Cell lines were imported under and transfection was approved under Environmental Protection Authority (EPA) approvals APP201858 and APP201859 respectively. All cell lines were grown in a humidified atmosphere of 5% CO₂ at 37°C. HEK293 and COS-7 cells were maintained in DMEM with 10% FBS, 1% penicillin/streptomycin and 1% glutamine (all Life Technologies, from Thermo Fisher Scientific, Auckland, New Zealand). M1 mouse CCD cells were maintained in DMEM:Ham’s F12 medium (without phenol red) supplemented with 10% FBS, 1% Glutamax, 1% penicillin/streptomycin and 1 mM dexamethasone (Life Technologies). Fischer rat thyroid (FRT) cells were maintained in Ham’s F12 (Sigma, New Zealand) supplemented with 10% FBS and 1% penicillin/streptomycin. Mouse CCD clone 1 (mCCD_cl1) cells were maintained in a defined medium as described in Gaeggeler et al. (2005). For experiments investigating the effect of aldosterone on p11 expression, M1 or mCCD_cl1 cells were grown to 80% confluency whereupon normal growth medium was replaced with serum and dexamethasone free medium. After 24 h, aldosterone (Sigma, to a final concentration of 10 nM) or a vehicle control (ethanol) was added to M1 or mCCD_cl1 cells for 1, 3, or 24 h. Cells were then lysed for immunoblotting as described below. ENaC plasmids, including HA (hemagglutinin, YPYDVPDYA) epitope tagged versions, were described previously (McDonald et al., 1995; Wiemuth et al., 2007; Fronius et al., 2010). FRT, HEK293 and COS-7 cells were transfected with either HA-tagged or wild type α-, β-, and γ-ENaC subunits using Lipofectamine 2000 (Thermo Fisher Scientific) or CaCl₂ precipitation methods.

Recombinant DNA experiments were conducted under EPA approval APP201859. Glutathione S-transferase (GST) pull-down assays were performed as described previously (Wiemuth et al., 2007). Briefly, p11 proteins tagged with GST (O’Kelly and Goldstein, 2008), or GST alone were produced in BL21 E. coli and purified on glutathione-agarose beads (Sigma, Auckland, New Zealand). COS-7 cell lysates containing HA-tagged α-, β-, or γ-ENaC were precleared for 1 h at 4°C with GST bound to glutathione-agarose beads. Cleared lysates were then incubated with 50 μg of GST-p11 or GST alone for 3 h at 4°C. The glutathione-agarose beads, fusion proteins and any interacting proteins were collected by centrifugation. After extensive washing with 1% Triton in PBS buffer, samples were resolved on a 8% SDS-PAGE gel, transferred to PVDF membrane (Roche Biochemicals, distributed by Sigma New Zealand) at 45 mA for 2 h (Hoefer semi-dry transfer). Membranes were probed with anti-HA antibodies (H6908, Sigma New Zealand, 1:1000) applied in blocking buffer (20 mM Tris, pH 7.4, 150 mM NaCl, 0.1% Tween 20, all Sigma) and 5% non-fat dry milk (Fonterra, New Zealand). The secondary antibody HRP-conjugated goat anti-rabbit (A6154, Sigma) was used at 1:10,000 dilution.

To evaluate endogenous p11 and annexin II protein expression in M1 or mCCD_cl1 cells, cells were solubilized in lysis buffer (138 mM NaCl, 20 mM Tris-HCl, 1% Triton X-100, 10 mg/ml PMSF and 3 μg/ml aprotinin, pH 7.4, all from Sigma) and protein concentration evaluated according to Bradford’s protocol (Bio-Rad Laboratories, Auckland, New Zealand). Protein samples were loaded onto a 15% SDS-PAGE and transferred onto PVDF membrane as above. The primary antibodies used were: rabbit anti-p11 at 1:1000 (ab76472, Abcam, Melbourne, VIC, Australia), or mouse anti-p11 at 1:1000 (ab89438, Abcam) and rabbit annexin II 1:5000 (41803, Abcam), overnight at 4°C.

For co-immunoprecipitation, HEK293 cells were co-transfected with p11-myc-DDK plasmid (RC204992, OriGene Technologies, Rockville, MD, United States) and plasmids encoding α-, β_HA-, and γ-ENaC subunits. After 24 h, cells were lysed in 1% Triton in 1x TBS (50 mM Tris-HCl, 150 mM NaCl, pH 7.4–7.5) with protease inhibitors. Mouse anti-p11 antibody was used to immunoprecipitate p11 along with protein G-agarose beads. The isolated immuno-protein complex was then washed and used for Western blot analysis (12% SDS-PAGE) with anti-HA and anti-p11.

Fischer rat thyroid cells were co-transfected with 1 μg each of plasmids encoding α-, β_HA, or γ-ENaC with 20 pmol control or p11 siRNA (Sigma-Aldrich, St. Louis, MO, United States). Cells were washed five times in ice-cold PBS 48 h post-transfection. Surface proteins were labeled as described in Butterworth et al. (2005). Briefly, cells were biotinylated in borate buffer (85 mM NaCl, 4 mM KCl, 15 mM Na₂B₄O₇, 10 μg/ml PMSF, 1 μg/ml pepstatin, 2 μg/ml leupeptin, 2 μg/ml aprotinin pH 9.0) with 1 mg/mL Sulfo-NHS-LC-Biotin (Life Technologies New Zealand Limited) for 20 min on ice with shaking. After removal of biotin, signal was quenched with PBS containing 10% FBS. Cells were washed again five times in ice-cold PBS and lysed in biotinylation lysis buffer (50 mM EDTA, 10 mM Tris, 1% NP-40, 0.4% sodium deoxycholate, pH 7.4) for 10 min on ice while shaking. Lysates were transferred to a 1.5 mL tube and centrifuged for 10 min at 16.1 rcf at 4°C. To equal amounts of proteins (as determined by Bradford assay), 100 μL NeutrAvidin (Pierce^TMNeutrAvidin^TMAgarose, Life Technologies New Zealand Limited) beads were added and incubated overnight at 4°C with rotation. Following incubation, samples were centrifuged at 0.4 rcf to pellet beads (surface), while supernatant (cytosolic) fractions were transferred to new tubes. Beads were washed four times with biotinylation lysis buffer, with centrifugation at 0.4 rcf and then boiled at 95°C for 5 min. Samples were then separated by 8% SDS-PAGE gel followed by immunoblot analysis.

HEK293 cells were co-transfected with p11-myc-DDK plasmid and plasmids encoding α-, β_HA-, and γ-ENaC subunits as above. A first round of immunoprecipitation was performed using either anti-HA antibody to immunoprecipitate βENaC-HA or anti-DDK (TA50011, OriGene Technologies) to immunoprecipitate p11-myc-DDK protein from the pre-cleared lysate. The isolated immuno-protein complex was then washed and subjected to a second round of immunoprecipitation using anti-DDK or anti-HA to isolate proteins associated with the βENaC-p11 complex from the first round of immunoprecipitation. The immune-protein complexes were then divided into two tubes in equal amounts, one was used for Western blot analysis (12% SDS-PAGE) with anti-HA and anti-myc DDK (TA50011, OriGene, United States) to confirm β-ENaC_HA and p11 expression respectively, while the other was processed for mass spectrometry (Figure 3A). Co-immunoprecipitated p11, ENaC and other interacting proteins were denatured in SDS-PAGE sample buffer, and proteins were separated on individual 12% SDS-PAGE gels using standard techniques. The gels were stained with Coomassie G250 and protein lanes cut into three fragments for liquid chromatography coupled with electrospray ionization linear ion trap [LC-ESI LTQ] Orbitrap tandem mass spectrometry [MS/MS]. Fractions were subjected to in-gel protein digestion with trypsin as described previously (Shevchenko et al., 1996). Each digested fraction was concentrated using a centrifugal vacuum concentrator and reconstituted in a 15 μl aqueous solution of 2% (vol/vol) acetonitrile (ACN) supplemented with 0.2% formic acid for LC-ESI LTQ Orbitrap analyses.

Peak lists were processed through the Proteome Discoverer 1.1 software (Thermo Scientific, San Jose, CA, United States) for all ESI LTQ Orbitrap data using the software’s default settings. All peak lists were then searched with an in-house Mascot server (version 2.1.0; Matrix Science) against an amino acid sequence database combining all predicted and translated p11 and all entries from the NCBI non-redundant sequence database, matching the taxa Homo sapiens. Mascot search settings allowed for full tryptic peptides with up to three missed cleavage sites and variable modifications of carbamidomethyl (C) and oxidation (M). The precursor and fragment mass tolerances were set to 10 ppm and 0.8 Da. To evaluate the false-discovery rate (FDR), all peak lists were searched against a decoy database using identical search settings. The decoy database was built using the decoy database tool at the Trans-Proteomic Pipeline (TPP; Seattle Proteome Center), comprising the reversed sequence entries of the aforementioned combined database. The FDR was calculated by determining the number of false-positive peptide hits from the decoy search versus the number of peptide identifications from the true search using the same Mascot score as a significance threshold.

Only peptide hits with an individual ion score of >25 (Mascot significance threshold at a p of < 0.05) were accepted as significant identifications. This resulted in an FDR of <0.02 for all searches. Significant protein identification required at least two significant peptide hits covering different sequences of the protein. In addition, a protein that was identified by single peptide-based protein identification in one experiment was also confirmed by a different peptide identification covering another sequence stretch in one of the other experiments was considered a significant multipeptide identification.

mCCD_cl1 cells were exposed to aldosterone or vehicle as above. Cells were lysed in Trizol (Life Technologies) and RNA was isolated according to manufacturer’s instructions. cDNA was prepared using the PrimeScript RT Reagent Kit (Perfect Real time, Medi’Ray, New Zealand) and qPCR was carried out with SYBR Premix Ex Taq (Tli Rnase H Plus) ROX plus (Medi’Ray, New Zealand). KiCqStart^TM primers for 18S (Fwd 5′-CAGTTATGGTTCCTTTGGTC-3′, Rev 5′-TTATCTAGAGTCACCAAGCC-3′), p11 (Fwd 5′-GGAAAATCAAAAGGACCCTC-3′, Rev 5′-CCTTCTGCTTCATGTTTACTAC-3′), Annexin II (Fwd 5′-CTCCAGAAAGTGTTCGAAAG-3′, Rev 5′-TTCTAATCAGGACCTTGTCTC-3′), and Sgk1 (Fwd 5′-ATCCTGAAGAAGAAAGAGGAG-3′, 5′-ATGTAGTCCAGGACAAAGTAG-3′) were from Sigma. Reactions were analyzed on a CFX Connect PCR machine (Bio-Rad). Analysis of gene expression was calculated using the following equation: ΔCt = 2^{-(Ctgene-Ctreference)}.

cDNAs of human ENaC subunits and human p11 (p11-myc-DDK) were linearized and cRNAs synthesized in vitro with a T7 mMESSAGE mMACHINE^TM transcription kit (Ambion, from Thermo Fisher Scientific). Defolliculated oocytes (Stages V–VI) isolated from Xenopus laevis in accordance with guidelines approved by The University of Otago Animal Ethics Committee were injected with 0.75 ng of cRNA of each ENaC subunit. For experiments investigating the effect of p11 co-expression on ENaC function, oocytes were co-injected with 0.75 ng of p11 cRNA. After injection, oocytes were maintained at 18°C in low Na⁺ modified Barth’s solution [83 mM NMDG, 5 mM NaCl, 1 mM KCl, 0.33 mM Ca(NO₃), 0.41 mM CaCl₂, 10 mM HEPES, and 1% penicillin/streptomycin; pH 7.4. all from Sigma] for 1–2 days before ENaC recordings.

Two-electrode voltage clamp recordings were performed on ENaC-expressing oocytes using microelectrodes filled with 1 M KCl connected to a Turbo TEC 05 amplifier (NPI, Tamm, Germany). Oocytes were bathed and continuously perfused in ND96 solution (96 mM NaCl, 1 mM KCl, 1.8 mM CaCl₂, 1 mM MgCl₂, and 5 mM HEPES, pH 7.4, all Sigma) After impalement, membrane potentials were allowed to stabilize before voltage clamping at -60 mV. Amiloride-sensitive, ENaC specific Na⁺ currents were then recorded as the difference in current before and after the addition of amiloride (10 mM final concentration, Tokyo Chemical Industry, from Lab Supply, Dunedin, New Zealand).

Fischer rat thyroid cells were seeded onto 12 mm Snapwell^TM membranes (COR3801, Corning via In Vitro Technologies, New Zealand) at a density of 4 × 10⁵ cells per Snapwell^TM and incubated overnight. Cells were transfected with 0.067 μg of plasmids encoding α-, β-, or gENaC and 20 pmol siRNA targeting p11, or control siRNA, using Lipofectamine^TM 3000 (Invitrogen). The transfection medium was replaced with FRT full-growth media supplemented with 10 μM amiloride 6 h post-transfection. After 72 h, epithelia were mounted onto a modified Ussing chamber connected to a multichannel V/A clamp (Physiologic Instruments, San Diego, CA, United States) via a DI-720 data acquisition system (DataQ Instruments). Data was recorded using the Acquire and Analyze 2.3 program (Physiologic Instruments) running on a PC. The apical and basolateral surfaces were bathed in 1x Ringer’s solution (in mM 135 NaCl, 2.4 K₂HPO₄, 10 HEPES, 1.2 CaCl₂, 1.2 MgCl₂, pH 7.4–7.5), kept at 37°C and bubbled with O₂. The epithelia were clamped under short-circuit conditions and the amiloride-sensitive short circuit current (I_sc-amil) was measured. The I_sc-amil was determined as the difference in current recorded prior to and after the addition of amiloride (5 μM final concentration) into the apical bathing solution. The relative I_sc-amiloride was obtained by normalizing the I_sc-amiloride to that of cells transfected with αβγ ENaC alone in a parallel experiment. Transepithelial resistance (Rt) was monitored by applying repetitive 5 mV pulses for 1 s at 120 s intervals.

All data are expressed as the mean ± SEM of n experiments. Statistical significance between two groups was determined using an unpaired t-test or one-sample t-test at the p level indicated. ANOVA was used for multiple comparisons with a post hoc Bonferroni’s test to identify significant differences between groups at the p level indicated. Data was analyzed for normal distribution using the D’Agostino-Pearson normality test on Prism 7 (GraphPad Software, United States).

Both the activity of ENaC at the membrane and the number of membrane resident channels are influenced by multiple regulatory proteins directly interacting with ENaC channel subunits (Snyder et al., 2002; Biasio et al., 2004; Condliffe et al., 2004; Butterworth, 2010). Since p11 interacts with several apical membrane proteins (van de Graaf et al., 2003; Borthwick et al., 2007), we investigated whether p11 directly binds to ENaC channel subunits. A GST-p11 protein was used in pull-down experiments with HA-tagged α-, β-, or γ-ENaC subunits. As shown in the lower panel of Figure 1, GST-p11 was able to pull-down all three ENaC subunits either as individual subunits or as a trimeric complex of α-, β_HA-, and γ-ENaC. In contrast, no interaction was detected between GST alone and ENaC (Figure 1, first lane, and Supplementary Figure S1). These results demonstrate that p11 is capable of binding to each ENaC subunit either as a single protein or to the heteromultimeric ENaC complex.

To further evaluate the interaction between ENaC and p11 and to confirm that this occurred in intact cells, HEK293 cells were co-transfected with p11-myc-DDK and α-, β_HA-, and γ-ENaC. Complexes between p11-myc-DDK and ENaC were immunoprecipitated with an anti-p11 antibody and protein G-agarose beads before being analyzed by Western blotting using an anti-HA antibody. In control samples where only p11 was expressed and protein precipitated using a mouse p11 antibody, no ENaC protein was detected by immunoblot using the anti-HA antibody. In contrast, in samples co-expressing p11 and ENaC, p11 was able to co-immunoprecipitate bENaC which was revealed by immunoblotting by anti-HA antibody (Figure 2 and Supplementary Figure S2). This data indicates that p11 and ENaC are able to interact in mammalian cells, in support of our in vitro pull-down results.

In order to identify putative interaction partners of the p11-ENaC complex, the immunopurified fraction of HEK293 lysates transfected with tagged ENaC and p11 constructs were isolated by gel electrophoresis, in-gel trypsinized and the resulting peptide fragments analyzed by mass spectroscopy (Figure 3A). Table 1 lists the proteins identified by mass spectrometric sequencing of MS spectrum peaks (Figure 3B) that are known to have roles in protein trafficking pathways trafficking pathways, and Supplementary Table S1 lists all proteins identified. Several tubulin proteins of the microtubule network required for transporting vesicles to the plasma membrane were detected with high Mascot and unique peptide scores. In addition, HSP-90 and annexin A6 proteins involved in exocytic trafficking associate with the p11-ENaC complex. Interestingly, the mass spectrometry analysis also identified the E3 ubiquitin protein ligase RNF2 protein confirmed to play a role in ubiquitination pathways (Xia et al., 2014) suggesting that the p11-ENaC complex persists at the plasma membrane before being targeted for endocytosis and subsequent recycling or degradation. Altogether, these results confirm that the p11-ENaC complex is able to interact with multiple proteins involved in controlling ENaC trafficking to and from the plasma membrane.

The epithelial Na⁺ channel has a pivotal role in Na⁺ reabsorption of the CCD of the kidney where its expression is upregulated by aldosterone. To verify whether CCD cells also endogenously express p11 and if aldosterone can increase this expression, we examined the level of p11 protein expression in the mouse M1-CCD cell line in the absence and presence of aldosterone. Under standard culture conditions, a robust expression of endogenous p11 protein was detected (Figure 4A and Supplementary Figure S3). Removal of serum from the media for 24 h resulted in p11 expression levels that were not significantly different from full medium (Figure 4A,B). To examine whether endogenous p11 expression was regulated by aldosterone, M1 cells were exposed to aldosterone (10 nM final concentration) or vehicle (ethanol) for 1, 3, or 24 h before quantifying p11 protein levels. Since typical housekeeping genes such as GAPDH and actin have been shown to be aldosterone sensitive in CCD cells (Perlewitz et al., 2010), we normalized p11 expression to the expression of annexin II which is not altered by aldosterone (Perlewitz et al., 2010). Similar to p11, annexin II was also endogenously expressed in M1 cells (Figure 4A) which is important given its role as a p11 interacting protein. Quantification of p11 protein level normalized to annexin A2 shows that p11 protein level was not significantly altered by short-term (1–3 h) aldosterone (10 nM) exposure, remaining at similar levels to serum free conditions (Figure 4B). However, incubating M1 cells in aldosterone (10 nM) for 24 h caused a higher p11 protein level to approximately the same level as observed in full medium although ANOVA did not detect a significant difference (Figure 4B). In contrast, treating cells with an ethanol vehicle for 24 h had no significant effect on p11 protein level (Figure 4B). These results demonstrate that CCD cells endogenously express p11 and annexin II, however aldosterone did not stimulate a significant increase in p11 protein synthesis.

As it is possible that the subclone of M1 cells used did not express the mineralocorticoid receptor (Náray-Fejes-Tóth et al., 2004), and because we found SGK1 was not markedly increased in the presence of aldosterone (data not shown), we repeated these experiments in mCCD_cl1 cells to test if aldosterone altered protein levels of p11 and annexin II. mCCD_cl1 cells were subjected to aldosterone (10 nM final concentration) or vehicle (ethanol) for 1, 3, or 24 h before quantifying p11 and annexin II protein levels using Western blot. Figure 4C,D and Supplementary Figure S4 shows that neither short or long term aldosterone treatment changed p11 or annexin II protein level. Finally, to test for a short term effect of aldosterone on p11 or annexin II mRNA (similar to the effect reported for SGK1, Náray-Fejes-Tóth et al., 1999), we subjected mCCD_cl1 to 1 h aldosterone treatment, and quantified mRNA levels using qRT-PCR. Figure 4E shows that although SGK1 mRNA was increased after 1 h aldosterone treatment the mRNA levels of p11 and annexin II were not changed.

Having demonstrated that p11 physically interacts with ENaC, we next investigated whether p11 affected ENaC channel function by heterologous expression of p11 and ENaC in Xenopus oocytes. Amiloride-sensitive current was used to quantify the amount of ENaC-specific current using two electrode voltage clamp recordings of ENaC expressing oocytes. Figure 5A shows that both baseline current and the amount of amiloride-sensitive current were higher in oocytes co-expressing ENaC and p11 compared to oocytes expressing ENaC alone. On average, p11 co-expression caused a significant increase in relative amiloride-sensitive current at -60 mV from 1.0 ± 0.09 (n = 17) in oocytes injected with ENaC RNA alone to 1.4 ± 0.12 (n = 18) in oocytes injected with both ENaC and p11 RNA (Figure 5B). Water injected control oocytes had no significant amiloride-sensitive current (data not shown). These results suggest that, in addition to interacting with ENaC channels, p11 exerts a positive effect on ENaC channel function at the plasma membrane.

To confirm and extend the observation that overexpression of p11 increases ENaC current in oocytes, we used a model mammalian epithelia, FRT epithelia, to test the effect of reduced p11 levels on ENaC current. Plasmids encoding the α-, β-, and γ ENaC subunits were introduced into FRT epithelia by transfection together with siRNA targeting p11 (si-p11), or a control siRNA (si-control). Figure 6A shows that epithelia expressing α-, β-, and γ- ENaC generated an amiloride-sensitive short-circuit current. In the presence of si-p11 the amiloride-sensitive current was reduced compared to si-control; and pooled data (Figure 6B) showed a significant reduction in ENaC current when p11 protein level was reduced (n = 6–7, p < 0.01). These results support the hypothesis that p11 is required for efficient delivery of ENaC to the cell surface but could also represent changes in ENaC endocytosis, open probability or proteolytic cleavage. Western blot analysis confirmed that si-p11 reduced p11 protein level in FRT cells by an average of 75% (n = 6–7, p < 0.01, also see Figure 7A for representative western blot showing p11 knockdown).

To further test the hypothesis that p11 regulates the cell surface population of ENaC we carried out cell surface protein biotinylation assays. FRT cells were transfected with plasmids encoding α-, β_HA-, and γ ENaC together with equal amounts of si-control, or si-p11 (20 or 40 pmol). After 24 h, the cells were incubated with cell-impermeable sulfo-NHS-LC-biotin to label proteins at the cell surface. After cell lysis, biotinylated proteins were captured on Neutravidin^TM beads while the supernatant was retained as the cytosolic fraction. Figure 7A shows successful detection of β_HA ENaC in the cell surface and cytosolic fractions. The amount of β_HA ENaC at the cell surface was significantly reduced (p < 0.01, n = 3) in the presence of si-p11 (40 pmol) indicating that p11 alters the cell surface population of ENaC (Figure 7A,B and Supplementary Figure S5).