INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

IN CULTURED CELLS derived from the renal medulla, urea and NaCl activate divergent signaling pathways leading to distinct profiles of gene transcription. Urea, in concentrations unique to the renal medulla, activates phospholipase C, induces the intracellular release of inositol 1,4,5-trisphosphate (IP₃), and activates protein kinase C (6). These events ultimately lead to activation of the extracellular signal-regulated kinase (ERK) cascade, which mediates urea-inducible transcription of immediate-early genes (4). This transcriptional event requires the coordinated actions of at least two transcription factors, the serum response factor (SRF) and the ERK-responsive transcription factor, Elk-1, operating through composite serum response element and Ets motif DNA consensus sequences (7).

The signaling pathways activated by NaCl are less clear. In renal epithelial cells, NaCl activates all three families of mitogen-activated protein kinase (MAPK) [including the ERK (15, 31), p38 (34), and jun kinase (JNK; 34) families] but fails to appreciably activate Elk-1-dependent transcription (7). Instead, NaCl appears to regulate transcription of tonicity-responsive genes through a recently characterized consensus element present in the 5' flanking regions of genes under osmotic control (10, 30).

Focal adhesion kinase (FAK), a soluble tyrosine kinase immunolocalizable to focal adhesions, is activated upon cell-matrix interaction or in response to one of a group of neuropeptide activators of G-protein-coupled receptors (13). Although lacking demonstrable src-homology 2 (SH2) and src-homology 3 (SH3) domains, FAK regulates a variety of signaling intermediates previously implicated in the osmotic stress response. For example, FAK indirectly activates the MAPK ERK (16, 29), and focal adhesion formation is associated with JNK activation (3, 21). The recently described related adhesion focal tyrosine kinase [RAFTK (1, 19), also known as PYK2 (18) and CAK (27)], a kinase structurally related to FAK, is responsive to diverse cell stressors. Like FAK, it lacks obvious protein interaction domains, and has also been implicated in the activation of ERK and JNK MAPKs (9, 33). In addition, both FAK (13) and RAFTK (1, 18) are renally expressed. In aggregate, these observations suggested a potential role for FAK and RAFTK in solute signaling in cells of the mammalian renal medulla.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Cell culture and treatment. Cells of the inner medullary collecting duct clonal mIMCD3 line (25) were maintained and passaged as previously described (5). For solute treatment, confluent, serum-deprived (0% FBS for 24 h) monolayers were treated with the gentle, drop-wise addition of an aliquot of concentrated solute in sterile water or sterile 150 mM NaCl.

Immunoprecipitation and immune-complex kinase assays for FAK and RAFTK. FAK antibody was purchased from Santa Cruz; anti-RAFTK was prepared as described (1). Immunoprecipitation and autophosphorylation reactions were performed as described (19). Briefly, monolayers of mIMCD3 cells (100 mm dishes) were washed twice with ice-cold PBS and lysed with 1 ml of ice-cold RIPA buffer (14). All subsequent manipulations were performed on ice or at 4°C. Lysates were clarified by brief centrifugation and then immunoprecipitated with anti-RAFTK or anti-FAK antibody and protein G beads (Pharmacia) for 1 h. Beads were pelleted, washed twice with RIPA buffer, washed twice with kinase buffer [20 mM HEPES (pH 7.4), 50 mM NaCl, 5 mM MgCl₂, 5 mM MnCl₂, and 0.1 mM sodium orthovanadate], raised in 20 µl of kinase buffer supplemented with 250 µCi/ml [-³²P]ATP, and incubated for 15 min at 25°C. Kinase reactions were resolved via SDS-PAGE and exposed for autoradiography. Resultant band intensity was quantitated densitometrically (Epson ES-1000C), and data were reduced using Image 1.57 software (National Institutes of Health). Data reflect means ± SE of 3-6 experiments, with a single determination per experiment.

Western analysis. Anti-FAK and anti-RAFTK immunoprecipitates were prepared from confluent, serum-deprived mIMCD3 monolayers as described above. Immunoprecipitates derived from an entire P100 dish were subjected to SDS-PAGE and electroblotted to polyvinylidene difluoride membrane (PVDF, Millipore). Membranes were probed with monoclonal antibody PY99 according to the manufacturer's directions (Santa Cruz Laboratories) and developed with the Renaissance enhanced chemiluminescence system (DuPont-NEN).

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Urea induces transient and NaCl induces prolonged activation of RAFTK. RAFTK, a newly described member of the FAK family, is activated in response to cell stressors, and has been implicated in signaling to multiple potentially stress-responsive MAPK pathways. For this reason, the ability of urea, as well as NaCl, to activate RAFTK, was investigated using an immune complex kinase/autophosphorylation assay, wherein RAFTK autophosphorylation in the presence of [-³²P]ATP was monitored autoradiographically. Urea (400 mM) increased RAFTK activity by 45% at 1 min of treatment, but not at any subsequent time-points examined (Fig. 1). Although in this series of experiments, urea appeared to activate RAFTK at 5 min as well, the difference from control did not achieve statistical significance. In marked contrast, NaCl (200 mM/400 mosmol/kgH₂O) activated RAFTK by between 40% and 70% at all time points examined to 30 min. At only the 15- and 30-min time points was NaCl-inducible RAFTK activity statistically greater than that inducible by urea treatment.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Related adhesion focal tyrosine kinase (RAFTK) activation, determined through immune-complex autophosphorylation assay, expressed as a function of time of treatment (in min) of serum-deprived confluent mIMCD3 monolayers with NaCl or urea (400 mosmol/kgH2O). Inset: raw data from a representative experiment. Lanes are as follows: lane 1, control; lane 2, urea, 1 min; lane 3, urea, 5 min; lane 4, urea, 15 min; lane 5, urea, 30 min; lane 6, NaCl, 1 min; lane 7, NaCl, 5 min; lane 8, NaCl, 15 min; and lane 9, NaCl, 30 min. Here and in Figs. 2-4, data depict means ± SE of 3-6 separate experiments, and significance is indicated as follows: * Statistically significant (P < 0.05) difference with respect to control (time 0); # Statistically significant (P < 0.05) difference (in the case of NaCl treatment) with respect to urea treatment. Where error bars are not visible, they are contained within the data point.

Urea and NaCl activate RAFTK in a dose-dependent fashion. Confluent, serum-deprived mIMCD3 monolayers were control treated, or treated with 200, 400, or 800 mosmol/kgH₂O NaCl or urea; RAFTK activity was assessed at 5 min of treatment. NaCl at 200 mosmol/kgH₂O modestly suppressed RAFTK activity (by 10%; Fig. 2), whereas 400 and 800 mosmol/kgH₂O NaCl increased RAFTK activity by 60% and 125%, respectively. 200 mosmol/kgH₂O urea modestly increased RAFTK activity (15%); 400 and 800 mM urea increased activity by 45% and 80%, respectively. Solute-inducible RAFTK activity was statistically greater than that of control at all concentrations except 200 mosmol/kgH₂O urea, and only at this concentration did the effect of urea differ from that of NaCl.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Effect of increasing concentrations of NaCl or urea (expressed in mosmol/kgH2O) upon RAFTK activation (measured as in Fig. 1) at 5 min of treatment. Inset: raw data from a representative experiment. Lanes are as follows (in mosmol/kgH2O): lane 1, control; lane 2, 200 urea; lane 3, 400 urea; lane 4, 800 urea; lane 5, 200 NaCl; lane 6, 400 NaCl; and lane 7, 800 NaCl.

FAK is activated more rapidly by NaCl than by urea. Because FAK and RAFTK are structurally related and may participate in related signaling events, the ability of the two solutes to activate each of these kinases was next evaluated. In general, FAK activation by NaCl and urea was considerably less than that of RAFTK. Specifically, NaCl (400 mosmol/kgH₂O) activated FAK by 25% to 40% at 1, 5, and 15 min of treatment (Fig. 3). Urea (400 mM), in contrast, failed to activate FAK until 15 min of treatment (35% activation), whereupon FAK activity remained elevated to at least 30 min of treatment. In this and in the subsequent series of experiments, solute induction of kinase activity appeared modest, owing in part to the relatively high basal activity of FAK and RAFTK in this cell line. A similar pattern was previously observed with solute-inducible ERK activation (4). In each individual experiment in a given series of experiments, the solute-inducible changes were qualitatively identical (direction of change), whereas magnitude of induction (e.g., fold activation) varied considerably between experiments.

View larger version (27K): [in this window] [in a new window]   Fig. 3.   Focal adhesion kinase (FAK) activation, determined through immune complex autophosphorylation assay, expressed as a function of time of treatment (in min) of serum-deprived confluent mIMCD3 monolayers with NaCl or urea (400 mosmol/kgH2O). Inset: raw data from a representative experiment. Lanes are as follows: lane 1, control; lane 2, urea, 1 min; lane 3, urea, 5 min; lane 4, urea, 15 min; lane 5, urea, 30 min; lane 6, NaCl, 1 min; lane 7, NaCl, 5 min; lane 8, NaCl, 15 min; and lane 9, NaCl, 30 min.

Effect of solute dose upon FAK activation. The ability of increasing doses of solute to activate FAK was evaluated at 5 min of treatment. This time-point was selected for consistency with other observations (see below). As expected, at this time-point, there was no appreciable activation of FAK in response to 200, 400, or 800 mosmol/kgH₂O urea (Fig. 4). In contrast, 400 mosmol/kgH₂O NaCl increased FAK activation by 30%; data at 200 and 800 mosmol/kgH₂O were consistent but did not achieve statistical significance.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Effect of increasing concentrations of NaCl or urea (expressed in mosmol/kgH2O) upon FAK activation (measured as in Fig. 3) at 5 min of treatment. Inset: raw data from a representative experiment. Lanes are as follows (in mosmol/kgH2O): lane 1, control; lane 2, 200 urea; lane 3, 400 urea; lane 4, 800 urea; lane 5, 200 NaCl; lane 6, 400 NaCl; and lane 7, 800 NaCl.

RAFTK tyrosine phosphorylation correlates with RAFTK activity. To corroborate the above data with RAFTK, anti-RAFTK immunoprecipitates were prepared from control-treated cells, as well as cells treated with 5 and 30 min of exposure to 800 mosmol/kgH₂O NaCl, and then resolved via SDS-PAGE and subjected to Western analysis with PY99, a monoclonal antibody recognizing phosphotyrosine residues (Fig. 5). These treatment conditions were chosen because they represented conditions of maximal RAFTK activation (Figs. 1 and 2). Tyrosine phosphorylation of RAFTK was increased by ~50% at 5 min of treatment and by greater than 100% at 30 min of treatment (Fig. 5; top). These data were consistent with the degree of RAFTK activation under these same conditions. In contrast, an increment in FAK tyrosine phosphorylation could not be reproducibly demonstrated under these conditions (Fig. 5; bottom). This latter finding is not surprising in light of the modest increase in FAK activity (20-40%) observed under these conditions in the in vitro kinase assay (which, as an enzymatic assay, should reflect an amplification of changes in phosphorylation state), and the relative insensitivity of western analysis to subtle changes in immunoreactivity, potentially as a consequence of the nonlinearity of enhanced chemiluminescence.

View larger version (80K): [in this window] [in a new window]   Fig. 5.   Effect of 5 and 30 min of NaCl (400 mM) treatment upon tyrosine phosphorylation of RAFTK (top; arrowhead) and FAK (bottom; arrowhead) in anti-RAFTK and anti-FAK immunoprecipitates (IP), respectively, as determined by anti-phosphotyrosine (PY99) immunoblotting (IB). Both RAFTK and FAK comigrated with 127-kDa molecular mass marker. C, control.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

These data indicate that both NaCl and urea activate the newly identified FAK-like tyrosine kinase, RAFTK, in a time- and dose-dependent fashion, in a renal medullary cell line derived from a tissue specifically adapted to endure osmotic fluctuations. The kinetics of induction by these solutes are markedly dissimilar: urea transiently activates RAFTK, whereas NaCl exerts its effect for greater than 30 min. In marked contrast, FAK is only modestly activated by NaCl and urea, as measured by in vitro autokinase assay. Interestingly, it is NaCl that exerts the early effect upon FAK, whereas the effect of urea is not demonstrable until fully 15 min of treatment. Therefore, whereas FAK and RAFTK were coordinately regulated by NaCl, they were differentially regulated by urea. These distinctions serve to underscore the dissimilar physiological effects of NaCl and urea upon cell signaling in cells of the renal medulla. In addition, they suggest the presence of potentially distinct regulatory mechanisms operative in solute stress signaling to each of the two members of the FAK family.

That urea and NaCl should differentially regulate FAK family kinases is consistent with an evolving understanding wherein these two solutes activate distinct signaling pathways and influence gene transcription through divergent mechanisms. Urea induces the activation of phospholipase C, the release of IP₃, and the activation of protein kinase C in renal medullary cells (6). In addition, urea activates the ERK MAPKs in a MAPK/ERK kinase (MEK)-dependent fashion (4). This latter pathway mediates urea-inducible transcription of immediate-early gene transcription factors, such as Egr-1, in an SRF- and Elk-1-dependent fashion (4, 7). NaCl, in contrast, although capable of activating ERK MAPKs (15, 31), positively regulates gene transcription through a unique, ERK-independent (17), mechanism. Although the signaling intermediates are unknown, NaCl treatment ultimately activates transcription through one of a family of related consensus sequences identified in the 5'-flanking region of genes encoding proteins instrumental in osmolyte synthesis and transport (10, 30). Further underscoring the dissimilar effects of these two solutes, we noted that NaCl activates the stress-responsive MAPKs, p38 and JNK, in renal medullary cells, whereas urea fails to do so (34).

Several previous lines of evidence have implicated members of the FAK family in stress signaling. In addition to activation through neuropeptide-responsive G-protein-coupled receptors, as well as through membrane depolarization and calcium ionophore treatment in neurally derived cells (18), RAFTK/PYK2, like the MAPKs JNK and p38, is also activated in response to pro-inflammatory cytokine treatment, ultraviolet (UV) irradiation, and sorbitol-induced osmotic stress in hematopoietic cells (33). The upstream signaling event initiated by these latter two physical stimuli remains obscure. FAK has received less attention in the context of response to cell stressors. Although regulated through mitogenic neuropeptides, much like RAFTK, and by Src transformation (28), FAK is also activated by adhesive glycoprotein-mediated signaling through integrins at focal adhesions (13). This latter mechanism is potentially regulated by physical stressors, such as osmotically inducible changes in cell volume, that can result in cytoskeletal reorganization (e.g., see Refs. 12 and 32) and alter dynamic cell-matrix adhesive contacts. Neither FAK family kinase has previously been examined in the context of urea stress or in detail in NaCl stress in a cell type adapted for chronic exposure to elevated and rapidly fluctuating osmolality.

The precise role that RAFTK signaling plays in the renal medullary cell response to NaCl and urea remains unclear. RAFTK is renally expressed and, in other tissues, directly regulates ion channel phosphorylation and activation (18). Regulation of this type affords an obvious precedent for potential physiological effector functions in the response to anisotonicity. RAFTK can also mediate the activation of ERKs by interaction of neuropeptide mitogens with their G-protein-coupled receptors (9, 18). How far upstream RAFTK functions in the ERK MAPK cascade is unknown; urea-inducible ERK activation, for example, requires activation of the ERK activator MEK, as determined through studies with the specific MEK inhibitor, PD-98059 (4). Similarly, the molecular target of RAFTK impingement upon the MAPK cascade that culminates in JNK activation, as well as the precise role of this event in NaCl-inducible JNK activation, remains unclear in the present model. Overexpression of a dominant negative-acting form of RAFTK blocked UV irradiation and osmotic induction of JNK in a cell type-specific fashion (33). A role for members of the Rho family of small GTPases [known potential activators of JNK (8, 20, 23)] in this pathway was suggested by the ability of dominant negative-acting mutants of this class to block RAFTK-inducible JNK activation (33).

The role of FAK in previously described solute-inducible signaling events also merits speculation. Although lacking confirmed substrates, FAK directly or indirectly (via an intermediary kinase) tyrosine-phosphorylates paxillin and tensin and thereby influences cytoskeletal organization (reviewed in Ref. 26). Like RAFTK, FAK activates the ERK MAPKs via a Grb2-dependent (and, therefore, Ras-dependent) pathway (16, 29). Through which molecular target, if any, in the ERK MAPK cascade FAK exerts this effect is unknown. Although FAK has thus far not been directly implicated in signaling to other stress-responsive MAPKs, JNK activation has been described in the context of integrin-mediated cell adhesion (3, 21). In addition, as described above, small GTPases of the Rho family (which include Rho, Rac, and Cdc42) regulate both formation of focal adhesions (22, 24) and activation of JNK (8, 20, 23). Therefore, as for RAFTK, FAK remains a potential signaling intermediate in the solute-inducible signaling cascades leading to ERK (15, 31) and JNK (34) activation in renal epithelial cells. Activation of protein kinase C, which accompanies urea treatment of renal medullary cells, can also activate focal adhesion formation (reviewed in Ref. 11), and presumably FAK. Furthermore, FAK physically associates with phosphatidylinositol-3 kinase (2), a lipid kinase activated by osmotic stressors (S. Soltoff, Z. Zhang, and D.M. Cohen, unpublished observation).

These data demonstrate that the tyrosine kinases, FAK and RAFTK, are differentially responsive to hypertonic (NaCl) and urea stress in cultured cells of the renal medulla. Each of these kinases has been implicated in regulating downstream signaling events associated with the osmotic stress response; it is possible that the differential downstream signaling effects of NaCl and urea are mediated, at least in part, by the differential upstream activation of FAK and RAFTK.