Central role for Rho in TGF-1-induced -smoothmuscle actin expression during epi - 《American Journal of Physiology》美国生理学杂志干细胞之家

干细胞之家 - 中国干细胞行业门户第一站 › 干细胞之家论坛 › 干细胞文献资源库 › 《American Journal of Physiology》美国生理学杂志 › Central role for Rho in TGF-1-induced -smoothmuscle ...



免疫细胞治疗专区	欢迎关注干细胞微信公众号

12 3 4 5 6 7 8 9 10 ... 25 下一页

返回列表

查看: 406775\|回复: 242	go Central role for Rho in TGF-1-induced -smoothmuscle actin expression during epi [复制链接]

轻羽

新手上路

Rank: 1

积分: 0
威望: 0
包包: 0

楼主

发表于 2009-4-21 13:36 |只看该作者 |倒序浏览 |打印

作者：AndrásMasszi, CaterinaDi Ciano, GáborSirokmány, William T.Arthur, Ori D.Rotstein, JiaxuWang, Christopher A. G.McCulloch, LászlóRosivall, IstvánMucsi,  AndrásKapus作者单位：1 Department of Surgery, The Toronto GeneralHospital and University Health Network, Toronto, Ontario M5G 1L7; Canadian Institutes of Health Research Group in MatrixDynamics, Faculty of Dentistry, University of Toronto, Toronto,Ontario, Canada M5S 3E2; Institute ofPathophysiology, Hungarian Academy o % k6 j6 w4 [# o+ S: X0 c; h  R
   ' G" J) w1 {5 S$ P, o0 Y3 M
   % t% D2 ?; N/ o2 H: W" t( X- Y7 E4 f


         8 C; \! N& m; ~2 Z$ V



                  4 v/ n4 n. t" R, _& u- Y* O



      【摘要】; r) R$ X" d" l$ z9 v
   Newresearch suggests that, during tubulointerstitial fibrosis, -smooth muscle actin (SMA)-expressing mesenchymal cells might derivefrom the tubular epithelium via epithelial-mesenchymal transition(EMT). Although transforming growth factor- 1 (TGF- 1 ) plays a key role in EMT, the underlying cellularmechanisms are not well understood. Here we characterizedTGF- 1 -induced EMT in LLC-PK 1 cells andexamined the role of the small GTPase Rho and its effector, Rho kinase,(ROK) in the ensuing cytoskeletal remodeling and SMA expression.TGF- 1 treatment caused delocalization and downregulationof cell contact proteins (ZO-1, E-cadherin, -catenin), cytoskeletonreorganization (stress fiber assembly, myosin light chainphosphorylation), and robust SMA synthesis. TGF- 1 induced a biphasic Rho activation. Stress fiber assembly was preventedby the Rho-inhibiting C3 transferase and by dominant negative (DN) ROK.The SMA promoter was activated strongly by constitutively active Rhobut not ROK. Accordingly, TGF- 1 -induced SMApromoter activation was potently abrogated by two Rho-inhibiting constructs, C3 transferase and p190RhoGAP, but not by DN-ROK. Truncation analysis showed that the first CC(A/T)richGG (CArG B) serumresponse factor-binding cis element is essential for the Rhoresponsiveness of the SMA promoter. Thus Rho plays a dual role inTGF- 1 -induced EMT of renal epithelial cells. It isindispensable both for cytoskeleton remodeling and for the activationof the SMA promoter. The cytoskeletal effects are mediated via theRho/ROK pathway, whereas the transcriptional effects are partially ROK independent. 5 A. x* k! D5 M/ i
      【关键词】 Rho kinase epithelialmesenchymal transdifferentiation transforming growth factor kidney proximal tubulecells" O% a- z; w  R$ ~
   INTRODUCTION* ?$ }) ^! f6 l& X* C9 p

TUBULOINTERSTITIAL FIBROSIS is the common pathomechanism whereby a variety ofchronic kidney diseases progress to end-stage renal failure ( 14, 48 ). The process is characterized by excessive deposition ofextracellular matrix (ECM) components and the consequent destruction ofnormal tissue architecture. A key factor in the underlying pathology isthe progressive accumulation of myofibroblasts, the main producers ofthe ECM ( 61 ). Indeed, both clinical studies and animalmodels indicate a strong positive correlation between the loss ofkidney function and the number of myofibroblasts or the expression of -smooth muscle actin (SMA; see Refs. 11, 44, and 68 ), a hallmark of the myofibroblastphenotype ( 18, 53 ). Despite the recognition of theircentral importance in disease progression, the origin of myofibroblastshas not been elucidated completely. Although in different organsmyofibroblasts may derive from smooth muscle cells ( 49 )and resident fibroblasts ( 23, 45 ), intriguing new studiessuggest that they may also arise from transdifferentiation of thetubular epithelium ( 19, 66 ). Recent studies provideevidence that epithelial-mesenchymal transition (EMT) contributes tothe generation of kidney fibroblasts during experimental renal fibrosis( 30, 42, 58 ) and that tubular epithelial cells have theability to transdifferentiate into SMA-positive mesenchymal cellstermed as myofibroblasts ( 19, 40, 66 ). During EMT,epithelial cells lose their polygonal morphology and adhesive cellcontacts and acquire fibroblast-like characteristics, includingelongated shape, expression of mesenchymal markers, and increasedmotility ( 7, 19, 58 ). Further transdifferentiation towardthe myofibroblast phenotype may ensue, as indicated by the expressionof SMA ( 40, 66 ). Importantly, similar processes appear toparticipate in the clinical pathogenesis of kidney fibrosis, asevidenced by the presence of cell populations that stain positively forboth epithelial and mesenchymal markers and express SMA ( 31, 47 ).

The multifunctional cytokine transforming growth factor(TGF)- 1 is a potent inducer of EMT in several tissues( 10 ) and has been shown to provoke SMA production invarious cell types ( 25, 40, 42 ). Moreover,TGF- 1 is both a product and an activator ofmyofibroblasts ( 46 ) and has been identified as a majormediator of kidney fibrosis ( 9 ). However, the mechanism whereby TGF- 1 treatment of epithelial cells triggers SMAexpression remains to be clarified.* E" v5 V3 p* j2 F$ v
+ w- K2 o8 b$ A7 f6 z0 w
A recent study has implicated the small GTPase Rho and its downstreameffector Rho kinase (ROK) in TGF- 1 -induced remodeling ofcell contacts in mammary epithelial cells ( 8 ). Theinvolvement of Rho in TGF- 1 signaling is of particularinterest since this small G protein is not only a major organizer ofthe cytoskeleton ( 24 ) but has been shown to regulate geneexpression ( 27, 34, 36, 50 ). Specifically, Rho has beenfound to be necessary for the constitutive expression of SMA in smoothmuscle cells ( 34 ). These observations led us tohypothesize that the Rho/ROK pathway might play an important role inthe TGF- 1 -induced SMA expression in kidney epithelial cells.! ?" _( w* f  j8 a- B- z! b2 L
* u4 x& N/ E5 J" M- ^: g
The regulation of the SMA promoter is complex ( 35 ) andshows substantial tissue specificity ( 25, 56 ).Importantly, its basal activity and inducibility markedly differbetween smooth muscle cells, fibroblasts, and endothelial cells( 25, 56 ). However, despite the increasingly recognizedpathological significance of SMA expression by epithelial cells, theregulation of the promoter has not been hitherto investigated in thiscellular context.: X* X% s$ F, [  B6 g
, Q. {3 F* N% k) g- d- w
To address these issues, we established a renal cell culture model inwhich TGF- 1 -induced EMT and SMA expression can beanalyzed reliably. Here we show that TGF- 1 -treatedLLC-PK 1 proximal tubule cells undergo EMT that manifests inloss of cell contacts, cytoskeleton remodeling, myosin light chain(MLC) phosphorylation, and SMA expression. TGF- 1 triggers a biphasic activation of Rho in LLC-PK 1 cells.Using various Rho- and ROK-interfering constructs, we provide evidencethat Rho, but not ROK, activation is indispensable for theTGF- 1 -induced SMA promoter activation. Our results showthat the first serum response factor (SRF)-binding cis -element (CArG B box) is essential for both Rhoinducibility and TGF- 1 responsiveness of the SMApromoter in LLC-PK 1 cells.  Y6 \, ^, U6 T& U9 R% Z0 z

METHODS2 _/ [/ d- ?1 l. Q9 l# v- Q
$ Y4 ~5 ~5 |3 A9 K$ Z& B
Materials and Antibodies* M; U8 T% O. o; {6 m* W3 b
! u1 \9 [. T: D& H  a
DMEM (1,000 mg/l glucose), Hank's balanced salt solution, PBS,FBS, penicillin/streptomycin, and trypsin were from GIBCO-BRL (Burlington, ON). We purchased human recombinant TGF- 1 from Sigma (St Louis, MO). The ROK inhibitor compound Y-27632 was fromCalbiochem (San Diego, CA). Anti- -catenin, anti-MLC,anti-phospho-MLC (specific for the diphosphorylated form of MLC),anti-Myc (9E10), anti-Rho, and anti-SRF antibodies were purchased fromSanta Cruz Biotechnology (San Francisco, CA). Anti-E-cadherin was fromBD Transduction Laboratories (Mississauga, ON); anti-cortactin andanti-ZO-1 were from Upstate Biotechnology (Lake Placid, NY);anti- -SMA (1E4) was from Sigma; and mouse anti-Histone antibody wasfrom Chemicon. Rhodamine and Alexa (488)-labeledphalloidin were from Molecular Probes (Eugene, OR). Horseradishperoxidase-conjugated anti-mouse and anti-rabbit IgG antibodies werepurchased from Amersham Biosciences (Uppsala, Sweden), and anti-goatantibody was from Santa Cruz. FITC and Cy3-labeled anti-goat,anti-rabbit, and anti-mouse secondary antibodies were from JacksonImmunoresearch Laboratories (West Grove, PA). Enhancedchemiluminescence reagent was from Amersham Biosciences.+ I6 P+ F3 h/ U
; F- Q: Q: ^7 @, m
Cells

LLC-PK 1 is a well-characterized and widely usedproximal tubular epithelial cell line from the pig ( 28 ).LLC-PK 1 cells (Cl 4 ) stably expressing therabbit AT 1 receptor were a kind gift from Dr. R. Harris( 12 ). Cells were kept in DMEM containing 10% FBS, 100 U/ml penicillin, and 100 µg/ml streptomycin in a humidified incubatorat 37°C under 5% CO 2. Cells were treated with 4 ng/ml TGF- 1 from 30% confluence for the indicated times.Morphological changes were detected by phase-contrast microscopy.' \: r7 {$ C$ |5 _% z; T
+ c- Z! ]( v8 a/ _
Plasmids2 T' Z% ^! \! \# L* w7 X

A 765-bp piece of the rat -SMA promoter ligated in apromoterless luciferase vector (PA 3 -Luc) designated asp765-SMA-Luc was a kind gift from Dr. R. A. Nemenoff( 21 ). The constructs p155/LacZ, p92/LacZ, and p56/LacZ(later designated as p155, p92, and p56) contain the first 155, 92, and56 bp of the rat -SMA promoter, respectively, and were inserted in a -galactosidase vector (pUC19/AUG). The original construct, p547/LacZwas obtained from Dr. G. K. Owens. The thymidine kinase-driven Renilla luciferase vector (pRL-TK; Promega) was used as aninternal control. The plasmid encoding the Myc-tagged constitutivelyactive (Q63L) form of RhoA was provided by Dr. G. Downey. Expressionvectors encoding the Myc-tagged constitutively active catalytic domainof p160 Rho-associated kinase I (ROK-CAT) and the dominant-negativeform of the kinase [ROK-RB/PH (TT) designated here as DN-ROK] were akind gift of Dr. Kozo Kaibuchi ( 43 ). Vectors encoding C3transferase ( 4 ) and p190RhoGAP ( 62 ) weredescribed previously. To generate p190RhoGAP-green fluorescence protein(GFP) expression plasmid, the p190RhoGAP insert was excised frompKH3-p190RhoGAP with Bam HI and Eco RI and thenligated in pEGFP-C1 that had been digested with Bgl II and Eco RI ( Bam HI and Bgl II arecompatible). pEGFP vector was from Clontech Laboratories (Palo Alto,CA), and pcDNA 3 was from Invitrogen (Burlington, ON).4 s, d) R+ s( h5 K( t! y
) M1 y3 b9 H. @, o& |* l. g: w
Transient Transfection and Reporter Enzyme Assays
. [4 ~4 _& H$ W: R6 j$ C" Z
Cells were plated on six-well plates 1 day before transfection.At 30% of confluence, cells were transfected with 1 µg of thecorresponding DNA using 2.5 µl FuGENE6 (Roche Molecular Biochemicals, Indianapolis, IN). For the SMA-luciferase construct (p765-SMA-Luc), cells were transfected with 0.5 µg promoter plasmid, 0.1 µg pRL-TK, and either 2 µg of the specific construct or 2 µg of the empty expression plasmid (pcDNA 3 ) per well. After a 16-htransfection period, cells were washed with Hank's balanced saltsolution and incubated in serum-free DMEM for 3 h. This wasfollowed by treatment with either vehicle or 10 ng/mlTGF- 1 (dissolved in 4 mM HCl and 0.1% albumin) for24 h. Subsequently, the cells were lysed in 250 µl of passivelysate buffer (Promega) and exposed to one cycle of freezing/thawing( 80°C/37°C), and the samples were clarified by centrifugation(13,000 rpm at 4°C, 5 min). Firefly and Renilla luciferaseenzyme activities were determined in an aliquot of the supernatantusing the Dual-Luciferase Reporter Assay Kit (Promega) and a BertholdLumat LB 9507 luminometer according to the manufacturer's instructions. Results are expressed as a normalized ratio obtained bydividing the firefly luciferase activity by the Renilla luciferase activity of the same sample. Each transfection was done in duplicate, and determinations for each group were repeated at least three times.For the -galactosidase-coupled SMA promoters, cells were transfectedwith 1 µg promoter construct, 0.1 µg pRL-TK, and 2 µg of eitherthe empty vector or the RhoAQ63L construct per well. We followed asimilar time schedule as for the luciferase constructs, but the cellswere scraped in 250 µl of 100 mM potassium phosphate buffer (pH 7.8)supplemented with 1 mM dithiothreitol. After three freeze-thaw cycles,the lysates were cleared by centrifugation, and -galactosidaseactivity was determined by the luminescent -galactosidase kit(Clontech Laboratories). Renilla luciferase activity wasmeasured from an aliquot of the same sample by the renilla luciferasereporter assay system kit (Promega) following the manufacturer'sprotocol. The endogenous -galactosidase activity of nontransfectedcells was determined and subtracted from the total values. Thetransfection-dependent -galactosidase activity was then normalizedto the Renilla luciferase activity of the same sample.Cotransfection efficiency was assessed by immunofluorescence. Cellswere transfected with GFP and the Myc-labeled construct of interest(RhoAQ63L, ROK-CAT, or DN-ROK) using 0.5 and 2 µg DNA, respectively,and the percentage of Myc-expressing cells in the GFP-expressingpopulation was determined by immunostaining the epitope tag. Inagreement with our previous data ( 59 ), 90% of greencells were Myc positive. In the case of C3 transferase, which was notlabeled with an epitope, we used a functional assay. Aftercotransfection of C3 transferase with GFP, we stained the cells withrhodamine phalloidin and compared the F-actin structure in GFP-positiveand -negative cells. Abnormal F-actin structure (stress fiberdisruption and major reduction in staining) was observed in 97% of thegreen cells. Transfection with GFP alone had no effect.

Western Blotting+ F0 x+ w+ V7 S7 {# B! N' x
2 z; C* m% L8 i; i
Cells cultured with or without TGF- 1 on 10-cmdishes were scraped in 800 µl ice-cold Triton buffer [30 mM HEPES(pH 7.4), 100 mM NaCl, 1 mM EGTA, 20 mM NaF, 1% Triton X-100, 1 mMphenylmethylsulfonyl fluoride (PMSF), 20 µl/ml protease inhibitorycocktail (Pharmingen BD Biosciences), and 1 mMNa 3 VO 4 ]. The samples were clarified bycentrifugation at 13,000 rpm for 5 min at 4°C. We added 2× Laemmlibuffer [375 mM Tris (pH 6.8), 10% SDS, 20% glycerol, 0.005% bromphenol blue, and 2% -mercaptoethanol] to the supernatants andboiled for 5 min. Protein concentrations were determined by theBradford method (Bio-Rad Laboratories, Hercules, CA). Equal amounts ofprotein were separated on 10% SDS-polyacrylamide gels with a Bio-RadProtean II apparatus and transferred to nitrocellulose membranes(Bio-Rad). Blots were blocked in Tris-buffered saline and 0.1% Tween20 (TBS-T) containing 5% albumin for 1 h. The membranes wereincubated with primary antibody (diluted in TBS-T containing 1%albumin) for 1 h, washed extensively, and incubated with the appropriate peroxidase-conjugated secondary antibody for another hour.After the final washes, immunoreactive bands were visualized byenhanced chemiluminescence reaction. The bands were quantified using aBio-Rad GS-690 Imaging Densitometer and the Molecular Analyst software(Bio-Rad), as in Ref. 32. Data are presented asrepresentative blots of at least three similar experiments.) E; g- c9 u4 g* G: _+ n* [
4 J: O( g4 [, _  F1 `7 ^
MLC Phosphorylation
' H3 y0 p- C3 S
To detect changes in MLC phosphorylation, the nonphosphorylated,mono- and diphosphorylated forms of MLC were separated using nondenaturating urea-glycerol PAGE. Cells plated on 10-cm dishes andgrown until confluence were serum starved for 3 h, treated withvehicle or 10 ng/ml TGF- 1 for 4 h, and then lysedin 1.5 ml of acetone containing 10% trichloracetate and 10 mMdithiothreitol. The lysates were spun down, and the resulting pelletwas washed in 1 ml pure acetone. The solvent was aspirated, and thepellet was air-dried at room temperature for 1 h. The samples werethen dissolved in 200 µl sample buffer [20 mM Tris (pH 8.6), 23 mM glycine, 8 M urea, 234 mM sucrose, 10 mM dithiothreitol, and 0.01% bromphenol blue] with periodic agitation and subjected tourea-glycerol PAGE using a 12% gel containing 40% glycerol, 20 mMTris (pH 8.6), and 23 mM glycine, as described previously( 59 ). Separated proteins were transferred tonitrocellulose membranes, and Western blotting was performed using ananti-MLC antibody.
8 Z  t9 A' ~$ z+ u( X5 U
Preparation of Glutathione-S-Transferase-Rho-Binding Domain Beadsand Measurement of RhoA Activity

This method is a pulldown affinity assay based on the ability ofthe Rho-binding domain (RBD) of Rhotekin (amino acids 7-89) toselectively bind the GTP-loaded form of Rho. The recombinant glutathione- S -transferase (GST)-RBD protein was preparedfrom Escherichia coli, as described previously( 5 ) with slight modification. Briefly, DH5 E. coli culture transformed with pGEX plasmid encoding GST-RBD waspelleted and dissolved in 10 ml STE buffer [10 mM Tris (pH 8.0), 150 mM NaCl, and 1 mM EDTA] supplemented with 5 mM dithiothreitol, 100 µg/ml lysozyme, 20 µl/ml protease inhibitory cocktail, and 1 mMPMSF. For completing bacterial lysis, we applied two cycles of FrenchPress (900 psi) and added 1% sarcosyl for 10 min. The supernatant waspurified by centrifugation, supplemented with 1% Triton X-100, andincubated with 1 ml gluthatione-Sepharose 4B beads (Amersham PharmaciaBiotech) for 1 h at 4°C by constant agitation. The beads werewashed three times with 10 ml of STE buffer containing 1% Triton X-100and then three times with STE buffer alone and stored at 4°C. Todetermine the short-term effects of TGF- 1 on Rhoactivity, confluent cells were serum starved for 3 h and thenexposed to 10 ng/ml TGF- 1 for the indicated timeperiods. To determine the long-term effect of TGF- 1, the cells were grown to 60% confluence and then exposed toTGF- 1 for 1-3 days. After treatment, the cells werewashed with ice-cold PBS, lysed, and scraped in 800 µl lysis buffer[100 mM NaCl, 50 mM Tris (pH 7.6), 20 mM NaF, 10 mM MgCl 2,and 1% Triton X-100, supplemented with 0.1% SDS, 0.5% sodiumdeoxycholate, 20 µl/ml protease inhibitor cocktail, 1 mMNa 3 VO 4, and 1 mM PMSF]. The detergent-insoluble fraction was removed by centrifugation, and thelysates were incubated with 10-15 µg of GST-RBD for 45 min at4°C. The beads were washed three times with 1 ml lysis buffer supplemented only with 1 mM Na 3 VO 4 and boiledin 25 µl of 2× Laemmli buffer for 5 min. The bead-associatedproteins were resolved by 15% SDS-PAGE, and the captured Rho proteinwas detected by Western blotting using an anti-Rho antibody.0 D' _; Y9 Q2 b& n0 Y6 i

Immunofluorescence Microscopy and Phalloidin Staining2 m, A8 o7 L. S4 ]! @

Cells were cultured on 25-mm coverslips and treated withTGF- 1 or vehicle for the indicated periods. After thecoverslips were washed with PBS, cells were fixed in 4%paraformaldehyde (PFA) for 30 min. PFA was quenched with PBS containing100 mM glycine, and the coverslips were washed thoroughly with PBS.Cells were permeabilized for 20 min in PBS containing 0.1% TritonX-100. For phalloidin staining, cells were incubated for 1 h withrhodamine-labeled phalloidin in 1:100 dilution. For E-cadherinstaining, cells were fixed and permeabilized in ice-cold methanol for 5 min. Nonspecific binding was blocked with 5% albumin in PBS for 1 h. Subsequently, the coverslips were incubated with the primaryantibodies for 1 h. After being washed six to eight times withPBS, samples were incubated with the fluorescently labeled secondaryantibodies for 1 h. The coverslips were washed and mounted onslides using Fluorescence Mounting Medium (Dako Diagnostics Canada,Mississauga, ON). Samples were viewed using a Nikon Eclipse TE200microscope (100× objective) coupled to a Hamamatsu cooledcharge-coupled device camera (C4742-95) controlled by the SimplePCI software.

Preparation of Nuclear Extracts and EMSAs
; k3 V2 V# L/ {0 Q/ l) w  Z/ {) M. ?
Nuclear extracts were prepared according to a modified Digman'smethod, as described previously ( 63 ). The double-stranded CArG B oligonucleotide (5'-GAGGTCCCTATAGGTTTGTG-3') was synthesized andpurified commercially (GIBCO-BRL). The EMSA probe was generated by endlabeling single-stranded oligonucleotide (20 µM) with 150 µCi[ 32 P]ATP (3,000 Ci/mmol; Mandel) using T4 polynucleotidekinase. The labeled single-stranded oligonucleotide was annealed andpurified from unincorporated nucleotide using ProbeQuant TM G-50 Micro columns (Pharmacia Biotech). For EMSAs, the samples were incubated for30 min at room temperature in 20 µl of a reaction mixture containing1× binding buffer [10 mM Tris · HCl (pH 7.5),100 mM KCl, 50 mM NaCl, 1 mM dithiothreitol, 1 mM EDTA, and 5%glycerol], ~50 pg (50,000 cpm) labeled probe, 10 µg nuclearextract, and 0.25 µg poly(dA-dT). Samples were separated byelectrophoresis on 4.5% polyacrylamide gels at 150 volts in 45 mM Trisborate and 1 mM EDTA. For supershift assays, 10 µg nuclear extractswere preincubated for 15 min with 2 µl SRF antibody.6 s+ n, o! t# b+ W# D( t9 A
* p: R6 q8 d+ S' C
Statistical Analysis. d5 @% Y: {8 I

Data are presented as representative blots from three similarexperiments or as means ± SE for the number of experiments( n ) indicated. Statistical significance was determined byStudent's t -test or ANOVA (one-way ANOVA; Prism, GraphPad Software)., X$ f5 W, @# w8 Q8 ~. A, w

RESULTS

Characterization of TGF- 1 -Induced EMT inLLC-PK 1 Cells  G! W8 l5 B, @+ K* \9 E1 x: [% s3 ^

Morphology and cell contact proteins. To assess whether TGF- 1 induces characteristic EMT inthe proximal tubule cell line LLC-PK 1, 20-30%confluent cultures were treated with 4 ng/ml TGF- 1, andthe subsequent changes were followed by phase-contrast andimmunofluorescence microscopy. Vehicle-treated control cells formedislands within which individual cells showed typical polygonalappearance and were tightly attached to each other (Fig. 1 A ). In contrast, cellstreated with TGF- assumed an elongated shape, and many cells lostcontact with their neighbors (Fig. 1 B ). Thesecharacteristics developed gradually; the effect was discernible after24 h, whereas after 3 days 80% of the cells exhibitedfibroblast-like shape. To visualize the reorganization of tightjunctions and adherent junctions, cells were immunostained for ZO-1 andfor E-cadherin and -catenin. In control cells, ZO-1 accumulated atthe cell boundary, forming a sharp narrow line, and showed a faintpunctate labeling in the cytosol (Fig. l C ). OnTGF- 1 treatment, the peripheral staining becamediscontinuous, and ZO-1 accumulated in rod-like structures that wereperpendicular to the cell membrane (Fig. 1 D ).TGF- 1 caused delocalization of E-cadherin and -catenin from the cell periphery (Fig. 1 E-H ). Moreover,increased cytosolic -catenin staining was observed that wasfrequently accompanied with enhanced perinuclear/nuclear labeling (Fig. 1 H ). Consistent with this, TGF- 1 increasedthe amount of -catenin in the nuclear fraction, as revealed byWestern blots (Fig. 1 I ). In addition to relocalization,TGF- 1 induced a marked reduction in the overallexpression of both adheren junction proteins (Fig. 1 J ).However, although a 3-day treatment resulted 80%) of E-cadherin (Fig. 1 J, top ), thedecrease in -catenin was of smaller magnitude (Fig. 1 J, bottom ), consistent with the observed cytosolic/nuclearaccumulation of this protein./ {3 H% G2 Y* s2 |- _/ X1 y

Fig. 1. Transforming growth factor (TGF)- 1 inducesdisassembly of cell-cell contacts and downregulation of junctionalproteins in LLC-PK 1 cells. Cells grown on tissue culturedishes ( A and B ) or glass coverslips( C-H ) were either treated with vehicle alone (Control; A, C, E, and G ) or exposedto 4 ng/ml TGF- 1 ( B, D, F, and H ) for 3 days. Morphological changes wereexamined by phase-contrast microscopy ( A and B ).Distribution of the tight junction component ZO-1 ( C and D ) and the adherens junction proteins E-cadherin( E and F ) and -catenin ( G and H ) were visualized by immunofluorescent microscopy, asdescribed in METHODS. For A and B ×10 and for C-H ×100objectives were used. I : cells were treated with vehicle( ) or 4 ng/ml TGF- 1 ( ) for 3 days, and nuclearlysates were prepared and then probed for -catenin ( top ).To test for loading of nuclear proteins, the same blot was stripped andreprobed with an anti-Histone antibody ( bottom ). The effectof TGF- 1 on the total amount of E-cadherin and -catenin was analyzed by Western blotting ( J ). Lysatesfrom cells treated as in I and containing an equal amount ofprotein were subjected to SDS-PAGE, blotted on nitrocellulose, andprobed with the corresponding primary and secondary antibodies.
1 Y" {+ H& Y* i/ D& T: Q/ j
Cytoskeletal reorganization. Next we studied the effect of TGF- 1 on cytoskeletalstructure. Control LLC-PK 1 cells exhibited a strongperipheral F-actin ring with slim central stress fibers (Fig. 2 A ), whereasTGF- 1 -treated cells showed a decrease in marginalF-actin but contained much thicker central stress fibers that weremostly oriented parallel to the long axis of the cells (Fig. 2 B ). TGF- 1 has been reported to increase MLCphosphorylation in endothelial cells ( 29 ), a processassociated with cell contact remodeling ( 22, 33 ). Therefore, we analyzed whether a similar phenomenon occurs in LLC-PK 1 cells. Staining for the diphosphorylated (active)form of MLC gave only background and nuclear labeling in controls (Fig. 2 C ) but visualized distinct cytosolic filaments inTGF- 1 -treated cells (Fig. 2 D ). This findingwas substantiated by biochemical means; with the use of urea-glycerolPAGE, we separated the nonphosphorylated and mono- and diphosphorylatedforms of MLC in lysates obtained from control andTGF- 1 -challenged cells. The various forms were visualized by Western blotting using an anti-MLC antibody that reactsindependently of phosphorylation status. Figure 2 G shows that TGF- 1 exposure resulted in the accumulation of thephosphorylated forms of MLC.4 A1 `1 K: l! o% z, h
& _2 ~# ?% g) m6 u5 \/ k
Fig. 2. TGF- 1 induces cytoskeleton reorganizationand myosin light chain (MLC) phosphorylation in LLC-PK 1 cells. LLC-PK 1 cells were treated with vehicle alone(Control; A, C, and E ) or 4 ng/mlTGF- 1 ( B, D, and F ) for3 days and then fixed and permeabilized. A and B :F-actin was visualized by rhodamine-phalloidin staining. C and D : distribution of the diphosphorylated form of MLC(ppMLC) was detected by immunoflourescence microscopy after stainingwith an antibody specific for ppMLC. This antibody gave nuclearlabeling that was not different in control andTGF- 1 -treated cells. Note the presence of a large numberof ppMLC-positive cytosolic fibers that were seen exclusively in theTGF- 1 -treated cells. (Bar: 10 µm). G : toseparate differentially phosphorylated forms of MLC, serum-depletedcells were left untreated ( ) or exposed to TGF- 1 for4 h ( ) and lysed. Lysates were subjected to nondenaturingurea-glycerol PAGE, followed by Western blotting with anonphospho-specific anti-MLC antibody, as described in METHODS. Under our conditions, eachform of MLC (non-, mono-, and diphosphorylated) ran as a doublet. Thiswas verified by using a monophospho-specific antibody and by inducingmaximal MLC phosphorylation with the phosphatase inhibitor calyculin A(data not shown). E and F : cortactin wasvisualized using a specific monoclonal anti-cortactin antibody. Arrowin F indicates cortactin accumulation in a largelamellopodium.
5 X5 M/ v& o, ]0 A7 `* W6 T
Mesenchymal transdifferentiation results in a motile phenotype,characterized by leading edge formation ( 51 ). To address whether TGF- 1 can elicit such an effect inLLC-PK 1 cells, we stained the cells for cortactin, asensitive marker of cortical cytoskeleton dynamics and actin-basedmotility ( 64 ). In untreated cells, cortactin was dispersedevenly in the cytosol with a faint accumulation along the entire cellperiphery (Fig. 2 E ). In TGF- 1 -treated cells,cortactin distribution became highly polarized, visualizing largelamellipodia that developed in many cells (Fig. 2 F ). This morphology is suggestive of increased migratory potential of the TGF- 1 -treated LLC-PK 1 cells.( H6 P3 x* A* D7 q) b  F6 o# g
$ R  E0 @' }: }8 Q1 R( @
SMA expression. -SMA expression is a marker of myofibroblasts, a cell type thatrepresents an advanced phase of EMT ( 11, 48, 68 ). We therefore investigated the effect of TGF- 1 SMA proteinexpression using Western blotting and immunfluorescence microscopy. NoSMA was detected in control LLC-PK 1 cells, whereasTGF- 1 exposure induced strong SMA expression by 3 days,which slightly increased further by 6 days (Fig. 3 A ). Accordingly,immunofluorescence images showed only weak background staining incontrol cells, whereas a 3-day exposure to TGF- 1 inducedintense labeling in 60% of the cells (see below). Importantly, thenewly synthesized SMA assembled in thick fibers (Fig. 3 B ).This filamentous pattern was the most robust morphological marker ofthe effect, since it was observed exclusively inTGF- 1 -treated cells and was present independent of theoverall magnitude of SMA expression.
( E4 A: K7 i3 c# F
Fig. 3. TGF- 1 induces -smooth muscle actin(SMA) expression and stimulates SMA promoter activity inLLC-PK 1 cells. A : cells were treated withvehicle ( ) or 4 ng/ml TGF- 1 for the indicated timesand lysed. Equal amounts of protein from these lysates were subjectedto Western blotting using an anti-SMA antibody. B : controlor TGF- 1 -treated (3 days) cells were fixed and stainedfor SMA. C : cells grown on 6-well plates were cotransfectedwith a 765-bp sequence of the rat SMA promoter coupled to the fireflyluciferase gene (p765-SMA-Luc, 0.5 µg/well) and with Renilla luciferase (pRL-TK, 0.1 µg/well), as described in METHODS. Later (1 day), the cellswere treated with vehicle (Control) or with 10 ng/mlTGF- 1 for 24 h. At the end of this period, thecells were lysed, and luciferase activities of the samples weremeasured by luminometry. TGF- 1 caused a 3.52 ± 0.26-fold ( n = 12) increase in the normalizedp765-SMA-Luc activity ( P" B6 J9 e# B* \' V9 A7 `
% ^3 s# U) k( k- W
To investigate the effect of TGF- 1 on SMA genetranscription in a kidney epithelial cell setting, LLC-PK 1 cells were cotransfected with a construct encoding a 765-bp sequence ofthe SMA promoter fused to the firefly luciferase gene( 21 ), along with Renilla luciferase. Thiscotransfection method provides a highly reliable way to correct forpotential differences in transfection efficiency. Figure 3 C shows that a 24-h exposure to TGF- 1 induced a 3.52 ± 0.26-fold ( n = 12) increase in SMA promoteractivity, indicating that it rapidly and efficiently stimulated thepromoter in LLC-PK 1 cells. Taken together, our data showthat TGF- 1 induced the transformation ofLLC-PK 1 proximal tubule cells from an epithelial to amesenchymal/myofibroblast-like phenotype. This EMT manifested incharacteristic shape changes, downregulation of tight and adherensjunction components, cytosolic and nuclear -catenin accumulation,F-actin reorganization, MLC phosphorylation, leading edge formation,and robust de novo SMA synthesis, presumably through transcriptionalactivation. In contrast to MCT cells ( 42 ), epidermalgrowth factor failed to induce any of the above changes inLLC-PK 1 cells (data not shown). This observation suggeststhat, in terms of EMT, LLC-PK 1 cells are selectivelyresponsive to TGF- 1.

Role of Rho in TGF- 1 -Induced SMA Expression andCytoskeletal Reorganization
$ M7 }2 f4 Z; n
Effect of TGF- 1 on Rho activity inLLC-PK 1 cells. The TGF- 1 -induced changes in cytoskeletal organizationand SMA expression raised the possibility that the small GTPase Rho maybe a central mediator of these effects, since Rho is known to increasestress fiber formation and MLC phosphorylation ( 24 ) andhas been shown to stimulate SMA expression in muscle cells ( 34 ). To test this hypothesis, first we measured whetherTGF- 1 induces detectable changes in the amount of active(GTP-bound) Rho in LLC-PK 1 cells. Cell lysates from controland TGF- 1 -treated cells were incubated with a GST fusionprotein containing the RBD of Rhotekin, which selectively captures theactive form of Rho ( 5 ). TGF- 1 exposurecaused a rapid and transient increase in the amount of GTP-Rho. Theeffect was visible after 1 min, peaked around 5 min, and decayedthereafter (Fig. 4 A ). Thecytoskeletal changes (i.e., stress fiber assembly and enhanced MLCphosphorylation) observed after a 3-day TGF- 1 treatmentraised the possibility that the basal Rho activity might be chronicallyelevated in the transformed cells. We tested whether, in addition toimmediate Rho stimulation, TGF- 1 treatment resulted in alater Rho stimulation. Consistent with this notion, we found that,after 24 h of TGF- 1 treatment, an elevation in Rhoactivity was noticeable again compared with the control. Thislate-onset Rho activation further increased and persisted throughoutthe whole course (3 days) of the experiment (Fig. 4 B ). ThusTGF- 1 induced a biphasic change in Rho activity inLLC-PK 1 cells; a rapid and transient response was followed by a chronic elevation.
: Y! P# `; E3 w" D
Fig. 4. Rho is activated by TGF- 1 in a biphasicmanner in LLC-PK 1 cells. A : cells grown on 10-cmtissue culture dishes were serum starved for 3 h and then treatedeither with vehicle ( ) or with 10 ng/ml TGF- 1 for theindicated times and lysed. To capture active Rho (i.e., GTP bound), thelysates were incubated with glutathione beads covered with aglutathione- S -transferase (GST) fusion protein containingthe Rho-binding domain of Rhotekin, as described in METHODS. Beads were then separatedby centrifugation, and the captured Rho was detected by Westernblotting (Active). To verify that equal amounts of Rho were subjectedto the assay, an aliquot of the cell lysates was taken from each samplebefore incubating the rest with the fusion protein, and the totalamount of Rho was determined by Western blotting (Total). In fiveseparate experiments, a 5-min treatment with TGF- 1 caused a 1.7 ± 0.3-fold ( P B : cells were cultured thepresence of 10 ng/ml TGF- 1 for the indicated times,after which the samples were processed as in A. Note that,after 24 h of TGF- 1 treatment, Rho activationbecame noticeable again and increased further by days 2 and 3 despite the lower total Rho content of the transformedcells." t8 i4 p1 E8 ^7 w$ e' |( c- h
. d1 ~/ g' o" P( V
Involvement of Rho in the TGF- 1 -induced responses. Next we addressed whether there is a causal relationship between theobserved Rho activation and the subsequent changes in SMA expressionand cytoskeleton organization. We applied two approaches. First, wetested whether constitutively active Rho elicits similar cytoskeletalresponses in LLC-PK 1 cells as observed afterTGF- 1 treatment. Second, we investigated whetherinterference with the activation of endogenous Rho could preventTGF- 1 -induced cytoskeletal and transcriptional effects.Cells were transfected with a construct encoding a GTPase-defective(and thereby constitutively active) Myc-tagged Rho mutant (RhoAQ63L).Later (2 days), the cells were doubly stained using anti-Mycantibody and either Alexa(488)-phalloidin oranti-diphospho-MLC antibody. RhoAQ63L-expressing cells showed abundantthick stress fibers (Fig. 5 A )and substantially increased labeling for diphospho-MLC (Fig. 5 B ). Having confirmed the efficiency of RhoAQ63L in exertingstrong cytoskeletal effects, we tested whether it can drive the SMApromoter by cotransfecting RhoAQ63L with the SMA reporter system.RhoAQ63L provoked a 4.72 ± 0.52-fold ( n = 14)increase in SMA promoter activity, indicating that Rho is a potentactivator of this construct in LLC-PK 1 cells (Fig. 5 C ).  T6 w1 u1 b( W# _" ]% r
* B1 s/ h3 z. e, s8 f5 R! ?
Fig. 5. Constitutively active RhoA induces cytoskeletal responsesand stimulates the SMA promoter in LLC-PK 1 cells. A and B : cells grown on coverslips weretransfected with a constitutively active mutant of Myc-tagged RhoA(RhoAQ63L). F-actin structure was visualized byAlexa(488)-phalloidin staining ( A ), whereas MLCphosphorylation was detected using anti-ppMLC goat primary andFITC-labeled anti-goat secondary antibody ( B ). To identifysuccessfully transfected cells, the samples were also stained with ananti-Myc mouse primary and a Cy3-labeled anti-mouse secondary antibody(Myc). Arrows on corresponding images ( top and bottom ) indicate identical cells. Note that RhoAQ63L causedrobust stress fiber assembly and led to the formation of ppMLC-positivefibers and increased peripheral ppMLC staining. C : cellsgrown on 6-well plates were cotransfected with p765-SMA-Luc (0.5 µg/well), pRL-TK (0.1 µg/well), and either 2 µg of the emptyvector (Control) or the active Rho mutant (RhoAQ63L), as described in METHODS. SMA promoter activity wasdetected 48 h later by luminometry. Active Rho induced a 4.72 ± 0.52-fold ( n = 14) increase in SMA promoter activity( P4 m( o. ~* k1 q' R; D

To interfere with endogenous Rho activity before TGF- 1 challenge, we used two constructs that inhibit Rho by distinctmechanisms. We expressed either Clostridium botulinum C3transferase (C3), which selectively ADP-ribosylates and therebyinactivates Rho ( 27, 65 ), or the GFP-tagged version ofp190RhoGAP that enhances the endogenous GTPase activity of Rho, therebyterminating its action ( 4 ). C3 or p190RhoGAP wascotransfected with the SMA reporter system, and 24 h later thecells were exposed to TGF- 1 for 1 day. Expression of C3or p190RhoGAP did not significantly change the basal luciferaseexpression but strongly inhibited the TGF- 1 -induced risein SMA promoter activity (Fig. 6 ). These findings show that Rho activity is indispensable for theTGF- 1 -induced upregulation of the SMA promoter.* }$ C7 n4 Z" k0 g# _9 i
9 o' U/ }% @# W2 g0 @
Fig. 6. Inhibition of Rho abrogates theTGF- 1 -induced stimulation of the SMA promoter. Cellswere cotransfected with p765-SMA-Luc (0.5 µg), pRL-TK (0.1 µg), andeither 2 µg of empty vector (Control) or C3 transferase (C3) orp190Rho GTPase-activating protein (GAP)-green fluorescence protein(GFP; p190RhoGAP). A day after transfection, the cells were serumstarved for 3 h and then exposed to vehicle (veh) or 10 ng/mlTGF- 1 for 24 h, as indicated. Subsequently,p765-SMA-Luc activity was determined. Values were normalized to thebasal activity measured in vehicle-treated control cells. TheTGF- 1 -induced degree of activation(TGF- 1 /veh) in each group was as follows: control:3.45 ± 0.38; C3: 1.72 ± 0.18; p190RhoGAP: 1.69 ± 0.25 (for control vs. C3 and control vs. p190RhoGAP P$ X4 b4 o. I3 G: w$ j1 h
) r' ]7 S9 h  b& I& B) t7 x0 [
To discern whether the TGF- 1 -induced cytoskeletalreorganization and in situ SMA protein expression are also Rhodependent, we compared the effects of TGF- 1 in theabsence and presence of Rho inhibition. Successful transfection withthe C3 construct was verified by expression of the cotransfected GFP(see METHODS ), whereas theexpression of p190RhoGAP-GFP could be visualized directly. Expressionof GFP alone did not interfere with the F-actin structure under controlor stimulated conditions (Fig. 7 A, left ). In contrast, C3 causeda complete loss of stress fibers in untreated cells and prevented theTGF- 1 -induced remodeling of the F-actin cytoskeleton (Fig. 7 A, right ). Furthermore, although GFP alonehad no effect on the diphospho-MLC staining (Fig. 7 B, left ), inhibition of Rho prevented theTGF- 1 -induced accumulation of diphospho-MLC (Fig. 7 B, right ). Importantly, TGF- 1 failed to induce normal SMA upregulation in C3-expressing cells,although it had a strong effect in nontransfected or onlyGFP-expressing cells (Fig. 8 A ). To quantify theseeffects, we determined the percentage of SMA-expressing cells afterTGF- 1 treatment in nontransfected, GFP-transfected, andGFP plus C3-transfected cells (Fig. 8 B ). A 3-day exposure toTGF- 1 induced SMA expression in 61 ± 9% ofcontrol cells. Expression of GFP alone resulted in a modest reductionin the percentage of SMA-positive cells (to 41 ± 6%), whereascotransfection with C3 abolished SMA expression. These findings clearlyindicate that C3 strongly inhibited SMA expression, and the absence ofSMA in individual C3-expressing cells cannot be accounted for by theless than complete transformation observed in the control cells.Furthermore, overexpression of p190RhoGAP also abrogatedTGF- 1 -induced SMA protein expression (Fig. 8 C ). These findings confirmed that Rho is required forendogenous SMA expression in LLC-PK 1 cells.* o1 i8 d( R$ r8 r, t

Fig. 7. Inhibition of Rho prevents the TGF- 1 -induced stressfiber assembly and MLC phosphorylation in LLC-PK 1 cells.Cells grown on coverslips were either cotransfected with C3 transferase(2 µg) and GFP (0.5 µg; to identify the C3-expressing cells) ortransfected with GFP (2 µg) alone, as indicated at top.Cells were then incubated without (control) or with 4 ng/mlTGF- 1 for 3 days and fixed. Samples were processed tovisualize F-actin using rhodamine-phalloidin ( A ) oranti-ppMLC and a Cy3-labeled anti-goat secondary antibody( B ). Arrows indicate identical areas on the correspondingimages obtained by the red and green filters. Note thatTGF- 1 induced robust stress fiber formation and MLCphosphorylation in nontransfected or only GFP-expressing cells, but itfailed to do so in C3-expressing cells.

Fig. 8. Insitu SMA expression is abolished by C3 transferase and p190RhoGAP. A : cells were transfected with GFP alone or cotransfectedwith C3 and GFP, exactly as described in Fig. 7. Cells were then leftuntreated or exposed to 4 ng/ml TGF- 1 for 3 days, fixed,and stained for SMA using monoclonal anti-SMA and a Cy3-labeledanti-mouse secondary antibody. Arrows point to successfully transfectedcells and indicate identical areas of the corresponding images. B : quantification of the inhibition of SMA expression by C3.The percentage of SMA expressors was determined in nontransfected,GFP-transfected, or GFP C3-transfected cells after 3 days ofTGF- 1 treatment. The results are from 3 independentexperiments. In each experiment, 50-100 cells were counted in eachcategory. C : cells were transfected with the p190RhoGAP-GFPconstruct (2 µg), the expression of which can be directly visualized.Treatments and staining for SMA were performed as in A.Arrows point to successfully transfected cells and indicate identicalareas of the corresponding images.( s; H; c* }) A+ \+ x
. @2 G6 P* r5 }. f: t$ K
ROK is a downstream effector of Rho that has been implicated as acentral mediator of Rho's cytoskeletal effects ( 3 ).However, the involvement of ROK in Rho-dependent gene transcription is controversial and remains to be elucidated ( 13, 34, 38, 50 ). We therefore aimed to assess the role of ROK in SMApromoter activity. In the first set of experiments, we expressed theMyc-tagged catalytic domain of ROK (ROK-CAT) that acts in aconstitutively active manner ( 43, 59 ). To verify thatROK-CAT expression was sufficient to exert functional effects, wecompared the actin cytoskeleton organization in control andROK-CAT-expressing (Myc-positive) cells. In ROK-CAT expressors, theactin skeleton showed significant alterations as follows: thick F-actinfibers were formed that radiated from central F-actin foci or patches(Fig. 9 A ). In contrast to themarked cytoskeletal effects, the same construct caused only marginalchanges in SMA promoter activity, inducing a increase inSMA-luciferase expression (Fig. 9 B ). Although ROK-CAT initself was insufficient to mimic the effect of Rho on the SMA promoter,it was conceivable that it might contribute to theTGF- 1 -induced effect. To address this, we transfectedcells with a Myc-tagged kinase-deficient ROK (DN-ROK) that has beenshown to act as a dominant-negative mutant ( 43, 59 ).DN-ROK caused almost complete stress fiber disassembly and stronglyinhibited the cytoskeletal reorganizing effects of TGF- 1 (Fig. 9 C ). In contrast, DN-ROK did not reduce the magnitudeof the TGF- 1 -induced SMA promoter activation and causedonly a slight inhibition ( 25%) when the TGF- 1 -induced increases were compared betweenmock-transfected and DN-ROK-transfected cells. Consistent with thisfinding, long-term pretreatment of the cells with the ROK inhibitorY-27632 failed to significantly change the SMA promoter activity inTGF- 1 -stimulated cells and caused only partialinhibition in the TGF- 1 -induced fold-increase in SMApromoter activation (Fig. 9 D ). Collectively, theseobservations suggest that ROK is neither sufficient nor absolutelyrequired for SMA expression, and a substantial part of theTGF- 1 effect on SMA transcription appears to be mediated by a Rho-dependent but ROK-independent mechanism.

Fig. 9. Differential involvement of Rho kinase (ROK) in theTGF- 1 -induced F-actin reorganization and SMA promoteractivation. A : cells were transfected with the Myc-taggedconstitutive active catalytic domain of Rho kinase (ROK-CAT, 1 µg).Later (3 days), the cells were fixed and doubly stained for F-actinusing Alexa(488)-phalloidin ( top ) and with ananti-Myc antibody ( bottom ) to identify ROK-CAT expressors. B : cells were cotransfected with p765-SMA-Luc (0.5 µg),pRL-TK (0.1 µg), and either 2 µg of empty vector (control) orROK-CAT. p765-SMA-Luc activity was determined 72 h later[control: 1 ± 0.03 vs. ROK-CAT: 1.49 ± 0.17 ( n = 6); P C :cells were transfected with a vector encoding the Myc-taggeddominant-negative form of ROK (DN-ROK, 2 µg), and either leftuntreated (control, top ) or exposed to 4 ng/mlTGF- 1 ( bottom ) for 3 days. After being fixed,the cells were doubly stained as in A. To clearly visualizethe differences in the F-actin structure of DN-ROK-expressing cells andtheir nonexpressing neighbors, the staining for ROK (red) and F-actin(green) is shown separately and also as merged images. Identicallyoriented arrows point to the same transfected cell. D : cellswere cotransfected with p765-SMA-Luc (0.5 µg), pRL-TK (0.1 µg), andeither empty vector (control) or the DN-ROK (2 µg) construct. Thecells were then treated with vehicle or 4 ng/ml TGF- 1 inthe absence or presence of 20 µM Y-27632, as indicated. After 24 h, luciferase activity was determined as above. Values were normalizedto the basal activity measured in vehicle-treated control cells. TheTGF- 1 -induced degree of activation(TGF- 1 /veh) in each group was as follows: control:2.95 ± 0.19; DN-ROK: 2.59 ± 0.12; Y-27632: 1.89 ± 0.05; control vs. DN-ROK is not significant; control vs. Y-27632, P2 u' S' t- H+ ], Y! O+ B4 @( V
; p- ~. |, u+ f$ e
The regulation of SMA expression is complex, since the promotercontains a variety of positive and negative regulatory sequences, therelative importance of which depends on the cellular context ( 25, 56 ). The following experiments were performed to identify theRho-dependent region of the SMA promoter in LLC-PK 1 cells and to investigate the relationship between Rho-inducible and TGF- 1 -responsive cis elements. We used a -galactosidase reporter system containing the first 155 bp of theSMA promoter (p155) and two truncations (p92 and p56) of this sequence.p155 has been shown to confer TGF- 1 responsiveness andprovide maximal transcriptional activity in various cell types( 25, 26, 56 ). It contains two CArG elements (B and A in 5' 3' direction, respectively) that are binding sites for the SRF, aTGF- 1 control element (TCE) whose trans factor has not yet been fully identified ( 1 ), and a TATAbox. The various constructs and the results obtained after theirtransfection are shown in Fig. 10 A. In LLC-PK 1 cells, TGF- 1 induced a 2.4 ± 0.07-fold( n = 6) increase in the reporter activity of the p155construct. Importantly, cotransfection with RhoAQ63L resulted in a5.16 ± 0.07-fold ( n = 6) activation of thepromoter. This level of stimulation is similar to that obtained usingthe 765-bp construct, suggesting that Rho responsiveness is confined tothe first 155-bp promoter region. The basal activity of p92, whichlacks the CArG B box, decreased by 50% compared with p155. Moreimportantly, the inducibility of this construct dramatically differedfrom p155; TGF- 1 caused only a slight rise in promoter activity (1.39 ± 0.05-fold, n = 6), whereas theeffect of RhoAQ63L was essentially abolished (1.25 ± 0.09-foldincrease). A further truncation involving the CArG A box (p56) resultedin complete loss of both basal activity and stimulation either byTGF- 1 or RhoAQ63L. These results unambiguously show thatthe CArG B box is required for the Rho inducibility of the SMApromoter. Furthermore, in agreement with earlier findings obtained withother cell types, the CArG box is also essential for theTGF- 1 responsiveness of the SMA promoter ( 25, 26 ).
5 @' v3 w& ~6 R7 q- ^1 |6 V$ ]6 Z
Fig. 10. CC(A/T)richGG (CArG) B box is essential for the Rhoinducibility of the SMA promoter in LLC-PK 1 cells. Cellswere transfected with 1 µg of the indicated SMA promoter construct(p155, p92, or p56) along with RhoAQ63L (2 µg) or empty vector pluspRL-TK (0.1 µg). Later (1 day), cells were exposed to vehicle or 10 ng/ml TGF- 1 for 24 h, as indicated. At the end ofthis period, the -galactosidase and renilla luciferase activities ofthe samples were determined as described in METHODS. Values were normalized tothe basal activity of p155 measured in vehicle-treated cells. The basalactivity of p92 was 54.11 ± 3.87% of p155, whereas it dropped to 0% in the case of p56 ( P 1 treatment orRhoAQ63L expression in each group compared with its own untreatedcontrol (veh) were as follows: p155, for TGF- 1 2.4 ± 0.07, for RhoAQ63L 5.16 ± 0.07; P 1 1.42 ± 0.09; P P = 0.04; p56, nondetectable. B : cellsexposed to vehicle or 4 ng/ml TGF- 1 for 72 h werelysed, and the lysates were probed by Western blotting using ananti-SRF antibody. C : EMSAs were performed on nuclearextracts from vehicle- and TGF- 1 -treated (72 h) cellsusing the CArG B box-specific probe, as described in METHODS. Where indicated, theextracts were incubated with anti-SRF-antibodies. Arrows on bottom show the TGF- 1 -induced binding of theprobe; arrow on top points to the supershifted band. In theabsence of anti-SRF antibody, the lower mobility band was not observed,even at longer exposure times.7 r7 s0 h; [- W0 M0 V, Y) M# p
$ X8 ~3 i# _: k$ o
CArG boxes are targets of SRF, a member of the MAD box transcriptionfactor family that is known to regulate SMA expression ( 35, 54 ). To assess whether changes in SRF content might contribute to TGF- 1 -induced, Rho-dependent SMA expression inLLC-PK 1 cells, we analyzed lysates from control andTGF- 1 -treated cells by Western blotting using ananti-SRF antibody. As shown in Fig. 10 B,TGF- 1 exposure (72 h) caused a marked increase in SRFcontent of the cells. The robust increase in cellular SRF maycontribute to the in situ activation of the SMA promoter. In support ofthis notion, TGF- 1 treatment resulted in enhancedbinding of the CArG B probe to nuclear extracts, and the addition of ananti-SRF antibody induced the appearance of a band with further reduced mobility.' I9 W7 \. }$ _0 b5 y
9 v  H- t. z+ `) L  i
DISCUSSION

To explore the signaling mechanisms involved intransdifferentiation of renal epithelial cells, we set up a cellculture model using LLC-PK 1 cells, a well-known and stableporcine proximal tubule cell line. We found that TGF- 1 exposure of LLC-PK 1 cells induces an authentic EMTcharacterized by tight and adherens junction disassembly, nucleartranslocation of -catenin, increased stress fiber formation, MLCphosphorylation, cortactin redistribution, and the expression of themyofibroblast marker SMA. These findings confirm and extend recentstudies that described various aspects of EMT in rat and human proximaltubule cells ( 19, 66, 67 ). Together these observationsimply that epithelial-mesenchymal transformation is a genuine responseof proximal tubule cells on chronic TGF- 1 exposure andsupport the intriguing concept that SMA-positive myofibroblast-likecells can derive from the kidney epithelium itself. Accordingly,TGF- 1 may promote renal fibrosis not only by stimulatingmyofibroblasts but also by inducing their formation from the tubularepithelium. Because myofibroblasts secrete TGF- 1 ( 46 ), this mechanism could create a positive feedback loopthat may contribute to the progressive nature of the fibrotic disease.9 k  ?: L: f: q" {0 s7 p
# q. b/ c1 O+ m* E
Having established the EMT model, we intended to identify key signalingevents underlying the cytoskeletal changes and particularly theTGF- 1 -induced SMA expression. We considered the smallGTPase Rho as a candidate to mediate the above TGF- 1 effects, since Rho 1 ) has a central role in stress fiberformation ( 24 ), 2 ) induces MLC phosphorylationvia both ROK-mediated direct phosphorylation and inhibition of MLCphosphatase ( 20 ), and 3 ) has recently beenimplicated in constitutive SMA expression by smooth muscle cells( 34 ). Consistent with this hypothesis, we found thatTGF- 1 causes a biphasic Rho stimulation inLLC-PK 1 cells: a rapid and transient peak, similar to thatobserved in mammary epithelial cells ( 8 12 h) elevation. While our manuscript was prepared forsubmission, a paper was published by Edlund et al. ( 15 )reporting similar biphasic Rho activation inTGF- 1 -stimulated prostate cancer cells. The biphasic Rhoresponse is consistent with and may explain the kinetics of otherTGF- 1 -induced signaling events, e.g., the bimodalactivation of c-Jun kinase, a process dependent on or potentiated byRho ( 6, 17 ). The mechanism(s) whereby TGF- 1 stimulates Rho remains to be clarified. The process may involve rapidactivation of guanine nucleotide exchange factors (GEFs) and/or theinhibition of GTPase-activating proteins (GAPs) or GDP dissociationinhibitors ( 60 ). Regarding the late phase, severaladditional mechanisms may be evoked. For example, TGF- 1 was found to stimulate the synthesis of NET1, a RhoA-specific GEF( 55 ), and to stabilize RhoB by inhibiting its degradation( 16 ). Importantly, cadherin disengagement has been shownto potently stimulate Rho activity ( 41 ). In our case, thismay be a crucial factor, since in LLC-PK 1 cellsTGF- 1 induced not only the delocalization of E-cadherin(as in mammary epithelial cells; see Ref. 8 ) but also adramatic decrease in the level of this protein.8 S6 \1 ~! V) B6 W

The functional significance of Rho activation is evidenced by the factthat inhibition of Rho prevented TGF- 1 -induced stress fiber formation and MLC phosphorylation. Similarly, DN-ROK abrogated F-actin reorganization upon TGF- 1 treatment. Theseresults indicate that the Rho/ROK pathway is indispensable forTGF- 1 -induced cytoskeleton remodeling. Presumably, anincrease in contractility is one of the major mechanisms whereby Rhopromotes EMT, since elevated cell tension itself has been shown tocontribute to contact remodeling ( 33 ) and fibroblastictransformation of epithelial cells ( 69 ).

The other crucial mechanism appears to be a Rho-dependent change ingene transcription. We found that active Rho strongly stimulated theSMA promoter in LLC-PK 1 cells, whereas two Rho inhibitoryconstructs prevented the TGF- 1 -induced promoteractivation. Interestingly, the dependence of theTGF- 1 -provoked SMA promoter activation on Rho and ROKmarkedly differed since ROK caused only marginal promoter activationand DN-ROK exerted only a slight inhibitory effect. In accordance withthis, pharmacological inhibition of ROK resulted in only partialreduction in the SMA response. Thus the TGF- 1 -inducedSMA activation is mediated by Rho-dependent but partiallyROK-independent mechanisms.4 L% M8 j  U: z% F9 B

Previous studies have shown that the first 125 bp of the SMA promoter(p125) drive maximal reporter expression in smooth muscle cells andprovide moderate basal activity in fibroblasts and endothelial cells( 25, 26, 56 ). This region contains two CC(A/T)richGG (CArGB and A) cis -acting elements and a recently described TCE upstream of a TATA box. The CArG motifs are binding sites for thetranscriptional activator SRF, whereas the TCE likely interacts withKruppel factor-like transactivators ( 1, 35 ).Interestingly, intact CArG boxes are essential for the highconstitutive expression of SMA in smooth muscle cells ( 26, 56 ), but their role is not critical for the basal expression infibroblasts and endothelial cells ( 25 ). In contrast, eachfunctional domain of p125 (or p155) was found to be required forTGF- 1 responsiveness ( 25, 26 ). Furtherupstream regions are responsible for suppressing expression innonmuscle cells and harbor muscle-type specific regulatory sequences.We found that, in epithelial cells, active Rho potently stimulated boththe longer (765-bp) and the minimal (p155) promoter constructs.Moreover, the level of activation was similar, suggesting that Rhotargets the p155 sequence and acts primarily by driving the minimalpromoter rather than removing a constraint acting on the upstreamsequences. Consistent with this notion, deletion of CArG B, whichreduced but did not eliminate basal transcription, entirely abolishedRho responsiveness. Importantly, Rho and TGF- 1 inducibility showed similar structural requirements, supporting theconcept that Rho is a key factor in mediating the effect ofTGF- 1 on SMA transcription. The fact that Rho acts through CArG boxes implicates SRF as the responsible transactivator. Wefound that TGF- 1 markedly increases SRF protein inLLC-PK 1 cells, and this effect is likely to contribute tothe dramatic rise in SMA synthesis. Our EMSAs support the involvementof SRF in the TGF- 1 -induced SMA response; however, theparticipation of additional cis -elements is also possible.

The mechanism whereby Rho activates or induces the expression of SRF isnot well understood. Rho has been suggested to stimulate the serumresponse element of the c- fos promoter and the constitutive expression of the SMA promoter through changes in actin organization ( 34, 57 ). It has been proposed that actin monomers inhibit the action of SRF, and Rho relieves this block by promoting actin polymerization. This mechanism may certainly contribute to the TGF- 1 -induced upregulation of the SMA promoter inLLC-PK 1 cells. However, additional factors are likely toparticipate, since inhibition of Rho or ROK had similar effects onTGF- 1 -induced cytoskeletal reorganization, but theRho-inhibitory constructs exerted a stronger inhibition on the SMApromoter than DN-ROK. Consistent with the notion of partialdissociation between cytoskeletal and transcriptional effects, variousRho effector loop mutants that fail to affect the cytoskeleton havebeen shown to induce strong activation of SRF-dependent transcription( 50 ). Moreover, we found that RhoN19, a dominant-negativemutant that disrupted stress fibers in LLC-PK 1 cells, didnot prevent the TGF- 1 -induced SMA response (data not shown). It must be emphasized that the mechanism of action of RhoN19 isdifferent from C3 transferase or p190RhoGAP, since the latter proteinsinactivate endogenous Rho itself, whereas RhoN19 competes foractivating factors and/or downstream effectors. Therefore, RhoN19 mayhave a differential capability to interfere with distinct Rho partnersand/or Rho isoforms. This notion is substantiated by the finding thatRhoN19 was unable to counteract SRF activation induced by certain Rhomutants ( 50 ) and had only a moderate inhibitory effect onthe phenylephrine-induced c-Fos serum response element activation inmyocytes ( 38 ). In contrast, RhoN19 potently inhibited the TGF- 1 -induced dissociation of E-cadherinin mammary epithelial cells ( 8 ). Considered together,these observations suggest that partially overlapping but distinctdownstream pathways are involved in the Rho-dependent cytoskeletalremodeling and SMA expression and that the cytoskeletal effects may beessential for cell contact remodeling but not for SMA induction.

Although Rho activation is a prerequisite for SMA synthesis, it may notbe sufficient in itself, since transfection of active Rho alone wasunable to induce significant increases in SMA immunoreactive protein inLLC-PK 1 cells (data not shown). Analogous observations weremade in fibroblasts, where injection of active Rho induced expressionof extrachromosomal SRF reporter genes but not chromosomal templates( 2 ). The most plausible explanation for this phenomenon isthe requirement for additional TGF- 1 inducible factors.TGF- 1 stimulates a multitude of signaling pathways anddrives gene expression via several transcriptional activators( 37, 39 ). Interestingly, TGF- 1 -inducedexpression of extra domain-A fibronectin was found to precede and berequired for SMA expression by myofibroblasts ( 52 ). Such requirement for additional factors mayalso explain why SMA protein expression is seen only after 2-3days of TGF- 1 treatment. Interestingly, our ongoingstudies suggest that -catenin signaling might also be necessary forefficient SMA expression.0 Y/ u9 _. K0 C* [9 D8 f: w0 ]3 R% a! b
, I% Q$ Q6 l# w, z. j
In summary, our work shows that Rho plays a critical role inTGF- 1 -induced cytoskeleton remodeling and SMA synthesisduring epithelial-mesenchymal/myofibroblast transdifferentiation.Future studies should define whether pharmacological interference with the Rho pathway might signify a therapeutically relevant approach tolessen organ fibrosis.; H. R+ U5 t  z- L9 J

ACKNOWLEDGEMENTS

We are indebted to Drs. K. Burridge, G. P. Downey, K. Kaibuchi, R. A. Nemenoff, and G. K. Owens for providingvarious constructs used in this study. We thank Dr. K. Szászi forvaluable discussions.
      【参考文献】) b% q' i/ @; @1 R, e$ u# ^0 y! {
1. Adam, PJ,Regan CP,Hautmann MB,andOwens GK. Positive- and negative-acting Kruppel-like transcription factors bind a transforming growth factor control element required for expression of the smooth muscle cell differentiation marker SM22alpha in vivo. J Biol Chem 275:37798-37806,2000 .
# ~! x" I0 ]1 F) E
$ D, Z# w) B$ B7 _. G
$ @/ s$ G5 t9 Y) I) M
2. Alberts, AS,Geneste O,andTreisman R. Activation of SRF-regulated chromosomal templates by Rho-family GTPases requires a signal that also induces H4 hyperacetylation. Cell 92:475-487,1998  .

5 j+ l$ d0 e  s; p/ A

3. Amano, M,Fukata Y,andKaibuchi K. Regulation and functions of Rho-associated kinase. Exp Cell Res 261:44-51,2000  .

5 [/ h: W2 x2 m, i. B' }
5 O5 J+ D, h& i8 T# ^  v6 F% H* R
4. Arthur, WT,andBurridge K. RhoA inactivation by p190RhoGAP regulates cell spreading and migration by promoting membrane protrusion and polarity. Mol Biol Cell 12:2711-2720,2001 .
4 m) j# ^4 V# S) v
) p# F8 h. p* f9 k" |

5. Arthur, WT,Petch LA,andBurridge K. Integrin engagement suppresses RhoA activity via a c-Src-dependent mechanism. Curr Biol 10:719-722,2000  .- U. v& c- _% T1 y
; L$ ~3 T: d  k" z/ j7 ^0 Z

6. Atfi, A,Djelloul S,Chastre E,Davis R,andGespach C. Evidence for a role of Rho-like GTPases and stress-activated protein kinase/c-Jun N-terminal kinase (SAPK/JNK) in transforming growth factor -mediated signaling. J Biol Chem 272:1429-1432,1997 .

4 c. T8 m" Z+ a. I- c: {, }0 V, x
7. Bakin, AV,Tomlinson AK,Bhowmick NA,Moses HL,andArteaga CL. Phosphatidylinositol 3-kinase function is required for transforming growth factor -mediated epithelial to mesenchymal transition and cell migration. J Biol Chem 275:36803-36810,2000 .

3 p8 Q4 l; d, K- [% ]: Q7 `

8. Bhowmick, NA,Ghiassi M,Bakin A,Aakre M,Lundquist CA,Engel ME,Arteaga CL,andMoses HL. Transforming growth factor- 1 mediates epithelial to mesenchymal transdifferentiation through a RhoA-dependent mechanism. Mol Biol Cell 12:27-36,2001 .

" X3 \$ N7 M: R0 a3 q# ?
: N, l9 P; E/ `7 p1 |1 f+ \
9. Border, WA,andNoble NA. TGF-beta in kidney fibrosis: a target for gene therapy. Kidney Int 51:1388-1396,1997  .' ~6 B' }: v, L3 U' o( A9 m1 ]; _
) c3 N) q6 \* A1 D/ i

5 \4 q3 Q9 {  a& X; B6 ]
10. Bottinger, EP,andBitzer M. TGF- signaling in renal disease. J Am Soc Nephrol 13:2600-2610,2002 .
+ _9 S) P1 ~* \5 y. `

4 F' @% s. F; ?3 m7 x$ f  w% C0 w
11. Boukhalfa, G,Desmouliere A,Rondeau E,Gabbiani G,andSraer JD. Relationship between alpha-smooth muscle actin expression and fibrotic changes in human kidney. Exp Nephrol 4:241-247,1996  .

5 M) W. r) |* ?0 F) L3 L

12. Burns, KD,andHarris RC. Signaling and growth responses of LLC-PK 1 /Cl4 cells transfected with the rabbit AT1 ANG II receptor. Am J Physiol Cell Physiol 268:C925-C935,1995 .# E/ X& ]- W3 V0 G7 p
) \2 T* w, t8 Q

13. Chihara, K,Amano M,Nakamura N,Yano T,Shibata M,Tokui T,Ichikawa H,Ikebe R,Ikebe M,andKaibuchi K. Cytoskeletal rearrangements and transcriptional activation of c-fos serum response element by Rho-kinase. J Biol Chem 272:25121-25127,1997 .

14. Eddy, AA. Molecular insights into renal interstitial fibrosis. J Am Soc Nephrol 7:2495-2508,1996 .2 }* U1 i4 i" V; O! V

15. Edlund, S,Landstrom M,Heldin CH,andAspenstrom P. Transforming growth factor- -induced mobilization of actin cytoskeleton requires signaling by small GTPases Cdc42 and RhoA. Mol Biol Cell 13:902-914,2002 .: S# x! D0 k; G/ s! N

16. Engel, ME,Datta PK,andMoses HL. RhoB is stabilized by transforming growth factor and antagonizes transcriptional activation. J Biol Chem 273:9921-9926,1998 .3 R% @- N9 p9 o4 h; `6 }: N

, V; L9 n3 l! A# ^! |% _+ o: \
/ g  U; f" t3 N* X+ X; g
17. Engel, ME,McDonnell MA,Law BK,andMoses HL. Interdependent SMAD and JNK signaling in transforming growth factor- -mediated transcription. J Biol Chem 274:37413-37420,1999 .* }2 j2 L% q8 V, F: A

18. Eyden, B. The myofibroblast: an assessment of controversial issues and a definition useful in diagnosis and research. Ultrastruct Pathol 25:39-50,2001  .- L: @( g  A" i, N- L- b2 P
8 V, m, }# r  _$ u% y
5 L- N# |  N2 Y8 l5 l  d$ P
' T# c$ B+ ?# O
19. Fan, JM,Ng YY,Hill PA,Nikolic-Paterson DJ,Mu W,Atkins RC,andLan HY. Transforming growth factor- regulates tubular epithelial-myofibroblast transdifferentiation in vitro. Kidney Int 56:1455-1467,1999  .
' _8 t& ~* b, ?

% i& ?4 \* @# w$ e
20. Fukata, Y,Amano M,andKaibuchi K. Rho-Rho-kinase pathway in smooth muscle contraction and cytoskeletal reorganization of non-muscle cells. Trends Pharmacol Sci 22:32-39,2001 .  M, D$ L: M- S$ d9 b# [, J3 \

21. Garat, C,Van Putten V,Refaat ZA,Dessev C,Han SY,andNemenoff RA. Induction of smooth muscle alpha-actin in vascular smooth muscle cells by arginine vasopressin is mediated by c-Jun amino-terminal kinases and p38 mitogen-activated protein kinase. J Biol Chem 275:22537-22543,2000 .- ^6 T# K8 l/ r; G

22. Goldberg, PL,MacNaughton DE,Clements RT,Minnear FL,andVincent PA. p38 MAPK activation by TGF- 1 increases MLC phosphorylation and endothelial monolayer permeability. Am J Physiol Lung Cell Mol Physiol 282

146-L154,2002 .4 b! [: E/ `8 ]

5 E+ b( S. }( N5 c  b
6 }7 x) K3 ~5 B* n
23. Grupp, C,Lottermoser J,Cohen DI,Begher M,Franz HE,andMuller GA. Transformation of rat inner medullary fibroblasts to myofibroblasts in vitro. Kidney Int 52:1279-1290,1997  .

5 `5 |$ |2 W! V3 B  @
, S. X7 y% r9 @
24. Hall, A. Rho GTPases and the actin cytoskeleton. Science 279:509-514,1998 ." W! m0 y/ R6 t+ j% g7 R
, V0 n* f9 @% U7 m2 l- f
0 @2 [+ V+ R9 [  D* Y9 S

25. Hautmann, MB,Adam PJ,andOwens GK. Similarities and differences in smooth muscle alpha-actin induction by TGF- in smooth muscle versus non-smooth muscle cells. Arterioscler Thromb Vasc Biol 19:2049-2058,1999 ." K: F+ B7 z2 v2 _8 A
8 z. M, r0 X, ]4 A
$ ~; p) q( w/ A" J' W$ Q; T
: T- @7 l  s/ P/ h7 A
26. Hautmann, MB,Madsen CS,andOwens GK. A transforming growth factor (TGF ) control element drives TGF -induced stimulation of smooth muscle alpha-actin gene expression in concert with two CArG elements. J Biol Chem 272:10948-10956,1997 .; e: I2 M, ~* |, v

27. Hill, CS,Wynne J,andTreisman R. The Rho family GTPases RhoA, Rac1, and CDC42Hs regulate transcriptional activation by SRF. Cell 81:1159-1170,1995  .* ]7 \1 N$ S& z9 H2 ]& l

28. Hull, RN,Cherry WR,andWeaver GW. The origin and characteristics of a pig kidney cell strain, LLC-PK. In Vitro Cell Dev 12:670-677,1976.8 t" x4 A* p4 t" _; V) _9 u4 {
8 s0 H( r& ?) o$ W; r  Y
5 @# v8 u9 \- [6 E' l- W, C6 B4 h2 v1 f

29. Hurst, VI,Goldberg PL,Minnear FL,Heimark RL,andVincent PA. Rearrangement of adherens junctions by transforming growth factor- 1 : role of contraction. Am J Physiol Lung Cell Mol Physiol 276

582-L595,1999 .
$ ^( l! R9 B! h3 G

30. Iwano, M,Plieth D,Danoff TM,Xue C,Okada H,andNeilson EG. Evidence that fibroblasts derive from epithelium during tissue fibrosis. J Clin Invest 110:341-350,2002  .- R3 O) ^5 l9 y. v+ p& r+ {0 M
2 }* k! B# D/ c" F6 f
7 |3 u& u% m' s1 c' \  @% \' I

31. Jinde, K,Nikolic-Paterson DJ,Huang XR,Sakai H,Kurokawa K,Atkins RC,andLan HY. Tubular phenotypic change in progressive tubulointerstitial fibrosis in human glomerulonephritis. Am J Kidney Dis 38:761-769,2001  .
* M- U  L9 s" `# z% ^9 Z
' z8 O2 m" y! C* B% n! ]
* c, Y1 g2 [; {; j8 j4 e& ?1 d
32. Kapus, A,Szaszi K,Sun J,Rizoli S,andRotstein OD. Cell shrinkage regulates Src kinases and induces tyrosine phosphorylation of cortactin, independent of the osmotic regulation of Na /H  exchangers. J Biol Chem 274:8093-8102,1999 .
# m9 m5 J5 {5 M5 _* G% I' R( U
  b, M6 G3 D9 _& y0 N3 ]* b

33. Krendel, M,Gloushankova NA,Bonder EM,Feder HH,Vasiliev JM,andGelfand IM. Myosin-dependent contractile activity of the actin cytoskeleton modulates the spatial organization of cell-cell contacts in cultured epitheliocytes. Proc Natl Acad Sci USA 96:9666-9670,1999 .

$ p3 N! @0 R* n9 ~( L! z. b* |9 w2 F
( E1 u0 R! a1 h9 e/ ~
34. Mack, CP,Somlyo AV,Hautmann M,Somlyo AP,andOwens GK. Smooth muscle differentiation marker gene expression is regulated by RhoA-mediated actin polymerization. J Biol Chem 276:341-347,2001 .
/ |: v  r5 @2 U: T8 N

35. Mack, CP,Thompson MM,Lawrenz-Smith S,andOwens GK. Smooth muscle alpha-actin CArG elements coordinate formation of a smooth muscle cell-selective, serum response factor-containing activation complex. Circ Res 86:221-232,2000 .1 g! @) @+ e) {1 w# a- O4 t+ z+ {: |

36. Marinissen, MJ,Chiariello M,andGutkind JS. Regulation of gene expression by the small GTPase Rho through the ERK6 (p38 gamma) MAP kinase pathway. Genes Dev 15:535-553,2001 .' f/ R' \# Z$ S* }8 m

; M. D/ m# n& x6 o
% x" t) R/ a$ ^7 A
37. Massague, J,andWotton D. Transcriptional control by the TGF- /Smad signaling system. EMBO J 19:1745-1754,2000  .

38. Morissette, MR,Sah VP,Glembotski CC,andBrown JH. The Rho effector, PKN, regulates ANF gene transcription in cardiomyocytes through a serum response element. Am J Physiol Heart Circ Physiol 278:H1769-H1774,2000 .
; \. ]4 W' F" v$ j6 Z. h4 {' l/ L
# C- H; s* J, Y( ?5 d' ~

39. Mucsi, I,Skorecki KL,andGoldberg HJ. Extracellular signal-regulated kinase and the small GTP-binding protein, Rac, contribute to the effects of transforming growth factor 1 on gene expression. J Biol Chem 271:16567-16572,1996 .4 I$ t9 s1 |' M
& E3 q: [" z7 Y: r% W" D

) n5 u% a% p" ~5 U
40. Ng, YY,Huang TP,Yang WC,Chen ZP,Yang AH,Mu W,Nikolic-Paterson DJ,Atkins RC,andLan HY. Tubular epithelial-myofibroblast transdifferentiation in progressive tubulointerstitial fibrosis in 5/6 nephrectomized rats. Kidney Int 54:864-876,1998  .& e# v+ z: H& r9 p3 o( E, I! A2 u
7 t& j/ ~- y- i! f4 \
0 ~3 [. h' f* A3 T8 T0 J/ ~9 m# S; m
6 ]" J  C# h! B: c* E7 d
41. Noren, NK,Niessen CM,Gumbiner BM,andBurridge K. Cadherin engagement regulates Rho family GTPases. J Biol Chem 276:33305-33308,2001 .

* E- |4 f3 d" j, N) h

42. Okada, H,Danoff TM,Kalluri R,andNeilson EG. Early role of Fsp1 in epithelial-mesenchymal transformation. Am J Physiol Renal Physiol 273:F563-F574,1997 .

43. Oshiro, N,Fukata Y,andKaibuchi K. Phosphorylation of moesin by rho-associated kinase (Rho-kinase) plays a crucial role in the formation of microvilli-like structures. J Biol Chem 273:34663-34666,1998 .

44. Pedagogos, E,Hewitson T,Fraser I,Nicholls K,andBecker G. Myofibroblasts and arteriolar sclerosis in human diabetic nephropathy. Am J Kidney Dis 29:912-918,1997  .; l+ @* l4 R% f' _
3 q# F* b/ ?' R! O2 b
6 V6 T2 H$ y& w9 r# R1 R+ Q- Z
$ K. _2 U) N& i; x5 M6 f
45. Petrov, VV,Fagard RH,andLijnen PJ. Stimulation of collagen production by transforming growth factor- 1 during differentiation of cardiac fibroblasts to myofibroblasts. Hypertension 39:258-263,2002 .% [  f2 R9 t6 K9 s

9 ~8 m- ]  Q! b5 Q  p% M+ t

46. Powell, DW,Mifflin RC,Valentich JD,Crowe SE,Saada JI,andWest AB. Myofibroblasts. I. Paracrine cells important in health and disease. Am J Physiol Cell Physiol 277:C1-C9,1999 .8 Z7 b( `/ e7 Y; ]5 [
8 U4 C9 I. }' Q' n. K4 c0 M

+ C5 G2 c7 Y3 ^
47. Rastaldi, MP,Ferrario F,Giardino L,Dell'Antonio G,Grillo C,Grillo P,Strutz F,Muller GA,Colasanti G,andD'Amico G. Epithelial-mesenchymal transition of tubular epithelial cells in human renal biopsies. Kidney Int 62:137-146,2002  .7 ^$ k4 N9 E' F' C) C8 ~! n8 f

" y" _: I$ x9 T
48. Remuzzi, G,Ruggenenti P,andBenigni A. Understanding the nature of renal disease progression. Kidney Int 51:2-15,1997  .

# D9 a( m- T$ T9 ~) g3 S# L
49. Ronnov-Jessen, L,Petersen OW,Koteliansky VE,andBissell MJ. The origin of the myofibroblasts in breast cancer. Recapitulation of tumor environment in culture unravels diversity and implicates converted fibroblasts and recruited smooth muscle cells. J Clin Invest 95:859-873,1995  .; A& K+ [9 E8 a0 A9 u

, T4 |& M% b, \, ]" ^
' q% i& a* J4 c
50. Sahai, E,Alberts AS,andTreisman R. RhoA effector mutants reveal distinct effector pathways for cytoskeletal reorganization, SRF activation and transformation. EMBO J 17:1350-1361,1998  ., F+ [# w9 A) s
$ R% D  m( o/ A3 A- z' K, j

2 h' K9 Q3 p+ V% n/ \; Q( Y- j) _
51. Savagner, P. Leaving the neighborhood: molecular mechanisms involved during epithelial-mesenchymal transition. Bioessays 23:912-923,2001  .7 \; p) m1 u0 w: U9 |: y
+ v) Z. \+ O9 _/ }. V* X) q. s/ n, h( ]+ J

52. Serini, G,Bochaton-Piallat ML,Ropraz P,Geinoz A,Borsi L,Zardi L,andGabbiani G. The fibronectin domain ED-A is crucial for myofibroblastic phenotype induction by transforming growth factor- 1. J Cell Biol 142:873-881,1998 .
9 B7 ^+ j( o; w$ [2 X# R- Y

5 W, v3 N* U' H: r. V
53. Serini, G,andGabbiani G. Mechanisms of myofibroblast activity and phenotypic modulation. Exp Cell Res 250:273-283,1999  .
) N  q" B- A! a9 z6 {  G3 V

: m8 r- |, }+ x. }  A0 ~- `! o
54. Sharrocks, AD. The ETS-domain transcription factor family. Nat Rev Mol Cell Biol 2:827-837,2001  .: C, t- D$ I( q* ^4 F

# ]6 G6 R- C' j, x
55. Shen, X,Li J,Hu PP,Waddell D,Zhang J,andWang XF. The activity of guanine exchange factor NET1 is essential for transforming growth factor- -mediated stress fiber formation. J Biol Chem 276:15362-15368,2001 .6 S. S) A+ c7 e  T6 w
( I/ t+ i2 u7 e- C

56. Shimizu, RT,Blank RS,Jervis R,Lawrenz-Smith SC,andOwens GK. The smooth muscle alpha-actin gene promoter is differentially regulated in smooth muscle versus non-smooth muscle cells. J Biol Chem 270:7631-7643,1995 .

  C5 Q* Y% u3 Y& q7 B$ W6 ]6 G
57. Sotiropoulos, A,Gineitis D,Copeland J,andTreisman R. Signal-regulated activation of serum response factor is mediated by changes in actin dynamics. Cell 98:159-169,1999  .
& ]9 u$ P+ Y& d. s, B2 ^
& l6 V- T/ N* u, z* H3 o0 K
5 P, T, }/ |7 W5 j" L  ~
58. Strutz, F,Okada H,Lo CW,Danoff T,Carone RL,Tomaszewski JE,andNeilson EG. Identification and characterization of a fibroblast marker: FSP1. J Cell Biol 130:393-405,1995 ." [2 ~& _7 M  [8 Z' }
+ ]8 I' d& Q, ~; }

59. Szaszi, K,Kurashima K,Kapus A,Paulsen A,Kaibuchi K,Grinstein S,andOrlowski J. RhoA and rho kinase regulate the epithelial Na /H  exchanger NHE3. Role of myosin light chain phosphorylation. J Biol Chem 275:28599-28606,2000 .7 c, f% C, N1 {4 L: c

; r  n" {5 ?0 F% |

60. Takai, Y,Sasaki T,andMatozaki T. Small GTP-binding proteins. Physiol Rev 81:153-208,2001 .3 }+ v4 b8 @3 @. f/ i1 J
4 z: z5 F1 f) L1 R4 ~0 N) w) l

61. Tang, WW,Van GY,andQi M. Myofibroblast and alpha 1 (III) collagen expression in experimental tubulointerstitial nephritis. Kidney Int 51:926-931,1997  .# j( o4 m) R- Q  d2 o* Z
7 o; Y$ E9 N3 K" d) G9 e
5 k* E, y. p# o; U4 b( ]' s

62. Tatsis, N,Lannigan DA,andMacara IG. The function of the p190 Rho GTPase-activating protein is controlled by its N-terminal GTP binding domain. J Biol Chem 273:34631-34638,1998 .4 |/ c" }+ ^. b/ S

, b! \1 d/ }9 `0 M" B5 r
63. Wang, J,Su M,Fan J,Seth A,andMcCulloch CA. Transcriptional regulation of a contractile gene by mechanical forces applied through integrins in osteoblasts. J Biol Chem 277:22889-22895,2002 .7 x& W$ A) c; O1 {7 i6 U; u

) K" B# Z% S5 C  f7 j
64. Weed, SA,andParsons JT. Cortactin: coupling membrane dynamics to cortical actin assembly. Oncogene 20:6418-6434,2001  .  T0 m( ^% q1 K$ |8 M/ s# |; ?5 ?3 L) s
9 \) \! r" Z- L! Q  v

65. Wilde, C,andAktories K. The Rho-ADP-ribosylating C3 exoenzyme from Clostridium botulinum and related C3-like transferases. Toxicon 39:1647-1660,2001 .

66. Yang, J,andLiu Y. Dissection of key events in tubular epithelial to myofibroblast transition and its implications in renal interstitial fibrosis. Am J Pathol 159:1465-1475,2001 ." j( V& o% _0 {) o, R/ j. I
3 x  E; g, [: g! p  M
4 Y2 H% z5 @! |

67. Yang, J,andLiu Y. Blockage of tubular epithelial to myofibroblast transition by hepatocyte growth factor prevents renal interstitial fibrosis. J Am Soc Nephrol 13:96-107,2002 .2 g5 Z. z; E8 b; T- _8 b" n% L

2 Q. e+ e3 R# r( V; |% B3 A
- s( A% Y, d( c8 D4 X% b5 B9 e) H
68. Zhang, G,Moorhead PJ,andel Nahas AM. Myofibroblasts and the progression of experimental glomerulonephritis. Exp Nephrol 3:308-318,1995  .2 F2 M# v0 K3 C5 H
" U, `% M9 f7 N- }3 I

/ H% }( C0 I# Q7 m, _6 n  \) ]+ \/ ^
69. Zhong, C,Kinch MS,andBurridge K. Rho-stimulated contractility contributes to the fibroblastic phenotype of Ras-transformed epithelial cells. Mol Biol Cell 8:2329-2344,1997 .

收藏0 分享0 顶0 踩0