Mitogen activated protein kinase and its activator are regulated by hypertonic stress in Madin Darby canine kidney cells

Mitogen-activated protein kinase and its

activator are regulated by hypertonic stress in

Madin-Darby canine kidney cells.

T Itoh, … , N Ueda, Y Fujiwara

J Clin Invest. 1994;93(6):2387-2392. https://doi.org/10.1172/JCI117245.

Madin-Darby canine kidney cells behave like the renal medulla and accumulate small organic solutes (osmolytes) in a hypertonic environment. The accumulation of osmolytes is primarily dependent on changes in gene expression of enzymes that synthesize osmolytes (sorbitol) or transporters that uptake them (myo-inositol, betaine, and taurine). The

mechanism by which hypertonicity increases the transcription of these genes, however, remains unclear. Recently, it has been reported that yeast mitogen-activated protein (MAP) kinase and its activator, MAP kinase-kinase, are involved in osmosensing signal

transduction and that mutants in these kinases fail to accumulate glycerol, a yeast osmolyte. No information is available in mammals regarding the role of MAP kinase in the cellular response to hypertonicity. We have examined whether MAP kinase and MAP kinase-kinase are regulated by extracellular osmolarity in Madin-Darby canine kidney cells. Both kinases were activated by hypertonic stress in a time- and osmolarity-dependent manner and

reached their maximal activity within 10 min. Additionally, it was suggested that MAP kinase was activated in a protein kinase C-dependent manner. These results indicate that MAP kinase and MAP kinase-kinase(s) are regulated by extracellular osmolarity.

Research Article

Find the latest version:

Mitogen-activated Protein Kinase

and Its Activator Are

Regulated by Hypertonic

Stress in Madin-Darby

Canine

Kidney Cells

Takahito Itoh,AtsushiYamauchi,AkikoMiyai, KenjiYokoyama,TakenobuKamada,Naohiko Ueda, andYoshihiro Fujiwara

First DepartmentofMedicine, Osaka UniversitySchoolofMedicine, Suita, Japan, 565

Abstract

Madin-Darby canine kidney cells behave like the renal medulla and accumulate small organic solutes (osmolytes) in a hyper-tonic environment. The accumulation of osmolytes is primarily dependent on changes in gene expression of enzymes that syn-thesize osmolytes (sorbitol) or transporters that uptake them (myo-inositol, betaine, and taurine). The mechanism by which hypertonicity increases the transcription of these genes, how-ever, remains unclear. Recently, it has been reported that yeast mitogen-activated protein (MAP) kinase and its activator,

MAPkinase-kinase, are involved in osmosensing signal trans-duction and that mutants in these kinases fail to accumulate glycerol, a yeast osmolyte. No information is available in mam-mals regarding the role of MAP kinase in the cellular response tohypertonicity. We have examined whether MAP kinase and MAP kinase-kinase are regulated by extracellular osmolarity in Madin-Darby canine kidney cells. Both kinases were acti-vated by hypertonic stress in a time- andosmolarity-dependent

manner and reached their maximal activity within 10 min. Ad-ditionally, it was suggested that MAP kinase was activated in a protein kinase C-dependent manner. These results indicate thatMAPkinase and MAP kinase-kinase(s) are regulated by extracellular osmolarity. (J. Clin. Invest. 1994. 93:2387-2392.) Key words: osmolyte * mitogen-activated protein kinase-kinase * hyperosmolarity-protein kinase C * transporter

Introduction

The renal medullaistheonly tissuein mammals thatnormally undergoeslarge changesinosmolarity.Thesechangesarepart

ofthe renalmechanism for producing aconcentrated or di-lutedurine. When the medulla ishypertonic, the cellsofthe medulla balance the high extracellular concentration of so-dium chloride byaccumulating high concentrations ofsmall

organicsolutes,termedosmolytes,throughcellularuptakeand

biosynthesis (forreviewsseereferences1and2).The

predomi-Address correspondencetoAtsushiYamauchi,TheFirst Department

of Medicine, OsakaUniversitySchoolof Medicine,2-2Yamadaoka, Suita,Japan, 565.

Receivedforpublication 4October 1993andinrevised form 24

February1994.

1.Abbreviationsused inthispaper: ERK, extracellularsignal-regulated kinase; GST,glutathioneS-transferase;MAP,mitogen-activated

pro-tein;MBP,myelinbasicprotein;MDCK,Madin-Darbycaninekidney;

MLCK, myosin light chainkinase;PKC, proteinkinase C; WT, wild-type.

nantrenalmedullary organic osmolytesaresorbitol,

myo-ino-sitol, betaine, glycerophosphorylcholine,and taurine, which do

not

perturb

cell

function,

incontrast to

high

concentrations of

electrolytes.Sorbitol isproducedfromglucose bytheenzyme

aldose reductase.Myo-inositol, betaine,and taurinearetaken

upinto cellsby Na-coupledtransporters( 1, 2).The cDNAs for aldose reductase and these osmolyte transporters have been

isolated, cloned, and characterized (3-6). It has been shown thatexpression of thesegenesincreases inresponse to

hyper-tonicstress(5, 7-11).

Hypertonicity alsoincreasesthelevels of mRNAencoding

theearlygenetranscription factors, Egr-l and c-fos in Madin-Darby canine kidney(MDCK)' cells(12). The expression of othergenesisalso affected.Na/K-ATPase expression increases in response to hypertonic stress in human renal cells ( 13).

HSP70 is induced by hypertonicstress in MDCK cells ( 12). Despite these studies, the signaling pathway ofhypertonic

stress from the extracellular environment tothe nucleus

re-mains unknown.

Mitogen-activated protein (MAP) kinase is known tobe activated quickly in response tovariousextracellularsignals, growthfactors, and vasoactive peptides (for reviewssee

refer-ences 14-17). MAP kinase is regulated by phosphorylation of itsownthreonine and tyrosine residues, which is catalyzedby

theupstreammolecule MAP kinase-kinase (for reviewsee

ref-erence 18). Recently,ithasbeenreportedthat theyeast coun-terpartsof MAPkinase and MAP kinase-kinaseareinvolved in osmosensing signal transduction and thatmutantsin these

ki-nases fail to grow normally ina hyperosmolar environment (19). However, noinformation is available regardingsuch a

role inhigher eukaryotes.

In this study, itwasdetermined thatMAPkinase and MAP kinase-kinase were regulated by extracellular osmolarity in MDCK cells. Several lines of evidence suggest that they

partici-pate in the cellularadaptation to osmotic stress in cultured renalepithelial cells.

Methods

Materials and chemicals. MDCK cells were a generous gift from the

Japanese CancerResearch Resources Bank. cDNAofratextracellular

signal-regulated kinase2 (ERK2) was cloned from a rat cDNA library

by PCRasdescribed previously (20).cDNAofthekinase-negative

ERK2(ERK2-K52R),whoselysine-52wassubstituted with arginine, was prepared by site-directed point mutagenesis using wild-type ERK2

(ERK2-WT)as atemplate,asdescribed previously (21 ). Each cDNA was constructedinto pGEX-2T (Pharmacia LKB Biotechnology Inc.,

Piscataway,NJ),and thenglutathione S-transferase (GST)-ERK2 fu-sionproteins,GST-ERK2-WT and GST-ERK2-K52R, were expressed

using Escherichia coliJM109. Each kinase fusionproteinwaspurified

asdescribed previously (20)and storedat-80°Cuntil use.Therat

cDNAlibrarywas agenerousgiftfrom Y. Takai(KobeUniversity,

Kobe, Japan). DME, raffinese,cytochalasin B, myelin basicprotein

(MBP), histone (type III-S), and protein A-Sepharose CL-4Bwere

purchasedfromSigmaImmunochemicals(St.Louis,MO).Aprotein

J. Clin. Invest.

© The American Society for Clinical Investigation, Inc.

0021-9738/94/06/2387/06 $2.00

kinaseinhibitor, staurosporin, was purchased from Kyowa Medex Co., Ltd., (Tokyo, Japan). An inhibitor of myosin light chain kinase (MLCK), ML-9, was purchased from BIOMOL Research Labs Inc. (Plymouth Meeting, PA). Anti-MAP kinase monoclonal antibody waspurchased from Zymed Laboratories, Inc. (South San Francisco, CA). Otherchemicalswere from commercial sources.

Cell culture and stimulation. MDCK cells were cultured in DME

supplementedwith 10%FCS,50 U/ml of penicillin, and 50Mg/mlof

streptomycin equilibrated with 5%C02/95%air at 370C.Cells were

passagedand grown to 80% confluent state and then cultured in

serum-freeDMEfor48 h.Afterrinsing twice with buffer A ( 10 mM Hepes [pH 7.4], 130 mM NaC1, 5.4 mM KCI, 1.2 mM CaC12, 1.2 mM

MgSO4, 0.1% dialyzed BSA),MDCK cells were incubated with buffer Afor 60min and then treated withbufferAthat was made hypertonic

by adding sodium chloride orraffinose forthe time indicated. For

depletion of protein kinase C (PKC),cellsin mediumweretreated with 130 nMofPMA for 16 h. For theexperimentsusingstaurosporin, ML-9, or cytochalasin B, cells were incubated in buffer A

supple-mented with these compoundsatthe indicated concentrations for 30

minandthenexposedtohypertonicstressintheirpresence. All

experi-ments wereperformed using 25-30passagesofMDCK cells.

Immunoblottingand immunoprecipitation. Immunoblotting and immunoprecipitation ofMAPkinasewereperformedasdescribed

pre-viouslywith aslight modification(22, 23). To detectsignalsof

immu-noblotting, biotin-labeledgoatanti-mouse IgGandavidin/biotin sys-tem were used(Vector Labs, Inc., Burlingame, CA).For

immunopre-cipitation,cellswerescraped offin1%SDS, 10mMTris/HCl, pH 7.5, andboiledat100°C for 5 min. After sonication, nucleiwereremoved by briefcentrifugationandanaliquot ofthe supernatant wasdilutedto afinalconcentration of 150mMNaCl,10 mMTris/HCl (pH 7.4),1%

Triton X-100, 0.1% SDS,0.5%NP-40,I mMEDTA,I mMEGTA,0.2

mMsodium orthovanadate,2MMp-amidinophenylmethanesulfonyl

fluoride. Thiswasused asthecelllysate. Immunoprecipitationwas

carriedoutbysequentially incubatingthecelllysatewithanti-MAP kinase antibodyat4°Cfor1h, with anti-mouse IgG (Zymed Laborato-ries, Inc.)at4°C forIh, and then withproteinA-Sepharose CL-4Bat

4°C for 30min.Aftercentrifugation,theprecipitatewaswashedand used astheimmunoprecipitate.

Preparation of cytosolfraction. Afterthe treatment of cellswith hypertonic bufferA,culture dishesweretransferredquicklyonice,and thebufferwasaspirated thoroughly.Cellswerescrapedoff with500 Ml/10-cm dish of ice-cold freshlypreparedhomogenizingbuffer(20

mMTris/HCl [pH 7.4],2 mMEGTA,10 mM,B-glycerophosphate, I

mM sodium orthovanadate, 1 mM DTT, 1 MM

p-amidinophenyl-methanesulfonylfluoride, aprotinin[ 100KIU/ml]). Homogenized by

briefsonication,thehomogenatewascentrifugedat125,000gat4°C

for30 min. The supernatantwascollectedasthecytosolfractionand storedat-80°C untiluse.

Assayofkinase activity. Activity ofMAP kinase wasmeasuredby

thein-gelkinase assay methodasdescribedpreviously (24, 25). Activ-ity ofMAPkinase-kinasewasmeasuredusingrecombinant MAP ki-nase asdescribedpreviouslywithaslightmodification (20,26, 27). Briefly,

20,ug

proteinofthecytosolfractionwasincubated with 500 ng ofGST-ERK2-WTin 100 Mlof reactionmixture(20mMTris/HCI [pH 7.4], 2mM EGTA, 10 mMMgCl2, 100MM ATP)at25°C for10 min.AfteraddingSDSsamplingbuffer forboilingandelectrophoresis,

activity of GST-ERK2 was measured by the in-gel kinase assay method. To study theeffect ofstaurosporinonMAPkinase-kinase,the

cytosol fraction was incubated with GST-ERK2-WT as described above,except that itwasperformedin the presence ofstaurosporinat

theconcentrationindicated. To detectphosphorylationofGST-ERK2

bythecytosolfraction,5Mgofproteinof thecytosolfractionwas incu-bated with 500 ngofGST-ERK2-WTorGST-ERK2-K52Rin 50 Mlof reactionmixtureusing 100 1M['y-32P]ATP(25MCi/ml)insteadof

nonlabeled ATP at25°Cfor 10 min. Thesamplewasappliedon SDS-PAGE, and the dried gel was exposed to XAR film (Eastman Kodak

Co., Rochester, NY). Radioactivity wasmeasuredby abioimaging analyzer(BAS2000; FujiPhoto FilmCo., Ltd., Tokyo, Japan).

Other procedures. Protein concentrations were measured as

de-scribedpreviously(28).

Results

MAP kinase activation by hypertonicstress. MBP

phosphory-lating activities markedly increased in the MDCK cell lysate, when cells were incubated with buffer A that was made hyper-tonic by adding sodium chloride to a totalosmolarity of1,000 mosmol/kg ofH20at 370C for 10min(Fig. 1, lanesIand2). Immunoblotting ofthe celllysatewith anti-MAPkinase mono-clonal antibody showed 49- and 42-kDbands (Fig. 1, lanes3 and 4). These kinases were exclusively immunoprecipitated from the celllysatewiththe same antibody(Fig. 1, lanes 5and

6). In addition, when histone was used as a substrate of the

in-gelkinase assay, these kinases did not phosphorylate it (data not shown). For these reasons, they were identified as MAP kinases of MDCK cells. The cytosol fraction from MDCK cells that were incubated with hypertonic buffer A also showed the

activation of the samekinases as the cell lysate (Fig. 1, lanes 7 and 8).To exclude the effects of detergents, we used the cytosol fraction as the sample for the following studies. The 49-kD

MAP kinase was activatedby hypertonicity in a

time-depen-dent manner. It reached its maximal activity 10minaftercells

were exposed to hypertonic stress (Fig.2). Half-maximal activ-ity was sustained at least up to 60 min. Similar results were obtained for the42-kD MAPkinase.

Osmolarity-dependent activation of MAP kinase. The 49-kD MAP kinase was activated by hypertonic stress ina osmo-larity-dependent manner (Fig. 3). This activity increased al-most linearly up to 1,000 mosmol/kg of H20 by adding so-dium chloride and decreased at 1,500 mosmol/kg ofH20. When using raffinose to makebuffer A hypertonic instead of sodium chloride, MAP kinase wasactivated more effectively and reached its maximum at 550 mosmol/kg of H20. Similar

A*(kD) - +

80

43.5-;

i.K

..}.. _

32.5-fpertonicStress

El

-+

-+

... . _ _

-MN,_

I.

IIE

1 2 3 4 5 6

7

8

Figure1. MAP kinase activationby

hypertonic

stress. MDCK cells

werestimulatedat

370C

for10 minwith isotonicor

hypertonic

buffer

A.

_{Hypertonicity}

wasinduced_by

adding

sodiumchloridetoatotal

osmolarityof_1,000

mosmol/kg

ofH20.For the

_in-gel

kinaseassay,

each_aliquotof thesamplewas

applied

onSDS-PAGE

using

10%

polyacrylamide gel

containing0.5_mg/mlMBP. Afterrenaturation ofthe_sample,itwasincubated with

['y-32P]ATP.

Phosphorylated

MBPwasvisualizedby

autoradiography.

Anti-MAP kinase

antibody

wasused forimmunoblottingand

immunoprecipitation.

Lanes I and

2,thein-gelkinaseassayof the cell

lysate;

lanes3 and4,

immuno-blottingof the cell_lysate;lanes 5 and6,the_in-gelkinaseassayofthe

immunoprecipitatefrom the celllysate;lanes7and8,thein-gel

ki-nase_assayof thecytosol fraction. The black and white arrowheads

representthe 49- and 42-kDMAPkinases,respectively.Molecular standardsareindicatedatthe left side. The resultsare

representative

Time

After

Stimulation

(min)

0

1

3

5

10

30

60

B

(pmol)

21.5

1.0

0

o0.5

C

0

0 20 40 1

Time After Stimulation

(min)

Figure2. Timecourseof MAPkinaseactivation inducedby hyper-tonicstress.MDCK cellswerestimulatedasdescribedinthelegend

toFig. 1.After the time indicated,thecytosolfractionwasprepared. Phosphate-32 incorporatedinto MBPwasmeasuredbyabioimaging analyzer. (A) Autoradiographyof thein-gelkinaseassay.The black andwhitearrowheads indicate 49- and 42-kD MAPkinases,

respec-tively. (B) Radioactivityofphosphate-32 incorporatedinto MBPby 49-kDMAPkinase. Theresultsarerepresentativeof three

indepen-dentexperiments.

after cellswere exposed to hypertonicstress. The activity of 49-kD MAP kinase decreased quickly and returned to near basal levelsat30minafter the shift(Fig. 5).MAPkinasewas activated when cellswerereexposedtohypertonic buffer Aat

30 min(Fig. 5). This result excluded thepossibilitythat the

changeofMAP kinaseactivitywasduetopermanent cell

dam-age. Similarresultswereobtained for the 42-kD MAP kinase

(datanotshown).

Effects

ofvarious inhibitors on MAP kinase activation. Some extracellularsignalsactivate MAP kinasethroughPKC

(forreviewseereference 15). To determine whether MAP ki-naseandMAPkinase-kinase(s) of MDCKcellswereactivated

by hypertonic stress through PKC, the experiment was

per-formed in the presenceor absence of PKC. When PKCwas

depleted byincubation of cells with 130 nM PMA for 16h,the activation of 49-kD MAP kinaseby hypertonicstresswas

abol-ished completely (Fig. 6 A). When cells were treated with

staurosporin,aninhibitorofPKC,for 30minand stimulated

withhypertonic buffer A containing staurosporin, the activa-tion of 49-kD MAP kinasewasinhibited inadose-dependent

manner (Fig. 6 B). The 50% inhibitory concentration was - 10-7 M. Additionally, staurosporin inhibited MAP kinase activation induced bythe cytosol fraction of cells exposedto

hypertonicstressaswell, and the 50%inhibitoryconcentration wasalso l0-- M(Fig. 6 B). An inhibitor of MLCK, ML-9,

showednoprevention of MAP kinase activation.Cytochalasin

B(20

_jiM)

alsohadnoeffectonMAPkinase activation (Fig. 6

B). Similarresultswereobtained for the 42-kD MAP kinase

(datanotshown).

Discussion

results were obtained for the 42-kD MAP kinase (data not

shown).

MAP kinase-kinase activation by hypertonic stress.

Be-cause of its molecular mass of 70 kD, a recombinant MAP

kinase, GST-ERK2, iseasy to distinguish from endogenous MAP kinases in polyacrylamide gel (20, 27). Sinceitshowsno basal activity, thechangeof itsactivityis detectedvery

sensi-tively (20, 27). For thesereasons,GST-ERK2isveryuseful for

detecting the activity ofMAP kinase-kinase (20, 27).

Incu-bated withthecytosol fractionfromcells exposedtohypertonic

stress,the activated GST-ERK2-WTwasdetected by the in-gel kinaseassay(Fig. 4 A). This indicated thatafactor(s) activat-ing GST-ERK2-WT was induced by hypertonic stress. This

factor(s)showed its maximal activitywhenthe cytosol fraction waspreparedfrom cellsexposedtohypertonicstressfor10min

(Fig. 4, A andC).Next,weexamined whether GST-ERK2was

phosphorylatedornotunder thesamecondition, because its activity was dependent upon its own phosphorylation

cata-lyzed by MAP kinase-kinase(s) (20). GST-ERK-WT was phosphorylated in parallel with its activity when itwas incu-bated with the cytosol fraction from cells exposedtohypertonic

stress for thetime indicated (Fig. 4 B). GST-ERK2-K52R,

whichwasdisruptedonits ATP-binding site and hadnokinase

activity (data not shown) (21), was also phosphorylatedby incubation with the cytosol fraction in thesameway(Fig. 4 B).

Therefore, thephosphorylationofGST-ERK2wasnotdueto

autophosphorylation. These results indicated that this fac-tor(s)was akinase, namelyMAPkinase-kinase(s).

MAPkinaseactivationwasreversible.Toconfirmwhether

MAPkinase activitywasregulated by extracellular osmolarity, hypertonic bufferAwaschangedtoanisotonic buffer 30 min

Renal medullary cells adapt to hyperosmolar environments under thephysiologicalconditions thatarepartof the

mecha-2

E

a

1.5

a.

so~~~~dIumchlord

0 0.5

a.

O 40V I I 6 1400

0 400 6~00 800

000" 12'00

1400

Osmolarity (mosm/kg

of

kO)

Figure 3. Osmolarity-dependency of 49-kDMAPkinaseactivation induced by hypertonicstress.MDCKcellswerestimulatedwithbuffer

A madehypertonic by adding sodium chlorideorraffinosetothe osmolarityindicated. After 10min,thecytosolfractionwasprepared, and MAPkinase activitywasassayedasdescribed inthelegendto

Fig.1.Theresultsarerepresentative ofthreeindependent experiments.

A

WT

_P

B

WT

_P

K52R

>

C

_-M

(pmol)

_1.51I

L

1.0

IL

s

0.5

0

0

Figure 4.Activation(

protein of thecytosol ERK2-WT or-K52R

Phosphorylated MBP assaymethod. (B)Ph

cytosol fractionwasi presence of

['y-32P]A

using 10%polyacryla

wasperformed.(C) F MBPbyGST-ERK2 GST-ERK2-WTand

tiveofthreeindepen(

nismofproducing

responses tohypert

responses aregene: initial response sui

TimeAfter Stimulation(min) (Fig. 1). The molecular masses were '-49 and 42 kD, respec-0 1 3 5 10 30 60 tively. These enzymes phosphorylated MBP, but not histone.

Therefore,weidentifiedthemasMAPkinases ofMDCK cells.

Another band with amolecularmassof -55 kD wasdetected \ \ \ \ \ in the cell lysate by the in-gel kinase assay (Fig. 1, lanes Iand

2), but since they phosphorylated histone other than MBP and

41

a

vwere notaffectedby

hypertonic

stress,theywerenotidentified inthisstudy. It has been reported that MAP kinase has several isozymes in various mammalian cells (14, 15), which are

re-W

ferred

to asthe ERK family. Both MAP kinases ofMDCK cells

mostlikelybelong to the ERK family. They showed maximal

activity 10minafter exposure to hypertonic stress (Fig. 2 ), and

thispeaktimeis comparable with the previous studies using growth factors (14, 15). MAP kinases of MDCK cells were

activatedin an osmolarity-dependent manner, which was lin-earupto 1,000mosmol/kgof H20when the osmolarity was

adjusted

bytheadditionof sodium chloride

(Fig.

3).The

activ-ity ofMAPkinasedecreased to 60% of the maximum at 1,500 mosmol/kg of H20, which seemed to be toxic for cultured 20 40 60 cells.

Raffinose,

aneutral_sugar,activated MAP kinasesmore

effectively, about twice as much at 500 mosmol/kg of H20 Time After Stimulation (min) (Fig. 3). Regarding the induction of an osmolyte transporter

ofMAPkinase-kinaseby hypertonic stress. 10 ,g geneby hypertonicity, similar differences induced by raffinose

Ifractionwasincubated with 500 ng of GST- and sodium chloride have been

reported ( 10).

The mechanism inthe reaction mixture at

250C

for 10min. (A) ofthis phenomenonremainstobeexamined.

'byGST-ERK2wasdetected by the in-gel kinase MAPkinase is known to be regulated by its own phosphor-iosphate-32 incorporatedintoGST-ERK2.The ylation that is catalyzed by MAP kinase-kinase. A recombinant incubated with GST-ERK2-WT or -K52R in the MAP kinase, GST-ERK2-WT, was phosphorylated and acti-TP.Each sample was applied on SDS-PAGE vated by the cytosol fraction from hypertonicity-treated cells

Lmide

gel without MBP, and autoradiography (Fig. 4, A-C). Autophosphorylation of GST-ERK2-WT was Zadioactivity of phosphate-32 incorporated into denied by the experiment using GST-ERK2-K52R with no

!-WT. Black and white arrowheads indicate

-K52R,respectively. The results arerepresent kinase activity (Fig. 4 B). Because we used GST-ERK2 asa

Ient

experiments, substrate for MAP kinase-kinase assay in excess ofendogenous

MAPkinases, which were quantified by Coomassie brilliant blue protein staining ofgel (data not shown), endogenous

aconcentrated urine(1,2).Avariety ofcell MAPkinaseswereunlikelytocompetitively inhibit the

activa-tonicstress has beenreported (1,2),and the tion and phosphorylation ofGST-ERK2-WT catalyzed by rallydivided intotwocategories.Oneisthe MAPkinase-kinase. Infact,endogenous MAP kinases in the ch asthe regulation ofcell volume, which cytosol fractionwerenotactivated furtherbyincubation with

doesnotneed geneexpression,and the otheristheinduction of

mRNAlevels and protein products of osmoregulatory genes suchasosmolytetransporters.Levels ofmRNAencoding early

genetranscription factors, Egr-1and

c-fos,

arealsoinduced by

hypertonic

stress( 12). Until recently, however, little informa-tionwasavailableregardingthe mechanismbywhichthe

signal

of extracellular

osmolarity

reached the nucleustoregulatethe

expression

ofthese genes.

MAPkinase and its

activator(s),

MAP

kinase-kinase(s),

arethoughttobekeyenzymes that transduce the extracellular

signals to the nucleus, since they are activated by various growth factors in vitro and inacellcycle-dependentmannerin vivo(forreviews seereferences 14, 15, 18). Veryrecently, it

hasbeenreported thatanMAP kinasehomologueofbudding

yeast, HOG 1, and itsactivator, PBS2, playanimportantrole in

the osmosensing pathway (19). Mutants lacking HOGI or

PBS2 showedabnormal cellgrowthonhyperosmolarmedium

( 19).Inthispaper,wereport the

regulation

ofMAPkinase and

MAPkinase-kinase by hypertonicstress in arenalepithelial

cellline,MDCK cells.

By immunoblottingand

immunoprecipitation

with anti-MAPkinase

antibody,

wedetectedtwobandsboth in the

lysate

and cytosol fractionfrom

hypertonicity-treated

MDCK cells

Figure 5. Effect of

nor-malizingthe

extracellu-2100

larosmolarityonMAP

*0

C i 0kinase-

activity.

Hyper-8

tonic bufferAwas

changedtoisotonic

75 \ bufferA30 min after

cellswereexposedto

/

\.'

hypertonic

stress

(so-diumchloride,afinal

|50- \ 1 / osmolarityof1,000

mosmol/kgof

H20)-/ \ .@30minafter

normaliz-C ₂₅-

I#@onlk

ing osmolarity,cells

underisotonic

condi-IL _{tionwerereexposedto}

a_____________________ hypertonicstress. At the

0 010₁₀ ~~~~~₂₀ ~~₃₀ . times indicated, the cy-_{tosolfractionwas}

pre-TimeAfterNormalizing(min) pared,and MAP kinase

activitywasassayedby

thein-gelkinase assay method. Theactivityat0 min afterswitchingbackwas_designatedas

(pmoQ

Figure 6. Effects of

A 2 PMA,staurosporin,and

a PKC(.) PKC(-) ML-9on49-kD MAP

8

_hkinase

activation

in-ducedby hypertonic

1 1

_ stress. (A) Effect of

PMA. MDCK cellswere

cultured in the presence

a.L O orabsenceof 130 nM

0 PMA, DME,0.1I% BSA

1 2 3 4 5 6 _{at370C for16hunder}

B

125

.

_{5% Co2.Afterwashing,} g100io they were stimulated for

10 min withbufferA

75 STAURO alone(lanes 1 and4),

SO

50W)

aand

supplementedwith

STAURO sodium chloride (a final

2at

T

osmolarity

of1,000

O c,11 , , . . . . mosmol/kg ofH20)

0

ld"1ld'cI

16' I67

106

1O6 (lanes 2and5)or 150 Concentratin (M) nMPMA (lanes 3 and 6).(B)Effectsofstaurosporin, ML-9,andcytochalasinB. MDCK cells were treated with the various concentrations ofstaurosporin,

ML-9, orcytochalasin Bfor 30 min and thenexposedtobufferA

madehypertonicbyadding sodiumchlorideto 1,000 mosmol/kgof

H20inthe presence of eachinhibitor.MAPkinaseactivitywas

mea-suredbythein-gel kinaseassaymethod,and inB,theactivityin the absenceof inhibitorswasdesignatedas100%.STAURO(MDCK),

the effectofstaurosporin onhypertonicity-induced49-kD MAP ki-naseactivationinMDCK cells;STAURO(Cell-free),theeffectof staurosporinonGST-ERK2 activation. Staurosporinwasaddedto

thecytosol fractionfromcellsexposedtohypertonicityin the absence

ofstaurosporin; ML-9,theeffect ofML-9onhypertonicity-induced

MAPkinase activation in MDCK cells;CyB, theeffectof

cytochala-sin Bonhypertonicity-inducedMAPkinaseactivationinMDCK cells. The resultsarerepresentative ofthreeindependent experiments.

GST-ERK2 (datanotshown), and the time course of GST-ERK2activationwas sodifferentfrom thatofMAPkinase that endogenous MAP kinase did not seem to contribute to the

phosphorylationand activation of GST-ERK2. These results

indicatethatMAPkinase-kinase(s) in MDCK cellswas acti-vated byhypertonicstress.

ComparedwithMAPkinase, MAPkinase-kinase seemed

tobe activatedmoreslowlyandtransiently(Fig. 4).Thismight

bepartly because of the higherbasal

activity

ofendogenous

MAPkinase. Nishida'sgroup(29)hasreportedthat MAP ki-nasephosphorylatesandactivatesMAP kinase-kinase and

pro-posed apositive feedback loop between thetwo. Iftrue,this

might explaintheacceleratedactivationofMAPkinase-kinase.

Furthermore, since active MAP kinase-kinase catalytically phosphorylatesandactivatesMAPkinase,theinitialresponse

ofMAP kinase-kinase may be small but enough to activate

MAPkinase.Alternatively, there isapossibilitythat

unidenti-fiedMAPkinaseactivator,whichmay bedifficulttodetectby

ourassay system, maybe involvedintheinitial activation of

MAPkinase. OnceMAPkinasewasactivated by hypertonic-ity, extracellular hypertonicitywasessentialto maintain its

ac-tivation, but not MAP kinase-kinase (Figs. 4 and 5). These

results suggest that MAP kinase isregulated by unidentified inactivators, possibly protein phosphatases, ratherthan MAP kinase-kinase inthelaterstages.

Toexaminetheinvolvement ofothersignalingmolecules,

pharmacological

studies were performed. MAP kinases of

MDCK cellsweremarkedlyactivatedby 150 nM PMAaswell

asbyhypertonicstress.Activationby 150nM PMAorby hy-pertonicitywereboth abolishedwhenPKCwasdepleted bythe

treatmentof cells with 130nMPMAfor 16h(Fig.6 A). This

suggests that hypertonic stress activated MAP kinases of

MDCK cells in a PKC-dependent manner. Staurosporin, which is sometimes usedas aninhibitorofPKC, didnot estab-lish that PKCwasinvolved inhypertonicity-inducedMAP

ki-nase activation because staurosporin inhibited arange of

ki-nasesincludingMAPkinase-kinase (Fig.6 B) (21, 26). After osmoticswellingorshrinkinginananisosmotic

envi-ronment,manycellsrecovertheiroriginalvolumebyadjusting

the concentration of intracellular

inorganic

ions and

osmo-lytes, which are referred to asregulatory volumeincrease or

decrease(1).Thechangeofcell volumeactivates

stretch-acti-vated ion channels (30), and it is postulated that thesensitivity ofthese channels is regulated by submembrane cytoskeletal

elements(31). It is reported thathypertonicstressinduces the

phosphorylation ofmyosin lightchain in cultured mesangial

cells (32). In light of these studies, we examined whether MLCKwasinvolved in thepathwayofhypertonicity-induced

MAPkinaseactivation.MDCK cellswerestimulatedwith hy-pertonicstressin thepresenceof ML-9,aninhibitor ofMLCK;

noinhibitionwasobserved inhypertonicity-inducedMAP

ki-naseactivation(Fig.6 B). We alsoexaminedtheeffect of

cyto-chalasin B, an inhibitor of microfilaments, on MAP kinase activation.When MDCK cellsweretreatedwith20

,M

of

cy-tochalasinB, it showed no effecton MAPkinaseactivation.

AlthoughcytochalasinBinhibits thefunctionof ion transport-ersinvariouscells (10, 33, 34), it is stillunknown whetherit inhibitsthe expression ofosmolytetransporters. Atleast, it is

unlikelythatmicrofilaments associate directlyto MAPkinase

activation.

Thereareseveral linesof evidenceshowing thatMAP

ki-nase regulates gene transcription and protein translation

di-rectlyorindirectly.First,MAPkinase activationis dependent

uponcellcycle( 14, 15 ). Second, MAPkinase is activatedby various growthfactors( 14, 15).Third,MAPkinase phosphor-ylates transcription factors, c-jun and c-myc in vitro (35). Fourth, MAPkinase translocates from cytoplasm to the

nu-cleus in response to growth factors (36). Fifth, MAPkinase

phosphorylates and activates S6 kinase II that is thoughtto

regulate thetranslational efficiency of40Sribosomal subunits ( 14). In light of these findings, a yeast study (19) and our

results, it ishypothesizedthathypertonicity-inducedMAP

ki-nasesofMDCK cellsmayregulate transcriptionand/or

trans-lation of osmolarity-associated genes such as osmolyte trans-porters and osmolyte synthetases. Alternatively, it has also

been reported that a variety ofvasoactive peptides activate

MAPkinasesin cultured cellssuch as vascular smooth muscle

cells and renalmesangial

cells(

16, 17). Since theyare

low-turn-overcellsinvivo, it is likelythat MAPkinasedoesnotregulate

transcriptionortranslation,but rather regulates cell physiologi-calfunctions suchasvasoconstriction andmesangial

contrac-tion. Anisosmotic stress inducesphosphorylation ofNa/K/ 2C1cotransporterandNa/K exchanger that are necessaryfor regulatoryvolumeincrease(37, 38). MAPkinasein MDCK cells may also regulate the functions of osmoregulatorygene

products directly or indirectly bytheir phosphorylation, but this has not yet been studied.

with theregulation ofosmoregulatorygenes ortheirproducts,

additional workswill berequiredtoconfirm a direct

relation-ship.

Acknowledgments

We thank Dr. Yoshimi Takai (KobeUniversity, Kobe, Japan) for

kindly providing the rat cDNAlibrary, and Dr. Joseph S. Handler (JohnsHopkins University, Baltimore, MD) for helpful advice.

Thisresearch was supported bygrants-in-aidforScientific Research

fromtheMinistryofEducation,Science,andCulture, Japan.

References

1.Chamberlin,M.E., and K.Strange. 1989. Anisosmoticcellvolume

regula-tion:acomparative view.Am. J.Physiol. 257:C159-C173.

2.Garcia-Perez,A., and M. B. Burg. 1991. Renal medullary organic osmo-lytes.Physiol.Rev.7:1081-1115.

3.Garcia-Perez,A., B.Martin,H.R. Murphy, S.Uchida,H. Murer, J.B. D.

Cowley,J.S. Handler, andM. B. Burg.1991.MolecularcloningofcDNAcoding

forkidney aldose reductase.J.Biol.Chem. 264:16815-16821.

4. Kwon, H. M., A.Yamauchi, S. Uchida, A. S. Preston, A.Garcia-Perez,

M.B.Burg,and J.S. Handler. 1992.Cloningofthe cDNAfora Na+/myo-inosi-tol cotransporter, ahypertonicitystressprotein.J.Chem. Biol.267:6297-6301. 5.Uchida, S., H.M.Kwon,A.Yamauchi,A.S.Preston, F. Marumo, and J. S.

Handler. 1992. Molecular cloning of the cDNA foranMDCK cellNa'-and

Cl--dependent taurinetransporter thatisregulated byhypertonicity.Proc. Nat!.

Acad. Sci. USA. 89:8230-8234.

6.Yamauchi, A., S. Uchida,H. M.Kwon,A.S. Preston,R. B.Robey,A.

Garcia-Perez,M. B.Burg,and J. S. Handler. 1992.CloningofaNa'-and Cl

-de-pendent betainetransporterthatis regulated byhypertonicity. J.BioL. Chem.

267:649-652.

7.Bondy, C. A.,S.L.Lightman, and S.L.Lightman. 1989.Developmental

andphysiological regulation of aldose reductasemRNAexpressionin renal me-dulla.Mot. Endocrinol. 3:1409-1416.

8.Cowley,B.D., J.D.Ferraris,D.Carper, andM.B.Burg. 1990.Invivo

osmoregulationof aldose reductase mRNA,protein,andsorbitol inrenal me-dulla. Am. J.Physiol. 258:F154-F161.

9.Smardo,F.L.,Jr.,M. B.Burg,and A.Garcia-Perez. 1990.Kidneyaldose

reductase gene transcription isosmotically regulated. J. Am. Soc. Nephrol.

1:744a.(Abstr.)

10.Yamauchi,A.,S.Uchida,A.S. Preston,H. M.Kwon,and J.S. Handler. 1993.Hypertonicity stimulatestranscriptionofgenefor Na+-myo-inositol co-transporterin MDCK cells.Am.J.Physiol.264:F20-F23.

11.Uchida, S.,A.Yamauchi,A.S. Preston,H. M.Kwon, and J. S. Handler. 1993. Mediumtonicityregulatesexpressionof the Na+-andCl--dependent be-taine transporter inMadin-Darbycaninekidneycellsbyincreasingtranscription ofthe transporter gene. J.Clin.Invest.91:1604-1607.

12.Cohen,D.M.,J.C. Wasserman, and S.R.Gullans. 1991. Immediate early

gene andHSP70expressioninhyperosmoticstressin MDCK cells. Am. J.

Phys-iol.261:C594-C601.

13. Yordy,M.R., andJ. W.Bowen.1993. Na,K-ATPaseexpressionand cell volumeduringhypertonicstressinhuman renal cells.KidneyInt.43:940-948.

14.Sturgill,T.W.,and J. Wu. 1991.Recent progress incharacterizationof

proteinkinasecascadesforphosphorylation of ribosomalproteinS6.Biochim.

Biophys.Acta.1092:350-357.

15. Cobb, M.H.,T.G. Boulton, and D.J. Robbins. 1991. Extracellular

signal-regulatedkinases:ERKsinprogress.CellRegul.2:965-978.

16.Simonson,M.S.,Y.Wang,J. M.Jones,and M. J. Dunn.1992.Protein

kinase C regulates activation ofmitogen-activatedproteinkinase andinduction of proto-oncogenec-fos byendothelin- 1. J. Cardiovasc.Pharmacol.20(Suppl.

12):S29-S32.

17.Tsuda,T., Y.Kawahara,Y.Ishida,M.Koide,K.Shii,and M.Yokoyama.

1992.AngiotensinIIstimulatestwomyelinbasicprotein/microtubule-associated

protein2 kinases incultured vascular smooth muscle cells. Circ. Res. 71:620-630.

18.Blenis,J. 1993.Signaltransduction via the MAPkinases:proceed at your ownRSK. Proc.NatL.Acad. Sci. USA. 90:5889-5892.

19.Brewster,J.L., T.Valoir,N. D. Dwyer,E.Winter,and M. C. Gustin. 1993. Anosmosensingsignal transduction pathway in yeast. Science(Wash. DC).259:1760-1763.

20.Itoh, T.,K.Kaibuchi,T.Masuda,T.Yamamoto,Y.Matsuura,A. Maeda, K.Shimizu,and Y. Takai. 1993. The post-translational processing of ras p21is

critical for its stimulation ofmitogen-activated protein kinase. J. Biol. Chem. 268:3025-3028.

21.Wu, J.,A. J.Rossomando,J.-H.Her,R. D.Vecchio,M. J. Weber, and T. W.Sturgill.1991.Autophosphorylationinvitroofrecombinant42-kilodalton

mitogen-activated protein kinase on tyrosine. Proc. Natl. Acad. Sci. USA. 88:9508-9512.

22.Oemar,B.S.,N. M.Law,andS.A.Rosenzweig. 1991.Insulin-like growth

factor-l inducestyrosyl phosphorylationofnuclear proteins. J. Biol. Chem. 266:24241-24244.

23.Pomerance,M.,F.Schweighoffer,B.Tocque, andM.Pierre.1992. Stimu-lation ofmitogen-activated proteinkinasebyoncogene rasp21 inXenopus oo-cytes. J. Biol. Chem.267:16155-16160.

24.Kameshita, I.,and H.Fujisawa.1989.Asensitive method fordetection of

calmodulin-dependentproteinkinase IIactivityinsodium dodecyl sulfate-polya-crylamide gel. Anal.Biochem. 183:139-143.

25.Gotoh,Y.,E.Nishida,S.Matsuda,N.Shiina,H.Kosako,K.Shiokawa, T.

Akiyama,K.Ohta,and H.Sakai. 1991. In vitroeffectsonmicrotubuledynamics ofpurifiedXenopusMphase-activatedMAP kinase.Nature(Lond.).

349:251-254.

26.Rossomando, A.,J.Wu, M. J.Weber,andT. W.Sturgill. 1992. The

phorbol ester-dependent activator ofthe mitogen-activated protein kinase

p42"I"isakinase withspecificityfor thethreonineandtyrosine regulatory sites.

Proc.Nat!. Acad.Sci. USA. 89:5221-5225.

27.Itoh, T.,K.Kaibuchi,T.Masuda,T.Yamamoto,Y.Matsuura,A.Maeda,

K.Shimizu,and Y. Takai. 1993. Aproteinfactorforrasp21-dependent activa-tion ofmitogen-activated protein (MAP)kinasethroughMAP kinase kinase. Proc.NatL Acad.Sci. USA. 90:975-979.

28.Bradford,M. M. 1976. Arapidandsensitivemethodforthe quantitation

ofmicrogram quantitiesofprotein utilizingtheprincipleofprotein-dyebinding. Anal.Biochem. 72:248-254.

29.Matsuda, S.,Y.Gotoh,and E.Nishida. 1993.Phosphorylation ofXenopus mitogen-activated protein (MAP)kinasekinasebyMAPkinase kinasekinase and MAP kinase. J.Biol.Chem.268:3277-328 1.

30.Lohr,J.W.,and J. J. Grantham. 1986. Isovolumetricregulationof iso-lated S2proximaltubules in anisotonic media. J.Clin.Invest. 78:1165-1172.

31.Guharay, F.,and F. Sachs. 1984.Stretch-activatedsingleion channel

currentsintissue-culturedembryonicchickskeletal muscle. J.Physiol. (Lond.).

352:685-701.

32.Takeda,M.,T. Homma, M. D. Breyer,N. Horiba, R. L.Hoover,S.

Kawamoto,I.Ichikawa,and V. Kon.1993. Volumeandagonist-induced

regula-tion ofmyosin light-chain phosphorylationinglomerular mesangialcells. Am. J.

Physiol.264:F421-F426.

33.Segal, S.,S. M.Hwang,J.Stem,and D.Pleasure. 1984. Inositoluptakeby

cultured isolatedratSchwann cells. Biochem.Biophys.Res.Commun. 120:486-492.

34.Biden,T.J.,and C. B.Wollheim. 1986. Activetransportofmyo-inositolin

ratpancreaticislets. Biochem. J. 236:889-893.

35.Leevers,S.J.,and C. J. Marshall. 1992. MAP kinaseregulation,the onco-gene connection. Trends Cell Biol. 2:283-286.

36.Chen, R.-H.,C.Sarnecki,and J. Blenis. 1992. Nuclear localization and

regulationof erk- and rsk-encodedproteinkinases.Mol.Cell. Biol.12:915-927. 37.Grinstein, S.,J. D.Goetz-Smith,D. Stewart,B. J. Beresford,and A. Mellors. 1986. Proteinphosphorylation duringactivation ofNa+/K+exchange

by phorbolestersandbyosmoticshrinking.J.Biol.Chem. 261:8009-8016. 38.Pewitt,E.B.,R. S.Hegde,M.Haas,and H. C.Palfrey.1990.The

regula-tion ofNa/K/2Clcotransportand bumetanidebindingin avianerythrocytesby