Deletion of the Prorenin Receptor from the Ureteric Bud Causes Renal Hypodysplasia

The role of the prorenin receptor (PRR) in the regulation of ureteric bud (UB) branching morphogenesis is unknown. Here, we investigated whether PRR acts specifically in the UB to regulate UB branching, kidney development and function. We demonstrate that embryonic (E) day E13.5 mouse metanephroi, isolated intact E11.5 UBs and cultured UB cells express PRR mRNA. To study its role in UB development, we conditionally ablated PRR in the developing UB (PRR UB−/−) using Hoxb7 Cre mice. On E12.5, PRR UB−/− mice had decreased UB branching and increased UB cell apoptosis. These defects were associated with decreased expression of Ret, Wnt11, Etv4/Etv5, and reduced phosphorylation of Erk1/2 in the UB. On E18.5, mutants had marked kidney hypoplasia, widespread apoptosis of medullary collecting duct cells and decreased expression of Foxi1, AE1 and H+-ATPase α4 mRNA. Ultimately, they developed occasional small cysts in medullary collecting ducts and had decreased nephron number. To test the functional consequences of these alterations, we determined the ability of PRR UB−/− mice to acidify and concentrate the urine on postnatal (P) day P30. PRR UB−/− mice were polyuric, had lower urine osmolality and a higher urine pH following 48 hours of acidic loading with NH4Cl. Taken together, these data show that PRR present in the UB epithelia performs essential functions during UB branching morphogenesis and collecting duct development via control of Ret/Wnt11 pathway gene expression, UB cell survival, activation of Erk1/2, terminal differentiation and function of collecting duct cells needed for maintaining adequate water and acid-base homeostasis. We propose that mutations in PRR could possibly cause renal hypodysplasia and renal tubular acidosis in humans.


Introduction
Congenital anomalies of the kidney and urinary tract (CAKUT) occur in 3-6 per 1000 live births and account for 31% of all cases of end-stage kidney disease (ESKD) in children in the United States [1]. All forms of CAKUT stem from abnormal kidney development [1,2]. Branching morphogenesis of the ureteric bud (UB) is a key developmental process that directs organogenesis of the metanephric kidney [3,4]. Terminal tips of branching UBs induce surrounding mesenchyme-derived nephron progenitors to differentiate into nephrons, thus forming the metanephric kidney [3,4]. Following completion of UB branching, UB-derived collecting ducts undergo terminal differentiation-acquisition of distinct epithelial cell types that perform specialized functions [4,5]. Notably, derangements in UB morphogenesis or UB cell differentiation result in CAKUT and distal renal tubular disorders, respectively [3][4][5][6][7][8].
The PRR is the cell-surface receptor for renin and prorenin, and an accessory subunit of the vacuolar proton pump H + -ATPase [9][10][11]. In the adult rat collecting duct, PRR is most abundant at the apical surface of type a intercalated cells (a-ICs) where it colocalizes with the H + -ATPase and may be activated in a paracrine fashion by prorenin or renin released by adjacent principal cells [12,13]. Moreover, H + -ATPase is required for the activation the extracellular signal-regulated kinase 1/2 (Erk1/2) induced by prorenin or renin in the collecting duct cells [12]. Critical role for H + -ATPase in development is evident from the observation that mutations in the genes encoding specific subunits of H + -ATPase in mice result in embryonic lethality or metabolic acidosis [14,16]. Given that pharmacologic inhibition of Erk1/2 decreases UB branching [17], disruption of PRR signaling in the UB may lead to aberrant UB morphogenesis and renal collecting system development. In addition, PRR may promote differentiation of H + -secreting intercalated cells in the developing collecting duct.
Here, we tested the hypothesis that targeted inactivation of the PRR in the UB epithelia in mice is essential for UB branching morphogenesis and collecting duct development. We demonstrate that Cre-mediated inactivation of the PRR targeted to the UB disrupts UB branching, reduces the number of nephrons and causes renal hypodysplasia. Reduced phosphorylation of Erk1/2 in the UB, widespread apoptosis of UB and medullary collecting duct cells, aberrant expression of Ret/Wnt11 UB morphogenetic program genes and collecting duct cell differentiation markers such as Foxi1, AE1, H + -ATPase and Aqp2 is observed in mutant kidneys. These findings demonstrate that PRR present in the UB epithelia performs essential functions during UB branching morphogenesis and collecting duct development via control of Ret/Wnt11 pathway gene expression, UB cell survival, activation of Erk1/2 signaling and differentiation of collecting duct cells involved in acid-base homeostasis and concentration of the urine.

Materials and Methods
Generation of UB-specific PRR-knockout Mice PRR-floxed mice were provided by Dr. Atsuhiro Ichihara (Keio University, Tokyo, Japan) [10]. To delete PRR conditionally in the ureteric bud (UB), we used the Hoxb7 Cre transgene, which drives Cre expression in the Wolffian duct and UB epithelium from E9.5 onwards [18]. The resulting Hoxb7 Cre+ /PRR flox/flox mice represent UB-specific PRR-knockout mice (PRR UB2/2 ). Control mice consisted of Hoxb7 Cre2/ PRR flox/flox and Hoxb7 Cre2/ PRR flox/+ (PRR UB+/+ ) littermates. UB-specific knockout of PRR was confirmed by qRT-PCR analysis which revealed an 80% decrease in PRR mRNA levels in E11.5 intact isolated UBs (iUBs) from PRR UB2/2 compared with PRR UB+/+ mice (0.1960.03 vs. 1.060, p,0.001). On P1, UB-specific knockout of PRR was confirmed by double immunostaining of PRR and Aqp2, which revealed no PRR expression in the collecting duct in PRR UB2/2 mice. All experiments involving mice were approved by Tulane Institutional Animal Care and Use Committee. All animal work involved in the generation of in situ hybridization probes by Dr. Jing Yu was approved by the University of Virginia Animal Care and Use Committee.

Immunohistochemistry and Histopathology
Kidneys were fixed in 4% PFA at 4uC and paraffin embedded. Immunostaining was performed by the immunoperoxidase technique using 4-mm sections with Vectastain Elite kit (Vector Laboratories, Burlingame, CA). Primary antibodies included anti-PRR (1:200, Santa Cruz), anti-Aqp2 (1:200, Santa Cruz), anti-H + -ATPase a4 (1:1000) [22], rabbit polyclonal anti-AE1 (1:200), antiphospho-Erk1/2 (1:200, Cell Signaling) and anti-Lotus Tetragonolobus Lectin (LTL) (1:400, Vector Laboratories). Whole intact E12.5 metanephroi from PRR UB2/2 and control mice (n = 6 mice and n = 12 kidneys per genotype) were processed for the whole mount immunofluorescence using anti-cytokeratin (1:200, Sigma) and anti-WT1 (1:100, Abcam) antibodies and the number of UB tips was counted. For immunofluorescence studies, secondary antibodies were detected with Alexa Fluor dyes (Invitrogen). Specificity of immunostaining was documented by the omission of the primary antibody. Left kidneys from P1 PRR UB2/2 and PRR UB+/+ mice (n = 3 mice per group) were cut in the longitudinal midplane, processed through the paraffin, and embedded on the cut surface. Kidneys were sectioned at 4-mm and stained with hematoxylin and eosin. The number of nephrons in each of 3 consecutive sections adjacent to the longitudinal midplane was counted and the mean number of nephrons per section per kidney was calculated. To determine the number of WT1-positive structures, we examined the intensity of WT1 immunostaining (1:100, Abcam) in P1 kidney sections (n = 3 mice per group) using Slide book 4.0 software (Intelligent Imaging Innovations, Denver, CO). Total number of H + -ATPase a4expressing cells in E18.5 collecting ducts was counted (n = 3 mice per genotype, 3 sections per kidney, 10 collecting ducts/section). More medullary domains of the kidney were chosen to count the number of H + -ATPase-expressing cells. All counts were performed in a blinded fashion.

In situ Hybridization (ISH)
Section ISH was performed on E14.5 and E18.5 PRR UB2/2 and control kidneys as previously described [21]. Mouse full length probes for Foxi1 and Aqp2 were a kind gift from Dr. Jing Yu (University of Virginia) [23]. 4 embryonic kidneys per group per probe were examined.

Cell Proliferation and Apoptosis Assays
Cell proliferation and apoptosis was examined in E13.5 and E18.5 kidney sections from PRR UB2/2 and control mice (n = 3 mice per genotype, 3 sections per kidney) as previously described Figure 2. Effect of targeted genetic inactivation of the PRR in the ureteric bud (UB) on UB branching and Erk1/2 phosphorylation in E12.5 metanephroi. A, B: Metanephroi were co-stained with anti-pancytokeratin antibody to visualize the UB (green) and anti-WT1 antibody to visualize metanephric mesenchyme (red). The number of UB tips is reduced in mutant PRR UB2/2 compared with PRR UB+/+ kidneys. C: Bar graph showing the effect of PRR deletion in the UB on the number of UB tips. D-I: Sections of E12.5 kidneys show that phospho Erk1/2 (pErk1/2) immunostaining (red) is reduced in the UB of mutant (G, I) compared with control (D, F) mice. J. Whole E12.5 kidney lysates were subjected to Western blotting with anti-pErk1/2 antibody. After stripping, the membrane was reprobed with anti-total Erk1/2 antibody (Erk1/2). Erk1/2 phosphorylation appears to be reduced in PRR UB2/2 compared with PRR UB+/+ kidneys. doi:10.1371/journal.pone.0063835.g002 [21]. Cell proliferation and apoptosis was assessed throughout the entire UB epithelium on E13.5. Anti-phospho-histone H3 (pH3) and anti-cleaved caspase-3 antibodies were used (Cell Signaling, Danvers, MA; 1:50). UBs and collecting ducts were visualized with anti-cytokeratin antibody (1:200, Sigma). The number of proliferating and apoptotic cells in the UB/collecting duct epithelia was normalized to the total number of DAPI-positive (Invitrogen) cells in each kidney section. The number of DAPI-positive cells was determined by Image J software (NIH).
Measurement of Serum Creatinine, Urine Volume, Osmolality and pH PRR UB2/2 and control mice (n = 4 mice per genotype) were housed in metabolic cages (Hatteras Instruments, Cary, NC), fed a standard chow and allowed free access to tap water. 24-hour urine was collected at baseline and after administration of NH 4 Cl (0.8 g/kg body weight in drinking water) for 48 hours and was processed immediately for pH analysis by the pH meter (Fisher Scientific) [8]. Plasma creatinine was measured by HPLC with picric acid (Jaffe method).

Western Blot Analysis
E12.5 kidneys from PRR UB2/2 and control mice (n = 4 kidneys per genotype per one pooled sample, n = 3 pooled samples per genotype) were pooled and homogenized in cold lysis buffer containing a cocktail of enzyme inhibitors. Proteins (60 mg/lane) were processed for Western blot analysis as previously described [24]. After blocking nonspecific binding, the membranes were incubated with the phosphospecific anti-phospho-Erk1/2 antibody (1:200, Cell Signaling). After stripping, membranes were reprobed with anti-total Erk1/2 antibody (Cell Signaling) to document equal protein loading. Immunoreactive bands were visualized using the enhanced chemiluminescence detection system (ECL, Amersham) as previously described [24].

Statistics
Statistical analyses were carried out upon all biologic replicates with Student's t test or a one-way ANOVA, followed by Bonferroni test. Data are presented as Mean6SEM. A p value of ,0.05 was considered statistically significant.

PRR Expression in Control and PRR UB2/2 Developing Kidneys
To determine the PRR expression during kidney development, we performed immunohistochemistry studies on sections of E13.5 and P1 mouse kidneys. Sections of E13.5 wild-type kidney reveal PRR immunostaining in the UB, nephron epithelia and in nascent glomeruli ( Figure 1A-D). While PRR UB+/+ mice expressed PRR in the collecting ducts and the mesenchyme on P1 ( Figure 1F-H), kidneys of PRR UB2/2 mice showed PRR labeling in the mesenchyme only and no expression in the collecting ducts ( Figure 1I-K). RT-PCR demonstrated apparently similar expression levels of PRR mRNA in the whole intact E13.5 wild-type mouse kidney, in intact UBs isolated from E11.5 wild-type mouse kidneys, and in immortalized UB cells grown in vitro ( Figure 1E). Thus, at early and later stages of metanephric development, PRR is expressed in both metanephric and UB lineages.

PRR Deletion Results in an Increased UB and Collecting Duct Cell Apoptosis
To identify the cellular mechanisms by which PRR deficiency in the UB could cause aberrant UB branching morphogenesis, we examined UB cell proliferation and apoptosis in PRR UB2/2 and control mice on E12. 5 Figure 5). These findings demonstrate an important role for the PRR in epithelial cell proliferation and survival during UB branching and collecting duct development, and suggest that enhanced UB cell apoptosis may account for UB branching defects observed in PRR UB2/2 mice.

PRR Deletion Results in Renal Hypodysplasia
Dissection of urogenital tracts of newborn mice showed that mutants had smaller kidney size (kidney length: 800620 vs. 1150632 mm, p,0.001) ( Figure 6). On P1, body weight did not differ in mutant and control mice (4.0860.11 vs. 4.1260.13 g, p = 0.82). In contrast, kidney weight (15.660.45 vs.

PRR UB2/2 Kidneys Display Aberrant Terminal Differentiation of the UB Epithelium and Altered Wnt mRNA Expression
To determine whether PRR regulates terminal differentiation of the collecting duct cells involved in water permeability and acidbase homeostasis, we examined the expression of winged helix transcription factor Foxi1, chloride-bicarbonate exchanger AE1, a-intercalated cell (a-IC)-specific H + -ATPase subunit a4 and water channel Aqp2 mRNA in PRR UB2/2 and control mice on E18.5. qRT-PCR showed decreased expression of Foxi1, AE1, H + -ATPase a4 and Aqp2 mRNA in mutants (Figure 7). These results were validated by in situ hybridization with Aqp2 and Foxi1 probes ( Figure 7) and by immunohistochemistry with anti-H + -ATPase a4 Given that Wnt signaling is prerequisite for proper metanephric development and that PRR is required for both canonical Wnt/bcatenin and frizzled (Fz)/planar cell polarity (PCP) signaling in Drosophila [26][27][28], we examined whether targeted deletion of UB PRR alters Wnt2b, Wnt4, Wnt7b and Wnt9b mRNA expression on E18.5 by whole-kidney qRT-PCR. Wnt2b, Wnt7b and Wnt9b mRNA levels were lower in PRR UB2/2 compared to PRR UB+/+ kidneys, whereas Wnt4 expression did not differ (Figure 7). These findings indicate that UB PRR may control UB/collecting duct development, in part, via the regulation of Wnt pathway.

PRR Deletion Impairs Kidney Function, Urinary Concentration and Acidification Ability
To test the functional consequences of the deficiency of the collecting duct PRR, we examined renal function, urinary volume, osmolality and pH on P30 during ad libitum water intake. Mutant mice had increased levels of serum creatinine on P30 (23760.13 vs. 59614 mmol/l, p,0.001). Despite similar water intake (2.560.37 vs. 2.260.36 ml, p = 0.52), 24-hour urine volumes were higher (3.260.32 vs. 2.260.24 ml, p,0.05) and urine osmolalities were lower (830639 vs. 16456160 mg, p,0.001) in mutants compared with controls ( Figure 9). These results demonstrate that PRR in epithelial cells of the collecting duct contributes to the concentration of the urine. Since H + -ATPase is important for secretion of protons into the tubular lumen, we next examined the ability of PRR-mutant mice to acidify the urine. While baseline urine pH did not differ between PRR UB2/2 and PRR UB+/+ mice (6.4060.05 vs. 6.5260.17, p = 0.51), it was higher following 48 hours of acidic loading with NH 4 Cl in mutants compared with controls (6.5160.13 vs. 6.0260.12, p,0.05) (Figure 9). Acidic load resulted in a decrease in urine pH in control (6.4060.05 vs. 6.0260.12, p,0.05) and no change in urine pH in mutant (6.5260.17 vs. 6.5060.13, p = 0.91) mice. These findings demonstrate a decreased capacity of PRR UB2/2 mice to handle an acidic load.

Discussion
Here, we provide the first in vivo evidence demonstrating that PRR is essential for UB branching and collecting duct development. PRR UB2/2 mice have profound defects in the UB lineage, including reduced UB branching, occasional cysts in the medullary collecting ducts, which ultimately result in a decreased number of nephrons and marked kidney hypoplasia. These abnormalities are secondary, at least in part, to aberrant apoptosis and proliferation of UB and collecting duct cells, reduced phosphorylation of Erk1/ 2, decreased expression of Ret, Wnt11, Etv4 and Etv5 in the UB epithelia. Mutant kidneys show decreased expression of Foxi1, AE1, H + -ATPase a4 and Aqp2, revealing that PRR in the collecting duct is required for the proper expression of genes needed for maintaining adequate water and acid-base homeostasis. Finally, mutant mice have polyuria with defects in urinary concentrating and acidification capacity.
A new role for PRR signaling in the UB lineage revealed in PRR UB2/2 mice is to stimulate UB branching by controlling Ret/ Wnt11 pathway gene expression and signaling via Erk1/2. Ret and Wnt11 function in a positive feedback loop to promote UB branching via induction of proliferation and migration of the UB tip cells by maintaining a balance between appropriate expression of glial-derived neurotrophic factor (GDNF) in the mesenchyme and of Ret/Wnt11 in the UB tips [29,30]. The Pea3 family of Ets transcription factors Etv4/Etv5 and Erk1/2 are indispensable for transmission of Ret signals to regulate UB branching [31,32]. Thus, decreased expression of Ret, its downstream targets, Wnt11, Etv4/Etv5, and reduced Erk1/2 phosphorylation in the UB of PRR UB2/2 mice are probably a major driving force behind UB branching defects. The observed increase in cell apoptosis in UB branches in PRR UB2/2 mice is consistent with the known function of Ret signaling to promote UB cell survival [33]. Since targeted inactivation of the PRR in mouse cardiomyocytes causes cardiomyocyte apoptosis, the effect of PRR depletion on UB-derived collecting duct cell survival observed in the present study may be, in part, Ret-independent [10]. In addition to aberrant cell proliferation and apoptosis, defects in collecting duct planar cell polarity/oriented cell division may explain presence of occasional collecting duct cysts in PRR mutants. This possibility is supported by the observations that treatment of Xenopus embryos with anti-PRR morpholinos causes a short body axis, smaller head size and a broader expression domain of Xnot, a hallmark of impaired convergent extension movements [26]. In addition, Drosophila PRR interacts biochemically with Fz receptor, which is required for PCP signaling, in human embryonic kidney HEK293T cells [26]. Presence of marked renal hypoplasia in PRR UB2/2 mice in the absence of changes in medulla/cortex ratio suggests that the observed phenotype is due predominantly to a proportional decrease in nephron endowment resulting from reduction in UB branching rather than to decreased elongation of medullary collecting ducts. Consistent with this interpretation, reductions in UB branching and nephron number are characteristic of human renal hypodysplasia, a form of CAKUT observed in 1 in 400 births [34].
Regulation of acid-base homeostasis by the kidney is essential for health. In the collecting duct, a-ICs secrete protons into the tubular lumen through apical membrane H + -ATPase functionally coupled to the basolateral membrane chloride-bicarbonate transporter anion exchanger 1 (AE1) [35]. Differentiation of ICs requires the winged helix transcription factor Foxi1, as Foxi1 UB2/ 2 mice develop distal renal tubular acidosis (dRTA) due to absence of ICs [8]. Given that PRR is an accessory subunit of the vacuolar proton pump H + -ATPase, which is expressed at the apical surface of collecting duct a-ICs [9,11,35], we strongly suspected that PRR deletion in the collecting duct would lead to aberrant H + -ATPase expression Indeed, expression of a-ICspecific H + -ATPase subunit a4 was reduced in PRR mutants. Reduction in H + -ATPase subunit a4 levels was accompanied by decreased expression of Foxi1 and AE1. Since Foxi1, reported to mediate differentiation of ICs from epithelial precursor UB cells, is necessary for expression of the H + -ATPase subunit a4 and AE1 in the collecting duct, decreased expression of ATPase subunit a4 and AE1 observed in this study is likely secondary to a reduction in Foxi1 levels [8,36]. Thus, PRR is a direct or indirect regulator of gene expression in a-ICs. The importance of these findings is underscored by the fact that disruption of a-IC function due to mutations in genes encoding H + -ATPase a4 subunit or AE1 results in inheritable forms of dRTA in humans [37,38]. ICs are present in the collecting duct in a random distribution among the majority principal cells (PCs). To determine whether PRR deletion in the collecting duct alters terminal differentiation of PCs, we examined expression of water channel Aqp2. Aqp2 mutations in humans cause nephrogenic diabetes insipidus, a disease where water resorption by the collecing duct is eliminated [39]. Abundance of Aqp2 mRNA and protein levels was diminished in PRR UB2/2 collecting ducts. Hence, PRR signaling may be also important in terminal differentiation and function of PCs. We found apparently normal localization of Aqp2 protein at the apical surface of PCs. Thus, PRR in the collecting duct does not appear to affect trafficking of Aqp2 in vivo.
To examine the impact of targeted PRR deletion on collecting duct function, we analyzed renal concentrating and acidification ability in PRR UB2/2 and PRR UB+/+ mice. Despite the presence of grossly morphologically normal, although relatively smaller compared with PRR UB+/+ mice, medulla and similar water intake, PRR UB2/2 mice were polyuric and had impaired urine-concentrating ability. These findings suggest that PRR regulates the urinary concentrating mechanism through direct affect on collecting duct epithelium via downregulation of Aqp2, a major determinant of urinary concentrating capacity [6]. Given that expression of IC genes important for urinary acidification was decreased in the collecting duct epithelium of PRR UB2/2 mice, we hypothesized that mutants had reduced ability to acidify the urine. Indeed, our findings revealed a decreased capacity to acidify the urine after acidic load in PRR mutants. Thus, collecting duct PRR is important in establishment and function of collecting duct cells involved in acid-base homeostasis. These features of the reduced H + -ATPase expression, coupled with a decreased ability to secrete protons into the urine in response to acidic loading, in PRR mutants are in agreement with the reported role of the PRR in podocytes, where targeted PRR depletion downregulates the expression of the V 0 c subunit of H + -ATPase resulting in an increased vesicular pH [40,41].
In summary, we demonstrate that UB/collecting duct PRR is essential for kidney morphogenesis and acquisition of function by collecting duct cells involved in water and acid-base homeostasis. Thus, PRR is a potential candidate for future genetic screening studies in patients with renal hypodysplasia or disorders of water resorption and dRTA.