Involvement of direct phosphorylation in the regulation of the rat parotid Na(+)-K(+)-2Cl- cotransporter.

We identify a 175-kDa membrane phosphoprotein (pp175) in rat parotid acini whose properties correlate well with the Na(+)-K(+)-2Cl- cotransporter previously characterized functionally and biochemically in this tissue. pp175 was the only phosphoprotein immunoprecipitated by an anti-Na(+)-K(+)-2Cl- cotransporter antibody and the only membrane protein whose phosphorylation state was conspicuously altered after a brief (45-s) exposure of acini to the beta-adrenergic agonist isoproterenol. Phosphopeptide mapping provided evidence for three phosphorylation sites on pp175, only one of which was labeled in response to isoproterenol treatment. The half-maximal effect of isoproterenol on phosphorylation of pp175 (approximately 20 nM) was in excellent agreement with its previously demonstrated up-regulatory effect on cotransport activity. Increased phosphorylation of pp175 was also seen following acinar treatment with a permeant cAMP analogue and with forskolin, conditions that have likewise been shown to up-regulate the cotransporter. Combined with earlier results from our laboratory, these data provide strong evidence that the up-regulation of the cotransporter by these agents is due to direct phosphorylation mediated by protein kinase A. AlF(-)4 treatment, which results in an up-regulation of cotransport activity comparable with that observed with isoproterenol (approximately 6-fold), caused a similar increase in phosphorylation of pp175. However, hypertonic shrinkage and treatment with the protein phosphatase inhibitor calyculin A, which also up-regulate the cotransporter (approximately 3-fold and approximately 6-fold, respectively) caused no change in the phosphorylation level. Furthermore, although acinar treatment with the muscarinic agonist carbachol results in a dramatic up-regulation of cotransport activity and a concomitant phosphorylation of pp175, no phosphorylation of pp175 was seen with the Ca(2+)-mobilizing agent thapsigargin, which is able to fully mimic the up-regulatory effect of carbachol on transport activity. Taken together, these results indicate that direct phosphorylation is only one of the mechanisms involved in secretagogue-induced regulation of the rat parotid Na(+)-K(+)-2Cl- cotransporter.

We identify a 175-kDa membrane phosphoprotein (pp175) in rat parotid acini whose properties correlate well with the Na ؉ -K ؉ -2Cl ؊ cotransporter previously characterized functionally and biochemically in this tissue. pp175 was the only phosphoprotein immunoprecipitated by an anti-Na ؉ -K ؉ -2Cl ؊ cotransporter antibody and the only membrane protein whose phosphorylation state was conspicuously altered after a brief (45-s) exposure of acini to the ␤-adrenergic agonist isoproterenol. Phosphopeptide mapping provided evidence for three phosphorylation sites on pp175, only one of which was labeled in response to isoproterenol treatment. The half-maximal effect of isoproterenol on phosphorylation of pp175 (Ϸ20 nM) was in excellent agreement with its previously demonstrated up-regulatory effect on cotransport activity. Increased phosphorylation of pp175 was also seen following acinar treatment with a permeant cAMP analogue and with forskolin, conditions that have likewise been shown to up-regulate the cotransporter. Combined with earlier results from our laboratory, these data provide strong evidence that the up-regulation of the cotransporter by these agents is due to direct phosphorylation mediated by protein kinase A. AlF 4 ؊ treatment, which results in an up-regulation of cotransport activity comparable with that observed with isoproterenol (ϳ6-fold), caused a similar increase in phosphorylation of pp175. However, hypertonic shrinkage and treatment with the protein phosphatase inhibitor calyculin A, which also up-regulate the cotransporter (ϳ3-fold and ϳ6-fold, respectively) caused no change in the phosphorylation level. Furthermore, although acinar treatment with the muscarinic agonist carbachol results in a dramatic up-regulation of cotransport activity and a concomitant phosphorylation of pp175, no phosphorylation of pp175 was seen with the Ca 2؉ -mobilizing agent thapsigargin, which is able to fully mimic the up-regulatory effect of carbachol on transport activity. Taken together, these results indicate that direct phosphorylation is only one of the mechanisms involved in secretagogue-induced regulation of the rat parotid Na ؉ -K ؉ -2Cl ؊ cotransporter.
Because of its experimental accessibility, relative homogeneity and rich hormonal responsiveness, the rat parotid gland is rapidly becoming one of the more popular mammalian experimental models for the study of the mechanism(s) and regula-tion of epithelial fluid and electrolyte secretion (1,2). Work from a number of laboratories has established that salt and water secretion by the acinar cells, which comprise the bulk of this gland, is due to transepithelial Cl Ϫ movement (1)(2)(3). The active step in this process is Cl Ϫ entry across the acinar basolateral membrane, a large component of which has been shown to be due to Na ϩ -K ϩ -2Cl Ϫ cotransport (4,5).
Consistent with its important role in secretion, we have shown that the activity of the rat parotid Na ϩ -K ϩ -2Cl Ϫ cotransporter is regulated by a number of physiological and other potentially physiologically relevant stimuli. We first demonstrated a substantial (ϳ6-fold) up-regulation of cotransporter activity following ␤-adrenergic stimulation and provided good evidence that this was due to a phosphorylation event mediated by cyclic AMP-dependent protein kinase (6). This up-regulation is paralleled in vivo by an increase in salivary flow seen when sympathetic (adrenergic) stimulation, arising, for example, from mastication, is superimposed on parasympathetic (muscarinic) stimulation (7), the main fluid secretory stimulus for the gland. In a later publication (8) we demonstrated that the rat parotid Na ϩ -K ϩ -2Cl Ϫ cotransporter is up-regulated (again ϳ6-fold) by aluminum fluoride (AlF 4 Ϫ ), an activator of G-proteins, and by calyculin A, a protein phosphatase inhibitor. Based on several factors, including diverse sensitivity to blockade of up-regulation by protein kinase inhibitors and the observation that AlF 4 Ϫ does not induce cAMP generation in the rat parotid, we have argued that the mechanisms of action of AlF 4 Ϫ and calyculin A on the cotransporter are different from that of ␤-adrenergic stimulation and from one another (8).
More recently (9) 1 we have shown that Na ϩ -K ϩ -2Cl Ϫ cotransport activity in these cells is also increased by muscarinic stimulation (Ͼ15-fold) and by hypertonic shrinkage (ϳ3-fold). Our data suggest that these latter effects are also unrelated to one another and unrelated to the effect of ␤-adrenergic stimulation (see "Discussion"). At this time our understanding of these up-regulatory events is still incomplete, and the physiological significance of some of these stimuli remains to be determined. However, our results clearly demonstrate that the rat parotid Na ϩ -K ϩ -2Cl Ϫ cotransporter is under tight regulatory control, in all likelihood by multiple intracellular signaling pathways, and thus that it provides a particularly rich experimental system for the study of transport regulation by hormonal and other stimuli.
In the present paper we explore these phenomena further by studying the effects of these various up-regulatory stimuli on the phosphorylation state of the Na ϩ -K ϩ -2Cl Ϫ cotransport protein itself. Although it is generally accepted that phosphorylation events play an important role in cellular signaling, relatively few studies have actually directly explored their possible * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  involvement in the regulation of facilitative membrane transport proteins. We show here that there is a good correlation between increased transport activity and increased transporter phosphorylation following ␤-adrenergic stimulation and AlF 4 Ϫ treatment of rat parotid acini, suggesting that the regulation of the cotransporter by these stimuli is due to direct phosphorylation. Somewhat surprisingly, however, this was not the case for the other stimuli studied, in spite of the fact that some of these agents have been shown to increase both the transport activity and the phosphorylation state of Na ϩ -K ϩ -2Cl Ϫ cotransporters in lower species (11,12). These observations indicate that the Na ϩ -K ϩ -2Cl Ϫ cotransporter in the rat parotid is regulated both via direct phosphorylation and via other, as yet unidentified, mechanisms.

EXPERIMENTAL PROCEDURES
Materials and Media-Male Wistar strain rats, weighing 250 -300 g, were purchased from Harlan Sprague-Dawley (Indianapolis, IN). Carrier-free 32 P i (10 mCi/ml) and [ 3 H]bumetanide (80.8 Ci/mmol) were obtained from Amersham Corp. Collagenase P, protease inhibitors, and dibutyryl cAMP were from Boehringer Mannheim. (Ϫ)Isoproterenol, phorbol 12-myristate 13-acetate (PMA), 2 V8 protease, and bovine serum albumin (number A6003) were purchased from Sigma. Calyculin A and forskolin were from Calbiochem. Phosphatidylserine (number 840032, from bovine brain; supplied in chloroform) was from Avanti Polar Lipids (Birmingham, AL). Molecular weight standards, prepoured 4 -20% SDS-PAGE gels, and prepoured 16% Tricine gels were obtained from Novel Experimental Technology (San Diego, CA). Protein G-Sepharose beads were from Pierce. All other chemicals were from standard commercial sources and were reagent grade or the highest purity available.
Protein concentration was measured with the BCA protein assay system (Pierce).
Preparation of Parotid Acini-Rats were anesthetized with diethyl ether and killed by cardiac puncture. The parotid glands were removed, dissected free of fat, and finely minced in a small volume of ice-cold digestion medium. The glands from two rats were suspended in 10 ml of digestion medium and incubated at 37°C with continuous agitation. The mince was dispersed by gently pipetteting 10 times with a 10-ml plastic pipette after 30 min and again after 45 min of digestion. The mince was then centrifuged (400 ϫ g, 15 s), the supernatant was discarded and replaced with the same volume of fresh digestion medium, and incubation was continued. After a total of 60 min of digestion the mince was pipetted 10 times with a 10-ml pipette fitted with a Rainin RC-200 blue pipette tip (Rainin Instruments, Emeryville, CA). Finally at times 75 and 85 min the mince was pipetted five times with a 10-ml pipette fitted with a Rainin RT-96 yellow tip. The resulting suspension was centrifuged and resuspended in 10 ml of PO 4 Ϫ -free-PSS (PSS without NaH 2 PO 4 ) and passed through a 450-m nylon screen (PGC Scientific, Gaithersburg, MD). The material in the filtrate was collected by centrifugation, resuspended in 10 ml of fresh PO 4 Ϫ -free PSS, and recentrifuged, then resuspended in 3-4 ml of PO 4 Ϫ -free PSS. 32 P i Labeling and Stimulation-Acini were incubated at 37°C in PO 4 Ϫ -free-PSS in the presence of 50 Ci/ml 32 P i for 30 min. The suspension was then placed on the bench for 1-2 min to allow the acini to settle, and the supernatant was carefully removed and discarded. The remaining material was diluted to 10 ml with PSS, washed once in the same volume, and finally resuspended in 3-4 ml of PSS at room temperature.
Aliquots (200 l) of labeled cells were incubated with the agents indicated at 37°C in siliconized glass tubes (Sigmacote number SL-2; Sigma). The incubation was terminated by the addition of 800 l of ice-cold stop solution and disruption by immersion in a Branson B-12 sonicator bath (Shelton, CT) as follows. Each sample was first sonicated for 30 s and then placed on ice. After disruption of all samples in this way, each sample was subsequently sonicated to clarity.
Preparation of Membrane Extract-Sonicated samples (1 ml total volume) were centrifuged at 1000 ϫ g for 10 min, and the resulting pellet was discarded. The supernatant was centrifuged at 100,000 ϫ g for 1 h. We refer to the supernatant and pellet from this high speed spin as the "cytosolic fraction" and the "particulate fraction," respectively. The particulate fraction was resuspended in 0.7 ml of extraction buffer (stop solution containing 0.3% Triton X-100 and no NaCl) and kept on ice for 30 min. This sample was then centrifuged at 100,000 ϫ g for 30 min. The supernatant and pellet from this second high speed spin are referred to as the "Triton extract" and the "Triton-insoluble fraction" respectively. As shown below (see "Results") the Triton extract was the only fraction that contained the phosphoprotein of interest (the Na ϩ -K ϩ -2Cl Ϫ cotransporter), and thus only this fraction was usually retained and analyzed.
Gel Electrophoresis and Autoradiography-SDS-PAGE was performed essentially as described by Laemmli (13) using a 4% polyacrylamide stacking gel and a 4 -20% (continuous gradient) polyacrylamide separating gel. Tricine-SDS (16%) electrophoresis was carried out according to Schagger and von Jagow (14). Samples were heated at 100°C for 2 min in sample buffer containing 2.5% SDS, 50 mM Tris-HCl (pH 6.8), 4% glycerol, 100 mM dithiothreitol, and 0.04% bromphenol blue and centrifuged before electrophoresis. Gels were stained with Coomassie Blue, dried, and visualized by autoradiography using Kodak X-Omat AR film (Eastman Kodak, Rochester, NY). Gels used in phosphopeptide mapping studies were washed twice for 15 min in 20% ethanol before drying to remove acetic acid. Autoradiographs were scanned using a Molecular Dynamics computing densitometer (Molecular Dynamics, Sunnyvale, CA) to quantitate 32 P i labeling. Linearity of the densitometric scans was confirmed using autoradiographic 14 C Micro-Scales (Amersham Corp.). Samples of Triton extract electrophoresed for quantitation of 32 P i labeling typically contained ϳ15 g of protein (the protein concentration of the Triton extract was ϳ0.5 mg/ml).
Production of Antiserum against Parotid Bumetanide Binding Protein-The following method was used to produce sheep antiserum directed against a rabbit parotid bumetanide binding protein previously identified in our laboratory as the bumetanide moiety (and perhaps all) of the Na ϩ -K ϩ -2Cl Ϫ cotransporter in this tissue (15). Deglycosylated bumetanide binding protein (M r ϳ135,000) was purified from rabbit parotid basolateral membranes as described previously (15). A suitable quantity of protein (100 g for the first injection and 30 -40 g for subsequent injections) was diluted to 1 ml with phosphate-buffered saline and combined with 1 ml of complete (first injection) or incomplete (subsequent injections) Freund's adjuvant. Injections were carried out at the NIH Animal Care Center (Ungulate Section) in Poolesville, Maryland. The primary injection was subcutaneous, and subsequent injections (3,8,19, and 36 weeks later) were intramuscular; all injections were done at multiple sites. Anti-bumetanide binding protein antibody titer in serum samples was monitored by Western blot analysis against purified bumetanide binding protein and rabbit parotid basolateral membranes. In these samples the antiserum strongly labeled proteins of M r 135,000 and 160,000 -175,000, respectively. 3 The range of molecular weights observed for the labeling of native membranes (160 -175 kDa) was related to the gel system used. With the large format Bio-Rad gels used in our previous publication (15) we observed an M r of 160,000 (in this same publication we showed that the parotid bumetanide binding protein had a native M r of 160,000 and a deglycosylated M r of 135,000), while with the prepoured Novex minigels used in the present work we consistently observed a M r of ϳ175,000.
Immunoprecipitation of 32 P i -Labeled Proteins-Protein G-Sepharose beads were prewashed twice with washing buffer (extraction buffer titrated to pH 8.6 with Tris and containing 0.1% SDS and 300 mM NaCl), once in low pH buffer (50 mM glycine-HCl, pH 2.3, containing 150 mM NaCl and 0.5% Triton X-100), and once in extraction buffer and then suspended in extraction buffer containing 1% ovalbumin.
A 400-l aliquot of Triton extract was incubated overnight at 4°C with immune or non-immune sheep serum (7 l/100 g of extract protein). Prewashed protein G-Sepharose beads (10 l of beads/l of serum) were then added. After 30 min of additional incubation, the beads were collected by centrifugation and washed six times with washing buffer. The tube was changed for the last spin. Protein retained by the washed beads was then eluted with 100 l of electrophoresis sample buffer.

Immunoprecipitation of [ 3 H]Bumetanide
Binding Activity-Triton extracts for bumetanide binding studies were prepared by a modification of the procedure given above. This modification was based on our previous observation that the bumetanide binding activity of the Na ϩ -K ϩ -2Cl Ϫ cotransporter could be preserved in detergent solutions by the addition of suitable exogenous lipids (16,17). The procedure was as follows. A "particulate fraction" was prepared as described above from cells that had not been labeled with 32 P i . However, instead of extraction buffer, the particulate fraction was resuspended in Buffer K (100 mM mannitol, 10 mM HEPES, 1 mM EDTA, 195 mM potassium gluconate, and 5 mM KCl buffered to pH 7.4 with Tris). This material was centrifuged again at 100,000 ϫ g for 1 h, and the resulting pellet was resuspended in Buffer K at a protein concentration of 2.5-3.5 mg/ml, fast frozen in aliquots, and stored above liquid nitrogen. On the day of the bumetanide binding experiment, frozen samples were thawed and diluted with Buffer K to a protein concentration of 2 mg/ml. This suspension was mixed with the same volume of 0.6% Triton X-100 in Buffer K, left on ice for 10 min, and then transferred to a glass tube in which a suitable volume of phosphatidylserine (final concentration 0.15%) had been evaporated. This mixture was sonicated to clarity in a Branson B-12 bath sonicator (ϳ60-s immersion) and centrifuged at 100,000 ϫ g for 1 h. The resulting supernatant, which is analogous to the "Triton extract" described above for 32 P i labeling studies, is referred to here as the "lipid-stabilized Triton extract." Immunoprecipitation of [ 3 H]bumetanide binding activity from the above lipid-stabilized Triton extracts was carried out using immune and nonimmune IgG preabsorbed onto protein G-Sepharose beads. This was done in order to avoid any possible interference of serum with the [ 3 H]bumetanide binding assay and to allow quantitation of protein remaining after immunoprecipitation (see "Results"). Protein G beads were washed twice with Buffer K containing 0.3% Triton X-100 and 0.15% phosphatidylserine (sonicated to clarity as above) and then resuspended in the same buffer containing 1% ovalbumin and incubated for 40 min with immune or nonimmune sheep serum (10 l of beads/l of serum; total volume ϳ300 l). The beads were then washed three times in Buffer K plus 0.3% Triton X-100 and 0.15% phosphatidylserine, added to the lipid-stabilized Triton extract (70 l of beads/100 g of extract protein), and incubated for 2 h at 4°C. After removal of the beads by centrifugation, [ 3 H]bumetanide binding activity remaining in the resulting supernatant was determined by the method given below.
[ 3 H]Bumetanide Binding Assay-Equilibrium bumetanide binding was measured using a nitrocellulose filtration assay as described previously (16,18). Briefly, a 20-l aliquot of sample was combined with 20 l of incubation medium consisting of either Buffer K containing 10 Ci/ml [ 3 H]bumetanide or the same medium with all potassium replaced by Na ϩ . After a 15-min incubation the reaction was terminated by the addition of ice-cold stop solution followed by Millipore filtration (HAWP 0.45 m). Other procedures were as described previously (16,18). [ 3 H]bumetanide binding observed in the absence of Na ϩ was subtracted from that observed in its presence to yield the Na ϩ -dependent component of binding. In previous studies (16,18) we have demonstrated that this Na ϩ -dependent component of binding represents the specific binding of bumetanide to its inhibitory site on the Na ϩ -K ϩ -2Cl Ϫ cotransporter.
Phosphopeptide Mapping-Triton extracts for phosphopeptide mapping studies were prepared as described above for 32 P i -labeling studies, except that the extraction buffer contained 1% Triton X-100 and the protein concentration of the extract was ϳ3 mg/ml. Following SDS-PAGE (ϳ90 g of extracted protein/lane) the band corresponding to the Na ϩ -K ϩ -2Cl Ϫ cotransporter (pp175, see "Results") was identified using autoradiography, cut from the dried gel, and swollen in 1 ml of a buffer containing 50 mM NH 4 HCO 3 (pH 8.0), 1 mM dithiothreitol, and 20 g/ml V8 protease. After 6 or 12 h of incubation at 37°C protease digestion was terminated by heating the sample to 100°C for 5 min. The liquid was set aside, and the gel slice was then sequentially incubated in 500 l of distilled water for 2 h, 500 l of distilled water for 1 h, 500 l of 0.1% SDS for 1.5 h, and 500 l of 0.1% SDS for 1 h. All of these samples were combined (total volume 3 ml) and dried in a Savant DNA Speed Vac (Savant Instruments Inc., Farmingdale, NY). This final V8 protease digest was then taken up in sample buffer and subjected to Tricine-SDS gel electrophoresis.
Data Analysis-All experiments were repeated three or more times with similar results. Data are given as means Ϯ S.E.

RESULTS
Evidence for Isoproterenol-dependent Phosphorylation of the Rat Parotid Na ϩ -K ϩ -2Cl Ϫ Cotransporter-As already mentioned, in a previously published report from our laboratory (6) we demonstrated that the Na ϩ -K ϩ -2Cl Ϫ cotransport activity of rat parotid acini is markedly (ϳ6-fold) and rapidly (within 40 s) up-regulated following ␤-adrenergic stimulation. In addition, our data indicated that this effect was mediated by protein kinase A. The first series of experiments presented below were undertaken in order to determine whether this effect could be attributed to a direct phosphorylation of the cotransport protein. In the experiment illustrated in Fig. 1 we compare protein phosphorylation patterns in 32 P i -labeled rat parotid acini incubated for 45 s in the presence (ϩ) or absence (Ϫ) of the ␤-adrenergic agonist isoproterenol (1 M). Following isoproterenol treatment the cells were disrupted, cytosolic (Cy) and particulate fractions were isolated, and the latter were further separated into Triton extracts (TE) and Triton-insoluble (TI) fractions (see "Experimental Procedures" for details). Fig. 1 shows an autoradiograph of an SDS-PAGE gel on which these various fractions were run. After this short period of incubation with isoproterenol, clear differences in the phosphorylation patterns of only two proteins were seen. The first corresponds to a relatively sharp band at M r ϳ95,000 in the autoradiographs of the cytosolic fractions (small arrow). This protein, whose phosphorylation is increased with isoproterenol treatment, is not considered further here. The second phosphoprotein appears as a rather diffuse band centered at M r Ϸ 175,000 in the autora- diographs of the Triton extracts (large arrow); its phosphorylation state is likewise markedly increased by isoproterenol, and its presence in the Triton extract indicates that it is an integral membrane protein.
A number of factors discussed in the remainder of the paper provide strong evidence that the 175-kDa phosphoprotein (pp175) identified above is (a major part or all of) the rat parotid Na ϩ -K ϩ -2Cl Ϫ cotransporter.
Demonstration of a Strong Correlation between Phosphorylation of pp175 and cAMP-dependent Up-regulation of the Parotid Na ϩ -K ϩ -2Cl Ϫ Cotransporter-In our previous report (6) we demonstrated that the half-maximal effect of isoproterenol for up-regulation of the rat parotid Na ϩ -K ϩ -2Cl Ϫ cotransporter was seen at ϳ20 nM, measured after 37.5 s of agonist incubation at 37°C. The dose response of isoproterenol for the phosphorylation of pp175, measured under essentially identical experimental conditions, is illustrated in Fig. 2A. The halfmaximal effect of isoproterenol for phosphorylation is also seen at ϳ20 nM, in excellent agreement with its effect on the upregulation of cotransport activity.
In parotid acinar cells, cAMP is thought to be the major intracellular messenger mediating the effects of ␤-adrenoreceptor stimulation. In our earlier work (6) we also demonstrated that significant up-regulation of cotransport activity was seen following acinar treatment with permeant analogues of cAMP and with forskolin, which increases intracellular cAMP by direct activation of the catalytic subunit of adenylate cyclase. Consistent with the effects of these agents on transport, in Fig.  2B we show that increased phosphorylation of pp175 is likewise seen when acini are treated with the permeant cAMP analogue dibutyryl cAMP and with forskolin.
The strong correlation established in Fig. 2 between the effects of isoproterenol and cAMP on phosphorylation of pp175 and their previously documented effects on Na ϩ -K ϩ -2Cl Ϫ cotransport activity (6) supports the hypotheses that pp175 is the rat parotid Na ϩ -K ϩ -2Cl Ϫ cotransporter and that its up-regulation by isoproterenol is due to direct phosphorylation.
Immunoprecipitation of pp175 by an Anti-Na ϩ -K ϩ -2Cl Ϫ Cotransporter Antibody-In order to further confirm that pp175 is indeed the rat parotid Na ϩ -K ϩ -2Cl Ϫ cotransporter we carried out several studies using a polyclonal antiserum raised against a rabbit bumetanide binding protein previously identified in our laboratory as the bumetanide binding moiety of the Na ϩ -K ϩ -2Cl Ϫ cotransporter in this tissue (see "Experimental Procedures" and Ref. 15). In Western blots of purified rat and rabbit parotid basolateral membranes this antiserum recognizes a broad band with M r ϳ175,000, which closely resembles the autoradiographic images of pp175 in Fig. 1. 4 In addition, as illustrated in Fig. 3, this antiserum quantitatively immunoprecipitates the Na ϩ -dependent component of bumetanide binding (the specific binding of bumetanide to the Na ϩ -K ϩ -2Cl Ϫ cotransporter) from a lipid-stabilized (see "Experimental Procedures") Triton extract of the rat parotid particulate fraction. In Fig. 3 we compare the Na ϩ -dependent component of bumetanide binding in the lipid-stabilized Triton extract with the activity of the same extract after immunoprecipitation with nonimmune serum (N) or immune serum (I) raised against the rabbit parotid bumetanide binding protein. No significant Na ϩdependent bumetanide binding activity remains in the supernatant after immunoprecipitation with immune serum (7.6 Ϯ 5.3% of control; n ϭ 3), while all binding is retained after precipitation with nonimmune serum. In addition, densitometric scans of Coomassie Blue-stained SDS-PAGE gels of the supernatants remaining after immune or nonimmune precipitation were virtually superimposable except for the region around M r ϳ175 kDa, where a broad minor protein band was missing from the sample treated with immune serum (not shown).
When this polyclonal antiserum was used in immunoprecipitation studies with Triton extracts from 32 P i -labeled cells it specifically precipitated pp175. This result is illustrated in Fig.  4. Here we compare autoradiographs of SDS-PAGE gels of material immunoprecipitated with nonimmune (N) or immune (I) serum from extracts of 32 P i -labeled acini pretreated with (ϩ) or without (Ϫ) 1 M isoproterenol. A single phosphoprotein appearing as a diffuse band centered at M r ϳ175,000 whose phosphorylation is markedly increased by isoproterenol treatment is seen in the precipitate from immune serum. No phosphoproteins were immunoprecipitated by nonimmune serum.

FIG. 2. Effects of isoproterenol concentration and cAMP on
the phosphorylation of pp175. 32 P i -Labeled acini were exposed to the concentrations of stimuli indicated for 45 s at 37°C and then Triton-extracts were isolated and analyzed by SDS-PAGE and autoradiography as in Fig. 1. The phosphorylation of the 175-kDa phosphoprotein (pp175) identified in Fig. 1 was quantified by scanning densitometry of the resulting autoradiographs. The pp175 phosphorylation determined in this way for each experimental condition has been normalized to the pp175 phosphorylation determined from a control (untreated) sample from the same preparation run on the same gel. The results shown are the means Ϯ S.E. of three or more independent experiments. A, phosphorylation of pp175 versus isoproterenol (ISO) concentration. B, phosphorylation of pp175 after acinar treatment with isoproterenol, the permeant cAMP analogue dibutyryl cAMP (DB-cAMP), and the activator of adenylate cyclase forskolin (FOR).
In addition, when the supernatants remaining after immunoprecipitation with immune serum were examined by SDS-PAGE and autoradiography, no 32 P i -labeled band at 175 kDa was detectable (not shown), indicating that all of the labeled protein at 175 kDa is recognized by the anti-Na ϩ -K ϩ -2Cl Ϫ cotransporter antibody.
Phosphopeptide Mapping- Fig. 5 shows the result of digestion of pp175 with V8 protease. In this experiment 32 P i -labeled pp175 from acini treated with (ϩ) or without (Ϫ) isoproterenol was incubated with V8 protease for 6 or 12 h as indicated (see "Experimental Procedures"), and the resulting digests were separated by Tricine-SDS gel electrophoresis and visualized by autoradiography. Three 32 P i -labeled bands are clearly seen in Fig. 5 at 16.6 Ϯ 1.1, 7.5 Ϯ 0.5, and 5.7 Ϯ 0.1 kDa, respectively (n ϭ 4 for these data and all those given below). Scanning densitometry revealed that only the 16.6-kDa peptide showed a significant change in phosphorylation with isoproterenol treatment (with pp175 from isoproterenol-treated acini, densities of the 16.6-, 7.5-, and 5.7-kDa bands determined after 6 h of digestion with V8 protease were 3.6 Ϯ 0.9, 0.97 Ϯ 0.10, and 1.21 Ϯ 0.16 times their levels without isoproterenol treatment, respectively; after 12 h of digestion these ratios were 2.74 Ϯ 0.48, 0.82 Ϯ 0.09, and 0.79 Ϯ 0.07, respectively).
On average little difference was found between the phosphorylation patterns observed after 6 and 12 h of V8 protease digestion. The density of labeling of the 7.5-kDa band did, however, decrease significantly between 6 and 12 h of protease treatment (20 Ϯ 5% and 31 Ϯ 9% decreases were found in digests from control and isoproterenol-treated cells, respectively), presumably indicating continued slow digestion of this peptide by V8 protease. No other significant increase or decrease in labeling with time of protease treatment was observed. In particular, paired t tests provided no evidence for a systematic shift of 32 P i from the 7.5-kDa to the 5.7-kDa peptide with time. This observation argues against the possibility that the latter peptide may be a digestion product of the former. Since all of the labeled protein at 175 kDa is recognized by our anti-Na ϩ -K ϩ -2Cl Ϫ cotransporter antibody (see above), all three of these labeled peptides are presumably associated with the transporter. Thus the results illustrated in Fig. 5 are consistent with the presence of at least three phosphorylation sites on pp175, only one of which is phosphorylated in response to isoproterenol treatment.
Effects of Other Stimuli on Phosphorylation of pp175-The effects of acinar treatment with AlF 4 Ϫ , calyculin A, and hypertonic medium on phosphorylation of pp175 are illustrated in Fig. 5. Interestingly, although treatment with the concentrations of AlF 4 Ϫ and calyculin A used in Fig. 6 both yield comparable up-regulations of cotransport activity to that produced by FIG. 3. Immunoprecipitation of Na ϩ -dependent bumetanide binding activity by antiserum against the parotid bumetanide binding protein. A lipid-stabilized Triton extract of the particulate fraction of rat parotid acini was prepared and subjected to immunoprecipitation using protein G-Sepharose beads preincubated with nonimmune sheep serum or immune serum raised against the rabbit parotid bumetanide binding protein (see "Experimental Procedures"). The Na ϩdependent component of [ 3 H]bumetanide binding (see "Experimental Procedures") in the lipid-stabilized Triton extract before immunoprecipitation (Control) and in the supernatant remaining after immunoprecipitation with immune (I) or nonimmune (N) serum are shown. The results are the means Ϯ S.E. of three independent experiments. The data from each experiment were normalized to the binding observed in the lipid-stabilized extract before immunoprecipitation (5.28 Ϯ 0.75 pmol/mg protein, n ϭ 3); the data were also corrected for the dilution arising from addition of protein G-Sepharose beads. Ϫ produces significant phosphorylation of pp175. This result is all the more surprising in view of the fact that calyculin A is a well-known inhibitor of protein phosphatases (19). Hypertonic shrinkage of acini with 80 mM sucrose, which has also been shown to upregulate cotransport activity in rat parotid acini (ϳ3 fold), 1 likewise results in no significant change in the phosphorylation level of pp175.
In the shark rectal gland experimental maneuvers that reduce intracellular chloride concentration have been shown to result in up-regulation of Na ϩ -K ϩ -2Cl Ϫ cotransport activity and phosphorylation of the Na ϩ -K ϩ -2Cl Ϫ cotransport protein (11). However, as also illustrated in Fig. 6, switching rat parotid acini to low chloride medium has no effect on the phosphorylation of pp175.
Recent work in our laboratory (9) has demonstrated a dramatic up-regulation of rat parotid Na ϩ -K ϩ -2Cl Ϫ cotransport activity by muscarinic stimulation (ϳ15-fold and ϳ25-fold after 30 s of stimulation with 1 M and 10 M of the muscarinic agonist carbachol, respectively). In Fig. 7 we show that this up-regulation is paralleled by increased phosphorylation of pp175. However, in additional experiments (9) we have shown that the up-regulatory effect of carbachol on the cotransporter can be duplicated by the microsomal Ca 2ϩ -ATPase inhibitor thapsigargin, which raises intracellular calcium concentration to levels comparable with that seen with carbachol, but without interacting with plasma membrane receptors, and without activating protein kinase C. We illustrate in Fig. 7 that treatment of acini with 1 M thapsigargin under conditions that yield an up-regulation of cotransport activity comparable with that produced by 1 M carbachol (Ref. 9 and data not shown), results in no significant phosphorylation of pp175. In addition, treatment of acini with the active phorbol ester PMA to activate protein kinase C yields no phosphorylation of pp175 in the presence or absence of thapsigargin (Fig. 7). DISCUSSION In this paper we identify a 175-kDa membrane phosphoprotein (pp175) in the rat parotid whose properties correlate well with the Na ϩ -K ϩ -2Cl Ϫ cotransporter characterized functionally in this gland (6) and with Na ϩ -K ϩ -2Cl Ϫ cotransporters previously identified biochemically in salivary glands (15) and other tissues (11, 20 -23). pp175 was the only membrane protein whose phosphorylation state was conspicuously altered after a brief (45-s) exposure to the cAMP-mobilizing secretagogue isoproterenol (Fig. 1). We have previously demonstrated that isoproterenol treatment results in a substantial up-regulation of Na ϩ -K ϩ -2Cl Ϫ cotransport activity in the rat parotid and provided strong evidence that this was due to a phosphorylation event mediated by protein kinase A (6). Consistent with the identification of pp175 as the Na ϩ -K ϩ -2Cl Ϫ cotransporter, the half-maximal effect of isoproterenol on phosphorylation of pp175 (ϳ20 nM; Fig. 2A) was in excellent agreement with its half-maximal effect on cotransport activity (6). Phosphopeptide mapping provided evidence for three phosphorylation sites on pp175 (Fig. 5), only one of which was labeled in response to isoproterenol treatment.
Increased phosphorylation of pp175 was also seen following acinar treatment with a permeant cAMP analogue and with forskolin ( Fig. 2B), conditions that have also been shown to up-regulate the cotransporter, presumably by the same mechanism as isoproterenol (6). In addition, pp175 was the only phosphoprotein immunoprecipitated by an antibody raised against the rabbit parotid Na ϩ -K ϩ -2Cl Ϫ cotransporter (Fig. 4). This antibody also quantitatively immunoprecipitated sodiumdependent bumetanide binding activity from a detergent extract of the rat parotid (Fig. 3) consistent with the expected properties of an anti-cotransporter antibody. Taken together with its molecular weight, which is in the expected range (150,000 -195,000) of previously identified Na ϩ -K ϩ -2Cl Ϫ cotransporters (15, 20 -23), the above results provide convincing evidence that pp175 is the Na ϩ -K ϩ -2Cl Ϫ cotransporter of the rat parotid.
The effects of cAMP-dependent secretagogues on the phosphorylation state of Na ϩ -K ϩ -2Cl Ϫ cotransporters recently identified in the shark rectal gland (Ref. 11; a 195-kDa phosphoprotein) and the avian salt gland (Ref. 23; a 170-kDa phosphoprotein) have also been studied. Consistent with the results presented here, in both these tissues a strong correlation between apparent up-regulation of transport activity and cotransporter phosphorylation was observed (11,23).
In their study of the phosphorylation of the shark rectal gland Na ϩ -K ϩ -2Cl Ϫ cotransporter, Lytle and Forbush (11) also showed that transport up-regulation by osmotic shrinkage is FIG. 6. Effects of AlF 4 ؊ , calyculin A, hypertonic shrinkage, and low Cl Ϫ medium on the phosphorylation of pp175. 32 P i -labeled acini were exposed to the stimuli indicated below, and the resulting phosphorylation of pp175 was quantitated and normalized as described in the caption to Fig. 2 7. Effects of carbachol, thapsigargin, and PMA on the phosphorylation of pp175. 32 P i -labeled acini were exposed to the stimuli indicated, and the resulting phosphorylation of pp175 was quantitated and normalized as described in the caption to Fig. 2. Acini were exposed to carbachol (CCh) for 40 s and to thapsigargin (ThG) and/or PMA for 120 s. The results shown are the means Ϯ S.E. of three or more independent experiments. accompanied by parallel increases in transporter phosphorylation. In addition, they showed that maneuvers that decrease intracellular chloride concentration in this tissue result in increased cotransporter phosphorylation. This latter observation is consistent with previous suggestions from this group that the decreased intracellular chloride concentration that accompanies secretion by the gland may itself play a role in the activation of the cotransporter (10,11,22). In the avian salt gland Torchia et al. (12,23) have demonstrated that Ca 2ϩ -mobilizing secretagogues also result in cotransporter phosphorylation. This is apparently due to the combined effect of increased intracellular calcium concentration and activation of protein kinase C, since it can be mimicked by the application of a Ca 2ϩ ionophore plus an active phorbol ester but not by either of these treatments alone (12). Treatment with the protein phosphatase inhibitor okadaic acid also resulted in cotransporter phosphorylation in the avian salt gland (12).
The above results from the shark rectal gland and avian salt gland are consistent with the hypothesis that direct phosphorylation, possibly at different sites by different stimuli, plays a central role in the regulation of Na ϩ -K ϩ -2Cl Ϫ cotransport activity in these tissues. However, the situation is clearly more complex in the rat parotid. Although the up-regulation of transport activity seen with isoproterenol and AlF 4 Ϫ (both ϳ6-fold) correlates well with the phosphorylation of pp175 induced by these agents (Figs. 2 and 6), this is not the case for treatment with the protein phosphatase inhibitor calyculin A or hypertonic shrinkage (Fig. 6). Despite the fact that both these latter treatments result in significant up-regulation of the cotransporter (ϳ6-fold and ϳ3-fold, respectively; Ref. 8 and Footnote 1), neither causes significant phosphorylation of pp175. Furthermore, although acinar treatment with the muscarinic agonist carbachol results in a dose-dependent phosphorylation of pp175 (Fig. 7), no phosphorylation is produced by the Ca 2ϩmobilizing agent thapsigargin, which is able to fully mimic the dramatic up-regulatory effect of carbachol on transport activity (ϳ15-fold for treatment with either 1 M carbachol or 1 M thapsigargin; Ref. 9). Moreover, treatment of acini with the active phorbol ester PMA to activate protein kinase C yielded no phosphorylation of pp175 in the presence or absence of thapsigargin (Fig. 7). Taken together, these latter results indicate that the phosphorylation of pp175 seen with muscarinic stimulation is not required for up-regulation of cotransport activity in the rat parotid and, in addition, suggest that this phosphorylation is not due to protein kinase C.
Finally, two observations made here indicate that the increase in cotransporter phosphorylation associated with decreased intracellular chloride concentration in the shark rectal gland (see above) is not seen in the rat parotid: (i) increased phosphorylation of pp175 is not observed after suspension of acini in low chloride medium (Fig. 6), and (ii) increased phosphorylation of pp175 is not observed after thapsigargin treatment (Fig. 6), which is expected to lead to a secretion-induced decrease in intracellular chloride concentration similar to that observed with muscarinic stimulation (4).
As already indicated, the experimental evidence available to date suggests that the up-regulations of the parotid Na ϩ -K ϩ -2Cl Ϫ cotransporter by treatment with ␤-adrenergic agonists, muscarinic agonists, hypertonic shrinkage, AlF 4 Ϫ , and calyculin A, all occur via different intracellular mechanisms. Briefly stated, this conclusion is based on the observations that the effects of ␤-adrenergic and muscarinic stimulation are secondary to increased intracellular levels of cAMP (6) and Ca 2ϩ (9), respectively, while the effects of hypertonic shrinkage, AlF 4 Ϫ , and calyculin A are apparently independent of both of these intracellular messengers (8). 1 The effects of these latter three stimuli on the cotransporter can, however, be distinguished by their sensitivities to inhibition by the compound K252a (K1 ⁄2 ϳ0.6 M, ϳ20 M, and Ͼ Ͼ20 M, respectively; Ref. 8 and Footnote 1). It is nevertheless always possible that the effects of some of these stimuli may be related. For example, two stimuli may act at different steps in the same up-regulatory pathway, resulting in the apparent differences discussed above.
We also considered the possibility that agents which resulted in an up-regulation of cotransport activity without a concomitant increase in the phosphorylation of pp175 might be acting via phosphorylation of another membrane-associated protein.
However, close examination of autoradiographs of Triton extracts from cells treated with thapsigargin, hypertonic shrinkage, or calyculin A did not reveal changes in the phosphorylation pattern of proteins at any molecular weight.
The results presented here support the hypothesis that the up-regulations of the rat parotid Na ϩ -K ϩ -2Cl Ϫ cotransporter by ␤-adrenergic stimulation and AlF 4 Ϫ treatment are due to direct phosphorylation of the transporter itself, whereas other mechanisms are clearly involved in the up-regulatory effects of muscarinic stimulation, hypertonic shrinkage, and calyculin A treatment.