Regulation of the Epithelial Sodium Channel by N4WBP5A, a Novel Nedd4/Nedd4-2-interacting Protein*

The amiloride-sensitive epithelial sodium channel (ENaC) plays a critical role in fluid and electrolyte homeostasis and consists of α, β, and γ subunits. The carboxyl terminus of each ENaC subunit contains a PPXY motif that is believed to be important for interaction with the WW domains of the ubiquitin-protein ligases, Nedd4 and Nedd4-2. Disruption of this interaction, as in Liddle's syndrome where mutations delete or alter the PPXY motif of either the β or γ subunits, has been shown to result in increased ENaC activity and arterial hypertension. Here we present evidence that N4WBP5A, a novel Nedd4/Nedd4-2-binding protein, is a potential regulator of ENaC. In Xenopus laevisoocytes N4WBP5A increases surface expression of ENaC by reducing the rate of ENaC retrieval. We further demonstrate that N4WBP5A prevents sodium feedback inhibition of ENaC possibly by interfering with the xNedd4-2-mediated regulation of ENaC. As N4WBP5A binds Nedd4/Nedd4-2 via PPXY motif/WW domain interactions and appears to be associated with specific intracellular vesicles, we propose that N4WBP5A functions by regulating Nedd4/Nedd4-2 availability and trafficking. Because N4WBP5A is highly expressed in native renal collecting duct and other tissues that express ENaC, it is a likely candidate to modulate ENaC function in vivo.

The apically localized amiloride-sensitive epithelial sodium channel (ENaC) 1 plays a critical role in fluid and electrolyte homeostasis and is widely expressed in absorptive epithelia such as the renal collecting duct, the colon, lung, and sweat and salivary ducts (1)(2)(3). ENaC is composed of three homologous subunits termed ␣, ␤, and ␥ that contain two transmembrane domains, a large extracellular loop, and short intracellular amino and carboxyl termini. The carboxyl terminus of each ENaC subunit contains a PPXY sequence (the PY motif), which when mutated or deleted in either the ␤ or ␥ ENaC subunits leads to Liddle's syndrome, an autosomal dominant form of hypertension (4 -6). Therefore, mutating just one PY motif from a single subunit of the multimeric ENaC complex is sufficient to lead to a disease phenotype. Studies in the Xenopus oocyte system show that the mutations in ENaC subunits, similar to those that cause Liddle's syndrome, result in increased amiloride-sensitive Na ϩ current (5)(6)(7)(8)(9)(10). This increase is mostly attributed to the presence of increased numbers of active Na ϩ channels in the cell membrane, although an increase in channel open probability (P o ) may also be a contributory factor (7, 10 -12).
The PY motifs in the carboxyl termini of ENaC subunits have been shown to interact with the WW domains of the two highly related ubiquitin-protein ligases, Nedd4 and Nedd4-2/ KIAA0439, that are known to be expressed in a number of different tissues (13)(14)(15)(16)(17)(18)(19)(20). Nedd4 and Nedd4-2 have been proposed to down-regulate Na ϩ channel activity in response to increases in intracellular Na ϩ by ubiquitinating the channel, leading to its endocytosis and degradation (17,(21)(22)(23). Recent studies (24,25) demonstrate that the serum-and glucocorticoid-regulated kinase regulation of ENaC is also mediated, at least in part, via the Nedd4-2 protein. Therefore, Nedd4 and Nedd4-2 are key negative regulators of ENaC, and a disruption of this regulation is the most likely cause of Liddle's syndrome. Nedd4 consists of a HECT type of ubiquitin-protein ligase domain, 3-4 WW domains, and a Ca 2ϩ and lipid binding domain (13,14,26). Nedd4-2 also contains a ubiquitin-protein ligase domain and 4 WW domains but lacks a Ca 2ϩ and lipid binding domain (19). The WW domains of Nedd4 and Nedd4-2 are required for substrate binding, whereas the ubiquitin-protein ligase domain is required for ubiquitination of substrates such as ENaC (26). Both Nedd4 and Nedd4-2 have been cloned from mammals, but so far only Nedd4-2 has been found to be present in Xenopus (13,14,19,20,27).
With the view of identifying additional substrates for Nedd4, we previously carried out a far Western screen using the WW domains of Nedd4 as a probe (28). By using this approach we identified a number of PY motif-containing proteins that interact with the WW domains of Nedd4. One of these, N4WBP5, is a novel Golgi-associated protein containing two PY motifs in the amino-terminal domain and three putative transmembrane domains in the carboxyl-terminal half of the protein (29). A closely related protein identified in data base searches, which we have named N4WBP5A, also contains two PY motifs and the three transmembrane domains. We show here that, similar to N4WBP5, N4WBP5A binds to Nedd4. Because Nedd4 is implicated in ENaC control, we were interested to test whether N4WBP5A has a role in Nedd4-mediated ENaC regulation. We report that N4WBP5A is highly expressed in collecting duct cells of the kidney and other tissues that express ENaC. We show that N4WBP5A interacts with Nedd4-2, in addition to Nedd4, and this interaction is mediated via PY motif/WW domain interactions. Most importantly, we demonstrate that in Xenopus oocytes, N4WBP5A has the ability to enhance ENaC surface expression by interfering with the Nedd4/Nedd4-2-mediated regulation of ENaC.

EXPERIMENTAL PROCEDURES
Sequence Analysis-Hydropathicity (Kyte and Doolittle) of the N4WBP5A protein was estimated using the Protscale Software at Ex-PASy Molecular Biology Server (SIB), and transmembrane (TM) domains were predicted using the TMHMM version 2.0 software (Center for Biological Sequence Analysis, Denmark).
Plamids and cDNA Constructs-The coding region of N4WBP5A was amplified by reverse transcriptase-PCR, with a carboxyl-terminal FLAG tag, and cloned into the EcoRI and HindIII sites of pcDNA3 (Invitrogen) to generate pcDNA3-N4WBP5A-FLAG. The generation of pCXN2-Nedd4 has been described previously (29). pcDNA3-Nedd4-2-FLAG was generated by PCR amplification of the mouse Nedd4-2 open reading frame with a FLAG tag engineered at the amino terminus and cloning into the EcoRI and XbaI sites of pcDNA3 (Invitrogen). PCR mutagenesis was used to create N4WBP5A-PY by altering tyrosine residues 56 and 82 to alanine.
The N4WBP5A-GST bacterial expression construct was generated by PCR amplification of the N4WBP5A cDNA encoding the amino-terminal 135 residues, lacking the initiation methionine, and cloned into pGEX-2TK (Amersham Biosciences) such that the N4WBP5A sequence was fused to the carboxyl-terminal of GST. The Nedd4-2-GST bacterial expression construct was generated by PCR amplification of a region of Nedd4-2, residues 110 -226, located between WW domains 1 and 2 and cloned into the BamHI/EcoRI sites of pGEX-2TK such that the Nedd4-2 sequence was fused to the carboxyl terminus of GST. Mouse Nedd4-2 single WW domain-GST constructs were generated by PCR amplification of each WW domain followed by cloning into either the EcoRI or BamHI/EcoRI sites of pGEX-2TK (Amersham Biosciences). PCR mutagenesis was used to mutate each WW domain such that the second conserved tryptophan was altered to phenylalanine and the conserved proline altered to alanine. The expression construct containing all four WW domains fused to GST was generated by PCR amplification and cloning into the EcoRI site of pGEX-2TK.
Antibody Production and Affinity Purification-GST fusion proteins were produced and purified as described previously (17). The purified N4WBP5A-GST and Nedd4-2-GST fusion proteins were used to inoculate two rabbits each (0.5 mg of protein/rabbit). Rabbits were boosted three times at 3-week intervals with 0.5 mg of protein/boost/rabbit. Sera were collected and tested on recombinant protein. The polyclonal antisera were passed through columns of GST coupled to cyanogen bromideactivated Sepharose 4 (Amersham Biosciences) to remove GST-specific antibodies. Flow-through from this step was affinity-purified against N4WBP5A-GST or Nedd4-2-GST coupled to cyanogen bromide-activated Sepharose 4. Bound proteins were eluted into Tris-HCl (pH 8.6) with 100 mM glycine (pH 2.5) and dialyzed against PBS. Affinitypurified antibodies were diluted in 50% glycerol and stored at Ϫ20°C.
Northern Blot Analysis-A full-length mouse N4WBP5A cDNA was used to probe a multiple tissue Northern (MTN) blot containing poly(A) ϩ RNA from mouse tissues (CLONTECH). A ␤-actin cDNA probe (CLONTECH) was used as a loading control probe.
Cell Lines and Transfections-M-1 cells were grown in Dulbecco's modified Eagle's medium supplemented with 5% fetal calf serum, and HEK 293T cells were grown in RPMI 1640, 10% fetal calf serum at 37°C with 5% CO 2 . Transfections were performed using FuGENE 6 (Roche Molecular Biochemicals) according to the manufacturer's instructions. For co-transfection of HEK 293T cells 1 g of each plasmid was used.
Immunoprecipitations and Far Western Analysis-Immunoprecipitations of HEK 293T cells were carried out the day after transfection, and M-1 cells were seeded at a density of 3 ϫ 10 6 cells/100-mm dish, and immunoprecipitations were carried out when cells reached confluency. Immunoprecipitations were performed by harvesting cells in lysis buffer (50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 10% glycerol, 1% Triton X-100, 10 mM EDTA, Complete Protease Inhibitor Mixture (Roche Molecular Biochemicals)), preclearing lysates with protein G-Sepharose (Amersham Biosciences) for 2 h, incubating lysates overnight at 4°C with the appropriate antibody (5 g/ml), and then for 2 h at 4°C with protein G-Sepharose. Immunoprecipitates were washed two times in lysis buffer and once in PBS, subjected to SDS-PAGE, and transferred to polyvinylidene difluoride. The following primary antibodies were used: anti-FLAG M2 monoclonal antibody (2 g/ml, Sigma), anti-Nedd4-2 (0.25 g/ml), and anti-Nedd4 monoclonal antibody (0.75 g/ ml). Secondary horseradish peroxidase-conjugated anti-mouse and anti-rabbit IgG antibodies (1:2000, Amersham Biosciences) were used, and detection of bound antibody was achieved using ECL (Amersham Biosciences). Equivalent loadings of purified GST fusion proteins were resolved on SDS-PAGE gels run in duplicate. One gel was stained with Coomassie, whereas the duplicate gel was transferred to nitrocellulose membrane (Schleicher and Schuell). A 32 P-labeled N4WBP5A probe was produced by direct labeling of the GST fusion protein with protein kinase A (New England Biolabs). Glutathione beads containing bound fusion protein were incubated with protein kinase A and [ 32 P]ATP in a buffer containing 20 mM Tris-HCl (pH 7.5), 100 mM NaCl, 12 mM MgCl 2 , and 1 mM dithiothreitol for 60 min at 4°C. Beads were washed 5 times in PBS, and labeled protein was eluted with glutathione buffer. The membrane was blocked in Hyb 75 (28) and hybridized with the 32 Plabeled N4WBP5A protein probe for 4 h at 4°C in Hyb 75. The membrane was washed three times in Hyb 75 and exposed to x-ray film.
Tissue Preparation for Immunofluorescence Microscopy-Kidneys of 6 -8-week-old male Wistar rats (n ϭ 3) were perfusion-fixed, and cryosections were prepared as described previously (31). Cryosections were brought to room temperature, rehydrated in 1ϫ PBS for 5 min, and then blocked for 30 min with 5% normal goat serum in PBS with 1% bovine serum albumin (BSA). The blocking solution was removed, replaced with an affinity-purified rabbit anti-NW4BP5A antibody (or corresponding negative controls, see below) diluted 1:100 in 1% BSA in PBS with 5% normal goat serum added, and the slides incubated overnight at 4°C in a humidified box. The slides were then incubated for 1 h at room temperature with an affinity-purified anti-rabbit IgG conjugated to Alexa 488 as secondary antibody (Molecular Probes, Europe Inc., Leiden, The Netherlands) diluted in 5% normal goat serum in 1% BSA in PBS. After incubation with the secondary antibodies, all sections were washed and mounted with Vectashield (Vector Laboratories Inc., Burlingame, CA). Negative control experiments were carried out by preincubating the N4WBP5A antiserum with an excess of antigenic peptide, by using the preimmune serum and by omitting the primary antibody. Immunofluorescence was visualized using a Zeiss Axioplan 2 microscope with 20 -100ϫ objectives. Images were acquired using a Hamamatsu digital camera and processed using the software package KS300 version 3.0 (Carl Zeiss Ltd., Welwyn Garden City, UK).
Isolation of Oocytes and Injection of cRNA-Xenopus laevis oocytes were prepared and injected as described (32,33). Defolliculated oocytes were injected with various cRNA combinations. For each ENaC subunit or Kir1.1a channel, 1 ng of cRNA was used, whereas 20 ng of cRNA were used for N4WBP5A, Nedd4-2, and CFTR. Injected oocytes were routinely kept in "high sodium" modified Barth's saline (88 mM NaCl, 2. Two-electrode Voltage Clamp Experiments-Unless stated otherwise, oocytes were studied 2 days after injection using the two-electrode voltage clamp technique as described previously (32,33). Oocytes were routinely clamped at a holding potential of Ϫ60 mV. The amiloridesensitive current (⌬I ami ) was determined by subtracting the corresponding current value measured in the presence of 2 M amiloride from that measured prior to the application of amiloride in a NaCl solution (in mM: 95 NaCl, 2 KCl, 1 CaCl 2 , 1 MgCl 2 , 10 HEPES, adjusted to pH 7.4 with Tris). Data are given as mean values Ϯ S.E.; n indicates the number of oocytes; N indicates the number of different batches of oocytes used; significance was evaluated by the appropriate version of Student's t test.
Surface Labeling of Oocytes-Experiments were essentially performed as described recently (33,34) using rat monoclonal anti-HA (clone 3F10, Roche Molecular Biochemicals) or mouse anti-FLAG M2 as primary antibodies (1 g/ml). Peroxidase-conjugated affinity purified F(ab) 2 fragment goat anti-rat IgG (Jackson ImmunoResearch) or peroxidase-conjugated sheep anti-mouse IgG (Chemicon) were used as secondary antibodies (2 g/ml). Chemiluminescence of individual oocytes placed in 50 l of Power Signal ELISA solution (Pierce) was quantified in a Turner TD-20/20 luminometer (Sunnyvale, CA) by integrating the signal over a period of 15 s. Results are given in relative light units (RLU).
Immunoblotting of Oocyte Proteins-Proteins from homogenized oocytes were separated by SDS electrophoresis and transferred to nitrocellulose filters. Blots were blocked in Tris-buffered saline containing 5% milk powder and 0.1% Nonidet P-40 at 4°C. For the detection of ENaC-HA, primary rat-anti-HA monoclonal antibody (100 ng/ml) and secondary peroxidase-conjugated goat anti-mouse antibody (160 ng/ml) were used. For the detection of N4WBP5A, affinity-purified antisera to N4WBP5A (250 ng/ml) and peroxidase-conjugated donkey anti-rabbit Ig (1:2000, Amersham Biosciences) were used. For the detection of FLAG-tagged N4WBP5A and N4WBP5, mouse anti-FLAG M2 (1 g/ml) and peroxidase-conjugated sheep anti-mouse IgG (200 ng/ml) were used. Nitrocellulose filters were incubated in primary antibody for 3 h and in the secondary antibody for 90 min at 20°C. Washes were in Tris-buffered saline, 0.1% Nonidet P-40. Detection was performed with the enhanced luminol reagent (PerkinElmer Life Sciences).

N4WBP5A Is a Putative Transmembrane Protein Expressed in Multiple
Tissues-N4WBP5A (GenBank TM accession number AAL05872) was identified in the GenBank TM data base as a protein highly homologous to N4WBP5. N4WBP5A contains two PY motifs in the amino-terminal half of the protein, and three potential transmembrane domains in the carboxyl-terminal region (Fig. 1, A and B). Both mouse and human homologues of N4WBP5A are present in the data base and share Ͼ90% identity with each other.
Northern blot analysis of RNA derived from mouse tissues suggested that N4WBP5A has two transcripts of 2.4 and 2.1 kb in most tissues (Fig. 1C). In testis, a smaller transcript of 1.3 kb was also detected. N4WBP5A expression was highest in the brain, liver, kidney, and testis with the 2.1-kb transcript predominating in brain, liver, and kidney. Lower level expression was apparent in heart, lung, spleen, and skeletal muscle (in descending order) (Fig. 1C). Similar results were obtained using a human multiple tissue expression array, which additionally showed N4WBP5A expression in colon, salivary gland, placenta, and trachea (data not shown).
N4WBP5A Is Expressed in Native Rat Distal Nephron-Intrarenal distribution of N4WBP5A was determined using immunofluorescence experiments on paraformaldehyde-fixed rat kidney cryosections. Fig. 1D shows that N4WBP5A-specific (i.e. peptide protectable) immunoreactivity is present in distal convoluted tubule and collecting duct cells (upper panel). N4WBP5A can be detected in the cytosol mainly at the basolateral side but also in perinuclear regions and within vesiclelike structures, with a punctate distribution pattern. Outer and inner medullary collecting ducts also exhibit strong intracellular N4WBP5A immunoreactivity (data not shown). Fig. 1D also shows that no N4WBP5A-specific immunofluorescence was detected in the glomeruli, proximal convoluted and proximal straight tubules, thin descending and ascending limbs, and in the parietal Bowman's epithelium. No significant fluorescence can be observed in sections incubated with peptide-absorbed antiserum (Fig. 1D, lower panel) when the preimmune serum was used and when the primary antibody was omitted (not shown).
N4WBP5A Interacts with Nedd4 and Nedd4-2-Like Nedd4, Nedd4-2, a close relative of Nedd4, has been shown recently (19,20,23) to mediate ENaC regulation. Nedd4-2 contains four WW domains, of which WW3 and WW4 interact with the ENaC subunits (20). To test whether N4WBP5A physically interacts with Nedd4 and Nedd4-2, we carried out immunoprecipitation experiments in HEK 293T cells ectopically expressing these proteins. HEK 293T cells were co-transfected with Nedd4 or Nedd4-2-FLAG and vector or pcDNA3-N4WBP5A-FLAG. Immunoprecipitations were performed with anti-Nedd4ab1 (N4) antibody that cross-reacts with both Nedd4 and Nedd4-2 or rabbit preimmune serum, and the blot was probed with anti-FLAG or an anti-Nedd4 monoclonal antibody. Both Nedd4 and Nedd4-2 were found to interact with N4WBP5A ( Fig. 2A). To determine whether endogenous N4WBP5A and Nedd4/ Nedd4-2 interact, immunoprecipitations from the mouse cortical collecting duct cell line M-1 were carried out. M-1 cells are known to endogenously express ENaC (35). By using the Nedd4-2 antibody, but not preimmune serum from the same animal, both Nedd4-2 and N4WBP5A were co-immunoprecipitated (Fig. 2B, lower panel). Because both antibodies (against Nedd4-2 and against N4WBP5A) were raised in rabbits, it is currently not possible to do co-localization immunocytochemistry using double staining. However, we analyzed N4WBP5A and Nedd4-2 localization in M-1 cells using side by side immunocytochemistry. As shown in Fig. 2C, Nedd4-2 is localized to the Golgi and throughout the cytoplasm, whereas N4WBP5A appears as punctated staining throughout the cytosol. Thus, both proteins are present in the cytoplasm of M-1 cells with some apparent overlap of the staining consistent with the coimmunoprecipitation data.
To characterize further the interaction between N4WBP5A and Nedd4-2, we produced GST fusion proteins containing either the wild-type or mutant Nedd4-2 WW domains. These proteins were used in far Western experiments using the amino-terminal domain of N4WBP5A containing both PY motifs as a probe (Fig. 2D). The fusion protein containing all four WW domains showed strong interaction with N4WBP5A (Fig. 2D, last lane). All four individual WW domains, but not their mutant counterparts, were also able to bind N4WBP5A. This result strongly suggests that the interaction between Nedd4-2 and N4WBP5A is mediated via PY/WW domain interactions as is the interaction between N4WBP5 and Nedd4 (29) and N4WBP5A and Nedd4. 2 Coexpression of N4WBP5A Specifically Stimulates ENaC Currents-Because N4WBP5A interacts with Nedd4 and Nedd4-2 and is highly expressed in nephron segments where ENaC is also present, we tested whether N4WBP5A can affect ENaC activity in the Xenopus oocyte system. As shown in sessed in ENaC/N4WBP5A oocytes and ENaC control oocytes at various times after cRNA injection. After 18 h the average ⌬I ami in ENaC/N4WBP5A was similar to that measured in control ENaC oocytes. However, a significant stimulatory effect of N4WBP5A on ⌬I ami was apparent 40 h after cRNA injection, and ⌬I ami was further enhanced 66 and 90 h after cRNA injection. The time course of ENaC stimulation by N4WBP5A was similar to that previously reported (36) for the stimulatory effect of the serum-and glucocorticoid-regulated kinase. Moreover, the finding that initially the two groups of oocytes were capable of expressing similar ENaC currents rules out the possibility that the observed stimulatory effect of N4WBP5A on ENaC currents was due to a nonspecific effect of the co-injection procedure. As N4WBP5A appears to regulate the Na ϩ feedback pathway (see results below), the lack of effect of N4WBP5A at 18 h may be due to the fact that at this time point ENaC expression was low, so that Na ϩ feedback was not operating. At later time points Na ϩ feedback was switched on leading to inhibition of ENaC activity in the group without N4WBP5A.
To investigate the specificity of the stimulatory effect of N4WBP5A on ENaC, we performed similar co-expression experiments with hCFTR or Kir1.1a (ROMK1) which like ENaC are expressed in renal collecting duct cells (37)(38)(39)(40)(41). However, N4WBP5A had no stimulatory effect on Kir1.1a K ϩ currents or on cAMP-activated CFTR currents (data not shown). The lack of effect of N4WBP5A on Kir1.1a or CFTR suggests that the stimulatory effect of N4WBP5A was specific for ENaC.
N4WBP5A Increases Surface Expression of ENaC-The stimulation of ⌬I ami in oocytes co-expressing N4WBP5A may be due to an increased P o of ENaC or to an increase in the overall number of ENaC channels expressed at the cell surface. To investigate this question we assessed ⌬I ami and surface expression of extracellular HA-tagged ENaC (ENaC-HA) using a chemiluminescence assay (33,34). The results shown in Fig. 4 demonstrate that in ENaC-HA/N4WBP5A oocytes surface expression and ⌬I ami were increased by about 4.2-and 3.7-fold, respectively, compared with ENaC-HA control oocytes. This indicated that an increased number of ENaC channels expressed at the cell surface was sufficient to explain the stimulation of ⌬I ami in oocytes co-expressing ENaC and N4WBP5A. The Western blot shown in Fig. 4 demonstrates that ENaC-HA and ENaC-HA/N4WBP5A oocytes express similar levels of ENaC-HA protein. Thus, increased protein synthesis was unlikely to be the cause of the stimulatory effect of N4WBP5A on ENaC surface expression.
N4WBP5A Reduces the Rate of ENaC Retrieval from the Plasma Membrane-N4WBP5A may increase surface expression of ENaC by enhancing its delivery to the cell surface or by inhibiting ENaC retrieval. To assess the rate of ENaC retrieval, we inhibited delivery of new channels to the plasma membrane by adding 18 M brefeldin A (BFA) to oocytes 2 days after injection with cRNA. BFA is a fungal metabolite that inhibits the secretory pathway of newly synthesized proteins without affecting clathrin-mediated endocytosis (42). Fig. 5 illustrates the effect of BFA on ⌬I ami in ENaC and ENaC/ N4WBP5A oocytes. In ENaC oocytes ⌬I ami decreased by about 85% within 4 h after addition of BFA (Fig. 5A), which is in good agreement with data reported previously (12). In non-treated ENaC oocytes ⌬I ami continued to increase throughout the 24-h period examined which suggests that the channel insertion exceeded channel retrieval during this period. After removal of BFA, ⌬I ami partially recovered demonstrating that the effect of BFA was non-toxic, specific, and reversible upon removal of the drug (Fig. 5A). In contrast to its dramatic inhibitory effect on ⌬I ami in ENaC oocytes, BFA had essentially no effect on ⌬I ami in ENaC/N4WBP5A oocytes (Fig. 5B). The resistance of ⌬I ami

FIG. 3. Time course of amiloride-sensitive whole-cell currents (⌬I ami ) in ENaC/N4WBP5A oocytes (open squares) and matched ENaC control oocytes (filled circles).
At time 0 oocytes from the same batch were injected either with cRNA for ␣␤␥ENaC alone (ENaC) or in combination with cRNA for N4WBP5A (ENaC ϩ N4WBP5A). After injection ocytes were kept in high sodium MBS. At the times indicated ⌬I ami (at Ϫ60 mV) was assessed in 10 oocytes from each group. ⌬I ami was significantly larger in ENaC/N4WBP5A oocytes compared with that in ENaC control oocytes 40 (*, p Ͻ 0.05), 66 (***, p Ͻ 0.001), and 90 h (***, p Ͻ 0.001) after injection. n.s., not significant. Similar time course experiments were performed in 3 additional batches of oocytes (data not shown).

FIG. 4. Effect of co-expression of N4WBP5A on surface expression of ENaC-HA and ⌬I ami .
A, experiments were essentially performed as described in Fig. 1. ⌬I ami and surface expression were determined 2 days after cRNA injection. Surface expression (filled bars) is expressed in RLU per 15 s per oocyte. 10 oocytes per group were used for the ⌬I ami measurements and also for detection of surface expression. B, Western blot analysis of total HA-tagged protein in oocyte homogenates from ENaC-HA oocytes and ENaC-HA/N4WBP5A oocytes. The bands at ϳ75 kDa indicate that similar levels of ENaC-HA protein are expressed in both groups. These bands were absent in noninjected control oocytes confirming that they represent the expression of ENaC-HA proteins.
to BFA in oocytes co-expressing N4WBP5A indicates that N4WBP5A stabilizes ENaC channels in the plasma membrane probably by preventing their endocytosis.
Stimulation of ⌬I ami by N4WBP5A Requires Sodium Loading of the Oocytes-An increase in intracellular Na ϩ concentration is known to cause feedback inhibition of ENaC by a complex regulatory pathway involving Nedd4/Nedd4-2 and resulting in channel retrieval from the plasma membrane (17,22,43). Because N4WBP5A may interfere with Nedd4/Nedd4-2 function, co-expression of N4WBP5A may prevent Nedd4/Nedd4-2dependent Na ϩ feedback inhibition and channel retrieval possibly resulting in enhanced ENaC activity. Therefore, we tested the Na ϩ dependence of the N4WBP5A effect by incubating ENaC/N4WBP5A and ENaC oocytes either in the usual high sodium (88 mM) or in low sodium (1 mM) MBS for 2 days after injection with cRNA. The I/V plots shown in Fig. 6A illustrate that after incubation in high sodium ⌬I ami was significantly higher in ENaC/N4WBP5A oocytes than ⌬I ami in ENaC control oocytes consistent with the data shown in Figs. 3-5. After 2 days in high sodium ⌬I ami (at Ϫ60 mV) was increased by 515 Ϯ 73% (n ϭ 160; N ϭ 16; p Ͻ 0.001) in ENaC/N4WBP5A oocytes compared with that in ENaC oocytes. In contrast, the stimulatory effect of N4WBP5A was essentially abolished in oocytes that were maintained in low sodium (Fig. 6B). In ENaC/ N4WBP5A oocytes incubated in low sodium, ⌬I ami averaged 96 Ϯ 5% that in control ENaC oocytes (n ϭ 30; N ϭ 3) (Fig. 6B). These findings indicate that the stimulatory effect of N4BP5A is dependent on maintaining the oocytes in the presence of high extracellular Na ϩ .
It is well known that ENaC-expressing oocytes kept in the presence of high extracellular Na ϩ will become severely Na ϩloaded (30). Indeed, in oocytes maintained in high sodium after injection with cRNA the reversal potential of ⌬I ami was Ϫ1 and Ϫ2 mV for ENaC and ENaC/N4WBP5A oocytes, respectively (Fig. 6A). These values indicate that the apparent intracellular Na ϩ concentration [Na ϩ ] i-app of these oocytes was essentially equal to that in the bath solution (95 mM). In contrast, in oocytes maintained in low sodium, the reversal potential of ⌬I ami was 21 mV for both ENaC and ENaC/N4WBPA oocytes (Fig. 6B). By using this reversal potential one can estimate a value of 41 mM for [Na ϩ ] i-app . It should be pointed out that this value probably reflects the Na ϩ concentration in a cytosolic compartment close to the plasma membrane and not necessarily the bulk Na ϩ concentration inside the cell. During voltage clamp experiments, entry of Na ϩ through ENaC acutely increases the Na ϩ concentration in the unstirred compartment close to the membrane. Thus, using the reversal potential of  group). B, I/V plots from oocytes incubated for 2 days in low sodium MBS (10 oocytes per group). To obtain I/V plots, voltage pulse protocols were performed using consecutive 400-ms step changes of the clamp potential from Ϫ60 to Ϫ120 mV up to ϩ60 mV in 20-mV increments. ⌬I ami was determined by subtracting the whole-cell current traces (using the last 100 ms) recorded in the presence of amiloride (2 M) from those recorded prior to its addition.
⌬I ami to calculate [Na ϩ ] i-app tends to overestimate the true intracellular sodium concentration (32). Nevertheless, these estimates clearly indicate that the intracellular sodium concentration was substantially lower in the cells incubated in low sodium than in those incubated in high sodium.
On average ⌬I ami was 876 Ϯ 121% (n ϭ 30, N ϭ 3) larger in ENaC oocytes incubated in low sodium compared with the corresponding ⌬I ami in ENaC oocytes incubated in high sodium. This is consistent with findings from a previous study that concludes that sodium-dependent down-regulation of ENaC can be prevented by maintaining oocytes in the presence of a low extracellular sodium concentration (11). The inability of N4WBP5A to stimulate ENaC activity in oocytes incubated in low sodium and its stimulatory effect in oocytes incubated in high sodium suggest that in the latter oocytes N4WBP5A prevents the sodium-dependent down-regulation of ENaC and thereby increases the number of ENaC channels at the cell surface. Indeed, in ENaC/N4WBP5A oocytes incubated in high sodium, ⌬I ami reached about the same level (14.76 Ϯ 2.21 A; n ϭ 30; N ϭ 3) as in ENaC oocytes incubated in low sodium (19.02 Ϯ 3.12 A; n ϭ 30; n ϭ 3). Thus, extracellular sodium removal and co-expression of N4WBP5A have a similar stimulatory effect on ENaC currents.
N4WBP5A Prevents Sodium Feedback Inhibition of ENaC-To further investigate whether N4WBP5A interferes with sodium feedback inhibition of ENaC, we combined ⌬I ami measurements with surface detection of ENaC-HA. Two days after cRNA injection and incubation in low sodium ⌬I ami in ENaC-HA/N4WBP5A was very similar to that in ENaC-HA control oocytes averaging 7.90 Ϯ 1.91 A (n ϭ 10) and 8.17 Ϯ 1.50 A (n ϭ 10), respectively (Fig. 7A). Oocytes were subsequently divided into two groups and either transferred to a bath solution containing 95 mM Na ϩ or further maintained in the presence of 1 mM extracellular Na ϩ . In ENaC-HA and ENaC-HA/N4WBP5A oocytes maintained in 1 mM Na ϩ ⌬I ami remained high at a relatively constant level. In contrast, exposure of ENaC-HA oocytes to 95 mM Na ϩ reduced ⌬I ami to 0.66 Ϯ 0.07 A (n ϭ 10) within 3 h demonstrating the presence of substantial sodium feedback inhibition of ENaC. Interestingly, in ENaC-HA/N4WBP5A oocytes exposure to 95 mM Na ϩ had a much smaller effect with ⌬I ami remaining at 6.51 Ϯ 0.77 A (n ϭ 10) after 3 h and at 5.41 Ϯ 0.48 A (n ϭ 10) after 9 h (Fig.  7A). These current measurements suggest that N4WBP5A prevents sodium-stimulated ENaC retrieval from the plasma membrane. This was confirmed by ENaC-HA surface expression measurements summarized in Fig. 7B. As long as the oocytes were maintained in the presence of 1 mM extracellular Na ϩ , surface expression of ENaC was similar in ENaC-HA oocytes compared with that in ENaC-HA/N4WBP5A oocytes. However, upon exposure to 95 mM extracellular Na ϩ surface expression was dramatically reduced in ENaC-HA oocytes, whereas it was largely preserved in ENaC-HA/N4WBP5A oocytes. These findings are consistent with the ⌬I ami data shown in Fig. 7A and confirm that acute sodium-stimulated ENaC retrieval is prevented by N4WBP5A.
Lack of Effect of N4WBP5A on ENaC with Liddle's Syndrome Mutation-The finding that N4WBP5A reduces sodium feedback inhibition was similar to the observation that mutations causing Liddle's syndrome reduce sodium-dependent downregulation of ENaC (11). Moreover, the lack of effect of BFA on ENaC/N4WBP5A oocytes was reminiscent of the previously reported finding (12) that BFA had almost no effect on the activity of ENaC with Liddle's syndrome mutation because the retrieval of the mutated channel by clathrin-mediated endocytosis was defective. Taken together, our results suggest that in the presence of N4WBP5A the wild-type ENaC channel adopts a similar phenotype as the channel with Liddle's syndrome mutation. We therefore hypothesized that ENaC with Liddle's syndrome mutation cannot be stimulated by N4WBP5A because endocytotic retrieval was already compromised in the mutated channel. Indeed, the data shown in Fig. 8 demonstrate that N4WBP5A failed to stimulate significantly ⌬I ami and channel surface expression in oocytes expressing ENaC with Liddle's syndrome mutation (␣␤ R564X ␥ ENaC). These findings suggest that the stimulatory effect of N4WBP5A requires the presence of an intact endocytotic retrieval mechanism that can be inhibited by N4WBP5A to enhance ENaC surface expression.
N4WBP5A Competes with Nedd4-2 for the Regulation of ENaC Endocytosis-Biochemical data suggest that N4WBP5A binds to Nedd4-2 (Fig. 2). Thus in Xenopus oocytes heterologously expressed N4WBP5A may sequester endogenous xNedd4-2. Sequestration of Nedd4-2 will reduce Nedd4-2-mediated endocytosis and degradation of ENaC resulting in an FIG. 7. Absence of sodium feedback inhibition in oocytes coexpressing ENaC-HA and N4WBP5A. Oocytes were incubated for 2 days in low sodium MBS. A, ⌬I ami was assessed for ENaC-HA/ N4WBP5A oocytes (filled circles) and for ENaC-HA matched control oocytes (open circles). Then oocytes were divided into a control group incubated continuously in low sodium MBS (dotted lines) and a group incubated in high sodium MBS (solid lines), and ⌬I ami was assessed after 3 and 9 h. Each point represents the mean of 10 oocytes. B, a chemiluminescence assay was used to detect surface expression of extracellular hemagglutinin-tagged ENaC (ENaC-HA) in oocytes co-expressing ENaC-HA and N4WBP5A (filled bars) and in ENaC-HA control oocytes (open bars). Similarly with the protocol in A, the oocytes were divided between a control group incubated continuously in low sodium MBS (0 h, low sodium MBS) and a group that was switched to high sodium MBS 3 h before the start of the surface assay and maintained in high sodium throughout the assay (6 h, high sodium MBS). Single oocyte chemiluminescence was measured from 10 oocytes per group, and mean values are expressed in RLU per 15 s per oocyte.
increased surface expression of ENaC. To confirm a functional competition between N4WBP5A and Nedd4-2 ( Fig. 9), we coexpressed ENaC with varying amounts of N4WBP5A and Nedd4-2. Co-injection of 20 ng of Nedd4-2 cRNA reduced ⌬I ami to 0.97 Ϯ 0.16 A (n ϭ 10) compared with ⌬I ami in ENaC control oocytes which averaged 2.77 Ϯ 0.19 A (n ϭ 10; p Ͻ 0.001). This is consistent with the well described inhibitory effect of Nedd4 on ENaC function (43). On the other hand, co-injection of 20 ng of N4WBP5A cRNA had the usual stimulatory effect with ⌬I ami averaging 7.74 Ϯ 0.78 A (n ϭ 10; p Ͻ 0.001). Interestingly, co-injection of 20 ng of N4WBP5A and 20 ng of Nedd4-2 cRNA resulted in ⌬I ami that averaged 3.33 Ϯ 0.36 A (n ϭ 10) and was similar to that in ENaC control oocytes. Thus, the stimulatory effect of N4WBP5A was cancelled by the inhibitory effect of Nedd4-2. Indeed, the stimulatory effect of N4WBP5A was suppressed by Nedd4-2 in a dosedependent manner (Fig. 9). Thus, it is conceivable that in native epithelia, which express both Nedd4-2 and N4WBP5A, the relative abundance of each protein will be critical for determining ENaC activity by regulating its retrieval rate.
The PY Motifs and the TM Domains of N4WBP5A Are Essential for Its Function-N4WBP5A has two PY motifs and three transmembrane (TM) domains (Fig. 1A). We used an N4WBP5A version in which both PY motifs were mutated (N4WBP5A-PY), and a carboxyl-terminal truncation mutant that has intact PY motifs but lacks all three TM domains (N4WBP5Astop) to establish the domains that are important for N4WBP5A function. Fig. 10A illustrates that both mutants failed to stimulate ENaC activity, although Western blots confirmed their expression. ⌬I ami averaged 1.64 Ϯ 0.27 A (n ϭ 10) in ENaC control oocytes and 16.11 Ϯ 2.44 A (n ϭ 10; p Ͻ 0.001) in ENaC/N4WBP5A oocytes. In contrast to the stimulatory effect of N4WBP5A, ⌬I ami averaged 1.53 Ϯ 0.22 A (n ϭ 10; p Ͼ 0.05) and 1.66 Ϯ 0.27 A (n ϭ 10; p Ͼ 0.05) in ENaC/N4WBP5Astop and ENaC/N4WBP5A-PY oocytes, respectively. These results suggest that the PY motifs of N4WBP5A are required for its stimulatory effect on ENaC. The PY motifs are likely to mediate N4WBP5As interaction with Nedd4-2 thereby preventing Nedd4-2 from down-regulating ENaC. However, our data also show that the presence of the PY motifs, which are preserved in N4WBP5Astop, was not sufficient to prevent Nedd4-2-mediated down-regulation of ENaC. In particular the TM domains were likely to be needed for the correct localization/stabilization of N4WBP5A in the appropriate cellular compartment to enable it to interact with Nedd4-2.
Absence of N4WBP5A Surface Expression and Lack of Effect of the Homologous Protein N4WBP5 on ENaC-To rule out a direct effect of N4WBP5A on ENaC at the level of the plasma membrane, we investigated whether FLAG-tagged N4WBP5A was transported to the cell surface using the chemiluminescence assay (Fig. 10B). The signals detected in ENaC/N4WBP5A, ENaC/N4WBP5A-PY, and N4WBP5A oocytes were very low and similar to the background signal measured in control oocytes expressing ENaC alone. In contrast a 1000-fold higher signal was detected in ENaC-FLAG oocytes that served as positive controls (7). The possibility that N4WBP5A was retained intracellularly by endogenous xNedd4-2 is unlikely, because the N4WBP5A-PY, which has no stimulatory effect and therefore does not seem to interact with Nedd4-2, was also not expressed at the cell surface (Fig.  10B). Interestingly, oocytes coexpressing ENaC and FLAGtagged N4WBP5, a protein closely related to N4WBP5A and also containing two PY motifs and three transmembrane domains (29), showed a significant chemiluminescence signal (Fig. 10B). These data indicate that N4WBP5A was not transported to the cell surface, whereas N4WBP5 was delivered to the cell surface efficiently. Importantly, co-expression of N4WBP5 with ENaC failed to stimulate significantly ⌬I ami (Fig. 10C) which in ENaC/N4WBP5 oocytes averaged 136 Ϯ 22% of that in ENaC control oocytes (n ϭ 80; n ϭ 8). We conclude that N4WBP5A was confined to an intracellular location where it interacted with Nedd4 in a complex manner to prevent Nedd4 from stimulating ENaC retrieval. On the other hand, N4WBP5, although structurally similar to N4WBP5A and with the ability to bind Nedd4-2 efficiently, 3 traffics to the plasma membrane but did not appear to affect Nedd4-ENaC interaction. The difference in the trafficking 3 A. Fotia and S. Kumar, unpublished data. Oocytes were all injected with 3 ng of cRNA for ENaC (1 ng for each subunit). In addition different amounts of N4WBP5A and Nedd4-2 were co-injected as indicated by the ratios N4WBP5A:Nedd4-2 (ng/ng) below the bars. After injection oocytes were incubated in high sodium MBS for 2 days. Each bar represents the mean ⌬I ami of 10 oocytes. n.s., not significant. behavior of the two homologues suggested that the intracellular localization of N4WBP5A was important for its interaction with Nedd4 and its stimulatory effect on ENaC surface expression. DISCUSSION We have described here the regulation of ENaC in Xenopus oocytes by a novel protein N4WBP5A which is highly expressed in native renal collecting duct and other tissues known to express ENaC. We have demonstrated that the increased ENaC activity in the presence of N4WBP5A is due to increased surface expression of ENaC. This effect is likely due to an increased stability of ENaC in the plasma membrane and does not appear to be dependent on the transport of newly synthesized ENaC to the surface.
The effect of N4WBP5A is dependent on the presence of high Na ϩ , suggesting that it acts to inhibit the sodium feedback pathway that involves Nedd4/Nedd4-2-mediated retrieval of ENaC from the plasma membrane. Whereas both Nedd4 and Nedd4-2 protein are present in many mammalian tissues, only a single homologue that is more similar to Nedd4-2 than to Nedd4 is known to be present in Xenopus (18 -20, 27). Recent results demonstrate that in Xenopus oocytes that contain endogenous xNedd4-2, Xenopus, mouse and human Nedd4-2 proteins, and to a much smaller extent rat and human Nedd4 proteins, can regulate surface expression of ENaC (19,20,23). Therefore, the relative abundance of Nedd4-2 and N4WBP5A proteins may affect levels of surface ENaC expression in vivo.
The stimulatory effect on ENaC requires intact PY motifs and transmembrane domains of N4WBP5A. It is likely that the PY motifs are needed to mediate the interaction of N4WBP5A with a WW domain containing protein. Our biochemical evidence that N4WBP5A PY motifs are required for its interaction with Nedd4/Nedd4-2 is consistent with N4WBP5A interfering with the activity of endogenous xNedd4-2. The additional requirement for intact TM domains suggests that these domains of N4WBP5A are also important to mediate Nedd4-2 sequestration. N4WBP5A is highly related to N4WBP5, a protein  Oocytes were incubated in high sodium MBS for 2 days after injection with cRNA. A, ⌬I ami was assessed in control ENaC oocytes (filled bar) and in ENaC oocytes co-expressing either the FLAG-tagged wild-type N4WBP5A or various N4WBP5A mutants (open bars). N4WBP5A D108X (N4WBP5Astop) lacks the three transmembrane domains of the carboxyl terminus, and N4WBP5A-PY had the two aminoterminal PY motifs mutated. Western blot analysis of total wild-type or mutant N4WBP5A protein confirmed similar levels of protein expression for both mutants. Note that the intense band at 12 kDa represents the expected protein size of N4WBP5Astop. B, surface expression of N4WBP5, N4WBP5A, and N4WBP5A-PY that were FLAG-tagged at their predicted extracellular carboxyl terminus. The chemiluminescence signal of ENaC/N4WBP5A, ENaC/N4WBP5A-PY, and N4WBP5A oocytes was similar to the background signal of ENaC control oocytes. In contrast, FLAG-tagged ENaC oocytes (ENaC-FLAG) used as positive controls gave at least a 1000-fold higher signal compared with back ground. N4WBP5/ENaC oocytes had also a high signal, indicating that the predicted extracellular carboxyl terminus of the N4WBP5/5A pro-largely associated with the Golgi complex (29), whereas N4WBP5A appears to be localized to post-Golgi vesicles. 4 The finding that N4WBP5 had no stimulatory effect on ENaC activity in the Xenopus oocyte system supports the conclusion that the effect of N4WBP5A is specific and that sequestration of Nedd4-2 requires more than the presence of binding domains because both proteins, N4WBP5 and N4WBP5A, bind Nedd4-2 with similar affinities via the PY motif/WW domain interaction. 3 How might N4WBP5A affect the Nedd4-2-mediated ENaC control? One possibility is that vesicles containing N4WBP5A sequester Nedd4-2 preventing its localization to ENaC at the plasma membrane. When necessary, Nedd4-2 may be released permitting translocation to the plasma membrane where it can associate with ENaC and mediate ubiquitin-dependent downregulation of the membrane-associated ENaC. This hypothesis is supported by the fact that Nedd4-2 can interact with both the ENaC subunits and N4WBP5A. Our data suggest that N4WBP5A itself is not transported to the cell surface. However, the vesicles containing N4WBP5A may traffic to the cytosolic side of the membrane allowing sequestration of Nedd4-2 when necessary, e.g. during hormonal stimulation of ENaC surface expression.
In addition to the tissues that express ENaC, N4WBP5A is also present in other cell types and tissues. Both Nedd4 and Nedd4-2 are also expressed in many tissues that lack ENaC (14,18,19,28). This suggests that N4WBP5A may also play a role in modulating the capacity of Nedd4/Nedd4-2 to regulate other cellular proteins that are controlled by Nedd4/Nedd4-2mediated ubiquitination. However, N4WBP5A has no stimulator effect on CFTR or Kir1.1a, both of which are apically located channel proteins expressed in a range of epithelia. This suggests that the stimulatory effect of N4WBP5A is rather specific for ENaC. Moreover, the lack of effect of the closely related protein N4WBP5 on ENaC function and the specific expression of N4WBP5A in nephron segments known to express ENaC make N4WBP5A a strong candidate for a physiologically relevant regulator of ENaC activity. So far the effect of N4BWP5A on ENaC activity is demonstrated in oocytes, and it remains to be determined that it also occurs in native renal epithelial cells. However, the finding that the majority of collecting duct cells show N4WBP5A-specific immunoreactivity clearly indicates that ENaC and N4WBP5A are co-expressed in native renal collecting duct principal cells. The hormonal regulation of ENaC activity in these cells is essential for the fine-tuning of renal sodium absorption and hence for body sodium balance and the long term regulation of arterial blood pressure. Thus, it will be an interesting topic for future research to investigate whether N4WBP5A is a component of the molecular mechanisms involved in ENaC regulation, e.g. by aldosterone.
The regulation of ENaC by N4WBP5A described here is somewhat reminiscent of the recently described control of CFTR plasma membrane expression by CAL, a PDZ domaincontaining, Golgi-associated protein (44). CAL favors retention of CFTR within the cell by interacting with the carboxyl terminus of CFTR, whereas Na ϩ /H ϩ exchanger regulatory factor, NHE-RF, promotes CFTR cell surface expression by competing with CAL for binding of CFTR. The similarity between CFTR regulation by CAL and ENaC regulation by N4WBP5A suggests that antagonistic competition between regulatory protein ligands may be an important new mechanism for regulating surface expression of membrane proteins.
In conclusion, we have demonstrated that the recently iden-tified Nedd4/Nedd4-2-binding protein N4WBP5A enhances ENaC surface expression most likely by preventing Nedd4-2mediated channel retrieval from the plasma membrane. N4WBP5A also attenuates Na ϩ feedback inhibition of ENaC which is known to involve the Nedd4/Nedd4-2 pathway (17,22,43). The effect of N4WBP5A is reminiscent of the action of the aldosterone-induced kinase and the serum-and glucocorticoidregulated kinase, which also increase surface expression of ENaC probably by PY motif-dependent binding to Nedd4-2 and its phosphorylation (24,25). Thus, as illustrated schematically in Fig. 11, Nedd4-2 seems to be an integrator of various pathways regulating ENaC activity. It remains a challenge for future research to elucidate the relative importance of N4WPB5A and its functional interrelationship with the different pathways and mechanisms involved in ENaC regulation under physiological and pathophysiological conditions.