ERK1/2 Controls Na,K-ATPase Activity and Transepithelial Sodium Transport in the Principal Cell of the Cortical Collecting Duct of the Mouse Kidney*

The collecting duct of normal kidney exhibits significant activity of the MEK1/2-ERK1/2 pathway as shown in vivo by immunostaining of phosphorylated active ERK1/2 (pERK1/2). The MEK1/2-ERK1/2 pathway controls many different ion transports both in proximal and distal nephron, raising the question of whether this pathway is involved in the basal and/or hormone-dependent transepithelial sodium reabsorption in the principal cell of the cortical collecting duct (CCD), a process mediated by the apical epithelial sodium channel and the basolateral sodium pump (Na,K-ATPase). To answer this question we used ex vivo microdissected CCDs from normal mouse kidney or in vitro cultured mpkCCDcl4 principal cells. Significant basal levels of pERK1/2 were observed ex vivo and in vitro. Aldosterone and vasopressin, known to up-regulate sodium reabsorption in CCDs, did not change ERK1/2 activity either ex vivo or in vitro. Basal and aldosterone- or vasopressin-stimulated sodium transport was down-regulated by the MEK1/2 inhibitor PD98059, in parallel with a decrease in pERK1/2 in vitro. The activity of Na,K-ATPase but not that of epithelial sodium channel was inhibited by MEK1/2 inhibitors in both unstimulated and aldosterone- or vasopressin-stimulated CCDs in vitro. Cell surface biotinylation showed that intrinsic activity rather than cell surface expression of Na,K-ATPase was controlled by pERK1/2. PD98059 also significantly inhibited the activity of Na,K-ATPase ex vivo. Our data demonstrate that the ERK1/2 pathway controls Na,K-ATPase activity and transepithelial sodium transport in the principal cell and indicate that basal constitutive activity of the ERK1/2 pathway is a critical component of this control.

The extracellular signal-regulated protein kinase 1 and 2 (ERK1/2) 1 pathway, also known as p42 and p44 mitogen-acti-vated protein (MAP) kinase pathway, plays a critical role in cellular responses to a wide variety of external stimuli including hormones, growth factors, and environmental stress (for review, see Refs. 1 and 2). The ERK1/2 pathway is a three-step kinase cascade in which ERK1/2 kinases are phosphorylated/ activated by two immediate upstream MAP kinases (MEK1 and MEK2 (MEK1/2)) that are in turn phosphorylated/activated by several MAP kinases (i.e. MOS, C-Raf, and B-Raf). Activated ERK1/2 (pERK1/2) phosphorylates various downstream effectors involved in differentiation, proliferation, apoptosis, survival, cellular metabolism, and ion transport. The MAP kinase cascade may also be constitutively activated independent of external stimuli or ligands. This basal activity is the result of the balance between positive and negative factors regulating MAP kinases activity; that is, phosphatases, subcellular localization, expression of scaffold proteins, and activity of other kinases. In the kidney, the ERK1/2 pathway is involved in the hormonal regulation of different apical and basolateral channels or transporters expressed along the nephron for maintaining the extracellular fluid homeostasis. In the proximal tubule the ERK1/2 pathway is involved in the internalization of the Na ϩ /phosphate co-transporter at the apical membrane and in the activation of Na,K-ATPase at the basolateral membrane (3,4). In the thick ascending limb the ERK1/2 pathway is also involved in the inhibition of basolateral Na ϩ /H ϩ exchange activity and HCO 3 Ϫ absorption by nerve growth factor (5). In the ␤-intercalated cells of the collecting duct, the ERK1/2 pathway plays a significant role in the isoproterenol-dependent activation of H,K-ATPase (6,7). In the principal cell of the collecting duct, the rate of sodium reabsorption is continuously adapted to maintain the osmolarity and the volume of the extracellular body fluid within a physiological range. Aldosterone and vasopressin play a major role in the control of sodium and water reabsorption in the principal cell. In this cell sodium enters the cell from the tubular lumen via an apical amiloride-sensitive epithelial sodium channel (ENaC) (8) and is then extruded out of the cell into the extracellular fluid by the basolateral Na,K-ATPase (9). Importantly, several studies have demonstrated that ENaC or Na,K-ATPase are directly or indirectly targets for ERK1/2 activity (4, 10 -13). Interestingly, immunohistochemical studies have recently revealed that the phosphorylated active form of ERK1/2 (pERK1/2) is expressed at detectable basal levels in the distal (but not in the proximal) nephron of normal rat kidney in vivo (14 -17). In the collecting duct Masaki et al. (16) have recently shown that pERK1/2 is expressed in the principal cell and not in the intercalated cells. These in vivo data suggest that basal activity of ERK1/2 may have a specific physiological relevance in the principal cell of the collecting duct, independent of hormonal or ligand-induced stimuli and distinct from the hormonedependent effects observed in intercalated cells.
To address this question we investigated the role of ERK1/2 activity on transepithelial sodium transport in the principal cell of ex vivo microdissected cortical collecting duct (CCD) tubules or in vitro in mpkCCD cl4 cells, a well differentiated cell culture model of the principal cell (18). The aim of this study was 2-fold; first, to assess the role of ERK1/2 activity in basal and hormone-stimulated transepithelial sodium transport in the principal cell, and second, to examine whether ENaC and/or Na,K-ATPase activity is controlled by pERK1/2. To inhibit the MEK1/2-ERK1/2 pathway in the principal cell we used three structurally unrelated inhibitors (PD98059, U0126, and SL327) of MEK1/2, allowing correlation of the physiological response (sodium transport) with the biochemical response (degree of phosphorylation of ERK1/2). We show that the MEK1/2-ERK1/2 pathway constitutively activates Na,K-ATPase but has no effect on ENaC and that the activation of Na,K-ATPase by the pERK1/2 pathway is required for both basal and aldosterone-or vasopressin-stimulated sodium transport. Our data also indicate that the activities of Na,K-ATPase and transepithelial sodium reabsorption in the principal cell are controlled by the constitutive or ligand-independent basal activity of ERK1/2.

MATERIALS AND METHODS
Isolation of Different Parts of the Nephron-Experiments were performed on male NMRI mice that had free access to water and standard laboratory chow until anesthesia. Isolated tubules were obtained by microdissection of the left kidney perfused with 40 g/ml Liberase Blendzyme 2 (Roche Applied Science) dissolved in DMEM/F-12 (1:1) medium (21041 medium, Invitrogen). Thin pyramids cut along the corticomedullary axis were incubated at 37°C for 40 min in aerated DMEM/F-12 (1:1) medium containing 40 g/ml Liberase. Microdissection was performed in ice-cold DMEM/F-12 (1:1) medium as described (19). The following structures were isolated: proximal tubule, medullary, and cortical portions of the thick ascending limb, distal convoluted tubule, connecting tubule, and cortical and outer medullary portions of the collecting duct. Tubular length was measured with an ocular micrometer (Zeiss), and pools of CCDs containing 10 -20 microdissected tubules with the total tubular length of ϳ10 mm/pool were transferred in 5 l of DMEM/F-12 (1:1) medium into 1.5 ml of Eppendorf tubes. Stimulation with hormones was performed by adding 5 l of doubly concentrated hormonal solutions in the same medium. The incubation was performed at 35°C. Reactions were stopped by the addition of 10 l of doubly concentrated protein sample buffer (4.8% Western Blotting on Microdissected Nephron Segments-Samples were heated at 95°C for 5 min and loaded and electrophoresed on a 13% SDS-PAGE. Proteins were then transferred to Protran nitrocellulose membrane (Schleicher & Schuell), and a Western blot was performed according to a standard procedure. Both anti-phospho-ERK1/2 monoclonal antibody (Santa Cruz) and anti-total-ERK1/2 monoclonal antibody (Zymed Laboratories Inc.) were used at 1:50 dilution. The antiactin antibody (Sigma) was used at a 1:250 dilution. The membrane was first probed with anti-phospho-ERK1/2 antibody and, after stripping, with anti-total-ERK1/2 and anti-actin antibodies. Blots were revealed with SuperSignal reagent (Pierce).
Cell Culture-The mpkCCD cl4 cells were routinely grown in plastic tissue culture flasks in a modified DMEM/F-12 (1:1) medium (31331 medium, Invitrogen) supplemented with 60 nM sodium selenate, 5 g/ml transferrin, 2.5 nM dexamethasone, 1 nM triiodothyronine, 10 ng/ml epidermal growth factor (EGF), 5 g/ml insulin, 11 mM D-glucose, 2% fetal calf serum, 10 mM HEPES, pH 7.4, 100 units/ml penicillin, 100 g/ml streptomycin (referred to as Complete medium) at 37°C in a 5% CO 2 , 95% air atmosphere. For electrophysiological and biochemical experiments, the mpkCCD cl4 cells were seeded and grown in the same medium on collagen-coated filters (Snapwell or Transwell, 0.4-m pore, 1 or 4.7 cm 2 , respectively; Corning Costar) until the cell monolayer reached transepithelial resistance greater than 000 ohm x cm 2 . After 5 days confluent cells were placed in identical medium but deprived of EGF, transferrin, and fetal calf serum for 5 days (referred to as Light medium). Eighteen hours before experiments filters were placed in DMEM/F-12 (1:1) medium supplemented with 11 mM D-glucose, 100 units/ml penicillin, and 100 g/ml streptomycin (referred to as Minimal medium).
Electrophysiological Measurements of Transepithelial Short Circuit Current (I sc ) in mpkCCD cl4 Cells Grown on Snapwell Filters-Transepithelial short circuit currents (I sc ) were recorded on confluent mpkC-CD cl4 cells grown on Snapwell filters and mounted in Ussing chambers, as described (20). Briefly, the epithelium was maintained under current-clamp conditions instead of voltage-clamp conditions to avoid high transepithelial Na ϩ flux, which could saturate the transport capacities of the cells. The transepithelial resistance (R te , kiloohms) was calculated from Ϯ10-A pulses of 20 ms duration elicited by a computercontrolled multichannel voltage-clamp apparatus (Physiological Instruments). The equivalent I sc (A/cm 2 ) was calculated from the measured transepithelial voltage (V te ) and the transepithelial resistance (I sc ϭ V te / R te ). In a standard experiment the I sc was simultaneously measured on six independently mounted Snapwell filters. The mean basal current of 26 filters used was 24.2 Ϯ l4.8 A/cm 2 (n ϭ 26).
ERK1/2 Western Blotting in mpkCCD cl4 Cells-Immediately after the last I sc measurement, cells were scraped into 80 l of Snapwell SDS-PAGE protein sample buffer. Samples were then sonicated for 15 and heated at 95°C for 5 min, and 20 l were loaded and electrophoresed on a 13% SDS-PAGE. Proteins were transferred for Western blotting, as described above. A pool of active or phosphorylated ERK1/2 was detected using anti-phospho-ERK1/2 monoclonal antibodies (Santa Cruz) at a 1:1000 dilution. After stripping, the same membrane was used for detection of the total pool of ERK1/2 using anti-ERK1/2 monoclonal antibody (Zymed Laboratories Inc.) at a 1:1000 dilution.
Measurements of Na,K-ATPase Current in Apically Permeabilized mpkCCD cl4 Cells-After incubation for at least 18 h in Minimal medium, confluent mpkCCD c14 cells were transferred to a Ussing chamber in the Na ϩ -free buffer containing 116 mM K ϩ gluconate, 1.8 mM CaCl 2 , 1.6 mM MgCl 2 , 0.8 mM KH 2 PO 4 , 2 mM D-glucose, 12 mM essential amino acids (free acids), 2 mM nonessential amino acids (free acids), 0.4 mM glutamine, 25 mM HEPES, 3 mM BaCl 2 , pH 7.4 (Buffer A). After a 7-min stabilization period, cells were apically permeabilized with 17.5 g/ml amphotericin B for 5 min, as described (21). The Na,K-ATPase was activated by apical and basolateral addition of Na ϩ -containing buffer (60 mM Na ϩ gluconate, 56 mM K ϩ gluconate, 1.8 mM CaCl 2 , 1.6 mM MgCl 2 , 0.8 mM KH 2 PO 4 , 2 mM D-glucose, 12 mM essential amino acids, 2 mM nonessential amino acids, 0.4 mM glutamine, 25 mM HEPES, 3 mM BaCl 2 , pH 7.4) (Buffer B). This raised the apical and basolateral Na ϩ concentrations to 30 mM. After a 3-min stabilization of the Na,K-ATPase current, MEK1/2 inhibitors were added to both apical and basolateral sides. The effect of the inhibitors was monitored for 5 min. At the end of this period 2 mM ouabain was added to the basolateral solution to completely block the pump current.
Measurements of ENaC Current in Basolaterally Permeabilized mp-kCCD cl4 Cells-After incubation for at least 18 h in Minimal medium, confluent mpkCCD c14 cells were transferred to a Ussing chamber in the 100 mM Na ϩ apical buffer containing 100 mM Na ϩ gluconate, 1.8 mM CaCl 2 , 1.6 mM MgCl 2 , 0.8 mM KH 2 PO 4 , 2 mM D-glucose, 12 mM essential amino acids, 2 mM nonessential amino acids, 0.4 mM glutamine, 25 mM HEPES, 10 M amiloride, pH 7.4 (Buffer C), and in the 100 mM K ϩ basolateral buffer containing 100 mM K ϩ gluconate, 1.8 mM CaCl 2 , 1.6 mM MgCl 2 , 0.8 mM KH 2 PO 4 , 2 mM D-glucose, 12 mM essential amino acids, 2 mM nonessential amino acids, 0.4 mM glutamine, 25 mM HEPES, 2 mM ouabain, pH 7.4 (Buffer D). After stabilization of cell monolayer resistance, cells were basolaterally permeabilized with 17.5 g/ml amphotericin B. Then the apical amiloride was washed out by replacement of the apical buffer by 50 ml of buffer C, free of amiloride (Buffer E). After stabilization of I sc (ϳ2-3 min) MEK1/2 inhibitors were added to both apical and basolateral sides. At the end of the experiment the ENaC activity was blocked by 10 M amiloride.
Cell Surface Biotinylation of Na,K-ATPase-Cell surface biotinylation was performed according to the protocol described by Planes et al., with minor modifications (22). mpkCCD c14 cells grown on Transwell filters (2 filters/condition) were placed on ice and washed three times with ice-cold phosphate-buffered saline solution complemented with 0.1 mM CaCl 2 and 1 mM MgCl 2 (Buffer F). Basolateral membrane proteins were biotinylated by a 1-h incubation at 4°C with 1.25 mg/ml NHS-ssbiotin (Pierce) freshly diluted in the biotinylation buffer (10 mM triethanolamine, 2 mM CaCl 2 , 150 mM NaCl, pH 7.5) (Buffer G) and stirred gently. Cells were rinsed with phosphate-buffered saline complemented with 0.1 mM CaCl 2 ,1 mM MgCl 2 , and 100 mM glycine, pH 7.5 (Buffer H), for 20 min at 4°C to quench unreacted biotin. The cells were then rinsed twice with Buffer F, scraped in ice-cold phosphate-buffered saline, and pelleted at 2000 rpm at 4°C. The pellets were solubilized for 45 min in 20 l of lysis buffer (1% Triton X-100, 150 mM NaCl, 5 mM EDTA, 50 mM Tris, pH 7.5 (Buffer I), containing protease inhibitors. The lysates were clarified by centrifugation at 14,000 ϫ g for 10 min at 4°C, and the supernatants were incubated overnight with streptavidinagarose beads to recover biotinylated proteins. The beads were then pelleted by centrifugation, and aliquots of supernatants were taken to determine the unbound, intracellular pool of proteins. Biotinylated proteins were eluted from the beads by heating at 56°C for 7 min in SDS-PAGE sample buffer before loading onto a 10% SDS gel. The a 1 subunit of Na,K-ATPase was detected by Western blotting using the antibody described in Girardet et al. (23). As a control for cell surface versus intracellular protein biotinylation, the same blots were probed with an anti-actin antibody (Sigma).
Measurement of Na,K-ATPase Activity in Microdissected CCDs-The hydrolytic activity of Na,K-ATPase was determined on pools of 4 -6 permeabilized CCDs by a radiochemical microassay, as previously described (24). Tubules were permeabilized by the classical procedure with a freezing-thawing step. Total ATPase activity was measured after the addition of a solution that contained 100 mM NaCl, 5 mM KCl, 10 mM MgCl 2 , 1 mM EDTA, 100 mM Tris-HCl, 10 mM Na 2 ATP, and 5 nCi/l [␥-32 P]ATP (DuPont) (2-10 Ci/mmol) at pH 7.4 (Buffer J). For Na ϩ ,K ϩindependent ATPase activity measurements, NaCl and KCl were omitted, Tris-HCl was 150 mM, and 2 mM ouabain was added. Na,K-ATPase activity (in pmol of ATP/ml of tubule/h) was taken as the difference between total and Na ϩ ,K ϩ -independent ATPase activities.
MEK1/2 Inhibitors-PD98059, U0126, and SL327 were prepared as 10 mM stock solution in Me 2 SO. An equivalent volume of Me 2 SO (0.5% for 50 M inhibitors concentration) was used for the control treatments.

RESULTS
Ex Vivo Expression of pERK1/2 along the Nephron-Mice used in these experiments had free access to water and standard laboratory chow and were not subjected to any special treatment. To assess the pattern of ERK1/2 and pERK1/2 expression along the nephron and to compare it to the previously published immunostaining pattern observed in vivo (14 -17), we microdissected different segments of the mouse nephron. Protein extracts from 10 glomeruli or 10 mm of microdissected tubules (ϳ3000 -5000 cells) were used for Western blotting with anti-pERK1/2, anti-ERK1/2, and anti-actin antibodies. As shown in Fig. 1, pERK1/2 is expressed at detectable levels in the distal convoluted tubule (DCT), connecting tubule (CNT), and CCD but not in the other more proximal nephron segments. On the same blot, the anti-ERK1/2 antibody detected a strong expression of total (active and inactive) pool of ERK1/2 in the proximal tubule but was not sensitive enough to detect ERK1/2 expression in other nephron segments. In microdissected CCDs, the presence of the active form of ERK1/2 in basal conditions was demonstrated in 15 independent experiments and was consistently detected independently of many possible variables including circadian rhythm (the time of mouse perfusion and of microdissection varied from early morning to late afternoon).
To examine whether the basal activity pERK1/2 observed in CCDs could be further stimulated by hormones known to increase transepithelial sodium transport, we stimulated microdissected CCDs with aldosterone (165 min, 10 Ϫ6 M) or vasopressin (90 min, 10 Ϫ8 M). As shown in Fig. 2, A and B, respectively, there was no significant change of pERK1/2 expression (as corrected for actin) induced by aldosterone or vasopressin (110.1 Ϯ 17.6% (n ϭ 5, not significant) or 103.2 Ϯ 7.6% (n ϭ 6, not significant) respectively, compared with unstimulated cells). These observations are consistent with the in vivo immunostaining data.
Expression of pERK1/2 in mpkCCD cl4 Cells-To study the role of the pERK1/2 pathway in regulating sodium reabsorption in the principal cell, we used the mpkCCD cl4 cell line derived from the principal cell of the mouse collecting duct (18). This cell line exhibits a high transmural resistance (Ͻ3000 Ohm⅐cm 2 ) and a high transepithelial Na ϩ transport, which is up-regulated by aldosterone and vasopressin (18). This phenotype is achieved by changing the culture condition in the function of growth or differentiation. For growth, the mpkCCD requires insulin, dexamethasone, T 3 , EGF, and fetal calf serum. For differentiation, the cells are then seeded on Transwell filters to establish polarized monolayers, allowing measurement of transepithelial sodium transport. The medium contained insulin, dexamethasone, T 3 , but no serum for 5 days. Finally, for the electrophysiological measurement, all factors can be removed. The monolayer can keep its sodium transport properties for days cultured in straight DMEM/F-12 medium without any serum or factors added. It is, therefore, an ideal system to study whether basal ERK 1/2 activity could be maintained in the absence of any potential external stimulus. It permits establishing direct relationships between the sodium transport (measured by the short circuit current method) and the biochemical response (by assessing the level of ERK1/2 and its active form). We first addressed the question of whether pERK1/2 is expressed in mpkCCD cl4 cells in basal conditions FIG. 1. Expression of pERK1/2 and ERK1/2 proteins along the nephron. Protein extracts from 10 glomeruli or 10 mm of microdissected nephron segments were loaded and electrophoresed and probed with an anti-pERK1/2 antibody, as described under "Materials and Methods." The same blot was stripped and re-probed with an anti-ERK1/2 antibody. After a second stripping, the blot was re-probed with an anti-actin antibody. Three independent experiments showed the same distribution of ERK1/2 and pERK1/2 along the nephron. MTAL and CTAL, respectively, medullary and cortical portions of the thick ascending limb; DCT, distal convoluted tubule; CNT, connecting tubule; OMCD, and outer medullary portion of the collecting duct. and whether pERK1/2 expression is regulated by aldosterone or vasopressin. In standard experimental conditions we used polarized mpkCCD cl4 cells grown on porous filters incubated in serum-and hormone-free "Minimal medium" for at least 18 h before the experiments (see "Materials and Methods"). As shown in Fig. 3, A and B, the mpkCCD cl4 cells in standard conditions express a detectable basal level of pERK1/2. Up to 3 h of stimulation with aldosterone (10 Ϫ6 M) and up to 4 h of stimulation with vasopressin (10 Ϫ8 M) did not significantly change the levels of pERK1/2 expression (Fig. 3, A and B, respectively) as observed ex vivo.
To examine whether pERK1/2 expression in mpkCCD cl4 cells could be regulated by other "classical" external stimuli, we incubated cells in three different culture media that contain (Complete medium, see "Materials and Methods") or do not contain (Light and Minimal media, see "Materials and Meth-ods") serum and EGF, two known stimulators of ERK1/2 activity. As shown in Fig. 3C, pERK1/2 levels were similar between cells incubated in Light or Minimal media. However, pERK1/2 level was significantly stimulated in cells incubated in the Complete medium (pERK1/2 expression in the Light medium was 148 Ϯ 38% of the pERK1/2 expression in the Minimal medium (n ϭ 4, NS); pERK1/2 expression in the Complete medium was 261 Ϯ 25% of the pERK1/2 expression in the Minimal medium (n ϭ 4, p Ͻ 0.05)). These results demonstrate that the absence of regulation of pERK1/2 expression by aldosterone and vasopressin in mpkCCD cl4 cells is not due to a maximal stimulation of ERK1/2 activity.
PD98059 Reversibly Inhibits Transepithelial Sodium Transport and ERK1/2 Phosphorylation in mpkCCD cl4 Cell-To further examine the role of the ERK1/2 pathway in transepithelial sodium transport in the principal cell, we assessed the effect of PD98059 (a MEK1/2 inhibitor) on transepithelial short circuit current (I sc ). PD98059 is known to inhibit the activation of MEK1 and MEK2 with IC 50 values of 2-7 and 50 M, respectively (25). I sc was measured on confluent mpkCCD cl4 cells grown on Snapwell filters and incubated in the Minimal medium for at least 18 h before experiments. The addition of PD98059 (50 M) to both basolateral and apical sides of unstimulated mpkCCD cl4 cells resulted in a rapid (within minutes) inhibition of basal I sc , a decrease in transepithelial potential difference, and an increase in transepithelial resistance ( Fig. 4A and Table I). This inhibition of I sc was slowly (within 30 min) but completely reversible without washout of the inhibitor. PD98059 did not produce any effect on I sc when ENaC activity was blocked by 10 M amiloride (Fig. 4B). The addition of PD98059 to apical or basolateral sides of mpkCCD cl4 cells produced the same inhibitory effect on I sc as when the inhibitor was added from both sides (data not shown).
Stimulation of mpkCCD cl4 cells with aldosterone or vasopressin significantly increased I sc with time courses characteristic of these hormones; the first significant induction of I sc with aldosterone was detectable after ϳ1 h of stimulation, and the maximal increase of I sc was reached after 3 h of stimulation ( Fig. 4C and Table I), whereas vasopressin rapidly (within minutes) increased I sc , which reached its maximal level after ϳ90 min of stimulation with the hormone (Fig. 4E and Table I). PD98059 added from both apical and basolateral sides of cells significantly and rapidly inhibited both aldosterone-and vasopressin-stimulated I sc , decreased the transepithelial potential difference, and increased the transepithelial resistance (Figs. 4, C and E, respectively, and Table I). This inhibition of aldosterone-and vasopressin-stimulated I sc was maintained over 30 min in the presence of PD98059 and was reversible after washout of the inhibitor (Figs. 4, D and F, respectively). As shown in Table I, the inhibitory effects of 50 M PD98059 were similar between basal and aldosterone-or vasopressin-stimulated I sc (ϳ33, ϳ30, and ϳ38% of inhibition, respectively). Importantly, this reversible inhibition of I sc produced by PD98059 in aldosterone-and vasopressin-stimulated mpkCCD cl4 cells paralleled a reversible decrease in ERK1/2 phosphorylation, as shown in Figs. 5, A and B, respectively. Collectively, these data demonstrate that both the basal and aldosterone-or vasopressinstimulated sodium transport in the principal cell are controlled by the ERK1/2 pathway. I sc Inhibition by PD98059 Correlates with a Decreased Activity of Na,K-ATPase-In the principal cell two major transporters involved in sodium transport are the apical amiloridesensitive sodium channel (ENaC) and the basolateral Na,K-ATPase. Thus, we addressed the question of whether ENaC or Na,K-ATPase activity is controlled by the pERK1/2 pathway. We first tested the effect of PD98059 and two other MEK1/2 inhibitors, namely U0126 and SL327, on Na,K-ATPase transport activity using a protocol developed by Summa et al. (21). In this protocol the apical membrane of mpkCCD cl4 cells is selec-tively permeabilized to monovalent cations with amphotericin B, thus allowing free diffusion of Na ϩ into the cell through the apical membrane. To check the efficiency of apical membrane permeabilization we tested the effect of amiloride (10 M) on I sc measured in permeabilized cells. These control experiments demonstrated that mpkCCD cl4 cells treated from the apical side with amphotericin B (5 min, 17.5 g/ml) do not express any residual amiloride-sensitive I sc , thus indicating that most of the cells are efficiently permeabilized (data not shown). To block the basolateral potassium conductance, all the experiments were performed in the presence of 3 mM Ba 2ϩ in the basolateral solution. In these conditions most of the I sc activated by replacement of Na ϩ -free by 30 mM Na ϩ -containing incubating media (see "Materials and Methods") is due to the activation of the Na,K-ATPase and is termed the Na ϩ pump current (I p ).
As shown in Fig. 6A and in Table II, PD98059 (50 M), U0126 (50 M), and SL327 (50 M) significantly inhibited the I p in unstimulated cells, albeit to a different extent. Importantly, the I p in permeabilized unstimulated cells is inhibited by U0126 with an IC 50 value of 26 M (Fig. 7), which corresponds well with the IC 50 of U0126 in in vitro assays (ϳ13 M) (26). As shown in Fig. 6B and in Table II, pretreatment of cells with aldosterone (180 min) resulted in a significant increase in I p (72 Ϯ 11% of increase (n ϭ 9, p Ͻ 0.001)), whereas I p measured in cells pretreated with vasopressin was not significantly different from that in unstimulated cells (Fig. 6C and Table II). Importantly, the I p in aldosterone-or vasopressin-stimulated cells could be blocked by PD98059 and U0126 to a similar extent as in unstimulated cells (Figs. 6, B and C, and Table II), thereby indicating that activity of the Na,K-ATPase in mpkC-CD cl4 cells is controlled by the MEK1/2-ERK1/2 pathway.
We were concerned by possible nonspecific direct effects of MEK1/2 inhibitors on the Na,K-ATPase. To examine this possibility we tested the effect of U0126 (50 M) on the endogenous Na,K-ATPase of Xenopus laevis oocytes. It is worth noting that the MEK1/2-ERK1/2 pathway in stage VI immature Xenopus oocytes remains inactive until its stimulation by progesterone, which induces oocyte maturation and activation of cell cycle progression. Therefore, this expression system is convenient for testing the possible direct nonspecific effects of MEK1/2 inhibitors on Na,K-ATPase. As shown in Fig. 6D, endogenous Na,K-ATPase is not sensitive to 50 M U0126 in Xenopus oocytes. The inhibitory effects of MEK1/2 inhibitors on Na,K-ATPase activity observed in mpkCCD cl4 cells are, therefore, unlikely to be due to a direct nonspecific inhibition of the Na,K-ATPase.
Effects of PD98059 and U0126 on ENaC Activity in Basolaterally Permeabilized mpkCCD cl4 Cells-To assess the role of  the pERK1/2 pathway in the control of ENaC activity, we permeabilized the basolateral membrane with amphotericin B (5 min, 17.5 g/ml) in a high basolateral K ϩ solution (100 mM K ϩ gluconate). Under these experimental conditions, the amiloride-sensitive sodium current (I Na ) depends only on the large Na ϩ gradient artificially maintained across the apical membrane and not on the Na,K pump activity at the basolateral membrane. This was checked by the addition of 2 mM ouabain to the basolateral solution, which had no detectable effect on the transepithelial current (data not shown). As shown in Fig. 8A, the addition of PD98059 or U0126 to basolaterally permeabilized unstimulated cells did not produce a significant effect on I Na . Pretreatment of cells with aldosterone or vasopressin significantly increased the I Na (326 Ϯ 50% increase with aldosterone (n ϭ 9, p Ͻ 0.001) and 169 Ϯ 7% increase with vasopressin (n ϭ 8, p Ͻ 0.001), Fig. 8, B and C, respectively). No change in I Na was observed upon treatment of aldosterone-or vasopressin-stimulated cells with PD98059 or U0126 (Fig. 8, B and C, respectively). These data indicate that the apical membrane Na ϩ conductance, and thus, ENaC activ- I p from aldosterone-stimulated cells that were not treated with MEK1/2 inhibitors is shown with filled squares (f). I p in the same experiment in control cells (without aldosterone and MEK1/2 inhibitors treatment) is shown as a dotted curve. D, Na,K-pump current (I p ) was measured as described earlier (40). U0126 was added for 4 min at a concentration of 50 M. Na,K-pump inhibition by 10 M ouabain was measured at the end. The mean Na,K-pump current values are normalized to the value measured just before U0126 addition, which was 74 Ϯ 10 nA (n ϭ 4). ity is not controlled by the ERK1/2 pathway in permeabilized mpkCCD cl4 cells.
Effects of PD98059 and U0126 on Cell Surface Expression of Na,K-ATPase-To address the question of whether MEK1/2 inhibitors decrease intrinsic Na,K-ATPase pump activity or its cell surface expression, we performed basolateral cell surface biotinylation of unstimulated mpkCCD cl4 cells treated or not with PD98059, U0126, or SL327 for 10 min and with PD98059 or U0126 for 30 min. The biotinylation was performed on cells grown on two Transwell filters, and total biotinylated proteins were precipitated with streptavidin-agarose beads and electrophoresed through SDS-PAGE. In parallel, one-third of intracellular proteins recovered after precipitation of biotinylated proteins were loaded on the same gel. The proteins were Western-probed with an anti-␣ 1 Na,K-ATPase subunit antibody. As shown in Fig. 9 (upper panel), none of these three inhibitors induced any significant changes in Na,K-ATPase expressed at cell surface (lanes 1-6) or in intracellular pools of Na,K-ATPase (lanes 7-12). Interestingly, the quantitation of cell surface versus intracellular pool of Na,K-ATPase demonstrated that in control cells a majority (63 Ϯ 2%, n ϭ 10) of Na,K-ATPase is expressed at the cell surface. As a control for the biotinylation procedure we re-probed the same Western blot with an antiactin antibody. As shown in Fig. 9 (lower panel), only intracellular protein extracts were positive for anti-actin staining (lanes 7-12), thus demonstrating that only cell surface-expressed proteins were biotinylated in our experiments. Collectively, these data indicate that inhibition of Na,K-ATPase current by PD98059, U0126, or SL327 results from a decreased intrinsic activity of Na,K-ATPase and not from a decrease in its cell surface expression.
Effect of PD98059 on Na,K-ATPase Activity and pERK1/2 Expression in Microdissected CCDs-We also checked whether the MEK1/2-ERK1/2 pathway controls the activity of the Na,K-ATPase in microdissected CCDs ex vivo. In these experiments we used microdissected CCDs preincubated or not with cAMP, a mediator of intracellular vasopressin signaling in the principal cell. As shown in Fig. 10A, 50 M PD98059 significantly inhibited the Na,K-ATPase activity in both untreated and cAMP-treated CCDs. This decrease in Na,K-ATPase activity was in parallel with a decreased ERK1/2 phosphorylation as determined by Western blot (Fig. 10B). Thus, the MEK1/2-ERK1/2 pathway is also involved in the control of Na,K-ATPase activity in microdissected CCDs. DISCUSSION In the present study we describe a novel mechanism that controls sodium reabsorption in the principal cell by a basal, hormone-independent activation of Na,K-ATPase by the ERK1/2 MAP kinase pathway.
Basal Activity of the ERK1/2 Pathway in Microdissected CCDs and in mpkCCD cl4 Cells-Several lines of evidence indicate that the principal cell of the kidney collecting duct exhibits a significant level of ligand-independent basal activity of the ERK1/2 pathway. First, Western blotting on proteins extracted from microdissected nephron segments revealed that pERK1/2 is expressed at significant basal levels in the distal nephron segments, including the collecting duct. Protein extracts from proximal tubules that express a significantly higher level of total pool of ERK1/2 were negative for staining with an anti-pERK1/2 antibody. Because the proximal tubule is the part of the nephron most sensitive to perfusion/microdissection stress, these results indicate that expression of pERK1/2 detected by Western blotting in the distal nephron is not the result of tubular stress during microdissection. Although we cannot rule out the possibility that a yet unidentified autocrine or humoral factor(s) continuously stimulates the ERK1/2 pathway in CCDs, our data are more compatible with a ligand-independent basal activation of ERK1/2. Indeed, activated ERK1/2 (pERK1/2) are usually rapidly dephosphorylated after removal of external stimuli, whereas in our experimental conditions incubation of microdissected CCDs in hormone-free medium for up to 3 h did not reduce the basal ERK1/2 phosphorylation. A possible mechanism for this basal activation of the ERK1/2 pathway in the principal cell could be the previously reported basal activation of G protein-coupled receptors or tyrosine kinase receptors (27,28). Second, significant basal expression of pERK1/2 was also detected in mpkCCD cl4 cells after 6 days of deprivation from serum and EGF. This strongly suggests that basal ERK1/2 activity in mpkCCD cl4 cells results from a ligandindependent basal activation of the cascade. Third, neither aldosterone nor vasopressin significantly affected ERK1/2 activity in microdissected CCDs or in mpkCCD cl4 cells. These results are different from those obtained by Hendron and Stockand (29) in A6 cells derived from X. laevis kidney in which aldosterone significantly increased the level of ERK1/2 phosphorylation via stimulation of expression and activity of KI-RasA. The reason for the difference is not known. A possible explanation is that Ras-dependent activation of ERK1/2 could be specific for amphibian cells, and the expression of Ras proteins in mammalian principal cells may not be regulated by aldosterone.
Our data are similar to recent observations made in tissues or cells showing that the basal activation of MAP kinase pathways may be responsible for sustained stimulation or inhibition of various ion transport processes. For example, in hepatocytes the basal activity of p38 MAP kinase is required for maintenance of cell volume through tonic inhibition of Na ϩpermeable ion channels (30). In dendrites of pyramidal neurons, the basal activity of ERK1/2 is responsible for sustained downregulation of A-type outward potassium currents (31,32).
Control of Transepithelial Sodium Reabsorption in the Principal Cell by the pERK1/2 Pathway-We provide the following lines of evidence that the basal activity of ERK1/2 directly controls the transepithelial sodium transport in mpkCCD cl4 cells.
First, PD98059, a MEK1/2 inhibitor, significantly and reversibly inhibits short circuit current (I sc ) in both unstimulated and aldosterone-or vasopressin-stimulated cells; this inhibition of I sc correlates with a small increase in transepithelial resistance and a decrease in transepithelial potential difference. Second, the inhibitory effect of PD98059 is absent in cells incubated in the presence of amiloride in the apical medium, thus demonstrating that sodium transport is the major target for the ERK1/2 pathway in the principal cell. Third, the inhibition of sodium transport by PD98059 correlates with a decrease in ERK1/2 phosphorylation. Fourth, the effect of PD98059 is specific for the ERK1/2 pathway. It is independent  8. Effects of PD98059 and U0126 on ENaC activity. A, unstimulated cells. Unstimulated mpkCCD cl4 cells were permeabilized with amphotericin B from the basolateral side, and the amiloride (Amil)-sensitive sodium current (I Na ) was measured as described under of whether the inhibitor is added to the apical or basolateral incubating medium, indicating that inhibition of I sc by PD98059 is not a direct nonspecific effect of the inhibitor on apical or basolateral channels and/or transporters. Moreover, MEK1/2 inhibitors have been reported to have several nonspecific targets. For example, PD98059 has been shown to inhibit the activity of COX-1, an enzyme involved in prostaglandin synthesis in the principal cell (33,34). Because prostaglandins are paracrine autacoids causing local regulation of sodium reabsorption in the principal cell, we were concerned by possible MEK1/2-unrelated effects of PD98059 on I sc . We, therefore, tested whether the effect of PD98059 on I sc is additive to that of resveratol, a specific COX-1 inhibitor. These experiments demonstrated that resveratol induces a small but significant decrease in I sc in mpkCCD cl4 cells that is fully additive to the decrease in I sc provoked by PD98059 (data not shown). Thus, the effect of PD98059 on I sc is not due to a nonspecific effect of this inhibitor on COX-1.
Effect of the ERK1/2 Pathway on Na,K-ATPase and/or ENaC Activity-Several recent studies have demonstrated that the functional effects of the ERK1/2 pathway on Na,K-ATPase and ENaC are highly dependent on the cellular context and on the duration and amplitude of ERK1/2 activation. For example, the Na,K-ATPase activity could either be stimulated or inhibited by ERK1/2-dependent phosphorylation of the ␣ 1 subunit (4,35). Long term stimulation of ERK1/2 has been shown to inhibit transcription of ENaC subunits in mouse collecting duct cells (mCT1) and in salivary epithelial cells (36 -38). Activation of the ERK1/2 pathway in the X. laevis oocyte expression system and in A6 X. laevis kidney cells has been shown to inhibit ENaC activity via different signaling mechanisms, including direct phosphorylation of ENaC subunits by ERK1/2 (10 -12).
Our present data demonstrate that in microdissected CCDs and mpkCCD cl4 cells, the constitutive basal activity of the ERK1/2 pathway is responsible for the control of sodium reabsorption via the control of Na,K-ATPase activity. In parallel experiments we did not find any evidence for a direct control of ENaC activity by this mechanism. Thus, these data indicate that in the principal cell the Na,K-ATPase is, as well as ENaC, a rate-limiting factor for transepithelial sodium reabsorption. However, it is probable that ENaC is inhibited secondarily, after inhibition of the Na,K-ATPase (39), as indicated by the observed increase of the transepithelial resistance. Importantly, the effects of MEK1/2 inhibitors on Na,K-ATPase activity and sodium transport were similar in unstimulated and aldosteroneand vasopressin-stimulated cells that exhibit significantly different current levels. This indicates that the basal activity of the ERK1/2 pathway in the principal cell plays a permissive role for Na,K-ATPase function. This mechanism may include either direct phosphorylation of Na,K-ATPase subunits by kinases belonging to the MEK1/2-ERK1/2 pathway or an indirect effect on the pump activity by phosphorylation of its modulatory proteins, for instance one of the members of the FXYD family. Acknowledgment-We thank Dr. K. Geering for the generous gift of the anti-␣ 1 Na,K-ATPase subunit antibody.
FIG. 9. Effects of PD98059 and U0126 on cell surface expression of Na,K-ATPase. Cells grown on two Transwell filters were surface-biotinylated as described under "Materials and Methods." A, Western blotting of cell surface and intracellular proteins with an anti-␣ 1 Na,K-ATPase subunit antibody. B, Western blotting of cell surface and intracellular proteins with an anti-actin antibody.
FIG. 10. Effect of PD98059 on Na,K-ATPase activity and pERK1/2 expression in microdissected CCDs. A, cell surface Na,K-ATPase activity was measured by freeze/thaw permeabilization of the CCD membrane. After permeabilization microdissected tubules were preincubated with 50 M PD98059 or 0.5% Me 2 SO for 45 min at 30°C and then incubated with or without 1 mM dibutyryl-cAMP (db-cAMP) for 15 min at 37°C. Values are the mean Ϯ S.E. from four animals. Statistical significance is calculated using Student's t test. B, microdissected CCDs were incubated with or without vasopressin (10 Ϫ8 M) for 120 min at 35°C. 50 M PD98059 was added during the last 30 min of incubation and analyzed for pERK by Western blot.