INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

UREA IS A POTENT DENATURANT of proteins (43) and nucleic acids (16), even at concentrations physiologically relevant to the renal medulla (25). NaCl represents the other principal constituent of excess medullary osmolality. The remarkable ability of cells of this tissue to survive and thrive in such a harsh environment is of considerable interest. Despite their presence in roughly equiosmolar amounts in the renal medulla in vivo under conditions of unlimited water access and intact antidiuretic hormone action, the signaling pathways engendered by urea and NaCl are distinct. Because urea, in marked contrast to NaCl, is potentially membrane permeant (25), fluctuations in urea concentration are less likely to cause acute changes in cell volume; therefore, the need for distinct signaling pathways in response to each of these solutes can readily be appreciated. In addition, states associated with pathological salt, water, and urea homeostasis may be associated with preferential intramedullary accumulation of either NaCl or urea (e.g., see Ref. 15).

In mIMCD3 renal inner medullary collecting duct cells, derived from the murine renal inner medulla, urea activates the receptor tyrosine kinase-specific phospholipase C (PLC) isoform, PLC-, as well as protein kinase C, and causes the intracellular release of inositol 1,4,5-trisphosphate (12). Urea also activates the mitogen-activated protein kinases (MAPKs), ERK1 and ERK2 (extracellular signal-regulated kinases), and their transcription factor substrate, Elk-1 (7). These events mediate the serum response element (SRE)- and Ets (E twenty-six specific) motif-dependent transcription (13) and translation of (8), and trans-activation by (9), immediate early gene transcription factors such as the zinc finger protein, Egr-1, by elevated concentrations of urea. Hypertonic NaCl, which activates transcription largely through a distinct DNA element (24, 33, 37) found within the 5'-flanking sequences of genes encoding osmolyte synthetic and transport proteins (35, 37, 39, 42), also activates the ERK MAPKs (26, 38). Despite the apparent convergence of urea and NaCl signaling upon MAPK activation, however, only NaCl activates the two other principal families of stress-responsive MAPKs, p38 and stress-activated protein kinase/jun kinase (SAPK/JNK), whereas urea fails to do so (44). In addition, although both NaCl and urea activate ERKs, only urea substantially activates ERK-dependent gene transcription (13).

Ribosomal S6 kinase (RSK; also known as p90 S6 kinase and MAPKAP kinase-1) is a growth-factor regulated serine-threonine kinase permissive for cell proliferation through an effect mediated at the G₀/G₁ transition (2). RSK is an ERK substrate (23) and therefore represents a potential physiological effector of urea signaling. Terada et al. (38) showed that the hypertonic stressor, NaCl, activated an S6 kinase activity in crude lysates prepared from salt-stressed MDCK cells (38); which of the disparately regulated isoforms of S6 kinase was responsible for this effect was not demonstrated. Because of the divergence between salt and urea signaling and the ability of urea to activate ERKs and the ERK substrate, Elk-1, the ability of urea to activate RSK and the dependence of this effect upon an intact ERK MAPK cascade were investigated in renal medullary cells.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Cell culture and solute treatment. mIMCD3 cells were maintained and passaged as previously described (8). Prior to each experiment, cells were taken out of serum for 24 h after reaching confluence (~4 days after passage). Solute treatment consisted of the gentle, drop-wise addition to monolayers of an aliquot of concentrated urea (4.2 M) or NaCl (2.25 M) in sterile water or an equal volume of NaCl (150 mM) in sterile water (sham treatment). Where indicated, PD-98059 (50 µM; Calbiochem) pretreatment for 30 min was used.

RSK immunoprecipitation and S6 kinase assay. Control and solute-treated mIMCD3 monolayers were lysed with 1 ml of S6 kinase lysis buffer [10 mM potassium phosphate (pH 7.05), 0.5% Triton X-100, 1 mM EDTA, 5 mM ethylene glycol-bis(-aminoethyl ether)-N,N,N',N'-tetraacetic acid (EGTA), 1 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride, 1 µg/ml leupeptin, 5 µg/ml aprotinin, and 10 µg/ml pepstatin A] and then immunoprecipitated with anti-RSK COOH-terminal polyclonal antibody (Santa Cruz Laboratories) for 60 min at 4°C. Immunoprecipitates were washed twice with ice-cold lysis buffer, twice with ice-cold kinase buffer [20 mM 3-(N-morpholino)propanesulfonic acid (pH 7.2), 25 mM -glycerophosphate, 1.25 mM EGTA, 200 µM sodium orthovanadate, 1 mM dithiothreitol; Ref. 19], and then subjected to an in vitro kinase assay using an S6 kinase-specific peptide substrate and [-³²P]ATP in the presence of inhibitors of other kinases, according to the manufacturer's directions (Upstate Biotechnology). Incubation was performed for 10 min at 30°C, after which the entire reaction was spotted on phosphocellulose paper, washed extensively with 0.75% phosphoric acid, and washed with acetone prior to drying and scintillation counting. Kinase activity was expressed in raw counts per 100-mm dish or counts relative to control. Unless otherwise indicated, data are presented as the mean of two wells in the same experiment (representative of at least 2 experiments).

RSK Western analysis. Detergent lysates from control and solute-treated mIMCD3 monolayers were prepared in radioimmunoprecipitation assay buffer (RIPA), immunoprecipitated (with anti-RSK antibody, 2 µg/ml; Santa Cruz Laboratories), subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and transferred to polyvinylidene difluoride as previously described (12). All incubations were performed at 25°C with gentle rocking. Membranes were blocked with 5% nonfat dry milk (NFDM) in phosphate-buffered saline (PBS)/0.1% Tween for 1.5 h and washed three times with PBS/0.1% Tween/1% NFDM; incubated with a 1:1,000 dilution of anti-p90 RSK antibody (Santa Cruz Laboratories) in PBS/1% NFDM for 1 h and washed three times with PBS/0.1% Tween/1% NFDM; and incubated with 1:4,000 dilution of the horseradish peroxidase-coupled appropriate secondary antibody (Pierce, Rockford, IL) in PBS/0.1% Tween/1% NFDM, and washed three times with PBS/0.3% Tween/1% NFDM. Detection was via enhanced chemiluminescence according to the manufacturer's directions (ECL; Amersham, Arlington Heights, IL).

Immunofluorescence. mIMCD3 monolayers were grown to confluence on glass slides and serum deprived for 24 h prior to sham treatment or treatment with 400 mosmol/kgH₂O solute. Monolayers were fixed with 3.7% formaldehyde in PBS, rinsed twice with PBS, permeabilized with 0.2% Triton X-100, blocked with 3% normal goat serum in PBS, rinsed with PBS, and incubated at 4°C overnight with 1:1,000 dilution of anti-RSK antibody (Santa Cruz Laboratories) in PBS + 3% normal goat serum. The following day, slides were washed with PBS, incubated with 1:250 dilution of secondary antibody [fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit immunoglobulin G] in PBS + 3% normal goat serum for 1 h at 25°C, washed four times with PBS, and mounted in 50% glycerol in PBS. Photomicrographs of representative fields (using a Nikon Microphot-FX) were obtained at ×400 under phase-contrast (see Fig. 4, B and D) and fluorescent (excitation 470-490 nm; see Fig. 4, A and C) illumination.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Urea activates RSK in a time- and dose-dependent fashion. Through the use of an immunoprecipitation-immune complex kinase assay, it was shown that urea (200 mM) activated p90 RSK S6 kinase activity in serum-deprived confluent mIMCD3 monolayers (Fig. 1). Increased activity was absent at 1 min of treatment, peaked (at 4-fold relative to control) at 5 min, and declined thereafter. In contrast, in response to an equivalent degree of hypertonic stress with NaCl (200 mosmol/kgH₂O), RSK activation increased less than 50% at 5 min and did not peak until fully 15 min of solute treatment, reaching 3.5-fold activation relative to baseline. Not shown, data obtained at 10 min of treatment were consistent with those obtained at 5 min of treatment. Only at 5 min (and, not shown, at 10 min) was the induction by urea greater than that of equiosmolar NaCl. The membrane-impermeant solute, mannitol, activated RSK to an equivalent degree as NaCl and with comparable kinetics (data not shown).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Effect of urea (200 mM) or NaCl (200 mosmol/kgH2O) upon ribosomal S6 kinase (RSK) activity, as measured by immune-complex kinase assay using anti-RSK immunoprecipitates (see text). Data points are the means ± SE of 3-4 experiments, each performed with wells treated in duplicate. * P < 0.05 relative to time 0. # P < 0.05 with respect to NaCl at the identical time point.

With respect to dose response, urea increased RSK activity approximately fourfold at 200, 400, and 800 mM at 5 min of treatment (Fig. 2). NaCl, in contrast, increased activity only 2.5-fold at 200, 400, and 800 mosmol/kgH₂O. After statistical analysis, activation by urea exceeded that of equiosmolar NaCl at only 200 and 800 mosmol/kgH₂O solute; in all experiments performed, however, activation by 400 mM urea also exceeded that of 400 mosmol/kgH₂O NaCl but did not achieve statistical significance, owing to interexperiment variability in the induction.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Effect upon p90 RSK S6 kinase activity of 5-min treatment of mIMCD3 renal inner medullary collecting duct cells with urea (200 mM) or NaCl (200 mosmol/kgH2O). Values are means ± SE of 3-6 experiments (each consisting of 1-2 wells/experiment) is depicted. * P < 0.05 relative to control (no solute or sham treatment). # P < 0.05 with respect to NaCl of equivalent osmolarity.

Urea and NaCl do not alter RSK protein abundance or immunoprecipitability. To confirm that urea was exerting no effect upon RSK protein abundance or immunoprecipitability during this time period and to thereby confirm that the observed changes in RSK activity were a consequence of kinase activation, Western analysis with anti-RSK antibody was performed. RSK immunoprecipitates were prepared from control, urea-, and NaCl-treated mIMCD3 cells as for the immune complex kinase assay and then resolved via SDS-PAGE for Western analysis. As shown in Fig. 3, a broad band of immunoreactivity at molecular mass of 90-100 kDa was detected, consistent with the expected migration of RSK on SDS-PAGE (23). There was no change in RSK abundance or anti-RSK immunoprecipitability at 5 min of treatment (200 mosmol/kgH₂O) with either urea or NaCl, as quantitated densitometrically. In addition, when equal amounts of whole cell lysate (40 µg) were resolved via SDS-PAGE (in contrast to immunoprecipitates) and subjected to anti-RSK Western analysis, there was again no change detected in RSK abundance over the short intervals examined (to 30 min).

View larger version (59K): [in this window] [in a new window]   Fig. 3.   Effect of control treatment (C) or treatment (200 mosmol/kgH2O for 5 min) with urea (U) or NaCl (N) upon RSK abundance in anti-RSK immunoprecipitates prepared from detergent lysates of mIMCD3 cells, as determined by Western analysis.

Urea-inducible RSK activation is associated with nuclear translocation of RSK. Because nuclear translocation accompanies RSK activation (5), the ability of urea to induce RSK nuclear translocation in mIMCD3 cells was investigated as a correlate of RSK activation. In control confluent, serum-starved, mIMCD3 monolayers, RSK immunofluorescence was confined to the cytosol (Fig. 4A). In contrast, following treatment with urea (400 mM for 5 min), cytosolic staining diminished markedly and was replaced by pronounced nuclear immunofluorescence (Fig. 4C). The fields depicted in Fig. 4, A and C, are also shown under phase-contrast illumination in Fig. 4, B and D, respectively, for comparison. When cells were treated with NaCl (400 mosmol/kgH₂O) for 5 min, translocation of RSK immunofluorescence from cytosol to nucleus was also evident, although it was less pronounced than with urea treatment (data not shown). When fixed monolayers were examined after staining with only secondary (FITC conjugated) antibody, in the absence of primary (RSK specific) antibody, negligible fluorescence was observed. Confluent monolayers were examined to maintain consistency with the conditions under which the S6 kinase assays were performed. Importantly, when subconfluent, non-serum-deprived cells were examined, abundant nuclear staining could be detected in selected cells under both control and urea-treated conditions, consistent with ongoing cell division and associated ERK activation.

View larger version (128K): [in this window] [in a new window]   View larger version (114K): [in this window] [in a new window]   View larger version (124K): [in this window] [in a new window]   View larger version (124K): [in this window] [in a new window]   Fig. 4.   Effect of no treatment (A and B) or urea treatment (400 mM for 5 min; C and D) upon RSK immunolocalization in confluent, serum-deprived mIMCD3 monolayers, as determined by immunofluorescence with anti-RSK antibody and fluorescein-conjugated secondary antibody. Photomicrographs were obtained under phase-contrast (B and D) or fluorescent (excitation wavelength 470-490 nm; A and C) illumination, at ×300 magnification.

Urea- and NaCl-inducible RSK activation is MEK dependent. Because RSK is a physiological substrate of ERK, the ability of the MEK-specific inhibitor of ERK activation, PD-98059 (1, 20), to block urea- and NaCl-inducible RSK activation was next evaluated to determine MEK/ERK dependence of this phenomenon. Urea (200 mM) increased RSK activity greater than fourfold at 5 min of treatment (Fig. 5A), consistent with Figs. 1 and 2. PD-98059 (30-min pretreatment; 50 µM) exerted no effect upon basal RSK activity, whereas the urea-inducible increment in RSK activity was inhibited fully 78%. Similarly, examining the time point of maximal NaCl-inducible RSK activation, 15 min of NaCl (200 mosmol/kgH₂O) treatment increased RSK activity twofold; PD-98059 (50 µM for 30 min) pretreatment exerted no effect upon basal RSK activity but suppressed NaCl-inducible activation by fully 100%.

View larger version (10K): [in this window] [in a new window]   Fig. 5.   A: effect of no pretreatment (PD-98059) or pretreatment with the MEK inhibitor, PD-98059 (+PD-98059; 50 µM for 30 min), upon RSK activity following 5 min of control treatment (control), or treatment with 200 mM urea (+Urea). B: effect of no pretreatment (PD-98059) or pretreatment with PD-98059 (+PD-98059; 50 µM for 30 min) upon RSK activity following 15 min of control treatment (control) or treatment with 200 mosmol/kgH2O NaCl (+NaCl). Data are means ± SE of at least 3 experiments. * P < 0.05 relative to untreated control. # P < 0.05 with respect to solute treatment in absence of PD-98059.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

These data demonstrate that, in cells of the renal medulla, urea activates the ERK substrate, RSK, and does so in a MEK-dependent fashion. The kinetics of urea-inducible RSK activation were distinct from those engendered by NaCl. Urea activates the ERK MAPKs, and urea-inducible gene transcription is mediated by ERK-dependent activation of the ERK substrate and transcription factor, Elk-1 (7). Urea-activated Elk-1 upregulates transcription from promoters containing its cognate DNA binding sequence, the Ets motif, in an appropriate context of SREs (13). The ability of urea to activate the ERK substrate and key regulatory kinase, RSK, underscores the functional importance of this pathway, as well as the likelihood of its representing an adaptive or protective response to urea in renal medullary cells. Additional putative ERK substrates include the enzyme phospholipase A₂ (31) and the transcription factor c-myc (34); although potentially operative in the response to urea stress, none has yet been investigated in this context.

The ability of the hypertonic stressor and other principal renal medullary solute, NaCl, to activate the ERKs in renal epithelial cells was demonstrated by Terada et al. (38) and Itoh et al. (26). Terada et al. (38) further showed that NaCl augmented total S6 kinase activity in crude lysates prepared from hypertonically stressed MDCK cells. Although no direct biochemical link was demonstrated, activation of this kinase temporally succeeded activation of an ERK-like MAPK (38). Relative contributions to this activity from p70 and p90 S6 kinase isoforms, with their attendant distinct regulatory mechanisms (see below), could not be distinguished in this fashion. In the present study, urea-inducible (and NaCl-inducible) RSK activation was directly demonstrated and was attributable to ERK activation because the specific MEK inhibitor, PD-98059 (1, 20), substantially abrogated the effect. In addition, this activation was correlated with the nuclear translocation of RSK.

The ability of urea and NaCl to similarly induce both ERK activation and RSK activation (albeit with dissimilar kinetics) was unexpected. First, urea and NaCl signal to activate transcription through divergent mechanisms. The membrane-permeant solute urea activates transcription through composite SRE/Ets motifs (13) in an ERK-dependent fashion (7). NaCl, a functionally impermeant (and therefore truly hypertonic) stressor, operates in an ERK-independent fashion (29) through one of a family of unique DNA regulatory elements specific to the 5'-flanking sequences of genes encoding proteins that govern osmolyte transport and synthesis (24, 33, 37). In addition, reflecting their dissimilar physiological effects, urea and NaCl induce the intracellular accumulation of distinct profiles of organic osmolytes; NaCl induces accumulation of all five of the principal intracellular organic osmolytes [including the polyols, sorbitol and myo-inositol; the methylamines, betaine and glycerophosphocholine (GPC); and the amino acid taurine] whereas urea results in the accumulation of only GPC (25). Furthermore, NaCl activates the stress-responsive MAPKs, p38 and SAPK/JNK, whereas urea fails to do so (44). Nonetheless, despite these remarkable differences, both urea and NaCl activate RSK, and in each case this activation occurs in an ERK-dependent fashion. Interestingly, RSK is also activated by other potentially cytotoxic stressors including heat shock (27) and ionizing radiation (28); much like NaCl treatment and in marked contrast to urea treatment, however, both of these stressors activate SAPK/JNK (reviewed in Ref. 6).

The role of RSK activation in the setting of hyperosmotic stress, inducible by either NaCl or urea, remains unclear. Substrates of RSK are few and include principally the S6 protein of the ribosomal 40S subunit, phosphorylation of which regulates translation and thereby traversal of the cell cycle (21, 23). With respect to translation, NaCl (14) but not urea treatment (10) globally suppresses protein synthesis in renal epithelial cells. The relationship between urea treatment and the cell cycle remains to be clarified. In renal epithelial MDCK and LLC-PK₁ cells, urea increased DNA synthesis without augmenting cell number (11); this phenomenon was not observed in the mIMCD3 cell line (8). In addition, the adenosine 3',5'-cyclic monophosphate (cAMP)-responsive element binding protein (CREB) undergoes serine phosphorylation in response to growth factor stimulation, an effect mediated in large part by RSK (3, 41). Importantly, urea treatment fails to activate transcription from a luciferase reporter gene driven by tandem repeats of the CRE, with which CREB interacts (data not shown). Although glycogen synthase kinase-3 is inhibited in vitro by activated RSK (22, 36), a different mechanism appears to be operative in intact cells (17). Other in vitro RSK substrates include c-fos (4, 5, 45), the G-subunit of protein phosphatase 1 (18, 30), and various histone proteins (5, 45); for none of these substrates has a physiological role for RSK-inducible phosphorylation been confirmed in any context.

Although it shares the same principal substrate as RSK, the p70 S6 kinase is regulated through a distinct mechanism. In contrast to RSK, which is activated by ERKs, p70 S6 kinase is activated by the phosphoinositide kinase and SH2 domain-containing receptor tyrosine kinase effector, phosphatidylinositol 3-kinase (40). Interestingly, urea treatment of mIMCD3 cells also resulted in the activation of this S6 kinase in a time- and dose-dependent fashion (S. Soltoff and D. M. Cohen, unpublished observations), however, these kinases are not coordinately regulated in most models. For example, p90 but not p70 S6 kinase is activated following thrombin stimulation of platelets (32).

These data establish that urea activates the S6 kinase, RSK, in a time- and dose-dependent fashion, and induces its nuclear translocation, thereby defining a second urea-inducible, MEK-dependent, signaling pathway in renal medullary cells.