diabetes mellitus is associated with Na⁺ and water retention and extracellular fluid volume expansion (9, 48). A principal site of renal salt and water reabsorption is the proximal tubule, where insulin receptors have been found in different species (16, 46, 61). Insulin is present in the plasma and glomerular ultrafiltrate and is degraded in the proximal tubule (32). Several studies have provided evidence that insulin decreases urinary Na⁺ excretion (24, 45, 47, 54,57). Baum (10) has shown that insulin directly stimulates volume absorption in rabbit proximal convoluted tubules. The stimulatory effect on the proximal tubule is associated with increased apical H⁺ secretion (40, 60) and EIPA-sensitive Na⁺ uptake (28), findings compatible with increased apical membrane Na⁺/H⁺ exchange activity. One postulate is that peripheral insulin resistance may be associated with relatively preserved insulin sensitivity in the kidney, and the price of hyperinsulinemia are renal NaCl retention and salt-sensitive hypertension (51, 56).

In the mammalian proximal tubule, >60% of the Na⁺absorption is mediated by apical brush-border membrane Na⁺/H⁺ exchange. Of the seven isoforms known to date, NHE3 is the only Na⁺/H⁺ exchanger isoform definitively shown to be expressed in the brush-border membrane of the renal proximal tubule, on the basis of antigenic (6, 13) and functional data (20, 62, 67). NHE3 mediates proximal tubule transcellular NaCl absorption via coupled transport with Cl⁻/base exchange (8, 63) as well as paracellular NaCl transport by lowering luminal HCO${}_3^{-}$concentration and elevating luminal Cl⁻ concentration (52). The importance of NHE3 in sustaining extracellular fluid volume is evident by the hypovolemia and hypotension seen in NHE3 null mice (55). Previous studies examining the effect of insulin on the proximal tubule did not specifically address the NHE3 isoform. The present study investigates the effects of insulin on apical membrane NHE3 activity, surface protein, total protein, and transcript levels in a cell line of the opossum kidney with proximal tubule characteristics (OKP cells). Because hydrocortisone has been shown to exert a permissive effect on the acid-induced activation of Na⁺/H⁺ exchange activity (5), we also examined glucocorticoid dependence of insulin-induced activation.

MATERIALS AND METHODS

Materials and supplies.

All chemicals were obtained from Sigma (St. Louis, MO), except for the following: acetoxymethyl derivative of 2′7′-bis(2-carboxyethyl)-5(6)-carboxyfluorescein (BCECF; Molecular Probes, Eugene, OR); NHS-SS-biotin and immobilized streptavidin (Pierce, Rockford, IL); and culture media (GIBCO BRL, Grand Island, NY).

Cell culture.

OKP cells (22) were passaged in high-glucose (450 mg/dl) DMEM supplemented with 10% FBS, penicillin (100 U/ml), and streptomycin (100 μg/ml). Before study, confluent cells were rendered quiescent by incubation in serum-free media [1:1 mixture of low-glucose (100 mg/dl) DMEM and Ham's F-12 ± 10⁻⁶M hydrocortisone] for 24–48 h. Human insulin (10⁻⁶to 10⁻¹⁰ M) was applied for the stated period of time before the assays.

Measurement of intracellular pH and Na⁺/H⁺exchange activity.

Continuous measurement of cytoplasmic pH (pH_i) was accomplished using the intracellularly trapped pH-sensitive dye BCECF, as described previously (4). Cells were loaded with 10 μM acetoxymethyl ester of BCECF for 35 min at 37°C, and pH_i was estimated from the ratio of fluorescence (λ_ex: 500 and 450 nm, λ_em: 530 nm, where ex is excitation and em is emission) in a computer-controlled spectrofluorometer (8000C, SLM Instruments, Urbana, IL, and RF-5000, Shimadzu, Kyoto, Japan). The BCECF excitation fluorescence ratio was calibrated intracellularly using K/nigericin as described (3). Na⁺/H⁺ exchange activity was assayed as the initial rate of the Na⁺-dependent pH_i increase after an acid load in the absence of CO₂/HCO₃, and results are reported as dpH_i/dt. Comparisons were always made between cells of the same passage studied on the same day. Intracellular buffer capacity was measured by pulsing with 20 mM NH₄Cl. Buffer capacity (β) was then calculated according to the formula β = [NH₄Cl]/ΔpH_i. Results for control and insulin-treated cells were not significantly different (34.5 vs. 34.4 mM, respectively).

NHE3 antigen.

Cells were rinsed three times with ice-cold PBS and Dounce-homogenized in isotonic Tris-buffered saline (150 mM NaCl, 50 mM Tris · HCl, pH 7.5, 5 mM EDTA) containing proteinase inhibitors [100 μg/ml phenylmethylsulfonyl fluoride (PMSF), 4 μg/ml aprotinin, 4 μg/ml leupeptin]. After nuclei removal (13,000g, 4°C , 5 min; Eppendorf 5415C, Hamburg, Germany), membranes were pelleted (109,000 g, 4°C, 20 min; Sorvall RCM 120EX, rotor S120 AT2–0130, Sorvall Products, DuPont, Wilmington, DE) and resuspended in Tris-buffered saline, and total protein content was determined by the method of Bradford. Fifteen micrograms of protein were diluted 1:5 in 5× SDS loading buffer (1 mM Tris · HCl, pH 6.8, 1% SDS, 10% glycerol, 1% 2-mercaptoethanol), size fractionated by SDS-PAGE (7.5% gel), and electrophoretically transferred to nitrocellulose. After blocking for 1 h (5% nonfat milk, 0.05% Tween 20 in PBS), blots were probed in the same buffer with a polyclonal anti-opossum NHE3 antibody [antiserum 5683, generated against a maltose-binding protein/NHE3 (amino acids 484–839) fusion protein] at a dilution of 1:300 (4). Blots were washed in 0.05% Tween 20 in PBS once for 15 min and twice for 5 min, incubated with a 1:10,000 dilution of peroxidase-labeled sheep anti-rabbit IgG, washed as above, and then visualized by enhanced chemiluminescence. NHE3 protein abundance was quantitated by densitometry (BioCapt software version 72.02s for Windows, Vilbert Lourmat, Marne la Vallée, France) and Scion Image Beta 3b, 1998 (Scion, Frederick, MD).

To measure plasma membrane NHE3, we used a surface biotinylation assay (23). Monolayers were rinsed three times with ice-cold PBS-Ca-Mg (PBS with 0.1 mM CaCl₂, 1.0 mM MgCl₂). Membrane proteins were then biotinylated by incubation of cells in 1.5 mg/ml NHS-SS-biotin in 10 mM triethanolamine (pH 7.4), 2 mM CaCl₂, and 150 mM NaCl for 90 min at 4°C. After labeling, plates were washed with 6 ml quenching buffer (PBS-Ca-Mg, with 100 mM glycine) for 20 min at 4°C. Cells were then lysed in RIPA buffer (150 mM NaCl, 50 mM Tris · HCl, pH 7.4, 5.0 mM EDTA, 1% Triton X-100, 0.5% deoxycholate, 0.1% SDS, 100 μg/ml PMSF, 5 μg/ml aprotinin, 5 μg/ml leupeptin), extracts were rocked for 30 min at 4°C and centrifuged at 12,000 g at 2°C for 10 min, and the supernatant was diluted to 3 mg/ml with RIPA buffer. Biotinylated proteins were then affinity precipitated with streptavidin-conjugated agarose, released by β-mercaptoethanol, and subjected to immunoblotting with anti-NHE3 antisera as above.

NHE3 transcript.

RNA was extracted using RNeasy (Qiagen, Valencia, CA). Fifteen micrograms of total RNA were size fractionated by agarose-formaldehyde gel electrophoresis and transferred to nylon membranes. The radiolabeled NHE3 probe was synthesized from full-length OKP NHE3 cDNA (7) and an 18S probe from a 752-baseSphI/BamHI fragment of mouse 18S rRNA (no. 63178, American Type Culture Collection, Rockville, MD) by the random hexamer method. Prehybridization, hybridization, and washing were performed as described previously (4). Filters were exposed to film overnight at −70°C, and labeling was quantitated by densitometry. Changes in NHE3 abundance were normalized for changes in 18S rRNA abundance.

Statistics.

All results are reported as means ± SE. Statistical analysis was performed using ANOVA unless stated otherwise, and n refers to the number of plates studied.

RESULTS

Insulin increases Na⁺/H⁺exchanger activity in OKP cells.

A typical tracing (Fig. 1A) shows that insulin stimulates Na⁺/H⁺ exchange activity. Figure 1Bshows the time course of the insulin effect. Acute incubation for 40 min does not significantly affect activity (+6%, not significant). At 2–24 h of incubation, insulin increases Na⁺/H⁺ exchange activity. This effect was dose dependent (Fig. 1C), with a detectable effect down to 10⁻⁸ M insulin and half-maximal stimulation at ∼10⁻⁷ M for both the acute (2 h) and chronic (24 h) effect. OKP cells express an EIPA-resistant Na⁺/H⁺ exchanger that is encoded by NHE3 (7). However, to exclude the possibility that the observed changes in dpH_i/dt may be mediated through an effect of insulin on another NHE isoform, we performed experiments in the presence of 100 μM HOE-642, which completely inhibits NHE1 or NHE2, but not NHE3. HOE-642 does not affect baseline or insulin-stimulated Na⁺/H⁺ exchange activity (Fig. 2), suggesting that the observed effect of insulin on Na⁺/H⁺ exchange is exclusively on NHE3.

We have previously shown that the activation of NHE3 in response to chronic acid incubation requires the synergistic effect of hydrocortisone (5). To test for glucocorticoid dependence, we examined the interaction between insulin and hydrocortisone. Hydrocortisone was introduced into the incubation medium during the periods of both serum deprivation (48 h) and insulin treatment (4–24 h). As shown in Fig. 3, hydrocortisone alone (10⁻⁹ M) has no effect on NHE3 activity but, when given with insulin (10⁻⁷ M), stimulated NHE3 activity to a level higher than with insulin (10⁻⁷ M) alone. Similarly, insulin itself (10⁻⁷ M) has a small stimulatory effect on NHE3 activity, but the effect is much greater in the presence of 10⁻⁹ M hydrocortisone (Fig.3A). Figure 3B shows the effect of varying doses of hydrocortisone added to OKP cells for 24 h with and without 10⁻⁶ M insulin. Insulin augmented the effect of hydrocortisone from 10⁻⁹ through 10⁻⁷ M. At a saturating dose of hydrocortisone, insulin had no further effect. These findings are in accordance with a synergistic effect between hydrocortisone and insulin on NHE3. The synergistic effect of hydrocortisone on the insulin-induced increase in Na⁺/H⁺ exchange is only visible when hydrocortisone was added at least 24 h before insulin addition (during the period of serum deprivation). If hydrocortisone was added simultaneously with insulin, minimal or no difference is observed between the insulin vs. insulin+hydrocortisone group (data not shown).

Insulin increases total and cell surface NHE3 protein abundance.

Changes in NHE3 activity can be associated with changes in total cellular NHE3 protein and/or changes in surface plasma membrane NHE3 protein. Figure 4A shows a typical blot depicting the effect of insulin in the presence or absence of 10⁻⁶ M hydrocortisone on OKP NHE3 total and cell surface protein abundance. Insulin does not affect total or cell surface NHE3 protein abundance after 12 h when NHE activity is clearly stimulated. In contrast, insulin at 24 h increased total cellular NHE3 antigen by 27% and surface NHE3 by 60%. The results are summarized in Fig. 4B. These results indicate that the early (8–12 h) and late (24 h) stimulation of NHE is mediated by distinct mechanisms. In the absence of hydrocortisone, the increase in cellular and surface NHE3 is variable and much less pronounced (Fig.4A). In the presence of hydrocortisone, the increase in NHE3 activity observed at 24 h is associated with increased total cell and surface antigen.

Insulin increases NHE3 transcript.

Insulin treatment of OKP cells for 24 h increases NHE3 transcript abundance (Fig. 5A). In contrast, insulin treatment for 12 h actually slightly decreases the level of NHE3 transcript (P = 0.021). Again, we determined the hydrocortisone dependence of the insulin effect on NHE3 transcript level at 24 h. Insulin alone increases NHE3 transcript level slightly by ∼43%. As shown before (5), hydrocortisone (10⁻⁶ M) by itself approximately doubles the NHE3 transcript level. Combined treatment with insulin and hydrocortisone results in another 2.4-fold increase in NHE3 mRNA compared with hydrocortisone alone.

DISCUSSION

The role of insulin in renal tubular salt and water handling has been previously implicated from clinical observations (24, 47,54) and tubule perfusion studies in animals (10, 40,60). The causal relationship between hyperinsulinemia and hypertension is still an issue of debate (30). Reaven (51) has suggested that even in states of hyperinsulinemia, additional factors other than insulin likely contribute to hypertension. Although it is controversial as to whether hyperinsulinemia leads to salt-sensitive hypertension, the present body of data are strongly supportive of a salt-retaining action of insulin on the kidney. Insulin stimulates Na⁺ transporters and Na⁺ absorption at both the proximal tubule (11, 28,40, 60) and thick ascending limb (39, 60).

The molecular mechanisms of the insulin-induced increase in Na⁺ transport have not been examined. The present study demonstrates that insulin directly stimulates the Na⁺/H⁺ exchanger NHE3 in OKP cells in a time- and concentration-dependent manner. The concentrations used in our experiments were higher than the circulating plasma levels. Therefore, we cannot exclude that some of the effects on the Na⁺/H⁺ exchanger NHE3 are mediated through the insulin-like growth factor-1 receptor. However, for NHE3 activity, significant stimulation by insulin was detectable down to a concentration of 10⁻⁸ M. The NHE1 isoform is ubiquitously expressed in the kidney (14), and there is indirect evidence supporting stimulation of NHE1 by insulin in cultured renal cells (27). Although NHE2 is expressed in the kidney (19, 59), its functional role is still enigmatic (20) and may only be activated under certain circumstances. We have ruled out the role of both NHE1 and NHE2 in mediating the insulin-induced increase in Na⁺/H⁺ exchange activity. We showed that insulin specifically upregulates proximal tubule NHE3, which likely mediates the increased proximal tubule Na⁺ absorption in response to insulin. The stimulation of NHE3 by insulin has two characteristics. First, it occurs in a biphasic fashion. Second, it is amplified by glucocorticoids.

Na⁺/H⁺ exchangers are regulated by a wide variety of agonists through vastly different mechanisms. Regulation at the level of transcription (4, 5, 7, 11, 17, 18, 37), translation (67), protein trafficking (2, 21, 23,25, 26, 33, 35, 36, 41, 43, 49, 68-71), phosphorylation (42, 50, 64-66, 72, 73), and binding to protein (12) or lipid regulators (1) has been implicated or proven. A single condition or agonist can regulate NHE3 at more than one step. This has been shown for acid incubation (4, 5, 7, 67, 68), parathyroid hormone (23,26), and dopamine (33, 65). The induction of NHE3 activation by insulin is time dependent, as a significant increase in dpH_i/dt was detectable only at 2 h and beyond. After 12 h, NHE3 activity is clearly increased whereas surface NHE3 protein abundance is still unchanged in insulin-treated cells. The possibility remains that the dpH_i/dtassay is more sensitive than the biotinylation assay. Alternatively, a more plausible explanation is that other posttranslational mechanisms may be operative and contribute to the stimulation of Na⁺/H⁺ exchange by insulin (23, 26,44). A biphasic response has previously been described for parathyroid hormone (23, 26) and dopamine (33,65) involving changes in transport activity of surface NHE3 followed by internalization of NHE3 protein. However, in those two situations, the decrease in NHE3 surface protein commences after a relatively short time. In the case of insulin, surface NHE3 activity is increased without changes in surface NHE3 protein for over 12 h. At present, the mechanism of how insulin induces and sustains this suppression of surface NHE3 transporters is unknown. After 24 h of incubation with insulin, one can see concomitant increases in surface and total NHE3 protein abundance that approximate but are not equal to the magnitude of increase in NHE3 activity. The slightly higher increase in surface NHE3 compared with total NHE3 may reflect an additional step, whereas the increased cellular pool of NHE3 is preferentially targeted to the cell membrane. Moreover, an increase in NHE3 protein is associated with an increase in NHE3 mRNA at 24 h. This pattern of coordinated upregulation at the levels of activity, surface protein, total protein, and mRNA is reminiscent of the effects of thyroid hormone on NHE3 (18).

The stimulation of intrinsic NHE3 activity in the early phase and the increase in NHE3 activity, protein, and mRNA in the late phase are all enhanced by hydrocortisone. At 10⁻⁹ M, where glucocorticoid itself has no effect on NHE3 activity (11), the presence of glucocorticoid allows insulin to exert its full action on NHE3, hence befitting the classic permissive role described in the pioneering manuscript by Ingle (34) a half-century ago. At 10⁻⁷ and 10⁻⁸ M, when corticosteroids themselves activate NHE3, the presence of insulin further increases NHE3 activity. At this point, hydrocortisone acts more like a biological amplifier as discerned by Granner (29). In a saturating dose of hydrocortisone (10⁻⁵ M), the addition of insulin no longer leads to further stimulation. Whether this is synergism, permission, or amplification, the interactive relationship (both positive and negative) between glucocorticoids and a variety of other agonists is pervasive in mammalian biology (53). In the liver, the ability of glucocorticoids to promote hepatic glycogen synthesis is “proinsulin” (58). In contrast, in skeletal muscle, glucocorticoid decreases the ability of insulin to stimulate glycogen synthesis (15). In the kidney, the acid-induced increase in Na⁺/H⁺ exchange can be abolished by adrenalectomy (38). We have shown that this is a direct effect of glucocorticoids because the acid-induced increase in NHE3 is dependent on the presence of hydrocortisone in the cell culture media during serum deprivation and acid incubation (5,31). Glucocorticoids may represent a more generally permissive agent for regulation of NHE3 in the kidney. The mechanism of the permissive effect of glucocorticoids is presently unknown.

In summary, we have shown that insulin activates the Na⁺/H⁺ exchanger NHE3 in OKP cells. This effect is biphasic in nature, with distinct mechanisms that involve increased activity of existing NHE3 proteins on the cell surface followed later by increased NHE3 transcript and total cellular and surface NHE3 protein. In both phases, the insulin-stimulated increase in NHE3 is enhanced by the presence of glucocorticoids. In conjunction with data from clinical and tubule perfusion studies, we propose that insulin stimulates NHE3 and proximal tubule Na⁺ absorption and contributes to the volume expansion and hypertension seen in insulin-resistance states.

FOOTNOTES

• Address for reprint requests and other correspondence: P. M. Ambühl, Renal Division, Univ. Hospital, Rämistrasse 100, CH-8091 Zürich, Switzerland (E-mail:patrice.[email protected]usz.ch).

• The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

• March 26, 2002;10.1152/ajprenal.00365.2001