INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

ALL HEALTHY ANIMALS CAN tolerate severe losses and/or depletion of water thanks to their capacity to concentrate urine. This is accomplished by high water absorption in the collecting duct mediated by an arginine vasopressin (AVP)-induced increase in water permeability in the apical plasma membrane (3). Discovery of the water channel proteins known as aquaporins (AQP) has led to a better understanding of the extraordinary capacity of renal tubules to adsorb water (17, 22, 25, 32). Several members of the AQP family are expressed in the kidney (7, 8, 11, 12, 30). One of these, aquaporin-2 (AQP2), is exclusively expressed in the renal collecting duct and appears to play a major role in vasopressin-regulated water transport across collecting duct cells (6, 7, 26, 27).

The precise mechanisms by which water permeability is regulated in the renal collecting duct are not yet known. However, protein phosphorylation is thought to be of importance (5, 14, 18, 19). In the collecting duct, AVP binds to adenylyl cyclase-coupled vasopressin (V₂) receptors, leading to activation of cAMP-dependent protein kinase A (PKA) (9, 29). AQP2 contains a single consensus PKA phosphorylation site at Ser²⁵⁶ (Arg-Arg-Gln-Ser) (6, 7). Moreover, it has recently been reported that, in vitro and in model cell lines, AQP2 is phosphorylated at this site by PKA and that this may lead directly or indirectly to increased water permeability (5, 14, 15, 18, 19). We have examined the phosphorylation of AQP2 in intact rat renal tissue. The results obtained indicate that phosphorylation of Ser²⁵⁶ of AQP2 is modulated by AVP or by other agents that activate the PKA pathway, suggesting that water permeability may be regulated by this process in collecting duct cells.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Animals. Male Sprague-Dawley rats weighing 200-250 g (B&K Universal, Sollentuna, Sweden) were used. The rats were fed a standard rat chow and had free access to normal drinking water. Rats were anesthetized with thiobutabarbital (8 mg/100 g body wt). Kidneys were rapidly removed and the papillae were excised immediately. All studies were started between 9 and 10 a.m.

³²P prelabeling. The preweighted papillae were incubated at 30°C in 2 ml of Krebs bicarbonate buffer containing (in mM) 124 NaCl, 4 KCl, 26 NaHCO₃, 10 glucose, 1.5 MgSO₄, 1.5 CaCl₂, 0.25 KH₂PO₄, and 1 sodium butyrate and oxygenated with 95% O₂-5% CO₂ (vol/vol). After 15 min, the medium was replaced with the same amount of fresh buffer containing 0.5 mCi [³²P]orthophosphoric acid (NEN Life Science Products, Boston, MA) and the tissue was incubated for 60 min. The papillae were washed twice with 2 ml of fresh buffer and incubated for an additional 1-10 min either with vehicle or in the presence of [Arg⁸]vasopressin (AVP), 1-desamino-[8-D-arginine]vasopressin (DDAVP), or forskolin plus 3-isobutyl-1-methylxanthine (IBMX) (Sigma-Aldrich Sweden, Stockholm, Sweden). After drug treatment, buffer was removed and tissue was rapidly frozen in dry ice and stored at 80°C.

Immunoprecipitation of AQP2. Preweighted papillae were sonicated in 1 ml of lysis buffer (20 mM Tris · HCl, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, and 0.2% BSA, pH 8.0) containing 1 mM EGTA, 50 mM NaF, and protease inhibitors (25 mM benzamidine, 100 µM phenylmethylsulfonyl fluoride, 5 µg/ml chymostatin, 5 µg/ml pepstatin A, 20 µg/ml leupeptin, and 20 µg/ml antipain). Peptides were from the Peptide Institute, Japan. For each time point, aliquots of the homogenates (2 µl) from both vehicle- and drug-treated samples were used for determination of total ³²P incorporation into trichloroacetic acid-precipitated proteins. The ³²P incorporation into the intact renal papilla was constant among the samples (data not shown). Five milligrams of preswollen Protein A-Sepharose CL-4B beads (Pharmacia Biotech, Uppsala, Sweden) were added to each tube and the samples were mixed for 20 h at 4°C. The sepharose beads and nonspecifically adsorbed proteins were removed by centrifugation for 15 s at 10,000 g. The supernatant was mixed for 90 min at 4°C with 8 µl of rabbit anti-serum LL126 (final dilution 1:400) raised against the final 22 amino acids of AQP2 (27). Samples were transferred to Eppendorf tubes containing 5 mg of preswollen Protein A-Sepharose beads and incubated for 1 h at 4°C. The beads were collected by centrifugation and washed once with 1 ml of lysis buffer; once with 1 ml of a buffer containing 20 mM Tris · HCl, 150 mM NaCl, 5 mM EDTA, 0.5% Triton X-100, 0.1% SDS, and 0.2% BSA (pH 8.0); once with 1 ml of a buffer containing 20 mM Tris · HCl, 500 mM NaCl, 0.5% Triton X-100, and 0.2% BSA (pH 8.0); and once with 1 ml of a buffer containing 50 mM Tris · HCl (pH 8.0). After the final wash, the beads were resuspended in 25 µl of SDS-PAGE sample buffer (50 mM Tris · HCl, 10% glycerol, 2% SDS, 10% 2-mercaptoethanol, and 0.01% bromphenol blue, pH 6.8), vortexed, and centrifuged. The recovered proteins were separated by SDS-PAGE on 12% acrylamide gels. Gels were dried, and ³²P incorporation into AQP2 was analyzed by autoradiography and by using a GS-250 Molecular Imager (Bio-Rad Laboratories, Sundbyberg, Sweden).

Two-dimensional phosphopeptide mapping. Gel pieces containing ³²P-labeled AQP2 were excised from dried gels, washed, and incubated with trypsin as described (2). Aliquots were spotted on a thin-layer cellulose plate (20 × 20 cm; in the middle and 4 cm from the bottom) and initially separated by electrophoresis at pH 3.5 in 10% acetic acid/1% pyridine until the dye front migrated 7 cm. Ascending chromatography was performed in 1-butanol/acetic acid/water/pyridine (15:3:12:10, vol/vol). Phosphopeptides were visualized by autoradiography.

Immunoblotting. The excised rat papillae were incubated at 30°C in saline solution containing (in mM) 137 NaCl, 5 KCl, 0.8 MgSO₄, 0.33 Na₂HPO₄, 0.44 KH₂PO₄, and 1 MgCl₂, with constant oxygenation with 95% O₂-5% CO₂ (vol/vol). After a 30-min preincubation, 10⁷ M AVP or vehicle was added and the tissue was incubated for an additional 5 min, then frozen in dry ice and stored at 80°C.

7

Double affinity purification of the anti-phospho-AQP2 antibody. To obtain antibody clones that exclusively recognize phosphorylated AQP2, we subjected the antisera raised against the phospho-AQP2 peptide to affinity purification using a two-step sequential affinity-purification protocol. For this purpose, both the phospho-AQP2 peptide mentioned above and a non-phospho-AQP2 peptide (1, 27) were used. The peptides (2 mg) were immobilized via covalent sulfhydryl linkage to activated agarose beads in columns (Sulfolink Immunobilization Kit No. 2; Pierce, Rockford, IL). As the first step of purification, 1 ml of the antiserum (anti-phospho-AQP2 serum) was applied to the column on which the non-phospho peptide was immobilized. The 1 ml antiserum preparation was applied on the same column three times (to obtain maximal extraction of potential clones that may recognize nonphosphorylated AQP2 and therefore need to be removed). The final eluate was then subjected to the second step of affinity purification. The preparation was applied to the column on which the phospho-AQP2 peptide was immobilized and recycled several times (to maximize extraction of antibody clones recognizing the phospho-AQP2 peptide and thus phosphorylated AQP2). Finally, the column was washed extensively and the column-bound antibody was eluted at low pH (1). This preparation was used for immunoblotting using the phospho-AQP2 peptide and the non-phospho-AQP2 peptide. Immunoblotting indicated that the final preparation of the affinity-purified antibody exclusively recognized the phospho-AQP2 peptide and did not cross-react with the non-phospho-AQP2 peptide (see Fig. 5).

Statistics. Values are presented as means ± SE. Comparisons between data were made by unpaired t-test. Values of P < 0.05 were considered significant.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

The phosphorylation of AQP2 was initially analyzed in ³²P-prelabeled rat renal papilla by immunoprecipitation followed by SDS-PAGE electrophoresis and autoradiography. Under control conditions, two phosphorylated polypeptides were detected that corresponded, respectively, to the nonglycosylated (29 kDa) and glycosylated (35-48 kDa) forms of AQP2, indicating that there was basal phosphorylation of AQP2 in intact rat renal tissue. The basal phosphorylation of AQP2 found in our experiments may be attributed to the presence of circulating AVP in the anesthetized rats or to the constitutive activation of adenylyl cyclase in the kidney papilla preparation. Addition of either AVP (10⁷ M) or DDAVP (10⁷ M) increased phosphorylation of both nonglycosylated and glycosylated forms of AQP2 (Fig. 1). In the presence of AVP, the increase in phosphorylation was rapid, reaching a maximal level within 1 min and remaining elevated throughout the time course of the incubation (Fig. 2). The effect of AVP was dose-dependent (Fig. 3), with a significant increase in phosphorylation being observed with 10 pM AVP compared with vehicle. A significant increase in phosphorylation of AQP2 was also observed after incubation with forskolin (10⁵ M) plus IBMX (5 × 10⁴ M), concentrations of which would increase cAMP levels (Fig. 4A).

View larger version (48K): [in this window] [in a new window]   Fig. 1.   Effect of arginine vasopressin (AVP) and 1-desamino-[8-D-arginine]vasopressin (DDAVP) on phosphorylation of aquaporin-2 (AQP2) in rat renal papillae. 32P-prelabeled rat renal papillae were incubated with vehicle (lane 1), 107 M AVP (lane 2), or 107 M DDAVP (lane 3) for 5 min at 30°C. An antibody specific for the 22 COOH-terminal amino acids of rat AQP2 was used for immunoprecipitation. Samples were separated by SDS-PAGE and analyzed by autoradiography. The 2 bands with apparent Mr values of 29 kDa and 35-48 kDa correspond to nonglycosylated and glycosylated AQP2, respectively.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Time course of AVP-stimulated AQP2 phosphorylation. Tissue was incubated with vehicle or 107 M AVP for various periods of time and samples were processed as described in legend to Fig. 1. 32P incorporation into AQP2 was analyzed by autoradiography and quantitation was carried out using a GS-250 Molecular Imager (Bio-Rad Laboratories). Data shown represent only glycosylated form of AQP2. Relative increase in 32P incorporation was same for nonglycosylated polypeptide. Data are expressed as percentage of 32P incorporated into AQP2 under control conditions for each respective time point. Values are means ± SE of 4-6 determinations.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   AVP stimulates phosphorylation of AQP2 in a concentration-dependent manner. Tissue was incubated with vehicle or various concentrations of AVP for 5 min and samples were processed as described in legend to Fig. 1. For each concentration of AVP, data are expressed as percentage of 32P incorporated into AQP2 under control conditions. Values are means ± SE of 3-6 determinations. At 10 pM AVP, 32P incorporation was increased 1.63-fold compared with vehicle (P < 0.05).

View larger version (36K): [in this window] [in a new window]   View larger version (36K): [in this window] [in a new window]   Fig. 4.   AVP and forskolin increase phosphorylation of the same site in AQP2. A: 32P-labeled papillae were incubated with vehicle (lane 1) or 105 M forskolin plus 0.5 mM 3-isobutyl-1-methyxanthine (IBMX) (lane 2) for 5 min. Samples were processed as described in legend to Fig. 1. B: two-dimensional tryptic phosphopeptide maps of phosphorylated AQP2. 32P-labeled AQP2 samples, prepared as shown in Fig. 4A, were excised from dried gels, washed, and digested with trypsin. Tryptic digests were separated by electrophoresis in first dimension and by chromatography in second dimension. O, origin; positive electrode is on left, negative electrode is on right. Phosphopeptides were visualized by autoradiography. Results shown are representative of 2 independent experiments. In other experiments, both peptides 1 and 2 were identified, but peptide 1 was present at higher level than peptide 2. This supports idea that peptides 1 and 2 are derived from alternative tryptic cleavage of a single phosphorylation site. A minor peptide could be detected in maps from highly phosphorylated samples that ran just below peptide 2. Level of this peptide varied in different samples and is likely to be a derivative of peptide 1 or 2.

Two-dimensional phosphopeptide mapping of ³²P-labeled AQP2 was carried out to further examine the basal and stimulated phosphorylation of the protein. Under control conditions, four phosphorylated peptides were detected (Fig. 4B, labeled 1-4). In the presence of AVP, there was a pronounced increase in phosphorylation of peptides 1 and 2, and an apparent decrease in phosphorylation of peptides 3 and 4. An essentially identical result was found for the AQP2 sample obtained from tissue incubated with forskolin.

To clarify the identity of the peptides phosphorylated, we prepared a phosphorylation state-specific antibody that specifically recognized the phosphorylated, but not the dephosphorylated, form of Ser²⁵⁶ (Fig. 5). Absorption of the double affinity-purified antibody with the immunizing phospho-peptide revealed a complete absence of labeling (data not shown). Immunoblotting studies of papillary tissue incubated under control conditions indicated that there was a measurable level of Ser²⁵⁶ phosphorylation in intact normally hydrated rats. Incubation with AVP significantly increased the phosphorylation of Ser²⁵⁶ in both the glycosylated and nonglycosylated forms of AQP2. Under AVP treatment, the level of phosphorylation was 218 ± 18% (n = 9) compared with a corresponding control (Fig. 6A). The total amount of AQP2 was not altered by AVP treatment (Fig. 6B).

View larger version (31K): [in this window] [in a new window]   Fig. 5.   Characterization of an antibody that specifically recognizes phosphorylated form of Ser256 in AQP2. Nonphosphorylated AQP2 peptide (residues 250-271; 50 ng in lane 3 and 200 ng in lane 4) or phosphorylated AQP2 peptide (P-AQP2; residues 253-262; 50 ng in lane 5 and 200 ng in lane 6) were separated by SDS-PAGE (12% acrylamide). Peptides were transferred to nitrocellulose membrane, and the phosphorylated form of AQP2 was detected using a phosphorylation state-specific antibody raised against a synthetic peptide corresponding to amino acids 253-262 of AQP2, which was chemically phosphorylated at Ser256. Antibody binding was detected by enhanced chemiluminescence. As a control, membranes from kidney inner medulla, prepared in the presence of protein phosphatase inhibitors (1 µg in lane 1 and 5 µg in lane 2), were also analyzed. The antibody labeled 2 bands of 28 and 35-45 kDa corresponding to the nonglycosylated and glycosylated forms, respectively, of basally phosphorylated AQP2.

View larger version (52K): [in this window] [in a new window]   Fig. 6.   Ser256 of AQP2 is phosphorylated in kidney papilla. A: papillae incubated with 107 M AVP for 5 min (lanes 2, 4, 6) or control papillae from the same rats (lanes 1, 3, 5) were homogenized and homogenates were suspended in Laemmli sample buffer and separated by SDS-PAGE (10 µg protein per lane). Proteins were transferred to Hybond-P PVDF membrane, and phosphorylated form of AQP2 was detected using phosphorylation state-specific antibody. Antibody binding was detected by enhanced chemiluminescence. Similar results were obtained in 3 independent experiments. B: reprobing of same membrane using antiserum that recognized both phosphorylated and nonphosphorylated forms of AQP2. Total amount of AQP2 was not altered by AVP treatment.

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

It is generally agreed that AVP exerts its antidiuretic effect by acting on V₂ receptors in the collecting duct and that the target protein is AQP2. The effect of AVP on AQP2 involves several steps. Under conditions of normal hydration, AQP2-containing vesicles are located intracellularly (23, 26, 27). After activation of V₂ receptors, the vesicles then fuse with the apical membrane, leading to a dramatic increase in the membrane water permeability (26). V₂ receptors are coupled to the adenylyl cyclase-cAMP-PKA pathway. AQP2 contains a single PKA consensus phosphorylation site (6, 7), and phosphorylation of the water channel has been suggested to play a role in AVP-dependent increases in water permeability in kidney collecting ducts (5, 14, 18, 19).

To date, studies of AQP2 phosphorylation have been carried out either in vitro (18, 19) or in cells expressing recombinant AQP2 (5, 14, 18). However, it was not known if the phosphorylation of AQP2 occurs in intact renal tissue or if the phosphorylation of Ser²⁵⁶ of AQP2 was regulated by AVP. The present study demonstrates that Ser²⁵⁶ of AQP2 is phosphorylated in response to the actions of AVP acting as a first messenger in intact rat renal tissue. DDAVP, the V₂ receptor agonist, and forskolin increased phosphorylation of AQP2. The two-dimensional phosphopeptide mapping studies indicated that AVP and forskolin treatment led to phosphorylation of the same site, suggesting that the effect of AVP on AQP2 phosphorylation was mediated via adenylyl cyclase and activation of PKA. By use of a phosphorylation state-specific antibody that specifically recognized the phosphorylated form of Ser²⁵⁶, we demonstrated that Ser²⁵⁶ was the major residue phosphorylated under control conditions and in response to incubation with AVP. Under basal conditions, AQP2 was phosphorylated to a low level at sites distinct from Ser²⁵⁶. The presence of basal phosphorylation of these sites leads to a slight underestimation of the efficacy of the AVP action in the results shown in Figs. 1-3. Moreover, in the immunoblotting study, where only Ser²⁵⁶ phosphorylation was detected, the calculated increase in AQP2 phosphorylation in response to AVP was significantly higher than that obtained by immunoprecipitation. The identity of the other phosphorylation site(s) is not known. There are four other potential sites for phosphorylation in AQP2, one for protein kinase C (PKC) and three for casein kinase II (6, 7). Future studies will hopefully reveal whether the minor site(s) phosphorylated under control conditions correspond to the PKC or casein kinase II sites, and whether phosphorylation of these sites might regulate AQP2.

The results obtained in this study demonstrate that the phosphorylation of AQP2 is regulated in a physiologically relevant manner in rat renal tissue. AQP2 is phosphorylated under basal conditions, but within 1 min after addition of AVP, the level of AQP2 phosphorylation is significantly increased. These results are consistent with data obtained using perfused collecting ducts, in which water permeability is altered within 1 min by AVP treatment (33). A significant increase in phosphorylation of Ser²⁵⁶ of AQP2 was observed at concentrations of 10¹¹-10⁷ M AVP. The plasma concentration of AVP is about 2 pM in normal hydrated rats, and increases up to 12 pM after 24-h dehydration (13, 34). Thus the present results support the concept that phosphorylation of AQP2 is one of the first steps leading to increased water permeability in the collecting duct cells. Recent studies of LLC-PK₁ cells expressing recombinant AQP2 (5, 14) have indicated that phosphorylation of AQP2 is one of a series of regulatory processes involved in the exocytosis of AQP2-containing vesicles (4, 16, 20, 28). The development of the phosphorylation state-specific antibody will hopefully help in future studies in defining the precise role of AVP-mediated phosphorylation of AQP2 in regulating its translocation and/or membrane insertion.

Several observations have demonstrated that sustained activation of V₂ receptors may lead to increased AQP2 synthesis, suggesting that AVP may play a role in the long-term regulation of AQP2 (1, 21, 22, 31). We recently reported that a single injection of DDAVP into rats resulted in significant accumulation of AQP2 mRNA within 6 h (35). It was also shown that AQP2 mRNA abundance is decreased in animals after treatment with a V₂-receptor antagonist (10). We have demonstrated that human AQP2 gene promoter activity was increased after V₂ receptor stimulation and activation of the PKA pathway in LLC-PK₁ cells (36). This effect was mediated by cAMP-responsive element (CRE) and AP-1 consensus sites present in the AQP2 promoter region. Vasopressin increased CRE binding (CREB) protein phosphorylation and CREB-CRE binding, as well as c-Fos mRNA and protein expression and c-Fos/c-Jun-AP-1 binding (36). Upregulation of CRE binding proteins was also demonstrated in dehydrated rats (24). All of these observations, together with data presented here, suggest that AVP regulates AQP2 both on a short- and long-term basis and that cAMP plays the central role in both processes. Such a dual effect of AVP would guarantee a sustained increase in urinary concentrating capacity after prolonged dehydration.