Significant Correlation Between Connective Tissue Growth Factor Gene Expression and Skin Sclerosis in Tissue Sections from Patients with Systemic Sclerosis

Significant Correlation Between Connective Tissue • Growth Factor Gene Expression and Skin Sclerosis tn Tissue Sections from Patients with Systemic Sclerosis Atsuyuki Igar ashi, Kiyoko Nashiro, Kanako Kikuchi,* Shinicbi Sato, * Hironobu Hm, * Gary R. Grotendorst;! and Kazuhiko Takehara* Departm e nt of Dermato logy, Tokyo University Drancl1 Hospital; *Department o f Dermatology, Uni versity of Tokyo Faculty of Medicine, Tokyo, J apa o; an d i"Depa rtm enr of Cell Biology and Anato m y, U ni ve rsity of Mjami School of M edicin e, Miami, Florida, U.S .A.

The role of some growth factors and cytokines in the pathogenesis of systemic sclerosis (SSe) has been suggested. In particular, the contribution of transforming growth factor f3 in the progression of skin sclerosis is suspected. Connective tissue growth factor (CTGF) was originally identified in human umbilical vein endothelial cells, and a recent study has revealed that human skin fibroblasts produce CTGF after stimulation with transforming growth factor {3. In the present study, the distribution of CTGF gene expression in tissue sections from patients with SSe was investigated by digoxigenin-labeled itt situ hybridization. Strong CTGF mRNA signals were observed in the fibroblasts in sclerotic lesions, espe-

s

ystcmi c sclerosis (SSe) is a co nnective tissue di sease characterize d by excessive fibrosis in the skin and man y internal o rgan s. Although th e pathogenesis of SSe is poorly understood, it ha s been suggested that growth factors and cytokin es released fi·om inflammatory cells in th e o rgan in volved play an importa nt ro le in the initiation of connective ti ssue fibrosis. Amon g these, tran sforming growth factor (3 (TGP-(3) has recently ga ined atte nti o n as an important factor in the p athoge nesis o f SSe [1 ]. TGF-(3 is secreted from macrophages and lymphocytes [2 ,3] and in creases the production of coll age n and fibronectin by fibroblasts in rdtro [4,5]. TGF-(3 is also known to h ave indirect mitogeni c activity o n fibroblasts, and this mitoge ni c activity appears to be de pend en t on the auto crine production of platelet-derived growth factor (PDGF)-related peptides [6- 8). ln clinical investigation s, several laboratories have reported the occ urrence of elevated expressio n of TGF-(3 in skin biopsies from patien ts with SSe by immun o histoche mi ca l staining or ;, sil11 h ybridization [9-14 ]. In o th er sclerotic s.kin disorders including m o rph ea, eos inophilic fasciitis, and keloids, expressio n of Manuscript rece ive d December 17, 1994 ; final revis io n rece ived May ·t 0, 1995; accepted for publication May 16 , 1995. R-eprint requests to: Atsuyuki lgaras lti , Department of Dermatology, Kanto Tc islun H ospital, 5-9- 22 Higas hi gota nda, Shin agawa-k u , Tokyo, 141,Japan . Abbreviations: CTGF, conn ective tissue growth factor; SSe, systemic sclerosis.

cially in the deep dermis, of the skin specimens from patients with SSe, whereas there was no expression in the skin from normal controls. The number of fibroblasts with positive hybridization signals was more abundant in the dermis from the sclerotic stage than in that from the inflammatory stage. Our findings indicate a correlation between CTGF gene expression and skin sclerosis and support the hypothesis that transfortning growth factor-{3 plays an important role in the pathogenesis of SSe, because transforming growth factor f3 is the only inducer for CTGF identified to date. Key words: TGF-f3/fiiJYoblasts/fibrosisli11 sit11 ltybridiz atioll. J Invest Dermatol 105:280-284, 1995

TGF-{3 ha s also been demonstrated [14-1 6]. Thus, TGF-(3 is suspected to pla y an important rol e in the pathogen esis of fibro sis. Co nnecti ve tissue growth factor (CTGF) is a cysteine-rich pe ptide originally identifi ed from hum an umbilical endothelial cells (HUVE) that exhibit PDGF-like ch emo tactic and mito ge ni c activities fo r m esen chym al ce Us and appea rs to be antigenically related to PDGF, although it has little peptide seq uen ce homology to either the PDGF A orB chain pe ptides [17] . Our recent study show ed that human for eskin fibroblasts produ ce hi gh levels of CTGF mRNA and pro te in after activation with TGF-(3 but not other growth factors including PDGF, e pide rm al g rowth factor, or basic fibroblast growth factor [1 8]. In th e wound-chamber model, coordinate TGF-{3 expression followed by CTGF in regenerating tissu e h as been o bserved , suggesting a cascade p1·ocess for control of tissue regeneration and re pair [1 8]. Thus, CTGF is a ca ndidate autocrine stim ulator released in response to TGF- (3 in skin fibrobla sts and appears to part icipate in the process of ski11 fi brosis. However, the co ntribution of CTGF to the disease process of scleroti c skin disorders has not been investiga ted previously. To determine the relationship between CTGF and th e disease process of SSe, we inves tigate d the express ion ofCTGF mRNA in biopsy specimens from SSe by i11 sit11 hybridi zation. Marked CTGF ge ne express ion was observed in the fibroblasts located in the sclerotic lesio n from patients with SSe, but not in sections from patients with SSe that showed no hi stologic sclerosis or from normal co ntro ls. Our findings indicated a signifi cant correlati on

0022-202X/95/S09.50 • SSD I0022- 202X(95)0025 1-F • Copyrigh t © 1995 by The Soc iety for Inv estigative Dermatology, Inc.

280

VOL. 105, NO . 2

AUGUST 1995

CTG F GENE EXPRESS ION IN SYSTEMI C SC LEROSIS

281

between CTGF gene expression and skin sclerosis, providing further evidence that TGF{3 is an important factor in the pathogenesis of fibrosis, as this factor is the onJy indu ce r of CTGF gene expression identified to date. MATERIALS AND METHODS Cell Culture Skin fib roblasts derived from normal co ntrol subjects were grown from expl ants of biopsies. Cell s were cultured in Dulbccco's modified Eagle's medium supplemented with 1 O'X, fetal bovin e serum. In c ub ation was maintain ed in T25 fla sks at 37°C in 95% air and 5% C0 2 . The cells were the n ex planted into multiweiJ chamber slides . After reaching co nfluen cy, cell s were treated with 10 ng/ml ofTGF-{31 (R & D Systems, Inc., Minneapolis, MN) for 4 h, rinsed with phosphate-buffered sa lin e (PBS) fixed with 4% paraformald eh yde in PBS , air-dried , and subj ected to i11 siw hybridization. Cells were used for experiments at the second or third passage.

Patients Besides fi ve nonnal skin contro ls, biop sies ·w ere take n fron1 15 patients with SSe (14 wome n , mean age 45, ran ge 14- 61 ) , one patient with nll.xed conn ective tissue disease, and on e with prin1ary Ra ynaud's dise ase . All p atients had Raynaud' s phenomenon as an initial sign of the di sease . SSe patients fulfilled the preliminao·y criteria for SSe propo sed by the American Rheumatism Association (ARA) [19). According to the classifi ca tion of LeRoy et nl [20) , 11 patients had diffuse cutaneous SSe and four pati ents had lim:ited cutan eo us SS e . All biopsy specimens were obtaiu ed from the dorsum o f the forearm . All sa mples were fixed in neutral buffered formalin , e mbedded in paraffin, and prepared for hemato>:y lin-eosin exam ination . On hematoxylin-eosin examination, 12 of 15 patients with SSe d emon strated histologic sclerosis in the dermis. Of th ese patients, three patients (three diffUse cutaneous SSe, mean age 42 , range 30-49; mean disease duration 4 years, range 1-10 years) showed clini ca l and histologic signs of the inflammatory stage of SSe, with a marked perivascular and periappendagea l Jytnphocyti c iufiltration in th e d ermis. Nine patients (two limi ted cutan eo us > SSe, seven dilfuse c utan eo us SSe, mean age 43, range 14-60; mean disease duration 4 years; range 4 months to 16 ye ars) re presented the scl ero ti c stage of SSe. T he dermal co llagen bundles were packed and thickened , and the denno-epidermal junctions were flattened . Dermal sclerosis extended into the subjacent adipose tissue. Three patients (two limi ted cutaneou s SSe. one di.ffi.ose cu taneous SSe, m ean age 57, range 53-61; m ean di sease duration 12 years, range 8 -18 years) showed no hi stologic sclerosis, re presenting the atrophic stage . T he mixed co nn ective tissue disease and primary R.aynaud' s ' disease biopsies showed a low level of perivascular lymphoid ce ll infil tration but no derma l sclerosis.

In Situ Hyln·idizatio rt Fonnalin-fi xed, paraffin-embedded skin tissues were inves tigated using a sli g ht modifi cati o n of the nonradioa ctive i11 situ hybridization tec hnique with digoxi.gen.in-labeled !t.N A probes . Briefl y, paraffin-embedded section s w e re c ut at a thickness of 4 p.m. mounted on silane-coated slides. deparaffini zed , and treated with 0.2 M H C l for 15 min fo llowed by 1.5 mg/ ml proteinase K digestion for 15 min at 37°C . T he sections w ere post-fixed with 4% paraformaldehydc in PBS for 5 min and , treated with PBS co ntaining 2 mg/ ml g lycine twice for 15 min each time. After rinsing with PBS , the samples were soa ked in two times standard saline citrate buffer (SSC) with 50% formamide and subjected to hybridization. A 2100-bp fragment from CTG F eDNA was subcloned into the EcoR.l site ofB lu escript pbagemid [17) and used for making probes. The DNA was linea rized using the restriction enzymes Xba ! (to prepaTe antisense strand) or Xhol (to pre pare sense strand). The probes were labeled with digoxi gcnin-11-UTP using a DI G RNA-labeling kit (Boehrin ger Mannhcim Bio-

A Figure 1. Iu situ hybridization of cultured huntan forearm skin fibroblasts from a normal control subject probed with CTGF antisense probe. A: nontreated. B: treated with 10 ng/ml ofTGF- {3 for 4 h. Cultured fibrobla sts were grown until confluency and trea ted with TGF-{3 as described. The cell s were fixed with 4% para fo rmalde hyd e in PBS and digoxigenin-label ed nonradioactiv e i11 si111 hybridization with CTGF antisense strand probe was performed as describ ed in th e text.

ization method emp loyed was effective. CuJtured human skin fibrobla sts are known to produ ce CTGF mRNA b y Northe m blot analysis after stimulation with TGF-{3 [18] . When confluent .fibroblasts fi-mn normal controls were h ybridized with antisense probe, only a tra ce level of staining was observed in non treated fibrobl asts (Fig lA), whereas a strong signal was seen in TGF-{3-trea te d fibroblasts (Fig lB) . Tl:tis r e sult indicates that this nonradioa c tiv e in situ hybridization me t hod effectively detec ts CTGF m.IlNA. No CTGF mRNA Expression in Normal Skin Fibroblasts Next, we investigate d CTGF gene express ion using biopsy specim e ns. As shown in Fig 2 , no CTGF mRNA was e xpre ssed in d e rmal fibroblasts in aJJ fiv e normal controls. The e ndothelial cells of the blood vessels w e re a lso n egativ e. A slight nonsp ecifi c backgrow1d staining w as observed in th e e pidermal k e ratino cyt es with both sense and antisense probes . CTGF mRNA Expression in the Fibroblasts in the Sclerotic Lesions ofSSc Specintens Wh e n tissu es fro m patie nts with SSe w e re examined b y i11 sit11 hybridization w ith the antisense strand CTGF probe, d e rmal fibroblasts were positiv e in aJI 12 cases that showed histologic sclerosis. Figure 3A is the hem atox y lin-eosin stained section from the sclerotic stage of SSe, showing th e

chemica , Gernlall)')-

The labeled RNA probe (final 1 p.g/ml) in a mixture containing 50% fo rrn amide, 10% dextran sulfnte, 1 X D enhardt's solu tion, 100 p.g/ ml tRN A. 5 X SSC, 0.25% sodium dodecyl sulf.,te, l mM eth ylenediaminctetraacetic acid and 50 mM NaH 2 P0.1 was pla ced o n the slid es and covered wi d~ a coverslip . Hybridization was performed in a humidified chamber for 18 hat 45 °C, after whi ch the specimens w ere washed in 2 X SS C with 50% fo rrnamide at 50°C. Unhybridized probes were digested in 2.5 p.g/ ml RNasc A, 500 mM NaCl, ·t mM e thylcnediaminotctraacetic acid . and 10 mM Tris HCI, pH 8. 0, for 15 min at 37°C. The slides were then wnshed for 15 mi.u in 2 X SSC and in 0.2 X SSC nvi ce at 50°C. After posthybridization washing, digoxi.geu.in-l:obeled probes w ere visuali zed ns described in the DIG nucleic ncid d etectio n kit pro toco l (Boehringe r Mannhe im Biochemica). T he sections were poststained with nuclcnr f., st red aud mounted. RESULTS Detection of CTGF mRNA in TGF-/3-Treated Cultured F ibroblasts We tried to d etermine whether the iu si/ 11 hybrid-

·A Figure 2. No CTGF mRNA expression in the normal skin. A: antisense strand probe . B: sense strand probe. Digo >.; genin-labeled i11 si111 hybridi zation of normal skin contro ls with CTG F gene probes w as preformed as described in the text. Bnr, 100 p.m.

282

IGARASHI E T AL

THE JOURNAL OF INVESTI GATIV E

.....

-

DEI~ATOLOGY

' -•. -'f

~~- -

- ... . t

~ .,

;

. •

~

.. ..

.' :...,

..

1.

""';'

•

· .... .-:;...

..,

··

·. . :

-

Figure 3. CTGF mRNA ex pression in the fibroblasts from patients with SSe in the sclerotic stage. Th e ti ssue from pati e nts with SSe in the scle roti c stage was h ybridi zed w ith C T GF antisense strand pro be (B, C, D) and sen se stra nd probe (E). A , sa m e lesion as B. is th e he matoxy lin-eosin stain ed sectio n . C and E arc hi g h m agnifi cation of the lower dermis. A rniii'S in D po int to va sc ular e ndothelial ce ll s nega tive for C TGF mRNA. Bnrs : B , 2 50 f..lm ; C and E, 1 00 p.,m ; D. 25 p.,m.

appare nt scle rotic chan ges of collagen bundl es in the whol e de rmi s. As sh own in Fig 3B, th e fibrobla sts disp laying CTGF mRNA sca ttered throughout th e dermis. Figure 3 C demonstrates that CTGF mR.NA-positive fibrob lasts w e re more prominent in the lower dermis. Figure 3D show s the hi gh mag nifi ca tion of fibro blasts expressing CTGF mRNA, whereas dennal vascu lar endoth elial cell s were n egative for CTGF mRNA. To co nfirm the specifi city of th e hybridi zation , in sit11 hybridi zatio n with CTGF sense strand probe was performed and no positive signal s were obtain ed (Fig 3E) .

DISCUSSION R ecent studies ha ve revea led the involvement of TGF-!3 in the progression of SSe [9-14]. We became inte rested in the CTGF gene expression in SSe skin tissues because CTGF production is

Fibroblasts with Positive Signals Are More Abundant in the Tissue from the Sclerotic Stage than That frott.l the Inflammatory Stage T issue from patients with SSe in the inflammatory stage also conta·in ed the fibroblasts positive for CTGF mRNA in the dermis , altho ugh the number of positive cell s was less th an th at in th e scl erotic stage. We o bserved no preferential distribution of CTGF mR.NA-positive fibrobbsts in the perivascular region s o r infl ammatory les io n s in any of the samples exa min ed (Fig 4). No CTGF mRNA Expression in the Tissue from the Atrophic Stage of SSe Figure 5A shows the hematoxylin-eosin stained sectio n of the biopsy speci m en in the atrophic stage of SSe that has no histolo gic scle rosis. As shown in Fig 5B, CTGF mR.NA express io n was n ot observed in that les ion. A lso, th e skin from the patients w ith M CTD and primary Raynaud 's di sease showed no expression of CTGF mRNA (data not shown) . T hese finding s su gges t that derma l sclerosis and CTGF ge ne expression arc significa ntl y co rrela ted.

Figure 4. CTGF mR.NA expression in the tissue section from patients with SSe in the inflammatory stage. The section w as h ybrid· ized with C TGF antisen se strand prob e. Arrows indi cate the pe riappcnd agcal lymph ocytic infiltntti o n. Note that the numb e r o f fibrob la sts with positive sig nal s is less than that in Fig 3C. Bnr, 100 p.,m .

VOL.

I 0 5, NO. 2

AUG U ST

I ~~5

F igure 5. No CTGF mRNA expression in the fibroblasts from patients with SSe in the atrophic stage . A: hematoxylin-eosin stained section shows the loose arrangement of collagen bundles. B: the adj acent section was hybridized wi th CTGF antisense strand probe. Bnr, 100 J.Lll1 .

u pregulated in cultured hum an skin fibrobla sts aft er stimul ation with TGF-/3 [1 8 ). CTGF is a secreted protein originall y identifi ed from human umbili cal endo th elial cells that exhibits PDGF-like ac tiv ity [17]. It ha s h o mology to th e 11-src-i.ndu ced CEF-1 0 pe ptide i d e ntifi ed in chi ck en embryo fibrobla sts [21). cyr6"/ trans cript clon ed fr o m mouse BALB/ c 3T 3 fibrobla sts after se rum indu cti o n [22), and th e se rum-indu ced.fisp1 2 [23 ] o r TG F-/3-induced /3IC M 2 trans cript iden tif ted in 3T3 cells [2 4]. CTGF is also relate d to th e noll ge n e pro duct that is overe xpressed in ne phroblastomas indu ced by m yelo blasto sis-associated virus [25], ye t th e biologic ro les o f t h ese CTGF-related ge ne produ cts arc poo rl y und ersto od . Our prese n t findin gs clearl y demon strated C T GF mRNA expression iJJ skin fibro bl as ts located in the scle roti c les io ns fi·om pa tie nts with SSe, whereas th ere was n o CTGF ge ne ex pressio n in sec tions tha t sh o w ed no histolog ic scl e rosis. These res ults indicate the existen ce o f a sig nifi ca nt correlatio n be tw ee n skin fib rosis and CTGF gen e expression, supportin g the hypo thes is that TGF-/3 pla ys an impo rtant ro le in the pathogen es is of SSe beca us e CTGF is in duced by T G F-/3 . C TGF mRNA-po sitive fibro blas ts w ere scatte red in scleroti c lesions and w ere pre fere nt ial.l y acc umul ated in th e d eep dermi s. Also, CTGF gen e expressio n wa s more pro min en t in t he dermi s of patients in the fibroti c stage than ti·om th ose in th e inflammatory stage. C onsidering th ese o bse rvations with th e previ o us report of increased collagen synthesis obse rved mainly in fibro bl as ts gro wn fr om the de e p de rmi s of patients wi th SSe 126] , the level o f CTGF m.R.NA express ion see m s to correlate with th e acti vity o f fibroblasts. It is une xp ec ted, how e ver, that CTG F tran sc rip ts were present in cases of SSe with long duration eve n g reater th an 10 years. This o b serv ation m ay indicate that the activity o f fibrobla sts in th e scleroti c lesio n o f SSe persists in the lo ng term, althou g h T GF-/3 see m s to ac t in th e ve ry ea rli es t stage of th e disease [9 ,1 0, 14]. W e sp eculate tha t unknown fa ctor(s) i11 ·11i11o may fun cti o n to m aintain th e CTGF gen e exp ression. A ltern atively, as th e he terogeneity o f dermal fibro blasts has bee n suggested [27) , ce rtain clon es may achieve pe rmanent or pe rsistent alterati o n in biosy nth e ti c activ ity and elude n o rm al TGF-/3 regul ati o n. Wh en dermal fibrobl as ts fro m pa ti.e nts with SSe w ere cultured i11 11irro , w e recentl y found that th e ex pression of C TGF in SSe fibroblas ts wa s greater th an that in n orma] fi bro blasts in the prese n ce of TGF-/3 both in immunobl o tting and re verse- tran sc riptase-pol ym erase ch ain reaction analys is , indicating th e in creased response of SSe fibroblast s to T GF-/3 [28] . It is cl ea r that SSe fibroblasts exp ress th e C TGF gene differentl y fro m norm al fibroblasts, ye t furth er inves tigati o n is n eed ed to clarify th e m echanism of expressi o n . D ermal blood vessel end o the lial cells did n o t rev eal CTGF

C T G I' GE N E EXPRESS IO N

IN SY ST EM IC SC LER O S IS

283

mKNA iJl tissue sectio n s, altho u g h CTGF is pro du ced by c ul tured hum an umbili cal en do th eli al ce ll s. It rem ain s un clear wh ether this di scre pan cy is du e to th e d iJfercnt ce ll types o r to th e diffe re nces between i11 l' itro and i11 l! il'o co nditi o n s. CTGF ge n e ex pressio n i.n inflammato t·y o r tum01·al diseases of bl ood vessels is n o w und er inves ti gatio n in o ur labo rato cy. T he regulati o n of CTGF ge ne an d pro a l co llage n ge ne expressio n appea rs to be simi lar, be cause bot h ge n es arc sti m ul ated by TG F-/3 and exp ressed in SSe skin t iss ues . Y et it re m ains unkn own w he ther the ex pressio n of th ese tw o genes is syn chro ni zed in respo n se to o th e r sti mttli . TG I0 -/3 is kn own to activate th e p ro a 1 coll agen ge ne via the nucl ear facto r 1 site in irs p ro m oter regio n [29] . H owever, th e existen ce o f TG F-/3-re spo n siv e elem ents in the promote r regio n o f the CTGF gen e rem ain s uncl ea r. In addi t io n , it wil.l be of in teres t to in ves tig ate the interactio n between co llagen and CTGF in th eir expressio n . CTGF is a n ovel m embe r of the P DGF-re lated gr owth fa ctor famil y that is selecti ve ly induced by TGF- /3 in hum an skin fi brobla sts and is suspec ted to pl ay an impo rtan t rol e dut·in g wo un d hea lin g and / or tiss u e regen eratio n . W e dem onstrated elevated C T G F gen e express io n in tb e skiJ1 fi·o m patie n ts w it h SSe, suggesting a path ogenic co ntribu tio n to fibrotic co nn ecti ve d iso rde rs in whi ch CTGF m ay act as an auto crine fa cto r in fi b rob las ts afte r stimul ati o n w ith TGF- /3. CTGF ge ne exp ress io n in othe r sclero ti c ski n di so rd ers such as k elo id and scar tiss u es is now und er exaJllinatio n , becau se th e bio logic role of CTGF is not yet full y und erstood.

R EFEREN CES I.

2.

3.

·1.

5.

6.

7.

8.

9.

10.

II .

12 .

13.

Lc fl o y EC . Edwin A. Sm ith M . Knhalc h 13, Tn ~ j:l ll O \\' S k a M . Si.lvcr RM : A stra te gy fo r de termini ng the p;tth o gc ncsis of sys rc mk scle rosis: is rransfo n nin g g rowth r:1 c to r {:3 d 1c answe r? A rtluiti.l· Rlli'/1111 32 :8 17- 825, 1989 Assoi;111 R.K . Flc u rdci ys BE. S1cvc n so n H C , M ilkr PJ . M ad rcs 01<. Rai n es EW . !"to ss R . ·-I o n1 MB : Ex pn.:ssio n an d sc ~ r e t io n o f rypc (3 rransfOrmin g gr ow th f.1 c to r h y :1 ctivarcd hmn:m ma c ro ph :~ g c s. P ml· t\.'nrl .Acud Sri USA R4 :6020G024. 1987 Kc hr!JH . Wakt, ficl d LM. l ~o b c rt.-; A B,Jakowkw S. A lvarcz-M o n M. Dc ryn c k R . Sporn MB . Fa uci AS : Prod lll:tiu n o f" tra nsfo nn ing gro wth f.1c to r {3 b y h uman T lym p ho cy tes a nd ir~ p o te n t ia l 1·n lc in th e rcg u lario n ofT ce ll g ro wth . ) Exp Ml'r/ 163: 1037- IOSU, 198 6 Varg a J. R ose nbl oo m J, Jin ll!l ll!Z SA : T ran sfo r111 ing g rowt h fit c ro r {3 (TG F/3) c:tu ~ e1" a p ·r ·iswnr in cre ase in ste a d y-s tate a1110 un ts fo r t) •pc I a nd t-yp e III coll;tgcll and fi hro Jt c c t·in m H. NA s in tt on tt a i!Junt an d cn n a l fi b roblast's. 13it~rlll'tJJ j 2 97 :597-GfH , 1987 R :tg h o w R . Pos rl ctl l\v~litc AE. J
284

IGARASH I ET AL

H igley 1-l , Pcrsich itte K, C hu S. Waegcll ·w. Van cheeswaro n R, Black C: llnmunocytoch cm ical loca lizatio n an d serologiC detectio n of transfonn ing growth f.1cror {31. Art/iritis Rlteum 37 :278-288. 1994 IS. PcltonenJ. Vaq;aJ, Sollbcrg S, U itroj , J imenez SA: Elevated expression of the genes for transforming b'TOwth f.1cto r-l3 l and type VI coHage n in d iffuse f.1sciitis assoc iated with the cosinophiliot-m yalgia syndrome.) l1 111est Derm atol96:20-25, 1991 16. Pelto n en J , Hsiao LL. Jaakkola S, Soli bergS. Aum ai lley M. Ti m p l R. C hu M-L. U itto J: Activatio n of collage n gene ex pression in kc loids: co- localization of type I and V I collage n tra nsform ing growth f.1ctor-{3 1 m R.NA.j .l111'est Dem111t11 f 97:240-248 . 1991 17. Bradham DM , lg:tras hi A. Potter R.L, G rotcndorst Glt: Coru1cctivc tiss ue growth fa ctor: a cystei ne-ri ch m itogen secreted hy hu ma n vascul ar cn dothc )jal ce lls is related to th e Sl:l....C ind uced immcd i :~t~! earl y ge ne product CEF-·t O.j Cell Bioi 11 4 :1 285- '1294, 1991 18. Igaras hi A. Okoch i H , Bradham DM, Grotcndorst GR: Regul atio n of connective tissue growth r.,ctor gene expression in hu man skin fibroblasts and during wound repair. Mol Bioi Cell 4 :637-645, 1993 19. American lt.hcumatism Association: Su bcommittee for Sclcrodcnna Criteria of the American lt.heumatism Association Diagnostic omd T herapeutic Criteria Comm.ittcc: preliminary cri teria for th e class i.fication of systemic sclerosis (scleroderma). Art/iritis Rltw111 23:581-590, 1980 2 0. Le l\.oy EC, lllack C. Fleischmajcr R., J ab lonska S. Krieg T, Medsgcr TA J r, Rowe ll N, Wollhcim F: Scleroderma (~)'stemi c sclerosis): classification, su bsets and pathogenesis. J R/icu wn tol 15:202-205. 1988

14.

THE J OU RNAL OF INVESTIGAT IVE DEllMATOLOGY

21 .

22 . 23.

24. 25.

26. 27.

28.

29.

Sim m o ns DL, Levy DB, Ya nn o ni Y, Erikson IlL: Identification of a ph orbol es te r-repressible v-src-inducib lc ge ne. Jlmc Nntl Acml Sci USA 86: t ·t 7H-1182 1989 • O'Brien TP, Yang GP. Sanders L, Lo u LF: Expression of cyr61, a growd1 f.1ctor-ind uciblc immediate early gene. Mo l Cell /Jiol 10:3569-3577. 1990 R)•scck ll...-P, MacDo nald-Bravo H. Mattei M-G . Brnvo R.: Structure. tnappi.n g and cxpressiou offisp- 12. a growth f.,ctor- in d uciblc gene encoding a cystein erich protei n. Cdl Crewtlt Diffi!!· 2:225-233 . 199 1 Brunn er A. C hi nn J, Ne u baue r M, Purc hio AF: Ide n tification of a gene familv regul ated by transforming growt h factor (3. DNA Cell /J ili/1 0:293-300, 199l Jo lio t V , Martin erie C, Da mb rinc G, Plass iart G. Brisac M, C rochet j. Perbal B· Provirnl rearrange m ents ;m d ·ovcrcxprcssion of a new cc l.lu.l ar gene (n ov) iJ~ ll lye lo bl :lstosis-associated virus type ·t- in d uced ncphro bl astomas. 1\r/ol Cell Bioi 12:10-2 1. 1992 Kr ieg T, Pcrlish J S, Flcischmajer R : Collagen synthesis by scleroderm a fibroblasts. 11 11 11 N l' r1wrl Sci 460:375-386, '1986 Krieg T. Perl ishj S, Flcischmajer R . Timpl R: Collage n synthesis in sclcrodern1 a: selection of fib roblast po pulntions during su bc ul tures. Arch Demwto/ Res 277:373-378, 1985 IGk uchi K. Kado no T, lim H. Sato S. Igarashi A. Nakagawa 1-l, Tamaki K Takc hara K: T he growth regu lation of sclcrodcm1a fibroblnsts: in creased res pons~ to transform ing growth f:1ctor {3 1. ) lutJt'St Demwtc>/1 05: 128-"132, 1995 Rossi P. Karscnry G. Robe rts All. Roc he NS . Sporn MB, de Crom bruggh c B: A nuclear f.1cto r 1 binding site m ediates the tr:m scriptional activ:1tion of type 1 collage n pro moter by transforming g rowth facto r-beta. Cd/ 52 :405-414, "1988