Stimulation of Fibroblast Cell Growth, Matrix Production, and Granulation Tissue Fortnation by Connective Tissue Growth Factor
Ken Frazier, Shawn Williams, Devashish Kothapallj, Helene Klapper, and Gary R . Gwtendorst Department of Cell Biology and Anatomy, University of Miami School of Medicine, Miami. Florida. U .S.A.
Connective tissue growth factor (CTGF) is a 36- to pulse labeling of cells with [35S]methionine. Subcu 38-kDa peptide that is selectively induced by trans taneous injection of TGF-J3 and CTGF into neonatal forming growth factor-J3 (TGF-J3) in fibroblastic cell NIH Swiss mice resulted in a large stimulation of types. We compared the biologic activities of CTGF granulation tissue and fibrosis at the site of injection. with TGF-J3 on fibroblasts in culture and in animal In sitll hybridization studies revealed that TGF-J3 models of fibroplasia. CTGF was active as a mitogen injection induced high levels of CTGF mRNA in the in monolayer cultures of normal rat kidney fibro dermal fibroblasts at the injection site, demonstrat blasts. CTGF did not stimulate anchorage-indepen ing that TGF-{3 can induce the expression of CTGF in dent growth ofNRK fibroblasts, however, or inhibit connective tissue cells ill vivo. No CTGF transcripts the growth of mink lung epithelial cells, distinguish were detected in the epidermal cells in either control ing CTGF's growth-regulatory activities from those or TGF-J3-injected skin or in fibroblasts in control ofTGF-J3. In NRK fibroblasts, both TGF-J3 and CTGF (saline-injected) skin. These results demonstrate that, significantly increased the transcripts encoding at like TGF-J3, CTGF can induce connective tissue cell type I collagen, as integrin, and fibronectm. Stimu proliferation and extracellular nlatrix synthesis. Key lation of type I collagen and fibronectin protein words: TGF-J3/collagell/wolUld 1.epairlfi/wosis. J Invest Del' synthesis by TGF-J3 and CTGF was confirmed by IIIatol 107:404-411, 1996
issue regeneration and repair proceed in a cascade fibroblast proli feration (DeLarco and Todaro, 1978; Assoian cf ai, fas hion beginning w ith a coagulati o n and infl amma 1984; Leof c/ ai, 1986; Soma and Grotendo rst, ] 989; Ishikawa e/ ,II, tOl'y phase, fo ll owed by granulation tissue formation 1990; Battegay e/ ai, 1990); (ii) elevated synthesis of extracellular and fi nally extraceIJul ar matrix deposition and termj matrix components inclu diJlg fibronecrin, typc I collagen, ill.tcgrins, nation of the respo nse. Peptide growth factors playa laminin, and glycosaminoglycalls productio n (Ignotz and Mas ceTntral role in this process. ft is li kely that f.1ctOI·S released by sague, '1986; Roberts eI ai, 1986; R aghow ct ai, 1987; Varga ct ai , pl atelets and inflamm atory cell s serve as initiators of the regenera 1987; Penttinen ct ai, 1988), and (iii) decrcased degradatio n of tion/repair response. Similarly, wound repair disorders, as wcll as extracellular m atrix due to direct inhibition of protease activity and organ-specifi c fibrosis, may be ca used by dysfunctional cascades. stimulation of the synthesis of protease iJlhibito rs (Laiho et ai, 1986; O ne of the principal regula tory factors that appears to function as an Lund et ai, 1987; Kerr ct ai, 1990). Previous studies have demon ini tiator in these processes is transforming growth factor-{3 strated that 'I large po rtion of the TGF-,B induction of matrix (TGF-(3) (Amcnto and Beck, 1991; Raghow, 1991; Wahl, 1992; protein synthesis is not shared by other growth factors such as R.oberts and Spo.", 1993). fibroblast growth factor (FGF) o r platelet deri ved growth factor TG F-{3 has been sho wn to act as a potent stimulatory signal for (PDGF) (Ignotz and Massague, 1986; R o berts el ai, 1986; Penttinen connective tissue formation during wound repair and in fibrotic et ai, 1988). conditions. Elevated TGF-{3 mRNA or protein Icvels have becn Connecti ve tissue growth factor (CTGF) is a cysteine-rich documented in tissue during normal wound repair (I garashi eI (II, mitogenic peptide that was o ri gin ally identificd as a growth factor 1993; Levine e ( ai, 1993), and fib rotic disorders ofch e skjn (Kulozik secreted by vascular endothe li al cell s in culture (Bradham el a/. et ai, 1990; Peltonen ct ai, 1990; Smith and LeRoy, 1990) and 1991 ). CTGF is selectively induced in fibroblasts aftel' acti vation internal organs and tissues (Nagy c ( ai, 1991 ; Kagami e/ ai, 1993; with TGF-{3 (Soma and Grotcndo rst, 1989; Igarashi ef ai, 1993). Bahadori et ai, 1995). T he increased fi brotic tiss ue has been CTGF is a m ember of a tamily of peptides that include serum attributed to severa l actions of TGF- {3, including: (i) increased induced gene products ceftO (Simmons et ai, 1989), cyr61 (O'Bri en et ai, 1990), flsp1 2/ {31G M 1 (Brunner el ai, 1991; R yseck et ai , '1991) Manuscript received February 23, 1996; revised April 29, 1996; accepted and a chi cken transforming gene, nov Oo li ot et 1992). CTGF fo r publication May 21.1 996. ai, Reprint requests to: Dr. Gary R . Grotcndorst, Department of Cdl also shares significant sequence homology w ith a Drosophila gene Biology and Anatomy, University of Miami School of Medicin e (R-124), product, twisted gastrulation (twg) (Mason et ai, 1994), that 1600 NW 10th Avenue, Miami, FL 33'136. determines cell fates during do rsa l/ventral pa ttern formation in the Abbreviations: CTGF, connective ti ss ue growth fa ctor; r TGF, recom cmbryo. Previo us studies have dcmonstratcd coordinate expression binant connective tisslI e growth flletor. of TGF-,Bl and CTGF in granul;ltion tiss ue bcds during wound
0022-202X/96/S10.50 • Copyri ght lid 1996 by T he Society for In vestigative Dermatology, Ill c.
404 VOL. 107, NO.3 SEPTEMB ER 19% e T C I: STIMULATION OF FIBROPLASIA 405 repair (Igarashi el n/, 1993) and found th at d e rmal fibroblasts in 49F fibroblasts as ta rget cell s as described previously (Soma and Groten sc1et-o d erma lesions overexpress CTGF (Ig arashi c( (/ 1, 1995). T h e do rst, 1989). Anchorage-independent growth assays 'were pcrfornlcd essen CTGF mRNA is selectively induced in fibroblasts b y TGF- {3, but tia ll y as described by Guadagno and Assoian (1991). not by oth e r g rowth fa c tors, su c h as epiderm a l growth f.,cto r Growth Inhibition Assays T he mink lung epithelial cell line. Mv1Lu (EGF), PDGF, or FGFs (acidic FGF, basic .FGF) (Igarashi el n/, (ATTC No. CCL 64), was used as C
u.. CTGF, based on amino acid compOS ition and peptide sequence u.. (!) m analys is. T he doublet pattern obscrved ill the rccombinant material en is similar to that of the natural product isolated fi'0l11 human skin (!) I- m fibroblast or endothelial cell -conditioned m edium (Fig 1). T he 0 t) l- LL. CTGF is greater th an 95% pure based o n quantitativc scans of the l- LL LL U C00l11l11a s .~ i e - s tain ed gels. T he basis fo r the 111ll.l tiple bands appears en (!) (!) W ~ LL. > 0 to be due to difFerences in glycosylation as detected by di(fcrences ~ l- I- en :::J a.. in reactivity with concanavalin A (Grotcndorst GR, Will iam s S, U u Segarul11 P, unpublish ed observatio ns). 2 a... a... I I a... Mitogenic Activity of CTGF We initiall y tested the biologic 97- activity of thc CTGF with NRK fibro blasts usin g a mitogcnic assay perfo rmcd on monolayer cultures. CTGF stimul ated a conccntra ti o n- dependent increase in DNA synthcsis in the m o nolayer cuJ tures, with a maximal stimll.lation of 6-fold over non- treatcd 45- cul tures at concen trations of 20 -50 ng rCTGF per ml (Fig 2A). T his activity was enhanced by the prescnce of small amounts of EGF (Fig 2A), although less than that observed w hcn EGF was 30- added with TGF-/3 (Fig 2C). Because CTGF exhibited a high aftinity for h eparin, and because previo us studies have demon strated that heparin-binding growth fa ctors such as acidic FGF can 21- exhibit much higher mitogenic activity in the presence of heparin , we cxamined the e ftects of hcpari.n on CTGF activity. Additio n of 10 j.Lg heparin per ml to the m edium resulted in a significant 14- increase in the amount of mitogenic activity exhibited by CTGF (3-fold over 110n-hepa.l;n-treated) with no significant chan ge in the concentrations of CTGF required to give half-maximal or m aximal Coomassie Immunoblo t anti-PDGF stimulation of DNA syn thesis (Fig 2B). Lastly, we compared the R-Z50 activity of TGF-/3 alone or in combination with either EGF or CTGF in the NR.K m on olaycr mitogenic assay (Fig 2C). TGF- /3 Figure 1. Gel electrophoretic analysis of CTGF. rCTGF produced in alon e exhibited li ttle activity in o ur DNA synthesis assay. A large High Five cel ls by a bacculovirus express ion system was purified by affini ty chromatography on heparin Sepharose. The purity of the rCTGF is sy nergistic stimulation occurred with low concentratio ns ofEGF, as demonstrated in lane 2 of the l
Inability of CTGF to Stimulate Anchorage-Independent of the baculoviral polyhedron gene promoter. CTGF was purified Growth or Act as a Growth Inhibitor for Epithelial Cells by a onc-step isolation procedure using heparin-Sepharose aftinjty TGF-/3 was o riginally identificd due to its ability to stimulate the chtom atography as described in Ma terials aHd Methods. T he purity of anchorage-independent growth of no rmal fibroblasts (Delarco and the rCTGF w as determined by SDS-PAGE and staining with Todaro, 1978). While other growth factors arc required for cell Coomassie Blue R.-250. As sho wn in F ig 1, two peptides are division under these conditions (Assoian e/ ai, 1984), no o ther detected that co-migrate with bands detectcd in a paralJ el immu growth factor h as been fo und to possess this unique activity. TGF-/3 noblot performed o n the same mate ri al. Both of these peptides are has also been shown to act as a spccifi c inhibitor of epitheli al cells
A B C 100 30 -{}- EGF 0.2 ng/ml '7 --+- + Heparin( 10 ug/ml) '7 '70 a ... T" 75 ~ )( >< >< - E E E 20 ..... CTGF 20 ng/ml Q. Q. 0. 0 0 0 >, >, >, -- .<: .<: J: "4 - ..... ~ ..... 10 J: ...,=i:. i- M -0- No Additions -0- No Additions
00 01"" 10 20 30 40 50 60 25 50 75 100 125 0 2 4 [CrGF] (ng/ml) [crGF] (ng/ml) [TGF-~] (ng/ml)
Figure 2. Mitogenic activity of CTGF. (A) Mitoge ni c activity of CTGP alone and in the presence of EGF fo r NR.K t'ibroblasts. Cell s were assayed a described in Materia ls alld Methods. The indicated amoun ts of CTGF were added to the media alone (0 ) or in the presence of EGF (0 .2 ng per ml , e). (B) Effect of heparin on the mitogeni c acti vity of CTGF on NRK ce ll s. Cell s Were assayed as in A. T he indicated amounts of CTGF were added alone (0 ) or in the presence of heparin (10 J.L g per ml, e) Synergism ofTGF-Bl and EGF but not CTGF to stimulate DNA synthesis of NRK ce lls in monolaye r cultures. T he indicated amoun ts ofpme TGF-/3 were added alone (0 ), in the presence ofCTGF (20 ng per ml, e) or EGF (0.2 ng per mi . 0 ). A n o", points to baseline of CTGF stimulation. Error bll rs, mean ::':: SO. VOL. '107. N O. 3 SEPTEMBER 1996 eTCl' STIMULATI ON O F FIl3ROPLASI A 407
Figure 3. Effect of TGF-f3 aud CTGF ou A B anchorage-iudependent growth of NRK fibroblasts and inhibition of growth of ci. -o- TGF-B 0 0.4 ::; "'? -+-CTGF MvlLu luug epithelial cells. (A) Anchor u age-independent gro\vth assays were per <='1 Ql . - 0
(Tucker ef ai, 1984), and vascular endothelial cells (Heimark ef ai, cultures of NRlZ cells. Treatment of cells with TGF- /3 and CTGF 1986; Muller e/ ai, 1987; Takehara et ai, 1987). We wanted to induced a marked upregulation of ai- type I collagen, fibronectin. compare the activity of CTGF with that of TGF-/3, both i.n an and as integrin transcripts as detected by northern blot analysis anchorage-indep endent growth assay with NRlZ fibroblasts and a (Fig 4). TGF-{3 also stimulated a large increase in the level of growth inhibition assay with MvlLu epithelial cell s. As expected, CTGF transcripts consistent with Ou r previous observations. No TGF-/3 stimulated the growth of NRlZ fibroblasts in a concentra detectable changes in CTGF message were observed in the CTGF tion-dependent matmer (Fig 3A). CTGF had 110 activity in this treated cultures. Control experiments examining ribosomal RNA assay. Furthermore, addition of CTGF with TGF-/3 did not aug by ethidium bromide staining indicated that equivalent amounts of ment the activity found with TGF-/3 alone. A similar outcome total RNA were present in each of the samples. Quantitative scans occurred in the growth-inhibitory assay using Mv1Lu cells as of the blots demonstrate a variation of:!: 15% in the actin signal, targets (Fig 3B). TGF-/3 was a potent and eifective inhibitor of which served as a second control for RNA loading. T he matrix proliferation of these lung epithelial cells in monolayer culture, protein-encoding genes (a l type I coUagell and fibronectin) exhibit whereas CTGF had no eifect either as a growth inhibitor or growth 4- to 12-fold levels of induction (400 to 1200%), and th e CTGF stiInulator. T hese data demonstrate that CTGF cannot substitute transcript was induced 20-fold (2000%) after TGF-/3 treatment for TGF-/3 in e ither of these assays and indicate that TGF-/3 exhibits compared to control cultures. T he transcript for as integrin was biologic activities that are distinct from those of CTGF. induced 3-fold by TGF-/3 and n early lS-fold by CTGF. T he rate of synthesis of collagen and fibronectin w as the n Elevation of Matrix Protein Transcripts and Matrix Protein examined in both TGF-{3- and CTGF-treated 1110nolayers of NRK Synthesis by rCTGF TGF-/3 stimulates the expression of col.la fibroblasts. Cells w ere starved in serum-free DMEM containing ITS gen, fibronectin, and integrin genes in fibroblasts. We wanted to for 18 h . Type I collagen and fibronectin w ere fol1owed by either compare the ability ofCTGF with TGF-/3 to elevate mRNA coding immwl0precipitation (fibronecti.n) or p epsin digestion (collagen), for 0'1 type I collagen, fibronectin, or 0'5 integrin in the monolayer followed by SDS-polyacrylamide gel electrophoresis and autora diography (Fig 5). T he results of these studies demonstrate that both type r collagen and fibronectin synthesis were stimulated by TGF-/3 and CTGF compared to control cul tures. Quantitative densitome tric analysis of the fi lms indicate that, based 011 the
NO ADD TGF-B CTGF TYPE I COLLAGEN Experiment # 1 2 1 2 1 2
FIBRONECTIN Fibronectin 0< 5 INTEGRIN
CTCF TYPE I collagen
ACTIN Figure 5. Stimulation of fibronectin and type I collagen synthesis by TGF-f3 and CTGF. ConAu ent monolayer cultures of NRK fibroblas ts Figure 4. Northern blot analysis for extracellular matrix protein were stimulated for 24 h with cither TGF-f3 (to ng per ml) or CTGF (20 ng mRNA in cultured fibroblasts. COIIHu ellt cultures of NIU( t1broblasts PCI' 1111) and pulse labeled for 2 h with [" Sjmethiol1.iJlc (50 !LCi per ml) in were placed in serum-free media fo r 24 h pl-ior to the addition of either methionine-frce media. Fibronectin and collagen synthesis was detcnnined buffer (control lane 1). lOng TGF-f3 per mi . or 10 ng rCTGF per ml for as described in JHateria/s al/ d Methods . Two separate experimcnts were 24 h. Northern blot analysis was then performed on total RNA as described pe rformcd and arc indicated as experimcnts I and 2. An"Ol/ls indicate in Materials am/ Met /lOds. fibroncctin (/0)1 pa/ll'I) or collagcn a -chains ( (W//O'll pal/d). 408 FRAZIER fiT AL T H E JO URNAL OF INVESTIGATIVE DERMATOLOGY
Fig ure 6. Skin and subcutis from NIH Swiss neonatal mice after injection of saline control or growth fac tors. Mjcc were injected one time daily for 3 d and the tissue was harvested and processed as described in Materials mill Metll ods. (A) saline; (B) 800 ng rPDGF BB; (C) 800 ng EGF; (D ) 800 ng TGF-~ ; (E) 400 ng rCTGF daily fo r 3 d. Epidermis (E) , dermis (D), hair fo llicles (F), and granulation ti ss ue (G) arc indicated. Scale bar, 200 J.L1ll . VOL. 107, NO.3 SEPTEMB ER 1996 CTGI' STIMULATION OF FIBROPLASIA 409
Figure 7. In sit" hybridization for CTGF tnRNA in TGF-fl-injected neonatal murine skin and subcutis. TGF-B1 (800 ng) "' average of the two exp eriments, TGF-,B stimulated a 3-fold in duced only a mil d granulation tiss ue response. similar to that crease in the synthesis of fibronectin and a 2.5-fold increase ill pl'oduced by CTGF levels of 50 ng/site. Injection of EGF did not collagen synthesis. CTGF stimulated a 5- fo ld increase in both res ult in detectable fibroplas ia, and the connective tissue in these fibronectin and collagen. T hus, with respect to the induction of tissue samples had an appearance similar to tbat of saline-injected matrix production, CTGF has actions that are similar to TGF-,B and controls. T he EGF-injected specimens, however, exhibited hyper distinct from other growth factors such as PDGF or FGF. keratosis of the epidermis and degenerative changes in the hair follicular epithelium, demonstrating th at the EGF was biologically Fibroplasia Induced by Subcutaneolls Injection of Growth active and produced efFects on the epidermal cells at the sites of Factors The injection of TGlo-/3 into the dermal/subcuticular injection . T he area of granulation tissue was meas\\l:ed in slides area of skin in neonatal NIH Sw.iss mice indu 'es a rapid and from the mice injected with TGF- /3, CTGF, or PDGF. T he average dramatic increase in the formation of granulation tissue prim<1xily area of fibroplasia in the specimens from the TGF-{3-injected mice composed of connecti ve tissue cells and large amounts of extracel 2 2 was 28.8 111111 , in the CTGF-injected mice it was 22.4 mm , and in lular matrix (Roberts ct nl, 1986). W e compared the actions of 2 the PDGF-iJ1i ected specimens it was 3.4 mm . TGF-{3 , PDGF BB, EGF, and CTGF in this ill "iv() model of fibrosis and wound repair (Fig 6). After two injections, gross nodules were TGF-f3 Induction of CTGF mRNA in Skin Fibroblasts [II apparent in both the TGF-{3- and CTGF-injected mice and con V iIIO. To determine whether the CTGF transcript was induced in tinued to enlarge throughout the remainder of the experimental tlle fibroblasts participating in the formation of the granulation protocol. Both TGF-,B aJld CTGF induced similar alterations of the tissue induced by TGF-,B injecti on, we performed ill silll hybridiza J10rmal tissue, including large increases in the number of connective tion for CTGF message on sections of the subcutis obtained fi'om tissue cells and extracellular matrix material. Dosage studies indi saline- and TGF-/3-inj ected mice (Fig 7). Animals were injected cated that the optimal concentration of CTGF for induction of with 800 ng of TGF-Bl followed by a second injection 24 h later. fibroplasia was 400 ng/site compared with 800 ng/site for TGF- /3, T he tissue from the injection site was harvested and prepared 8 h w hich is consis tent w ith the previously reported dose response for after the second injection. Control experim ents using saline only TGF-,B injection in mice. The amount of granulation tissue formed were performed in parallel. A strong signal was detected in the at sites injected w ith 800 ng of CTGF was consistently Jess than in TGF-{3-injected dermis w hen hybridization was performed lIsing those with 400 ng/site. Lower concentrations of CTGF induced an antisense CTGF l;boprobe (Fig 7A). Hyb,;dizatiol1 controls of significant granulation tissue formation with some effect detectable TGF-{3-inj ected tissue sections with sense strand CTGF riboprobes at amounts as low as 50 ng per injection. Prolonging the interval were negative, with no significant ba.ckground hybridization (Fig from the last injection to sacrifice and histologic examination to 7 d 7C). N o CTGF transcl;pts were present in either noniJ1iected skill in both the TGF-{3- and CTGF-injected mice resulted in resorption (not shown) or saline control-iI1iected skin si tes (Fig 7B). T he only of granulation tissue. Injection of PDGF BB at 800 ng/site pro- cells expressing the CTGF transCl;pts were cO IUJ ective tissue cell s 41 0 FRAZIER fiT At THE J OURNAL OF INVESTIGATIVE OElWATOLOGY (fibroblasts). N o cell s in the epidermis were positive for CTGF between TGF-13 and the increased deposition of collagen . O ne trn nscripts in either con trol or TGP-f3-injected skin. Cell s in the significant problem with linking the presence ofTGF-{3 transcripts deeper regions of the skin such as adipocytes were also negative for in a tissue with TGF-13 activity is that TGF-{3 is made in a latent CTGF transcripts in both control and TGF-f3-injected skin sites. form that must be processed to obtain the active protein (Lawrence Surprisingly, although the CTGF gene is constitutively expressed et ai, 1985; Pircher et ai, 1986; O dekon et ai, 1994). Most of the by vascular (h uman umbilical vein) endotheli al cell s in culture, we TGF-{3 in tissue and produced by cell s in cul ture is in this latent did not detect CTGF transcripts in cl earl y identifiable endothelial form (Flaumenhaft c( ai, 1993). Consequently, the presence of cell s in either large vessels or capillari es. T hese resul ts indicate that transcripts does not always correlate well with the amount of active the pt;ncipal so urce of CTGF ill /J i/Jo are connective tissue cell s TGF-{3 present (Flanders cf ai, 1989). Because the m ethods em activated w ith TGF-{3. ployed to effectively extract growth facto rs fro m tissue also convert latent TGF-{3 to the active form, it is difficult to determine how DISCUSSION much active TGF-{3 is present in a tiss ue . Our previo us studies on T he data from o ur studies indicate that CTGF acts to stimulate cell fi broblastic cells ill culture and the data reported in thjs manuscript division and extracellular matrix production by NRK fibroblasts in demo nstrate tha t the expressio n of CTGF is dependent on the culture. The mjtogenic activity of CTGF is augmented by EGF and presence of active TGF- {3. W e have not found factors other than heparin but not by TGF-{3. With respect to m atrix protein syntlle TGF-{3 that induce a signific3 ntlevel of expression of the CTGF sis, we find that CTGF stimulates an increase in the expression of gene. Furthermore, other genes, such as PAJ-1, w hose transcription a l-type I coll agen and fibroll ectin mRNAs, w hich results in is stimulated by TGF-{3 are equally induced by other growth factors increased synthesis of these proteins. Injection of CTGF into the such as PDGF (Grotendorst, unpublished observations). T hese skin of neonatal mice stimulates the formation of a m atrix-rich studies suggest that the presence of CTGF transcripts m ay be useful granulation tissue. Compared to other known growth fa ctors, as a marker for the presence of active TGF-{3 in a tissue . CTGF's biologic actions are most similar to t1lOse of TGF-{3. In summary , these data iJldicate tbat CTGF acts to stimulate Nonetheless, CTGF does not substitute for TGF-{3 in anchorage fibroblastic cell proliferation and extracellular matrix synthesis. independent growth assays, nor does it act to inhibit the growth of Because CTGF is selectively induced in connective tissue cell s by epitheUal cell s whose growth is inhibited by TGF-{3. Thus, CTGF TGF-{3, it is attractive to speculate that CTGF m ay function as an appears to mimic some of the actions of TGF-{3 with regard to the autocrine mediator for some of the biologic actions of TGP-{3 on stimulation of connective tissue cell growth and the synthesis of fi brobl asts. T he restricted activity of CTGF on ti broblasts is extracellular matrix but is not a substitute for TGF-{3 in other consistent with this model. It is also consistent w ith the o bserva assays. tions that the expressio n of the TGF- {3 and CTGF genes are Previous studies have indicated that TGF-{3 can stimulate the coordinately linked during norm al wound repair. Furthermore, in a growth of various fibroblastic cell types by the induction ofPDGF wide variety of fibrotic conditions in which TGF-{3 is found to be genes (Leof et ai, 1986; Battegay et ai, 1990; Islukawa et ai, 1990 ). expressed, such as scleroderma, fibrotic liver disease, and human T his m ay be a part of the mechani sm for TGF-{3 stimulation of atherosclerotic plaques, we fmd high levels of expression of the connecti ve tissue cell growth , how ever, it does not appear that CTGI: gene (Grotendo rst, unpubUshed o bservations). T hese obser PDG F plays a direct role in the stimulation of extracellular matrix vations support a cascade model for connective ti ss ue formation, protein production induced by TGF- {3 (Penttinen et ai, 1988). T hus, shared by nOl"mal and pathologic processes. In t1u s model, ini tiators whereas i" lIillo studies indicate that PDGF alo ne can accelerate acting during the acute response to il~u ry induce fa ctors whose granulation tissue form ation in normal animals (Grotendo rst e ( ai, fi.ll1ction is to maintain the repair process and all ow for the 1985), the type of ti ss ue induced by PDGF injectio n has much less maturation of the forming connective tissue. In this regard, the extracellular matrix than the tissue induced by TGF-{3 or CTGF pleotropic actio ns ofTGF-{3 make it we ll suited to act as an initiator injection. We feel there is a significant physiologic difference of the repair process. T he restricted actions of CTGF suggest this between the injection model [initiall y described by R.o berts eI a1 peptide fun ctions as an autocrine and paracrine signaling molecule (1986) and employed in the studies described here] and wouud to m ain tain and perhaps ampli fy and syn chronize the response of models. Tn the injection model, only a miru1l1al amount of trauma fibroblasti c cell s in the tissue. occurs at the site, and growth factors must act alone or in concert with factors tha t are constitutively present in interstitial fluids. In contrast, growth factors added to wound models in animals with normal heaUn g act in an environment containing a milieu of factors T his resenrch '"ilSs llp/ lO rled by Nnli,,,,n/ IlI s,i",'es ~r Hen/", C rall' C I'II/J 7223, n w hose primary functio ns are to stimulate ti ssue regeneratio n and dellc/op", ellltll al/lnr(lfr"'" 'he S), I,/es'cr Cnll cer Cell IeI', alld n gra lll Irom FibroCell , repair. T hus, agents that may play important roles during wound Te,e. (10 C.R.C.). KF. is slIpporlca II)' N ",ioll nllllslitlllcs ,u Henl,I, Traillillg C rnlll heati ng would not necessarily stimulate a large amount of connec RR07057-0"l A ., nwl till iustillllioll nl ji:llol/lship from th" Ulli llc",i' )' of Min",i. tive tiss ue formation in the injection model. T hi s may explain the PDGF and EGF results reported here. T hese constraiJlts make it all the more remarkable that CTGF induces the formati on of connec R EFERENCES tive tissue in a manner that closely resembles that induced by TGF-{3. Such o bservations support the possibility that CTGF may Amcl'l ro EP, Ucck LS: TG)=-{3 and wound healing. C UM Feill/Ifl S}'mp 157: 11 5-129, 1991 Asso iall RK. GrotclI dorst GR. Miller DV, Spo rn MB: Cellular trallsrormatioll by playa role in TGF-f3-mediated formation of granulation tissue. coordinatcd :lcrioll of three peptide growth factors fr0111 IWlllilll platelets nfllre T he coordinate expression of TGF-{3 and CTGF during the 309:804 -80G, 1984 wound repair cascade (Igarashi et ai, 1993) belps to functionally link AlISubel I' M. Brent R. Kingstoll I<-E . Moore DD, Scidmall J G. Smith .IA. Struhl K the inflammatory phase (where TGF-{3 is expressed) and the (cds.) Puri fi C:ltioll and tr;msfcctioll ofbaculoviral DNA for gC ll cl'ating recombi nant viruses. In: 1-1 Pi wl1 ica-Wonns. C ffrrrttf Protocols ;1/ IV/Illen t/a,. Biol"... ~}' . 1101. 2. granulation ti ss ue formation phase (where CTGF is expressed) of Wile)' Interscience. New York. pp 16. 10. 1-8. 1990 this complex process. During the course of examination of the Ba hadori L, Milder J , Gold L. Botll e), M: Acti ve macrop hage '1 ssociated TGF-~ expressio n of growth fa ctor genes in fibrotic disorders we have co-locaHzes with type I proco ll agcl1 gene expression in atherosclerotic human found CTGF expressio n to be closely linked to the expression of pulmoll ar), arteries. A lii) 1'" 11,,,/ 146: 11 40-'11 49, 1995 TGF-{3. For example, CTGF is overcxpressed in fibroblasts in the 13attcga}f EJ. R 'lines EW . Seife rt RA. 13 o w el1 -Popc Dr. R o ss It: Transformin g: g l'OWrll fa cto r bera induces bimoo,,) proli feration of connective 6s.'me c cJJ.~ via complex dermis of patients with scl eroderma (I ga rashi et ai, 1995), a systemic control of :m autocrine platelet-de ri ved growth fa ctor loop. Cdl 63:5 15-524 . fibrotic disorder characterized by the overproducti on of coll agen 1990 (LeRoy e/ ai, 1988). Overexpress ion of TGF-{3 also has been I3r~ld hal11 OM. Ig:trshi A. Potter n... L . Grotcndorst GR: CO Hll cnivc tiSS lI C growth fa ctor: a cysteine-ri ch mitogen secreted by human vascular endo thelial cell s is rehired (0 reported in scleroderma lesions (Kulozik e/ ai, ·1990; Peltonen et ai, til e s rtC-iJldliced il11ll1Cdi:"ll:c c olrl), gell c pro duc t CEF- l O. J Cell Bitl/ 1 '14 : 1285- 1990; Smith and LeRoy, 1990), suggesting a ca usative association '1294. 1991 VOL. '107. NO. 3 SEPTEMBER 1996 CTGF STIM ULATION O F FII3ROPLASIA 411 .Brunne r A. C hinll .I . Ncub'lU cr M. Purchio AF: Id cll tific:lrioll o f" a gene r.1mi ly Lund Lit. Riccio A. Andreasen PA. Nielscn LS. Kristcnsen P, Laiho M. Sakscla O. rcgulated by lr:msl;mlling growth f.,cto r bem. 0 A Cdl Wol 10:293-300. 199 1 l3la si F. Dano K: Transfo rming gro wth f;lCt"or-13 is a strong :md f.'1st acting positive DcLarco J E. T odaro GJ: Growth r.,ctors frolllllturinc san': UI11 :l virlis-tr:msfo rlllcd cell s. regula tor o ftllc lc\'cl oftypc- I plasmillogcll activator-inhibito r mRNA in \ '(f 1-38 I'm( NII II Arml Sri USA 75 :4001-4005. 1978 hUl1l all IUll g fi hroblasts. EMBO) 6: 1 2~ 1 - 1 286 , 1987 Fava RA. Casey "n ". W ilcox J . PeltOIl RW. Moses H L. Nanll ey LB : S)'llthcsis or Mason ED. Konrad KO. W ehb CD. Marsh J L: Dorsal \IIidlille rate in Dros"pit il a tr:ntsfOr111 ill ~ growth fa ctor-beta t by IIlcgakaryocyrcs and its locaiiz:lt'o l'l to cmbryos rcquil-cs twisted ,-~ (,s t"" 'ntioll, a gene c ll coding a sccreted protein rclarcd to I11 cgakar)locytc :l nd platelet 'llpha-grallulcs. Hltwd 76: 194(1-1955, 11)90 1I1I1Itall C0 I111 ccrive tiss ue growdl t:lcro r. Gnu:s 1)('" 8: 1489-150 1. 1994 Fla nders KC. Thompson NL. Cissel OS. Van O bberghen Schillillg E. Baker CC. Kass Muller C . Dehrc lt s J. Nussbaumer U. Bo hlen P. Birchmeier \X/ : Inhibito ry action of M E. Ellin g:s worth Lit, roherts A U. Spo rn MU: T r ~11lsf o nnillg growth fa ctor TGF-{3 011 elldOlhclial cell s. 1'"",. NaIl Acad Sri USA 8 4 :56 1l0- 560 ~. 1987 bc ra- I: histochc lllic: 1i localization with antibodies to difl-crcnt cpitopcs.J C dt BioI Nagy P. Schafr Z. L~pi s K: Immullo histochcmical detecti o n ofTCF-13 1 in fIbro ti c li vcr lOH:(,53-(,60. 1989 disease. HCI' (lI o l '\~ I ' 14: 269-273. 199 1 Flnunlcnhafr R. l{ ojil11:t S. Abc M. RHki.1l DU: ACti V:ltio ll of I:ltcnt tran sformin g O'l3ricll T I'. Yang G P. Sanders L. L:l y LF: Ex p rc~s i oll of cy-r6 ·1. a growth nlclOr g rowth fa cto r beta . II dll Pllnrlllnmi 24:5 1-76. 1993 inducible illllllediate carl )' gene. 11'/,,1 Cdl Wei 10:3569-3577. 1990 Grotendorst GR. Okochi H. Hayashi N: A Il ovel TGF-{:l respoll se clemellt COil trois the Odckon LE. Blasi F, Rifkin DB: Rcquircl11e nt" for re cepto r- bo und urokinase in c~ pl - e~ s i o n oftll c cO llllcctive tiss li e growth f;,c l'Or (e TC 1::) gcne . Cell Growlll DffFer plasmill-depe l1d cl1 t cellul ar cO ll versi01 1 of Iat cnt tra nsforming g rowth f.1c (" o r bcta 7:469--180. 1996 to transforming growth fa ctor beta.) Cell 1'11),si,,1 1 58:39S-~07. 1 99~ Grotcndo rst GR. Martin GR. Pencev O. Sodek J . Harvey AK: Stimulation of Ogawa Y. Seyedin SM: Purification or transfo rllling growth ractors {:lJ and (32 frO I1l gr::wl1lacion I-iss uc formation h), platelet-dcrived growth f.1 ctor in normal and bovin e bOll c and cell c ulture assays. In : 0 Barncs. JP M ~ th c r . GI-I Sato (cds.) diabetic rats.) Clill 11I I'<,s l 76:2323-2329. 1985 Mrtl",ds ,!(E"z ),,,,ot.\U, V,,1198. Acadelllic Press. New York. 1991. PI' 3 17-327 G u ad,lg no T M. Assoinn R..K: G 1 IS contro l of ancho ra ge independent growth ill the fibrohlast cell cycle. ) Cdl Will 1'15: 1572-1575. 199 1 Pel toncnJ. Kahari L. Jaakkola S. Ka hari V-M. VargaJ. Uitto J . Jimcnez SA: Evoluation of TGF-J3 and r)rpc I procoll agcn gene expression in f-ibro ti c skin di seases by ill I-IejITH,rk RL. Twardzik DR. Schwartz SM: Inllibirion of c1ldothclial regcncration hy type-beta tntllsrormillg growth f:,ctor rrom pl:ltclets. Sricllrc 233: 1078-1080. -