Chloride, not sodium, stimulates expression of the γ subunit of Na/K-ATPase and activates JNK in response to hypertonicity in mouse IMCD3 cells

Abstract

Hypertonicity induced by NaCl, but not by urea or mannitol, up-regulates expression of the γ subunit of Na/K-ATPase in cells of the murine inner medullary collecting duct line (IMCD3) by activation of the Jun kinase 2 (JNK2) pathways. We examined the ionic mediators of the osmosensitive response. An increase in osmolality to 550 milliosmoles per kg of water (mosmol/kgH₂O) for 48 h by replacement of NaCl with choline chloride did not prevent the up-regulation of the γ subunit. Neither Na⁺ ionophores nor inhibitors of cellular Na⁺ uptake altered the up-regulation of the γ subunit or JNK activation. Changes in cell cation concentrations driven by incubation in low-K⁺ medium were effective in up-regulating the α₁ subunit of Na/K-ATPase but did not have any effect on the γ subunit. The replacement of NaCl with choline chloride did not down-regulate γ-subunit expression in cells adapted to hypertonicity. In contrast, the replacement of NaCl with sodium acetate, or pretreatment of cells with the Cl^- channel inhibitor 5-nitro-2-(3-phenylpropyl-amino)benzoic acid (NPPB) completely blocked γ-subunit up-regulation, inhibited JNK activation, and caused a significant decrement in cell survival in hypertonic but not isotonic conditions. In adapted cells, replacement of 300 mosmol/kgH₂O NaCl with sodium acetate resulted in down-regulation of the γ subunit. In conclusion, we describe a Na⁺-independent, Cl^--dependent mechanism for hypertonicity-mediated activation of the JNK and the subsequent synthesis of the γ subunit of Na/K-ATPase, which are necessary for cellular survival in these anisotonic conditions.

The hypertonicity of the mammalian inner medulla requires the cells that inhabit it to adapt to this environment. Such an adaptation is critical to the operation of the renal concentrating mechanism. It is now established that this adaptation involves the activation of ion transport systems and inert organic osmolytes (1, 2). In addition, it has been recognized that the maintenance of cell viability causes a coordinated response in which a number of proteins, many critical to cell viability, are activated. Among these proteins are heat shock proteins (3), cyclooxygenase 2 (4), the atrial natriuretic peptide receptor (5), and the cystic fibrosis transmembrane conductance regulator (CFTR) (6). We have described that the α₁ and β subunits of Na/K-ATPase are up-regulated by tonicity (7); more recently, we have reported that such up-regulation is also the case with the renal-limited γ subunit of the enzyme (8). The γ subunit appears to localize with the α₁ subunit of Na/K-ATPase in several nephron segments (9). The two splice variants (γ_a and γ_b) have differential expression in various nephron segments (10). The γ_a variant has been reported recently to be present in the inner medulla,† whereas both γ_a and γ_b appear to be present under hypertonic conditions in cultured IMCD3 cells (8). We have reported also that the c-Jun N-terminal kinase (JNK) member of the mitogen-activated protein (MAP) kinase family, and particularly the JNK2 isoform as well as the phosphatidylinositol 3-kinase pathways, appears to be involved in the regulation of the γ-subunit synthesis (8). The observation that inhibiting these signaling pathways by either genetic (12) or pharmacologic (8) means causes a parallel decrement in γ-subunit synthesis and cell survival strongly suggested the important role of this protein in the survival process. Furthermore, our observation that we could not adapt inner medullary collecting duct cells chronically to hypertonicity with mannitol alone (7) and the failure to induce γ-subunit synthesis with this impermeant sugar (8) led us to hypothesize that it is one of the components that is involved in the regulation of this subunit of Na/K-ATPase. Therefore, the purpose of the present experiments was to define the ionic mediator responsible for osmosensitive response and to assess whether the same or a different mediator is responsible for the c-JNK activation that occurs in response to hypertonicity.

Experimental Procedures

Materials. Cell culture medium, serum, and antibodies were obtained from Invitrogen. Antibodies to the α₁ subunit of Na/K-ATPase were purchased from Upstate Biotechnology (Lake Placid, NY). Antibodies to the γ subunit of Na/K-ATPase (γ_a and γ_b, splice variants) were provided by Steven Karlish (The Weizmann Institute of Science, Rehovot, Israel). Ion channel inhibitors, such as amiloride and 5-nitro-2-(3-phenylpropyl-amino)benzoic acid (NPPB), sodium ionophores (gramicidin and monensin), as well as alternative chloride and sodium salts were obtained from Sigma. Osmolality was determined with an Advanced Instruments MicroOsmometer (model 3300, Nor-wood, MA).

Cell Culture. The established murine inner medullary collecting duct cell line (IMCD3) was provided by Steve Gullans (Harvard University, Boston). Cells were propagated as described previously (7). For cells adapted to hypertonic conditions, the medium osmolality was adjusted to the appropriate values with a sterile 5 M NaCl stock solution. When media with other salts were used, extreme care was taken to avoid pH changes. For exposure of cell cultures to low-K⁺ conditions, a low-K⁺ medium was prepared as described by Bowen and McDonough (13). Cell survival was measured by using the CellTiter 96 cell proliferation assay (Promega), as described previously (8).

Treatment and Inhibitors. Confluent cell cultures were preincubated for 24 h in low-serum medium (0.4% FBS). The medium was then replaced with low-serum medium containing the appropriate concentration of inhibitor (1,000× stock solution in DMSO). After 2 h, the osmolality was increased to 550 milliosmoles per kg of water (mosmol/kgH₂O) (osmotic shock) by the addition of sterile 5 M NaCl solution. Conversely, IMCD3 cultures adapted to 600 mosmol/kgH₂O were returned to isotonic media (300 mosmol/kgH₂O) or 600 mosmol/kgH₂O by employing alternative salts. Cultures were harvested after further incubation for 48 h for determination of α₁ and γ subunits, or for 15 min before determination of JNK activity. For UV activation of JNK, cells were exposed for 1 min to UV radiation, as described previously (7), followed by a 30-min incubation.

JNK Activity Measurement. JNK activity was measured as described (7) with the following modification. After polyacrylamide gel electrophoresis (PAGE), proteins were detected by copper staining (14), the bands corresponding to c-Jun were excised and incubated in scintillation vials with 5 ml of 0.1 M acetic acid for 1 h, and radioactivity was counted by Cerenkov radiation.

Western Analysis. Cell lysates were obtained as described previously (7). Protein content was determined with the BCA protein assay (Pierce). Either 50 μg (α₁ subunit) or 200 μg (γ subunit) of protein per lane was loaded in 7.5% (α₁ subunit) or 16.5% (γ subunit) polyacrylamide gels. Electrophoresis in Tricine/Tris/SDS buffer system was performed as described previously (15). Electroblotting to poly(vinylidene difluoride) (PVDF) transfer membrane (NEN), blocking, incubation with antibodies, and washing were done as described previously (7). Alkaline phosphatase was detected by incubating with Lumi-Phos reagent (Pierce). Chemiluminescence was recorded and analyzed as described (7).

Statistics. Results were analyzed with the INSTAT software package (GraphPad, San Diego). Multiple group comparisons used an analysis of variance. A value of P < 0.05 was considered significant.

Results

Studies on the Role of Na⁺. Effect of replacement of NaCl with choline chloride on the acute increment in the γ subunit. In line with previous reports, we were unable to detect the γ subunit of Na/K-ATPase in cells grown in isotonic conditions. Exposure of IMCD3 cells to 550 mosmol/kgH₂O of NaCl caused a clear expression of the protein. However, this stimulation was not altered when the concentration of Na⁺ was not increased because choline chloride provided the same robust response. This observation suggests that an increment in Na⁺ concentration is not necessary for stimulation of the γ-subunit synthesis.

Effect of decrement in [Na⁺] on the down-regulation of the γ subunit in IMCD3 cells adapted chronically to hypertonicity. We have shown previously that IMCD3 cells adapted to live chronically in high tonicity environments express the γ subunit and that a return to isotonic conditions (150 mM NaCl) down-regulates the protein (8). We investigated whether this down-regulation is mediated by a decrease in Na⁺ from 300 to 150 mM. As depicted in Fig. 2, the replacement of 150 mM Na⁺ by choline while total osmolality remained at 600 mosmol/kgH₂O did not decrease significantly the expression of the γ subunit.

Fig. 2.

Effect of cation substitution on γ-subunit expression in adapted IMCD3 cells. Confluent IMCD3 cells adapted to 600 mosmol/kgH₂O were changed to media substituting choline for sodium or reducing added NaCl to isotonic conditions. Cell cultures were harvested after 48 h of incubation for Western blot analysis. Data represent the mean and SEM (n = 4). The differences seen in the levels of γ_a between NaCl and choline chloride were not statistically significant (P > 0.27). A representative blot is shown.

Effect of inhibition of Na⁺ transport and Na⁺ ionophores on hypertonicity mediated α₁- and γ-subunit expression. Inner medullary collecting duct cells transport Na⁺ by an amiloride-sensitive pathway (16, 17). We tested whether inhibition of the epithelial Na⁺ channel alters the ability of hypertonicity to induce the γ subunit. As is depicted in Fig. 3, this ability was not altered because the stimulation was not altered by concentrations of amiloride that exceed those inhibiting Na⁺ uptake (17) in these cells. In contrast, this concentration of amiloride prevented an increase in the α₁ subunit of the Na/K-ATPase in response to hypertonicity. Likewise, we tested whether agents that increase cell Na⁺ (Na⁺ ionophore) stimulate γ-subunit expression. Neither gramicidin (18) nor monensin (19) had this effect (data not shown).

Fig. 3.

Effect of epithelial Na⁺ channel inhibitor on the stimulation of the α₁ and γ subunit of Na/K-ATPase upon osmotic shock (H) with NaCl. Confluent IMCD3 cell cultures were preincubated for 2 h with 20 μM amiloride followed by osmotic shock (H) with added NaCl to 550 mosmol/kgH₂O for 48 h. Western blot analysis data represent the mean and SEM (n = 4). The increment detected between control and amiloride-treated cells are significant neither for γ_a (P > 0.38) nor for γ_b (P > 0.07). In contrast, H mediated α₁-subunit expression was abolished in the presence of amiloride. A representative blot for the γ subunit is shown.

Effect of physiologically driven increment in cell [Na⁺] on γ-subunit up-regulation. Bowen and McDonough (13) demonstrated previously that exposure of kidney cells to low K⁺ medium (0.25 mM) produces a rapid change in cell ion concentration. Intracellular Na⁺ increases, whereas K⁺ decreases simultaneously with the up-regulation of the Na/K-ATPase. As depicted in Fig. 4, a time-dependent up-regulation of the α₁ subunit is detected upon exposure of IMCD3 cells to 0.25 mM K⁺-containing medium. In contrast, the γ subunit is not detected even after 72 h of incubation. At that time, the γ subunit is noticeably present in cells challenged with NaCl (see Fig. 4).

Fig. 4.

Effect on the expression of α₁ and γ subunits of Na/K-ATPase in IMCD3 cells upon challenge with low K⁺ medium. Confluent IMCD3 cultures were incubated with medium containing 0.25 mM K⁺ for 24, 48, and 72 h or challenged by the addition of NaCl to 550 mosmol/kgH₂O for 72 h. Protein samples (30 μg per lane for the α₁ subunit and 200 μg per lane for the γ subunit) were analyzed by Western blot. Data represent the mean and SEM (n = 4). A 32% increment in the level of α₁ subunit was determined after 24 h of incubation in low-K⁺ medium. In comparison with control cells this difference is highly significant (P < 0.01). In contrast, the low-K⁺ medium does not cause expression of the γ subunit. A representative blot for the γ subunit is shown.

Effect of inhibition of Na⁺ transport on the hypertonicity-stimulated JNK kinase activation. Because hypertonicity is well recognized as a stress signal leading to activation of all MAP kinase family members, we tested whether this activation would be altered when cellular Na⁺ uptake is inhibited. As is depicted in Fig. 5, amiloride did not alter JNK activation by hypertonicity. Also, in data not shown, the above-mentioned Na⁺ ionophores, gramicidin and monensin, did not activate the kinase in isotonic conditions.

Fig. 5.

Effect of epithelial Na⁺ channel inhibitor on the stimulation of JNK activity in IMCD3 cells by hypertonicity (H). Confluent IMCD3 cell cultures were preincubated for 2 h with 20 μM amiloride inhibitor followed by osmotic shock (H) with added NaCl to 600 mosmol/kgH₂O for 15 min. Cell lysates were obtained and JNK activity was measured as described in Experimental Procedures. Data represent the mean and SEM (n = 6). The difference between control and amiloride treatment is not significant.

Taken together, these experiments provide little evidence for any role of alterations in cell Na⁺ concentration having an impact on Na/K-ATPase γ-subunit synthesis.

Studies on the Role of Chloride. Effect of replacing NaCl with sodium acetate on the acute up-regulation of the γ subunit. The results obtained upon this substitution are in stark contrast with those obtained with substitution for Na⁺ (Fig. 1). As depicted in Fig. 6, the expression of the γ subunit that is observed clearly at 48 h when tonicity is raised to 550 mosmol/kgH₂O is absent in cells exposed to this tonicity with sodium acetate. In these experiments, pH of the medium was adjusted to prevent any significant deviation.

Fig. 1.

Effect of cation substitution during osmotic shock of IMCD3 cells on γ-subunit stimulation. Stock solutions of sodium chloride or choline chloride were added to culture media to reach a final osmolality of 550 mosmol/kgH₂O. Cell cultures were harvested after 48 h for Western blot analysis. Data in the histogram represent the mean and SEM (n = 4). Differences between sodium chloride and choline chloride were not statistically significant. A representative blot is shown in Lower.

Fig. 6.

Effect of anion substitution during osmotic shock of IMCD3 cells on γ-subunit stimulation. Stock solutions of sodium chloride or sodium acetate were added to culture media to reach a final osmolality of 550 mosmol/kgH₂O; special care was taken to keep medium pH constant. Cells were harvested after 48 h for Western blot analysis. Data represent the mean and SEM (n = 4). A representative blot is shown.

Effect of decrement in [Cl^-] on the down-regulation of the γ subunit in cells adapted chronically to hypertonicity. To determine whether a high Cl⁻ concentration is required to maintain the expression of the γ subunit in the cells adapted to live at 600 mosmol/kgH₂O, the Cl⁻ concentration was decreased from 300 to 150 mM by replacing chloride with acetate. As is depicted in Fig. 7, this maneuver was associated with down-regulation of the γ subunit similar to that seen when isotonic conditions are present, despite the fact that total osmolality was maintained at 600 mosmol/kgH₂O.

Fig. 7.

Effect of anion substitution on γ-subunit expression in adapted IMCD3 cells. Confluent IMCD3 cell cultures adapted to 600 mosmol/kgH₂O were changed to media substituting acetate for chloride or reducing NaCl addition to isotonic conditions. Cell cultures were harvested after 24 h of incubation for Western blot analysis. Data represent the mean and SEM (n = 4). A representative blot is shown. The difference between isotonic and sodium acetate conditions was not significant (P > 0.17). The difference between NaCl-adapted and either isotonic or sodium acetate conditions was very significant (P < 0.005).

Effect of Cl^- channel inhibitor on the expression of the α and γ subunits of Na/K-ATPase. To define further the role of chloride as a potential mediator of the osmolality-mediated up-regulation of the γ subunit, we used the chloride channel inhibitor NPPB. As depicted in Fig. 8, this agent prevents hypertonicity-stimulated γ-subunit production. At 48 h, the concomitant incubation with 10 or 25 μM NPPB caused a highly significant decrement (P < 0.001) in the expression of this subunit, whereas no significant decrement was evident for the α₁ subunit of the enzyme.

Fig. 8.

Effect of chloride channel inhibitor on the expression of α₁ and γ subunits of Na/K-ATPase in IMCD3 cells upon challenge by osmotic stress. Confluent IMCD3 cultures were preincubated with increasing concentrations of NPPB followed by challenge by the addition of NaCl to 550 mosmol/kgH₂O. Cell cultures were harvested for Western blots after 48 h of incubation. Data represent the mean and SEM (n = 4). No significant difference in the α₁ subunit was noted (P > 0.37), whereas a significant difference was detected between the no NPPB and 10 or 25 μM NPPB for both γ-subunit variants (P < 0.001).

Effect of chloride channel inhibitor on hypertonicity and UV-light-stimulated JNK activation. Fig. 9 depicts the results of three studies in which the activation of JNK by hypertonicity was examined in the presence and absence of increasing concentrations of NPPB. The 50% inhibition is observed at 50 μM. The concentration at which inhibition of γ-subunit synthesis was observed caused only a modest decrease in JNK activity. However, the former experiment reflects 48-h exposure to the drug at the maximum concentration tolerated by the cells for this length of time. The much shorter exposure required for the kinase assay permits the use of high drug concentration. In contrast to the inhibition observed in response to the osmotic stress, even the largest concentration of NPPB (100 μM) used caused no change in UV-stimulated JNK activation.

Fig. 9.

Effect of chloride channel inhibitor on JNK activity under different stimuli. Confluent cell cultures were preincubated for 2 h with increasing concentrations of NPPB followed by either osmotic shock to 600 mosmol/kgH₂O with added NaCl, or 1 min of UV irradiation. Cell cultures were harvested after 15 min for JNK activity measurements. Data represent mean and SEM (n = 3). ANOVA demonstrates a very significant difference at all NPPB concentrations (P < 0.0001) over the control (no NPPB) in contrast to differences by UV irradiation, which were not significant.

Effect of chloride channel inhibition on the survival of IMCD3 cells in hypertonic conditions. To assess whether the inhibition of the γ subunit associated with the exposure to NPPB altered the survival of the cells, we undertook measures of cell viability. As depicted in Fig. 10, the exposure to NPPB decreased the survival of cells under hypertonic conditions at concentrations as low as 10 μM. The drug failed to affect cell survival in isotonic conditions.

Fig. 10.

IMCD3 cells were grown to 95% confluence in 24-well plates followed by a change to low-FCS medium (0.4%) for 48 h, then treated for 2 h with NPPB at the indicated concentration and challenged for 24 h by increasing the osmolality of the medium to 550 mosmol/kgH₂O with NaCl. Cell survival was determined by using the CellTiter 96 assay, as described in the text. Data points represent the mean and SEM of four independent experiments normalized to untreated cells. ANOVA reveals a statistically significant (P < 0.0001) difference in survival between cells in isotonic and hypertonic conditions.

Taken together, and in marked contrast to the experiments designed to change Na⁺ concentration, the studies in which we altered the concentration of Cl⁻ had a strong impact on the synthesis of the γ subunit and cell survival.

Discussion

The ever-changing, hypertonic environment of the mammalian inner medulla calls for the cells living therein to adapt to this inhospitable milieu. The cellular responses that permit the survival of the cell have been the subject of considerable attention in the last decade. This response calls for the generation of a number of inert osmolytes coupled to the uptake of Na⁺ (20, 21). The initial step in the adaptive process involves both the activation of MAP kinase pathways (22) and the synthesis of cellular transport proteins (1). In the experiments described in this article, we attempted to link these early events with the subsequent up-regulation of the γ subunit of Na/K-ATPase that we recently described as a potentially vital osmo-regulated response (8).

Our initial studies strongly suggested that the initial MAP kinase response is osmotically driven (22). However, our subsequent studies demonstrated that the ultimate chronic adaptation to the hypertonic environment could not be accomplished solely by other solutes in the absence of added NaCl (7). Thus, we undertook the present studies to determine the ionic mediator responsible for the activation of pathways leading to the γ-subunit synthesis. Despite the fact that both microdissected IMCD3 cells (23) and cultured IMCD3 cells (17, 24) have the epithelial Na⁺ channel, we could gather no evidence that Na⁺ entry into the cells was involved in the regulation of the γ-subunit expression. Thus, neither the acute substitution of NaCl for choline chloride nor the exposure of the cells to the epithelial Na⁺ channel inhibitor, amiloride, prevented the increment in γ-subunit expression when cells were exposed to 550 mosmol/kgH₂O of NaCl for 48 h. In contrast, amiloride prevented α₁-subunit up-regulation under hypertonic conditions. Furthermore, the substitution of NaCl for choline chloride did not cause a decrement in the high level of the γ subunit observed in cells adapted chronically to hypertonic conditions. Increasing intracellular Na⁺ concentration either by sodium ionophores or by physiological maneuvers (low-K⁺ media) was insufficient to up-regulate the γ subunit in cells living under isotonic conditions. We observed that, in parallel with these results, treatment with amiloride did not alter the JNK kinase activating effect of hypertonicity. Taken together, these experiments provide no evidence for a role of increased cell Na⁺ as the mediator of the γ-subunit response. These results could be considered unexpected because maneuvers that increase cell Na⁺, such as depletion of extracellular K⁺ or the use of Na⁺ ionophores, have been described to increase Na/K-ATPase activity (13). It is, thus, very possible that the mechanisms that control the synthesis of the α₁ and β subunits of Na/K-ATPase are distinct from those mechanisms involved in the regulation of the γ-subunit synthesis.

The present experiments provide strong evidence that the effector ion in the control process may be chloride. Thus, unlike the foregoing substitution of Na⁺, the substitution of Cl⁻ by another anion caused a marked decrement of the γ subunit in osmotically adapted cells despite the maintenance of medium tonicity at 600 mosmol/kgH₂O. This substitution also prevented entirely the synthesis of the new protein following acute exposure to hypertonicity. Furthermore, the chloride channel blocker NPPB (25), at concentrations within the IC₅₀ (26), prevented hypertonicity-mediated increment in γ-subunit expression. The effect of NPPB on the γ subunit is specific because no significant decrement in α₁ subunit of Na/K-ATPase was observed. This observation provides yet another line of evidence for separated regulation of the α₁ and γ subunits of the enzyme. The doses of NPPB used were also associated with a decrement in cell survival and in hypertonic media, but not in isotonic conditions. Finally, NPPB prevented hypertonicity-, but not UV-light-stimulated, JNK activation. The concentrations used to inhibit the activation of the kinase by 50% (50 μM) cannot be sustained by cells for the length of time necessary to study the expression of the γ subunit. It is of interest that the concentration used for such studies caused only a modest acute suppression of JNK activation. We have shown previously, by employing a dominantnegative mutant of JNK2, that this activation is ultimately necessary for full expression of the γ subunit (8). It must be noted also that there are four JNK2 splice variants that could be activated differentially by chloride currents and/or that are involved to different degrees in the up-regulation of the γ subunit. Nevertheless, the present experiments, employing NPPB (Figs. 8, 9, 10), add to earlier evidence that links the importance of intact γ-subunit synthesis to cell survival; the present experiments further suggest that Cl⁻ may initiate the adaptive cascade.

The presence of electrogenic Cl⁻ secretion, which is sensitive to chloride channel blockers, has been described previously in cultured IMCD3 cells (23). This channel may be the one that allows Cl⁻ entry into the cell and sets into motion the sequence of events involving JNK activation and subsequent osmoprotective proteins expression, which is critical to cell survival. It must be noted that, despite the presence of a NaK/2Cl cotransporter in these cells (27), we could not mimic the above-described results seen with the chloride channel inhibitors, with bumetanide. Kizer et al. (24) also saw very modest inhibition of Cl⁻ entry in these cells when exposed to either furosemide or bumetanide. It is likely that this cotransporter plays a secondary role in IMCD3 cells. That this is likely to be the case also in vivo is supported by the inhibition of 8-Br-cAMP mediated fluid secretion in perfused inner medullary collecting ducts by chloride channel inhibitors (23), although in this setting inhibition was also seen with bumetanide. A role for chloride channels in osmoregulation has been suggested by their activation by hypertonicity in neuroblastoma cell lines (28), human intestinal cells (29), and HeLa cells (30). Chloride uptake also plays a critical role in the function of the juxtaglomerular apparatus, at least concerning the regulation of the cyclooxygenase 2 expression (31, 32). Whether a similar Cl⁻-dependent mechanism is operant in cyclooxygenase 2 expression by collecting duct cells (4) has not been explored.

In summary, our studies provide strong evidence for the presence of a chloride-dependent, but not sodium-dependent activation of JNK kinase in response to hypertonicity with a subsequent profound impact on the expression of the γ subunit of Na/K-ATPase. Our studies provide further evidence, suggested partially in our previous studies (8), for independent regulation of the γ and α₁ subunits of the Na/K-ATPase, with the former being Cl⁻ dependent and the latter Na⁺ dependent. We speculate that there may be a distinct advantage for the cells to have two different mechanisms for regulation of the α₁/β and the γ subunits. Because the γ subunit appears to reduce sodium affinity (33), if the γ subunit responded to an increase in cellular sodium as the α₁ and β subunits do, the overall effect on pump rate would be abrogated. We furthermore demonstrated that a critical role for a chloride-dependant pathway, most likely involving the γ subunit localized recently to the mammalian inner medulla,† is important to the survival of IMCD3 cells in hypertonic conditions. Because this process is crucial to the intact operation of the urinary concentrating system, it is attractive to postulate that the concentrating defect observed with some hypoglycemic agents such as glyburide (11) that block Cl⁻ channels could be related to a failure of the hypertonicity adaptive mechanism.