Characterization of a novel phosphorylation site in the sodium–chloride cotransporter, NCC

Abstract

The sodium–chloride cotransporter, NCC, is essential for renal electrolyte balance. NCC function can be modulated by protein phosphorylation. In this study, we characterized the role and physiological regulation of a novel phosphorylation site in NCC at Ser124 (S124). Novel phospho-specific antibodies targeting pS124-NCC demonstrated a band of 160 kDa in the kidney cortex, but not medulla, which was preabsorbed by a corresponding phosphorylated peptide. Confocal microscopy with kidney tubule segment-specific markers localized pS124-NCC to all distal convoluted tubule cells. Double immunogold electron microscopy demonstrated that pS124-NCC co-localized with total NCC in the apical plasma membrane of distal convoluted tubule cells and intracellular vesicles. Acute treatment of Munich–Wistar rats or vasopressin-deficient Brattleboro rats with the vasopressin type 2 receptor-specific agonist dDAVP significantly increased pS124-NCC abundance, with no changes in total NCC plasma membrane abundance. pS124-NCC levels also increased in abundance in rats after stimulation of the renin–angiotensin–aldosterone system by dietary low sodium intake. In contrast to other NCC phosphorylation sites, the STE20/SPS1-related proline–alanine-rich kinase and oxidative stress-response kinases (SPAK and OSR1) were not able to phosphorylate NCC at S124. Protein kinase arrays identified multiple kinases that were able to bind to the region surrounding S124. Four of these kinases (IRAK2, CDK6/Cyclin D1, NLK and mTOR/FRAP) showed weak but significant phosphorylation activity at S124. In oocytes, ³⁶Cl uptake studies combined with biochemical analysis showed decreased activity of plasma membrane-associated NCC when replacing S124 with alanine (A) or aspartic acid (D). In novel tetracycline-inducible MDCKII-NCC cell lines, S124A and S124D mutants were able to traffic to the plasma membrane similarly to wildtype NCC.

Key points

• The sodium–chloride cotransporter, NCC, is essential for renal electrolyte balance and its function can be regulated by protein phosphorylation

• Here we report the role and regulation of a novel phosphorylation site in NCC at Ser124

• Ser124 phosphorylation plays a role in mediating full NCC transport activity, but does not seem to be involved in NCC trafficking

• Various physiological stimuli such as vasopressin and aldosterone regulate the abundance of the Ser124 phosphorylation status and other phosphorylation sites in NCC

• Unlike other known phosphorylation sites in NCC, the STE20/SPS1-related proline–alanine-rich kinase and oxidative stress-response kinases (SPAK and OSR1) were not able to phosphorylate NCC at Ser124

• The results demonstrate that phosphorylation of NCC is a major factor in determining the function of NCC under various physiological conditions

Introduction

The renal thiazide-sensitive sodium–chloride cotransporter, TSC or NCC, is a member of the SLC12 family of electroneutral cation-coupled Cl⁻ cotransporters. This family also includes the Na⁺–K⁺–2Cl⁻ cotransporters, NKCC1 and NKCC2. NCC is expressed in the apical plasma membrane and subapical compartments of the distal convoluted tubule (DCT), and is the major NaCl transport pathway in this segment, absorbing between 5 and 10% of the glomerular filtrate (reviewed by Gamba, 2005). The important role of NCC in cardiovascular and renal physiology and pathophysiology is emphasized by the autosomal recessive disease Gitelman's syndrome, which results from genetic mutations in NCC and is characterized by hypokalaemia, hypomagnesaemia, metabolic alkalosis and hypocalciuria (Mastroianni et al. 1996; Simon et al. 1996; Lemmink et al. 1998; Monkawa et al. 2000).

Knowledge of how NCC is regulated is slowly emerging. Recently, attention has focused on the regulatory role of NCC phosphorylation. NCC contains several conserved phosphorylation sites in the amino terminal end. In mouse and rat NCC, phosphorylation at threonine53 (T53), T58 and serine71 (S71) (corresponding to T55, T60 and S73 in human NCC) are essential mediators of NCC activity (Pacheco-Alvarez et al. 2006; Richardson et al. 2008). For example, replacing T58 with a phosphorylation-deficient alanine prevents activation of NCC in response to hypotonic low Cl⁻ conditions in cultured HEK 293 cells and in Xenopus laevis oocytes (Pacheco-Alvarez et al. 2006; Richardson et al. 2008). Phosphorylation of T53 and T58 (and T48) is mediated via the STE20 (sterile 20)-like kinases SPAK (STE20/SPS1-related proline–alanine-rich kinase) and OSR1 (oxidative stress-responsive kinase-1). SPAK and OSR1 can directly phosphorylate NCC through interaction with a single N-terminal RFXI motif (Richardson et al. 2008). Whether other kinases could be involved remains unknown. Phosphorylation of NCC at T53, T58 and S71 can be enhanced by a variety of physiological stimuli, such as arginine vasopressin (AVP) (Mutig et al. 2010; Pedersen et al. 2010), ANGII (Talati et al. 2010; van der Lubbe et al. 2011) and aldosterone (Chiga et al. 2008; Vallon et al. 2009).

Feric et al. (2011) identified a novel phosphorylation site, S124, in the amino terminus of rat NCC via phosphoproteomic profiling of renal cortical membrane proteins. Bioinformatic studies demonstrated that this site is moderately conserved amongst other species, but is not conserved in the family members NKCC1 and NKCC2 (Feric et al. 2011). In this study, we examine the regulation of NCC via phosphorylation at S124. We demonstrate that pS124-NCC is associated with the apical plasma membrane and intracellular vesicles of DCT cells, where its abundance is increased by AVP treatment or dietary low salt conditions. Mutation of the S124 site results in decreased NCC activity. Furthermore, the kinases SPAK and OSR1 are unlikely to be involved in phosphorylation of S124-NCC.

Methods

Antibodies

Novel rabbit polyclonal antisera against NCC (Slc12a3) phosphorylated at S124 (pS124-NCC) were generated against a proprietary sequence by PhosphoSolutions (Aurora, CO, USA). The immune serum from two rabbits (nos. 3034 and 3035) was double affinity purified using both non-phosphorylated and phosphorylated peptides, resulting in pS124-specific antibodies. Other NCC antibodies used were pT53, pT58 and pT53/58 NCC (Pedersen et al. 2010), and a polyclonal rabbit antibody against total NCC (SPC-402D; StressMarq, Biosciences Inc., Victoria, Canada). Furthermore, a total AQP2 antibody (N-20; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), a mouse monoclonal antibody against calbindin D-28K (Research Diagnostics, Acton, MA, USA) and polyclonal sheep antibodies against total SPAK/OSR1 and pS325-OSR1 (Zagorska et al. 2007) were used.

Animal studies

In this study we utilized samples from previously published or new studies performed in our laboratory. All animal protocols (including Xenopus oocyte studies described below) comply with the European Community guidelines for the use of experimental animals and were performed in accordance to licences for the use of experimental animals issued by the Danish Ministry of Justice. Before each experiment, rats had free access to standard rat chow and water.

Protocol I: standard rat kidney tissue samples

Munich–Wistar rats were anaesthetized with isoflurane, the right kidney was clamped and removed for immunoblotting, and the left kidney was perfusion fixed with 3% paraformaldehyde in 0.1 m sodium cacodylate, pH 7.4, through the abdominal aorta (Pedersen et al. 2010).

Protocol II: short-term dDAVP infusion of Brattleboro rats

Homozygous Brattleboro rats deficient in AVP were treated i.v. with 1 ng dDAVP (deamino-Cys-1, d-Arg-8 vasopressin; Sigma-Aldrich, St Louis, MO, USA) in 200 μl saline per animal, and eight saline-injected rats served as controls. The rats were anaesthetized after 15 or 60 min (four rats for each group), the right kidney removed for protein preparation and the left kidney perfusion fixed as described above. Between injection of dDAVP and fixation of the kidney, the 60 min group had free access to water but not food. The 15 min group was kept under anaesthesia (Pedersen et al. 2010).

Protocol III: short-term dDAVP infusion of Munich–Wistar rats

Four rats were treated with s.c. injections of 1 ng dDAVP (Sigma-Aldrich) in 200 μl saline per animal, and four vehicle-injected rats served as controls. After 30 min, rats were anaesthetized, the right kidney was removed and the left kidney was perfusion fixed (Fenton et al. 2008).

Protocol IV: water loading/restriction of Munich–Wistar rats

Rats were kept in metabolic cages and fed a gelled diet containing either 15 ml of water (restricted) or 45 ml of water (loaded). After 5 days, the rats (five per group) were anaesthetized and kidneys were processed for immunoblotting and perfusion-fixation (Fenton et al. 2008).

Protocol V: altered dietary sodium intake in Munich–Wistar rats

Rats (five per group) were housed in metabolic cages and fed a gelled diet containing 15 g of 0.013% Na⁺ (low, Altromin no. 10103600), 0.24% Na⁺ (normal) or 3.37% Na⁺ (high) (extra NaCl added to the ‘low’ diet to get ‘normal’ and ‘high’ diets) and 10 ml of water. Rats had free access to water throughout the study. Urine was collected on the last 2 days for analysis. After 5 days, the rats were anaesthetized, blood was collected from the inferior vena cava and the kidneys were processed for immunoblotting and perfusion-fixation. The collected urine was centrifuged at 1000 g for 1 min to clear sediments. Osmolality was measured on an Advanced Wide-Range Osmometer 3W2 (Advanced Instruments Inc., Norwood, MA, USA). Blood plasma was prepared by collecting blood in Li-Heparin tubes, inverting 8–10 times, followed by centrifugation at 3000 g for 5 min at 4°C within 1 h after collection. Urine and plasma electrolyte levels (sodium, potassium, chloride and creatinine) were analysed by the Clinical Pathology Laboratory at the Medical Research Council (Harwell, UK). Plasma aldosterone levels were assessed by the Coat-a-Count Aldosterone Kit according to the manufacturer's protocol (Siemens, Denmark).

Protocol VI: dietary sodium restriction of Munich–Wistar rats

Rats were housed in metabolic cages and fed a gelled diet containing either 15 g of 0.013% Na⁺ (low, Altromin no. 10103600) or 0.24% Na⁺ (normal, Altromin no. 10100000) and 10 ml of water. Rats had free access to water throughout the study. After 3 days, the rats (four per group) were anaesthetized and kidneys were processed for immunoblotting and perfusion-fixation.

Immunoblotting and dot-blotting

Peptides of equal length that were either phosphorylated or non-phosphorylated at specific residues (pS124-NCC, S124-NCC, pS127-NCC or pT53-pT58-NCC) were synthesized by PhosphoSolutions. The peptides (5 ng) were spotted onto nitrocellulose membranes and probed with antibodies no. 3034 or no. 3035 (corresponding to the pS124-NCC antibodies raised in two different rabbits) overnight at 4°C. Antibody–antigen reactions were detected using ECL Plus Western Blotting Detection Reagents (GE Healthcare, Denmark). The preparation of tissue samples and immunoblotting were as previously described (Fenton et al. 2007). A horseradish peroxidase-conjugated secondary antibody (Dako P448, goat anti-rabbit IgG) was used at 1:5000, and antibody–antigen reactions were visualized using SuperSignal West Femto Chemiluminescent Substrate (Thermo Scientific, Denmark). Semi-quantitative data were obtained by analysis of band densities and calculated as relative abundance ratios for each individual sample for each time point or stimulant. All reported values are mean ± SEM. One-way ANOVA or Student's unpaired t tests were used for statistical comparisons where appropriate. Values were considered statistically significant at P < 0.05.

Immunohistochemistry, confocal laser scanning microscopy and immunogold electron microscopy

Procedures for preparation of tissue for light and confocal microscopy have been described previously (Fenton et al. 2007). A peroxidase-conjugated secondary antibody (P448; Dako) was used for detection of labelling for light microscopy, and goat anti-rabbit Alexa488- or 555-conjugated secondary antibodies (Invitrogen) for confocal microscopy. A Leica TCS SL confocal microscope was used for imaging. Pre-incubation experiments were performed by combining pS124-NCC antibody with a 10-fold molar excess of either S124-phosphorylated or non-phosphorylated synthetic peptides followed by incubation overnight at 4°C, before application to tissue sections. The double immunogold electron microscopy technique has previously been described in detail (Moeller et al. 2009a). Labelling of total NCC and pS124-NCC antibody was visualized with goat anti-rabbit IgG conjugated to 5 nm or 15 nm colloidal gold particles, respectively.

Quantification of apical NCC abundance

To assess the degree of apical membrane associated with total NCC following AVP treatment, two different methods were employed.

Confocal laser scanning microscopy

Tissue sections from Protocol II were assessed following labelling with total NCC (SPC-402D). Images were obtained using an HCX PL FLUOTAR 100×/1.30–0.60 oil-immersion objective. The dynamic range of the acquisition using a Leica TCS SL confocal microscope (PMT offset and gain, sampling period, and averaging) was set for each tissue section to obtain the maximal signal-to-noise ratio without saturated pixels. Multiple regions of interest (two per tubule, five tubules per animal), representing 5 μm single cross-sections through labelled DCT cells from the apical cell pole through to the basolateral side were selected using ImageJ software. Using the plot profile function, total pixel intensities in the initial 0.4 μm (representing the most apical NCC labelling) were analysed by calculating the area under the curve using Graphpad prism software.

Quantitative immunogold electron microscopy

All tissue processing, staining and counting procedures have been described in detail previously (Sandberg et al. 2006). Briefly, tissue generated from Protocol III were assessed following labelling with total NCC (SPC-402D), with a minimum of five complete DCT cells from different tubules per animal analysed from sections oriented approximately at right angles to the apical cell membrane and showing negligible background over mitochondria and nuclei. Gold particles within 40 nm of the apical plasma membrane were classified as apical and all other gold particles were classified as intracellular. All quantitative data were analysed using Student's unpaired t test, with P < 0.05 considered significant.

Non-radioactive kinase assays

Kinases and recombinant proteins were obtained from the Division of Signal Transduction, University of Dundee, UK. Assays were modified from Filippi et al. (2011). Either constitutively active SPAK (T233E) (1 μg), active OSR1 (T185E) (1 μg), inactive SPAK (D212A) (1 μg) or inactive OSR1 (D164A) (1 μg) kinases were incubated with 5 μg MO25α and a peptide (0.5 μg, TGANSEKSPGC) corresponding to the region surrounding S124-NCC or a purified recombinant protein corresponding to the first 100 amino acids of NCC (Richardson et al. 2008), in 1× assay buffer (50 mm Tris-HCl, pH 7.5, 0.1 mm EGTA, 0.1% 2-mercaptoethanol, 10 mm MgCl₂, 0.1 mm ATP) at 30°C for 1 h. Subsequently, 1.5 μl of each reaction was used in dot-blotting analysis, with pS124-NCC and pT53-pT58-NCC peptides as positive controls. Additional non-radioactive assays using SPAK and OSR1 were performed using the ADP-Glo assay kit from Promega (Madison, WI, USA) according to the manufacturer's protocol. For these assays, the S124-NCC peptide was used at 200 or 500 μm, and a CATCHtide peptide (CRRHYYYDTHTNTYYLRTFGHNTRR) was used as a positive control.

Bioinformatics

To identify possible kinases responsible for S124-NCC phosphorylation, custom analysis of rat NCC (Uniprot P55018) was performed by Kinexus (Vancouver, Canada) using a proprietary kinase predictor algorithm. The algorithm forms the basis of the open-access online repository of known and predicted human phosphorylation sites (http://www.phosphonet.ca/).

Protein kinase microarrays

Protein kinase microarrays were performed by Kinexus Bioinformatics. Briefly, a biotinylated peptide for the region surrounding S124-NCC and corresponding to biotin-ETGANAEKAPGEPVR was synthesized (underlined residue corresponds to S124). In this peptide, the S124 site was modified to an alanine residue to create a pseudo-substrate peptide that would be better bound to kinases on the microarray in the presence of ATP. These peptides were incubated with the KPKM-200 microarray at 10 and 50 μm and sites of peptide–kinase interaction were detected using fluorescent streptavadin.

Protein kinase profiling assays

A radioisotope assay format was used for evaluation of potential protein kinases that target S124-NCC. Synthetic peptides used for the kinase profiling assays corresponded to ETGANSEKSPGEPVR (underlined residue corresponding to S124), or specific control substrates. Protein kinase assays (in duplicate) were performed in a final volume of 25 μl containing 5 μl of diluted active protein kinase target (∼10–50 nm final protein concentration), 2.5 μl of stock solution of substrate (10–250 μm), 10 μl of kinase assay buffer or protein kinase activator in kinase assay buffer and 4.5 μl of [³³P]ATP (250 μm stock solution, 0.8 μCi). The assay was initiated by the addition of [³³P]ATP and the reaction mixture was incubated at ambient temperature for 20–40 min depending on the protein kinase target. After the incubation period, the assay was terminated by spotting 10 μl of the reaction mixture onto a multiscreen phosphocellulose P81 plate. The multiscreen phosphocellulose P81 plate was washed three times for approximately 15 min each in a 1% phosphoric acid solution. Radioactivity on the P81 plate was counted in the presence of scintillation fluid in a Trilux scintillation counter. Blank controls were set up for each protein kinase target, which included all the assay components except the addition of the appropriate substrate (replaced with an equal volume of assay dilution buffer). The corrected activity for each protein kinase target was determined by removing the blank control value.

³⁶Cl⁻ uptake studies in X. laevis oocytes

If not otherwise stated, all reagents were from Sigma-Aldrich. A rat NCC cDNA (a gift from Dr G. Gamba, Universidad National Autónoma de Mexico, Mexico City, Mexico) corresponding to the full open reading frame was subcloned into the pXOOM vector (Jespersen et al. 2002). Mutations in NCC were introduced using site-directed mutagenesis (Stratagene) and standard methodologies. Phosphorylation ‘deficient’ or ‘mimicking’ mutations were obtained by substituting serine or threonine at S124, T48, T53 and T58 with alanine (A) or aspartic acid (D), respectively. All constructs were verified by sequencing. The vectors containing the cDNAs were linearized downstream from the poly-A segment and in vitro transcribed using mMessage Machine (Ambion) according to the manufacturer's instructions. cRNA was extracted with MEGAclear (Ambion). Oocytes were surgically removed from anaesthetized frogs and the follicular membrane was removed by incubation in Kulori medium (90 mm NaCl, 1 mm KCl, 1 mm CaCl₂, 1 mm MgCl₂, 5 mm Hepes, pH 7.4, 182 mosmol l⁻¹) containing 10 mg ml⁻¹ collagenase (Type 1; Worthington) and 1 mg ml⁻¹ trypsin inhibitor (Sigma) as previously described (Fenton et al. 2010). cRNA was microinjected into the defolliculated X. laevis oocytes (50 ng RNA per oocyte) as previously described (Moeller et al. 2009b). After the final collection, the frogs were killed by decapitation during anaesthesia. For uptakes, a minimum of 10 oocytes were used for each construct per experiment. The oocytes were incubated overnight in 100 mm sodium gluconate, 2 mm KCl, 1 mm CaCl₂, 1 mm MgCl₂ and 10 mm Hepes, pH 7.4. The experiment was initiated by placing the oocytes in 100 mm NaCl, 2 mm KCl, 1 mm CaCl₂, 1 mm MgCl₂ and 10 mm Hepes, pH 7.4, containing 10 μm bumetanide (to block endogenous NKCC1) and 3 μCi ml^{−1 36}Cl⁻ for 20 min. The oocytes were subsequently washed four times in ice-cold 100 mm ChCl, 2 mm KCl, 1 mm CaCl₂, 1 mm MgCl₂ and 10 mm Hepes. Each oocyte was dissolved in 10% SDS, added to Optifluor scintillation fluid (Packard) and counted in a Tri-carb scintillation counter (Packard). For control experiments, 100 μm metolazone (from DMSO-based stock solution of 15 mm) was added to the uptake media (Supplementary Fig. 1). Data were assessed by one-way ANOVA, followed by Bonferroni's multiple comparisons test where appropriate. P < 0.05 was considered significant.

Preparation of oocyte membranes

Oocytes were prepared as described in detail previously (Moeller et al. 2009b). For oocyte outer membrane preparations, 10 oocytes of each construct were used per experiment. Oocytes were washed in Mes-buffered saline (MBSS: 80 mm NaCl, 20 mm Mes, pH 6.0), transferred to MBSS with 0.005% subtilisin A and incubated at room temperature for 10 min. Subsequently, the oocytes were incubated at 4°C for 60 min in cold MBSS with 1% Ludox AM-30. Finally, the oocytes were incubated in cold MBSS with 0.1% polyacrylic acid for 60 min at 4°C. The oocytes were washed in MBSS after each incubation step, all performed with mild agitation. Oocytes were homogenized in cold homogenization buffer (HbA) (5 mm MgCl₂, 5 mm NaH₂PO₄, 1 mm EDTA, 80 mm sucrose, 20 mm Tris, pH 7.4) containing protease inhibitors and subjected to differential centrifugation at 14, 22, 31 and 42 g for 30 s, adding new HbA to the pelleted membrane after each centrifugation for washing. After a final centrifugation at 16,000–17,000 g for 20 min, the membrane-containing pellet was dissolved in gel sample buffer (0.25 m SDS, 30% glycerol, 250 mm Tris, bromophenol blue, 0.4 m dithiothreitol) and heated for 15 min at 65°C. Oocyte total membrane preparations were performed by homogenizing two oocytes per construct in HbA, centrifuging at 250 g for 10 min and centrifuging the supernatant at 16,000–17,000 g for 20 min before resuspending the pellet in gel sample buffer.

Generation of a NCC-inducible MDCKII cell line

Rat NCC (rNCC) cDNA was amplified using specific primers that included an additional Kozak sequence and an AflII restriction site at the 5′ end and a NotI restriction site at the 3′ end. The PCR product was subcloned into the pcDNA5/FRT/TO/TOPO vector (Invitrogen). Phosphorylation-deficient/mimicking mutations were introduced at the S124 site (S124A or S124D) using site-directed mutagenesis (Stratagene) and standard methodologies. A tetracycline-inducible MDCK type II cell line containing a Flp Recombinase Target (FRT) site in its genome was used as the host cell line (Nejsum et al. 2011). These cells were cotransfected with the pcDNA5/FRT/TO/TOPO-rNCC constructs and the pOG44 encoding Flp recombinase using lipofectamine 2000 (Invitrogen), and positive clones were selected using 500 μg ml⁻¹ hygromycin B. After clonal selection, cells were grown in Dulbecco's modified Eagle medium GlutaMAX with 10% DBS, 5 μg ml⁻¹ blasticidin (selection for tet repressor in the host line) and 150 μg ml⁻¹ hygromycin B. PCR on cell genomic DNA was used to verify incorporation of rNCC into the genome. To induce NCC protein expression in MDCKII-rNCC cells, 10 μg ml⁻¹ tetracycline was added for a minimum of 16 h prior to experiments.

Cell surface biotinylation assays

Cells were grown on six-well filter plates (Costar, 0.4 μm) to 100% confluency for several days. Cells were washed in ice-cold PBS-CM (10 mm PBS, 1 mm CaCl₂, 0.1 mm MgCl₂, pH 8.0) and incubated with mild agitation for 90 min at 4°C in ice-cold biotinylation buffer (10 mm triethanolamine, 2 mm CaCl₂, 125 mm NaCl, pH 7.5) containing 1.5 mg ml⁻¹ final concentration of sulfosuccinimidyl 2-(biotinamido)-ethyl-1,3-dithiopropionate (Sulfo-NHS-SS-biotin; Pierce) added to either the apical or the basolateral compartment. Subsequently, cells were washed in ice-cold quenching buffer (PBS-CM, 50 mm Tris-HCl, pH 8) followed by two washes in PBS-CM. Cells were scraped and samples centrifuged at 4000 g for 5 min at 4°C. Cell pellets were dissolved in 500 μl lysis buffer (150 mm NaCl, 5 mm EDTA, 50 mm Tris-HCl, pH 7.5, 1% Triton X-100) with 5 μg ml⁻¹ leupeptin, 100 μg ml⁻¹ phefablock and 10 μg ml⁻¹ phosphatase inhibitor mixture (Sigma), and sonicated. The samples were spun at 10,000 g for 5 min at 4°C and a fraction of the supernatant was retained for total NCC protein estimates. The remaining supernatant was transferred to spin columns containing NeutrAvidin gel slurry (Pierce) and incubated for 60 min at room temperature with end-over-end mixing. After extensive washing, biotinylated proteins were eluted in gel sample buffer and heated for 15 min at 60°C. All experiments were performed in duplicate or triplicate.

Results

Specificity of pS124-NCC antibodies

The specificity of the novel antibodies for pS124-NCC was initially determined by dot blotting (Fig. 1A). The pS124-NCC no. 3035 antibody showed specificity for only the phosphorylated peptide corresponding to the region surrounding the S124 residue. The pS124-NCC no. 3034 antibody recognized the pS124-NCC peptide, but also to a lesser extent the non-phosphorylated peptide. The pS124-NCC no. 3035 antibody was used in all subsequent studies.

Figure 1. Specificity of the pS124-NCC antibody.

A, dotblot analysis demonstrates that the anti-pS124-NCC no. 3035 antibody only recognizes a phosphorylated S124 NCC peptide (*phosphorylated amino acid) and not other related peptides. B, immunoblotting of rat kidney samples with anti-pS124-NCC. Immuoblotting of whole kidney (WK), cortex (CTX) or inner medulla (IM) homogenates from Munich–Wistar rats with anti-NCC and anti-pS124-NCC antibody. A band of ∼160 kDa is observed in WK and CTX but not in IM.

Localization and distribution of pS124-NCC in the kidney

Immunoblots of rat kidney samples probed with the pS124-NCC antibody demonstrated a smeared band of ∼160 kDa in whole kidney and cortex samples, but not in inner medulla (Fig. 1B). A similarly sized band was observed on an identical blot probed with a total NCC antibody. No band was observed at 160 kDa on a similar blot that was probed with the pS124-NCC antibody previously incubated overnight with the immunizing peptide (data not shown), demonstrating specificity of the antibody.

To investigate the distribution of pS124-NCC in normal rat kidney, we initially performed immunoperoxidase microscopy using paraffin-embedded kidney samples from normal Munich–Wistar rats. pS124-NCC labelling was detected along the whole DCT (Fig. 2A). At higher magnification, the labelling was observed at the apical plasma membrane domain of DCT cells (Fig. 2B). To confirm that for immunohistochemistry, and at the concentration used in our studies, the pS124-NCC antibody was specific, we performed immunolabelling controls. Pre-absorption of the pS124-NCC antibody with a synthetic phosphopeptide corresponding to the targeted phosphorylation site completely abolished labelling, whereas pre-absorption with a synthetic non-phosphorylated peptide corresponding to the same region did not affect pS124-NCC labelling (not shown), demonstrating specificity of the antibody for immunolabelling.

Figure 2. Localization of pS124-NCC in Munich–Wistar rat kidney.

A, using immunohistochemistry, anti-pS124-NCC antibody shows labelling of all DCT segments. B, higher magnification revealing distinct pS124-NCC labelling of the apical membrane of DCT cells. C, confocal laser scanning micrograph demonstrating that pS124-NCC (green) colocalizes with total NCC (red). D and E, pS124-NCC (green) partially colocalizes with calbindin (red), a late DCT and connecting tubule (CNT)-specific marker. F, pS124-NCC (green) does not colocalize with the CNT and collecting duct principal cell-specific marker, aquaporin-2 (red). G, double immunogold electron micrograph demonstrating total NCC (small arrows) and pS124-NCC (large arrows) in the apical domain of DCT cells. Labelling was associated with the apical plasma membrane (upper left panel), but also intracellular vesicles and the Golgi apparatus (lower right panel). Gold particle diameter: 5 nm (pS124-NCC, large arrows) and 15 nm (total NCC, small arrows).

Double immunofluorescence labelling of Munich–Wistar rat kidney sections and confocal laser scanning microscopy revealed co-localization between pS124-NCC and total NCC (Fig. 2C), indicating that at the resolution of confocal microscopy, pS124-NCC is distributed similarly to total NCC. Double labelling of pS124-NCC and the late DCT and connecting tubule (CNT)-specific marker calbindin demonstrated only partial cellular co-localization (Fig. 2D and E). Some kidney tubules were only labelled with pS124-NCC or calbindin alone, suggesting that pS124-NCC has a greater abundance in the early than in late DCT. At the high concentration of antibody used for double labelling, some non-specific staining was observed in a subset of proximal tubules. No co-localization of pS124-NCC was observed with the CNT and collecting duct principal cell-specific marker aquaporin-2 (Fig. 2F).

To confirm the subcellular distribution of pS124-NCC we used immunogold electron microscopy on Munich–Wistar rat kidney sections (Fig. 2G). Both total NCC and pS124-NCC were associated with the apical plasma membrane of DCT cells, but also with subapical intracellular vesicles and the Golgi apparatus (Fig. 2G, insets). This is in contrast to pT53- and pT58-NCC, which have previously been shown only to be associated with the apical plasma membrane (Pedersen et al. 2010).

Physiological regulation of pS124-NCC

Previous studies have demonstrated that NCC phosphorylation can be regulated in vivo by, for example, AVP (Mutig et al. 2010; Pedersen et al. 2010), ANGII (van der Lubbe et al. 2011) and aldosterone (Chiga et al. 2008). Thus, we examined the regulation of pS124-NCC under two different conditions: increased AVP or reduced dietary salt intake. To assess the effects of AVP on pS124-NCC abundance, three different studies were performed. Initially, AVP-deficient Brattleboro rats were treated with the V₂ receptor-selective AVP analogue dDAVP, for either 15 or 60 min. Immunoblots of whole kidney homogenates showed an ∼50% increase in pS124-NCC abundance (Fig. 3A), but no significant effect on total NCC, following acute dDAVP administration in vivo. Maximum effects were apparent after 15 min, and no further increase in pS124-NCC was seen after 60 min administration. These results were confirmed by immunohistochemistry on kidney sections from corresponding Brattleboro rats (Fig. 3B). After vehicle administration, only modest pS124-NCC labelling was seen, which clearly increased following either 15 or 60 min of dDAVP treatment. At the resolution of light microscopy, the large increase in pS124-NCC occurred only at the apical plasma membrane. Similar results were obtained from a separate study in Munich–Wistar rats. Following 30 min of dDAVP treatment, there was a significant increase in abundance of pS124-NCC, whereas total NCC levels did not change (Fig. 4). In contrast, following 5 days of water restriction and prolonged increases in circulating AVP, no significant increases in pS124-NCC levels were observed (not shown).

Figure 3. Phosphorylation of NCC at S124 increases with acute dDAVP in Brattleboro rats.

A, left, immunoblotting with anti-NCC and anti-pS124-NCC antibodies on whole kidney homogenates from Brattleboro rats treated with dDAVP(+) or saline(−) for 15 or 60 min. Representative blots for pS124-NCC and total NCC after 15 min are shown. Right, collated data of pS124-NCC band densities (AU, arbitrary units) relative to control (mean ± SEM), n= 4. *P < 0.05 from control as assessed by Student's unpaired t test. B, immunohistochemistry of pS124-NCC in corresponding kidney sections. Increased pS124-NCC labelling was observed after both 15 and 60 min dDAVP administration. PT, proximal tubule.

Figure 4. Phosphorylation of NCC at S124 increases with acute dDAVP in Munich–Wistar rats.

A, immunoblotting with anti-NCC and anti-pS124-NCC antibodies on whole kidney homogenates from Munich–Wistar rats treated with dDAVP or saline for 30 min. Representative blots for pS124-NCC and total NCC are shown. B, collated data. Data are band densities relative to control (mean ± SEM), n= 4. *P < 0.05 from control as assessed by Student's unpaired t test. AU, arbitrary units.

Currently, there are contrasting data on the acute effects of AVP on NCC subcellular distribution. One study, performed in Brattleboro rats, concluded that AVP exposure in vivo does not result in increased total NCC on the apical plasma membrane of DCT cells (Pedersen et al. 2010). A contrasting result from Brattleboro rats indicated that AVP significantly increases the apical membrane abundance of NCC (Mutig et al. 2010). In an attempt to resolve this controversy, we quantitatively assessed whether acute AVP treatment altered NCC apical plasma membrane abundance in two different models, Munich–Wistar rats and Brattleboro rats. Quantitative immungold electron microscopy (Fig. 5A) demonstrated that, compared to controls, 30 min dDAVP treatment of Munich–Wistar rats had no significant effect on total NCC plasma membrane abundance, with the percentage of gold particles representing NCC in the membrane remaining relatively constant at 60% of total cellular abundance (Fig. 5B). Additionally, quantitative confocal laser scanning microscopy indicated that neither 15 nor 60 min of dDAVP treatment of Brattleboro rats had any significant effect on the plasma membrane abundance of total NCC (Fig. 5C). Together, these results indicate that acute AVP exposure does not cause trafficking of NCC.

Figure 5. Short-term vasopressin treatment does not increase NCC abundance in the apical plasma membrane of DCT cells.

A, immunogold labelling of total NCC in Munich–Wistar rat kidney sections. Rats were treated for 30 min with saline (upper panel) or dDAVP (lower panel). Under control conditions or after dDAVP treatment, gold particles representing NCC were observed in the apical plasma membrane (red dots) and also in intracellular vesicles (blue dots). Insets show high magnification of apical plasma membrane (APM). Scale bar = 2.5 μm. B, no significant differences were observed in the number of gold particles associated with the apical plasma membrane following dDAVP treatment. C, quantitative confocal laser scanning microscopy indicated that neither 15 nor 60 min dDAVP treatment of Brattleboro rats had any significant effect on the plasma membrane abundance of total NCC.

The renin–angiotensin–aldosterone system (RAAS) is a potent regulator of NCC (Kim et al. 1998; Sandberg et al. 2007; Chiga et al. 2008; Rozansky et al. 2009; San-Cristobal et al. 2009; Vallon et al. 2009). To examine its role in regulation of pS124-NCC, we placed rats on either a low, normal or high Na⁺ diet for 5 days to activate the RAAS, which was confirmed by significant differences in plasma aldosterone levels between the groups (Table 1). Other physiological parameters obtained during the study are summarized in Table 1. Under these conditions, there were small but significant increases in the abundances of pS124-NCC and total NCC (Fig. 6A and B), which were confirmed qualitatively using immunohistochemistry (Fig. 6C). Increases in abundance were also observed for other previously characterized NCC phosphorylation sites at T53 and T58 (Pacheco-Alvarez et al. 2006; Richardson et al. 2008), but total levels of OSR-1/SPAK, kinases that regulate NCC phosphorylation at the upstream T53, T58 and S71 sites (Moriguchi et al. 2005; Richardson et al. 2008), were not changed. Examination of pOSR-1/pSPAK abundances (active forms of the kinases) proved inconclusive due to technical difficulties with antibodies (not shown). Similar observations concerning pS124-NCC levels were obtained in a separate study where rats were fed a low Na⁺ diet for 3 days (Supplementary Fig. 2).

Table 1.

Physiological parameters

Parameter	Low sodium diet	Normal sodium diet	High sodium diet
Start weight of rat (g)	176 ± 3.9	179 ± 4.1	184 ± 5.5
End weight of rat (g)	178 ± 2.1	182 ± 3.2	184 ± 7.7
Food intake (mg food (g body weight)−1 day−1)	122 ± 2.0	122 ± 1.6	123 ± 2.8
Water intake (μl (g body weight)−1 day−1)	65.1 ± 10†	127.1 ± 21	416 ± 43‡
Serum values
Aldosterone (pmol l−1)	760 ± 59†	435 ± 124	118 ± 55
Sodium (mmol l−1)	139 ± 0.4	141 ± 1.7	139 ± 1.5
Potassium (mmol l−1)	4.74 ± 0.2	4.21 ± 0.2	4.6 ± 0.2
Chloride (mmol l−1)	103 ± 0.6	105 ± 1.8	100 ± 1.0
Creatinine (μmol l−1)	23.7 ± 1.0	25.2 ± 0.6	24 ± 0.7
Urine values
Urine volume (μl (g body weight)−1 day−1)	41.5 ± 8.4†	88.9 ± 16.4	341.7 ± 14.0‡
Sodium (mmol l−1)	14 ± 1†	82 ± 15*	301 ± 15‡
Sodium excretion (μmol day−1)	103 ± 13†	1142 ± 64	18,215 ± 904‡
Potassium (mmol l−1)	220 ± 27†	120 ± 21*	28 ± 2.8‡
Potassium excretion (μmol day−1)	1473 ± 89	1694 ± 83	1848 ± 133
Chloride (mmol l−1)	17 ± 2†	91 ± 17*	309 ± 16‡
Chloride excretion (μmol day−1)	137 ± 45†	1270 ± 65	18,683 ± 996‡
Creatinine (μmol l−1)	6774 ± 779†	3121 ±553*	694 ± 22‡
Creatinine excretion (μmol day−1)	41.2 ± 1.5	44.0 ± 2.4	43.4 ± 1.1
Osmolality (mosmol kg−1)	1214 ± 147†	653 ± 118*	512 ± 24
Osmolar excretion (mosmol day−1)	8.2 ± 0.7†	9.1 ± 0.4	30.5 ± 1‡

Results are given as mean ± SEM. †P < 0.05, low salt versus high salt. *P < 0.05, low salt versus normal salt. ‡P < 0.05, low salt versus high salt. Data were analysed by one-way ANOVA followed by Bonferroni's multiple comparisons test.

Figure 6. Phosphorylation of NCC at S124 increases following dietary sodium restriction of Munich–Wistar rats.

A, representative immunoblotting with anti-NCC, anti-pS124-NCC, anti-pT58-NCC, anti-pT53-NCC and anti-OSR1/SPAK antibodies on whole kidney homogenates from Munich–Wistar rats fed a low sodium diet (0.013% Na⁺), a normal control diet (0.24% Na⁺) or a high sodium diet (3.37% Na⁺) for 5 days. B, collated data. Data are band densities relative to normal control diet (mean ± SEM), n= 5. *P < 0.05 compared to normal Na⁺ control, #P < 0.05 compared to low Na⁺ diet as assessed by one-way ANOVA followed by Bonferroni's multiple comparisons test. AU, arbitrary units. C, qualitative differences in total NCC and pS124-NCC labelling intensities were observed using immunohistochemistry.

Role of SPAK and OSR1 in S124 phosphorylation

The sequences surrounding S71 [EHYANS*ALPGEP] (conserved amino acids underlined, * is phosphoserine) and S124 [ETGANS*EKSPGEP] of NCC are similar, suggesting that the same kinase may phosphorylate both serine residues. As the kinases SPAK and OSR1 can directly phosphorylate NCC and target S71-NCC (Moriguchi et al. 2005; Yang et al. 2007), we examined the role of SPAK/OSR1 in S124-NCC phosphorylation. Non-radioactive kinase assays with constitutively active and inactive SPAK and OSR1 mutants, in the presence of MO25α to enhance kinase activity (Filippi et al. 2011), were performed. A recombinant protein corresponding to the first 100 amino acids of NCC (NCC[1–100]) was readily phosphorylated by active SPAK and OSR1 at the T58 site (Fig. 7), confirming previous studies (Richardson et al. 2008). In contrast, neither SPAK nor OSR1 significantly phosphorylated a synthetic 10 amino acid S124-NCC peptide (TGANSEKSPG) (Fig. 7), or a synthetic 21 amino acid S124-NCC peptide (Biotin-GLVEDETGANSEKSPGEPVRF) using this assay. As peptides are much poorer substrates for kinases than proteins, we performed further SPAK and OSR1 protein kinase assays using the CATCHtide substrate as a control (to confirm that SPAK/OSR1 can phosphorylate a peptide substrate in our assay) and an ADP-Glo assay that measures the generation of ADP by the protein kinase reaction leading to an increase in luminescence signal. OSR1 increased luminescence 250 and 350%, and SPAK 320 and 450%, above background using 200 and 500 μm CATCHtide peptide, respectively. No increases in luminescence above background were observed using similar assays and a synthetic 15 amino acid S124-NCC peptide (ETGANSEKSPGEPVR) substrate. Together, these results suggest that neither SPAK nor OSR1 is responsible for S124 phosphorylation and that other protein kinases are involved.

Figure 7. SPAK and OSR1 do not phosphorylate NCC at S124 in vitro.

Kinase assays were performed using S124-NCC peptide and NCC[1–100] mixed with MO25α and constitutively active or inactive SPAK or OSR1 kinases (inactive: SPAK D212A and OSR1 D164A, active: SPAK T233E and OSR1 T185E) for 60 min at 30°C. The active SPAK and OSR1 kinases were able to phosphorylate the NCC[1–100] protein but not the pS124-NCC peptide. pT53/58-NCC and pS124-NCC peptides were used as controls for antibodies.

Identification of kinases involved in phosphorylation of S124

Initial bioinformatics analysis did not identify any protein kinases that were strong candidates to phosphorylate S124-NCC. Thus, we used protein kinase microarrays to identify protein kinases that bind to a synthetic non-phosphorylated peptide corresponding to the region surrounding the S124 residue. Over 200 different protein kinases were assayed in triplicate (Supplementary Fig. 3). The serine–threonine kinases that the peptide bound to and gave the greatest signal strengths or greatest signal based on percentage change from control were ALK1, TRKC, AKT2, CAMK2b, CDK6, HIPK3, PLK1, ROCK1, RAF1 and PKCz. For several of the top kinases that exhibited binding to the synthetic peptide, activity profiling was undertaken using the non-phosphorylated S124 peptide to identify whether the peptide was a good substrate for these protein kinases (Table 2). Relative to their activity towards known control substrates (peptides or proteins), the majority of protein kinases showed no significant activity towards the S124 peptide. Of the 24 protein kinases analysed, only four (IRAK2, CDK6/Cyclin D1, NLK and mTOR/FRAP) showed weak but significant activity against the S124-NCC peptide.

Table 2.

[³³P]ATP kinase assays demonstrate weak activity of selected kinases on a synthetic peptide corresponding to the region surrounding S124 of NCC

Protein kinase	Control substrate	% Activity against ETGANSEKSPGEPVR
ETGANSEKSPGEPVR at 10 μm
ERK5	Myelin basic protein	1
NLK	Myelin basic protein	9*
p38δ	Myelin basic protein	2
ERK2	Myelin basic protein	3
mTOR/FRAP	4EBP1	9*
p38γ	Myelin basic protein	2
ETGANSEKSPGEPVR at 250 μm
AKT2	AKT (SGK) synthetic peptide	7*
AKT3	AKT (SGK) synthetic peptide	4
AMPK(A1/B1/G1)	SAMStide synthetic peptide	7*
CDK4/Cyclin D3	RB protein	4
CDK5/p38	Histone H1	1
CDK6/Cyclin D1	RB protein	10*
CDK6/Cyclin D3	RB protein	6*
GSK3α	GSK3	1
HIPK3	Myelin basic protein	4
IRAK2	Modified AKT peptide	13*
P70s6k	S6K synthetic peptide	3
PDK1	PDKtide synthetic peptide	6*
PKCζ	CREBtide synthetic peptide	2
RAF1	Unactive MEK1 & ERK1	2
ROCK1	S6K synthetic peptide	3
RSK1	S6K synthetic peptide	1
RSK3	RSK synthetic peptide	1
ZAP70	Poly (Glu/Tyr, 4:1) synthetic peptide	2

The results are presented as the percentage activity change of the kinase against the synthetic S124-NCC peptide, compared to control substrate. *Significant activity (% > 5).

Role of S124 site in NCC activity

³⁶Cl⁻ uptake studies were performed following expression of various NCC mutants in X. laevis oocytes. An initial ³⁶Cl⁻ uptake time course study on wt-NCC-expressing oocytes showed that uptake was relatively linear across the time points examined (not shown), and all subsequent studies were performed using a 20 min uptake. The presence of bumetanide, which can inhibit endogenously expressed Na⁺–K⁺–Cl⁻ cotransporter NKCC1 (Shetlar et al. 1990; Gamba et al. 1994), had no significant effect on NCC-induced ³⁶Cl⁻ uptake (Supplementary Fig. 1), whereas ³⁶Cl⁻ uptake was almost completely inhibited by the thiazide-like diuretic metolazone. Together, these results indicate that the ³⁶Cl⁻ uptake observed under our experimental conditions is mediated exclusively by NCC.

To elucidate any role of S124 in NCC activity, ³⁶Cl⁻ uptakes of various NCC mutants with respect to their plasma membrane abundance were examined. Oocytes expressing an S124A-NCC mutant, which cannot be phosphorylated at S124, had significantly lower uptake compared to wt-NCC (Fig. 8A). A T48A-T53A-T58A-NCC mutation almost completely abolished ³⁶Cl⁻ uptake to the levels of uninjected oocytes (Fig. 8A). Normalization of uptakes relative to NCC membrane abundance (Fig. 8C) indicated that the reduced uptake was not due to decreased NCC protein at the membrane. Further studies on S124D-NCC and T48D-T53D-T58D-NCC mutants gave similar results (Fig. 8D–F). The introduction of so-called ‘phospho-mimicking’ S-to-D mutations at these sites reduced ³⁶Cl⁻ uptakes to a similar extent as S-to-A mutations. Although the S-to-D mutations did not mimic NCC phosphorylation (expected ³⁶Cl⁻ uptakes would be equal to or greater than wt-NCC), these results do highlight the importance of an intact S124 for full NCC activity. Furthermore, although not an ideal system to study protein trafficking, comparison of the NCC plasma membrane abundance with total membrane abundance (Supplementary Fig. 4) supported the conclusion that the decreased uptake activity in the various NCC mutants was not due to reduced trafficking of NCC, but rather a direct effect on NCC transport activity.

Figure 8. Role of S124 in NCC function.

A, representative ³⁶Cl⁻ uptakes from X. laevis oocytes microinjected with 50 ng wt, S124A or T48A-T53A-T58A mutant NCC cRNA. Data are from a single experiment and presented as means ± SEM. *P < 0.05 from NCC, as assessed by one-way ANOVA followed by Bonferroni's multiple comparisons test. B, immunoblotting of NCC on plasma membranes prepared from oocytes represented in A. C, normalized data of ³⁶Cl⁻ uptakes relative to plasma membrane NCC abundance. Data are collated from four individual experiments. D, representative ³⁶Cl⁻ uptakes from X. laevis oocytes microinjected with 50 ng wt, S124A, S124D or T48D-T53D-T58D mutant NCC cRNA. Data are from a single experiment and presented as means ± SEM. *P < 0.05 from NCC as assessed by one-way ANOVA, followed by Bonferroni's multiple comparisons test. E, immunoblotting of NCC on plasma membranes prepared from oocytes represented in D. F, normalized data of ³⁶Cl⁻ uptakes relative to plasma membrane NCC abundance. Data are collated from four individual experiments.

Characterization of the tetracycline-inducible cell line MDCKII-FRT-TO-rNCC

A suitable polarized kidney cell line model to study the role of NCC mutations in NCC function has not previously been established. We attempted to overcome this problem by generating an MDCK type II cell line with tetracycline-inducible NCC, S124A-NCC or S124D-NCC expression, allowing us to assess the role of S124 in NCC trafficking. Initial characterization studies of the novel MDCKII-FRT-TO-rNCC cell line determined that 10 μg ml⁻¹ tetracycline and a stimulation period of 10–22 h were optimal conditions for adequate NCC induction (Fig. 9A and B). Following an initial induction period of 16 h, NCC protein abundance returned to background levels after 24 h (Fig. 9D), and hence prolonged experiments require daily induction of tetracycline. Repeated daily stimulation with tetracycline caused a decrease in NCC protein abundance after 5 days (Fig. 9C). Localization of NCC in both the apical and the basolateral plasma membranes was demonstrated by side-specific cell surface biotinylation of confluent MDCK-FRT-TO-rNCC cell monolayers. However, NCC abundance in the basolateral membrane was 5-fold greater than in the apical membrane (Fig. 9E). In MDCK-FRT-TO-rNCC-S124A/D cell lines, surface biotinylation showed no significant difference between abundances of NCC or S124 mutants in the plasma membrane (Fig. 9F and H), suggesting that S124 is not required for NCC trafficking. Furthermore, the levels of pT58-NCC were not different between the cell lines (Fig. 9G and I), suggesting that the reduced activity of S124 mutants observed in oocytes is probably not due to alterations in phosphorylation levels at this site, which is essential for NCC function (Pacheco-Alvarez et al. 2006; Richardson et al. 2008).

Figure 9. Phosphorylation at S124 does not affect NCC trafficking in mammalian cells.

A, tetracycline (Tet) increases NCC abundance in a dose-dependent manner. Cells were induced for 16 h. B, time course of NCC induction with 10 μg ml⁻¹ tetracycline. C, repeated daily stimulation with tetracycline decreased NCC after 5 days. D, following an initial induction period of 16 h using 10 μg ml⁻¹ tetracycline, NCC protein was followed over time and returned to background levels after 24 h, at which point re-induction increased NCC levels. E, side-specific cell surface biotinylation of confluent MDCK-FRT-TO-rNCC cell monolayers demonstrated NCC in both the apical and the basolateral (BLT) plasma membranes. NCC abundance was greater in the basolateral membrane than in the apical membrane. F, surface biotinylation of NCC-, S124A-NCC- and S124D-NCC-expressing cell lines. G, quantitative analysis of four independent biotinylation experiments showed no significant effect of mutating the S124 site on NCC membrane abundance. H, assessment of pT58-NCC levels in the NCC-, S124A-NCC- and S124D-NCC-expressing cell lines. I, quantitative analysis of three independent experiments demonstrated that levels of pT58-NCC were not significantly different between the cell lines.

Discussion

The post-translational modification phosphorylation plays a vital regulatory role in NCC function. Although several studies have examined the function and regulation of the conserved amino-terminal phosphorylation sites T53, T58 and S71 (corresponding to T55, T60 and S73 in human NCC), little is known about an additional conserved phosphorylation site, S124, recently identified via phosphoproteomic profiling (Feric et al. 2011). Thus, the aim of this study was to elucidate the function, distribution and physiological regulation of pS124-NCC and determine any potential role in DCT sodium reabsorption.

The novel phospho-specific antibodies targeting pS124-NCC that we developed were useful and specific in a variety of applications, including immunoblotting, immunohistochemistry and immunogold electron microscopy, allowing us to assess the distribution and regulation of pS124-NCC. In the DCT, pS124-NCC was distributed similarly to total NCC; both were associated with the apical plasma membrane, but also with intracellular vesicles and the Golgi apparatus in the apical domain of the cell. The distribution of total NCC confirms previous studies (Sandberg et al. 2006, 2007; Pedersen et al. 2010). The association of pS124-NCC with intracellular vesicles and Golgi apparatus is in contrast to our previous studies on two other phosphorylated forms of NCC, namely pT53- and pT58-NCC (Pedersen et al. 2010), which are associated only with the apical plasma membrane, suggesting a specific role in NCC function. Indeed, T53 and T58 are critical for full NCC transport activity but not plasma membrane targeting (Pacheco-Alvarez et al. 2006; Richardson et al. 2008; Pedersen et al. 2010). Similar results were obtained in our current studies, where although T48A-T53A-T58A-NCC mutants trafficked similarly to wildtype NCC to the oocyte plasma membrane, their transport activity was almost completely abolished. In contrast, ³⁶Cl⁻ uptakes in oocytes expressing an S124A-NCC mutant were reduced to a much smaller extent compared to the T48A-T53A-T58A-NCC mutants. The reduced uptake activity in the S124A mutants occurred without any significant differences in plasma membrane expression compared to wildtype NCC. These data, combined with the intracellular distribution of pS124-NCC, suggests that although S124 plays a role in full NCC activity, it not as critical as other phosphorylation sites for mediating NaCl transport.

Serine-to-aspartic acid (S-D) mutations of NCC at S124 or T48-T53-T58 had similar oocyte uptake activities as serine-to-alanine (S-A) mutations, suggesting that rather than mimicking phosphorylation, the S-D mutations acted like phospho-site deletions at these sites. One possible explanation is our use of aspartic acid (D) as the phospho-mimicking amino acid rather than glutamic acid (E). We have previously shown that S-D mutations are a good mimic of phosphorylated serine residues (Moeller et al. 2009b) and in an earlier oocyte study, a T58D-rNCC mutant had higher Na⁺ uptake than wt-NCC (Glover et al. 2009). In contrast, T-E mutations have also been demonstrated not to mimic phosphorylated threonines in NKCC1 (Gagnon et al. 2007), so data generated using ‘phospho-mimicking’ substitutions should be interpreted with caution. An alternative explanation is the role of T48. To our knowledge, no one has studied the effect of phosphorylation at T48 on NCC activity and previous studies on corresponding residues in shark and mouse NKCC1 (T179 or T201, respectively) are not conclusive (Vitari et al. 2006; Gagnon et al. 2007). It is plausible that the significant decrease in Cl⁻ uptake in the ‘phospho-mimicking’ triple mutant (T48D-T53D-T58D-rNCC) could be explained by an inhibitory effect of phosphorylation at T48 that is not observed in the triple-A mutant (T48A-T53A-T58A-rNCC). Studies of single T48A- and T48D-NCC mutants would be required to resolve this possibility.

Various reports of immortalized DCT cell lines have demonstrated thiazide-sensitive NaCl fluxes, suggesting native expression of NCC (Gesek & Friedman, 1992; Peng et al. 1999; Ko et al. 2012). However, a polarized kidney cell model with stable NCC expression that will allow examination of NCC mutations has previously not been established. To examine the role of S124 in NCC trafficking, we generated various wildtype or mutant NCC cell lines that were based on a tetracycline-inducible MDCKII cell model containing FRT sites (Nejsum et al. 2011). These cells incorporated a single gene copy of NCC or mutant NCC into the same insertion site in the MDCK genome, allowing more direct comparisons of NCC function. These cells displayed tetracycline-inducible NCC expression in a dose- and time-dependent manner. However, in these cells NCC was targeted to both the apical and the basolateral plasma membrane, suggesting that different NCC targeting mechanisms are present compared to native DCT cells where NCC is only apically localized. In these cells, there were no significant differences in NCC membrane targeting or pT58-NCC levels when comparing wildtype NCC with S124 A/D NCC mutants, suggesting that the pS124 site does not play a role in trafficking of NCC to the plasma membrane. However, the anomaly of basolateral and apical NCC expression has to be taken into consideration when drawing solid conclusions from these data.

Assessing the combined data from the oocyte uptake studies, cell studies and immunogold electron microscopy together, it is difficult to draw solid conclusions regarding the precise role of S124 for regulation of NCC function. Although S124 mutants have equal abundances in the plasma membrane, indicating no significant role for S124 in NCC trafficking to the plasma membrane, the site is phosphorylated inside the Golgi apparatus. Thus, one could speculate that S124 plays some role in the transport of NCC through the Golgi apparatus to trafficking vesicles, similar to what has been proposed for the S256 phosphorylation site in the water channel aquaporin-2 (Procino et al. 2003). Additionally, the small effect of S124 mutation on NCC transport activity, rather than the large effects observed with mutation of other phosphorylation sites, suggests that phosphorylation of S124 may be able to modulate sodium transport in small steps, allowing ‘fine tuning’ of sodium balance in the DCT without an absolute ‘on–off’ switch.

The protein kinases SPAK and OSR1 phosphorylate NKCC1, NKCC2 and NCC at residues conserved among all three cotransporters (T53 and T58). Furthermore, the SPAK/OSR1-docking motif RFX[I/V] present in all three cotransporters suggests a similar mechanism of regulation and activation of the sodium-dependent branch of the SLC12 family (Moriguchi et al. 2005; Vitari et al. 2005, 2006; Gagnon et al. 2007; Yang et al. 2007, 2010; Richardson et al. 2008). The amino acid sequences surrounding S71 and S124 are similar, suggesting that the same protein kinase may phosphorylate both sites (Feric et al. 2011). As S71 in NCC has been demonstrated to be a phosphoacceptor site for SPAK/OSR1, it was plausible that these kinases also target S124. However, using multiple approaches we were not able to demonstrate SPAK/OSR1-dependent phosphorylation at S124. This was surprising, as the S124 site is within both the RFTI sequence (amino acids 19 and 22) and the RFGW sequence (amino acids 135–138), which are potential binding sites for the conserved C-terminal binding domain of SPAK and OSR1 (Richardson et al. 2008). Alternatively, based on their involvement in mediating WNK4 effects on NCC (Zhou et al. 2012), the extracellular signal-regulated kinase (ERK) 1/2 is also a possible candidate for targeting S124-NCC. However, although both ERK1 and ERK2 scored above the cut-off criteria for significant kinase binding to the NCC peptide in our kinase arrays, the S124 site does not really conform with the known specificities of these kinases, and the synthetic peptide used for the arrays contains the neighbouring S127 site that is, based on bioinformatics, a potential substrate for ERK. Furthermore, ERK2 had almost zero activity against S124 (Table 2), arguing against a direct role for ERK1/2 on S124 phosphorylation. Protein kinase profiling demonstrated that only the protein kinases IRAK2, CDK6/Cyclin D1, NLK and mTOR/FRAP had weak activity towards the S124 site. However, we are unable to rule out that the low activity of the majority of protein kinases towards the S124 site did not result from using peptide substrates rather than recombinant proteins. Peptides are poor substrates for kinases as the surrounding amino acids are not well constrained in the short sequences compared with the full-length protein and the increased entropy of peptides makes them less likely to be presented in an optimal three-dimensional structure for the kinase active site.

To maintain tight control over Na⁺ reabsorption in the DCT, NCC function is modulated by various physiological stimuli (Mutig et al. 2010; Pedersen et al. 2010; van der Lubbe et al. 2011). In our current studies, we assessed the role of AVP or the RAAS to regulate pS124-NCC in vivo. In line with previous studies on pT53- and pT58-NCC (Chiga et al. 2008; Vallon et al. 2009; Mutig et al. 2010; Pedersen et al. 2010), acute dDAVP administration increased the abundance of pS124-NCC. These results further emphasize that, at least in rat, the DCT is an AVP responsive segment and that AVP is a potent regulator of NCC activity and Na⁺ transport in the DCT. In contrast, prolonged high AVP levels resulting from water restriction did not increase pS124-NCC levels, suggesting that the site is more important for short-term regulation of Na⁺ transport. Recently, there has been debate over whether short-term AVP exposure results in increased abundance of NCC in the plasma membrane of DCT cells (Mutig et al. 2010; Pedersen et al. 2010). The current studies, utilizing two different animal models and two different forms of quantification, indicate that acute exposure to AVP does not increase the apical abundance of total NCC, despite the large increases in NCC phosphorylation. Thus, it appears that unlike other physiological stimuli such as ANGII (Sandberg et al. 2007), AVP specifically exerts its effects via alterations in the activity of NCC already present in the DCT plasma membrane.

Mice fed a low salt diet have increased phosphorylation of NCC at T53, T58 and S71, alongside increased pSPAK/OSR1 (Chiga et al. 2008). Furthermore, increased ANGII or aldosterone levels independently increased pT53- and pT58-NCC and SPAK/OSR1 phosphorylation in rats (van der Lubbe et al. 2011, 2012). In line with this, following dietary salt restriction of Munich–Wistar rats and increases in plasma aldosterone levels, pS124-NCC levels were increased, alongside pT53-NCC, pT-58-NCC and total NCC levels, suggesting that the phosphorylation of NCC at S124 follows a similar aldosterone-regulated pathway as other pNCC sites. However, no significant changes were observed in total OSR-1/SPAK levels in our rats and examination of pOSR-1/pSPAK abundances proved inconclusive due to technical difficulties with antibodies. Furthermore, although the WNK-OSR1/SPAK cascade is a potent regulator of NCC (Moriguchi et al. 2005; Vitari et al. 2005; Yang et al. 2007; Chiga et al. 2008), our kinase assays failed to demonstrate phosphorylation of S124-NCC by SPAK/OSR1. One could speculate that another kinase is involved in the WNK-OSR1/SPAK cascade for phosphorylation of S124-NCC, or that S124-NCC phosphorylation is mediated directly or indirectly by alterative aldosterone-regulated pathways (Rozansky et al. 2009; Vallon et al. 2009). In support of this last possibility, alterations in dietary salt have been demonstrated to affect NCC protein abundance in an aldosterone-dependent mechanism via the WNK4-ERK1/2-mediated pathway (Lai et al. 2012).

In summary, in addition to previously examined sites, phosphorylation of NCC at S124 has an important role in regulating NCC function, probably via directly mediating NCC activity and not by modulating NCC plasma membrane targeting. Our studies suggest that various physiological stimuli can alter sodium transport in the DCT via regulated multi-site phosphorylation of NCC, and that alternative pathways to the WNK-SPAK/OSR1 pathway are likely to be involved.

Glossary

AVP

arginine vasopressin

CNT

connecting tubule

DCT

distal convoluted tubule

ERK

extracellular signal-regulated kinase

FRT

Flp Recombinase Target

NCC

sodium–chloride cotransporter

OSR1

oxidative stress-responsive kinase-1

RAAS

renin–angiotensin–aldosterone system

SPAK

STE20/SPS1-related proline–alanine-rich kinase

Author contributions

L.L.R., N.M. and R.A.F. designed studies; L.L.R., M.A., N.B.P., N.M. and R.A.F. performed studies; L.L.R., N.M. and R.A.F. analyzed data; L.L.R. and R.F. wrote the manuscript. All authors approved the final version of the manuscript.

Supplementary material

Supplementary Fig. 1

Supplementary Fig. 2

Supplementary Fig. 3

Supplementary Fig. 4