Phosphorylation of UT-A1 on serine 486 correlates with membrane accumulation and urea transport activity in both rat IMCDs and cultured cells

Abstract

Vasopressin is the primary hormone regulating urine-concentrating ability. Vasopressin phosphorylates the UT-A1 urea transporter in rat inner medullary collecting ducts (IMCDs). To assess the effect of UT-A1 phosphorylation at S486, we developed a phospho-specific antibody to S486-UT-A1 using an 11 amino acid peptide antigen starting from amino acid 482 that bracketed S486 in roughly the center of the sequence. We also developed two stably transfected mIMCD3 cell lines: one expressing wild-type UT-A1 and one expressing a mutated form of UT-A1, S486A/S499A, that is unresponsive to protein kinase A. Forskolin stimulates urea flux in the wild-type UT-A1-mIMCD3 cells but not in the S486A/S499A-UT-A1-mIMCD3 cells. The phospho-S486-UT-A1 antibody identified UT-A1 protein in the wild-type UT-A1-mIMCD3 cells but not in the S486A/S499A-UT-A1-mIMCD3 cells. In rat IMCDs, forskolin increased the abundance of phospho-S486-UT-A1 (measured using the phospho-S486 antibody) and of total UT-A1 phosphorylation (measured by ³²P incorporation). Forskolin also increased the plasma membrane accumulation of phospho-S486-UT-A1 in rat IMCD suspensions, as measured by biotinylation. In rats treated with vasopressin in vivo, the majority of the phospho-S486-UT-A1 appears in the apical plasma membrane. In summary, we developed stably transfected mIMCD3 cell lines expressing UT-A1 and an S486-UT-A1 phospho-specific antibody. We confirmed that vasopressin increases UT-A1 accumulation in the apical plasma membrane and showed that vasopressin phosphorylates UT-A1 at S486 in rat IMCDs and that the S486-phospho-UT-A1 form is primarily detected in the apical plasma membrane.

urea plays a central role in the generation of a concentrated urine (4, 8). Several studies show that maximal urine-concentrating ability is decreased in protein-deprived animals and humans and is restored by urea infusion (reviewed in Ref. 21). In addition, UT-A1/UT-A3 (3, 5), UT-A2 (27), and UT-B knockout mice (16, 29, 30) were each shown to have urine-concentrating defects. Thus, any hypothesis regarding the mechanism by which the kidney concentrates urine needs to include some effect derived from urea and urea transporters.

Vasopressin [AVP; also known as antidiuretic hormone (ADH)] is the primary hormone regulating urine-concentrating ability (reviewed in Refs. 21, 22). Vasopressin, when added to the bath of a perfused rat terminal inner medullary collecting duct (IMCD), results in binding to V₂ receptors, stimulating adenylyl cyclase, increasing cAMP production, and increasing facilitated urea transport (18, 23, 24, 26). Vasopressin regulates urea transport through two cAMP-dependent signaling pathways: protein kinase A (PKA) and exchange protein activated by cAMP (28). The mechanism involves vasopressin-stimulated increases in both the phosphorylation and apical plasma membrane accumulation of the UT-A1 urea transporter (1, 2, 10, 14, 28, 31).

Vasopressin phosphorylates UT-A1 at serines 486 and 499 (2). Mutation of both serine residues eliminates vasopressin stimulation of UT-A1 apical plasma membrane accumulation and urea transport in heterologous expression systems (2). Serine 486 was identified as a vasopressin-sensitive phosphorylation site in rat IMCDs using a proteomics approach (9).

Vasopressin increases UT-A1 phosphorylation in fresh suspensions of rat IMCDs (31). However, this finding was obtained by metabolically labeling IMCDs with ³²P and thus could not distinguish between effects at serines 486 and 499 (2). To begin to assess the effect of phosphorylation at S486 in rat inner medulla (IM), we developed a phospho-specific antibody to S486-UT-A1.

METHODS

Animals.

All animal protocols were approved by the Emory University Institutional Animal Care and Use Committee. Male Sprague-Dawley rats (Charles River Laboratories, Wilmington, MA), weighing 100–150 g, received free access to water and standard rat chow (Test diet 5001, Purina) containing 23% protein.

Sample preparation.

Kidneys were removed and the IM was collected. For direct Western analysis of cell lysate, IMs were homogenized in ice-cold isolation buffer (10 mM triethanolamine, 250 mM sucrose, pH 7.6, 1 μg/ml leupeptin, and 2 mg/ml PMSF), and then SDS was added to a final concentration of 1% (11–13). To prepare IMs for metabolic radiolabeling, IMs were cut into small pieces ∼1 × 1 × 1 mm and preincubated in phosphate-free DMEM. To prepare tissue for immunoprecipitation, IMs from each rat were homogenized in 750 μl of isolation buffer, and then 750 μl of RIPA buffer were added and the sample was briefly dounced further. The homogenate was sheared with a 28-gauge insulin syringe and centrifuged to remove insoluble particulates. Some sample was reserved for preimmunoprecipitation controls.

Phosphorylation.

Metabolic labeling with ³²P-orthophosphate was performed as previously published (31). Briefly, IM pieces were incubated in phosphate-free DMEM containing 0.2 mCi/ml of ³²P-orthophosphate. After the 3-h labeling period, IMCDs were incubated for a further 30 min with 100 nM vasopressin or 10 μM forskolin. IMCD cell lysates were prepared and UT-A was immunoprecipitated as described below. Precipitated proteins were separated by SDS-PAGE and radiolabeled UT-A1 was determined by autoradiography of the dried gel.

Primary antibodies.

Two primary antibodies are used in these studies: 1) total UT-A1 was assessed with a polyclonal antibody to the COOH terminal of UT-A1 (15) and 2) a new phospho-specific antibody to UT-A1 phosphorylated at the serine 486 that was prepared against the phosphopeptide PRRKS(P)VFHIEW[C]-NH₂ position (PhosphoSolutions, Aurora, CO). Both primary antibodies were peptide affinity-purified before use.

Immunoprecipitation.

IM samples were incubated at 4°C overnight with 10 μl/sample of either the COOH-terminal UT-A1 antibody or the phospho-S486-UT-A1 antibody. Protein A sepharose beads (Pierce) were added for 2 h and then washed 7× with RIPA (1 ml/wash). Proteins were removed from the beads and solubilized by boiling in Laemmli buffer for 1 min, and then analyzed by Western blot or autoradiography as appropriate.

Western blot analysis.

Proteins (20 μg/lane) were size-separated by SDS-PAGE on 10% gels and then electroblotted to polyvinylidene difluoride membranes (Immobilon, Millipore, Bedford, MA). After being blocked with 5% nonfat dry milk for 1 h, blots were incubated with primary antibody overnight at 4°C. Blots were washed three times in tris-buffered saline with 0.5% Tween 20 and then incubated for 2 h with Alexa Fluor 680-linked anti-rabbit IgG (Molecular Probes, Eugene, OR). Bound secondary antibody was visualized using infrared detection with the LICOR Odyssey protein analysis system (Lincoln, NE).

Site-directed mutagenesis.

Serines 486 and 499 were mutated from rat UT-A1 using the QuickChange site-directed mutagenesis kit (Agilent Technologies, Santa Clara, CA) according to the manufacturer's instructions (2). The serine (S) residue was substituted with an alanine (A) residue. The oligonucleotide 5′-CCCAGGCGTAAGGCCGTGTTCCATATCGAGTGG-3′ was used to mutate the first serine residue at S486. Oligonucleotide 5′-TCATCCATCCGGAGGAGGGGCAAAGTGTTTGGAAAAAG-3′ was used to mutate the second serine residue at S499. A double mutant (S486A/S499A) construct was also generated. The constructs were verified by nucleotide sequence analysis (Macrogen, Rockville, MD).

Generation of mIMCD3 cell lines expressing wild-type or S486A/S499A UT-A1.

mIMCD3 cells that stably expressed wild-type or S486A/S499A-mutated UT-A1 were generated in the same manner as we previously generated UT-A1-MDCK cells (6). First, mIMCD3 cells were transfected with pFRT/lacZeo using Lipofectamine 2000 (Invitrogen) and grown for 10 days in 100 μg/ml Zeocin to select for cells stably expressing the FRT recombination site, followed by growing up one of the selected cell clones. This mIMCD3-FRT cell line was then cotransfected with pOG44, a plasmid for transiently expressing the Flp recombinase, and with pcDNA5-FRT-UT-A1 or pcDNA5-FRT-UT-A1M. The latter plasmid contained the S486A/S499A-mutated UT-A1 gene and a FRT site for homologous recombination with the cell's FRT site, as well as a hygromycin resistance gene. Cells that underwent the recombination and incorporation of the UT-A1 gene and became resistant to hygromycin were then cultured in DMEM containing 400 μg/ml hygromycin for clonal selection.

Urea influx assays in mIMCD3 cells expressing UT-A1 or S486A/S499A UT-A1.

Transepithelial tracer urea fluxes at 37°C were measured as described previously (6, 7). The measurements consisted of adding tracer urea to the apical medium (containing 5 mM cold urea) in the Transwell insert to start the flux experiment. The insert was then moved in 3-min intervals from one well of a 12-well culture plate to the next, and the radioactivity accumulated during this interval in the basolateral medium was used to calculate the rate of urea flux. Dimethylthiourea (1.5 mM) was added to the basolateral medium at the end of the flux period to inhibit all urea transporter activity and test for the possibility of nonspecific leak formation.

Immunohistochemistry.

Kidneys from rats that were injected intraperitoneally with vasopressin were perfusion fixed with 4% paraformaldehyde and then embedded in paraffin sectioned into 4-μm slices. Tissue slices were stained overnight with primary antibodies diluted in PBS (1:1,000 anti-COOH-terminal UT-A1 and 1:200 anti-pser486-UT-A1) and incubated 2 h with Alexa fluor 488-linked goat anti-rabbit IgG (Invitrogen, Carlsbad, CA) for fluorescent confocal microscopy, as described previously (1, 2, 14). Slices for confocal micrographs were also treated with the nuclear stain DAPI. Confocal microscopy was performed with a Zeiss LSM 510 META ZEN confocal microscope and LSM ZEN imaging software.

Statistics.

Data are presented as means ± SE; comparisons were made with one-way ANOVA and Fisher's LSD posttest. P < 0.05 was considered significant.

RESULTS

Phospho-specific antibody to S486-UT-A1.

We developed a phospho-specific antibody to S486-UT-A1 using an 11 amino acid peptide antigen starting from amino acid 482 that bracketed the phosphorylated serine 486 in roughly the center of the sequence. A cysteine was added to the NH₂ end to allow carrier protein linkage. The affinity-purified antibody was tested for nonspecific associations by peptide preadsorption in a Western blot analysis (Fig. 1). Both the 97- and 117-kDa glycoprotein forms of UT-A1 are identified with the phospho-specific S486-UT-A1 antibody. Greater than 95% of the signal is obliterated by preadsorption of the antibody with 1 μg/ml of the immunizing peptide.

Fig. 1.

Phospho-specific antibody to phospho-S486-UT-A1 specifically recognizes the phosphorylated urea transporter in rat inner medullary collecting ducts (IMCDs). Western analysis of rat IMCDs immunoprecipitated with total UT-A1 antibody and probed with phospho-S486-UT-A1 antibody without (A) or with (B) preadsorption of the antibody with 1 μg/ml immunizing peptide. LICOR scans were performed at the same sensitivity level and gels were loaded identically. Both 97- and 117-kDa phosphorylated UT-A1 are designated by arrows. Shown are 3 samples representative of a total n = 6 for this determination.

mIMCD3 cell lines expressing wild-type or S486A/S499A UT-A1.

We developed two stably transfected mIMCD3 cell lines (20): one expressing wild-type UT-A1 and one expressing a mutated form of UT-A1, S486A/S499A, that is unresponsive to PKA (2). The wild-type UT-A1-mIMCD3 cells show a robust increase in urea influx in response to forskolin (Fig. 2), as one would expect from a cAMP-mediated process (25). In contrast, the S486A-S499A-UT-A1-mIMCD3 cells show no increase in urea influx in response to forskolin (Fig. 2).

Fig. 2.

Urea flux assay of mIMCD3 cells stably transfected with wild-type UT-A1 (A) or S486A/S499A UT-A1 (B). Cells were stimulated with 5 μM forskolin (FSK) and urea flux was determined each 3 min over a 40-min period. Dimethylthiourea (DMTU) was added at the end to inhibit urea transporters and determine basal (nontransporter mediated) flux rate. Data points are the averages of 3 determinations ± SE.

To verify that the loss of forskolin-stimulated urea flux was not due to the absence of a UT-A1 protein, Western analysis was performed on both mIMCD3 cell lines. Figure 3 shows that UT-A1 was successfully produced in both wild-type UT-A1-mIMCD3 cells (A) and in S486A/S499A-UT-A1-mIMCD3 cells (B).

Fig. 3.

Wild-type UT-A1 or S486A/S499A-UT-A1 is made in the stably transfected mIMCD3 cells. Western analysis of mIMCD3 cells that are stably transfected with wild-type UT-A1 (A, C) or S486A/S499A UT-A1 (B, D) probed with antibody to total UT-A1 (A, B) or with phospho-S486 UT-A1 antibody (C, D). Shown are 2 representative samples of a total of n = 6 determinations.

The loss of activity in the S486A/S499A-UT-A1-mIMCD3 cells could be due to mutation of S486, S499, or both. To test whether the loss of activity correlates with the loss of phosphorylation at S486, the Western blots were probed with the phospho-S486-UT-A1 antibody. The antibody identified the 97-kDa UT-A1 protein band in the wild-type UT-A1-mIMCD3 cells (Fig. 3C) but saw no UT-A1 protein in the S486A/S499A-UT-A1-mIMCD3 cells (Fig. 3D).

Phospho-S486-UT-A1 in rat IMCD.

Since forskolin increases total UT-A1 phosphorylation (31), we tested whether it would increase UT-A1 phosphorylation at S486 using our new phospho-S486-UT-A1 antibody. Forskolin increased the abundance of phospho-S486-UT-A1 (Fig. 4A). As a positive control, we measured total UT-A1 phosphorylation by ³²P incorporation and showed that forskolin increased UT-A1 phosphorylation (Fig. 4B). The increases are not due to an overall increase in UT-A1 protein since the whole cell lysate from rat IMCDs that were treated with forskolin and then analyzed by Western blot (Fig. 4C) shows no increase in total UT-A1 protein.

Fig. 4.

General and S486-UT-A1 phosphorylation are both increased by forskolin in rat IMCD. Rat IMCD was metabolically labeled with ³²P, and then treated for 30 min with 10 μM forskolin. Total UT-A1 was immunoprecipitated. Precipitated protein was separated by SDS-PAGE and analyzed by Western blot using the phospho-S486-UT-A1 antibody (A) and by autoradiography of the dried gel (B). Whole cell lysate from IMCDs without (control) or with forskolin treatment probed for total UT-A1 showed no significant increase with treatment (C).

To detect only the UT-A1 protein phosphorylated by forskolin at S486, we immunoprecipitated phospho-S486-UT-A1 from the whole cell lysate of rat IMCDs using the phospho-S486-UT-A1 antibody. There was a 192 ± 23% increase in the abundance of phospho-S486-UT-A1 precipitated from the forskolin-treated IMCDs relative to control IMCDs (Fig. 5).

Fig. 5.

P-ser486-UT-A1 antibody immunoprecipitates more protein from forskolin-treated rat IMCD tissue than from untreated control tissue. The p-ser486-UT-A1 Ab was used to immunoprecipitate phosphorylated UT-A1 from rat IMCD tissue that was treated with forskolin for 30 min. A: Western blot of immunoprecipitated samples probed with pser486-UT-A1 Ab. B: bar graph showing band densities as a percent of control. Data are means ± SE, *P < 0.05, n = 4.

Plasma membrane phospho-S486-UT-A1.

The forskolin-stimulated increase in UT-A1 phosphorylation has been implicated in UT-A1 accumulation in the plasma membrane (2). To determine specifically whether phospho-S486-UT-A1 is in the plasma membrane of rat IMCDs, we examined the biotinylated protein population by Western blot. We probed for total UT-A1 (Fig. 6A, top) or for phospho-S486-UT-A1 (Fig. 6A, bottom). Forskolin increased the plasma membrane accumulation of phospho-S486-UT-A1 in rat IMCDs (Fig. 6B).

Fig. 6.

Phosphorylated UT-A1 in the biotinylated protein population of IMCDs treated with (+) or without (−) forskolin was assessed. Phospho-S486-UT-A1 antibody was used to probe biotinylated proteins on a Western blot. A: triplicate determinations per condition. B: bars provide means ± SE of n = 9 samples per condition. *P < 0.01.

To complement the biotinylation studies, we used the phospho-S486-UT-A1 antibody to stain IMCD slices and compared UT-A1 localization in rats that were treated without (control) or with arginine vasopressin for 45 min before death (Fig. 7). UT-A1 is seen throughout the cell in IMCDs from control rats probed for total UT-A1 (Fig. 7A). While the biotinylation studies indicate that there is an increase in UT-A1 membrane accumulation on cAMP stimulation, it is not very apparent when looking at total UT-A1 (Fig. 7B). The phospho-S486-UT-A1 antibody detects much less cytosolic UT-A1 (Fig. 7C), suggesting that the phospho-S486 protein is mostly maintained in the plasma membrane. In IMCDs from the vasopressin-treated rats, virtually all of the phospho-S486-UT-A1 appears in the apical plasma membrane (Fig. 7D).

Fig. 7.

Confocal microscopy of phospho-S486-UT-A1 and total UT-A1 in IMCDs of control and vasopressin-treated rats. Paraffin sections stained with antibody to total UT-A1 (A, B) and p-486-UT-A1 (C, D) were also stained with DAPI to show nuclei. Micrographs were acquired at a magnification of ×63 and acquisition settings were optimized and then matched for each pair of samples that were stained with the same antibody. B and D: results from animals treated with vasopressin (+AVP).

DISCUSSION

We previously showed that vasopressin phosphorylates UT-A1 at S486 and S499 (2). Mutation of both serine residues eliminates vasopressin stimulation of UT-A1 apical plasma membrane accumulation in transiently transfected LLC-PK₁ cells and urea transport in Xenopus laevis oocytes (2). In the present study, we confirmed and extended these findings through the generation of a new phospho-specific antibody to S486-UT-A1 and two new mIMCD3 cell lines: wild-type UT-A1-mIMCD3 cells and a cAMP-unresponsive form, S486A/S499A-UT-A1-mIMCD3 cells.

We developed a phospho-specific antibody to S486-UT-A1 to begin to assess the effect of UT-A1 phosphorylation at S486. We verified the specificity of this antibody by showing that no bands were detected when the antibody was preadsorbed with the immunizing peptide or in the S486A/S499A-UT-A1-mIMCD3 cells. We previously showed that vasopressin increases P-32 incorporation into UT-A1 (1, 31). In the present study, we reconfirmed this finding, and showed that the abundance of phospho-S486-UT-A1, detected using our new phospho-specific antibody, was increased. Thus, we can conclude that vasopressin increases the phosphorylation of UT-A1 at S486 in rat IMCDs.

Vasopressin also increases UT-A1 phosphorylation at S499 in transfected LLC-PK₁ cells. However, determining whether vasopressin increases UT-A1 phosphorylation at S499 in rat IMCDs will require the development of an S499 phospho-specific antibody.

Previous immunohistochemical studies using an antibody to UT-A1 showed that the protein was present in the apical membrane and throughout the cytoplasm (17, 19). It was always somewhat surprising that these studies did not show better definition of staining at the apical membrane (17). While vasopressin clearly increases plasma membrane accumulation when assessed by surface biotinylation (14), similar evidence by immunohistochemistry was lacking. In the present study, our new S486 phospho-specific antibody shows clear labeling of the apical plasma membrane. In addition, apical membrane phospho-S486-UT-A1 increases in response to vasopressin. Thus, the present findings confirm our previous studies that used surface biotinylation to show that vasopressin increases the apical membrane accumulation of UT-A1 (1, 14). In addition, the present findings suggest that the phospho-S486-UT-A1 protein is preferentially located in the apical plasma membrane.

In contrast to the Western blot results, it may appear that there is less pS486-UT-A1 with vasopressin than without it in the confocal microscopy studies (Fig. 7). Confocal microscopy is very useful for assessing localization but not as useful for protein quantitation. By necessity, the confocal samples come from two separate animals, one treated and one not treated with vasopressin. In addition, the concentration of pS486-UT-A1 in the apical membrane with vasopressin may give the sense that there is less signal compared with the more diffuse staining without vasopressin.

In summary, we developed stably transfected mIMCD3 cell lines expressing wild-type and cAMP-unresponsive forms of UT-A1 and an S486-UT-A1 phospho-specific antibody. Since mIMCD3 cells are mouse in origin (20), they will facilitate the use of mouse-specific molecular reagents for the study of UT-A1. Using the S486-UT-A1 phospho-specific antibody, we confirmed that vasopressin increases UT-A1 accumulation in the apical plasma membrane. In addition, it allows us to extend previous findings by showing that UT-A1 is phosphorylated at S486 by vasopressin in rat IMCDs and that the S486-phospho-UT-A1 form is primarily detected in the apical plasma membrane.

GRANTS

This work was supported by the National Institutes of Health Grants R01-DK-41707 and P01-DK-61521.

DISCLOSURES

No conflicts of interest are declared by the authors.