Gastrin Induces Sodium-Hydrogen Exchanger 3 Phosphorylation and mTOR Activation via a Phosphoinositide 3-Kinase-/Protein Kinase C-Dependent but AKT-Independent Pathway in Renal Proximal Tubule Cells Derived From a Normotensive Male Human

Abstract

Gastrin is natriuretic, but its renal molecular targets and signal transduction pathways are not fully known. In this study, we confirmed the existence of CCKBR (a gastrin receptor) in male human renal proximal tubule cells and discovered that gastrin induced S6 phosphorylation, a downstream component of the phosphatidylinositol 3 kinase (PI3 kinase)-mammalian target of rapamycin pathway. Gastrin also increased the phosphorylation of sodium-hydrogen exchanger 3 (NHE3) at serine 552, caused its internalization, and decreased its expression at the cell surface and NHE activity. The phosphorylation of NHE3 and S6 was dependent on PI3 kinases because it was blocked by 2 different PI3-kinase inhibitors, wortmannin and LY294,002. The phosphorylation of NHE3 and S6 was not affected by the protein kinase A inhibitor H-89 but was blocked by a pan-PKC (chelerythrine) and a conventional PKC (cPKC) inhibitor (Gö6976) (10 μM) and an intracellular calcium chelator, 1,2-bis-(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid, tetra(acetoxymethyl)-ester, suggesting the importance of cPKC and intracellular calcium in the gastrin signaling pathway. The cPKC involved was probably PKCα because it was phosphorylated by gastrin. The gastrin-mediated phosphorylation of NHE3, S6, and PKCα was via phospholipase C because it was blocked by a phospholipase C inhibitor, U73122 (10 μM). The phosphorylation (activation) of AKT, which is usually upstream of mammalian target of rapamycin in the classic PI3 kinase-AKT-p70S6K signaling pathway, was not affected, suggesting that the gastrin-induced phosphorylation of NHE3 and S6 is dependent on both PI3 kinase and PKCα but not AKT.

Studies in the mid-1970s showed that an oral sodium load produces a greater natriuretic response than an iv sodium load, suggesting roles of hormones secreted by the gut in this process, but the identity of such hormones is not known (1). Although an increase in blood pressure (BP) after an increase in salt intake (pressure-natriuresis) is accepted as an important mechanism in the regulation of sodium balance (2, 3), BP does not regulate normal sodium excretion (4). Several hormones have been suggested to mediate the increase in sodium excretion after an oral sodium load, independent of atrial natriuretic peptide and aldosterone (5). These candidate gastrointestinal hormones include guanylin, uroguanylin, cholecystokinin, and gastrin. Although, Guca2b^−/− mice have an impaired ability to excrete an acute oral sodium load, their blood pressures are only slightly increased, and salt sensitivity is similar to that of Guca2b^+/+ littermates (6). Furthermore, circulating guanylin levels are not increased by a sodium load and that “uroguanylin's renal effects depend on its local synthesis and release in the kidney” (5), indicating that uroguanylin is unlikely to be the major hormone responsible for postprandial natriuresis. Cholecystokinin (CCK) is natriuretic (7), but circulating CCK levels are not increased by a duodenal infusion of isotonic saline (8) and CCK is not taken up by renal tubules (9). Food, which contains sodium, increases serum gastrin levels (10, 11) and circulating gastrin levels are 10- to 20-fold higher than CCK (12). Of all the gut hormones, gastrin is the one that is taken up to the greatest extent by renal proximal tubules (9). Gastrin is important in the regulation of sodium balance and BP because mice lacking gastrin (ie, Gast^−/− mice) are hypertensive (10) and salt-sensitive (P. Jose, Y. Yang, C. Zeng, C. Escano, L. Asico, S. Cuevas, V.A. Villar, X. Wang, J.R. Pisegna, I. Wank, R.A. Felder, I. Armando, unpublished observations). However, the mechanisms by which gastrin regulates sodium transport in the kidney are not known. Because the gastrin receptor, CCKBR, is highly expressed in the renal proximal tubule (13), and because the renal proximal tubule is 1 site of increased sodium transport in human essential hypertension (14, 15), we concentrated our studies using renal proximal tubule cells from humans.

CCK and gastrin, encoded by 2 distinct genes, are synthesized and secreted by I cells in the upper intestine and G cells in the gastric antrum, respectively, but they share the same receptors, CCKA receptor (CCKAR) and CCKB receptor (CCKBR), which belong to the family of G protein-coupled receptors. Sulfated CCK binds to CCKAR with a 500- to 1000-fold higher affinity than sulfated gastrin or nonsulfated CCK, whereas the CCKBR binds and responds to gastrin or CCK with similar affinity, regardless of the state of sulfation. Because the levels of plasma gastrin are much higher than CCK, the CCKBR can be considered the gastrin receptor (16–18).

CCKAR mRNA has been shown to be expressed in the murine and human kidney, but the renal expression of CCKAR protein has not been reported (16). In contrast, CCKBR has been shown to be present in the guinea pig, mouse, and rat kidney at both the mRNA and protein level; it is expressed to a greater extent in proximal and distal tubules and collecting ducts and to a lesser extent in glomeruli (Refs 10, 13, and 19 and Y. Yang, S. Evans, C. Escano, L. Asico, S. Zhang, S. Cuevas Gonzales, V.A. Villar, X. Wang, J.R. Pisegna, I. Wank, I. Armando, P.A. Jose, unpublished studies). Gastrin, via the CCKBR, has been reported to increase sodium and potassium excretion in the isolated, perfused rat kidney (20). The intrarenal infusion of CCK in anesthetized rats has also been shown to induce diuresis and natriuresis (21). CCK also inhibits Na⁺/K^+-ATPase activity in the intestinal mucosa (22). The inhibition of renal and intestinal sodium transport by gastrin is in contrast to its ability to increase H⁺/K⁺-ATPase activity in gastric parietal cells (23, 24) and the ability of CCK to increase Na⁺/K⁺-ATPase activity in pancreatic acinar cells (25–27). These studies suggest tissue specificity of the regulation of ion transport by gastrin and CCK.

Normal ion transport in the GI tract and kidney is critical in the maintenance of body water and sodium homeostasis. Na⁺/H⁺ exchanger (NHE)-3 located at the apical membrane is the major ion transporter responsible for luminal sodium transport in the renal proximal tubule and intestine (15, 28–30). Acute stimulation or inhibition of NHE3 by hormones, growth factors, and second messengers (15, 28–30) is associated with redistribution of NHE3 to the plasma membrane, in the case of stimulation, and redistribution from the plasma membrane to the subapical area, in the case of inhibition (28–29, 31–34). Therefore, guided by these reports, we studied the mechanism by which NHE3 is regulated by gastrin. We now report that in cultured human renal proximal tubule cells, gastrin increased the phosphorylation of NHE3 at serine 552, caused its internalization, and decreased its expression at the cell surface and its activity, in agreement with the effect of high-salt diet, serotonin, and dopamine in nonhuman renal proximal tubule cells (33, 35–37). Gastrin also increased the phosphorylation of S6, a downstream component in the phosphatidylinositol 3 kinase (PI3 kinase)-mammalian target of rapamycin (mTOR) signaling pathway. We studied S6 phosphorylation and used it as a marker for PI3 kinase activation throughout our study. In addition its natriuretic effect, gastrin has been demonstrated to stimulate the growth of renal proximal tubule cells (13), the mechanism of which is not known. In this study we studied both NHE3 phosphorylation and S6 phosphorylation induced by gastrin and show that the gastrin-induced phosphorylation of NHE3 and S6 was dependent on both PI3 kinase and conventional PKC (cPKC) but not the classical AKT pathway. We hypothesize that the natriuretic effect of gastrin relies on the inhibition of sodium transporters such as NHE3, whereas its growth-stimulating effect depends on the stimulation of the mTOR pathway, as reflected by S6 phosphorylation. These 2 seemingly independent events are both regulated by gastrin.

Materials and Methods

Chemicals and antibodies

Materials and sources are the following: human gastrin (Genscript, Piscataway, New Jersey); Gö6976 and H-89 (LC Labs, Woburn, MA); 2′,7′-bis(carboxyethyl)-5(6)-carboxyfluorescein (BCECF) and bis-(o-aminophenoxy) ethane-N,N,N′,N′-tetraacetic acid, tetra(acetoxymethyl)-ester (BAPTA-AM; Invitrogen, Grand Island, New York); AKT inhibitor V (EMD Biosciences, Billerica, Massachusetts); L365,260, phosphatase inhibitor cocktail (Roche, Indianapolis, Indiana); antiphospho-NHE3 antibody (Novus, Littleton, Colorado); anti-NHE3 antibody (Dr Mark Knepper, National Heart, Lung, and Blood Institute, Bethesda, Maryland); Path-Scan antibody and antibodies against phospholipase C (PLC), protein kinase C (PKC) isoform, AKT, and phospho-AKT (Cell Signaling Technology, Danvers, Massachusetts); anti-phospho-PKCα antibody and secondary antibodies (Santa Cruz Biotechnology, Santa Cruz, California); primers for human CCKAR and CCKBR and human CCKBR cDNA clone (OriGene, Rockville, Maryland); Pierce cell surface protein isolation kit (Thermo Scientific, Rockford, Illinois), and U73122, chelerythrine, and other chemicals, including protease inhibitor cocktail (Sigma-Aldrich, St Louis, Missouri).

Cell culture

Male human proximal tubule cells (NT 16) were initially cultured and differentiated, as described (38), and then switched to DMEM with 4.5% fetal bovine serum, antibiotics, and antimycotics (Invitrogen) for further growth. Cells were serum starved for 3 hours before treatment. For cells treated with gastrin and a particular inhibitor, the inhibitor was added 5 minutes before the addition of gastrin. Most inhibitors (including L365, 260, BAPTA-AM, U73122, Gö6976, and wortmannin) were dissolved in dimethylsulfoxide, whereas chelerythrine and gastrin were dissolved in sterile deionized water. The corresponding amount of vehicle was added in the vehicle-treated (control) groups.

RNA extraction and quantitative PCR (qPCR)

Total RNA, extracted from NT16 cells (RNeasy Plus minikit; QIAGEN, Valencia, California) was reverse transcribed (Bioline, Taunton, Massachusetts) and real-time PCR performed (SYBR green PCR; AB Applied Biosystems, Carlsbad, California).

Immunoblot analyses

Cells were collected into lysis buffer, with both protease and phospatase inhibitor cocktail, and then sonicated for 30 seconds. Equal amounts of protein were loaded onto 4%–20% or 10%–20% tricine gels (Invitrogen) for SDS-PAGE, transferred onto nitrocellulose membranes (Bio-Rad Laboratories, Hercules, California), and then probed with primary antibody (overnight/4 C). The blots were then incubated with the corresponding horseradish peroxidase-conjugated secondary antibody (Santa Cruz Biotechnology) (2 hours/22 C) and developed with enhanced chemiluminescence reagent (Santa Cruz Biotechnology). The autoradiographs were scanned and the intensity of the signals (including a blank region) was quantified (Image J; National Institutes of Health, Bethesda, Maryland). Equal loading of samples was confirmed by glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or α-tubulin on the same membrane. Control and treatment values were corrected for blank values and were finally expressed as relative percentage changes over controls.

Biotinylation of cell surface proteins

Cell surface protein was isolated using Pierce isolation kit, as per the manufacturer's instructions (Thermo Scientific, Rockford, Illinois).

Sodium-hydrogen exchanger activity

Intracellular pH was measured with the pH-sensitive fluorescent probe, BCECF (39). Cells grown to confluence on 96-well plates were loaded with 12 μM BCECF-AM in control solution (141 mM NaCl; 5.4 mM KCl; 1 mM CaCl₂; 0.4 mM KH₂PO₄; 0.5 mM MgCl₂; 0.4 mM MgSO₄; 0.3 mM Na₂HPO₄; 10 mM HEPES; 0.6 mM glucose, pH 7.4). The plates were rinsed with the control solution to remove the BCECF-containing solution and fluorescence measured with excitation wavelengths of 485 and 355 nm, and intensity of emission was measured at 535 nm (Victor 3; PerkinElmer, Norwalk, Connecticut). The cells were then exposed to 20 mM NH₄Cl (121 mM NaCl; 5.4 mM KCl; 1 mM CaCl₂; 0.4 mM KH₂PO₄; 0.5 mM MgCl₂; 0.4 mM MgSO₄; 0.3 mM Na₂HPO₄; 10 mM HEPES; 0.6 mM glucose; 20 mM NH₄Cl, pH 7.4) for 2 minutes, and the fluorescence was measured with excitation wavelengths at 485 and 355 nm. Then the cells were exposed to the control solution and the fluorescence measured in the same manner. The sodium-hydrogen exchanger activity was expressed as the initial rate of fluorescence change at 485 nm (normalized with reading at 355 nm) during the first 2 minutes after the start of the pH recovery.

Statistical analyses

One-way ANOVA was used to compare more than 2 groups with 1 treatment, whereas a 2-way ANOVA was used to compare more than 2 groups with more than 1 treatment. A post hoc test used a Tukey's multiple comparison test with a P value < .05 deemed significant.

Results

CCKBR expressed in human renal proximal tubule cells is active

To confirm the expression of the CCK receptor in renal proximal tubule cells, total RNA was extracted from the human renal proximal tubule (NT16) cells and subjected to real-time RT-PCR with predesigned commercial qPCR CCKAR and CCKBR primer pairs (Origene). We were able to detect CCKBR but not CCKAR mRNA (Figure 1A). To determine whether CCKBR is active and its signaling pathway, we used the Path-Scan antibodies (Cell Signaling Technology) and found that gastrin induced the phosphorylation of S6, a downstream target of the PI3 kinase-mTOR pathway (Figure 1B). Cotreatment with the CCKBR-specific inhibitor L365,260 blocked the phosphorylation of S6 induced by gastrin (Figure 1, B and C).

Figure 1.

CCKBR is expressed in human renal proximal tubule cells and is active. A, CCKBR, not CCKAR, was detected by qPCR with commercial predefined primer sets in total RNA extracted from NT16 cells, and the products were subjected to agarose gel electrophoresis. B, NT16 cells were treated with gastrin at the indicated concentrations, alone or in combination with 1 μM CCKBR inhibitor L365,260 (1 μM) for 3 hours. Samples were subjected to immunoblotting with antiphospho-S6 and GAPDH antibodies; blots are representative of 3 independent experiments. C, Quantification of phospho-S6 levels in NT16 cells after treatment with gastrin at the indicated concentrations, alone or in combination with the CCKBR inhibitor L365,260 for 3 hours as in B (n = 3). * P < .05 vs others.

Gastrin treatment induced a concentration-dependent increase in NHE3 and S6 phosphorylation

To determine whether and how gastrin influences NHE3, we treated NT16 cells with gastrin alone or in combination with the PI3 kinase inhibitor, wortmannin (1 μM). Gastrin increased NHE3 phosphorylation at serine 552 in a concentration-dependent and biphasic manner (Figure 2, A and B). Our preliminary dose-response study (data not shown) showed that gastrin treatment increased S6 and NHE3 phosphorylation to the maximum level at approximately 300–500 nM, and such phosphorylation was blunted at 1000 nM. There was also a parallel increase in S6 phosphorylation, whereas phospho-P44/42 was unchanged (Figure 2, A and B). S6 is a well-known downstream component of the PI3 kinase-AKT (protein kinase B)-mammalian target of rapamycin complex 1 pathway (27). The gastrin-mediated phosphorylation of NHE3 and S6 was significantly blocked by pretreatment with wortmannin, suggesting that PI3 kinase activity was essential in this process (Figure 2, A and B). Total NHE3 level was not altered by gastrin treatment (Figure 2A), suggesting that the increase in NHE3 phosphorylation was not due to a change in total NHE3 expression. To further prove the specificity of PI3 kinase inhibition by wortmannin (1 μM), we also treated cells with another commonly used PI3 kinase inhibitor, LY294,002, and obtained similar results (data not shown). Because gastrin induced the phosphorylation of both NHE3 and S6, we studied them side by side as internal control for each other throughout our study.

Figure 2.

Gastrin treatment induces a concentration-dependent increase in NHE3 phosphorylation as well as S6 phosphorylation, which is PI3 kinase dependent. A, NT16 cells were treated with gastrin at the indicated concentrations both alone and in combination with wortmannin (1 μM) for 3 hours. The samples were then subjected to immunoblotting with antiphospho-NHE3, phospho-P44/42, phospho-S6, eukaryotic translation initiation factor 4E (elF4E), and total NHE3; blots are representative of 3 independent experiments. B, Quantification of phospho-NHE3 expression in NT16 cells after treatment with gastrin at the indicated concentrations both alone and in combination with wortmannin (1 μM) for 3 hours as in A (n = 3; each group is assigned a letter, and different letters indicate significant difference with P < .05; eg, a is different from b, c, and d; b is different from a, c, and d; c is different from a, b, and d; d is different from a, b, and c).

Gastrin reduced the cell surface expression of NHE3

The acute regulation of NHE3 by hormones usually involves changes in its subcellular redistribution. For example, the phosphorylation of NHE3 resulting in inhibition of its activity is associated with its internalization (34–37). We asked whether gastrin may regulate NHE3 by similar mechanisms. Gastrin treatment, for 3 hours at 100 and 300 nM, did not change total NHE3 (Figure 2A) but significantly decreased the cell surface NHE3 (Figure 3A, 3B). The quality of the cell surface labeling was verified with integrin-β3, a cell surface protein marker, which was present in the cell surface protein product but not in the flow-through samples (data not shown). We tried but failed to detect any signal with antiphospho-NHE3 antibody using the purified biotinylated samples. However, the flow-through and crude biotinylated samples (before passing through the avidin beads) had increased phospho-NHE3 levels after gastrin treatment (Figure 3C), consistent with the results shown in Figure 2A. These results indicate that the gastrin-induced increase in NHE3 phosphorylation is accompanied by a decrease in its cell surface expression, indicative of an increase in NHE3 protein internalization.

Figure 3.

Gastrin treatment decreases cell surface expression of NHE3 and activity. A, NT16 cells were treated with gastrin at the indicated concentrations and cell surface proteins were biotinylated, purified, and subjected to immunoblotting for total NHE3. B, Quantification of cell surface NHE3 expression in NT16 cells after treatment with gastrin at the indicated concentrations (n = 3). * P < .05 vs gastrin 0 nM. C, The crude biotinylated samples of NT16 cells treated with gastrin as mentioned above in A, as well as flow-through after the avidin beads binding, were subjected to immunoblotting for phospho-NHE3. D, NT16 cells were treated with gastrin at the indicated concentrations for 3 hours, and NHE activity was measured as described in Materials and Methods (n = 6). * P < .05 vs gastrin 0 nM.

Gastrin treatment decreased NHE activity

Because activation of NHE3 at the cell surface increases sodium transport, its internalization should be associated with a corresponding decrease in its apparent activity, which can be measured by the rate of pH recovery after acidification with NH₄Cl (33–37).

As shown in Figure 3D, gastrin treatment, at the indicated concentrations for 3 hours, decreased the rate of Na⁺/H⁺ exchange, as reflected by the decreased rate of pH recovery.

NHE3 and S6 phosphorylations were PKC dependent

To determine whether the phosphorylation of NHE3 and S6 caused by gastrin is PKC dependent, NT16 cells were treated with gastrin, alone or in combination, with a pan-PKC inhibitor chelerythrine. Gastrin increased both NHE3 and S6 phosphorylation, which was completely prevented by cotreatment with chelerythrine (Supplemental Figure 1, A and B, published on The Endocrine Society's Journals Online website at http://endo.endojournals.org). These results indicate that PKCs were also involved in the signal transduction of NHE3. We found that eukaryotic translation initiation factor 4E in the Path-Scan antibody mixture (Cell Signaling Technology), which is supposed to be a control, was also slightly but positively affected by gastrin treatment (Supplemental Figure 1A), so we used other housekeeping genes as internal controls as described in the Materials and Methods.

We also tested the effects of the protein kinase A (PKA) inhibitor H-89; PKA inhibition did not block the phosphorylation of NHE3 or S6 induced by gastrin (data not shown), suggesting that PKA is not directly involved in this aspect of gastrin signal transduction.

Gastrin-induced NHE3 and S6 phosphorylation was dependent on intracellular calcium

NT16 cells treated with gastrin showed increased phosphorylation of NHE3 and S6 (Figure 4A), similar to those shown in previous figures, but pretreatment with BAPTA-AM, an intracellular calcium chelator, decreased both the basal and gastrin-mediated phosphorylation of NHE3 and completely blocked basal and gastrin-mediated phosphorylation of S6 (Figure 4). Because the activation of cPKCs is dependent on increased intracellular calcium, we next examined the effects of a specific cPKC inhibitor.

Figure 4.

Gastrin-induced phosphorylation of NHE3 and S6 is dependent on intracellular calcium. A, NT16 cells were treated with gastrin at the indicated concentrations both alone or in combination with the intracellular calcium chelator BAPTA-AM (25 μM) for 3 hours, and samples were subjected to immunoblotting with antiphospho-NHE3, phospho-S6, eukaryotic translation initiation factor 4E (elF4), and GAPDH antibodies; blots are representative of 3 independent experiments. B, Quantification of phospho-S6 levels in NT16 cells after treatment with gastrin at the indicated concentrations both alone or in combination with intracellular calcium chelator BAPTA-AM (25 μM) for 3 hours as in Figure 5A (n = 3; each group is assigned a letter, and different letters indicate significant difference with P < .05; eg, a is different from b, c, and d; b is different from a, c, and d; c is different from a, b, and d; d is different from a, b, and c).

Gastrin-induced NHE3 and S6 phosphorylation was blocked by Gö6976, a cPKC inhibitor

NT16 cells were treated with gastrin alone or in combination with the cPKC inhibitor Gö6976 (10 μM). The basal and gastrin-mediated increase in the phosphorylation of NHE3 and S6 was decreased by the cotreatment with the cPKC inhibitor (Figure 5, A and B). As a control, BAPTA-AM treatment also decreased the phosphorylation of NHE3 under these conditions, similar to those shown in Figure 4.

Figure 5.

Gastrin-induced phosphorylation of NHE3 and S6 is blocked by the cPKC inhibitor Gö6976 as well as by the PLC inhibitor U73122. A, NT16 cells were treated with gastrin alone or in combination with either BAPTA-AM (25 μM) or Gö6976 (10 μM) for 3 hours, and samples were subjected to immunoblotting with antiphospho-NHE3, phospho-S6, and α-tubulin antibodies; blots are representative of 3 independent experiments. B, Quantification of phospho-S6 in NT16 cells after treatment with gastrin at the indicated concentrations as in Figure 6A (n = 3; each group is assigned a letter, and different letters indicate significant difference with P < .05; eg, a is different from b, c, and d; b is different from a, c, and d; c is different from a, b, and d; d is different from a, b, and c). C, NT16 cells were treated with gastrin at the indicated concentrations, alone or in combination with PLC inhibitor U73122 (10 μM) for 3 hours, and samples were subjected to immunoblotting with antiphospho-NHE3, phospho-S6, and GAPDH antibodies; blots are representative of 3 independent experiments. D, Quantification of phospho-NHE3 levels in NT16 cells after treatment with gastrin at the indicated concentrations, alone or in combination with PLC inhibitor U73122 (10 μM) for 3 hours as in Figure 6C (n = 3; each group is assigned a letter, and different letters indicate significant difference with P < .05; eg, a is different from b, c, and d; b is different from a, c, and d; c is different from a, b, and d; d is different from a, b, and c).

Gastrin-induced NHE3 and S6 phosphorylation was dependent on PLC

Because gastrin increased the phosphorylation of NHE3 and S6 via activation of cPKC and PKC is activated by PLC, we next determined the effect of the PLC inhibitor U73122 on gastrin-mediated phosphorylation of NHE3 and S6. Gastrin induced the phosphorylation of both NHE3 and S6, as shown in the previous studies, but such phosphorylation was attenuated by the PLC inhibitor U73122 (Figure 5, C and D), suggesting that PLC, which is usually upstream of PKC, is also involved in the gastrin signaling pathway.

Gastrin induced the phosphorylation of S6 but not AKT

AKT is generally accepted as upstream of mTOR and downstream of PI3 kinase in the classical PI3 kinase-mTOR signaling pathway. Because the phosphorylation of S6 is downstream of AKT in some systems (40, 41), we asked whether AKT was involved in the gastrin-mediated signaling pathway. NT16 cells were treated with vehicle or gastrin, and phosphorylation of S6 and AKT was determined by immunoblotting. Total S6, which was used as an internal control, was not affected by gastrin, but gastrin did increase phospho-S6 (Figure 6A). In contrast, gastrin had no effect on the phosphorylation of AKT (Figure 6A). These results suggest that PKC, instead of AKT, is upstream of mTOR and S6 in gastrin-CCKBR signaling, which is similar to epithelial growth factor receptor signaling to mTOR through PKC and independently of AKT in glioma cells (42).

Figure 6.

Gastrin induced phosphorylation of PKCα but not phosphorylation of AKT, which is blocked by the PLC inhibitor U73122. A, NT16 cells were treated with gastrin at the indicated concentrations for 3 hours, and samples were subjected to immunoblotting with antiphospho-AKT(Ser473), phospho-S6(Ser235/236), eukaryotic translation initiation factor 4E (elF4E), and S6 antibodies; blots are representative of 3 independent experiments. B, NT16 cells were treated with gastrin at the indicated concentrations for 3 hours, and samples were subjected to immunoblotting with antiphospho-PKCα, which may also cross-react with PKCβII (Cell Signaling Technology), phospho-S6(Ser235/236), and elF4E antibody; blots are representative of 3 independent experiments. C, NT16 cells were treated with gastrin at the indicated concentrations for 3 hours, and samples were subjected to immunoblotting with a specific antiphospho-PKCα (Santa Cruz) and α-tubulin antibodies; blots are representative of 3 independent experiments. D, Quantification of phospho-PKCα in NT16 cells after treatment with gastrin at the indicated concentrations as in Figure 7C (n = 3; each group is assigned a letter, and different letters indicate significant difference with P < .05; eg, a is different from b, c, and d; b is different from a, c, and d; c is different from a, b, and d; d is different from a, b, and c).

Gastrin-mediated increase in PKCα was blocked by the PLC inhibitor U73122

NT16 cells were treated with gastrin and the cells were lysed with buffer containing both proteinase and phosphatase inhibitor cocktails. Gastrin treatment increased the phosphorylation of S6 and PKCα, as revealed with both an antibody that recognizes both phospho-PKCα and phospho-PKCβ II (Figure 6B) and an antibody that recognizes only phospho-PKCα (Figure 6, C and D). The gastrin-induced phosphorylation of PKCα was reduced by the PLC inhibitor U73122 (10 μM), suggesting that PLC is upstream of PKCα.

Discussion

We now report that CCKBR mRNA but not CCKAR is expressed in human renal proximal tubule cells, consistent with those reported in the guinea pig and rodents (Refs 10, 13, and 19 and X. Wang, R.A. Felder, P.A. Jose, unpublished studies). Moreover, we show for the first time that gastrin increased NHE3 phosphorylation but deceased NHE3 cell surface expression in these human renal proximal tubule cells. Gastrin also caused phosphorylation of S6. A CCKBR antagonist, L365,260 (1 μM) inhibited gastrin-induced phosphorylation of NHE3 and S6. These results indicate that CCKBR is expressed in those cells, and it is active.

We used the Path-Scan antibodies (Cell Signaling Technology) to figure out which pathway is activated by gastrin. As indicated above, S6 is phosphorylated by gastrin. S6 phosphorylation, as a consequence of gastrin treatment, has been reported in other systems. S6 is a downstream component of the PI3 kinase-AKT-mTOR pathway, and therefore, we used S6 phosphorylation as a marker of gastrin signaling through CCKBR in subsequent experiments. Our initial time-course and concentration-response experiments showed that gastrin induced a concentration-dependent and biphasic increase in S6 phosphorylation, with the maximum effect reached at approximately 300–500 nM but blunted at 1000 nM. The time-course experiments also indicated that the maximum effect of gastrin on S6 phosphorylation was reached and maintained at 2–3 hours (data not shown). We chose to observe the effects of gastrin within this time frame because this is also the time frame at which the natriuresis after a meal occurs (43).

We studied the effect of gastrin on NHE3 because it is the major sodium transporter in the luminal membrane of renal proximal tubule cells, and its phosphorylation and internalization induced by other hormones have been studied (34–37, 44). The phosphorylation of NHE3 at serine 661 has been reported to be associated with an increase in NHE3 activity (45), whereas the phosphorylation of NHE3 at serines 552 and 605 results in an inhibition of NHE3 activity (34, 35, 46, 47). The phosphorylation of NHE3, at serine 552 and 605, results in its redistribution away from the tips of the microvilli to the base of the microvilli (33, 37, 48). We found with cell surface protein biotinylation that gastrin treatment reduced the cell surface expression of NHE3, without altering the total NHE3 expression in the whole-cell lysate and was associated with a decrease in NHE activity. In addition to the apical NHE3, there are other NHE isoforms expressed in the renal proximal tubule including NHE1 at the basolateral membrane, which is involved in pH regulation, and NHE8 at the apical membrane (30). The change in NHE activity may be the sum of NHE1, NHE3, and NHE8 activities. Therefore, we also tested the effects of gastrin treatment on NHE1 expression. Interestingly, gastrin treatment also quickly reduced the level of NHE1 (Liu, T., unpublished data,). These results suggested that both NHE3 and NHE1 are targets of gastrin, and the coordinated inhibition of NHE1 and NHE3 is responsible for the reduction of sodium transport induced by gastrin. The effect of gastrin on NHE8 remains to be determined.

Having demonstrated that gastrin increased the phosphorylation and internalization of NHE3, which resulted in decreased activity, we next explored the gastrin-CCKBR signal transduction pathway using both S6 phosphorylation and NHE3 phosphorylation as end points. Because S6 phosphorylation is downstream of PI3 kinase (40), we tested whether gastrin-induced NHE3 phosphorylation was also downstream of PI3 kinase. This was the case because the PI3 kinase inhibitor wortmannin prevented the phosphorylation of NHE3 by gastrin. Another commonly used PI3 kinase inhibitor, LY294,002, produced similar results (data not shown), suggesting that the wortmannin-mediated blockade of NHE3 and S6 phosphorylation was probably through PI3 kinase.

We also tested the effects of PKC inhibitors on gastrin-mediated NHE phosphorylation because gastrin is known to activate PKC in other tissues (49, 50). Cotreatment with chelerythrine, a pan-PKC inhibitor, decreased the basal and gastrin-induced phosphorylation of NHE3 and S6, suggesting that PKC is essential for the phosphorylation of NHE3 and S6. This is reminiscent of the acute regulation of NHE3 phosphorylation by PKC in other systems (51, 52).

Because there are at least 12 different isoforms of PKC, we next determined which PKC isoform is responsible for the gastrin-mediated phosphorylation of NHE3 and S6. Gö6976 inhibits the conventional PKCs, which include PKCα, -βI, -βII, and -γ. This cPKC inhibitor reduced the basal and blocked gastrin-induced phosphorylation of NHE3 and S6. Because the activation of cPKC requires an increase in intracellular calcium, we also tested the effect of decreasing intracellular calcium concentration. When intracellular calcium concentration was decreased with the chelator BAPTA-AM, the basal and gastrin-induced phosphorylation of NHE3 and S6 was blocked. To further prove the involvement of cPKC in this process, we first used an antibody that is unique to phosphorylated PKCα and -βII. After gastrin treatment the phosphorylation of PKC was increased. To confirm that gastrin could increase the phosphorylation of PKCα, we used another antibody that is unique to phospho-PKCα and found that its phosphorylation was also increased by gastrin in a concentration-dependent manner. Moreover, when the cells were cotreated with U73122, a PLC inhibitor, the basal and gastrin-induced phosphorylation of PKCα was blocked or reduced. These results strongly indicated that PKCα was involved in the gastrin-induced phosphorylation of NHE3 and S6 and that PKCα was downstream of PLC. Gastrin-induced phosphorylation of PKCα is consistent with studies of NHE3 regulation by other groups because it has been shown that elevating intracellular calcium causes an acute inhibition of NHE3 activity by decreasing its plasma membrane expression in a PKCα-dependent manner (53). In our study, PKCα was also needed for the phosphorylation of NHE3 caused by gastrin. Although we did not test whether PKC inhibition blocked the internalization of NHE3, Lee-Kwon et al (53) have demonstrated that PKC was needed to decrease the plasma membrane expression of NHE3. However, the ability of PKC to phosphorylate NHE3 was not studied in that report. Our current study provides the first evidence that gastrin increases NHE3 phosphorylation and internalization and decreases NHE activity.

We found that phosphorylation of PLCγ1, but not PLCγ2, was increased by gastrin treatment (data not shown). This finding is consistent with the reports that gastrin induced inositol trisphosphate formation through PLCγ1 in rat colonic epithelial cells (54) and that CCKBR is associated with PLCγ1 (55). Phosphorylated PLCγ has been shown to bind directly with NHE3 and is required for calcium regulation of exchanger activity (56), which are consistent with our findings.

In this study, we showed that both the PI3 kinase and PKC pathways are involved in gastrin-CCKBR signaling, but how these 2 pathways are connected is not known. AKT is generally accepted as downstream of PI3 kinase but upstream of S6. Our results showed that both PI3 kinase and PKC inhibitors completely blocked the phosphorylation of S6, suggesting that these 2 pathways acted in a sequential manner rather than 2 parallel pathways. Because the phosphorylation of S6 is downstream of AKT in some systems, we also tested whether gastrin-induced S6 phosphorylation involved AKT. Although gastrin increased the amount of phosphorylated S6, the amount of phosphorylated AKT was not affected. The selective AKT inhibitor V (1 μM) also failed to block the gastrin-induced phosphorylation of S6 (data not shown), so the gastrin-induced phosphorylation of NHE3 and S6 is AKT independent. We did not study how PI3 kinase interacted with PLC after gastrin treatment and will be studied in the future, but we suggest that PI3 kinase very likely acts upstream of PLC. It has been shown in other systems that PI3 kinase increases the level of phosphatidylinositol 4,5-trisphosphate, which in turn facilitates the binding of PLC to the membrane, in which it catalyzes the conversion of phosphatidylinositol 4,5-bisphosphate to diacylglycerol and inositol trisphosphate, both of which are needed for PKCα activation. Activation of PI3 kinase has been shown to cause PLCγ pleckstrin homology domain-mediated membrane targeting and activation of PLCγ (57). Therefore, it is likely that PI3 kinase acted upstream of PLCγ1 in the gastrin-NHE3 pathway. It is not uncommon for PI3 kinase to interact with PKC pathways in other systems. Based on our results, we propose a hypothetical signaling pathway for gastrin-induced phosphorylation of NHE3 and a decrease in renal sodium transport (Figure 7).

Figure 7.

Hypothetical gastrin-CCKBR signal transduction pathway. Gastrin signaling through CCKBR causes PI3 kinase activation and the production of phosphatidylinositol 4,5-trisphosphate (PIP3), which in turn increases the binding of PLC to the plasma membrane. PLC then activates PKCα through both diacylglycerol (DAG) and calcium, and PKCα, but not AKT, causes the phosphorylation of NHE3 and S6.

In this study, we also showed that there was a parallel activation of the mTOR pathway, as reflected by phosphorylation of S6, which is dependent on cPKC (most likely PKCα) and calcium but not AKT. The activation of mTOR is associated with good nutritional conditions and cells tend to grow when mTOR is activated (58). Gastrin, whose secretion is stimulated after food intake, may reflect the availability of nutrients (59), and gastrin has been demonstrated to stimulate the growth of renal proximal tubule cells (13), which is consistent with S6 phosphorylation shown here and in another study (in revision). It is possible that gastrin activates mTOR, as a reflection of good nutritional status and signals cells to grow but also down-regulates NHE3 and Na⁺/K⁺-ATPase and therefore sodium transport, as a negative feedback mechanism. Gastrin may play multiple roles in renal proximal tubule cells. The function of gastrin in other regions of the kidney in which its receptor is expressed could be the subject of future studies.