Proc. Natl. Acad. Sci. USAVol. 93, pp. 11242-11247, October 1996Physiology

Urea-inducible Egr-1 transcription in renal inner medullarycollecting duct (mIMCD3) cells is mediated byextracellular signal-regulated kinase activationDAVID M. COHENDivision of Nephrology, Oregon Health Sciences University and Portland Veterans Affairs Medical Center, PP262, 3314 SW U.S. Veterans Hospital Road,Portland, OR 97201

Communicated by Maurice B. Burg, National Heart, Lung, and Blood Institute, Bethesda, MD, July 16, 1996 (received for review January 19, 1996)

ABSTRACT Urea (200-400 milliosmolar) activates tran-scription, translation of, and trans-activation by the immedi-ate-early gene transcription factor Egr-l in a renal epithelialcell-specific fashion. The effect at the transcriptional level hasbeen attributed to multiple serum response elements and theiradjacent Ets motifs located within the Egr-1 promoter. Elk-i,a principal ternary complex factor and Ets domain-containingprotein, is a substrate of the extracellular signal-regulatedkinase (ERK) mitogen-activated protein kinases. In the renalmedullary mIMCD3 cell line, urea (200-400 milliosmolar)activated both ERK1 and ERK2 as determined by in-gelkinase assay and immune-complex kinase assay of epitope-tagged ERK1 and ERK2. Importantly, urea did not affectabundance of either ERK. Urea-inducible Egr-1 transcriptionwas a consequence of ERK activation because the ERK-specific inhibitor, PD98059, abrogated transcription from themurine Egr-1 promoter in a luciferase reporter gene assay. Inaddition, activators of protein kinase A, including forskolinand 8-Br-cAMP, which are known to inhibit ERK-mediatedevents, also inhibited urea-inducible Egr-1 transcription. Fur-thermore, urea-inducible activation of the physiological ERKsubstrate and transcription factor, Elk-1, was demonstratedthrough transient cotransfection of a chimeric Elk-1/GAL4expression plasmid and a GAL4-driven luciferase reporterplasmid. Taken together, these data indicate that, in mIMCD3cells, urea activates ERKs and the ERK substrate, Elk-1, andthat ERK inhibition abrogates urea-inducible Egr-1 tran-scription. These data are consistent with a model of urea-inducible renal medullary gene expression wherein sequentialactivation of ERKs and Elk-1 results in increased transcrip-tion of Egr-l through serum response element/Ets motifs.

Cells of the mammalian renal medulla are unique in theirperpetual exposure to an elevated and rapidly fluctuatingambient urea concentration. Because urea is a potent dena-turant of both protein (1) and nucleic acid (2), the ability ofthese cells to withstand hundreds of millimolar urea is ofconsiderable interest. Unlike the better-studied renal medul-lary solute, NaCl, urea is readily membrane-permeant and isunlikely to engender a marked decrement in cell volume (3).In addition, whereas hyperosmotic NaCl induces the accumu-lation of a full complement of osmotically active intracellularorganic solutes or organic osmolytes (including polyols, meth-ylamines, and amino acid analogs), urea treatment resulted inthe accumulation of only a single organic osmolyte-themethylamine, glycerophosphorylcholine (3). Therefore, hyper-osmotic NaCl and urea engender distinct cellular responses.

Several genes encoding proteins essential for the synthesis oruptake of organic osmolytes or osmolyte precursors, includingthe Na+/betaine (4, 5) and Na+/myoinositol (6) cotransport-ers, as well as the enzyme aldose reductase (7), are transcrip-

tionally up-regulated in response to hyperosmotic NaCl. Incontrast, only a single gene has been shown to be transcrip-tionally activated by urea. In the mIMCD3 cell line, derivedfrom microdissected terminal inner medullary collecting ductof mice transgenic for the large T antigen of simian virus 40 (8),expression of the immediate-early gene (IEG) transcriptionfactor, Egr-1, is up-regulated at the mRNA and protein levelsin response to physiologically relevant concentrations of urea(9, 10). This response appears to be unique to cells of renalepithelial origin (9), and is a consequence of enhanced tran-scription (10).

Signaling of hyperosmotic stress inducible by functionallyimpermeant solutes such as NaCl has received considerableattention in prokaryotes (11) and yeast (12-14). The yeastmitogen-activated protein kinase (MAPK), HOG1, is essentialfor osmotic tolerance in Saccharomyces cerevisiae (12). Inhigher eukaryotes, parallel kinase cascades activate membersof the three principal families of MAPKs including membersof the mitogen-responsive extracellular signal-regulated kinase(ERK) family (15), and the "stress-responsive" jun kinase/stress-activated protein kinase (JNK/SAPK; refs. 16 and 17)and p38 families (18-21). Recent evidence suggests that thehypertonic stressor, NaCl, activates ERK-like MAPKs in therenal epithelial MDCK cell line (22, 23); however, theseMAPKs are not required for transcriptional activation of genesencoding osmolyte transporters (24).

In the present study, we show that a renal-specific hyperos-motic (but nonhypertonic) stressor, urea, specifically activatesERK1 and ERK2, as determined by in-gel kinase assay and animmune complex kinase assay. In addition, urea-inducibletranscription of the Egr-1 gene is ERK-dependent because thespecific inhibitor of ERK activation, PD98059 (25, 26), abro-gated the effect, as did activators of cAMP-dependent proteinkinase (PKA). Consistent with a model of ERK-mediatedurea-inducible Egr-1 transcription, urea also activated Elk-1, aprincipal ERK-responsive Ets domain-containing protein andtranscriptional activator of IEGs.

METHODSCell Culture and Solute Treatment. mIMCD3 cells were

maintained in DMEM/F12 medium (Life Technologies,Grand Island, NY) supplemented with 10% FBS (JRH Bio-sciences, Richmond, CA) as described (10). Cells were growth-suppressed in DMEM/F12 without serum for 24 h beforetreatment with medium supplemented with urea to a finalconcentration of 200 or 400 milliosmolar as described (27).

Abbreviations: ERK, extracellular signal-regulated kinase; HA, influ-enza hemagglutinin; IEG, immediate-early gene; PK, protein kinase;MAPK, mitogen-activated PK; JNK/SAPK, jun kinase/stress-activated PK; mIMCD3, murine terminal inner medullary collectingduct cell line; NFDM, nonfat dry milk; SRE, serum response element;SRF, serum response factor.

11242

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement" inaccordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Apr

il 16

, 202

0

Proc. Natl. Acad. Sci. USA 93 (1996) 11243

In-Gel Kinase Assay. In-gel kinase assay was performed asdescribed (28, 29). Monolayers were washed with ice-coldHepes-buffered saline and scraped into 0.5 ml of ice-coldextraction buffer (20 mM Tris, pH 7.5/5 mM EGTA/0.5%Triton X-100/50 mM f3-glycerophosphate/1 mM polymethyl-sulfonyl fluoride/2% aprotinin/6 mM DTT/1 mM sodiumorthovanadate), and homogenized with 20-30 strokes in aDounce homogenizer. Lysate was centrifuged at 1000 x g for3 min and supernatant was recentrifuged at 400,000 x g for 15min. Lysate (150 [kg per lane) was subjected to electrophoresisthrough a 10% polyacrylamide gel containing 0.5 mg/mlmyelin basic protein (Sigma), overlaid with a 3% stacking gel.The entire gel was washed twice with 100 ml of 20% propanolin 50 mM Tris (pH 8.0) at 25°C for 1 h, once with 250 ml ofbuffer A (50 mM Tris, pH 8.0/5 mM 2-mercaptoethanol) at25°C for 1 h, twice with 100 ml of 6 M guanidine HCl at 25°Cfor 1 h, and five times with 250 ml of buffer A + Tween 20(0.04%) at 4°C for a total of 16 h. The gel was then incubatedwith 10 ml of kinase buffer [40 mM Hepes, pH 8.0/2 mMDTT/0.1 mM EGTA/5 mM MgCl2/25 ,utM ATP/25 ,uCi[,y-32P]ATP (1 Ci = 37 GBq)] at 22°C for 1 h, washed five timeswith 500 ml of Wash buffer (5% trichloracetic acid/1% sodiumpyrophosphate) at 25°C, dried on Whatman paper, and sub-jected to autoradiography.

Western and Northern Blot Analyses. For Western blotanalysis, lysates were prepared, subjected to electrophoresis,and transferred to polyvinylidene difluoride as previouslydescribed (10). All incubations were performed at 25°C withgentle rocking. Membranes were blocked with 5% nonfat drymilk (NFDM) in 1 x PBS/0.1% Tween 20 for 1.5 h and washedthree times with SIB (lx PBS/0.1% Tween 20/1% NFDM)and then either incubated with 1:5000 dilution of anti-MAPK(Zymed) or 1:500 dilution of anti-Egr-1 (Santa Cruz Biotech-nology, Santa Cruz, CA) in lx PBS/1% NFDM for 1 h, andwashed three times with SIB, or incubated with 1:4000 dilutionof the horseradish peroxidase-coupled appropriate secondaryantibody (Pierce) in SIB, and washed three times with 1xPBS/0.3% Tween 20/1% NFDM. Detection was via enhancedchemiluminescence (Amersham) according to the manufac-turer's directions. Total RNA was prepared according to themethod of Gough (30) and subjected to electrophoresis andNorthern blotting as described (27). Full-length murine Egr-1cDNA was kindly provided by V. P. Sukhatme (Beth IsraelHospital, Boston, MA).

Epitope-Tagged ERKs, Stable Transfection, and in VitroKinase Assay. Hemagglutinin (HA)-epitope-tagged ERK1was kindly provided by M. J. Weber (Univ. of Virginia,Charlottesville) (31). Epitope-tagged ERK2 was prepared asfollows: cDNA encoding ERK2 [kindly provided by D. Charest(Kinetek Biotechnology, British Columbia) and S. L. Pelech(Kinetek Biotechnology, British Columbia) (32)] was ampli-fied by PCR to contain NotI/XhoI restriction site ends thatpermitted cloning into NotI/XhoI-digested HA3pcDNA3(provided by J. Epstein, Brigham and Women's Hospital,Boston). HA3pcDNA3 contains three tandem repeats of theHA tag (YPYDVPDYA) subcloned into the XhoI/ApaI-digested pcDNA3 (Invitrogen). Stable transfectants ofmIMCD3 cells were prepared through electroporation (seebelow) and selected in Geneticin (Life Technologies). ERKactivation was quantitated through a modification of previ-ously published methods (31, 33). Monolayers were washedtwice with ice-cold PBS, lysed in lysis buffer (1% NonidetP-40/150 mM NaCl/10 mM Tris, pH 8.0/1 mM polymethyl-sulfonyl fluoride/2mM sodium orthovanadate/10 mM sodiumpyrophosphate/0.4 mM EDTA/10 mM NaF/2 ,ug/ml aproti-nin/2 gg/ml leupeptin/10 mM N-nitrophenyl phosphate), andcleared at 14,000 rpm in a microfuge at 4°C for 3 min. Afterthe addition of S ,lI of 12CA5 antiserum (Babco, Berkeley,CA) and 40 ,ul of 50% solution of washed protein G-Sepharosebeads (Pharmacia), the lysate was incubated with inversion at

4°C for 2-4 h. Immunoprecipitates were washed twice with 0.8ml of lysis buffer, twice with 0.8 ml of TBS, and once withkinase buffer (25 mM Hepes, pH 7.5/10 mM MgCl2/2 mMMnCl2/1 mM DTT). Beads were resuspended in 40 ,tl ofcomplete kinase buffer {kinase buffer + 50 ,tM [y32P]ATPand 0.2 ,ug/,ul myelin basic protein (Sigma)}, and incubated at30°C for 20 min. One volume of 2x Laemmli sample buffer wasadded and the reaction was boiled for 3 min before resolutionby SDS/10% polyacrylamide gel. Gels were dried and exposedto film; band intensity was quantitated through scanningdensitometry.

Transient Transfection and Reporter Gene Analysis. Fortransfection via electroporation, mIMCD3 cells were grown to80-90% confluence, trypsinized, resuspended in warmedcomplete medium, pelleted at 1000 g x 5 min, washed withice-cold DMEM/F12, repelleted, and then resuspended inice-cold DMEM/F12 at a working concentration of -5 x 106cells per ml. Cell suspension (0.5 ml) was added to 20 ,ug ofluciferase reporter plasmid and 5 ,ug of the cytomegalovirus-Gal vector (for normalization) in ice-cold electroporationcuvettes (Invitrogen), incubated on ice for 10 min, electropo-rated at 960 ,uF and 260 V (GenePulser, Bio-Rad), andincubated on ice for a final 10 min. Cells were diluted 1:20 withwarmed complete medium and plated. After 24 h, cells wereplaced in serum-free medium; at 48 h, cells were treated for 6 hwith the desired condition before harvest. ,3-Galactosidaseactivity was determined using standard methods (34). Tomeasure luciferase activity, individual wells of six-well plateswere washed with ice-cold PBS and lysed with 150 ,ul ofLuciferase lysis buffer (125 mM Tris, pH 7.6/0.5% TritonX-100). Lysate (100 ,ul) was incubated with 200 ,tl of 5 mMATP in luciferase buffer (25 mM glycylglycine/15 mM MgSO4,pH 7.8) and 100 gl of luciferin (60 ,tg/ml; Analytical Lumi-nescence Laboratory, San Diego) in Luciferase buffer in anautomated luminometer (Berthold, Nashua, NH), counted for30 s, and normalized to ,B-galactosidase activity. Data areexpressed as mean + SEM, except where noted. Egr-1-Luc wasa Sall fragment (=1.2 kb) of the murine Egr-1 promoter(provided by V. P. Sukhatme; ref. 35), subcloned into the Sallsite of pXP2 (36). Elk-1/GAL4 chimeric expression plasmid(37) and 5 x GAL4-Luc reporter plasmid (38) were kindlyprovided by R. Maurer (Oregon Health Sciences University,Portland, OR).

RESULTSWhen mIMCD3 cells in culture were treated with 200 or 400mM urea for 10 min, up-regulation of the myelin basicprotein-phosphorylating activity (i.e., MAPK activity) of pro-teins migrating at -44 and 42 kDa was noted, consistent withthe Mr of ERKs 1 and 2, respectively (Fig. 1A). By scanningdensitometry, these up-regulations were -twofold for ERK1and threefold for ERK2. In addition, up-regulation of severalas-of-yet unidentified bands was also noted. Hypertonic NaCl(200 milliosmolar) also increased ERK activity to an approx-imately comparable degree (data not shown).To ascertain whether these up-regulations in ERK activity

were a consequence of increased activity, per se, or increasedERK abundance, Western blot analysis was performed.Whole-cell lysates from control and urea-treated (200 or 400mM x 10 min) mIMCD3 cells were subjected to SDS/PAGE,transferred to nylon, and probed with a polyclonal antiserumthat recognizes both ERK 1 and ERK2 (Fig. 1B). Bands wereevident at -44 and 42 kDa, consistent with ERK1 and ERK2,respectively; neither band was up-regulated by this brief ureatreatment.

Because the in-gel kinase assay only demonstrates up-regulation of an unspecified MAPK of a given Mr, and becausea non-ERKMAPK may comigrate with ERK1 or ERK2, it wasimportant to definitively demonstrate ERK activation by urea.

Physiology: Cohen

Dow

nloa

ded

by g

uest

on

Apr

il 16

, 202

0

Proc. Natl. Acad. Sci. USA 93 (1996)

A

I43D

31D

C U U

200 400

B

71D

43D

31D

C U U

200 400

FIG. 1. (A) Urea-inducible MAPK activity as quantitated by in-gelkinase assay. Lysates prepared from mIMCD3 cells, treated for 10 minwith control medium or medium supplemented with 200 or 400 mMurea, were subjected to SDS/PAGE through a gel impregnated withmyelin basic protein. (B) Effect of urea upon ERK abundance.Western blot analysis of lysates prepared from mIMCD3 cells treatedas inA and probed with anti-ERK antibody (Zymed). (A and B) Openarrows indicate migration ofMr markers; solid arrows at 44 and 42 kDaindicate presence of ERK1 and ERK2, respectively.

For this reason, a complementary strategy using epitope-tagging and an immune-complex kinase assay was employed.Stable transfectants of mIMCD3 were selected that harboredcDNAs encoding either ERK1 or ERK2 linked to an HA-epitope tag. HA-ERK1-transfected or HA-ERK2-transfectedcells were control- or urea-treated for 10 min. Lysates were

prepared and epitope-tagged ERK was immunoprecipitatedwith anti-HA antibody. ERK activity of immunoprecipitateswas evaluated in an in vitro immune complex kinase assay.ERK1 activity was increased twofold (Fig. 2A) and ERK2activity was increased threefold (Fig. 2B), consistent with thein-gel kinase data. Therefore, urea activated both ERK1 andERK2.Because ERKs phosphorylate and activate Elk-1 (39-41),

the heterodimerization partner of serum response factor(SRF), and because urea-inducible expression of the IEGEgr-1 is mediated by SRF:Elk-1 interaction with serum re-

sponse element (SRE)/Ets motifs (42), the specific role ofERK activation in urea-inducible Egr-1 expression was evalu-ated. The model of Egr-1 expression consisted of transienttransfection of mIMCD3 cells with a luciferase reporter gene

c u c u

FIG. 2. Effect of urea upon ERK activation. mIMCD3 cells, stablytransfected with HA-epitope-tagged ERK1 (A) or ERK2 (B), weretreated with control medium or medium supplemented with 200 mMurea for 10 min before harvest. Lysates were immunoprecipitated withanti-HA antibody and subjected to an in vitro immune complex kinaseassay as described.

driven by 1.2 kb of the murine Egr-1 promoter (35). Althoughcrude pharmacological inhibitors ofERK activation have beendescribed previously, the recent identification of a specificMEK inhibitor (and therefore, inhibitor of ERK activation),PD98059 (25, 26), enabled the assessment of the contributionof ERK activation to urea-inducible Egr-1 expression. Cellswere pretreated for 30 min with PD98059 (50 ,uM or 100 AM)or the equivalent amount of dimethyl sulfoxide vehicle beforethe 6-h treatment with control medium or medium supple-mented with 200 mM urea (Fig. 3). Urea increased reportergene activity -"15-fold. Vehicle alone slightly increased bothcontrol and urea-inducible luciferase activity (i.e., Egr-1 tran-scription). PD98059 potently suppressed urea-inducible Egr-1transcription in a dose-dependent fashion, without signifi-cantly affecting basal activity. In addition, shown in the insetto Fig. 3, pretreatment with PD98059 abrogated the urea-inducible increase in Egr-1 mRNA abundance as determinedby Northern analysis. Not shown, a similar effect upon urea-inducible Egr-I mRNA abundance was obtained followingpretreatment with forskolin. Also not shown, NaCl (200milliosmolar) failed to substantially up-regulate Egr-l mRNAabundance; hence, the effect of PD98059 upon NaCl-inducibleEgr-1 expression could not be evaluated. These latter data areconsistent with the marginal increase in Egr-1 transcriptionobserved in response to hypertonic NaCl in the mIMCD3 cellline (42).Although PD98059 is purported to exhibit specificity for the

ERKs, it was necessary to corroborate these findings. Activa-tors of PKA inhibit ERK activation, in a putative Raf-dependent fashion (43, 44). We therefore examined the abilityof the PKA activators, forskolin and 8-Br-cAMP, to inhibiturea-inducible Egr-1 transcription, as measured by the lucif-erase reporter gene assay. Both forskolin (100 AM) and8-Br-cAMP (3 mM) abrogated urea-inducible Egr-1 transcrip-tion (Fig. 4), eorroborating the data obtained with PD98059.Furthermore, both agonists inhibited Egr-1 protein expressionin response to urea as determined by Western blot analysis;data for forskolin are depicted in the inset to Fig. 4.Because urea activates Egr-1 transcription in an SRE/Ets-

dependent fashion, and because activation of the principalMAPK-responsive Ets-domain-containing protein, Elk-1, canmediate trans-activation through this composite element, itwas hypothesized that urea would activate Elk-1. To demon-strate activation of this physiological substrate of ERK inresponse to urea treatment, transient transfection of mIMCD3

... soz;:pilmr

11244 Physiology: Cohen

A

Dow

nloa

ded

by g

uest

on

Apr

il 16

, 202

0

Proc. Natl. Acad. Sci. USA 93 (1996) 11245

6000 -

'a 4000 -

73-2

2000 -

Control

1 2 3 4

.T.

+ UreaFIG. 3. Effect of the ERK inhibitor, PD98059, upon urea-inducible Egr-I transcription and mRNA abundance. mIMCD3 cells were transiently

transfected with a luciferase reporter vector driven by 1.2 kb of the murine Egr-1 promoter (35). Urea-inducible reporter gene activity was abrogatedby 30-min pretreatment with PD98059 (50 ,uM and 100 ,uM), but not by vehicle (dimethyl sulfoxide) alone. (Inset) PD98059 abrogates urea-inducibleEgr-1 mRNA expression. Northern blot analysis of Egr-1 mRNA abundance following 30 min of control treatment (lanes 1 and 2) or 30 min ofurea (200 mM) treatment (lanes 3 and 4) in the absence (lanes 1 and 3) or presence (lanes 2 and 4) of 30 min of pretreatment with PD98059 (50,uM). Data depicted are mean ± SEM of at least three determinations and are representative of three separate experiments.

cells with a modified Elk-1 expression system was undertaken.An expression plasmid encoding the activation domain of thetranscription factor (and ERK substrate), Elk-1, linked to theDNA-binding domain of the yeast transcription factor GAL4(kindly provided by R. Maurer; ref. 37), was transientlytransfected into mIMCD3 cells, in conjunction with a lucif-erase reporter vector driven by 5 repeats of the yeast GAL4DNA binding site (5 x GAL4-Luc, (38)). This chimericElk-1/GAL4 assay permits a direct assessment of Elk-1 acti-vation (in the absence of an adjacent SRE or associated SRF)in response to a physiological stimulus. As seen in Fig. 5, whentransfected cells were treated with urea, Elk-1 activity wasincreased threefold; however, in cells transfected with the 5 xGAL4 reporter and only empty expression vector, there was noresponse to urea. [This urea effect was sensitive to a 30-minpretreatment with PD98059 (50 ,M) (data not shown).]Therefore, urea activated the transcription factor and ERKsubstrate, Elk-1.

6000

_4000

300-j

2000

Control + Urea

FIG. 4. Effect of PKA activation upon urea-inducible Egr-1transcription. mIMCD3 cells, transiently transfected as in Fig. 3, werepretreated for 30 min with forskolin or 8-Br-cAMP, before receivingcontrol medium or medium supplemented with 200 mM urea for 6 h.Data are normalized to expression of f3-galactosidase activity from acotransfected cytomegalovirus-Gal expression vector. Data depictedare mean ± SEM of three determinations. (Inset) Western blotanalysis ofEgr-I protein abundance in response to control (lanes 1 and3) or urea treatment (200 mM; lanes 2 and 4) in the presence (lanes3 and 4) or absence (lanes 1 and 2) of forskolin pretreatment.

DISCUSSIONIn cultured cells derived from the murine inner medulla,elevated but physiologically relevant concentrations of themembrane-permeant solute, urea, activated ERK1 and ERK2,as well as the principal ERK-responsive Ets domain-containing protein and IEG inducer, Elk-1. Previous workexamining kinase signaling of hyperosmotic stress has focusedupon hypertonicity induced by functionally impermeant sol-utes. Terada (23) and Itoh (22) showed that hypertonic NaCland raffinose activated ERK-like MAPKs in the renal epithe-lial MDCK cell line. In their studies, MAPK activation was notlinked to the acquisition of a solute-inducible phenotype orpattern of gene expression, and was ascribed to the hypertonicnature of the stressor. Interestingly, although activated byhypertonicity, ERKs appear not to play a role in the transcrip-tional up-regulation of genes encoding transporters essentialfor the adaptive response to hypertonicity (24). Because it isreadily membrane-permeant, urea is not considered a hyper-tonic stressor (3). Further support for the presence of dissim-ilar signaling pathways for NaCl and urea is reflected in their

250

EL control200 ;

Il+ Urea

z 150-

0~~~- 100-

50-

0GAL4-Elk-1 pcDNA3

Vector

FIG. 5. Urea-inducible Elk-1 activation. mIMCD3 cells were tran-siently cotransfected with an expression vector encoding the activationdomain of Elk-1 fused to the DNA-binding domain of GAL4 (Elk-1/GAL4), or empty vector alone (pCDNA3), as well as a GAL4-driven luciferase reporter vector. Cells were treated for 6 h withcontrol medium or medium supplemented with 200 mM urea. Datadepicted are means ± SEM of at least three determinations in a singleexperiment, and are representative of three separate experiments.

E no add

E] vehicle (lo)

* vehicle (hi)

* PD98059 (50 uM)

* PD98059 (100 uM)

_. _E.l no add

E: forskolin1 2 3 4 U 8-Br-cAMP

--T l

0 1.. mm.. I

Physiology: Cohen

Dow

nloa

ded

by g

uest

on

Apr

il 16

, 202

0

Proc. Natl. Acad. Sci. USA 93 (1996)

differential sensitivity to the tyrosine kinase inhibitor,genistein. Itoh (22) reported that genistein failed to inhibithypertonicity-inducible signaling, whereas the urea signalingresponse to IEG transcription is quite genistein-sensitive (45).In addition, the present data link urea-inducible ERK activa-tion to a downstream trans-activating event (Elk-1 activation)and transcriptional event (Egr-1 transcription).

In addition to ERKs, other MAPKs are potentially respon-sive to hypertonic and hyperosmotic stressors in cells of highereukaryotes in vivo. p38, the mammalian homolog of the yeastosmoresponsive MAPK, HOG1, is activated by hypertonicityin mammalian cells (19, 46). Hypertonicity also activatedJNK/SAPK in Chinese hamster ovary cells, and osmotic shockof yeast cells resulted in both phosphorylation and activationof transfected mammalian JNK/SAPK (47). In addition, bothJNK (47) and p38 (19) can complement a HOG1-deficientyeast strain and restore osmotic tolerance.

Urea-inducible transcription of the Egr-1 gene is likelymediated by SRE/Ets motifs (42). At least two Ets motif-binding (Ets domain-containing) proteins have been identifiedand cloned; whereas Elk-1 (48, 49) has been implicated intransducing MAPK-mediated signals, SAP-1 (50) appears tofunction through a distinct pathway (51). For this reason,attention was focused upon urea-inducible Elk-1 activation.Because Elk-1 functions in concert with SRF in activatingtranscription from promoters bearing adjacent SREs and Etsmotifs (52, 53), and because SRF may be independentlyactivated through another ERK-independent pathway (54), areporter gene assay with a chimeric activator (37, 38) was usedto examine Elk-1 activation in isolation. The concept ofcompartmentalization of transcription factor structure intofunctional "modules" is well-described. Fusion of the Elk-1activation domain to the DNA-binding domain of the yeasttranscriptional activator, GAL4, permits quantitation ofGAL4-driven reporter gene activity as an index of Elk-1activation. Urea activated Elk-1 by this assay, consistent withthe proposed model of ERK-mediated Egr-1 transcription byurea.

In addition to being phosphorylated by ERKs, the transcrip-tion factor Elk-1 can also be phosphorylated and activated byJNK/SAPKs (55, 56). Although urea activated ERKs, itremained possible that urea-inducible Elk-1 activation was aconsequence of JNK/SAPK activation and not of ERK acti-vation. Urea-inducible Egr-l transcription is potently inhibit-able by the MEK inhibitor (and therefore, ERK inhibitor)PD98059. PD98059 appears to be quite specific for ERK-mediated events, failing to inhibit activation of JNK/SAPK ina cell culture model (57). Therefore, it would appear that Elk-1activation is ERK-mediated in response to urea.The present data, in conjunction with previous observations,

begin to delineate a signaling pathway activated by urea in cellsof the renal medulla. Urea, through a putative tyrosine kinasephosphorylation event (e.g., activation of a receptor tyrosinekinase) activates the tyrosine kinase-specific PLC isoform,PLC-y (45). Activation of PLC--y results in cleavage of PIP2and liberation of IP3. IP3-inducible Ca2+ release, in conjunc-tion with diacylglycerol, activates PKC (45), which ultimatelyresults in transcription of the IEG, Egr-1. The mechanismthrough that elevated ambient urea concentration inducesERK activation remains obscure. It is likely that ERK activa-tion in this, as in diverse other, contexts is a consequence ofRas-induced activation of the MAPKKK, Raf (reviewed inrefs. 15 and 58). Raf, in turn, activates the ERK activator,MEK. Alternatively, activated PKC may directly activate Raf(59). MEK is likely to be involved in the present model becausethe MEK inhibitor, PD98059, inhibited urea-inducible Egr-Itranscription. Urea-inducible ERK activation then results inactivation of the ternary complex factor, Elk-1, which activatestranscription of Egr-I through SRE/Ets motifs. This putativesequence of events serves to underscore the active role taken

by physiological concentrations of urea in signaling to nuclearevents in cells of the renal medulla.

We thank M. J. Weber for HA-ERK2, D. L. Charest and S. L. Pelechfor human p44erkl cDNA, J. Epstein for HA3pCDNA3, V. P.Sukhatme for murine Egr-1 promoter and Egr-1 cDNA, A. Saltiel andParke-Davis Pharmaceutical for PD98059, R. Maurer for Elk-1/GAL4-pCDNA3 and 5 x GAL4-Luc plasmids, and Bruce Magun forhelpful suggestions. This work was supported by National Institute ofDiabetes and Digestive and Kidney Diseases Grant DK02188.

1. Yancey, P. H., Clark, M. E., Hand, S. C., Bowlus, R. D. &Somero, G. N. (1982) Science 217, 1214-1222.

2. Cox, R. A. & Kanagalingam, K. (1968) Biochem. J. 108,599-610.3. Garcia-Perez, A. & Burg, M. B. (1991) Physiol. Rev. 71, 1081-1115.4. Uchida, S., Yamauchi, A., Preston, A. S., Kwon, H. M. & Han-

dler, J. S. (1993) J. Clin. Invest 91, 1604-1607.5. Takenaka, M., Preston, A. S., Kwon, H. M. & Handler, J. S.

(1994) J. Biol. Chem. 269, 29379-29381.6. Yamauchi, A., Uchida, S., Preston, A. S., Kwon, H. M. & Han-

dler, J. S. (1993) Am. J. Physiol. 264, F20-F23.7. Smardo, F. L., Burg, M. B. & Garcia-Perez, A. (1992) Am. J.

Physiol. 262, C776-C782.8. Rauchman, M. I., Nigam, S., Delpire, E. & Gullans, S. R. (1993)

Am. J. Physiol. 265, F416-F424.9. Cohen, D. M. & Gullans, S. R. (1993) Am. J. Physiol. 264,

F593-F600.10. Cohen, D. M., Chin, W. W. & Gullans, S. R. (1994)J. Biol. Chem.

269, 25865-25870.11. Csonka, L. N. & Hanson, A. D. (1991) Annu. Rev. Microbiol. 45,

569-606.12. Brewster, J. L., de Valoir, T., Dwyer, N. C., Winter, E. & Gustin,

M. C. (1993) Science 259, 1760-1763.13. Maeda, T., Wurgler-Murphy, S. M. & Saito, H. (1994) Nature

(London) 369, 242-245.14. Maeda, T., Takekawa, M. & Saito, H. (1995) Science 269,

554-558.15. Davis, R. (1993) J. Biol. Chem. 268, 14553-14556.16. Derijard, B., Hibi, M., Wu, I.-H., Barrett, T., Su, B., Deng, T.,

Karin, M. & Davis, R. (1994) Cell 76, 1025-1037.17. Kyriakis, J. M., Banerjee, P., Nikolakaki, E., Dai, T., Ruble,

E. A., Ahmad, M. F., Avruch, J. & Woodgett, J. R. (1994) Nature(London) 369, 156-160.

18. Lee, J. C., Laydon, J. T., McDonnell, P. C., Gallagher, T. F.,Kumar, S., Green, D., McNulty, D., Blumenthal, M. J., Heys,J. R., Landvatter, S. W., Strickler, J. E., McLaughlin, M. M.,Siemens, I. R., Fisher, S. M., Livi, G. P., White, J. R., Adams,J. L. & Young, P. R. (1994) Nature (London) 372, 739-746.

19. Han, J., Lee, J.-D., Bibbs, L. & Ulevitch, R. J. (1994) Science 265,808-811.

20. Rouse, J., Cohen, P., Trigon, S., Morange, M., Alonso-Llamaz-ares, A., Zamanillo, D., Hunt, T. & Nebreda, A. R. (1994) Cell78, 1027-1037.

21. Freshney, N. W., Rawlinson, L., Guesdon, F., Jones, E., Cowley,S., Hsuan, J. & Saklatvala, J. (1994) Cell 78, 1039-1048.

22. Itoh, T., Yamauchi, A., Miyai, A., Yokoyama, K., Kamada, T.,Ueda, N. & Fujiwara, Y. (1994) J. Clin. Invest. 93, 2387-2392.

23. Terada, Y., Tomita, K., Homma, M. K., Nonoguchi, H., Yang, T.,Yamada, T., Yuasa, Y., Krebs, E., Sasaki, S. & Marumo, F.(1994) J. Biol. Chem. 269, 31296-31301.

24. Kwon, H. M., Itoh, T., Rim, J. S. & Handler, J. S. (1995) Biochem.Biophys. Res. Commun. 213, 975-979.

25. Dudley, D. T., Pang, L., Decker, S. J., Bridges, A. J. & Saltiel,A. R. (1995) Proc. Natl. Acad. Sci. USA 92, 7686-7689.

26. Lazar, D. F., Wiese, R. J., Brody, M. J., Mastick, C. C., Waters,S. B., Yamauchi, K., Pessin, J. E., Cuatrecasas, P. & Saltiel, A. R.(1995) J. Biol. Chem. 270, 20801-20807.

27. Cohen, D. M., Wasserman, J. C. & Gullans, S. R. (1991) Am. J.Physiol. 261, C594-C601.

28. Gotoh, Y., Nishida, E., Yamashita, T., Hoshi, M., Kawakami, M.& Sakai, H. (1990) Eur. J. Biochem. 193, 661-669.

29. Kameshita, I. & Fujisawa, H. (1989) Anal. Biochem. 183, 139.30. Gough, N. M. (1988) Anal. Biochem. 173, 93-95.31. Her, J.-H., Lakhani, S., Dent, P., Sturgill, T. & Weber, M. J.

(1993) Biochem. J. 295, 25-31.32. Charest, D. L., Mordret, G., Harder, K. W., Jirik, F. & Pelech,

S. L. (1993) Mol. Cell. Bio. 13, 4679-4690.

11246 Physiology: Cohen

Dow

nloa

ded

by g

uest

on

Apr

il 16

, 202

0

Proc. Natl. Acad. Sci. USA 93 (1996) 11247

33. Wu, J., Rossomando, A. J., Her, J. H., Vecchio, R. D., Weber,M. J. & Sturgill, T. W. (1991) Proc. Natl. Acad. Sci. USA 88,9508-9512.

34. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) MolecularCloning: A Laboratory Manual (Cold Spring Harbor Lab. Press,Plainview, NY), 2nd Ed.

35. Tsai-Morris, C., Cao, X. & Sukhatme, V. P. (1988) Nucleic AcidsRes. 16, 8835-8846.

36. Nordeen, S. K. (1988) BioTechniques 6, 454-458.37. Roberson, M. S., Misra-Press, A., Laurance, M. E., Stork, P. J. &

Maurer, R. A. (1995) Mo. Cell. Biol. 15, 3531-3539.38. Sun, P., Enslen, H., Myung, P. S. & Maurer, R. A. (1994) Genes

Dev. 8, 2527-2539.39. Hill, C. S., Marais, R., John, S., Wynne, J., Dalton, S. & Treisman,

R. (1993) Cell 73, 395-406.40. Janknecht, R., Ernst, W. H., Pingoud, V. & Nordheim, A. (1993)

EMBO J 12, 5097-5104.41. Marais, R., Wynne, J. & Treisman, R. (1993) Cell 73, 381-393.42. Cohen, D. M., Gullans, S. R. & Chin, W. W. (1996)J. Biol. Chem.

271, 12903-12908.43. Wu, J., Dent, P., Jelinek, T., Wolfman, A., Weber, M. J. &

Sturgill, T. W. (1993) Science 262, 1065-1069.44. Burgering, B. M., Pronk, G. J., van Weeren, P. C., Chardin, P. &

Bos, J. L. (1993) EMBO J. 12, 4211-4220.45. Cohen, D. M., Gullans, S. R. & Chin, W. W. (1996) J. Clin. Invest.

97, 1884-1889.

46. Derijard, B., Raingeaud, J., Barrett, T., Wu, I.-H., Han, J.,Ulevitch, R. J. & Davis, R. J. (1995) Science 267, 682-685.

47. Galcheva-Gargova, Z., Derijard, B., Wu, I.-H. & Davis, R. J.(1994) Science 265, 806-808.

48. Hipskind, R. A., Rao, V. N., Mueller, C. G., Reddy, E. S. &Nordheim, A. (1991) Nature (London) 354, 531-534.

49. Rao, V. N., Huebner, K., Isobe, M., Ar-Rushdi, A. & Croce,C. M. (1989) Science 244, 66-70.

50. Dalton, S. & Teisman, R. (1992) Cell 68, 597-612.51. Hipskind, R. A., Buscher, D., Nordheim, A. & Baccarini, M.

(1994) Genes Dev. 8, 1803-1816.52. Hill, C. S. & Treisman, R. (1995) Cell 80, 199-211.53. Treisman, R. (1994) Curr. Opin. Genet. Dev. 4, 96-101.54. Hill, C. S., Wynne, J. & Treisman, R. (1995) Cell 81, 1159-1170.55. Whitmarsh, A. J., Shore, P., Sharrocks, A. D. & Davis, R. J.

(1995) Science 269, 403-407.56. Zinck, R., Cahill, M. A., Kracht, M., Sachsenmaier, C., Hipskind,

R. A. & Nordheim, A. (1995) Mol. Cell. Biol. 15, 4930-4938.57. Alessi, D. R., Cuenda, A., Cohen, P., Dudley, D. T. & Saltiel,

A. R. (1995) J. Biol. Chem. 270, 27489-27494.58. Marshall, C. J. (1994) Curr. Opin. Genet. Dev. 4, 82-89.59. Kolch, W., Heidecker, G., Kochs, G., Hummel, R., Vahidi, H.,

Mischak, H., Finkenzeller, G., Marme, D. & Rapp, U. R. (1993)Nature (London) 364, 249-252.

Physiology: Cohen

Dow

nloa

ded

by g

uest

on

Apr

il 16

, 202

0