1/2; A ); CBP (DCT2, CNT, and CCD principal cells; B ); Na/Ca exchanger (NX1; DCT2, CNT principal cells; C ); and H + -ATPase (CCD intercalated cells; D ). Claudin-7 is found primarily at the tight junctions (arrowheads) in DCT1/2. In the CNT, it is expressed at the basolateral membrane, more strongly in intercalated cells (ic) than in connecting tubule cells (cnt). In the CCD, it is expressed at the basolateral membrane, more strongly in - and -intercalated cells ( -ic and -ic) than in principal cells (pc). Scale bar = 25 µm.

Fig. 8. Nephron segment localization of claudin-7 in the renal inner medulla. Dual-color immunofluorescence was performed using anti-claudin-7 antibody (green) and antibodies to the following tubule markers (red): AQP1 (thin descending limb); ClC-K (ClC-K1 in the thin ascending limb); and AQP2 (CD principal cells). Claudin-7 localizes to the basolateral membranes of late thin descending limbs (open arrowheads) and of IMCD principal (solid arrowheads) and intercalated (arrows) cells. No colocalization with ClC-K1 was observed. Scale bars = 25 µm.

Fig. 9. Colocalization of claudin-7 and -8 to the same nephron segments. Low-magnification views of the kidney cortex ( A ) and medulla ( B ) are shown. Dual-color immunofluorescence was performed using the 933 anti-claudin-8 antibody (green) and anti-claudin-7 antibody (red).

DISCUSSION

We have used immunofluorescence methods to determine the intrarenal sites of claudin-7 and -8 expression in mice. We believe that our claudin antibodies were specific for several reasons. First, both antibodies were raised against peptide sequences at the COOH terminus of the claudin protein, which is a region with very little homology between the claudin isoforms. Second, each antibody was shown by Western blotting to react with its cognate isoform but not with the other isoform. Third, by Western blotting both antibodies were remarkably clean and reacted predominantly with a single band of the expected size in a mouse kidney membrane protein preparation. Fourth, where possible (for the claudin-8 antibody), control immunostains using nonimmune and peptide-blocked sera were performed and found to be negative.

Differences in claudin-8 localization compared with a prior study. We found that claudin-8 was expressed in the entire ASDN and in the distal portion of the TDL. Our results disagree with those of Kiuchi-Saishin et al. ( 25 ). Like us, they found claudin-8 in DCT and collecting duct, but they also concluded that it was in thin ascending limbs, based on colocalization to tubules staining positively with anti-ClC-K. A possible reason for the discrepancy could be that those investigators mistook ClC-K staining of other distal tubule segments for the thin ascending limb. Anti-ClC-K recognizes both ClC-K1 and ClC-K2 and so has a complex pattern of staining including not only the thin ascending limb ( 41 ) but also the thick ascending limb of Henle, DCT, connecting tubule, corticial collecting duct, and possibly even the outer medullary collecting duct ( 26, 42 ). Our conclusion that claudin-8 is not in thin ascending limbs is based on its absence from tubules that we identified based on several rigorous criteria: the tubules were in the inner medulla, stained positively with anti-ClC-K, did not stain for AQP2, and had the morphological appearance of thin limbs by light microscopy.

On the other hand, we found claudin-8 expression in the thin descending limb, where it colocalized with AQP1-stained tubules, while Kiuchi-Saishin et al. ( 25 ) did not. The thin descending limb consists of at least three morphologically and probably functionally distinct segments in all species except the rabbit ( 23 ): the descending limbs of short-looped nephrons (SDL), the early segments of thin descending limbs of long-looped nephrons (TDL1), which are largely in the inner stripe of outer medulla, and the TDL2, which are confined to the inner medulla. The TDL1 has a very low transepithelial resistance ( 46 ), a 10-fold higher passive Na + permeability than either TDL2 or SDL ( 23 ) and, like the proximal tubule, expresses claudin-2 ( 10 ), which is likely to be a high-conductance paracellular cation pore ( 1, 13 ). Not surprisingly, we found no claudin-8 in the TDL1, which we defined by localization to the inner stripe of outer medulla, strong AQP1 staining, and contiguity with the S3 segment of proximal tubules. We did find claudin-8 in AQP1-stained tubules in the inner medulla, which we believe to be TDL2. We speculate that the absence of colocalization with thin descending limbs reported by Kiuchi-Saishin et al. ( 25 ) may be because their observations were made in the outer medulla.

Possible roles of claudin-8 in the ASDN and TDL2. Our finding that claudin-8 is expressed at the tight junction in the ASDN and TDL2, which are known to be the most electrically tight segments of the nephron, is entirely consistent with its postulated role as a cation barrier ( 48 ). The ASDN has to have a low paracellular permeability to cations so that transtubular gradients for Na +, K +, and H + set up by active, transcellular transport are not dissipated. It is even conceivable that as aldosterone stimulates active transcellular cation transport in these segments, it may concurrently augment the paracellular cation barrier to cope with the steeper transtubular cation gradients, perhaps by regulation of claudin-8.

It is thought that the TDL2 needs a low paracellular ion permeability to generate the inner medullary concentrating gradient. According to the Kokko-Rector-Stephenson passive model ( 30 ), the high permeability of the TDL2 to water and its low permeability to Na + and Cl - allow it to reabsorb water and concentrate NaCl within the tubule. This generates the diffusion gradient for NaCl reabsorption in the highly Na + - and Cl - -permeable thin ascending limb, thus contributing to the high interstitial osmolarity in the inner medulla. Claudin-8 presumably plays a role by contributing to the paracellular cation barrier in the TDL2.

All epithelia that have so far been studied express multiple claudin isoforms ( 2, 8, 12, 25, 34, 44 ). Claudin isoforms interact with each other both within the tight junction strands of the same cell and between the strands of adjacent cells ( 15 ). Thus the paracellular permeability properties of any epithelia probably depend on the precise combination of claudins expressed. Of the other claudin isoforms that have been mapped in the renal tubule, claudin-4 is known to be in the collecting duct, and claudin-3 in DCT and collecting duct ( 25 ). At least five other claudin isoforms have yet to be studied. Thus it is likely that the paracellular cation barrier in the ASDN and TDL2 is constituted not solely by claudin-8, but by a combination of claudin-8 with multiple other claudin isoforms.

A prediction of our results is that inherited or acquired inactivation or loss of claudin-8 expression in the kidney could potentially lead to salt wasting, hyperkalemic metabolic acidosis, and a urinary concentrating defect. Circulating renin and aldosterone concentrations would likely be secondarily elevated, and the electrolyte disorders would be resistant to exogenous administration of mineralocorticoids. These are classic features (except the concentrating defect) of pseudohypoaldosteronism type I ( 18 ). While autosomal recessive and dominant forms of pseudohypoaldosteronism type I have been identified that are due to mutations in the epithelial Na + channel ( 7 ) and the mineralocorticoid receptor ( 16 ), respectively, there is genetic heterogeneity in the autosomal dominant form ( 43 ), and claudin-8 might be a candidate gene for some families with this form of the disease. Hyperkalemic distal renal tubular acidosis, renal salt wasting, and urinary concentrating defects are also characteristic of certain chronic tubulointerstitial diseases with a propensity for the inner medulla, such as sickle cell disease ( 3 ) and obstructive uropathy ( 4 ). We speculate that acquired defects in claudin-8-mediated barrier function may underlie some of these disorders.

Basolateral localization of claudin-7. We found that claudin-7 was expressed in the same nephron segments as claudin-8, but was largely confined to the basolateral membrane. There are now many reports that claudins can localize not only to the tight junction, but are often additionally found at the basolateral membrane ( 17, 34 ). Furthermore, there is precedent for claudins being expressed exclusively at the basolateral membrane. For example, claudins-3 and -5 are located only at the basolateral surface in the gastric epithelium ( 34 ). The only other tissue in which claudin-7 expression has so far been examined is the airway epithelium, where it is also expressed at the lateral membrane ( 8 ). Thus our finding of claudin-7 localization at the basolateral membrane of renal tubules is likely to be correct and not simply an artifact of antibody cross-reactivity.

There are several possible reasons why claudins might be located at the basolateral membrane instead of, or in addition to, localization at the tight junction. Basolateral claudins might simply represent a storage pool capable of being recruited to the apical junctional complex when needed. Alternatively, basolateral claudins may have some other novel function such as participating in cell-cell or cell-matrix adhesion or signaling. The role of claudin-7 in basolateral membranes of tight distal nephron segments is unknown.

GRANTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants RO1-DK-062283 (to A. S. L. Yu) and DK-48522 (to the USC Center for Liver Diseases for the Microscopy Subcore) and by a Norman S. Coplon Extramural Grant from Satellite Research.

ACKNOWLEDGMENTS

We are grateful to the many investigators listed in Table 1 that have generously donated nephron segment-specific antibodies to us over the years for this and other studies and to Alicia McDonough for helpful discussions.

【参考文献】
  Amasheh S, Meiri N, Gitter AH, Schoneberg T, Mankertz J, Schulzke JD, and Fromm M. Claudin-2 expression induces cation-selective channels in tight junctions of epithelial cells. J Cell Sci 115: 4969-4976, 2002.

Ban Y, Dota A, Cooper LJ, Fullwood NJ, Nakamura T, Tsuzuki M, Mochida C, and Kinoshita S. Tight junction-related protein expression and distribution in human corneal epithelium. Exp Eye Res 76: 663-669, 2003.

Batlle D, Itsarayoungyuen K, Arruda JA, and Kurtzman NA. Hyperkalemic hyperchloremic metabolic acidosis in sickle cell hemoglobinopathies. Am J Med 72: 188-192, 1982.

Batlle DC, Arruda JA, and Kurtzman NA. Hyperkalemic distal renal tubular acidosis associated with obstructive uropathy. N Engl J Med 304: 373-380, 1981.

Brown D, Hirsch S, and Gluck S. Localization of a proton-pumping ATPase in rat kidney. J Clin Invest 82: 2114-2126, 1988.

Campean V, Kricke J, Ellison D, Luft FC, and Bachmann S. Localization of thiazide-sensitive Na + -Cl - cotransport and associated gene products in mouse DCT. Am J Physiol Renal Physiol 281: F1028-F1035, 2001.

Chang SS, Grunder S, Hanukoglu A, Rosler A, Mathew PM, Hanukoglu I, Schild L, Lu Y, Shimkets RA, Nelson-Williams C, Rossier BC, and Lifton RP. Mutations in subunits of the epithelial sodium channel cause salt wasting with hyperkalaemic acidosis, pseudohypoaldosteronism type 1. Nat Genet 12: 248-253, 1996.

Coyne CB, Gambling TM, Boucher RC, Carson JL, and Johnson LG. Role of claudin interactions in airway tight junctional permeability. Am J Physiol Lung Cell Mol Physiol 285: L1166-L1178, 2003.

Dowland LK, Luyckx VA, Enck AH, Leclercq B, and Yu ASL. Molecular cloning and characterization of an intracellular chloride channel in the proximal tubule cell line, LLC-PK 1. J Biol Chem 275: 37765-37773, 2000.

Enck AH, Berger UV, and Yu AS. Claudin-2 is selectively expressed in proximal nephron in mouse kidney. Am J Physiol Renal Physiol 281: F966-F974, 2001.

Ethier JH, Kamel KS, Magner PO, Lemann J, and Halperin ML. The transtubular potassium concentration in patients with hypokalemia and hyperkalemia. Am J Kid Dis 15: 309-315, 1990.

Florian P, Amasheh S, Lessidrensky M, Todt I, Bloedow A, Ernst A, Fromm M, and Gitter AH. Claudins in the tight junctions of stria vascularis marginal cells. Biochem Biophys Res Commun 304: 5-10, 2003.

Furuse M, Furuse K, Sasaki H, and Tsukita S. Conversion of zonulae occludentes from tight to leaky strand type by introducing claudin-2 into Madin-Darby canine kidney I cells. J Cell Biol 153: 263-272, 2001.

Furuse M, Sasaki H, Fujimoto K, and Tsukita S. A single gene product, claudin-1 or -2, reconstitutes tight junction strands and recruits occludin in fibroblasts. J Cell Biol 143: 391-401, 1998.

Furuse M, Sasaki H, and Tsukita S. Manner of interaction of heterogeneous claudin species within and between tight junction strands. J Cell Biol 147: 891-903, 1999.

Geller DS, Rodriguez-Soriano J, Vallo Boado A, Schifter S, Bayer M, Chang SS, and Lifton RP. Mutations in the mineralocorticoid receptor gene cause autosomal dominant pseudohypoaldosteronism type I. Nat Genet 19: 279-281, 1998.

Gregory M, Dufresne J, Hermo L, and Cyr D. Claudin-1 is not restricted to tight junctions in the rat epididymis. Endocrinology 142: 854-863, 2001.

Hanukoglu A. Type I pseudohypoaldosteronism includes two clinically and genetically distinct entities with either renal or multiple target organ defects. J Clin Endocrinol Metab 73: 936-944, 1991.

Hulter HN, Ilnicki LP, Harbottle JA, and Sebastian A. Impaired renal H + secretion and NH3 production in mineralocorticoid-deficient glucocorticoid-replete dogs. Am J Physiol Renal Fluid Electrolyte Physiol 232: F136-F146, 1977.

Imai M. Function of the thin ascending limb of Henle of rats and hamsters perfused in vitro. Am J Physiol Renal Fluid Electrolyte Physiol 232: F201-F209, 1977.

Imai M and Nakamura R. Function of distal convoluted and connecting tubules studied by isolated nephron fragments. Kidney Int 22: 465-472, 1982.

Imai M, Taniguchi J, and Yoshitomi K. Transition of permeability properties along the descending limb of long-loop nephron. Am J Physiol Renal Fluid Electrolyte Physiol 254: F323-F328, 1988.

Imai M and Yoshitomi K. Heterogeneity of the descending thin limb of Henle's loop. Kidney Int 38: 687-694, 1990.

Kaplan MR, Plotkin MD, Lee WS, Xu ZC, Lytton J, and Hebert SC. Apical localization of the Na-K-Cl cotransporter, rBSC1, on rat thick ascending limbs. Kidney Int 49: 40-47, 1996.

Kiuchi-Saishin Y, Gotoh S, Furuse M, Takasuga A, Tano Y, and Tsukita S. Differential expression patterns of claudins, tight junction membrane proteins, in mouse nephron segments. J Am Soc Nephrol 13: 875-886, 2002.

Kobayashi K, Uchida S, Mizutani S, Sasaki S, and Marumo F. Intrarenal and cellular localization of CLC-K2 protein in the mouse kidney. J Am Soc Nephrol 12: 1327-1334, 2001.

Liddle GW. Aldosterone antagonists. Arch Intern Med 102: 998-1004, 1958.

Loffing J, Loffing-Cueni D, Valderrabano V, Klausli L, Hebert SC, Rossier BC, Hoenderop JG, Bindels RJ, and Kaissling B. Distribution of transcellular calcium and sodium transport pathways along mouse distal nephron. Am J Physiol Renal Physiol 281: F1021-F1027, 2001.

Loffing J, Summa V, Zecevic M, and Verrey F. Mediators of aldosterone action in the renal tubule. Curr Opin Nephrol Hypertens 10: 667-675, 2001.

Masilamani S, Knepper MA, and Burg MB. Urine concentration and dilution. In: Brenner and Rector's The Kidney (6th ed.), edited by Brenner BM. Philadelphia, PA: Saunders, 2000, p. 595-635.

Morita K, Furuse M, Fujimoto K, and Tsukita S. Claudin multigene family encoding four-transmembrane domain protein components of tight junction strands. Proc Natl Acad Sci USA 96: 511-516, 1999.

Nielsen S, DiGiovanni SR, Christensen EI, Knepper MA, and Harris HW. Cellular and subcellular immunolocalization of vasopressin-regulated water channel in rat kidney. Proc Natl Acad Sci USA 90: 11663-11667, 1993.

Plotkin MD, Kaplan MR, Verlander JW, Lee WS, Brown D, Poch E, Gullans SR, and Hebert SC. Localization of the thiazide sensitive Na-Cl cotransporter, rTSC1 in the rat kidney. Kidney Int 50: 174-183, 1996.

Rahner C, Mitic LL, and Anderson JM. Heterogeneity in expression and subcellular localization of claudins 2, 3, 4, and 5 in the rat liver, pancreas, and gut. Gastroenterology 120: 411-422, 2001.

Reilly RF, Shugrue CA, Lattanzi D, and Biemesderfer D. Immunolocalization of the Na + /Ca 2+ exchanger in rabbit kidney. Am J Physiol Renal Fluid Electrolyte Physiol 265: F327-F332, 1993.

Reuss L. Tight junction permeability to ions and water. In: Tight Junctions, edited by Cereijido M. Boca Raton, FL: CRC, 1991, p. 49-66.

Sabolic I, Valenti G, Verbavatz JM, Van Hoek AN, Verkman AS, Ausiello DA, and Brown D. Localization of the CHIP28 water channel in rat kidney. Am J Physiol Cell Physiol 263: C1225-C1233, 1992.

Sands JM, Nonoguchi H, and Knepper MA. Hormone effects on NaCl permeability of rat inner medullary collecting duct. Am J Physiol Renal Fluid Electrolyte Physiol 255: F421-F428, 1988.

Stokes JB. Ion transport by the cortical and outer medullary collecting tubule. Kidney Int 22: 473-484, 1982.

Tsukita S and Furuse M. Pores in the wall: claudins constitute tight junction strands containing aqueous pores. J Cell Biol 149: 13-16, 2000.

Uchida S, Sasaki S, Nitta K, Uchida K, Horita S, Nihei H, and Marumo F. Localization and functional characterization of rat kidney-specific chloride channel, ClC-K1. J Clin Invest 95: 104-113, 1995.

Vandewalle A, Cluzead F, Bens M, Kieferle S, Steinmeyer K, and Jentsch TJ. Localization and induction by dehydration of ClC-K chloride channels in the rat kidney. Am J Physiol Renal Physiol 272: F678-F688, 1997.

Viemann M, Peter M, Lopez-Siguero JP, Simic-Schleicher G, and Sippell WG. Evidence for genetic heterogeneity of pseudohypoaldosteronism type 1: identification of a novel mutation in the human mineralocorticoid receptor in one sporadic case and no mutations in two autosomal dominant kindreds. J Clin Endocrinol Metab 86: 2056-2059, 2001.

Wang F, Daugherty B, Keise LL, Wei Z, Foley JP, Savani RC, and Koval M. Heterogeneity of claudin expression by alveolar epithelial cells. Am J Respir Cell Mol Biol 29: 62-70, 2003.

Yang L, Leong PK, Chen JO, Patel N, Hamm-Alvarez SF, and McDonough AA. Acute hypertension provokes internalization of proximal tubule NHE3 without inhibition of transport activity. Am J Physiol Renal Physiol 282: F730-F740, 2002.

Yoshitomi K and Imai M. Electrophysiological characterization of upper portion of descending limb of long-looped nephron. Am J Physiol Renal Fluid Electrolyte Physiol 260: F311-F316, 1991.

Yu AS. Claudins and epithelial paracellular transport: the end of the beginning. Curr Opin Nephrol Hypertens 12: 503-509, 2003.

Yu AS, Enck AH, Lencer WI, and Schneeberger EE. Claudin-8 expression in MDCK cells augments the paracellular barrier to cation permeation. J Biol Chem 278: 17350-17359, 2003.

作者单位：Division of Nephrology, Department of Medicine, and Department of Physiology and Biophysics, University of Southern California Keck School of Medicine, Los Angeles, California 90033