INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

MITOGEN-ACTIVATED PROTEIN (MAP) kinases are key intermediates in the signal transduction pathways activated by a wide variety of extracellular stimuli (10, 14, 50). Several distinct subgroups of MAP kinases have been identified and cloned in mammalian cells, including extracellular signal-regulated kinase (ERK) (12), c-Jun NH₂-terminal kinase (JNK) (17, 39), and p38 MAP kinase (19, 31, 40, 45). Each subgroup is activated via a signaling cascade involving the sequential phosphorylation of protein kinases, resulting in the activation of a MAP kinase kinase (MEK or MKK) that activates MAP kinase through dual phosphorylation on threonine and tyrosine (1, 14, 16, 32, 50). Although overlap exists between the MAP kinase signaling pathways, the different MAP kinase subgroups are activated by distinct MEK/MKKs, have different substrate specificities, and are regulated by different extracellular stimuli, suggesting that they subserve distinct physiological functions. The ERK pathway is stimulated primarily by growth factors and tumor promoters (1, 12), whereas the JNK and p38 MAP kinase pathways are activated by pro-inflammatory cytokines and environmental stress (31, 39, 40, 44, 45, 48, 50). Included in the latter category is hyperosmotic stress, which rapidly activates the JNK and p38 MAP kinase pathways in mammalian cell lines (6, 7, 20, 31, 42, 44). One complication in defining the physiological relevance of these osmotically activated pathways is that most cells of the body are not normally exposed to a hyperosmotic environment. An exception is the cells in the renal medulla, which routinely are exposed to a hyperosmotic interstitial fluid in vivo and are subjected to rapid and large variations in osmolality during changes in H₂O balance due to the operation of the urinary concentrating mechanism (37). Studies of renal cells in culture have shown that short-term hyperosmolality can activate the ERK, JNK, and p38 MAP kinase pathways (6, 34, 38, 51, 56). However, there have been no studies of the osmotic regulation of MAP kinases in intact renal tubules; thus the role of these pathways in the physiological responses of medullary tubules to osmotic stress has not been defined.

The medullary thick ascending limb (MTAL) of the loop of Henle is located in the renal outer medulla and plays a key role in sodium and water homeostasis (37). The MTAL also participates in the regulation of acid-base balance by reabsorbing a sizable fraction of the HCO₃ filtered at the glomerulus (24). The proton secretion required for this HCO₃ absorption is mediated virtually completely by apical membrane Na⁺/H⁺ exchange (30). Recently, we showed that both apical membrane Na⁺/H⁺ exchange and transepithelial HCO₃ absorption are osmotically regulated: hyperosmolality inhibited apical membrane Na⁺/H⁺ exchange activity by decreasing its sensitivity to intracellular H⁺, thereby inhibiting HCO₃ absorption (23, 54). These inhibitory effects were blocked by the protein-tyrosine kinase inhibitors genistein and herbimycin A, suggesting a key role for tyrosine phosphorylation in the hyperosmotic response (25). However, the molecular identities of the specific signaling proteins involved in the hyperosmotic inhibition are unknown, and the mechanisms of osmotic regulation of Na⁺/H⁺ exchange remain poorly defined (15, 25, 52). In view of their rapid activation by hyperosmotic stress in other systems, MAP kinases could be components of the signal transduction pathway by which hyperosmolality inhibits apical Na⁺/H⁺ exchange activity and HCO₃ absorption in the MTAL. At present, the role of MAP kinases in the osmotic regulation of ion transport in epithelial cells is not understood.

The purpose of the present study was to examine the regulation of ERK, JNK, and p38 MAP kinase by hyperosmolality in the MTAL of the rat and to determine whether activation of these MAP kinase pathways plays a role in the hyperosmotic inhibition of HCO₃ absorption. To accomplish these goals, methods were developed for direct measurement of MAP kinase activities using immune complex assays in microdissected MTALs. The results demonstrate that hyperosmolality activates p38 MAP kinase and ERK but has no effect on JNK. We also show that eliminating the activation of p38 MAP kinase and ERK does not prevent hyperosmotic inhibition of HCO₃ absorption, suggesting that the osmotic regulation of apical membrane Na⁺/H⁺ exchange (NHE3) occurs via a signaling pathway distinct from these MAP kinase pathways.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Materials. Stock solutions of genistein (20 mM) and PD98059 (10 mM) were prepared in DMSO. These agents were diluted into experimental solutions to final concentrations given under RESULTS. Equal concentrations of DMSO were added to control solutions. Genistein and PD98059 were purchased from Research Biochemicals International (Natick, MA), [-³²P]ATP was from Amersham Life Science (Arlington Heights, IL), and acrylamide, bis-acrylamide, and glycine were from Bio-Rad (Hercules, CA). Polyclonal antibodies against p38 MAP kinase and JNK were raised in rabbits as described previously (44, 55). The anti-JNK antibody recognizes both the 46-kDa (JNK1) and 55-kDa (JNK2) isoforms (48). Rabbit polyclonal antibody directed against ERK (p42/p44 MAP kinase) was obtained from Upstate Biotechnology, Lake Placid, NY (anti-rat MAPK R2). Recombinant GST-c-Jun (amino acid residues 1-79) and GST-activated transcription factor 2 (GST-ATF2; residues 1-109) fusion proteins were purified by affinity chromatography using GSH-Sepharose as described (17, 44). Myelin basic protein (MBP) and all other chemicals were obtained from Sigma Chemical (St. Louis, MO).

Tissue preparations for MAP kinase assays. Two tissue preparations from kidneys of male Sprague-Dawley rats (70-120 g; Taconic, Germantown, NY) were used to study MAP kinase activities: 1) the inner stripe of the outer medulla and 2) microdissected MTALs. The rats were anesthetized with pentobarbital (50 mg/kg ip), and both kidneys were removed and sliced in ice-cold control solution (see below). The inner stripe of the outer medulla (the region of the kidney highly enriched in MTALs) was cut from the slices and dissected at 10°C into thin strips of 20-40 tubules. The tissue strips were then divided into two to four samples of equal amount, and the samples were incubated in control or hyperosmotic solution at 37°C for various times as described under RESULTS. The control solution contained (in mM) 146 Na⁺, 4 K⁺, 122 Cl, 25 HCO₃, 2.0 Ca²⁺, 1.5 Mg²⁺, 2.0 phosphate, 1.2 SO²₄, 1.0 citrate, 2.0 lactate, and 5.5 glucose (final osmolality = 290 mosmol/kgH₂O). The hyperosmotic solution was prepared by adding 300 mM mannitol to the control solution (final osmolality = 590 mosmol/kgH₂O). In some experiments, tissue was preincubated at 37°C in control solution containing inhibitor (genistein or PD98059) or vehicle (DMSO) for 1 h prior to exposure to hyperosmolality (see RESULTS). All solutions were equilibrated with 95% O₂-5% CO₂ (pH 7.4). The tissue samples were bubbled with this gas mixture throughout incubation for mixing and to maintain the oxygen tension and pH of the solutions. The incubation solutions and protocols were chosen to reproduce those used previously in in vitro microperfusion experiments (25). After incubation, the tissue was suspended in ice-cold Triton lysis buffer (20 mM Tris, pH 7.4, 137 mM NaCl, 2 mM EDTA, 1% Triton X-100, 25 mM -glycerophosphate, 1 mM sodium orthovandadate, 2 mM sodium pyrophosphate, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, and 10 mg/ml leupeptin), homogenized using a Dounce pestle, and lysed for 4 h at 4°C on an orbital shaker. The cell lysates were then stored at 80°C until assayed for MAP kinase activity. Paired comparisons with fresh lysates showed that freezing had no effect on MAP kinase activities of control or hyperosmotic samples. Protein concentrations were determined using a bicinchoninic acid assay (Micro BCA kit; Pierce, Rockford, IL).

MTALs were dissected from the inner stripe of the outer medulla at 10°C in control solution without enzymatic digestion of the tissue as previously described (23, 29). On a given day, MTALs were dissected over a 4- to 6-h period from two to three rats, and the pooled tubules were divided into two groups of equal total length. The groups of tubules were then incubated at 37°C for 15 min in control or hyperosmotic solution as described above for inner stripe tissue. After incubation, the MTALs were solubilized in ice-cold Triton lysis buffer for 4 h at 4°C and the lysates stored at 80°C. To obtain sufficient protein for immunoprecipitation and assay of MAP kinases (10-30 µg total protein per sample), 150-350 mm of tubule were required. This quantity usually was obtained by combining samples from 2 days of dissection; however, in some experiments, a sufficient number of tubules were obtained in a single day. MAP kinase activities measured in dissected MTALs were similar to those in rapidly excised inner stripe tissue; thus microdissection and in vitro incubation did not alter the activities of the enzymes. As discussed previously (5), the inner stripe preparation has the advantage that it yields sufficient protein to permit study of multiple kinases under several experimental conditions but has the disadvantage that it contains nephron segments other than the MTAL (although these account quantitatively for only a minor percentage of the total protein). Microdissected tubules permit direct study of protein kinase activities in a pure preparation of MTALs isolated without enzymatic digestion; however, the technical difficulty of dissection precludes their use for multiple protocols with numerous enzymes. Use of both preparations in parallel permits a comprehensive analysis of the activities of multiple protein kinases, plus direct confirmation of key observations in freshly dissected MTALs. Studies of the regulation of protein kinase C isoforms (5) and MAP kinases (present study) demonstrate that results obtained with the inner stripe preparation reflect accurately changes observed in dissected MTALs.

Immunoprecipitation and immune complex kinase assays. MAP kinases were immunoprecipitated by incubation for 1 h at 4°C with 2 µg of anti-p38, anti-JNK, or anti-ERK antibody prebound to either Pansorbin (Calbiochem-Novabiochem, La Jolla, CA) or protein A-agarose (Santa Cruz Biotechnology, Santa Cruz, CA). Equal amounts of protein (200 µg for inner stripe preparation; 10-30 µg for microdissected MTAL) were used for immunoprecipitation within each experiment. The immune complexes were washed twice with Triton lysis buffer and twice with kinase buffer (50 mM HEPES, pH 7.4, 50 mM -glycerophosphate, 50 mM MgCl₂, 1 mM dithiothreitol, and 0.2 mM sodium orthovanadate). The immune complex assays were performed at 30°C for 18 min in a total volume of 40 µl of kinase buffer that contained 2 µg substrate, 20 µM ATP, and 40 µM [-³²P]ATP (6,000 Ci/mmol). Substrates were GST-ATF2 for p38, GST-c-JUN for JNK, and MBP for ERK (44, 55). The reactions were terminated by rapid centrifugation and removal of the supernatant, which was combined with an equal volume of 2× Laemmli buffer and boiled for 2 min. Phosphorylated substrates were resolved by SDS-PAGE and detected by autoradiography. Autoradiograms were digitized, and band intensities quantified by densitometry (ImageQuant; Molecular Dynamics, Sunnyvale, CA). When sample protein or immunoprecipitating antibody was excluded from the assays, there was no detectable substrate phosphorylation.

After the completion of assays, immune complexes were recovered, and equal amounts of p38 MAP kinase, JNK, or ERK were verified within experiments by immunoblotting with the same antibodies used for immunoprecipitation. The immunoreactive bands were detected by enhanced chemiluminescence (ECL; Amersham Life Science, Arlington Heights, IL) and quantified by densitometry. Kinase activities were normalized for any differences in the amount of MAP kinase in immune complexes.

Tubule perfusion and measurement of HCO₃ absorption rates. MTALs were dissected, transferred to a bath chamber on the stage of an inverted microscope, and perfused in vitro at 37°C using concentric glass micropipettes as previously described (25, 29). The tubules were perfused and bathed with the same control solution as used in the MAP kinase experiments. Experimental modifications of the perfusion and bath solutions are described under RESULTS. The protocol for study of transepithelial HCO₃ absorption has been described (23, 25). One to three 10-min tubule fluid samples were collected for each experimental period studied in a given tubule. Absolute rates of HCO₃ absorption (J_HCO3, pmol · min¹ · mm¹) were calculated from fluid flow rates, and the difference between total carbon dioxide concentrations was measured in perfused and collected fluids (23).

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Hyperosmolality differentially activates MAP kinases. The effects of hyperosmolality on MAP kinase activities in the inner stripe of the outer medulla are shown in Fig. 1, A and B. Inner stripe tissue was incubated in control or hyperosmotic (300 mM added mannitol) solution for 15 min, and then MAP kinase activities were determined in immune complex assays using the substrates ATF2 for p38, MBP for ERK, and c-Jun for JNK (see METHODS). Hyperosmolality increased p38 MAP kinase activity 2.3-fold and ERK activity 2.0-fold but had no effect on JNK activity¹ (1.1-fold) (control vs. hyperosmotic, Fig. 1). Based on the observations that genistein blocks hyperosmotic inhibition of HCO₃ absorption in the MTAL (25) and inhibits the activation of MAP kinases by various stimuli in other systems (4, 21, 47, 53, 56), we tested the effects of genistein on osmotic regulation of MAP kinases in the inner stripe. Pretreatment with genistein (7 µM) for 1 h did not affect the basal activity of p38 MAP kinase, ERK, or JNK (control vs. genistein, Fig. 1B). However, genistein pretreatment significantly inhibited the stimulation of p38 MAP kinase and ERK by hyperosmolality (hyperosmotic vs. genistein + hyperosmotic, Fig. 1, A and B). These findings support the involvement of tyrosine kinase pathways in the hyperosmotic activation of p38 MAP kinase and ERK.

View larger version (30K): [in this window] [in a new window]   Fig. 1.   Effects of hyperosmolality on mitogen-activated protein (MAP) kinase activities in inner stripe of outer medulla. A: inner stripe tissue was preincubated in control solution for 1 h and then either maintained in control solution (C) or exposed to hyperosmotic solution (H) (300 mM added mannitol) for 15 min. In most experiments, additional samples were preincubated with 7 µM genistein for 1 h prior to hyperosmotic exposure (G+H). Activities of p38 MAP kinase, extracellular signal-regulated kinase (ERK), and c-Jun NH2-terminal kinase (JNK) were measured in immune complex assays using the indicated substrates (see METHODS). The phosphorylated substrates were resolved by SDS-PAGE and detected by autoradiography. Autoradiograms are of representative experiments. ATF2, activated transcription factor 2; MBP, myelin basic protein. B: substrate phosphorylation was quantified by densitometric analysis, and kinase activities are presented as % of control activity (C) measured in the same experiment. Data are means ± SE for 3-6 independent experiments in each group (except G for JNK, n = 1). * P < 0.01 vs. C. # P < 0.05, G + H vs. H.

Although the MTAL comprises the majority of the tissue mass of the inner stripe (5), this region contains other nephron segments (outer medullary collecting duct; thin descending limb) that may contribute to MAP kinase activities. To determine whether the changes observed in the inner stripe reflect the properties of the MTAL, we measured MAP kinase activities directly in microdissected MTALs. MTALs were incubated for 15 min in control and hyperosmotic solution, and then MAP kinase activities were measured in immune complex assays as described in METHODS. Hyperosmolality increased p38 MAP kinase activity (2.1-fold) and ERK activity (2.0-fold) but had no effect on JNK activity (1.0-fold) (Fig. 2). These results demonstrate that hyperosmolality activates p38 MAP kinase and ERK, but not JNK in the MTAL, and confirm that the changes observed in the inner stripe of the outer medulla reflect changes in the MTAL.

View larger version (57K): [in this window] [in a new window]   Fig. 2.   Effects of hyperosmolality on MAP kinase activities in microdissected medullary thick ascending limbs (MTALs). MTAL were incubated in control (C) or hyperosmotic (H) (300 mM added mannitol) solution for 15 min, and then MAP kinase activities were measured by immune complex assays (left) as described in METHODS. Substrates were the same as in Fig. 1. Equal amounts of p38 MAP kinase, ERK, and JNK in control and hyperosmotic immunoprecipitates were confirmed in each experiment by immunoblotting with the same antibodies used for immunoprecipitation (right). Hyperosmolality activated p38 MAP kinase 2.1-fold and ERK 2.0-fold relative to control but had no effect on JNK (1.0-fold). Each experiment was performed twice with identical results.

Time course of MAP kinase activation. The time course of p38 MAP kinase and ERK activation is shown in Fig. 3, A and B. Inner stripe tissue was incubated in hyperosmotic solution (300 mM added mannitol) for the indicated times, and then MAP kinase activities were measured by immune complex assays as described in METHODS. Tissue samples incubated in control solution for the same time periods served as time controls in each experiment. p38 MAP kinase activity increased within 5 min of exposure to hyperosmolality (1.7-fold) and remained elevated for up to 60 min (1.8-fold). In contrast, ERK activity was unchanged at 5 min (1.0-fold), increased at 15 min (1.7-fold), and returned to basal levels at 30 min (1.2-fold). Thus hyperosmolality induced a sustained increase in p38 MAP kinase activity but activated ERK transiently. Hyperosmolality had no effect on JNK activity at time points up to 30 min (data not shown).

View larger version (28K): [in this window] [in a new window]   Fig. 3.   Time course of p38 MAP kinase and ERK activation. A: inner stripe tissue was incubated in hyperosmotic solution (300 mM added mannitol) for the times indicated, and then the activities of p38 MAP kinase and ERK were measured in immune complex assays with ATF2 (p38) and MBP (ERK) as substrates. Phosphorylated substrates were resolved by SDS-PAGE and detected by autoradiography. Autoradiograms are for representative experiments. Assays for 30 vs. 60 min are from the same gel but were analyzed in experiments separate from the 0-30 min experiments. All samples were preincubated for 1 h in control solution prior to hyperosmotic exposure; control activities at 0 min were measured immediately following this preincubation. B: substrate phosphorylation was quantified by densitometric analysis, and kinase activities are presented as % of activity measured in control solution at the same time point. Data are means ± SE for 3-7 independent experiments in each condition. * P < 0.05 vs. control.

PD98059 blocks hyperosmotic activation of p38 MAP kinase and ERK. The activation of MAP kinases by a variety of stimuli is mediated via MAP kinase kinases (MEK/MKKs), which are upstream dual-specificity kinases that phosphorylate and directly activate MAP kinases (16, 50). To investigate the role of MEKs in the hyperosmotic activation of MAP kinases in the MTAL, we tested the effects of PD98059, a selective MEK inhibitor (2). Inner stripe tissue was preincubated in control solution for 1 h with 15 µM PD98059. The tissue was then either maintained in control solution or exposed to hyperosmolality (300 mM added mannitol) for 15 min in the continued presence of the inhibitor. Identical samples were studied in each experiment without inhibitor. Pretreatment with PD98059 blocked the hyperosmotic activation of both p38 MAP kinase and ERK (Fig. 4, A and B). In control solution, PD98059 decreased basal ERK activity slightly (~20%) and had no effect on basal p38 MAP kinase activity (cont  vs. cont +, Fig. 4). PD98059 had no effect on JNK activity in control or hyperosmotic solution (data not shown). These data suggest that the activation of ERK and p38 MAP kinase by hyperosmolality involves PD98059-sensitive MEK(s).

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Effects of PD98059 on MAP kinase activation. A: inner stripe tissue was preincubated for 1 h in control solution in absence (cont ) or presence (cont +) of 15 µM PD98059. Tissue was then either maintained in control solution or exposed to hyperosmolality (300 mM added mannitol) for 15 min in absence (hyper ) or the continued presence (hyper +) of the inhibitor. Samples were assayed for p38 MAP kinase activity with ATF2 as substrate and ERK activity with MBP as substrate as described in METHODS. Phosphorylated substrates were resolved by SDS-PAGE and detected by autoradiography. Autoradiograms are of representative experiments. B: phosphorylated substrates were quantified by densitometric analysis, and p38 MAP kinase (open bars) and ERK (solid bars) activities are presented as % of activity measured in control solution in the absence of inhibitor (cont ). Data are means ± SE for 4 separate experiments in each group. * P < 0.05 vs. cont .

PD98059 does not block hyperosmotic inhibition of HCO₃ absorption. Hyperosmolality produced by the addition of mannitol or NaCl inhibits apical membrane Na⁺/H⁺ exchange and HCO₃ absorption in the MTAL (25, 54). To investigate whether activation of MAP kinases plays a role in this inhibition, we examined the effects of PD98059 on HCO₃ absorption (Fig. 5). MTAL were perfused and bathed in vitro with control solution, and then 15 µM PD98059 was added to the bath solution for 1 h (a time sufficient to block hyperosmotic activation of MAP kinases, Fig. 4). Bath addition of PD98059 reduced the basal rate of HCO₃ absorption slightly, from 13.1 ± 1.1 to a stable value of 11.5 ± 1.0 pmol · min¹ · mm¹ (control vs. PD98059, n = 4). In the continued presence of the inhibitor, increasing osmolality in the lumen and bath solutions by addition of 300 mM mannitol or 75 mM NaCl decreased HCO₃ absorption by 40%, from 10.5 ± 0.8 to 6.4 ± 0.6 pmol · min¹ · mm¹ (PD98059 vs. PD98059 + Hyper, n = 7; P < 0.001). This inhibition was reversible and is similar to that observed in previous studies in the absence of PD98059 (25, 30). The latter was confirmed in six additional control experiments in the present study: increasing osmolality in the absence of PD98059 by the addition of mannitol or NaCl decreased HCO₃ absorption from 11.8 ± 0.6 to 7.2 ± 0.7 pmol · min¹ · mm¹ (P < 0.001). Thus the inhibition of HCO₃ absorption by hyperosmolality is virtually identical in the absence and presence of the inhibitor. These data, together with the results of Fig. 4, demonstrate that activation of p38 MAP kinase and ERK is not necessary for hyperosmotic inhibition of HCO₃ absorption.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Effects of PD98059 on HCO3 absorption. MTAL were isolated and perfused in vitro, and HCO3 absorption rates (JHCO3) were measured as described in METHODS. Tubules were perfused and bathed initially in control solution (cont), and then 15 µM PD98059 was added to the bath for 1 h. In the continued presence of the inhibitor, perfusion and bath solutions were made hyperosmotic by addition of 300 mM mannitol () or 75 mM NaCl (). Data points are average values for individual tubules; lines connect paired measurements made in the same tubule. JHCO3 was measured in the initial control period prior to the addition of PD98059 in 4 of 7 experiments. Mean HCO3 absorption rates are given in the text.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The adaptive responses of cells to hyperosmotic stress include changes in the activities of ion transporters, increased expression of enzymes and transporters responsible for accumulation of organic osmolytes, and the induction of immediate early genes and heat shock proteins (9, 11, 13, 37). Hyperosmolality activates several MAP kinase pathways in mammalian cells, but their role in the physiological responses to osmotic stress is not understood. Previously, we demonstrated in the MTAL of the rat that hyperosmolality inhibits transepithelial HCO₃ absorption through inhibition of apical membrane Na⁺/H⁺ exchange (25, 54), effects that may be important physiologically for limiting delivery of HCO₃ to the medullary interstitial fluid during antidiuresis (23, 24). In the present study, we examined the osmotic regulation of MAP kinases in the MTAL and determined their possible role in this inhibitory response. We found that hyperosmolality differentially regulates MAP kinases in the MTAL: it increases the activity of p38 MAP kinase and ERK but has no effect on JNK. We also show that blocking the activation of p38 MAP kinase and ERK does not prevent the inhibition of HCO₃ absorption by hyperosmolality. As discussed below, these findings indicate that MAP kinase pathways are not involved in hyperosmotic inhibition of the apical membrane Na⁺/H⁺ exchanger isoform NHE3.

The molecular mechanisms involved in osmotic regulation of MAP kinase pathways are a subject of recent intense investigation. A tissue in which these osmosensing pathways are likely to be of particular physiological relevance is the renal medulla, where cells normally are exposed to a variable and profoundly hyperosmotic environment (37). In the rat, the osmolality of the renal medulla can vary from 290 mosmol/kgH₂O to more than 1,500 mosmol/kgH₂O in response to changes in the diuretic state of the animal. The osmolalities used in the present study (290-590 mosmol/kgH₂O) thus represent a reasonable estimate of the range of values expected to surround the MTAL in the outer medulla in vivo. We have shown previously that virtually identical tyrosine kinase-dependent inhibition of HCO₃ absorption is observed in the MTAL regardless of whether osmolality is increased by the addition of mannitol or NaCl, the physiological solute responsible for the hyperosmolality of the renal outer medulla in vivo (25). Thus the changes in the activities of membrane transporters and signaling pathways that we observed reflect physiological responses to osmotic stress, independent of the particular solute used to produce hypertonicity. The effects of osmotic stress on the activities of MAP kinases have been studied in renal cell lines (6, 34, 38, 51, 56), but no studies have examined the osmotic regulation of MAP kinase pathways in intact renal tubules. Therefore, a goal of our study was to develop methods to measure the activities of distinct MAP kinases in native MTALs, an approach essential to define the physiological relevance of these signaling pathways. We show that p38 MAP kinase, ERK, and JNK activities can be assayed in microdissected MTALs and that these tubules are viable for physiological studies of the regulation of MAP kinase activities. Furthermore, our results demonstrate the feasibility of using immune complex assays to study protein kinase activities in microdissected renal tubules, an approach that should prove useful to uncover the regulation and functional interactions of a wide variety of signaling pathways in intact and precisely defined nephron segments.

Hyperosmolality differentially activates MAP kinases in the MTAL. ERK, JNK, and p38 MAP kinase were detected by immunoblotting and by enzyme assay (Fig. 2), indicating that all three pathways are constitutively expressed in the MTAL. Hyperosmotic stress can activate all three pathways simultaneously in renal cell lines (6, 34, 38, 51, 56). However, we found that hyperosmolality differentially activates MAP kinases in intact MTALs; p38 MAP kinase and ERK activities were increased, whereas JNK activity was unchanged. Hyperosmolality induced a rapid (5 min) and sustained (up to 60 min) activation of p38 MAP kinase, suggesting that this pathway plays an important role in the early adaptive response of the MTAL to changes in osmolality. p38 MAP kinase is a mammalian homolog of HOG1, a stress-activated MAP kinase in yeast that is coupled to increased synthesis of the osmolyte glycerol, which permits growth in hypertonic medium (8). p38 MAP kinase also is involved in the activation of MAPKAP kinase, which phosphorylates small heat shock proteins (19, 45). These findings suggest that the p38 MAP kinase pathway could play a role in renal medullary cells in the accumulation of compatible organic osmolytes (9) or in the induction of heat shock proteins (9, 13), processes that stabilize macromolecules in the early response to osmotic shrinkage.

Hyperosmolality also activated ERK, but with a time course different from that of p38 MAP kinase (Fig. 3). Activation of ERK was less rapid than activation of p38 MAP kinase and was transient, with activity returning to basal levels within 30 min. A similar transient activation of ERK by hypertonicity has been reported in MDCK cells in association with the downstream activation of S6 kinase (51). At present, the role of the ERK pathway in the physiological response of mammalian cells to osmotic stress is unknown. Eliminating activation of ERK did not prevent the hyperosmotic activation of gene transcription for myo-inositol and betaine transporters in MDCK cells (38) or increased inositol uptake in inner medullary collecting duct IMCD-3 cells (6). Thus there is no evidence for involvement of the ERK pathway in mediating the adaptive accumulation of organic osmolytes in renal cells. As discussed below, ERK also does not appear to be involved in the osmotic regulation of Na⁺/H⁺ exchange.

The lack of effect of hyperosmolality on JNK activity was unexpected in view of the rapid and nearly universal activation of this pathway by osmotic stress in a wide variety of mammalian cell lines (6, 7, 20, 36, 42, 44). The absence of JNK activation by hyperosmolality in our experiments is not due to technical limitations, because we readily observed an increase in JNK activity in response to anisomycin. We also have observed a marked increase in JNK activity in both the inner stripe of the outer medulla (18) and in microdissected MTALs (J. F. Di Mari, R. Safirstein, and D. W. Good, unpublished observations) in response to 10-min ischemia-reperfusion using the same methods. In other systems, the JNK and p38 MAP kinase pathways share common activation by a variety of environmental stresses (14, 39, 44, 45, 50), can be activated by upstream MKK4 (16, 41, 46), induce apoptosis in PC12 cells (55), and complement yeast strains defective in HOG1 expression (20, 31), indicating significant overlap and functional redundancy in the two pathways. In the MTAL, however, we find that the p38 MAP kinase and JNK pathways are functionally distinct: hyperosmolality activates p38 MAP kinase but not JNK (Figs. 1 and 2), whereas ischemia and reperfusion activates JNK but not p38 MAP kinase (18). Thus, in the MTAL, p38 MAP kinase is activated as an early response to osmotic stress; JNK is activated as an early response to oxidative stress induced by reperfusion injury (18). Possible explanations for selective osmotic activation of the p38 MAP kinase stress pathway include activation of upstream regulators specific for p38 MAP kinase (MKK3 or MKK6), osmotic activation of secondary feedback events that shut down activation of the JNK pathway, and/or a lack of activation of adaptor proteins necessary for functional regulation of the JNK pathway (10, 16, 50). Our results do not rule out the possibility that late (> 30 min) activation of JNK may occur in the MTAL and contribute to long-term osmotic adaptations such as regulation of gene transcription. However, JNK appears to play no role in the effect of short-term hyperosmolality to inhibit HCO₃ absorption.

The separate regulation of ERK, JNK, and p38 MAP kinase is achieved through protein kinase cascades that result in the activation of distinct MAP kinase kinases (MEK/MKKs), which then directly and selectively activate MAP kinases (10, 16, 31, 50). In yeast and in mammalian cells, the osmotic activation of MAP kinases occurs through activation of these upstream kinase cascades. For example, hyperosmolality activates the Raf-MEK pathway that activates ERK (36, 42, 51), the MEKK-MKK4 pathway that activates JNK (42, 43), and MKK3, a direct and selective activator of p38 MAP kinase (16). We found in the MTAL that activation of ERK by hyperosmolality was prevented by PD98059, a selective inhibitor of MEK1, a kinase that directly phosphorylates and activates ERK (2). Hence, the osmotic activation of ERK likely is mediated through activation of MEK. PD98059 also blocked hyperosmotic activation of p38 MAP kinase, which is not a direct substrate for MEK1. It is possible that PD98059 may inhibit upstream MAP kinase kinases that are selective for p38 MAP kinase, such as MKK3 (16) or MKK6 (32). Alternatively, PD98059 may cause activation or inhibition of upstream components of the ERK pathway that prevents osmotic activation of p38 MAP kinase through an as yet unidentified cross-regulatory mechanism.² Resolution of these issues will require identification of the upstream activators of p38 MAP kinase and ERK in the MTAL and analysis of possible crosstalk between these two pathways. Our results establish, however, that PD98059 blocks hyperosmotic activation of p38 MAP kinase in the intact MTAL, thus providing an effective means by which to assess the possible involvement of this pathway in the physiological responses to osmotic stress.

Hyperosmolality activates MAP kinases and inhibits HCO₃ absorption via distinct pathways. Hyperosmolality markedly inhibits HCO₃ absorption in the MTAL (23, 25). This inhibition is blocked by the protein-tyrosine kinase inhibitors genistein and herbimycin A, suggesting that tyrosine phosphorylation is an important component of the signaling pathway that leads to transport inhibition (25). Genistein also inhibited hyperosmotic activation of ERK and p38 MAP kinase (Fig. 1), consistent with the possible involvement of these signaling pathways in the transport inhibition. Further studies revealed, however, that eliminating the activation of p38 MAP kinase and ERK with the selective MEK inhibitor PD98059 did not prevent inhibition of HCO₃ absorption by hyperosmolality (Figs. 4 and 5). We also have found that pretreatment of MTALs for up to 2 h with the selective p38 MAP kinase inhibitor SB203580 (15 µM) (40) blocks hyperosmotic activation of p38 MAP kinase but does not affect the inhibition of HCO₃ absorption (data not shown). Thus activation of p38 MAP kinase, ERK, or JNK is not required for hyperosmotic inhibition of HCO₃ absorption, indicating that the osmotic regulation of HCO₃ absorption occurs via activation of a signal transduction pathway distinct from these MAP kinase pathways. These results also indicate that genistein blocks hyperosmotic inhibition of HCO₃ absorption through mechanisms other than its action to inhibit MAP kinase activation. At this point, we do not know whether genistein may inhibit the activation of two independent osmosensing signaling pathways or whether it inhibits a single pathway (possibly at the level of the osmotic sensor) that diverges to regulate distinct pathways coupled to MAP kinase activation and HCO₃ transport inhibition. The osmotic sensor(s) that initiate intracellular signals in response to hyperosmotic stress in mammalian cells have not been identified.