THE JOURNAL OF BIOIQCICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 269, No. 49, Issue of December 9, pp. 31296-31301, 1994 Printed in U.S.A.

Sequential Activation of Raf-1 Kinase, Mitogen-activated Protein ( M A P ) Kinase Kinase, MAP Kinase, and S6 Kinase by Hyperosmolality in Renal Cells*

(Received for publication, April 5, 1994, and in revised form, August 15, 1994)

Yoshio Terada$, Kimio Tomita, Miwako K. HommaO, Hiroshi Nonoguchi, Tiaxin Yang, Takehisa Yamada, Yasuhito YuasaO, Edwin G. Krebsn, Sei Sasaki, and Fumiaki Marumo From the Second Department of Internal Medicine and the $Department of Hygiene and Oncology, Tokyo Medical and Dental University, Tokyo 113, Japan and the IDepartment of Pharmacology, Howard Hughes Medical Institute, University of Washington, Seattle, Washington 98195

In the renal medulla during antidiuresis, the extracel- lular fluid becomes hyperosmotic. Madin-Darby canine kidney (MDCK) epithelial cells adapt in hyperosmotic conditions and serve as a useful tissue culture model for cellular responses to hyperosmolality. We demonstrate that hyperosmolality stimulates phospholipase C, Raf-1 kinase mitogen-activated protein ( M A P ) kinase kinase, MAP kinase, and S6 kinase activities and that it increases phosphorylation of Raf-1 kinase, and p42 MAP kinase in MDCK cells. Stimulation of these kinases is osmolality- dependent (from 300 to 600 mosmkg H,O). The time course of activation is sequential; the peak stimulation for Raf-1 kinase is at 5 min, at 10 min for MAP kinase kinase and MAP kinase, and at 20 min for S6 kinase. The activation of Raf-1 kinase and MAP kinase is inhibited by phorbol 12-myristate 13-acetate pretreatment in the presence of calphostin C or H-7. Tyrosine kinase inhibi- tors (genistein, herbimycin) do not significantly sup- press hyperosmolality-induced MAP kinase activity. The increase of Ins-1,4,5-P, levels by hyperosmolality sug- gests that activation of these kinases is mediated at least partially via activation of phospholipase C. Thus, hyperosmolality stimulates the serine/

threonine kinases, Raf-1 kinase, MAP kinase kinase, MAP kinase, and 56 kinase, via predominantly protein kinase C-dependent, tyrosine kinase-independent path- ways in MDCK cells.

When a normal human is dehydrated and the urine is highly concentrated, the osmolality in the renal medulla increases to over 1000 mosmkg H,O (1). The Madin-Darby canine kidney (MDCK)' epithelial cell line is considered to have characteris- tics of distal nephron segments and can tolerate extremes of osmolality and, therefore, is a useful model for cellular re- sponses to hyperosmolality. MDCK cells grow well in hyper-

entific Research in Japan 05837007, 04454234, and 04670383 and by * This work was partly supported by Grants-in-Aid for General Sci-

the Mochida Memorial Foundation for Medical and Pharmaceutical Research in 1992. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$. To whom correspondence and reprint requests should be addressed: Second Department of Internal Medicine, Tokyo Medical and Dental University, 5-45, Yushima l-chome, Bunkyo-ku, Tokyo 113, Japan. Tel:

The abbreviations used are: MDCK, Madin-Darby canine kidney; M A P , mitogen-activated protein; MAPK, MAP kinase; M A P K K , MAP kinase kinase; MEKK, MAPKK kinase; DTT, dithiothreitol; PMA, phor- bo1 12-myristate 13-acetate; H-7 l-(5-isoquinolinesulfonyl)-2-meth- ylpiperazine.

813-3813-6111; F a : 813-3818-7177.

tonic medium from 300 to 915 mosmkg H,O and accumulate betaine, inositol, and glycerophosphorylchlorine under hyper- osmolar conditions (2,3,4). Two types of observation have been made in cells treated with hypertonic medium. In one case, acute changes in cell function such as activation of transporters in response to hyperosmolality have been studied (5). In the other case, gene regulation in response to hyperosmolality has been studied (4, 6). However, how these two mechanisms are connected and what signaling systems are used to trigger these responses are not fully defined. To provide a more complete picture of the signals generated by hyperosmolality, we inves- tigated signaling pathways and protein kinase cascades from the cell membrane to the nucleus. Recently, Brewster et al. (7) reported that hyperosmolality activates members of the mito- gen-activated protein kinase (MAPK) and MAP kinase kinase ( M A P K K ) families in yeast (7). However, they did not deter- mine how MAPKK is activated (7). It is not clear whether the activation of MAPK by hyperosmolality is a general biological phenomenon, because not all regulatory pathways, especially upstream of MAPKK, are the same in mammalian cells and yeast (7). "illy et al. (8) reported that protein tyrosine phospho- rylation is involved in hyposmotic stimulation in a human in- testinal cell line. Although many growth factors and hormones are reported to activate MAPK, it is not known what steps are involved in signal transduction caused by hyperosmolality in mammalian cells.

Recently, models of signal transduction pathways have been proposed in which a number of molecules act in series to acti- vate MAP kinases (9-13). MAPKK appears to be a convergence point for two pathways, one originating at receptor and non- receptor tyrosine kinases where information travels through ~21"" to Raf-1 and then to MAPKK and the other originating at G-protein-coupled signaling systems through MEKK WAl"AP kinase) to MAPKK. Following activation of MAPKK by either pathway, MAP kinases are activated, and they in turn may activate pp90rsk, transcription factors, or a variety of other po- tential substrates (14-17). Protein kinase C could be activated by either the tyrosine kinase or G-protein-dependent pathways through phospholipase Cy or - p and may activate either Raf-1 kinase or MEKK (12, 13). In this study, we investigated the hyperosmolality-activated signaling pathway from the cell membrane to the cytosolic kinases that are known to be ki- nases for transcription factors and other regulatory molecules. We demonstrate that hyperosmolality induced by NaCl or raf- finose activates phospholipase C, Raf-1 kinase, MAPKK, MAPK, and S6 kinase. Therefore, we have defined a signaling pathway in MDCK cells activated by hyperosmolality from the cell membrane to second messenger-independent serine/ threonine kinases.

31296

Activation of MAP Kinase in MDCK Cells 31297

MATERIALS AND METHODS

Cell Culture and Preparation of Extracts-MDCK cells originally purchased from the American Type Culture Collection (Rockville, MD) were grown in a defined medium: a 5050 mixture of Dulbecco's modified Eagle's medium with 5 m~ glucose and Coon's modified Eagle's medium supplemented with 10 m~ HEPES, 5 PM triiodothyronine, 50 1.1~ hydro- cortisone, 10 I" Na,SeO,, transferrin at 5 pg/ml, insulin at 5 pg/ml, prostaglandin E, at 25 ng/ml, 2 mM L-glutamine, penicillin a t 50 IU/ml, and streptomycin at 50 pglml. The osmolality of the medium was ad- justed to 300 mosmkg H,O by adding NaHCO, at 30 mM. The medium was made hypertonic by adding NaCl or raffinose to 400,500, and 600 mosmkg H,O. After varying lengths of incubation time at 37 "C in hyperosmotic medium, cells were scraped into a total volume of 0.5 ml of extraction buffer containing 50 mM p-glycerophosphate, pH 7.3, 1.5 mM EGTA, 0.1 mM Na,VO,, 1 mM DTT, 10 pg/ml leupeptin, 10 pg/ml aprotinin, 2 pg/ml pepstatin A, and 1 mM benzamidine. Cells in extrac- tion buffer were sonicated for 20 s (Sonicator, Central Co., Tokyo, Japan) and centrifuged at 100,000 x g for 20 min at 4 "C (Beckman TL-100 centrifuge). Supernatants (-0.3 mg/ml protein) were stored frozen at

Polyclonal Antibodies-Antibodies against MAPK were prepared as described (18). The residues used for the antigenic peptides are con- served in both p42 MAPK and p44 MAPK. Antibodies against Raf-1 protein were raised against a C-terminal peptide (amino acid sequence, CTLTTSPRLPW). For immunization, the peptide was conjugated to keyhole limpet hemocyanin with glutaraldehyde. Adult male New Zea- land White rabbits were immunized subcutaneously with the conjugate emulsified with either complete or incomplete Freund's adjuvant four times at 2-week intervals. The rabbit sera were tested for the presence of antipeptide antibodies by enzyme-linked immunoassay using plates coated with peptide only. Positive samples were further tested for spec- ificity to Raf-1 protein by Western blotting, using NIH3T3 cells overex- pressing the protein (data not shown). The anti-phosphotyrosine anti- body was purchased from UBI (Lake Placid, NY).

Kinase Assays-MDCK cell cytosolic extracts (1.0 pg) were mixed with 4 pl of 2.0 mg/ml myelin basic protein for MAPK assays or 4 pl of 18.0 mg/ml S6 peptide (RRRLSSLRA) (UBI) for S6 kinase assays. Phos- phorylation reactions were initiated by adding 8.3 p1 of kinase assay buffer containing 50 mM p-glycerophosphate, pH 7.3, 3.75 mM EGTA, 0.15 mM Na,VO,, 1.5 mM DTT, 30 p~ calmidazolium, 6 p~ protein kinase inhibitor peptide, 30 mM MgOl,, 0.1 mM ATP, and 0.3 mM [y32PlATP (specific activity -2000 cpdpmol) and incubated at 30 "C for 15 min. The reaction was terminated by spotting 20 p1 of assay mixture onto P-81 phosphocellulose filter paper (Whatman, Maidstone, England), which was washed several times with 175 mM phosphoric acid, and radioactivity was counted. In filter paper assays of kinase activity in cytosolic extracts, levels of phosphate incorporation measured in the absence of substrate were subtracted from values in the presence of substrate to correct for nonspecific phosphorylation.

MAPKKAssays-MDCK cytosolic extracts (1.0 pg) were mixed with 5.2 pl of 4.8 pg/ml MAPK protein (UBI), and phosphorylation reactions were initiated by adding 8.3 pl of the kinase assay buffer described above without Iy3zPlATP. After incubation at 30 "C for 10 min, myelin basic protein phosphorylation was initiated by adding 4 pl of 0.33 mg/ml myelin basic protein and 0.3 mM [Y-~'P]ATP (specific activity -2000 cpdpmol) and incubated at 30 "C for 15 min. The reaction was termi- nated by spotting 20 pl of assay mixture onto P-81 phosphocellulose filter paper (Whatman), and radioactivity was counted. In this MAPKK assay, levels of phosphate incorporation measured in the absence of exogenous MAPK protein were subtracted from values in the presence of exogenous MAPK protein to correct for nonspecific phosphorylation.

Immunoprecipitation-After varying lengths of incubation time at 37 "C in hyperosmotic medium, cells were lysed into a total volume of 0.5 ml of lysis buffer containing 10 mM /3-glycerophosphate, pH 7.4, 1.0 mM EDTA, 5.0 mM EGTA, 0.15 mM Na,VO,, 2.0 m~ DTT, 20 m~ HEPES, pH 7.4, 0.5% (v/v) Nonidet P-40, 0.1% (w/v) sodium, deoxy- cholate, 1 m~ phenylmethylsulfonyl fluoride (PMSF), 10 pg/ml aproti- nin, and 10 pg/ml leupeptin. Lysates were allowed to remain on ice for 30 min and were then centrifuged at 15,000 x g for 30 min at 4 "C. Supernatants were incubated for 2 h at 4 "C with 3 pl of rabbit anti- phosphotyrosine antibody (UBI) and protein G-Sepharose. Immune complexes were recovered by centrifugation. Pellets were washed three times with 50 m~ Tris, pH 8.5,150 mM NaCl, 0.02% NaN,, 1.0% Nonidet P-40,0.5% sodium deoxycholate, 0.5 mM Na,VO,, 5 mM Dm, and 1 mM phenylmethylsulfonyl fluoride. Pellets were solubilized in SDS sample buffer. The samples were resolved by SDS-polyacrylamide gel electro- phoresis. The proteins were then transferred to an Immobilon P mem-

-80 "C.

u 600mOsm/ kgHpO

cells. MDCK cells were incubated in hyperosmotic medium (600 FIG. 1. Hyperosmolality increased Ins-1,4,5-P, levels in MDCK

mosmkg H,O using NaCl or raffinose) for 5 min. Ins-1,4,5-P, levels

hyperosmotic medium (600 mosmkg H,O) (n = 5, mean f S.E.). were measured after incubation with control (300 mosmkg H,O) or

brane (Daiichikagaku, Tokyo, Japan). To detect MAPK, the membrane was incubated with anti-MAPK antibody at a dilution of 1500 in Tris- buffered saline containing 0.1% Tween 20 for 2 h at 37 "C and visualized with 9-labeled goat anti-rabbit IgG (0.5 pCi/ml). 1251-labeled MAP kinase protein bands were visualized by autoradiography. Raf protein was immunoprecipitated with a rabbit anti-Raf-1 protein. Immune com- plexes were obtained using exactly the same protocol as described above. The pellets were used in immune complex kinase assays accord- ing to Siege1 et al. (19) in a 50-pl volume of 10 mM Tris-HC1, pH 7.5, 10 mM MnCl,, 150 mM NaCl, 10 mM MgCl,, 2 mM DTT, and 1.2 mM [y-32PlATP (specific activity -2000 cpndpmol), for 5 min at 37 "C. Re- actions were terminated by adding SDS and boiling for 5 min. The samples were then resolved by SDS-polyacrylamide gel electrophoresis, and phosphorylation of Raf-1 protein was detected by autoradiography.

Down-regulation of Protein Kinase C and Drosine Kinase Actiui- ties-% deplete PMA-sensitive protein kinase C, we preincubated MDCK cells with 100 nM phorbol 12-myristate 13-acetate (PMA) for 24 h before hyperosmotic incubation. To inhibit protein kinase C activ- ity, we incubated MDCK cells in the presence of 100 I" calphostin C or 10 p~ l-(5-isoquinolinesulfonyl)-2-methylpiperazine (H-7) for 30 min before hyperosmotic incubation. To inhibit tyrosine kinase activity, we incubated MDCK cells in the presence of 20 1.1~ genistein or 1 mM herbimycin for 30 min before hyperosmotic incubation. These inhibitors were also added to the hyperosmotic medium. The cells were lysed and centrifuged, and supernatants were assayed for MAPK activity, as de- scribed above.

Inositol 1,4,btriphosphate Measurements-MDCK cells were incu- bated with hypertonic medium (600 mosmkg H,O by adding NaCl or raffinose) for 5 min at 37 "C. Incubations were terminated by aspirating the medium, washing with phosphate-buffered saline, and immediately adding 1 ml of trichloroacetic acid and then 1 ml of distilled water. Both fractions were collected in a tube. The extracts were centrifuged at 5000 x g for 10 min, and the resulting supernatant was washed 4 times with diethyl ether saturated with distilled water to remove trichloroacetic acid. The pH of the supernatant was adjusted to pH 7.4 with 10% NaHCO,, and aliquots of the supernatant were used for Ins-1,4,5-P3 assay kit (Amersham Corp.), as described by Fukami et al. (20).

Statistics-The results were given as means 2 S.E. The differences were tested using analysis of variance. p 0.05 was considered significant.

RESULTS

Hyperosmolality Increased Ins-1,4,5-P3 levels in MDCK Cells (Fig. 1)-The increase in protein kinase C-like activity has been observed in lymphocytes in response to osmotic shrinkage

31298 Activation of MAP Kinase in MDCK Cells

A Timehid kDa 0 5 10 20 30 60

97 - e"00 ... CRaf-1

66 -

B mOsm/kg H20 mOsm/kg H2O by NaCl "

by raffinose

0 0 0 0 0 0 0 0 kDa a z s g

97 - CRaf-l

66 - FIG. 2. Hyperosmolality stimulated the phosphorylation of

Raf-1 kinase in MDCK cells. A, MDCK cells were preincubated with NaCl at 600 mosm/kg H,O for 5, 10, 20, 30, and 60 min. B, the osmo- lality of the medium was increased by NaCl or raffinose to 400,500, and 600 mosmkg H,O in MDCK cells for 10 min. Autophosphorylation of Raf-1 kinase was detected in anti-Raf-1 kinase immunoprecipitates incubated with [y121ATP.

(21). Therefore, we studied whether hyperosmolality induced by NaCl or raffinose was able to enhance Ins-1,4,5-P3 formation to detect the activity of phospholipase C. When osmolality was increased with NaCl or raffinose to 600 mosmkg H,O in MDCK cells, Ins-1,4,5-P3 level was significantly increased in hyperos- motic conditions with NaCl or raffinose, respectively, by 5 min (Fig. 1). There was no statistically significant difference be- tween NaCl and raffinose in the levels of Ins-1,4,5-P3 at 600 mosmkg H,O. Values are means f S.E. of 5 observations.

Hyperosmolality Stimulated the Phosphorylation of R a f l Kinase in MDCK Cells (Fig. 2)"I'he activation of phospho- lipase C leads to an increased Ins-1,4,5-P3 level and protein kinase C activity. Protein kinase C was reported to activate Raf-1 kinase by direct phosphorylation (22). Therefore, we studied whether hyperosmolality induced by NaCl or raffinose was able to activate Raf-1 kinase. To examine hyperosmolality- stimulated autophosphorylation of Raf-1 kinase, MDCK cells were preincubated with NaCl or raffinose at 400, 500, or 600 mosmkg H,O in MDCK cells for 10 min. Autophosphorylation of Raf-1 kinase was detected in anti-Raf-1 kinase immunopre- cipitants incubated with [-y-32P]ATP. Fig. 24 shows the time course of autophosphorylation of Raf-1 kinase in response to 600 mosmkg H,O in MDCK cells. Autophosphorylation of Raf-1 kinase increased to a maximum level at 5 min and then gradually decreased up to 60 min after hyperosmotic incuba- tion. As shown in Fig. 2B, autophosphorylation of Raf-1 kinase is osmolality-dependent (300,400,500, and 600 mosmkg H,O by NaCl or raffinose).

Hyperosmolality-induced Stimulation of MAPKK Activity Pig. 3)"Raf-1 kinase is reported to phosphorylate MAPKK (14). We investigated the time course for activation of MAPKK by NaCl or raffinose a t 600 mosmkg H,O at 5, 10,20,30, and 60 min. MAPKK activity was stimulated within 5 min and reached a maximum level at 10 min (Fig. 3A). This stimulation was dependent on the osmolality (300, 400, 500, and 600 mosmkg H,O) (Fig. 3B). There was no statistically significant difference between NaCl and raffinose in the activation of MAPKK in 400,500, and 600 mosmkg H,O. Hyperosmolality-stimulated MAPKActivity and Phosphoryl-

ation of p42 MAP Kinase Protein (Fig. 4)"we studied whether hyperosmolality induced by NaCl or raffinose was able to acti- vate MAPK activity. When osmolality was increased with NaCl

A B

0 1 0 5 10 20 30 60 300 400 500 600

Time (min) mOsm/kgH,O

FIG. 3. Stimulation of MAPKK activity by hyperosmolality. A, MDCK cells were incubated in hyperosmotic medium (600 mosmkg H,O using NaCl (0) or raffinose (0)) for 5, 10, 20, 30, and 60 min. MAF'KK activity was measured after preincubation with or without MAF'K protein. B, this stimulation was dependent on the osmolality (300,400,500, and 600 mosmkg H,O). At each osmolality, there was no statistically significant difference in stimulation of MAPKK by NaCl or raffinose. A and B, n = 5, mean 2 S.E.

A 300 r

0 0 300 400 500 600 0 5 IO 20 30 60

mOsm kgH.0 Tlme (mm)

C 2 mOsm/kg HpO mOsm/kg H,O 2 by NaCl 3" - by raffinose

49.5- 49.5.

32.5- 32.5.

c

phorylation of p42 MAF'K protein. A, the osmolality of the medium FIG. 4. Hyperosmolality stimulates MAPK activity and phos-

was increased by NaCl(0) or raffinose (0) to 400,500, and 600 mosmkg H,O in MDCK cells, and MAPK activity was measured. At each osmo- lality, no difference was measured between NaCl and raffinose (n = 5, mean e S.E.). B, time course of MAPK activity stimulated a t 600 mosmkg H,O by NaCl (0) or raffinose (0) (n = 5, mean f S.E.). C, MDCK cells were incubated in hyperosmotic medium for 10 min. Qro- sine phosphorylation of MAF'K was observed in anti-phosphotyrosine immunoprecipitates probed with anti-MAF'K. The left lane is a Western blot with anti-MAF'K antibody of MDCK cell lysate from untreated cells.

or raffinose to 400,500, and 600 mosmkg H,O in MDCK cells, MAPK activity was stimulated in an osmolality-dependent manner (Fig. 4A). The time course of MAPK stimulation is shown in Fig. 4B. MAPK activity was increased with 600 mosmkg H,O with NaCl or raffinose within 5 min and reached a maximum level at 10 min after the hyperosmotic incubation (Fig. 4B). There was no statistically significant difference be- tween NaCl and raffinose in the activity of MAPK a t 400,500, and 600 mosmkg H,O and in the time course.

Activation of MAP Kinase in MDCK Cells 3 1299

0 5 10 20 30 60 300 400 500 600 Time (mi& m0sm/kgH20

MDCK cells were incubated with hyperosmotic medium (600 mosmkg FIG. 5. Stimulation of S6 kinase activity by hyperosmolality. A,

H,O with NaCl (0) or raffinose (0)) for 5, 10, 20, 30, and 60 min. S6 kinase activity was measured by incubation with or without Rsk sub- strate. B, this stimulation was dependent on the osmolality (300, 400, 500, and 600 mosmkg H,O). At each osmolality, there was no statisti- cally significant difference in stimulation of S6 kinase by NaCl or raf- finose. A and B, n = 5, mean 2 S.E.

To examine MDCK cells expressing both forms of MAPK, immunoblot of crude cell lysate was performed. As shown in the left lane of Fig. 4C, clear bands were detected at 42 kDa and 44 kDa. To further examine the hyperosmolality-stimulated tyrosine phosphorylation of MAPK, MDCK cells were preincu- bated with NaCl or raffinose at 400,500, or 600 mosmkg H,O for 10 min. Tyrosine phosphorylation of p42 MAPK was in- creased in an osmolality-dependent manner in anti-phosphoty- rosine immunoprecipitants probed with anti-MAPK antibody (Fig. 4C). Tyrosine phosphorylation of p44 MAPK was not de- tected in our experimental conditions. Hyperosrnolality-induced Stimulation of S6 Kinase Activity

(Fig. 5)-S6 kinase is known to be phosphorylated and acti- vated by MAPK (14-17). S6 kinase may act as a mediator between signal transduction pathways and intranuclear events. Thus, we measured the S6 kinase activity induced by hyperosmolality. To investigate the time course for activation of S6 kinase by NaCl or raffinose at 600 mosmkg H,O in MDCK cells, we examined S6 kinase activity at 5, 10, 20, 30, and 60 min. S6 kinase activity was stimulated from 5 min and reached a maximum level at 20 min (Fig. 5A). This stimulation was dependent on the osmolality (300, 400, 500, and 600 mosmkg H,O) (Fig. 5B). There was no statistically significant difference between NaCl and raffinose in the activity of S6 kinase in 400, 500, and 600 mosmkg H,O.

Effects of PMA, Calphostin C, H-7, Genistein, and Herbimy- cin on Hyperosmolality-stimulated MAPK Activity in MDCK Cells (Fig. 6)-We examined whether MAPK activation by hy- perosmolality is induced via protein kinase C-independent or -dependent pathways. We used protein kinase C inhibitors, calphostin C and H-7, and PMA-sensitive protein kinase C depletion by PMA pretreatment. As shown in Fig. 6 A , PMA and hyperosmolality activated MAPK in MDCK cells that were not pretreated with PMA. However, when MDCK cells were pre- treated with 100 nM PMA for 24 h, hyperosmolar stimulation induced at 500 mosmkg H,O by NaCl or raffinose was de- creased by approximately 86.5 ? 10.8% or 88.6 * 13.1%, respec- tively ( n = 5, mean & S.E.) as compared with that seen without pretreatment. PMA stimulation of MAPK activity decreased significantly in PMA-pretreated MDCK cells. Prior treatment with calphostin C for 30 min, a potent and specific protein kinase C inhibitor, also inhibited hyperosmolality-induced MAPK activation by NaCl or raffinose by approximately 76.7 12.5% or 79.8 * 10.9%, respectively ( n = 5, mean i: S.E.). H-7 also inhibited hyperosmolality-induced MAPK activation by

NaCl or raffinose by approximately 83.5 2 11.9% or 80.1 2

12.4%, respectively (n = 5, mean 2 S.E.). These data demon- strated that hyperosmolality-induced MAPK activation was mainly dependent on the activation of PMA-sensitive protein kinase C.

PKC and Raf-1 kinase could be activated by either the tyro- sine kinase or G-protein-coupled receptor pathways (12, 13, 22). In order to distinguish between these two possibilities, we used two tyrosine kinase inhibitors, genistein and herbimycin. We tested the effects of genistein and herbimycin on EGF- stimulated MAPK activity in MDCK cells. As shown in Fig. 6B, incubation of MDCK cells with M epidermal growth factor in the presence of 40 p~ genistein resulted in a significant decrease (87.9 12.7%) of stimulated MAPK activity. One mM herbimycin also caused a significant decrease (84.6 & 14.1%) of epidermal growth factor-stimulated MAPK activity. We studied hyperosmolality-stimulated MAPK activity in MDCK cells in- cubated for 30 min with genistein and incubated for 60 min with herbimycin before incubation with hyperosmotic medium. As shown in Fig. 6 A , incubation of MDCK cells with hyperos- motic medium (by NaCl or raffinose) in the presence of 40 p~ genistein resulted in a small decrease (12.7 & 5.8% or 11.1 &

5.1%, respectively) of stimulated MAPK activity. One mM her- bimycin caused a small decrease (8.9 * 4.1% or 7.5 & 3.8%) of hyperosmolality-stimulated MAPK activity.

DISCUSSION

The major findings of the present study are that hyperosmo- lality stimulates phospholipase C, MAPKK, MAPK, and S6 kinase activity and increases phosphorylation of Raf-1 kinase, p42 MAPK, and pp9WSk, in MDCK cells. These stimulations of kinase activity show a dose dependence on osmolality. The time course of activation is sequential; Raf-1 kinase peaks a t 5 min, MAPKK and MAPK peak at 10 min, and S6 kinase peaks at 20 min. This activation of MAPK is inhibited 80% by PMA pretreatment or in the presence of calphostin C or H-7. Tyro- sine kinase inhibitors (genistein and herbimycin) do not inhibit hyperosmolality-induced MAPK activity. This evidence sug- gests that the activation of serinekhreonine cascades by hyper- osmolality is largely dependent on PMA-sensitive protein ki- nase C and independent of tyrosine kinase.

Recently, PKC was reported to activate Raf-1 kinase by di- rect phosphorylation (22). Raf-1 kinase was also known to be phosphorylated by p2lrQs, which is activated by receptor or non- receptor tyrosine kinases (23). MAPKK was reported to be phosphorylated by Raf-1 kinase and MEKK (12, 13, 14). Thus, hyperosmolality may activate phospholipase C and protein ki- nase C or protein kinase C-like kinase, which then phosphoryl- ate Raf-1 kinase. The Raf-1 kinase then activates MAPKK, and the MAPKK activates MAPK. Our time course results support this sequential activation. Our experiments using tyrosine ki- nase inhibitors and protein kinase C inhibitors showed that this activation of serinehhreonine cascades by hyperosmolality is via predominantly PMA-sensitive protein kinase C-depend- ent, tyrosine kinase independent pathways. We will discuss the possible mechanism and physiological function of the hyper- osmolality-stimulated signal transduction pathway from cell membrane to the nucleus.

These results raise questions as to how hyperosmolality stimulates serinehhreonine kinase cascades in MDCK cells. When cells are subjected to a sudden increase in medium os- molality, the first event that occurs is the rapid flux of water across the membrane and a decrease in cell volume (5,24). This shrinkage alters intracellular ionic and solute concentrations and may change several factors (5, 24). Uchida et al. (25) re- ported that potassium and sodium content increased in hyper- tonic medium in rabbit renal medullary cells. The cell shrink-

31300

A 350

Activation of MAP Kinase in MDCK Cells

B

300

150

100

* 11

* - * I 1 - I

Msm/kgH# 500 bv NaCl 300 I 500 bv raffinose

P M A - - - pretreatment

+ - + - +

L FIG. 6. Effects of PMA, calphostin C, €I-7, genistein, and herbimycin on hyperosmolality-stimulated MAP kinase activities in

MDCK cells. MDCK cells were incubated in hyperosmotic medium (500 mosmkg H,O with NaCl or raflinose) in the presence or absence of M PMAfor 24 h. MDCK cells were preincubated with calphostin C, H-7, genistein, or herbimycin in hyperosmotic medium (500 mosmikg H,O with NaCl or rafinose) ( n = 5, mean r S.E., *, p < 0.05).

age activates the Na-K-2C1 cotransporter and the Na-H and C1-HCO, exchangers (5, 24). In shark rectal gland, cell shrink- age changes the phosphorylation of Na-K-2C1 cotransporter (26). Hyperosmolality induced by NaCl may directly change intracellular sodium and chloride concentration. We must dis- tinguish whether the serinekhreonine kinase cascade re- sponses result from hyperosmotic change or change of a specific solute concentration. The addition of raffinose elicited almost identical activation of serinekhreonine kinase cascades, indi- cating that a change in osmolality rather than the concentra- tion of a specific solute is the stimulus that elicits this response.

Our data showed that hyperosmolality increased Ins-1,4,5-P3 levels in MDCK cells. This evidence suggests that cellular re- sponses to hyperosmotic changes are mediated at least par- tially via phospholipase C. Grinstein et al. reported that Na/H exchanger in lymphocytes is activated by cellular shrinkage in hypertonic medium and that it can also be activated by phorbol diester-induced activation of protein kinase C (21). They con- cluded that cellular shrinkage causes activation of a protein kinase that is similar, but not identical, to protein kinase C (21). We showed that PMApretreatment, H-7, and calphostin C inhibited hyperosmolality-induced MAP kinase activation. These results indicate that PMA-sensitive protein kinase C is upstream of Raf-1 kinase and MAP kinase in hyperosmolality- induced signal transduction in MDCK cells. In MDCK cells, the presence of the a, p, 6, E, and 5 PKC isozymes are reported (27). Thus, hyperosmolality may activate phospholipase C and PMA- sensitive protein kinase C or protein kinase C-like kinase, which then phosphorylate Raf-1 kinase. The Raf-1 kinase acti- vates MAPKK, the MAPKK activates MAPK, and the MAPK then phosphorylates S6 kinase. Our time course results sup- port this sequential activation. Our results using genistein and herbimycin indicate that the tyrosine kinase pathway is not significantly involved in hyperosmolality-induced MAPK acti- vation. Thus, our most favored hypothesis is that the alteration of ionic and solute concentration in the cell may activate phos- pholipase C or directly activate protein kinase C and that the sequential activation of serinekhreonine kinase cascades is in-

duced by hyperosmotic change in MDCK cells. Next, we shall address the possible physiological functions of

serinelthreonine kinase cascades in hyperosmotic changes in MDCK cells. Cell shrinkage activates some transport systems such as Na-K-2C1 cotransporter and the Na-H and C1-HCO, exchangers (5, 24) in MDCK cells. Thus, one possible physio- logical role of the activation of serinekhreonine cascades is to phosphorylate and regulate these transporters and exchangers.

Our present study demonstrates for the first time that hy- perosmolality activates S6 kinase in MDCK cells. S6 kinase may act as a mediator between the signal transduction path- ways and intranuclear events (28). Thus, it is a fascinating hypothesis that serinekhreonine kinase cascades mediate the signal transduction pathway from hyperosmotic stress to tran- scriptional induction of some genes, which are induced by hy- perosmolality.

Many growth factors and hormones induce MAP kinase ac- tivation (9, 17). Now we shall discuss the difference between mitogenic signals and hyperosmotic signals, and cellular re- sponses. Cohen et al. (29) reported that raising osmolality by adding 100 m~ NaCl increased Egr-1 and c-fos mRNA levels but that it suppressed [,H]thymidine incorporation by 59% in MDCK cells. This inhibition of [,H]thymidine incorporation mitigates a generalized nonspecific transcriptional up-regula- tion. In addition to inhibiting the rate of DNA synthesis, NaC1- induced hyperosmotic stress inhibited the rate of protein syn- thesis (Le. [,H]leucine incorporation) by 43% (29). Growth factors such as epidermal growth factor were reported to in- crease DNA and protein synthesis in MDCK cells (30). Thus, the phenomenon of activation of MAP kinase is similar for hyperosmotic stress and a variety of growth factor signaling, but the cellular responses are quite different if we consider DNA and protein synthesis. Wang et al. (31) reported that phor- bo1 ester activates MAP kinase but does not stimulate cell growth in rat mesangial cells. It is not clear what kind of signal-transducting mechanisms may differ for hyperosmotic stress and stimulation of growth factors in the stimulation of DNA and protein synthesis.

Activation of MAP Kinase in MDCK Cells 31301

Acknowledgments-We thank Dr. R. Tyler Miller (Division of Ne-

Texas), Dr. Jeff M. Sands (Emory University School of Medicine), and phrology, University of Texas Southwestern Medical School, Dallas,

Dr. Shinichi Uchida (Tokyo Medical and Dental University) for critical reading of this manuscript and useful suggestions.

REFERENCES 1. Beck, F., Dorge, A,, Rick, R., and Thurau, K. (1985) Pflueg. Arch. Eur. J.

2. Uchida, S., Green, N., Coon, H., Triche, T., Mims, S., and Burg, M. B. (1987)

3. Nakanishi, T., and Burg, M. B. (1989) Am. J. Physiol. 257, C7954801 4. Burg, M. B., and Garcia-Perez, A. (1992) J. Am. SOC. Nephrol. 3, 121-127 5. Parker, J. C. (1993) Am. J. Physiol. 265, C1191-Cl200 6. Garcia-Perez, A., and Burg, M. B. (1991) J. Membr. Biol. 119, 1-13 7. Brewster, J. L., de Valoir, T., Dwyer, N. D., Winter, E., and Gustin, M. C. (1993)

8. Tilly, B. C., van den Berghe, N., Tertoolen, L. G. J., Edixhoven, M. J., and

9. Roberts, R. M. (1992) Nature 380,534-535

Physiol. 405, Suppl. 1, S28-S32

Am. J. Physiol. 253, C23eC242

Science 259,1760-1763

de Jonge, H. R. (1993) J. Biol. Chem. 268, 19919-19922

10. Ahn, N. G., and Krebs, E. G. (1990) J. Biol. Chem. 265, 11495-11501 11. Posada, J., and Cooper, J. A. (1992) Science 255, 212-215 12. Lange-Carter, C. A,, Pleiman, C. M., Gardner, A. M., Blumer, K. J., and

13. Crews, C. M., and Eriksan, R. L. (1993) Cell 74, 215-217 14. Hughes, D. A., Ashworth, A,, and Marshall, C. J. (1993) Nature 364,349-352 15. Nguyen, T. T., Scimeca, J. C., Fillow, C., Peraldi, P., Carpentier, J. L., and Van

Johnson, G. L. (1993) Science 280, 315-319

Obberghen, E. (1993) J. Biol. Chem. 268, 9803-9810

16.

17.

18.

19.

20. 21.

22.

23. 24.

25.

26. 27.

28. 29.

30.

31.

Anderson, N. G., Maller, J. L., Tonks, N. K., and Sturgill, T. W. (1990) Nature 343,651-653

Thomas, G., Martin-Perez, J., Siegmann, M., and Otto, A. M. (1982) Cell 30, 235-242

Seger, R., Ahn, N. G., Posada, J., Munar, E. S., Jensen, A. M., Cooper, J. A,, Cobb, M. H., and Krebs, E. G. (1992) J. Biol. Chem. 267,14373-14381

Siegel, J. N., Klausner, R. D., Rapp, U. R., and Samelson, L. E. (1990) J. Biol. Chem. 265, 18472-18480

Fukami, K., and Takenawa, T. (1989) J. Biol. Chem. 264, 14985-14989 Grinstein, S., Goetz-Smith, J. D., Stewart, D., Beresford, B. J., and Mellors,A.

Kolch, W., Heidecker, G., Kochs, G., Hummel, R., Vahidi, H., Mischak, H., (1986) J. Biol. Chem. 261, 8009-8016

Vojtek, A. B., Hollenberg, S. M., and Cooper, J. A. (1993) Cell 74, 205-214 Finkenzeller, G., Marme, D., and Rapp, U. R. (1993) Nature 364, 249-252

Seldin, D. W., and Giebisch, G. (1992) The Kidney: Physiology and Pathophysi-

Uchida, S., Garcia-Perez, A,, Murphy, H., and Burg, M. B. (1989) Am. J.

Lytle, C., and Forbush B., I11 (1992) J. Biol. Chem. 267, 25438-25443 Cardone, M. H., Smith, B. L., Song, W., Mochly-Fiosen, D., and Mostov, K. E.

Ward, G. E., and Kirschner, M. W. (1990) Cell 61, 561-577 Cohen, D. M., Wasserman, J. C., and Gullans, S. R. (1991)Am. J. Physiol. 261,

Hatguel-DeMouzon, S., Csermely, P., Zoppini, G., and Kahn, C. R. (1992)

Wang, Y., Simonson, M. S., Pouysseger, J., and Dunn, M. J. (1992) Biochem. J.

ology, 2nd Ed., Ravan Press, New York

Physiol. 256, C614-C620

(1994) J. Cell Biol. 124, 717-727

C594-C601

J. Cell Physiol. 150, 180-187

287,589-594