INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE SODIUM/HYDROGEN EXCHANGER (NHE) is the major transporter of sodium across the luminal membrane of the proximal tubule (1). Six NHE isoforms are found in mammals, and all are expressed in renal tissue, with the exception of the NHE5 isoform, which is found in the central nervous system, testes, spleen, and skeletal muscle (1). The NHE3 isoform is the only NHE isoform expressed in rat renal brush-border membranes (BBM). Therefore, it is responsible for the amiloride-sensitive Na⁺ transport in BBM vesicles (BBMV) (2-4, 6, 7, 22, 41). NHE3 activity is regulated by phosphorylation/dephosphorylation processes and membrane recycling in intact cells (14, 23, 42, 43). NHE3 can also be regulated by G proteins independent of cytoplasmic second messengers; however, the G protein subunits involved in this regulation are not known (5, 10). Therefore, the primary aim of these experiments was to study G protein subunit regulation of NHE3 activity in renal BBMV in a system devoid of cytoplasmic components and second messengers.

Several hormones, e.g., parathyroid hormone, dopamine, angiotensin II, and norepinephrine, have been shown to regulate luminal NHE3 activity via their receptors, which are all coupled to heterotrimeric G proteins. Dopamine receptors of the D₁-like family, found in brush-border and basolateral membranes of renal proximal tubules (29) and coupled to G_s, inhibit NHE3 activity by increasing protein kinase A (PKA) activity (10-12). Parathyroid hormone, another receptor linked to G_s, inhibits NHE3 activity via PKA and protein kinase C (PKC) in opossum kidney and other cells (18, 39). Conversely, ₂-adrenergic and angiotensin II receptors, found in BBM of renal proximal tubules (27, 36) and coupled to G_i, stimulate NHE3 activity by decreasing cAMP production (8) and/or by increasing products of cytochrome P-450 arachidonic acid metabolism (27). However, in isolated renal proximal tubular BBM, G protein-coupled receptors continue to regulate NHE3 activity in the absence of cytosolic components and in the presence of inhibitors of adenylyl cyclase, PKA and PKC (5, 10). In addition, 5'-O-(3-thiotriphosphate) (GTPS) inhibits NHE3 activity in BBMV, presumably by stimulating G proteins (5, 10). These heterotrimeric G proteins are composed of G, , and  subunits, where the  subunits bind and hydrolyze GTP, whereas the  and  subunits exist as a tightly bound dimer. The G subunit and / dimer directly regulate effector proteins, including enzymes and ion channels (25, 28). Some effectors, e.g., adenylyl cyclase, are regulated by both the G protein  subunits and / dimers (37). G protein subunits may also regulate NHE3 activity in renal BBM independent of second messengers (5, 10, 32); however, the G protein subunit mediating this regulation is not known. Therefore, a secondary aim of this study is to determine the role of the G protein subunits _s, _i, and / in the second messenger-independent regulation of NHE3 activity in renal BBM by D₁-like receptors.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Preparation of renal BBM. Male Wistar-Kyoto (WKY) rats (Taconic, Germantown, NY), 9-16 wk old, fed on regular Purina rat chow diet were anesthetized with an intraperitoneal injection of pentobarbital sodium (50 mg/kg body wt), and the kidneys were harvested. We wanted to measure the activity of NHE3, the NHE isoform expressed in BBMV (2, 4, 41). BBMVs were isolated from the outer two-thirds of the cortex, which was separated from the medulla to avoid inclusion of medullary rays and the medullary thick ascending limb of Henle. These maneuvers ensured isolation of nephrons that do not express NHE1 (3, 4, 6, 7, 22). The presence of NHE1 in basolateral membranes of deeper nephrons could contaminate the BBM preparation and thus confound interpretation of our results (22). These BBMs were devoid of NHE1 protein, as determined by immunoblotting (data not shown), in agreement with other reports (3, 22). BBMs were isolated by MnCl₂ precipitation and differential centrifugation, as previously reported (10, 12, 31). The BBMs were suspended in (mM) 150 KCl with 25 2-(N-morpholino)ethanesulfonic acid (MES) and adjusted to pH 5.5 with 4 KOH ("inside" buffer). The BBMs spontaneously form vesicles while incubating on ice for 60 min. Subsequently, depending on the experiment, drugs, or their vehicle, and antibodies, or their appropriate control (heat-denatured antibodies to assure the same sodium and protein concentrations among groups), were added to the BBMs.

Measurement of NHE3 activity. NHE3 activity was assayed by measuring the 3-s 100 µM 5-(N-methyl-N-isobutyl)-amiloride (MIA)-sensitive ²²Na⁺ uptake (difference between ²²Na⁺ uptake in the absence of amiloride or its analogs and ²²Na⁺ uptake in presence of amiloride or its analogs) as described (10-12, 21, 30-32, 39). The validity and reliability of this assay have been well established in many laboratories, including our own (5, 10-12, 27, 32, 39). In agreement with Wu et al. (41), NHE activity in BBMV was due to NHE3 isoform. NHE activity in our BBMV preparations was resistant to 5-(N-ethyl-N-ispopropyl)-amiloride (EIPA); NHE1 is 100 times more sensitive to the inhibitory effect of EIPA, relative to its effect on the NHE3 isoform (19, 30). Uptake of ²²Na⁺ into BBMV was measured at 24°C by using the Millipore rapid filtration technique with 0.65-µm nitrocellulose filters (10-12). The BBMV (after vesicle formation) were incubated with receptor agonists/drugs for 30 min before ²²Na⁺ uptake. When antibodies were used, they were added to BBM before vesicle formation (i.e., 90 min before ²²Na⁺ uptake). When drugs were used with the antibodies, the BBM were exposed to the antibodies 60 min before vesicle formation, and the receptor agonists/drugs were added after vesicle formation 30 min before ²²Na⁺ uptake, as stated previously. ²²Na⁺ uptake was then determined by mixing 20 µl of the membrane vesicle suspension (150-350 µg protein) and 30 µl of uptake medium and incubating for 3 s at 24°C. The final concentration was (in mM) of 142 KCl, 14.7 KOH, 10 MES, 9 HEPES, and 1 NaCl (containing 0.1 to 0.2 µCi of ²²Na⁺), pH 7.5. Three seconds after mixing, transport was halted by adding 2 ml of ice-cold stop buffer (in mM) 150 KCl, 15 HEPES, 0.1 MIA, pH 7.5. The studies were performed in the presence of an outwardly directed pH gradient (pH_in = 5.5, pH_out = 7.5) and an inwardly directed Na⁺ gradient ([Na⁺]_out = 1 mM, [Na⁺]_in = 0 mM). Drug or antibodies that required access to the interior of the vesicles were added to the membrane suspensions before vesicle formation, as described above and previously reported (6, 11, 12, 39).

Pertussis toxin treatment of BBM. After the isolation of BBM, the membranes were suspended in an incubation buffer containing (in mM) 100 Tris(hydroxymethyl)-aminomethane hydrochloride, 40 dithiothreitol, 20 thymidine, 5 ethylenediaminetetraacetic acid, 1 MgCl₂, 1 adenosine-5'-triphosphate, and 1 -nicotinamide adenine dinucleotide (NAD), pH 7.4. The suspension was centrifuged at 18,000 rpm for 20 min and then resuspended in the above buffer. Pertussis toxin (PTX) was preincubated in incubation buffer for 10 min at 30°C to activate PTX (16); buffers without PTX were treated similarly as controls. Subsequently, the BBMs were incubated in incubation buffer with or without PTX (25 µg PTX/mg protein) at 30°C for 1 h to ADP ribosylate G_i (16). The incubation buffer- and PTX-treated membranes were centrifuged at 18,000 rpm for 20 min. The pellet was washed and resuspended in a volume of inside buffer (equal to 100 times the pellet volume) and centrifuged at 18,000 rpm for 20 min (2×). The resulting pellet was resuspended with inside buffer to an approximate protein concentration of 1 mg/ml. Protein concentrations were determined by the Lowry method. The BBMV were then used for transport studies as described in the preceding section.

ADP ribosylation studies. The BBMs were treated in a manner similar to that described for PTX, except that ³²P-labeled NAD was used (16). The BBMs were solubilized in Laemmli Tris-glycine and SDS-PAGE denaturing, reducing buffer (catalog #LC2675, Novex, San Diego, CA) and boiled for 3 min. The suspensions of solubilized BBMV with and without PTX were loaded on a 12% Tris-glycine gel and electrophoresed at 100 V for 120 min.

Immunoprecipitation studies. Three types of experiments were performed. In the first set of experiments, BBM were treated with GTPS (10³ M) or an equal volume of incubation buffer containing (in mM) 150 NaCl, 10 MgCl₂, and 20 Tris · HCl, pH 7.5 for 30 min at room temperature and then centrifuged for 20 min at 18,000 rpm, 4°C. In the second set of experiments, plasma membranes or immortalized renal proximal tubules cells from WKY rats (40) were incubated with vehicle, GTPS (10³M), or guanosine 5'-O-(2-thiodiphosphate) (GDPS; 10³M) for 30 min at 37°C. In the third set of experiments, immortalized renal proximal tubules cells were incubated with vehicle or fenoldopam (5 × 10⁶ M) for 10 min at 37°C. The cells were lysed with ice-cold lysis buffer [PBS with 1% Nonidet P-40, 0.1% SDS, 0.5% deoxycholate, 1 mM phenylmethylsulfonyl fluoride, 10 µg/ml aprotinin, and 10 µg/ml leupeptin (1 mM sodium vanadate was added in the immortalized proximal tubule studies)] and centrifuged at 15,000 rpm for 20 min. The lysates were then incubated with 4 µl of antibody against the protein of interest or 4 µl of normal rabbit IgG (as a control) for 1 h at 4°C. Thereafter, 20 µl of 25% protein A agarose (Santa Cruz Biotech, Santa Cruz, CA) were added and incubated overnight on a rocking platform at 4°C. The immunoprecipitates were pelleted, washed with lysis buffer, suspended in sample buffer, and boiled for 10 min. In some studies using anti-NHE3 antibody for immunoprecipitation and anti-G_s for immunoblotting, 2-mercaptoethylamine (100 mM) was used instead of -mercaptoethanol. In this experimental setup, the samples were incubated for 120 min at 24°C to cleave IgG into 75-kDa fragments. This allowed better visualization of proteins with molecular sizes of ~50 kDa, e.g., G_s; similar results were obtained using 150 mM 2-mercaptoethylamine incubated for 30 min at 37°C.

Western blotting. The proteins were separated by electrophoresis on 7.5% SDS-PAGE and then electrophoretically transferred onto nitrocellulose membranes. The transblots were probed with the indicated antibodies and detected by hydrogen peroxidase-labeled secondary antibody and chemiluminescence detection reagents. Quantification of the immunoblots was performed; the density of the area of each immunoblot was quantified using Quantiscan (Biosoft, Ferguson, MO).

Statistical analysis. Data are expressed as means ± SE. Differences within groups were analyzed by analysis of variance for repeated measures, followed by Scheffé's test; paired t-test was used when only two groups were compared. Differences among groups were analyzed by one-way analysis of variance, followed by Scheffé's test; if only two groups were compared, unpaired t-test was used.

Drugs. The chemicals used were MIA, EIPA, SKF-81297, and UK-14304 (Research Biochemicals, Natick, MA); 5,8,11,14-eicosatetraynoic acid (ETYA; Biomol Research Laboratories, Plymouth Meeting, PA); PTX, GTPS, and G_s standard (Calbiochem, La Jolla, CA); 2-mercaptoethylamine-HCl (Pierce, Rockford, IL). All other reagents were bought from Sigma (St. Louis, MO). Anti-NHE1 monoclonal antibodies were purchased from Chemicon, Temecula, CA. Anti-NHE3 antibodies were raised against synthetic oligopeptides from the amino acid sequence of rat NHE3 (amino acids 633-646; Research Genetics, Huntsville, AL) (2).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

NHE3 activity in BBMV. The NHE isoform responsible for a particular NHE activity can be determined by its sensitivity to amiloride analogs (30). NHE3 is relatively insensitive to EIPA, whereby the EIPA-sensitive isotypes are fully inhibited by 5 × 10⁷ M EIPA (30). NHE3 is the only known NHE isoform found in the BBM of renal proximal tubules that is relatively insensitive to EIPA (2-4, 6, 7, 22, 41). Therefore, we determined the effect of varying concentrations of EIPA (10⁹-10⁵ M) on ²²Na⁺ uptake into BBMV. ²²Na⁺ uptake into BBMV was not affected by 10⁶ M EIPA, indicating that all of the measured NHE activity in BBM was due to NHE3 and not to NHE1 (30) or other EIPA-sensitive NHE isotypes.

²²Na⁺ uptake at 1-2 h was assumed to represent equilibrium values and also served as an index of vesicle size (21). In the current and previous reports, no differences among drug/antibody or vehicle-treated membranes at 1-2 h were noted, indicating that vesicle sizes were similar among the groups.

Basal NHE3 activity in BBMV was 1.95 ± 0.16 nmol ²²Na · mg protein¹ · min¹. Pretreatment of BBM with dopamine or D₁-like agonists inhibited the amiloride-sensitive Na⁺ transport (3-s uptake) up to 70% in a dose- and time-dependent manner (Fig. 1A, Refs. 10, 12). In contrast, receptor ligands had no effect on EIPA or MIA-insensitive Na⁺ transport. Previous experiments had established that dopaminergic inhibition of amiloride-sensitive Na⁺ transport cannot be explained by an increase of the amiloride-insensitive Na⁺ transport rate, collapse of the proton gradient driving Na⁺/H⁺ exchange, or a decrease in vesicle size (9, 12). The amiloride-insensitive Na⁺ transport includes all contributions of Na⁺ movement through Na⁺ cotransporters, such as Na²⁺-glucose transporter (SGLT)-1, SGLT-2, sodium-amino acid cotransporters, and sodium/phosphate cotransporters. The amiloride-insensitive Na⁺ transport was unchanged by pretreatment with dopamine or D₁-like agonist under the experimental conditions of Figs. 1, A and B, 2, 3, 4, A-C (absence of phosphate, glucose, or amino acids in the incubation buffer). Therefore, the decrease in the amiloride-sensitive Na⁺ transport rate reflects inhibition of NHE3. This inhibition in BBM is seen in the absence of added ATP and GTP.

View larger version (2018K): [in this window] [in a new window]   Fig. 1.   A: best-fit analysis of dose response inhibition by dopamine and D1-like receptor agonists, SKF-81297 and fenoldopam, on sodium-hydrogen exchanger (NHE) 3 activity in brush-border membrane vesicles (BBMV) isolated from rat renal cortex (n = 3-7/drug concentration). NHE3 activity was significantly inhibited by all concentrations, except for fenoldopam at 5 × 107 M. B: inhibitory effect of SKF-81297 (5 × 106 M) on NHE3 activity in BBMV incubated with and without antibodies for the Gs subunit (1:100 dilution). * P < 0.05 vs. others, # P < 0.05 vs. D1-like agonist, ANOVA, Scheffé's test.

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Effect of pertussis toxin (PTX), D1-like receptor agonist SKF-81297 (SKF; 5 × 106 M), and the 2-adrenergic agonist UK-14304 (UK; 5 × 1010M) on NHE3 activity in BBMV. * P < 0.05 vs. respective controls, # P < 0.05 vs. PTX, $ P < 0.05 SKF vs. UK + SKF, ANOVA, Scheffé's test.

View larger version (28K): [in this window] [in a new window]   Fig. 3.   Effect of D1-like receptor agonist SKF-81297 (5 × 106 M) and 2-adrenergic agonist UK-14304 (5 × 1010 M) and anti-Gi antibody (1:100 dilution) on NHE3 activity in BBMV. * P < 0.05 vs. control, # P < 0.05 vs. D1-like agonist, ANOVA, Scheffé's test.

View larger version (221618K): [in this window] [in a new window]   Fig. 4.   A: effect of anti-Gcommon antibodies and SKF-81297 on NHE3 activity. * P < 0.05 vs. others, # P < 0.05 vs. D1-like agonist or D1-like agonist + anti-Gcommon antibodies, ANOVA, Scheffé's test. B: effect of anti-common antibodies on 2-adrenergic agonist UK-14304-mediated reversal of D1-like receptor agonist-mediated inhibition of NHE3 activity. * P < 0.05 vs. D1-like receptor agonist + 2-adrenergic agonist, ANOVA, Scheffé's test. C: effect of anti-common antibodies on D1-like receptor agonist-mediated inhibition of NHE3 activity in PTX-treated BBMV. * P < 0.05 vs. control, # P < 0.05 vs. D1-like agonist alone, ANOVA, Scheffé's test.

G_s mediates D₁-like agonist inhibition of NHE3 activity in BBMV. Although, NHE3 activity is regulated by phosphorylation/dephosphorylation processes and membrane recycling (14, 23, 42, 43), it can be regulated also by G proteins, independent of cytoplasmic second messengers (5, 10). In agreement with previous studies, dopamine and two different D₁-like agonists, SKF-81297 and fenoldopam, decreased NHE3 activity in BBMV (Fig. 1A), presumably independent of cAMP (10, 11). Other investigators have shown that dopamine does not stimulate cAMP production in BBMV unless ATP is added (32). Moreover, inhibition of adenylyl cyclase, PKA, and PKC activities did not prevent the inhibitory action of fenoldopam on NHE3 activity in rat BBMV (10). Because dopamine can stimulate phospholipase A₂ activity in BBM (32), we also studied the effect of ETYA on NHE3 activity at a concentration (10⁴ M) known to inhibit phospholipase A₂, lipoxygenase, and monoxygenase activities (38). In our study, the inhibitory effect of SKF-81297 was not affected by ETYA (data not shown). These results demonstrate that the inhibitory effect of D₁-like agonists on NHE3 activity in BBMV was not caused by PKA or PKC activation or by eicosanoids.

GTPS incorporated inside BBMVs inhibited NHE3 activity (5, 10). Thus NHE3 activity can be inhibited by G proteins independent of cytoplasmic second messengers (5, 10). In the current study, we confirm the ability of GTPS (10³ M) to decrease NHE3 activity (control amiloride-sensitive ²²Na⁺ uptake = 2.99 ± 0.24 vs. 2.06 ± 0.16 nmol Na⁺ · mg protein¹ · min¹ with GTPS, P < 0.05, paired t-test). In contrast, GTPS added to the "outside" of BBMV did not significantly affect NHE3 activity (2.52 ± 0.24 nmol ²²Na⁺ · mg protein¹ · min¹, P > 0.05, t-test) (9).

The inhibition of NHE3 activity by the D₁-like agonist SKF-81297 was caused in part by G_s, because antibodies directed against this G protein subunit (1:100 dilution) partially reversed the D₁-like inhibition (SKF-81297 + heat-denatured anti-G_s antibody = 59 ± 5%, n = 12, SKF-81295 + anti-G_s antibody = 36 ± 4%, n = 8) of NHE3 activity by ~39 ± 5%; anti-G_s antibody by itself (control) was without effect (Fig. 1B). Heat-denatured anti-G_s antibody had a minimal effect on NHE3 activity when compared with vehicle treatment alone (data not shown). All groups were compared with a vehicle control, as well as a control consisting of heat-denatured anti-G_s antibodies in the place of active antibodies, to ensure ion and protein concentrations are the same between groups. The anti-G_s antibody used in this study has been shown to be specific (33). In the current report, the anti-G_s antibody recognized a recombinant 45-kDa G_s standard (not shown) as well as 45- and 52-kDa proteins in immunoblots of BBM (see, for example, Fig. 6A). The ability of anti-G_s antibody to partially attenuate the inhibitory effect of SKF-81297 on NHE3 activity was caused by anti-G_s activity, because heat-denatured anti-G_s antibody (control studies) had no effect on D₁-like inhibition of NHE3 activity (Fig. 1B). The inability of anti-G_s antibody to completely block the inhibitory action of SKF-81297 may be related to the fact that only a limited amount of antibody (1:100) can be "loaded" inside the vesicle; 1,000-fold dilution of anti-G_s antibody had no effect, whereas 1:10 dilution resulted in variable uptake of ²²Na⁺ (data not shown).

G_i does not mediate the ability of an ₂-adrenergic agonist to reverse the D₁-like inhibition of NHE3 activity in BBMV. Norepinephrine and angiotensin II oppose D₁-like agonist inhibition of NHE3 activity in the luminal membrane of renal proximal tubule cells caused, in part, by inhibition of adenylyl cyclase activity (8). UK-14304 by itself had no effect (Fig. 2), probably because NHE3 activity must be inhibited first to demonstrate a stimulatory effect (8). However, in BBMV where cAMP cannot be generated, UK-14304, an ₂-adrenergic agonist, was able to partially reverse (22-55%) the D₁-like agonist inhibition of NHE3 activity (Fig. 3). The G_i subunit did not appear to be involved in the current experimental setup, because anti-G_i-3 antibodies (1:100 dilution) did not affect the ability of UK-14304 to counteract the D₁-like agonist inhibition of NHE3 activity (Fig. 3). G_i-3 subunit is the predominant G_i isoform expressed in renal proximal tubules (35). The anti-G_i-3 antibody used in this study has been shown to be specific (34). In the current report, the anti-G_i-3 antibody recognized a 41-kDa protein that corresponded with the protein AD³²P ribosylated by PTX in BBM (data not shown). Lower dilutions of anti-G_i antibody (1:10 dilution), as with the lower dilutions of anti-G_s antibody, resulted in variable uptakes of ²²Na (data not shown). In addition, anti-G_i antibodies did not influence basal or D₁-like agonist inhibition of NHE3 activity (Fig. 3). All groups were compared with a vehicle control as well as a vehicle plus heat-denatured anti-G_i (1:100 dilution) antibody control to ensure that concentrations of ions and proteins are the same between groups.

The inability of anti-G_i-3 antibodies to affect NHE3 activity may have been due to the presence of other G_i subunit isoforms. Therefore, to prove that G_i subunits did not mediate the ₂-adrenergic agonist attenuation of D₁-like agonist inhibition of NHE3 activity, we performed additional studies in BBM treated with PTX, a toxin known to inactivate G_i through ADP ribosylation (16). Incubation of BBM with PTX at a concentration that ADP ribosylated a 41-kDa protein had no effect on basal NHE3 activity compared with the vehicle (control) or ₂-adrenergic agonist [UK-14304 (5 × 10¹⁰ M)]-treated group. The vehicle control groups were studied concurrently and consisted of BBMs incubated in the same buffers without PTX. PTX treatment did not affect basal NHE3 activity; it also did not alter any UK-14304 effect. Thus, in our studies, G_i did not influence NHE3 activity. However, after PTX treatment, there was no longer a difference in NHE3 activity between D₁-like receptor agonist treatment alone and D₁-like receptor agonist plus ₂-adrenergic receptor treatment (Fig. 2). This was not caused by a reduction in the stimulatory effect of the ₂-adrenergic agonist on NHE3 activity but, rather, by a diminution (43% ± 5%) in the inhibitory effect of the D₁-like agonist [SKF-81297 (5 × 10⁶ M)].

/ dimer mediates the ability of ₂-adrenergic agonist to reverse the D₁-like inhibition of NHE3 activity in BBMV. Because G_i did not mediate the "stimulatory" effect of ₂-adrenergic receptor activation on NHE3 activity in BBMV, we determined whether / subunits were involved in the ₂-adrenergic agonist antagonism of D₁-like agonist inhibition of NHE3 activity. Antibodies against _common slightly decreased NHE3 activity in the basal state (compared with the vehicle control or heat-denatured anti-_common antibody control) but had an insignificant effect on the inhibition of NHE3 activity by the D₁-like agonist (compared to D₁-like agonist alone, control group; Fig. 4A). All groups were compared with a vehicle control and to another control consisting of heat-denatured anti-_common (1:100 dilution) antibodies in place of nondenatured antibodies to ensure the same ion and protein concentrations between groups (Fig. 4, A-C). However, anti-_common antibodies (1:100 dilution) attenuated the ₂-adrenergic agonist UK-14304 (5 × 10¹⁰ M) reversal of the inhibitory action the D₁-like agonist SKF-81297 [5 × 10⁶ M, n = 4 (71 ± 5%), 5 × 10⁸ M, n = 2 (75% ± 5%)] (Fig. 4B). The anti-_common antibody used in these studies has been shown to be specific (13, 34). In the current report, the anti-_common antibody recognized a 35-kDa protein in immunoblots of BBM (see below). / Subunits were also involved in the diminished inhibitory effect of the D₁-like agonist on NHE3 activity caused by PTX, because anti-_common antibodies enhanced the inhibitory effect of SKF-81297 after PTX treatment of BBM (Fig. 4C).

To determine the mechanism of the /-dependent effect of PTX treatment of BBM on D₁ receptor-mediated NHE3 inhibition, the G_s, G_i, and  subunits were immunoprecipitated with their respective antibodies from vehicle- and PTX-treated BBM. The immunoprecipitates were electrophoresed and then immunoblotted using the antibody for  subunits (Fig. 5). There was no difference in the amount of  subunits (n = 5) immunoprecipitated in the control and PTX-treated BBM. PTX treatment tended to increase the amount of / dimers associated with G_i subunits (n = 5) and significantly decreased the amount of / dimers associated with G_s (n = 3).

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Effect of PTX treatment on G/ dimer coupling to Gi and Gs subunits. Antisera specific for the common (n = 5), Gi (n = 5), and Gs (n = 3) were used to immunoprecipitate the respective subunits from BBM (inset) treated with vehicle (lanes 2-4) or with PTX (25 µg/mg protein) (lanes 5-7) and immunoblotted with anti-common antibody. Lane 1, BBM (50 µg protein) immunoblotted with anti-common antibody (no immunoprecipitation); lanes 2 and 5, immunoprecipitation with anti-common antibody; lanes 3 and 6, immunoprecipitation with anti-Gi3 antibody; lanes 4 and 7, immunoprecipitation anti-Gs antibody. All blots were scanned and analyzed by densitometry as described in the METHODS. Compared with vehicle treatment, PTX treatment decreased amount of  subunits immunoprecipitated with the Gs subunit. * P < 0.05 vs. vehicle-treated BBM, ANOVA, Scheffé's test.

There is physical linkage between NHE3 and G proteins. Three sets of studies were performed to determine the linkage between NHE3 and G_s. In the first set of studies, renal BBMs were treated with vehicle, GTPS, or GDPS. The BBMs were then immunoprecipitated with anti-NHE3 antibody and immunoblotted with anti-G_s antibody. Figure 6A demonstrates two bands (45 and 52 kDa) that represent G_s observed after immunobloting rat renal BBM with the anti-G_s antibody. The amount of G_s that coimmunoprecipitated with NHE3 was increased by GTPS treatment [Fig. 6A, compare lanes 6 and 7, the GTPS-treated BBM, with lanes 2 and 3, the vehicle (control)-treated BBM]. In contrast, the inactive analog of GDP, GDPS (Fig. 6A, lanes 4 and 5), had no effect [Fig. 6A, compare lanes 2 and 3, the vehicle-treated BBM, (control), lanes 4 and 5, the GDPS-treated BBM (negative control), and lanes 6 and 7, the GTPS-treated BBM]. We have reported that GTPS but not GDPS enhanced the inhibitory effect of D₁-like agonists on NHE3 activity in BBMV (12). The densitometric analyses of three other G_s studies are shown in Fig. 6B.

View larger version (2419K): [in this window] [in a new window]   Fig. 6.   Regulated linkage between NHE3 and Gs. A: immunoblots of NHE3 and Gs. In the first set of studies, immunoprecipitation was performed using anti-NHE3 antibodies and immunoblotted with anti-Gs antibodies. BBMs from rat renal cortex were treated with vehicle (lanes 2 and 3), guanosine 5'-O-(2-thiodiphosphate) (GDPS; 103 M, lanes 4 and 5), or guanosine 5'-O-(3-thiotriphosphate) (GTPS; 103 M, lanes 6 and 7) for 30 min and immunoprecipitated with anti-NHE3 antibodies and immunoblotted with anti-Gs antibodies. Lane 1 is immunoblot for Gs showing 2 molecular sizes, 45 and 52 kDa. The densitometric analysis from 3 other studies are shown in B. In a second set of studies, immunoprecipitation was performed using anti-Gs antibodies and immunoblotted with anti-NHE3 antibodies. Membranes or cells from immortalized renal proximal tubule cells (lanes 11-13) were treated with vehicle alone (lane 11), GTPS (103 M; lane 12), or GDPS (103 M; lane 13). The densitometric analysis from 3 studies are shown in B. Linkage between NHE3 and Gs is specific, because no band of the appropriate size (85 kDa) was seen when immunoprecipitant was IgG (lane 8). Moreover, a band of the appropriate size (85 kDa) was seen when membranes were immunoprecipitated with anti-NHE3 antibody and then immunoblotted with anti-NHE3 antibody (lane 9); this band corresponded to NHE3 band seen after immunoblotting of renal tubular cell membranes with anti-NHE3 antibody (lane 10). In a third set of studies, immortalized renal proximal tubule cells were incubated for 10 min with vehicle alone (lane 15) or the D1-receptor agonist fenoldopam (5 × 106 M; lanes 14 and 16). The whole cell lysates were immunoprecipitated with anti-Gs antibodies (lanes 15 and 16) or with normal rabbit IgG (lane 14) and immunoblotted with anti-NHE3 antibodies (lanes 14-16). Densitometric analyses are shown in B. Specificity of immunoprecipitate is shown by absence of immunoblotable NHE3 using normal rabbit IgG as the immunoprecipitant (lane 14). B: densitometric analyses of effect of GTPS (103 M) and GDPS (103 M) on amount of NHE3 linked to Gs in renal BBM (n = 5/group). Effect of GTPS (103 M) (n = 5), GDPS (103 M) (n = 5), and fenoldopam (5 × 106M) (n = 3) on amount of NHE3 linked to Gs treatment of renal proximal tubule cells before membrane lysis. * P < 0.05 vs. respective controls, ANOVA for repeated measures, Scheffé's test or paired t-test when only 2 groups were compared (e.g., control vs. fenoldopam-treated cells).

In the second set of studies, cell membranes from immortalized renal proximal tubule cells were immunoprecipitated with anti-G_s antibody and then immunoblotted with anti-NHE3 antibody. The reversal of the immunoprecipitant/immunoblotting antibody (compared with the first set of experiments) was performed to demonstrate the specificity of G_s/NHE3 coupling. Incubation of cell membranes from immortalized proximal tubule cells of WKY rats with GTPS increased the amount of NHE3 linked to G_s (Fig. 6A, lane 12, the GTPS-treated membranes, vs. lane 11, the vehicle-treated membranes). The inactive analog of GDP, GDPS, did not affect the amount of NHE3 linked with G_s (Fig. 6A, lane 13, the GDPS-treated membranes, vs. lane 12, the GTPS-treated membranes, or lane 11, the vehicle-treated membranes). The densitometric analyses of three other studies are shown in Fig. 6B. The linkage between NHE3 and G_s was specific because no band of the appropriate size (85 kDa) was seen in the control using rabbit IgG as the immunoprecipitant (Fig. 6A, lane 8). However, in an additional control for specificity, a band of the appropriate size (85 kDa) was seen when the membranes were immunoprecipitated with anti-NHE3 antibody and then immunoblotted with anti-NHE3 antibody (Fig. 6A, lane 9); this band corresponded to the NHE3 band seen after immunoblotting renal proximal tubular cell membranes with anti-NHE3 antibody (Fig. 6A, lane 10). Differences in sizes between lanes 9, 12, and 16 (immunoprecipitation) from lane 10 (immunoblot) were small and may be due to the persistent partial association of immunocomplex fragments with the immunoprecipitated NHE3 protein.

In the third set of studies, immortalized renal proximal tubule cells were incubated with the D₁-like agonist fenoldopam for 10 min before cell lysis. These studies were performed to determine if G_s binds to NHE3 in the intact cell where cytoplasmic second messengers can be generated. The cell lysates were immunoprecipitated with anti-G_s antibody and then immunoblotted with anti-NHE3 antibody. Treatment of immortalized proximal tubule cells from WKY rats with the D₁-like agonist fenoldopam (5 × 10⁶ M for 10 min) increased the amount of NHE3 linked to G_s (Fig. 6A, lane 16, the fenoldopam-treated cells, vs. lane 15, the vehicle control-treated cells); the results were specific because no band of the appropriate size was seen when the lysates were immunoprecipitated with a rabbit IgG control instead of anti-G_s antibody (Fig. 6A, lane 14). The densitometric analyses of these studies are shown in Fig. 6B.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

NHE3 activity is regulated by phosphorylation/dephosphorylation processes and membrane recycling in intact cells (14, 23, 42, 43). NHE3 activity can also be regulated by G proteins independent of cytoplasmic second messengers, but the G protein subunits involved in this regulation have not been reported (5, 10, 17). We now demonstrate a physical linkage between NHE3 and G_s and that this coupling increases after activation of regulatory pathways known to inhibit NHE3 activity in both rat renal BBMV and in the intact renal proximal tubular cell. In addition, we show that NHE3 activity can be regulated also by G/ dimers.

Activation of G protein-coupled receptors, e.g., D₁-like receptors, by its agonist results in the mobilization of G protein subunits that may then interact with effectors, including enzymes (adenylyl cyclase, phospholipases) (37), ion channels (25), or transporter proteins (5, 10, 32). In keeping with the linkage of D₁-like receptors and adenylyl cyclase activation, the immunoprecipitation studies demonstrated that the G_s subunit couples with NHE3 in a regulatory manner in intact cells (12). PKA has been shown to inhibit NHE3 activity whether naturally or heterologously expressed in kidney and other cell lines (12, 18, 39). PKC has also been reported to decrease NHE3 activity (18). However, the inhibitory action of dopamine on renal proximal tubular luminal NHE3 activity is mediated by PKA but not by phospholipase C signal transduction products (11, 12, 32). Eicosanoids generated from cytochrome P-450 metabolism may also participate in the regulation of NHE activity in BBMV (27, 32). In the current study, the inhibitory effect of the D₁-like receptor on NHE3 activity in rat renal BBMV was not affected by ETYA (data not shown), an inhibitor of arachidonic acid metabolism. Dopamine can also inhibit NHE1 activity (26). However, expression and activity of NHE1, which is found in basolateral membranes, was not detected in our BBM preparation. There are no other NHE isoforms in BBMs of renal proximal tubules; indeed the NHE isoform responsible for sodium transport in the BBM of renal proximal tubules has been reported to be NHE3 (1, 2, 4, 41).

Dopamine, via D₁-like receptors, has also been shown to inhibit NHE activity in rat renal proximal tubules independently of PKA (10, 32) and PKC (10); this second messenger-independent regulation of NHE activity is G protein linked (10, 32). The current studies confirm and extend these observations. With the use of renal BBMVs that are devoid of cytoplasmic second messengers and adenosine triphosphate (10), we find that dopamine and two different D₁-like agonists inhibit NHE3 activity in a concentration-dependent manner. In the current report, we determined which G protein subunit is linked to the inhibitory effect of D₁-like receptors on NHE3 activity in BBM independent of cytoplasmic second messengers. Immunoprecipitation studies using proximal tubule cell membranes and rat renal BBM demonstrated that stimulation of D₁-like receptors or direct activation of the heterotrimeric G proteins with GTPS resulted in a robust increase in the amount of G_s associated with the NHE3. In addition, anti-G_s antibodies reduced the D₁-like receptor agonist-mediated NHE3 inhibition. The mechanism by which PTX treatment decreased the inhibitory effect of the D₁-like agonist on NHE3 activity is probably an indirect one. The stabilization of G_i// heterotrimers by PTX (16) may have limited the amount of / dimers available to form heterotrimers with G_s and thereby decreased its coupling to D₁-like receptors. At any rate, these studies provide concrete evidence of a direct regulatory role for G_s, independent of second messengers, in the D₁-like-mediated inhibition of NHE3 activity in rat renal BBMV.

The G_i subunit had no discernible role in the regulation of basal or D₁-like agonist-mediated inhibition of NHE3 activity in BBMV. The ₂-adrenergic agonist, by itself, did not affect NHE3 activity, an action that was not modified by PTX. However, we confirmed that ₂-adrenergic agonist stimulation counteracted the inhibitory effect of a D₁-like agonist on NHE3 activity in BBMV. The modest ₂-adrenergic receptor agonist-mediated reversal (22-55%) of the D₁-like receptor-mediated inhibition may be related to the lower expression of ₂-adrenergic receptors relative to D₁-like receptors in rat renal BBM [D₁-like receptor maximal binding (B_max) = 206 fmol/mg protein, ₂-adrenergic receptor B_max = 108 fmol/mg protein] (15, 36). UK-14304 exerts its actions via ₂-adrenergic receptors, and, under physiological conditions, the ₂-adrenergic receptor is linked to G_i (8). UK-14304 has been shown to stimulate NHE3 activity via ₂-adrenergic receptors/G_i in intact opossum kidney cells (8). Because ₂-adrenergic receptors are linked to G_i, the reversal of the inhibitory action of the D₁-like agonist should have been blocked by antibodies to G_i3; G_i3 is the most abundant G_i subunit isoform in renal proximal tubules (35). However, the antibodies to G_i-3 did not affect the ability of an ₂-adrenergic agonist to counteract the inhibitory effect of a D₁-agonist on NHE3 activity. The inability of anti-G_i-3 antibodies to affect NHE3 activity was not due to the presence of other isoforms of the G_i subunit, because PTX treatment of BBM did not affect basal NHE3 activity or alter any ₂-adrenergic agonist effect. Thus the PTX and the anti-G_i antibody studies suggest that G_i did not mediate the ability of the ₂-adrenergic agonist to attenuate the inhibitory action of D₁-like receptors on NHE3 activity in BBMV. Taken together, the failure of PTX and anti-G_i antibodies to reduce the ₂-adrenergic agonist effect suggests a model that includes a PTX-insensitive G protein. In light of the results obtained using the anti-_common antibody (discussed below), this pathway appears to involve release of / subunits, which preferentially couple an as yet unidentified heterotrimeric G protein.

Because activation of G_i could not explain the ability of the ₂-adrenergic receptor agonist to reverse the inhibitory effect of the D₁-like agonist on NHE3 activity in BBMV, we turned our attention to the G protein / subunits. In BBM, antibodies to _common alone had a small but significant effect on basal NHE3 activity, which suggests that / dimers have a tonic stimulatory effect on NHE3 activity. The lack of a substantial effect of anti-_common antibodies in the D₁-like receptor inhibition of NHE3 activity is likely due to the fact that maximum NHE3 inhibition by the D₁-like agonist has already been achieved. In addition, the inhibitory effect of G_s may normally predominate over the stimulatory effect of / dimers. However, antibodies to _common reversed the ability of the ₂-adrenergic agonist to attenuate the inhibitory effect of the D₁-like agonist. These data implicate / dimers in the effect of PTX on D₁-like receptor action. However, the effect of PTX seen here is difficult to rationalize according to present models of PTX action, in which it is seen as an uncoupler of receptors and their respective subclass of G proteins (16). Nevertheless, the functional effect of PTX treatment of BBMs and the reversal of this effect by anti- antibodies were highly reproducible in this experimental system. The most economical interpretation of these results would be that / dimers released from a pool of PTX-insensitive G proteins by the ₂-adrenergic agonist bind and stimulate NHE3 activity, opposing the effect of G_s. Moreover, this model is consistent with preliminary studies from our laboratory suggesting that NHE3 can physically couple to G/ (24).

The mechanism by which PTX treatment is able to counteract the inhibitory effect of D₁-like agonists on NHE3 function is more obscure. Certainly this must be an indirect effect, because PTX is unable to modify G_s directly. A clue to a possible mechanism emerged from the immunoprecipitation data of Fig. 5, where PTX treatment reduced the amount of G_s-associated /. It is likely that G_s and G_i couple to the same pool of / isoforms. Therefore, after PTX-induced stabilization of G_i// heterotrimers, the amount of / available for G_s// heterotrimer formation would diminish. Such a scenario would result in a decreased coupling G_s// heterotrimers with the D₁-like receptor. These uncoupled D₁-like receptors could no longer respond to agonist stimulation resulting in a decreased D₁-like agonist inhibition of NHE3 activity after PTX treatment. Anti-_common antibodies could have enhanced the D₁-like receptor inhibition of NHE3 activity after PTX treatment by removing the stimulatory effect of / dimers associated with NHE3 and thereby allowing an unopposed inhibitory action of G_s. Whether regulation of G_s-associated / is a physiological mechanism in the D₁-like receptor interaction with NHE3 or is simply a pathological consequence of PTX treatment remains uncertain.

In summary, D₁-like agonists via G_s, independent of cytoplasmic second messengers, inhibit NHE3 activity in what appears to be a direct regulatory coupling of the two proteins. / dimers released on activation of ₂-adrenergic receptors can act to oppose the G_s inhibitory effect most likely at the level of NHE3. Our findings do not rule out that PTX-insensitive G subunits implicated here in ₂-adrenergic signaling may also contribute to stimulation of NHE3. Further research will also be required to determine if separate functional pools of G/, observed in intact cells (20), exert regulatory effects both at the level of G_s and the effector NHE3. Proteins that regulate NHE3 activity, such as NHE regulatory factor and NHE3 kinase A regulatory protein, are unlikely to be involved in our experiments using BBMV, because the regulation of NHE3 by these proteins involves cAMP (14).

Perspectives