Decorin Attenuates Gliotic Scar Formation in the Rat Cerebral Hemisphere

tpadswww1umhttahbwotwpewotipfier

fit

etf

Experimental Neurology 159, 504–510 (1999)Article ID exnr.1999.7180, available online at http://www.idealibrary.com on

0CA

Decorin Attenuates Gliotic Scar Formation in the RatCerebral Hemisphere

Ann Logan,* Andrew Baird,† and Martin Berry‡*Department of Medicine, University of Birmingham, Birmingham B15 2TT, United Kingdom; †SelectiveGenetics Gene Therapeutics,

San Diego, California 92121; and ‡Department of Anatomy and Cell Biology, AGKI (Guy’s Campus), London, SE1 9RT, United Kingdom

Received September 18, 1998; accepted June 5, 1999

halct2pabaapsTaaedmgcceptTmwsttTgtTnl

glT

The transforming growth factor-bs (TGF-bs) are po-ent fibrogenic factors implicated in numerous CNSathologies in which fibrosis and neural dysfunctionre causally associated. In this study, we aimed toemonstrate significant inhibition of fibrogenesis, glialcarring, and inflammation in penetrating incisionalounds of the rat brain using the proteoglycan decorin,hich effectively inhibits TGF-b activity. Adult ratsere assigned to two treatment groups each receiving

4 daily intraventricular injections of 10 ml total vol-me of: (i) saline plus 0.3% autologous rat serum 5 30g protein); or (ii) saline plus 30 mg recombinantuman decorin. On day 0 of the experiment, a stereotac-ically defined unilateral incisional lesion was placedhrough the cerebral cortex into the lateral ventriclend, after 14 days, brains were processed for immuno-istochemical analysis of the lesion site. Specific anti-odies were used to visualize the deposition within theound of matrix molecules and the extent and naturef reactive astrocytosis and inflammation. Quantita-ive and qualitative image analysis of the fibrous scaras performed in sections from a defined anatomicallane through the wound to detect the antifibroticffects of decorin treatment. Treatment of woundsith decorin led to a marked attenuation of all aspectsf CNS scarring including matrix deposition, forma-ion of an accessory glial limiting membrane, andnflammation. Our findings suggest that decorin isotentially applicable to a number of human CNSbrotic diseases to arrest the deposition of excessivextracellular matrix components and maintain and/orestore functional integrity. r 1999 Academic Press

Key Words: central nervous system; injury; scarring;brosis; decorin; scar inhibition; gliosis; cytokines;ransforming growth factor b.

INTRODUCTION

Fibrosis is caused by an excessive deposition ofxtracellular matrix which often compromises the func-ion of the tissue involved. The transforming growth
actor-bs (TGF-bs) are potent fibrogenic proteins which b
504014-4886/99 $30.00opyright r 1999 by Academic Pressll rights of reproduction in any form reserved.

ave been implicated in a broad diversity of biologicalctions including enhancement of wound healing, stimu-ation of extracellular matrix synthesis, modulation ofell proliferation, modulation of inflammatory cell infil-ration, immunosuppression, and neuroprotection (3,2). TGF-bs have been implicated in numerous CNSathologies in which fibrosis and neural dysfunctionre causally associated, for example, in posttraumarain and spinal cord scarring (18–20), postsurgicalrachnoiditis (18), hemorrhagic stroke (14), and sub-rachnoid haemorrhage (13). TGF-bs may also promotelaque development in Alzheimer’s disease and Downsyndrome (28). In all of these conditions, the levels ofGF-b are raised in both the cerebrospinal fluid (CSF)nd also locally in damaged neural tissue. For example,fter CNS traumatic injury we have demonstrated anlevation of TGF-b1 and TGF-b2 isoforms, initiallyerived from hematogenous cells and later supple-ented by endogenous local synthesis by neurons and

lia in the damaged neuropil, and by choroid plexusells of the impaled lateral ventricle leading to raisedytokine levels in the CSF (18, 20). The most directvidence for a fibrogenic role for TGF-bs in the patho-hysiology of CNS fibrosis comes from experiments inhe lesioned brain. On the one hand, raised levels ofGF-bs are correlated with the deposition of scaraterial in such lesions, while immunoneutralizationith TGF-b antibodies markedly inhibits fibrogenic

carring (18, 20). It is implicit from these observationshat attenuation of either excess or inappropriate ma-rix deposition in a variety of CNS pathologies usingGF-b-related antifibrotic agents may limit the patho-enic process, with anticipated clinical benefits. Whilehe principle of inhibition of fibrogenesis in the CNS byGF-b neutralization is established, exploitation of aaturally occurring regulator of TGF-b bioactivity to

imit fibrogenesis may have therapeutic advantages.Decorin is a small secreted dermatan sulfate proteo-

lycan found as a natural component of the extracellu-ar matrix of most tissues, including those of the CNS.his molecule neutralizes all isoforms of TGF-b by
inding the ligand to its core protein (29) and, there-

ffisadOtpmcfirhsw

D

cfrni

A

alid2mspcµspvpwlhwwpafls

H

s

n(aama1fi1lDd

Q

woauMoigdc

S

ivf*

C

sgwr(p1lsa(tctbwg

505DECORIN INHIBITS SCARRING IN THE CNS

ore, could be of therapeutic value in the treatment ofbrotic CNS disease. After CNS injury, decorin isynthesized by astrocytes in the damaged neuropil (23),nd this may represent an endogenous attempt toown-regulate cytokine activity within the wound.thers have demonstrated in rodents that administra-

ion of decorin, either by injection of the recombinantrotein or by transfer of the decorin gene into skeletaluscle, reduces the kidney fibrosis that occurs as a

onsequence of glomerulonephritis (2, 11) or the lungbrosis induced by bleomycin (6). In this study, weeport that intraventricular injections of recombinantuman decorin significantly inhibits fibrogenesis, glialcarring, and inflammation in penetrating incisionalounds of the rat brain.

MATERIALS AND METHODS

ecorin

Recombinant human decorin was prepared fromulture medium of Chinese hamster ovary cells trans-ected with human decorin cDNA (29). The ability of theecombinant human decorin used in this study toeutralize the activity of endogenous and exogenous TGF-bn bioassays has been characterized elsewhere (29).

nimal Surgery

Surgical procedures and animal care were licensednd carried out according to British Home Office guide-ines. Stereotactic lesioning of the cerebral cortex andntraventricular cannulation were executed exactly asescribed by us elsewhere (17, 18, 20). Adult female00- to 250-g Wistar rats were assigned to two treat-ent groups of 5 animals each receiving: (i) 10 µl/day

aline plus 0.3% autologous rat serum to give a finalrotein content of 30 µg/10 µl to negate the proteinoncentration of the test reagent; or (ii) 30 µg/10gl/day decorin in saline. On day 0 of the experiment, atereotactically defined unilateral incisional lesion waslaced through the cerebral cortex into the lateralentricle at the same time as ipsilateral placement of aermanent intraventricular cannula. Reagents (10 µl)ere perfused into the lesion site by daily intraventricu-

ar injections through the cannulae for 14 days underalothane anaesthesia. The body weights of each animalere monitored daily and no significant differencesere noted between the treatment groups at any timeoint (data not shown). After 14 days post lesion (dpl),nimals were killed and their brains processed foruorescent immunohistochemical analysis of the lesionite.

istology and Immunohistochemistry

Brains were processed into polyester wax and 7-µm
ections of the lesion site stained by fluorescence immu- (
ohistochemistry to detect glial fibrillary acidic proteinGFAP)-positive astrocytes (using a polyclonal rabbitnti-GFAP antibody from Dakopatts, Ely, Cambs, UK,t a dilution of 1:250), ED1-positive macrophages andicroglia (using a monoclonal mouse anti-rat ED1

ntibody from Serotec, Oxford, UK, at a dilution of:200), fibronectin (using a polyclonal rabbit anti-bronectin antibody from Dakopatts at a dilution of:100), and laminin (using a polyclonal rabbit anti-aminin antibody from Sigma Chemical Co., Poole,orset, UK, at a dilution of 1:100). The method isescribed in detail elsewhere (17, 18, 20).

uantitation of Immunohistochemistry

In all cases, the effects of each treatment on the CNSounding response were quantified by image analysisf fluorescently labeled sections taken from a definednatomical plane through the lesion site (17, 18, 20)sing a Leitz confocal microscope linked to a Bio-RadRC500 laser scanning system. The relative intensity

f fluorescence in terms of the mean integrated pixelntensity was expressed for each animal as the inte-rated fluorescent intensity per square micrometer inefined and exactly equivalent subpial, mid- and deep-ortical areas at a constant magnification.

tatistics

Significant differences in the specific immunoreactiv-ty measured per unit area in lesions of decorin-treatedersus saline-treated rats were calculated using single-actor ANOVA to give a significance levels of *P , 0.05,*P , 0.01, or ***P , 0.005.

RESULTS

ellular Effects of Decorin on CNS Scarring

By 14 dpl, in rats treated with saline, all the fibroticcars contracted into a dense permanent trilaminarlial/fibrotic complex (Figs. 1ai–1aiv and 2), the core ofhich was laid down by meningeal fibroblasts and was

ich in fibronectin (Figs. 1aii, 2, and 3B) and collagennot shown). Residual macrophages and microglia occu-ied the core and bordering viable neural tissue (Figs.aiv, 2, and 3D), both of which were separated by aaminin-rich basal lamina of the glia limitans of thecar (Figs. 1aiii, 2, and 3C) and undercoated by end-feetnd processes of reactive astrocytes of the glia limitansFigs. 1ai, 2, and 3A). These three layers extendedhroughout the lesion and became contiguous with theomplementary laminae of the glia limitans externa athe pial surface of the cerebrum. The basement mem-ranes of neuropil blood vessels were also visualizedith the anti-laminin antibody and revealed the angio-enic response in the tissue surrounding the wound
Figs. 1 and 2).

(fip

506 LOGAN, BAIRD, AND BERRY

FIG. 1. Cellular effects of decorin treatment on cerebral wounds. Coronal sections through a defined plane of the site of a cerebral lesionBregma 12.5 mm) at 14 dpl comparing the effects of infusion of (ai–aiv) saline with (bi–biv) decorin on (i) GFAP-positive astrocytes, (ii)bronectin, (iii) laminin, and (iv) ED1-positive macrophages and microglia. Sections were stained with fluorescent conjugated antibodies; theath of the lesion is marked by residual autofluorescent macrophages/microglia. Bar, 100 µm.

tcccwwsopabl3wttrs1tl(

Dtd

mifiolvWrmbbed

a

fiB


By contrast, in every decorin-treated rat at 14 dpl,he glia limitans of the scar (comprising abutted astro-yte processes and a basal lamina) was absent from theortical lesion, except in the outermost layers of theortex where it was reconstructed to become continuousith the glia limitans externa, thereby sealing off theound and limiting normal wound closure to this

uperficial site. In the deeper cortical layers, the numer-us activated astroglia neither organized into the ex-ected limiting membrane of the scar (Figs. 1bi and 2nd quantified in Fig. 3A) nor laid down a laminin-richasal lamina in the wound (Figs. 1biii, 2, and 3C) andittle or no fibronectin was deposited (Figs. 1bii, 2, andB). Hence, the cut neuropil surfaces became apposedithout the intervention of a glial/matrix scar. Decorin

reatment also suppressed the angiogenic response inhe neuropil around the wound, so that few woundelated blood vessels were seen, although the normalmaller diameter microvasculature was apparent (Figs.and 2). Finally, there was a significant reduction in

he numbers of EDI-positive macrophages and microg-ia in the wound and juxtaposed neuropil at 14 dpl

FIG. 2. Image analysis of the cortical glial/matrix scar. Representbronectin, laminin, GFAP, and ED1 immunofluorescence at 14 dpl inar, 100 µm.

Figs. 1biv, 2, and 3D), indicating immunosuppression. t

espite the general immunosuppression, the post-rauma inflammatory cysts were still apparent in theecorin-treated cerebral lesions (Fig. 1).

DISCUSSION

The technique of cerebral injury in rats, fully docu-ented by us elsewhere (17), provides a well-character-

zed experimental model of wound healing and CNSbrosis in particular, in which the antifibrotic efficacyf decorin can be tested. The lesion penetrates theateral ventricle, allowing reagents injected into theentricular cerebrospinal fluid to perfuse the lesion.ithin these standard lesions a sequential cellular

esponse occurs characterized by hemorrhage, inflam-ation, the formation of a trilaminar glial–matrix scar

y astrocytes and transiently invading meningeal fibro-lasts, and an abortive regeneration response of sev-red axons, all of which are essentially complete by 14pl in the rat.This study demonstrates that acute-phase treatment

fter wounding with recombinant human decorin leads

ve high-power images from the confocal analysis to illustrate specificfined cortical areas of saline- versus decorin-treated cortical wounds.

atide

o a marked attenuation of all aspects of CNS fibrotic

swgwtTiatggtitgamteuoeguw

Tata

rmsatmoratbtosieprt

((ssc *P ,


carring. The spatially graded response to decorin fromound depths to cerebrum surface may reflect theradient of wound perfusion, since the proteoglycanas delivered from the ventricles through the base of

he wound. While decorin is known to directly blockGF-b activity in vivo and in vitro, our study does not

dentify the mechanism of decorin’s antifibrotic activity,nd there are several possible means through whichhe molecule can achieve this end. For example, thelycosaminoglycan chain of decorin has many sulfateroups, which confer the potential to bind many posi-ively charged proteins and cytokines which may benvolved in CNS fibrogenesis. However, the characteris-ics of decorin suppression of CNS inflammation andlial/mesenchymal scarring are similar to those seenfter TGF-b immunoneutralization in the same animalodel (18, 20), suggesting that the actions of decorin in

his instance are specifically coupled to suppression ofndogenous TGF-b bioactivity. The effects observed arenlikely to be attributable to nonspecific proteoglycanr protein effects since the delivery into the wound ofquivalent amounts of serum protein in the controlroup had no effect on scar formation when compared tontreated lesioned animals (data not shown). Hence,

FIG. 3. Quantitative image analysis of the wounding response afsolid bars) and decorin (hatched bars) on the levels of (A) GFAP-immuD) ED1-positive macrophages and microglia at 14 dpl. Images of fluoite (as illustrated in Fig. 2) were digitized and the integrated fluorubpial, mid-cortical, and deep-cortical areas. Significant reductionompared to saline-perfused control wounds are indicated as follows:

hile the influence of decorin may not be restricted to m

GF-b suppression, it seems likely that these cytokinesre the primary target of decorin in CNS lesions andhe means by which decorin achieved its antifibrogenicctivity.All of the activities of decorin observed in this study

elate closely to the known actions of TGF-b in experi-ental models. TGF-bs are known to induce chemokine-

is and chemotaxis in vitro and in vivo (26), up-regulatestrocyte production of monocyte chemoattractant pro-ein-1 (a potent chemoattractant and stimulator ofonocytes) (8), and affect cell migration via modulation

f expression of integrins, a major class of cell adhesioneceptors and cell adhesion molecules (7, 10). Thepparent immunosuppression by decorin observed inhis model of CNS injury, evidenced by the reduction inoth immunoreactive macrophages and microglia inhe wound, is very similar to the immunosuppressionbserved in CNS lesions after TGF-b2 immunoneutrali-ation (20). Parodoxically, neutralization of the TGF-b1soform in the same experimental model resulted in annhanced inflammatory response, with microglia beingarticularly responsive (18). Thus, while TGF-bs areecognized as general immunosuppressors, in the ini-ial stages of inflammation they stimulate monocyte

cerebral lesion. The effects of perfusing cerebral wounds with salinepositive astrocytes, (B) immunoreactive fibronectin, (C) laminin, andcently stained sections through a defined coronal plane of the lesionent intensity for each antigen was quantified in exactly equivalentf quantified elements in cerebral wounds after decorin treatment

0.05, **P , 0.01, and ***P , 0.005.

ternoresesc

o

igration (26) and it may be this TGF-b activity which

Tbmmi

miacacwbetpgab

ptitoTdersa9meastt

ipgsaawaanstin

r

iicbw

DdCI

1

1

1

1


GF-b2-specific antibodies, and by implication decorin,lock. The immunosuppression noted with decorin treat-ent may be therapeutically advantageous, since itay reduce the risk of treatment initiating adverse

mmunological reactions.Others have shown that TGF-bs affect astrocyteorphology, proliferation, migration, and interaction

n vitro (22). TGF-b is a potent astrocytic chemotacticgent, it reduces cell–cell contacts and increases focalontacts, both precursors for cell migration (5), and itlso stimulates the development of highly branchedellular processes and multicellular colonies, all ofhich are essential components of glia limitans mem-rane formation (4, 15, 23). Immunoneutralization ofither the TGF-b1 or the TGF-b2 isoform does not limithe extent of reactive gliosis after injury, but doesrevent the organization of astrocytes into the limitinglial membrane of the CNS scar (18, 20). Decorin hasn identical effect suggesting that the proteoglycanlocks TGF-b-mediated astrocyte scar formation.The fibrogenic effects of TGF-bs in various models of

eripheral tissue injury are well documented (3) andhe antifibrotic effects of decorin have been linked tonfluences on TGF-b activity (2, 6, 11). These observa-ions support our suggestion that the antifibrotic effectsf decorin in the CNS reflects inhibition of endogenousGF-b activity and explain the marked reduction ineposition in CNS wounds of every matrix molecule wexamined. The antifibrogenic actions of decorin mayelate to astrocytes as well as the meningeal fibroblastsince both cell types are TGF-b responsive and producerange of matrix molecules in response to injury (1, 4,

, 24, 27). The fibrogenic actions of TGF-bs are probablyediated by up-regulation of synthesis of multiple

xtracellular matrix molecules (1, 9, 25, 27) and prote-se inhibitors (16, 21), down-regulation of the expres-ion of matrix-degrading proteases (12), and modula-ion of integrin expression and consequent fibroblastrafficking into the wound (7, 10).

The apposition of the cut neuropil surfaces observedn decorin-treated lesions suggests the absence of thehysical, and perhaps biochemical, barrier that theliotic scar represents to regenerating axons. In thistudy, regeneration was not rigorously studied usingxonal tracing methods; however, histologically, noxons were seen to traverse the scar-inhibited lesionhich confirms our previous finding after TGFb1 (18)nd TGFb2 (20). This observation presumably reflects,t least in part, the limiting supply of appropriateeurotrophic factors required for regeneration. Weuggest that, in addition to inhibiting cicatrix forma-ion, a strategy to mobilize the axon growth machinerys required in order to achieve vigorous and sustainedeuron regeneration across a transection site.In summary, our results demonstrate efficacy of

ecombinant human decorin in vivo, which we suggest

nhibits endogenous TGF-b bioactivity thereby reduc-ng CNS fibrogenesis. Our findings suggest that decorinould be developed as a therapeutic antifibrotic agent,roadly applicable to a number of human CNS diseasesith excessive extracellular matrix deposition.

ACKNOWLEDGMENTS

The authors are indebted to Telios Pharmaceuticals Inc., Saniego, California, for providing and characterizing the recombinantecorin used in these studies. We are also grateful to Jonathonarlile for assistance with the histology. This work was funded by the

nternational Spinal Research Trust and the Wellcome Trust.

REFERENCES

1. Baghdassarian-Chalaye, D., D. Toru-Delbauffe, J. M. Gaveret,and M. Pierre. 1993. Effects of transforming growth factor b1 onthe extracellular matrix and cytoskeleton of cultured astrocytes.Glia 7: 193–202.

2. Border, W. A., and N. A. Noble. 1994. Natural inhibitor oftransforming growth factor b protects against scarring inexperimental kidney disease. Nature 360: 361–364.

3. Cui, W., and R. J. Ackhurst. 1996. Transforming growth factorbs: Biochemistry and biological activities in vivo and in vitro. InGrowth Factors in Health and Disease (D. LeRoith and C.Bondy, Eds.), pp. 319–356. JAI Press, Greenwich, CT.

4. Flanders, K., G. Ludecke, J. Renzing, C. Hamm, D. Cissel, andK. Unsicker. 1993. Effects of TGFb’s and bFGF on astroglial cellgrowth and gene expression in vitro. Mol. Cell. Neurosci. 4:406–417.

5. Gagelin, C., M. Pierre, J-M. Gaveret, and D. Toru-Delbauffe.1995. Rapid TGFb1 effects on actin cytoskeleton of astrocytes:Comparison with other factors and implications for cell motility.Glia 13: 283–293.

6. Giri, S. N., D. M. Hyde, R. K. Braun, W. Gaarde, J. R. Harper,and M. D. Pierschbacher. 1997. Antifibrotic effect of decorin in ableomycin hamster model of lung fibrosis. Biochem. Pharmacol.54: 1205–1216.

7. Heino, J., and J. Massague. 1989. Transforming growth factor bswitches the pattern of integrins expressed in MG-63 humanosteosarcoma cells and causes a selective loss of cell adhesion tolaminin. J. Biol. Chem. 36: 21806–21811.

8. Hurwitz, A., W. Lyman, and J. Berman. 1995. Tumour necrosisfactor alpha and transforming growth factor b upregulateastrocyte expression of monocyte chemoattractant protein-1. J.Neuroimmunol. 57: 193–198.

9. Ignotz, R. A., and J. Massague. 1989. Transforming growthfactor b stimulates the expression of fibronectin and collagenand their incorporation into the extracellular matrix. J. Biol.Chem. 261: 4337–4345.

0. Ignotz, R. A., J. Heino, and J. Massague. 1989. Regulation of celladhesion receptors by transforming growth factor b: Regulationof vitronectin receptor and LFA-1. J. Biol. Chem. 264: 389–392.

1. Isaka, Y., D. K. Brees, K. Ikegaya, Y. Kaneda, E. Imai, N. A.Noble, and W. A. Border. 1996. Gene therapy by skeletal muscleexpression of decorin prevents fibrotic disease in rat kidney.Nature Med. 2: 418–423.

2. Kerr, L. D., D. B. Miller, and L. M. Matrisian. 1990. TGFb1inhibition of transin/stromelysin gene expression is mediatedthrough a Fos binding sequence. Cell 61: 267–278.

3. Kitazawa, K., and T. Tada. 1994. Elevation of transforming
growth factor-b in cerebrospinal fluid of patients with communi-

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2


cating hydrocephalus after sub-arachnoid haemorrhage. Stroke25: 1400–1404.

4. Krupinski, J., P. Kumar, S. Kumar, and J. Kaluza. 1996.Increased expression of TGF-b in brain tissue after ischemicstroke in humans. Stroke 27: 852–857.

5. Labourdette, G., J. Thiery, P. Laeng, F. Perraud, D. Lawrence,and B. Pettman. 1990. Transforming growth factor b1 modu-lates the effects of basic fibroblast growth factor on growth andphenotypic expression of rat astroblasts in vitro. J. Cell Biol.144: 473–484.

6. Laiho, M., O. Saksela, P. A. Andreasen, and J. Keski-Oja. 1986.Enhanced production and extracellular deposition of the endo-thelial-type plasminogen activator inhibitor in cultured humanlung fibroblasts by transforming growth factor b. J. Cell Biol.103: 2403–2410.

7. Logan, A., and M. Berry. 1994. Cellular response of centralnervous system tissue to invasive therapeutic measures. Meth-ods Neurosci. 21: 3–19.

8. Logan, A., M. Berry, A. M. Gonzalez, S. A. Frautschy, M. B.Sporn, and A. Baird. 1994. Effects of transforming growth factorb1 on scar production in the injured central nervous system.Eur. J. Neurosci. 6: 355–363.

9. Logan, A., S. A. Frautschy, A. M. Gonzalez, M. B. Sporn, and A.Baird. 1992. Enhanced expression of transforming growth fac-tor b1 in the rat brain after a localised cerebral injury. BrainRes. 587: 216–225.

0. Logan, A., J. Green, A. Hunter, R. Jackson, and M. Berry. 1999.Inhibition of glial scarring in the injured rat brain by arecombinant human monoclonal antibody to transforminggrowth factor-b2. Eur. J. Neurosci. 11: 2367–2374.

1. Lund, L. R., A. Riccio, P. A. Andreasen, L. S. Nielsen, P.Kristensen, M. Laiho, O. Saksela, F. Blasi, and K. Dano. 1987.Transforming growth factor-b is a strong and fast acting posi-

tive regulator of the level of type-1 plasminogen activator-inhibitor mRNA in WI-38 human lung fibroblasts. EMBO J. 6:1281–1286.

2. Prat, B. M., and J. M. McPherson. 1997. TGF-b in the centralnervous system: Potential roles in ischemic injury and neurode-generative diseases. Cytokine Growth Factor Rev. 8: 267–292.

3. Stichel, C. C., J. Kappler, U. Junghans, A. Koops, H. Kresse, andH. W. Muller. 1995. Differential expression of the small chondroi-tin/dermatan sulfate proteoglycans decorin and biglycan afterinjury of the adult rat brain. Brain Res. 704(2): 263–274.

4. Toru-Delbauffe, D., D. Baghdassarian-Chalaye, J. M. Gaveret,F. Courtin, M. Pomerance, and M. Pierre. 1990. Effects oftransforming growth factor b1 on astroglial cells in culture. J.Neurochem. 54: 1056–1061.

5. Varga, J., J. Rosenbloom, and S. A. Jimenez. 1987. Transforminggrowth factor beta causes a persistant increase in steady-stateamounts for type I and type III collagen and fibronectin mRNAsin normal dermal human fibroblasts. Biochem. J. 297: 597–604.

6. Wahl, S. M., D. A. Hunt, L. M. Wakefield, N. McCartney-Francis, L. M. Wahl, A. B. Roberts, and M. B. Sporn. 1987.Transforming growth factor b induces monocyte chemotaxis andgrowth factor production. Proc. Natl. Acad. Sci. USA 84: 5788–5792.

7. Wrana, J. L., J. Sodek, R. L. Ber, and C. G. Bellows. 1986. Theeffects of platelet-derived transforming growth factor-b on nor-mal diploid gingival fibroblasts. Eur. J. Biochem. 159: 69–76.

8. Wyss-Coray, T., E. Masliah, M. Mallory, L. McConlogue, K.Johnson-Wood, C. Lin, and L. Mucke. 1997. Amyloidogenic roleof cytokine TGF-b1 in transgenic mice and in Alzheimer’sdisease. Nature 389: 603–606.

9. Yamaguchi, Y., D. M. Mann, and E. Ruoslahti. 1990. Negativeregulation of transforming growth factor-b by the proteoglycandecorin. Nature 346: 281–284.

Documents

Decorin Attenuates Gliotic Scar Formation in the Rat Cerebral Hemisphere