Deletion of Cdh16 Ksp-cadherin leads to a developmental delay in the ability to maximally concentrate urine in mouse

Abstract

Ksp-cadherin (cadherin-16) is an atypical member of the cadherin superfamily of cell adhesion molecules that is ubiquitously expressed on the basolateral membrane of epithelial cells lining the nephron and the collecting system of the mammalian kidney. The principal aim of the present study was to determine if Ksp-cadherin played a critical role in the development and maintenance of the adult mammalian kidney by generating and evaluating a mouse line deficient in Ksp-cadherin. Ksp-null mutant animals were viable and fertile, and kidneys from both neonates and adults showed no evidence of structural abnormalities. Immunolocalization and Western blot analyses of Na⁺-K⁺-ATPase and E-cadherin indicated that Ksp-cadherin is not essential for either the genesis or maintenance of the polarized tubular epithelial phenotype. Moreover, E-cadherin expression was not altered to compensate for Ksp-cadherin loss. Plasma electrolytes, total CO₂, blood urea nitrogen, and creatinine levels were also unaffected by Ksp-cadherin deficiency. However, a subtle but significant developmental delay in the ability to maximally concentrate urine was detected in Ksp-null mice. Expression analysis of the principal proteins involved in the generation of the corticomedullary osmotic gradient and the resultant movement of water identified misexpression of aquaporin-2 in the inner medullary collecting duct as the possible cause for the inability of young adult Ksp-cadherin-deficient animals to maximally concentrate their urine. In conclusion, Ksp-cadherin is not required for normal kidney development, but its absence leads to a developmental delay in maximal urinary concentrating ability.

NEW & NOTEWORTHY Ksp-cadherin (cadherin-16) is an atypical member of the cadherin superfamily of cell adhesion molecules that is ubiquitously expressed on the basolateral membrane of epithelial cells lining the nephron and the collecting system. Using knockout mice, we found that Ksp-cadherin is in fact not required for kidney development despite its high and specific expression along the nephron. However, its absence leads to a developmental delay in maximal urinary concentrating ability.

INTRODUCTION

Ksp-cadherin is an atypical member of the cadherin superfamily of Ca²⁺-dependent cell adhesion molecules that is preferentially expressed in the kidney, where it is localized to the basolateral membrane of epithelial cells lining the tubular nephron and the collecting system (1, 2). Ksp-cadherin together with LI-cadherin forms the so-called “7 D” subgroup of the cadherin superfamily (3). Members of this subgroup have seven extracellular cadherin repeat domains, a single membrane-spanning domain, and a truncated cytoplasmic domain that lacks the canonical interaction sites for catenin binding found in classical cadherins like E-cadherin or N-cadherin. Unlike Ksp-cadherin, LI-cadherin is not expressed in the mammalian metanephric kidney (4). Cadherins have been implicated in a wide range of biological functions including cell-cell adhesion, tissue polarization, tissue maintenance, tumorigenesis, tissue remodeling, and organ development (for a review, see Ref. 5).

Ksp-cadherin is expressed relatively late in the development of the metanephric kidney (1). In the early stages of renal development, Ksp-cadherin expression is confined to the basolateral membrane of the ureteric bud and its derivatives, where it is largely coexpressed with E-cadherin. Ksp-cadherin has not been detected in the mesenchymal condensate, renal vesicle, or early proximal or distal tubular anlagen. It is not until a recognizable capillary loop-stage glomerulus is clearly evident that Ksp-cadherin expression is detected in the tubular nephron. In the mature mouse kidney, Ksp-cadherin is coexpressed with E-cadherin on the basolateral membrane of all segments of the tubular nephron and the collecting system. Cell adhesion assays with transfected cells have confirmed that Ksp-cadherin is capable of mediating mechanical adhesion via homotypic interactions in a manner similar to that proposed for E-cadherin (6), but as with other members of the cadherin superfamily, the actual function of Ksp-cadherin in the kidney in vivo is yet to be determined.

Given the uncertain function of Ksp-cadherin and the ubiquitous nature of its expression in the tubular epithelium of the human, mouse, and rabbit kidney, the principal aim of the present study was to generate an animal model deficient in Ksp-cadherin and to use that model to determine if Ksp-cadherin plays a critical role in either the development or maintenance of the mammalian kidney. Ksp-null mutant animals were viable and fertile, and kidneys from both null mutant neonates and adults showed no evidence of structural abnormalities. Plasma electrolyte, blood urea nitrogen (BUN), and creatinine values were unaffected by loss of Ksp-cadherin, but null mutant animals displayed a subtle but significant developmental delay in their ability to maximally concentrate urine. A detailed analysis of the principal proteins involved in salt and water movement in the kidney suggests that the defect may be linked to misexpression of aquaporin-2 (AQP2) in the inner medullary collecting duct (IMCD) of young adult mutant animals.

MATERIALS AND METHODS

Generation of Ksp-Cadherin-Null Mice

The targeting construct was designed to introduce a premature in-frame stop codon 12 nucleotides downstream from the native translation initiation site of Ksp-cadherin (Cdh16) (Fig. 1A). The goal was to preserve the original Kozak consensus sequence and native translation initiation site to reduce the likelihood of translation proceeding from alternate downstream initiation sites and the generation of Ksp-cadherin protein fragments that may possess confounding functional activity. In all, 1,818 nucleotides spanning exons 2−6 would be deleted if homologous recombination was successful and replaced with a neomycin selection cassette oriented in the opposite direction.

Figure 1.

Generation of Ksp-cadherin-null mice. A: targeting construct and predicted homologous recombination event showing locations for restriction enzyme digests (EcoR1, Pst1, and Not1), Southern 5′ (PB-1) and 3′ (PB-2) DNA probes, and PCR screening primers (W1/W2, D1/D2, H1/H2, and K1/K2). B: confirmation of homologous recombination of the 5′ end of the targeting construct with wild-type (WT) Ksp-cadherin by Southern blot analysis with DNA probe PB-1. C: confirmation of homologous recombination of the 3′ arm of the targeting construct with WT Ksp-cadherin and successful excision of the diphtheria toxin A-negative selection cassette by Southern blot analysis with DNA probe PB-2. D: PCR genotyping strategy using the combination of primer pairs W1/W2 and K1/K2 to simultaneously detect WT and mutant Ksp-cadherin alleles, respectively, in genomic DNA isolated from 8- to 12-day-old mice. E: confirmation of ablation of the Ksp-cadherin gene product by Western blot analysis of whole kidney homogenates with an anti-Ksp-cadherin polyclonal antibody (1).

Ksp-cadherin homologous regions of the targeting vector were generated by PCR from mouse 129/SvJ genomic DNA (Jackson Laboratories, Bar Harbor, ME). The 5′ homologous arm is 1,445 nt in length, begins just upstream of the first exon, introduces a NotI restriction site, and was generated by PCR between 5′- CCAAGCGGCCGCTCCCAGTCCTTTCCTTAGCTCCAC-3′ and 5′- CCGGTCTAGACAGAGATCATGGTCAAGGCAGAC-3′. The 3′ homologous arm is 5,273 nt in length, spans exons 6−15, and was generated by PCR between 5′- CACAATCGATTCCGACCTTCGCTTCCACATTC-3′ and 5′- CACAGTCGACTGTAGGGACCATTCCTGTTGGC-3′. The positive neomycin selection cassette (PGK-Neo), 5′ and 3′ Ksp-cadherin homologous arms, and negative selection diphtheria toxin A cassette (DT) were assembled on a pBluescript II SK backbone (Stratagene, La Jolla, CA). The DT cassette was ligated into the 3′ end of the construct just outside of the 3′ homologous arm.

NotI-linearized targeting vector was electroporated into 129/SvJ ES cells by the Yale Animal Genomics Service. G418-resistant colonies were screened by PCR with 5′-specific (H1: 5′- ACATGAGACAGGTCAGAGCCTCAC-3′ and H2: 5′- AATGGAAGGATTGGAGCTACGG-3′) and 3′-specific (D1: 5′- ATCGCCTGACACGATTTCCTGC-3′ and D2: 5′- TCGTACCACGGGACTAAACCTG-3′) primer pairs to detect homologous recombination of the targeting construct with the wild-type (WT) allele. DNA samples from colonies giving a 1.7-kb product with the 5′-H1/H2 primer pair and no product with the DT-specific D1/D2 primer pair were digested independently with either EcoRI or PstI and then probed by Southern blot with an external 5′ probe (PB-1) or an internal 3′ probe (PB-2), respectively, to confirm homologous recombination. Successful homologous recombination, as indicated by labeling of a 4.6-kb EcoRI restriction fragment with probe PB-1 (Fig. 1B) and a 3.5-kb PstI restriction fragment with probe PB-2 (Fig. 1C), was confirmed in only one embryonic stem (ES) cell clone.

C57Bl/6J blastocysts were injected with the positive ES cell clone and transferred to pseudopregnant CD-1 female mice by the Yale Animal Genomics Service. Male chimeric offspring were mated with 129/SvJ females through eight rounds of selective breeding to generate congenic heterozygotes containing a single copy of the mutant allele. Heterozygotes were crossed, and homozygous null mutants were born at the expected Mendelian frequency, indicating that the null mutation was not embryonic lethal. Offspring were initially genotyped by PCR and Southern blot analysis. However, once verified congenic null and WT lines were established, routine genotyping was performed by PCR with a cocktail containing primer pairs W1/W2 (5′- TCAGGAGACTCAAACACGGCAG-3′ and 5′- TGGAAGCGAAGGTCGGAGTTAG-3′) and K1/K2 (5′- GTCTTCTGTTGCTGAAGGAATCGTC-3′ and 5′- GGTGGGATTAGATAAATGCCTGCTC-3′) to identify WT and mutant Ksp-cadherin alleles simultaneously in a single PCR (Fig. 1D). Western blot analysis of whole kidney homogenates with an antibody directed against the COOH-terminal region of Ksp-cadherin confirmed complete loss of expression of Ksp-cadherin in homozygous null mice (Fig. 1E).

Measurement of Plasma Electrolytes, BUN, Creatinine, and Urine Osmolality

Plasma samples were collected from 12-wk-old male WT and Ksp-cadherin-null (Ksp-null) mice by retroorbital bleed. Plasma concentrations of Na⁺, K⁺, Cl⁻, and total CO₂ were determined by the Yale Mouse Metabolic Phenotyping Center using a Roche Diagnostics COBAS MIRA system. Plasma creatinine concentrations were determined by targeted liquid chromatography multiple-reaction monitoring (LC-MRM) on a 4000 QTRAP mass spectrometer (Yale George M. O’Brien Kidney Center). Plasma BUN values were determined spectrophotometrically with a diacetylmonoxime-based assay kit (Stanbio Laboratory, Boerne, TX). Urine osmolality of spot urine samples was measured on a Wescor 5100c vapor pressure osmometer (Wescor, Logan, UT).

Tissue Preparations and Western Blot Analysis

Total homogenates were prepared from kidneys of 12-wk-old adult male WT and Ksp-null mutant mice. Kidneys were homogenized with a Thomas-style homogenizer and a serrated pestle in PBS containing protease (Halt, Thermo) and phosphatase inhibitors (PhosStop, Roche) and stored at −70°C. Homogenates were solubilized in SDS-sample buffer and subjected to standard SDS-PAGE, electrophoretic transfer, and Western blot analysis as previously described (7). Briefly, transfers were stained with Ponceau S or Coomassie brilliant blue, digitized on a flatbed scanner, and then subjected to densitometric image analysis with ImageJ (8) to provide an estimate of relative protein loading for each lane on the transfer. Protein transfers were then blocked in blotting buffer (1× PBS containing 5% dried milk and 0.1% Tween 20) for 2 h at room temperature and then incubated with primary antibody diluted in blotting buffer for 16 h at 4°C. Transfers were then washed five times (10 min each wash) with blotting buffer at room temperature, incubated with the appropriate horseradish peroxidase-conjugated secondary antibody diluted in blotting buffer for 1 h at room temperature, washed five times (10 min each wash) with blotting buffer at room temperature, and finally incubated with an enhanced chemiluminescence detection reagent (Clarity, Bio-Rad) in accordance with the manufacturer’s protocol. Antibody labeling was captured on film. The films were digitized on a flatbed scanner and subjected to densitometric analysis. Primary antibodies and the respective dilutions used in the study are provided in Supplemental Table S1 (all Supplemental material is available at https://doi.org/10.6084/m9.figshare.13119911). The secondary antibodies used were affinity-purified horseradish peroxidase-conjugated donkey anti-mouse, donkey anti-rabbit, and bovine anti-goat (Jackson Immunoresearch, 1:20,000 dilution).

Renal Histology and Immunocytochemistry

Kidneys destined for paraffin embedding were dissected free of extraneous tissue, decapsulated, bisected, and immersion fixed in Bouin’s solution for 24 h before being embedded. Paraffin sections were used for hematoxylin and eosin, endomucin, and lotus lectin staining. Four-micrometer paraffin sections were deparaffinized and subjected to antigen retrieval [10 mM Tris, 1 mM EDTA, 0.05% Tween 20 (pH 9), autoclaved 121°C 30-min liquid cycle] before being blocked [5% donkey serum, 2.5% BSA, and 0.5% Triton X-100 in Tris-buffered saline (pH 7.5) for 1 h at room temperature] and overnight incubation in primary antibody (rat anti-endomucin, AB106100, Abcam, 1:250). Slides were then washed and incubated in the appropriate Alexa Fluor-conjugated secondary antibody (Life Technologies) and fluorescein-conjugated tetragonolobus lectin (LTL; FL-1321, Vector Laboratories, 1:100). After being washed, slides were mounted and imaged on a Nikon A1R confocal microscope. Image quantification was performed using Fiji software (9).

Kidneys destined for epon embedding were cleared with PBS; perfusion fixed with 2% paraformaldehyde, 750 mM lysine, and 10 mM sodium periodate (PLP) in phosphate buffer (pH 7.4); and then embedded in epon as previously described (10). Epon-embedded sections were used for immunocytochemical labeling experiments requiring a higher degree of cellular resolution. One-micrometer PLP-fixed epon sections were etched, subjected to antigen retrieval, and then labeled with the appropriate primary and fluorochrome-labeled secondary antibodies (Alexa Fluor 488 and 594, Molecular Probes). The primary antibodies used for immunolocalization are provided in Supplemental Table S1.

Histopathological Analysis

Hematoxylin-and-eosin-stained sections of the thyroid and kidney were evaluated through the Yale Comparative Pathology Research Core (Department of Comparative Medicine) by a researcher (C. J. Booth) blind to experimental manipulation and genotype. Images were examined on an Axio Imager A1 microscope (Zeiss Microsystems), photographed using an Axiocam MRC 5 (Zeiss Microsystems) camera running Axiovision 40 × 64 version 4.9.1.0 (Zeiss Microsystems), and optimized using Adobe Photoshop Creative Cloud version 21.1.1 (San Jose, CA).

Apical versus diffuse cytoplasmic labeling of AQP2 in IMCD cells was evaluated by a researcher (R. B. Thomson) blind to subject and genotype. ImageJ (8) was used to generate pixel intensity plot profiles along lines manually drawn from the basolateral border to the apical border of individual IMCD cells. Distinct apical localization was defined as an increase in pixel intensity at the apical border of 2× or greater relative to the cytoplasm. In total, 55–88 collecting duct (CD) cross sections per animal (from multiple random tissue sections per animal, 4 animals per group) were evaluated to arrive at the percentage of total IMCDs with distinct apical AQP2 labeling.

Quantitative Transcript Analysis

Kidneys were harvested, dissected free of extraneous tissue, decapsulated, snap frozen in liquid N₂, and stored at −80°C. Crude total RNA was isolated from whole kidneys with TRIzol reagent (Life Technologies) and then further purified with RNeasy Plus Mini Kits (Qiagen). cDNA was reverse transcribed from 3 μg of total RNA per reaction with SuperScript III First-Strand Synthesis SuperMix for the quantitative RT-PCR kit (Invitrogen). Transcript abundance analysis was performed by quantitative real-time PCR (qPCR) using Bimake SYBR Green and a Bio-Rad CFX96 Real-Time System with a C1000 thermal cycler. Primer efficacies were confirmed by template serial dilution, and all primers are provided in Supplemental Table S2. Transcript abundance was normalized to expression of hypoxanthine guanine phosphoribosyltransferase with the comparative threshold cycle (dCt) method and reported as 2^−dCt.

Statistical Analysis

All values are presented as means ± SE. Quantitative assessments of Western blot results were performed by densitometry, and all values were normalized to the respective total protein as assayed by Ponceau or Coomassie brilliant blue staining before statistical analysis with an unpaired two-tailed Student’s t test (GraphPad Prism).

Animals

All animal protocols and procedures used in this study were approved by the Yale University Institutional Animal Care and Use Committee. Mice were euthanized by intraperitoneal injection of pentobarbital sodium before all tissue procurements. All mice were maintained under a standard light-dark cycle with free access to standard laboratory chow and, unless otherwise noted, group housed on ground corn cob bedding with free access to water. Mice used for the urine concentrating study were singly housed in standard cages on rolled cellulose bedding (Enrich-a-Nest) for 1 wk before water deprivation. Water was withheld for 24 h from 12:00 PM to 12:00 PM the following day.

RESULTS

Generation of Ksp-Cadherin-Null Mice

The 129/SvJ homozygous null mutants had no overt or obvious phenotypic abnormalities. They were fertile and their body weights, posture, grooming, and rates of mortality were indistinguishable from those of WT littermates. Preliminary evaluations of renal histology and plasma electrolytes of 12-wk-old male Ksp-cadherin-null mutants (data not shown) failed to disclose obvious differences with that of an age- and sex-matched cohort of WT mice, suggesting that renal function in adult 129/SvJ mice was not compromised by the lack of Ksp-cadherin.

To address the possibility that genetic background could potentially be masking the influence of the Ksp-null mutation, the Ksp-null allele was transferred from the 129/SvJ mouse strain to the C57Bl/6J mouse strain. Homozygous male Ksp-null 129/SvJ mice were crossed with female C57Bl/6J mice. The progeny were subjected to eight rounds of selective breeding to generate congenic heterozygote individuals. Congenic heterozygotes were then crossed to generate homozygous Ksp-null and Ksp-wt (WT) animals to be used as founders for the respective Ksp-null and Ksp-wt C57Bl/6J mouse lines. Genotypes of the respective mouse lines were confirmed by PCR of tail biopsy genomic DNA with primers W1/W2 and K1/K2, as shown in Fig. 1, A and D. All subsequent data reported in the present study were obtained with homozygous Ksp-null and Ksp-wt male mice bred independently on the C57Bl/6J genetic background.

Phenotypic Analysis of C57Bl/6J Ksp-Null Mutants

The objective of the present study was to determine if Ksp-cadherin specifically and directly played an essential role in the development, maintenance, and function of the mouse kidney. Ksp-cadherin was originally described as having a kidney-specific distribution in the adult mouse (2), but in 2007, Calì et al. (11) reported that they had also identified Ksp-cadherin on the plasma membrane of mouse thyrocytes. The well-documented link between thyroid function and the development and function of the mammalian kidney (12) raised the distinct possibility that a null mutation of Ksp-cadherin could indirectly affect the kidney through its action on the thyroid independent of its null expression in the kidney.

To address this possibility, we performed an assessment of the impact of Ksp-cadherin deletion on the mouse thyroid before investigating potential effects on the kidney. Circulating total thyroxine levels measured in plasma from 12-wk-old male Ksp-null and WT animals were not significantly different (Supplemental Fig. S1A). Likewise, a blinded evaluation of hematoxylin-and-eosin-stained paraffin-embedded thyroid/parathyroid glands from the same animals failed to identify structural abnormalities in the thyroids from Ksp-null animals (Supplemental Fig. S1B). These results strongly suggested that Ksp-cadherin deletion had no effect on the thyroid of mutant animals and that any potential renal phenotype is likely due to Ksp-cadherin deletion in the kidney per se and not a consequence of altered thyroid function.

As previously observed with the 129/SvJ genetic background, adult C57Bl/6J Ksp-null mice did not appear overtly different from age- and sex-matched WT controls, and adults from both groups had similar levels of fertility, mortality, and longevity. Values for plasma electrolytes, total CO₂, BUN, and creatinine were also not significantly different between WT and Ksp-null 12-wk-old adult mice (Table 1), suggesting that under baseline conditions, overall renal function in adult mice is not compromised by loss of Ksp-cadherin.

Table 1.

Plasma electrolyte, blood urea nitrogen, and creatinine values and urine osmolality for 12-wk-old adult male wild-type and Ksp-null mice

	n	Wild-Type	Null	P Value
Na+, mM	9	145.1 ± 0.6	145.9 ± 0.5	0.2640
K+, mM	9	3.67 ± 0.09	3.55 ± 0.05	0.2395
Cl−, mM	9	112.2 ± 0.5	112.5 ± 0.3	0.4320
Total CO2, mM	9	26.09 ± 0.55	25.63 ± 0.32	0.4804
Blood urea nitrogen, mg/dL	12	25.63 ± 1.39	24.08 ± 0.49	0.2809
Creatinine, mg/dL	10	0.040 ± 0.002	0.035 ± 0.003	0.2020
Urine osmolality, mmol/kg	5	1482 ± 103	899 ± 121	0.0063*

Values are means ± SE. n is the number of animals in each group. *Significantly different by t test at P < 0.05.

Ksp-cadherin is expressed relatively late in renal development. It is not until a functional capillary loop-stage glomerulus is clearly evident that Ksp-cadherin is stably expressed on the basolateral membrane of all tubular epithelial cells of the nephron and the collecting system (1). Because the nephron is largely formed before Ksp-cadherin expression is detected, we did not expect that the null mutation of Ksp-cadherin would result in renal agenesis or extreme renal dysmorphia. Rather, we considered it more likely that the lack of Ksp-cadherin might result in alterations in the manifestation of the final stages of renal development and/or in the maintenance of the terminally differentiated renal phenotype.

To assess the potential impact of the Ksp-cadherin null mutation on renal development, we performed a detailed histological evaluation of 1-μm epon sections prepared from P1 neonatal kidneys from both Ksp-wt and Ksp-null animals. Deletion of the Ksp-cadherin allele had no overt effect on the structure of the neonatal mouse kidney. The nephrogenic zones of P1 WT and Ksp-null kidneys were of comparable size and contained similar densities of the respective nephron, CD, and glomerular precursors. S-shaped bodies in Ksp-null animals showed clear differentiation of proximal and distal anlage and the vascular cleft and were indistinguishable from similarly staged structures in WT animals (Fig. 2).

Figure 2.

Phase-contrast image of S-shaped bodies from P1 neonatal wild-type and Ksp-cadherin-deficient mice. The arrow indicates a vascular cleft. Bar = 25 μm. DA, distal anlage; PA, proximal anlage; UD, ureteric duct.

We also examined the expression of two representative proteins, E-cadherin and Na⁺-K⁺-ATPase, that are known to display polarized epithelial localization at temporally distinct points in renal development (13). In the nephrogenic zone of neonatal WT mice, E-cadherin is highly expressed on the basolateral membrane of all clearly recognizable epithelial structures including renal vesicles, comma- and S-shaped bodies, and the ureteric duct (14). Na⁺-K⁺-ATPase, on the other hand, is expressed at very low levels in the tubular precursors in the nephrogenic zone (13) and is typically not expressed at levels above background until after expression of Ksp-cadherin is detected and the epithelium is relatively well differentiated (Supplemental Fig. S2). In kidney sections from PLP-fixed P1 neonatal mice, Ksp-cadherin deficiency had no effect on immunolocalization or the level of expression of either E-cadherin or Na⁺-K⁺-ATPase (Fig. 3), and both proteins showed staining patterns similar to those described in Refs. 14 and 15. Taken together, these results indicated that Ksp-cadherin is not essential for the development of either the nephron or the collecting system and that it does not appear to play a role in the establishment of the polarized epithelium.

Figure 3.

E-cadherin and Na⁺-K⁺-ATPase immunolocalization in the nephrogenic region of P1 neonatal wild-type and Ksp-cadherin-deficient mice. A and B: E-cadherin (antibody 20874-1-AP) staining of ureteric ducts in wild-type and Ksp-null neonates, respectively. Bar = 40 μm. C and D: Na⁺-K⁺-ATPase (antibody A5) staining of ureteric ducts in wild-type and Ksp-null neonates, respectively. Bar = 50 μm.

Kidneys from 12-wk-old adult Ksp-null and WT mice also exhibited no discernible differences in the structural organization of nephrons and the collecting system, the delineation of corticomedullary borders, or the general proportion of cortical mass to medullary mass (Fig. 4). Estimates of proximal tubular density as evidenced by staining with Lotus tetragonolobus agglutinin (Supplemental Fig. S3) were also not significantly different between the two groups. Again, these results strongly suggested that Ksp-cadherin is not critically important for any stage of nephron formation or nephron segment delineation.

Figure 4.

Hematoxylin and eosin-stained kidneys from 12-wk-old male wild-type and Ksp-cadherin-deficient mice. Arrows in the bottom images indicate transitions from the cortex to OSOM and from the ISOM to IM, respectively. Bar = 200 μm. IM, inner medulla; ISOM, inner stripe of the outer medulla; OSOM, outer stripe of the outer medulla.

To address the possibility that Ksp-cadherin may play a role in the maintenance of the terminally differentiated polarized epithelial phenotype in the mature kidney, we used antibodies directed against E-cadherin and the α-subunit of Na⁺-K⁺-ATPase to assess epithelial cell polarity in kidney sections from adult Ksp-null and WT animals. In the adult WT mouse kidney, both E-cadherin and Na⁺-K⁺-ATPase have distinct highly polarized basolateral membrane localizations and a very dynamic range of expression along the entire length of the nephron and the collecting system (15).

Figure 5 shows a representative image depicting the typical staining patterns seen for E-cadherin and Na⁺-K⁺-ATPase in multiple sections per animal from over 20 adult animals in each group. In all cases, the staining patterns confirmed the polarized cellular distribution of both proteins in the adult kidney and clearly demonstrated that there is not a discernible difference in the immunolocalization patterns or intensity of staining along the length of the nephron and collecting system of either E-cadherin or Na⁺-K⁺-ATPase in kidneys from Ksp-null and WT animals. The magnification and image exposure used for E-cadherin in Fig. 5 were specifically chosen to illustrate that the null mutation of Ksp-cadherin did not induce a compensatory change in the level of expression of E-cadherin at any point along the length of the nephron or collecting system. Supplementary high-magnification images of anti-E-cadherin antibody staining of proximal tubules of WT and Ksp-cadherin-null animals (Supplemental Fig. S4) further supported the conclusions that Ksp-cadherin expression is not essential for the normal polarized plasma membrane expression of E-cadherin and that E-cadherin expression is not grossly altered to compensate for Ksp-cadherin loss.

Figure 5.

E-cadherin and Na⁺-K⁺-ATPase immunolocalization in kidneys from 12-wk-old male wild-type and Ksp-cadherin-deficient mice. Antibodies 20874-1-AP and A5 were used for E-cadherin and Na⁺-K⁺-ATPase, respectively. Bar = 80 μm. D, distal tubule; G, glomerulus; P, proximal tubule.

In agreement with the immunolocalization study, Western blot analysis of E-cadherin and Na⁺-K⁺-ATPase expression levels in whole kidney homogenates from similar cohorts of adult animals (Fig. 6) showed no significant effect of ablation of the Ksp-cadherin allele. The Western blot analysis was expanded to include villin, Na⁺/H⁺ exchanger isoform 3 (NHE3), and α-epithelial Na⁺ channel (ENaC), representative proteins known to have segment-specific expression patterns, to further assess a potential role for Ksp-cadherin in nephron segment delineation. Villin is expressed in the brush border of the proximal tubule (16), NHE3 is expressed in both the proximal tubule and the loop of Henle (17), and α-ENaC is expressed in the late distal tubule and CD (18). Again, the null mutation of Ksp-cadherin had no significant effect on the expression of these proteins, further strengthening the conclusion that under baseline conditions, Ksp-cadherin is not essential for either the development or maintenance of the structure of the mammalian kidney. Furthermore, both immunolocalization and Western blot analysis clearly showed that the expression of E-cadherin is not modified in any way to compensate for Ksp-cadherin loss.

Figure 6.

Western blot analysis of pan-tubular and segment-specific markers in whole kidney homogenates from 12-wk-old male wild-type (WT or W) and Ksp-cadherin-deficient [knockout (KO) or N] mice. A: composite Western blot. The following antibodies were used: Ksp-cadherin, anti-Ksp; E-cadherin (E-cad), 20874-1-AP; Na⁺-K⁺-ATPase, 55187-1-AP; villin, sc-7672; Na⁺/H⁺ exchanger isoform 3 (NHE3), 3H3; and α-epithelial Na⁺ channel (ENaC), SPC-403. B: total protein stains of gels probed for A used for protein loading normalization calculations. C: normalized densitometry analysis of the results shown in A. n = 6 animals for each group. MW, molecular weight.

Ksp-Null Mutants Have a Urine Concentrating Defect

Although we did not observe a significant difference in any of the plasma parameters that were monitored, we did observe a modest but significant difference in urine osmolality in group-housed age-matched adult WT and Ksp-null mutant animals (Supplemental Fig. S5), suggesting the possibility that Ksp-null mutants may have an altered urine concentrating capability.

To rule out the possibility that the urine osmolality difference observed between the two groups was the result of differential water consumption, urine osmolality, blood electrolyte, BUN, and gross body weights of age-matched individually housed adult WT and Ksp-null animals were measured before and after imposition of 24 h of water restriction. Measurements were made at the same time of day (12:00 PM) 1 wk apart to preclude effects of circadian rhythm and blood sampling volume loss.

As previously observed, there were modest but significant differences in urine osmolality between WT and Ksp-null animals with free access to water (Fig. 7A). After 24 h of water restriction, animals from both groups were clearly capable of producing highly concentrated urine, but Ksp-null animals were not capable of concentrating their urine to the same extent as WT animals. We observed a similar difference in maximum urine concentrating capability between WT and null mice in both male and female animals (Supplemental Fig. S6). Concomitant with the predicted higher diuretic water loss, Ksp-null animals had significantly greater weight loss (Fig. 7B) and higher BUN levels (Fig. 7D) after water restriction than their WT cohorts. Although both groups showed a modest increase in blood Na⁺ levels after water restriction (Fig. 7C), the values were not significantly different between WT and Ksp-null mutant animals under either condition.

Figure 7.

Urine concentrating capability of 12-wk-old male wild-type (WT) and Ksp-cadherin-deficient [knockout (KO)] mice with either free access to water or after 24 h of water restriction. n = 6 animals for each group. A: urine osmolality. B: percent weight loss. C: blood Na⁺ concentration. D: blood urea nitrogen (BUN).

Molecular Basis of the Ksp-Null Urine Concentrating Defect

The ability of the mammalian kidney to produce a concentrated urine in response to water restriction is critically dependent on the generation of an axial corticopapillary osmotic gradient (countercurrent multiplication), an ability to maintain that osmotic gradient while providing adequate blood flow to both the outer and inner medulla (countercurrent exchange), and an ability to independently regulate the water permeability of the CD in both the cortex and medulla (for a review, see Ref. 19). The observation that Ksp-null animals were able to generate appreciably concentrated urine in response to water restriction (Fig. 7) implies that the basic elements of the urine concentrating mechanism and their responsiveness to vasopressin are largely intact. We considered that the inability to maximally concentrate the urine is likely due to an inability to either generate a sufficient osmotic gradient (countercurrent multiplication) in the medulla or modulate the water permeability sufficiently in the CD to facilitate maximum water reabsorption.

As seen in Figure 4, we did not observe significant differences in the relative proportions of the cortex, medulla, or papilla between Ksp-null animals and their WT counterparts or an overt difference in the nephron and CD structure. It is unlikely, therefore, that the concentrating defect is directly due to gross malformations of the nephron or collecting system in Ksp-null animals. The renal epithelium-specific expression of Ksp-cadherin also makes it unlikely that its null mutation would have an impact on vasopressin levels or on the vascular structures of the kidney and hence countercurrent exchange. However, given the extreme importance of the vasa recta to the maintenance of the corticopapillary osmotic gradient, we performed a rudimentary evaluation of the peritubular capillary mass and the general architecture of the renal vasculature by labeling paraffin-embedded kidney sections with an antibody directed against endomucin (Supplemental Fig. S7). Endomucin is an endothelium-specific type 1 membrane glycoprotein (20). In the mouse kidney, endomucin labels the ascending vasa recta (AVR), peritubular and glomerular capillaries, and all large venous vessels (21). We did not observe a significant difference in endomucin staining in either the cortex or medulla of the two groups, consistent with a lack of effect on the peritubular capillaries and AVR. As part of a qPCR survey of osmotically pertinent transcripts in kidneys from Ksp-null and WT animals (Fig. 8), we also failed to detect a significant difference in transcript levels of urea transporter (UT)-B between the two groups. UT-B expression is limited to the endothelial cells of the descending vasa recta (22). Similar UT-B transcript levels in the two groups is consistent with a lack of effect of the Ksp-null mutation on the formation of the descending vasa recta and, taken with the endomucin labeling study, strongly suggested that Ksp-cadherin is not involved with either the formation or maintenance of the renal vasculature.

Figure 8.

Quantitative transcript analysis of whole kidney total RNA samples isolated from 12-wk-old male wild-type (WT) and Ksp-cadherin-deficient [knockout (KO)] mice. n = 6 animals for each group. Aqp, aquaporin; Avpr2, arginine vasopressin receptor 2; ClC-K1, Cl⁻ channel; dCT, transcript to hypoxanthine guanine phosphoribosyltransferase expression ratio expressed as 2^−dCt (where Ct is threshold cycle); NKCC2, Na⁺-K⁺-2Cl⁻ cotransporter; ROMK, renal outer medullary K⁺ channel; UT, urea transporter.

The molecular mechanisms driving the formation of the axial medullary osmotic gradient and the resultant movement of water have been studied extensively (for a review, see Ref. 19). Briefly, the development of the axial osmotic gradient in the outer medulla is largely attributed to countercurrent multiplication in the loop of Henle. Isotonic tubule fluid entering the outer medulla is gradually concentrated by passive water loss as it descends through the medulla and is then systematically diluted in the thick ascending limb by the active reabsorption of NaCl into the interstitium. The axial osmotic gradient is further enhanced in the inner medulla by the interstitial accumulation of urea.

The relative difference in urine osmolality between Ksp-null and WT animals was proportionally similar in animals with free access to water and those on the 24-h water-restricted regimen (see Fig. 7). Therefore, we performed our investigation into the source of the urine concentrating defect on 12-wk-old animals that were maintained under basal conditions with free access to water. Transcript analysis of the principal proteins involved in the movement of salt [Na⁺-K⁺-2Cl⁻ cotransporter (NKCC2)a, NKCC2b, NKCC2f, uromodulin, renal outer medullary K⁺ channel (ROMK), and Cl⁻ channel CLCK1], urea (UT-A1, UT-A2, UT-A3, and UTB), and water [AQP1 + 2 and vasopressin 2 receptor (V2R)] (Fig. 8) and Western blot analysis of AQP2 (Fig. 9A), NKCC2 (Fig. 9B), uromodulin (Fig. 9C), and UT-A1 (Fig. 9D) all failed to detect differences in relative expression between Ksp-null and WT animals, indicating that the concentrating defect was likely not due to altered expression of the proteins involved in the generation and maintenance of the corticomedullary osmotic gradient.

Figure 9.

Western analysis of aquaporin 2 (AQP2), Na⁺-K⁺-2Cl⁻ cotransporter (NKCC2), uromodulin, and urea transporter (UT)A1 in whole kidney homogenates from 12-wk-old male wild-type (WT or W) and Ksp-cadherin-deficient [knockout (KO) or N] mice. A: Western blot of AQP2 (antibody: AB3274). B: Western blot of NKCC2 (antibody: SPC-401D). C: Western blot of uromodulin (antibody: K90071C). D: Western blot of UTA1 (antibody: SPC-406). Top: raw Western blot images of respective proteins. Middle: total protein Ponceau stain of the blots probed in the respective top blots used for protein loading normalization. Bottom: normalized densitometric analysis of the respective top Western blots. n = 6 animals for each group. MW, molecular weight.

Likewise, we did not observe differences in the immunolocalization of NKCC2 and AQP2 in either the cortex or outer medulla of Ksp-null and WT animals (Figs. 10 and 11, respectively). We did, however, observe a striking difference in AQP2 localization in the inner medulla (Fig. 12). In the WT IMCD, moderate levels of diffuse AQP2 labeling were observed throughout the cytoplasm, but distinct apical staining of AQP2 was prominent in over 90% of the IMCDs surveyed. In contrast, in Ksp-null animals, AQP2 labeling was largely diffuse with clear evidence of distinct apical localization present in <20% of total IMCDs surveyed. This suggested that the urine concentrating defect observed in Ksp-null animals may be partially due to impaired AQP2 delivery to the apical membrane and a consequent inability to reabsorb water in response to the osmotic gradient generated in the inner medulla.

Figure 10.

Immunolocalization of total (tot) and phosphorylated (P_i) Na⁺-K⁺-2Cl⁻ cotransporter (NKCC2) in the cortex and outer medulla (OM) of kidneys from 12-wk-old male wild-type and Ksp-cadherin-deficient mice. Immunofluorescent staining superimposed on phase-contrast images showed exclusive thick ascending limb staining in the OM and cortex (B) and the thin portion of the distal straight tubule in proximity to the glomerulus (A, cortex). Total NKCC2 labeled with antibody T9 and phosphorylated NKCC2 labeled with antibody R5. Bars = 30 μm. C, collecting duct; G, glomerulus; P, proximal tubule.

Figure 11.

Immunolocalization of aquaporin 2 (AQP2) in the cortex and outer medulla of kidneys from 12-wk-old male wild-type and Ksp-cadherin-deficient mice. Left: immunofluorescent images. Right: immunofluorescent image superimposed on phase-contrast images. Bars = 30 μm. The antibody used was sc-9882. C, collecting duct; D, distal tubule; G, glomerulus; P, proximal tubule; T, thick ascending limb.

Figure 12.

Immunolocalization of aquaporin-2 (AQP2) in the inner medulla of kidneys from 12-wk-old male wild-type (WT) and Ksp-cadherin-deficient [knockout (KO)] mice. A: immunofluorescent image of inner medullary collecting ducts (IMCDs) stained with anti-AQP2 antibody sc-9882. Bars = 20 μm. B: fraction of total IMCD cross-section profiles with distinct apical AQP2 localization (n = 4 animals per group; 55-88 IMCD profiles from multiple random tissue sections per animal).

The Ability to Maximally Concentrate Urine Is Developmentally Delayed in Ksp-Null Animals

The urine concentrating defect observed in 12-wk-old Ksp-null animals is strikingly reminiscent of the limited urine concentrating capability observed in neonates (for a review, see Ref. 23). In animal studies, the limited urine concentrating capacity of neonatal kidneys appears to be largely due to an attenuated response to antidiuretic hormone in the medullary CD [for a review, see Gattineni and Baum (24)], an effect that would manifest itself by reduced levels of AQP2 on the apical membrane. This and the role played by other members of the cadherin superfamily in tissue development in general prompted us to question if the urine concentrating defect that we observed in 12-wk-old Ksp-null animals may be due to an extremely delayed maturation of renal function. We addressed this possibility by repeating the water restriction study with 10-mo-old Ksp-null and WT animals. Under conditions of either free access to water or 24 h of water deprivation, there were no significant differences in urine osmolality between these older WT and Ksp-null animals (Fig. 13). The lack of difference did not appear to be due to the attenuation of the urine concentrating capability of WT animals, as might be predicted for older animals, but rather an enhanced urine concentrating capability in 10-mo-old Ksp-null mutants.

Figure 13.

Urine concentrating capability of 10-mo-old male wild-type (WT) and Ksp-cadherin-deficient [knockout (KO)] mice with either free access to water or after 24 h of water restriction. n = 6 animals for each group.

As already observed in younger adults, whole kidney AQP2 protein expression levels in 10-mo-old animals maintained under basal conditions with free access to water were not significantly different between WT and Ksp-cadherin-null mice (Supplemental Fig. S8). However, in contrast to the diffuse cytoplasmic AQP2 immunolocalization observed in 12-wk-old Ksp-cadherin-null animals, AQP2 localization in IMCD cells of 10-mo-old animals now had, in accordance with that observed in WT animals, a prominent apical component in over 90% of the IMCDs surveyed (Fig. 14). The age-related enhancement of urine concentrating capability along with normalization of IMCD AQP2 immunolocalization supported the hypothesis that the null mutation of Ksp-cadherin induced a significant developmental delay in concentrating ability that improved with age of the animals.

Figure 14.

Immunolocalization of aquaporin-2 (AQP2) in the inner medulla of kidneys from 10-mo-old male wild-type (WT) and Ksp-cadherin-deficient [knockout (KO)] mice. A: immunofluorescent image of inner medullary collecting ducts (IMCDs) stained with anti-AQP2 antibody sc-9882. Bars = 20 μm. B: fraction of total IMCD cross-section profiles with distinct apical AQP2 localization (n = 4 animals per group; 55–88 IMCD profiles from multiple random tissue sections per animal).

DISCUSSION

The principal aim of the present study was to determine if Ksp-cadherin is critically essential for the development and function of the mammalian kidney by creating a mouse line deficient in Ksp-cadherin. We successfully generated a Ksp-cadherin-deficient mouse line as confirmed by restriction enzyme and Southern blot analysis of genomic DNA and Western blot analysis of whole kidney homogenates. Ksp-null mutant mice on two different genetic backgrounds were morphologically indistinguishable from matched WT cohorts at all stages of postnatal development and displayed no evidence of impaired fertility or life expectancy. Kidneys from both neonate and adult Ksp-cadherin-deficient mice were indistinguishable from WT cohorts and showed no evidence of altered expression levels of a broad panel of tubule segment-specific markers by both Western blot and quantitative transcript analysis.

Although we did not observe structural alterations in either the developing or adult kidney in Ksp-cadherin-deficient animals, we did observe a subtle but significant developmental delay in the ability to fully concentrate urine. Twelve-week-old Ksp-cadherin-deficient animals concentrated their urine to only 67%− 85% of that achieved by similarly aged WT animals. We did not conduct a detailed time-course evaluation, but by 10 mo of age, the urine concentrating capacity of Ksp-cadherin-deficient mice had increased to levels that were equivalent to those observed with adult WT animals.

It is well known that neonates have a diminished urine concentrating capability relative to adults. Human full-term infants do not achieve adult levels of urine concentrating capacity until they are at least 6–12 mo of age (23), and rodents are not able to fully concentrate their urine until they are 4–5 wk of age (25). The diminished urine concentrating capacity of neonates has been linked to reduced medullary tonicity, lower levels of AQP2 expression, and a blunted response to vasopressin (24). We did not measure medullary tonicity in Ksp-cadherin-deficient animals and cannot rule out the possibility of a reduced osmotic gradient in the medulla, but by 12 wk of age, Ksp-cadherin-deficient animals did have typical adult levels of expression of the transport proteins that contribute to formation of the osmotic gradient, and they were consuming a standard adult chow that would make available adequate urea to contribute to inner medullary tonicity. We also did not observe reduced levels of total kidney AQP2 expression and saw no evidence of mislocalization of AQP2 in either the cortex or the outer medulla. The apparent mislocalization of AQP2 in the inner medulla of Ksp-cadherin-deficient animals that we did observe could account for the reduced urinary concentrating capability that the null mutants exhibited at 12 wk of age.

At this point, it is unclear how lack of Ksp-cadherin could specifically affect the cellular localization of AQP2 in the inner medulla, especially given the apparently normal localization of AQP2 in the cortex and outer medulla. It is unlikely to be a simple matter of aberrant cell polarity in the inner medulla. In our detailed survey of Na⁺-K⁺-ATPase and E-cadherin expression in 12-wk-old mice, we did not detect a difference in staining patterns between Ksp-cadherin-null and WT animals at any point along the length of the nephron or the collecting system and found no evidence for a lack of cell polarity in the inner medulla of the null mutants. Similarly, given the simultaneous apical expression of AQP2 in the cortex and outer medulla of the 12-wk-old Ksp-null mutants, and the observation that apical expression in the inner medulla is ultimately only delayed, it is extremely unlikely that apical localization of AQP2 could be critically dependent on a direct interaction with Ksp-cadherin. It is conceivable that null expression of Ksp-cadherin could mimic some subtle aspect of the early stages of renal development before Ksp-cadherin is normally expressed in the kidneys of WT animals, leading to the delay in maturation of urine concentrating capability that we observed in Ksp-cadherin-null mutant animals.

Aboudehen et al. (26) reported that mice with a CD-specific deficiency in hepatocyte nuclear factor-1β (HNF-1β) had a similar urine concentrating defect. HNF-1β is a transcription factor expressed in the kidney, genitourinary tract, pancreas, and liver. In the kidney, it is expressed in the epithelial cells lining the tubular nephron and collecting system, and it has been shown to significantly modulate the expression of Ksp-cadherin (27). Unlike Ksp-cadherin-deficient animals, kidneys from CD HNF-1β-deficient mice had reduced levels of UT-A1/3 and elevated levels of AQP2. However, as observed in Ksp-cadherin-null mutants, AQP2 had a largely cytoplasmic localization in medullary CD cells of CD HNF-1β-deficient mice rather than the predominantly apical localization observed in matched WT animals. Because HNF-1β drives the expression of multiple transcripts in the collecting system, it is difficult to ascribe the urine concentrating defect to altered expression of any one protein, and it likely reflects the summation of the interactions among multiple pathways. However, the phenotype of the urine concentrating defect driven by Ksp-cadherin deficiency is entirely contained within that observed with CD HNF-1β deficiency. Knowing that HNF-1β modulates the expression of Ksp-cadherin raises the possibility that deficient expression of Ksp-cadherin may be linked to mistargeting of AQP2 and delayed maturation of urine concentrating capability in both mutant models.

Ksp-cadherin is expressed at relatively high levels in every segment of the mouse nephron and collecting system, and given its ability to facilitate cell-cell adhesion in in vitro cell culture assays (6), it is surprising that Ksp-cadherin deficiency had no observable effect on the architecture of the mouse kidney. It is tempting to speculate that there is biological redundancy and that another cadherin or closely related protein may be upregulated to compensate for the loss of Ksp-cadherin. The logical candidate to compensate for Ksp-cadherin loss is E-cadherin. Multiple cadherins have been identified in both the developing and the mature kidney (14, 28–32), but only E-cadherin is coexpressed with Ksp-cadherin at every stage of development in which Ksp-cadherin has been detected and only E-cadherin is coexpressed with Ksp-cadherin in every segment of the nephron and collecting system in the adult mouse kidney. We specifically looked for compensation by E-cadherin in Ksp-cadherin-deficient animals but could find no evidence for upregulation of its expression either by immunolocalization studies in developing and adult kidneys or by Western blot analysis in adult kidneys. It is possible that compensation by E-cadherin is not accompanied by changes in levels of expression or that other cadherins or proteins might be involved. Alternatively, it is possible that Ksp-cadherin does not participate directly in either the development or maintenance of the structural integrity of the tubular epithelium by mediating cell adhesion per se.

The similar lack of a structural phenotype in the kidneys of a proximal tubule-specific E-cadherin-deficient mouse model led Zheng et al. (33) to propose that in the kidney, the role of E-cadherin as a signal transduction molecule may be of more consequence than its purported role as a structural cell-cell adhesion molecule. Likewise, null mutations of P-cadherin (34), R-cadherin (29), cadherin-11 (35), and cadherin-6 (31) all failed to have significant structural consequences for the adult mouse kidney. It is, therefore, important to consider the potential role(s) played by cadherins in the kidney beyond mediating structural cell-cell adhesion.

Despite this detailed survey of renal structure and function in Ksp-cadherin-deficient mice, we are still left pondering its function. The extremely late onset of expression of Ksp-cadherin in nephrogenesis and the developmental delay in urine concentrating capability in Ksp-cadherin-deficient animals hint at involvement in establishing, maintaining, or signaling terminal differentiation, but at this point, such a role is purely speculative. Proximal tubule-specific E-cadherin-deficient animals also exhibited no clear structural alterations or functional deficiencies under basal conditions (33). It was only after imposition of renal injury via unilateral ureteral obstruction that a potential antifibrotic role for E-cadherin became apparent. It is likely that exposure to specific physiological or pathological challenges will be similarly required to uncover the role(s) played by Ksp-cadherin in the kidney. Indeed, an important limitation of our study is that additional defects in kidney structure or function may not have been detected under the conditions examined, but such defects might become magnified in response to physiological or pathological stresses.

In conclusion, we have successfully generated a Ksp-cadherin-deficient mouse model. The mutant animals are healthy and fertile and are grossly phenotypically indistinguishable from a matched WT cohort. Kidneys from neonatal Ksp-cadherin-deficient animals show no evidence of structural abnormalities, disorganization of nephrogenic progenitors, or compromised metanephrogenesis, and the structure of the kidneys of adult mutant animals is similarly unaffected. Plasma electrolytes, total CO₂, BUN, and creatinine levels were comparable in Ksp-null mutant and WT animals, although mutants displayed a subtle but significant developmental delay in their ability to maximally concentrate urine. Ksp-cadherin is clearly not required for nephron formation, nephron segment delineation, or baseline renal function. Perhaps as the developmental delay in urine concentrating ability suggests, it may play a role in establishing or signaling the terminal differentiation of renal tubular epithelial cells.

SUPPLEMENTAL DATA

All Supplemental Material: https://doi.org/10.6084/m9.figshare.13119911.

GRANTS

This work was supported by the National Institute of Diabetes and Digestive and Kidney Diseases Grants P01DK017433 and P30DK079310 (to P.S.A. and T.W., George M. O’Brien Kidney Center at Yale) and P30DK114857 (to B.R.T., Northwestern University George M. O’Brien Kidney Research Core Center). S.B. received funding from the thematic network grant TRENAL of the Deutscher Akademischer Austauschdienst.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

R.B.T., D.W.D., and P.S.A. conceived and designed research; R.B.T., D.W.D., S.B., B.T. and T.W. performed experiments; R.B.T., D.W.D., S.B., B.T., and C.J.B. analyzed data; R.B.T., D.W.D., S.B., B.T., C.J.B., and P.S.A. interpreted results of experiments; R.B.T., B.T., and C.J.B. prepared figures; R.B.T. and P.S.A. drafted manuscript; R.B.T., D.W.D., S.B., B.T., C.J.B., F.K., T.W., and P.S.A. edited and revised manuscript; R.B.T., D.W.D., S.B., B.T., C.J.B., F.K., T.W., and P.S.A. approved final version of manuscript.