co-localization of immunoreactive transforming growth factor-beta1 and decorin in bronchial biopsies...

J. Pathol. 186: 410–415 (1998)

CO-LOCALIZATION OF IMMUNOREACTIVETRANSFORMING GROWTH FACTOR-BETA1 AND

DECORIN IN BRONCHIAL BIOPSIES FROMASTHMATIC AND NORMAL SUBJECTS

. 1*, . 2, . 1 . 1

1University Medicine, Southampton General Hospital, Southampton, U.K.2University Pathology, Southampton General Hospital, Southampton, U.K.

SUMMARY

Airway wall remodelling is an established pathological feature of asthma but its causes are not well understood. One cytokine ofpotential relevance is transforming growth factor-beta1 (TGF-â1). The immunolocalization of TGF-â1 and of its small bindingproteoglycan decorin have been examined in the airways of normal subjects and atopic asthmatics. Bronchial biopsy specimens wereobtained by fibreoptic bronchoscopy, processed into glycolmethacrylate resin, and stained immunohistochemically using specificantibodies. Immunoreactive TGF-â1 was principally localized extracellularly in association with subepithelial connective tissue. Somestaining of bronchial epithelial cells was also evident, but otherwise there was little intracellular staining. The overall pattern ofimmunohistochemical staining was indistinguishable in biopsy specimens from asthmatic and control subjects. Comparison of adjacentsections demonstrated the co-localization of immunoreactivity for TGF-â1 and decorin in the mucosa. It is concluded thatimmunoreactive TGF-â1 in human airways is principally extracellular and that matrix-associated TGF-â1 is likely to be bound at leastin part to decorin. This interaction may provide a reservoir of TGF-â1 that can be released in an active form in response to appropriatestimuli. ? 1998 John Wiley & Sons, Ltd.

KEY WORDS—transforming growth factor-beta; TGF-â1; decorin; asthma; bronchoscopy; biopsy

INTRODUCTION

Airway wall remodelling is an established pathologi-cal feature of asthma (reviewed in ref. 1). Morphometricstudies have demonstrated structural changes in theairway wall in this disease, affecting both large and smallairways and involving every tissue compartment, includ-ing the adventitia, submucosa, glands, and airwaysmooth muscle. Particular attention has focused on thelamina reticularis of the subepithelial basement mem-brane, which in asthma is greatly thickened by thedeposition of connective tissue proteins, principally col-lagen types III and V and fibronectin.2 Recent studieshave reported correlations between the depth of theconnective tissue layer and clinical and physiologicalindices of asthma severity.3,4

Little is known, however, about the mechanisms lead-ing to these structural alterations. In other chronicinflammatory diseases, the generation and release ofpotent growth and activating factors for fibroblasts areimportant, and transforming growth factor-beta(TGF-â) has emerged as one of a number of mediatorswhich have been implicated in repair following tissueinjury. The mammalian TGF-â family comprisesthree isoforms denoted TGF-â1, TGF-â2, and TGF-â3(reviewed in ref. 5). TGF-â1 is an extremely potentstimulus to formation of the extracellular matrix (ECM).In vitro, it increases the synthesis by fibroblasts of many

CCC 0022–3417/98/120410–06$17.50? 1998 John Wiley & Sons, Ltd.

components of the ECM, including collagen types I andIII, fibronectin, vitronectin, tenascin, and proteo-glycans.5 In addition, it decreases the synthesis ofenzymes that degrade the ECM, such as collagenaseand stromelysin, and increases the synthesis of inhibi-tors of these enzymes, including tissue inhibitorof metalloproteinase-1 (TIMP-1) and plasminogenactivator inhibitor type-1 (PAI-1).5

On account of these properties, we have hypothesizedthat TGF-â1 may be implicated in airway wall remod-elling in asthma. In support of this, we have recentlyreported that basal levels of this growth factor inbronchoalveolar lavage (BAL) fluid are increased inasthma and that these levels increase further in responseto allergen exposure.6 The principal aim of the presentstudy was to examine the immunohistochemical locali-zation of TGF-â1 in bronchial biopsy specimensobtained by fibreoptic bronchoscopy from healthy sub-jects and from atopic asthmatics. As receptors forTGF-â are expressed by almost all cells,7 it is believedthat mechanisms must exist to prevent unregulatedactivity of this growth factor in vivo. One means bywhich this is probably achieved is the rapid removal ofactive TGF-â1 from the extracellular milieu by seques-tration in the ECM. Many ECM components have thecapacity to bind TGF-â, but in this context binding todecorin is of particular relevance because this interactionhas been shown to neutralize the biological activity ofthe growth factor both in vitro8 and in vivo.9 A secondaim of the present study was, therefore, to examine thelocalization of immunoreactive decorin in bronchial

*Correspondence to: Dr A. E. Redington, Department of Respirat-ory Medicine, Second Floor, Thomas Guy House, Guy’s Hospital,London SE1 9RT, U.K. E-mail: [email protected]

Received 31 July 1997Revised 3 February 1998

Accepted 7 July 1998

411TGF-â AND DECORIN IN ASTHMA

biopsy specimens and its relation to immunoreactiveTGF-â1.

MATERIALS AND METHODS

Subjects

We studied four non-atopic non-asthmatic normalsubjects (two males and two females, mean age 20 years,range 19–21 years) and five subjects with clinicallystable, mildly symptomatic, atopic asthma (two malesand three females, mean age 34 years, range 20–54years). All asthmatics were receiving therapy withinhaled â2-agonists, but no subject in this group hadreceived treatment with inhaled or oral corticosteroidsfor at least 4 weeks before the study. Normal subjectshad no history of symptoms suggestive of asthma orrhinitis and were on no regular medication other thanoral contraceptives. All subjects were non-smokers andnone had experienced recent symptoms of upper respir-atory tract infection. Fibreoptic bronchoscopy was per-formed according to our previously published protocol10

and bronchial biopsy specimens were obtained fromproximal airway carinae. The study was approved by theSouthampton Joint University and Hospitals EthicsCommittee, and all subjects gave their written informedconsent.

Tissue processing and immunohistochemistry

Biopsy specimens were placed immediately into ice-cooled acetone solution containing 20 m iodoaceta-mide and 2 m phenyl methyl sulphonyl fluoride asprotease inhibitors, and fixed overnight at "20)C.The following day, biopsy specimens were embeddedin glycolmethacrylate (GMA) resin (Polysciences,Northampton, U.K.) as previously described.11 Tissueblocks were stored at "20)C in air-tight conditionsprior to analysis. Immunohistochemical staining wasperformed as previously described.11 Briefly, 2 ìm sec-tions were cut using an ultramicrotome (OM U3;Reichertz, Vienna, Austria), floated onto 0·2 per cent(v/v) ammonia solution for 1 min, transferred to poly--lysine-coated slides, and air-dried at room temperaturefor 1–4 h. Sections were pretreated with a solution of 0·1per cent sodium azide and 0·3 per cent hydrogen perox-ide for 30 min to inhibit endogenous peroxidase activity.After 3#5 min washes in Tris-buffered saline (TBS), pH7·6, blocking medium, consisting of Dulbecco’s modifiedEagle’s medium, 10 per cent fetal calf serum and 1 percent bovine serum albumin, was applied for 30 min.Sections were then incubated with the primary anti-bodies at previously determined dilutions. The primaryantibodies used were mouse IgG1 monoclonal antibody(MAb) TB21 directed against TGF-â1 (Serotec, Oxford,U.K.) and rabbit polyclonal antiserum against decorin(Chemicon International Inc. Temecula, CA, U.S.A.).The anti-decorin antiserum was raised against a15-amino acid peptide derived from the N-terminalsequence of human decorin and is reported to show nocross-reactivity with collagens I, III, IV or V, or

? 1998 John Wiley & Sons, Ltd.

fibronectin. After overnight incubation at room tem-perature, bound antibodies were labelled using bioti-nylated rabbit anti-mouse Fab fragments (Dako Ltd.)and the procedure was completed using a streptavidin–biotin–peroxidase detection system (Dako Ltd.). Amino-ethylcarbazole was used as a chromogen and sectionswere counterstained with Mayer’s haematoxylin. Adja-cent sections were examined in order to investigate therelationship between the immunohistochemical stainingpatterns obtained with the two antibodies. Controlexperiments were performed in which the primaryantibody was either omitted or replaced by an ir-relevant mouse IgG1 MAb (MOPC21; Sigma ChemicalCompany), or by non-immune rabbit serum (AnimalFacility, Southampton General Hospital), as appropri-ate, used at the same concentration as the correspondingprimary antibody.

RESULTS

No significant immunostaining was detected in con-trol sections in which the primary antibody was omittedor replaced with either mouse myeloma protein (Fig. 1a)or normal rabbit serum. In biopsy specimens fromboth asthmatic (Fig. 1b) and control (Fig. 1c) subjects,immunoreactive TGF-â1 was principally localized extra-cellularly in association with the subepithelial connectivetissue. Immunoreactivity was present throughout thedepth of this layer, apart from a zone immediately belowthe epithelial basement membrane where immuno-staining was either absent or less prominent. Somestaining of bronchial epithelial cells was evident (Fig.1c), but staining of the subepithelial basement mem-brane was rarely seen. There was little intracellularstaining of cells in the subepithelial tissue, althoughpericellular immunoreactivity was frequently seen inrelation to cells at this location (Fig. 2). The walls ofsmall blood vessels in the submucosal tissue were stainedstrongly (Fig. 3). Airway smooth muscle was absentfrom most biopsy specimens, but where present this wasnot immunostained (Fig. 4a). The overall pattern ofimmunohistochemical staining was indistinguishable inbiopsy specimens from asthmatic and control subjects.Immunoreactivity for decorin was evident throughoutthe subepithelial connective tissue in biopsy specimensfrom both control subjects and asthmatics. Immuno-staining of vessel walls and pericellular staining patternswere also evident, although these features were lessprominent than was the case with TGF-â1. Airwaysmooth muscle was not immunostained. Comparisonof adjacent sections demonstrated the co-localizationof immunoreactivity for TGF-â1 and decorin at sub-epithelial connective tissue sites (Figs 4a and 4b).

DISCUSSION

In the present study, we have shown that immuno-reactive TGF-â1 is present in bronchial biopsy speci-mens from atopic asthmatic and healthy controlsubjects. In both cases, a major site of localization was

J. Pathol. 186: 410–415 (1998)

412 A. E. REDINGTON ET AL.

the ECM, although immunostaining of bronchial epi-thelial cells was also observed. The pattern of stainingappeared to be specific, as no significant immuno-staining was detected in control sections in which theprimary MAb was either omitted or replaced by anirrelevant mouse IgG1 myeloma protein. Overall, theappearances were indistinguishable in biopsy specimensfrom asthmatic and control subjects.

The present findings may be compared with otherreports that have examined the expression of TGF-â inanimal and human airways. In rodents, immunohisto-chemical studies have localized immunoreactive TGF-â1


to bronchial epithelial cells and to the underlying con-nective tissue in large conducting airways.12,13 In thosestudies, TGF-â1 mRNA was found to be expressedprimarily in airway smooth muscle and in connectivetissue cells, but was not detected in the bronchial epi-thelium.12,13 In human airways, use has been made oftwo polyclonal antisera raised against TGF-â1, denotedLC(1–30) and CC(1–30), which react with intracellularand extracellular forms of the molecule, respectively.14,15

Using these two antisera, Aubert et al.16 localized LC(1–30) immunoreactivity to the bronchiolar epithelium, thesubmucosa, and airway smooth muscle in post-mortemand resected lung specimens. With CC(1–30), on theother hand, strong immunoreactivity was associatedwith the subepithelial connective tissue and airwaysmooth muscle, whereas the bronchial epithelium wasnot stained. In agreement with our present findings, nodifferences were evident between asthmatic and control

Fig. 1—Representative immunohistochemical staining patterns. (a) Control sections in which the primary MAb was replaced by mouse IgG1

myeloma protein. Sections of bronchial biopsy specimens from (b) a non-asthmatic subject and (c) an atopic asthmatic immunostained withanti-TGF-â1 MAb TB21. In both cases, immunoreactivity is predominantly extracellular and matrix-associated

Fig. 2—Immunohistochemical localization of TGF-â1 in a sectionfrom a control subject: immunostaining of bronchial epithelial cellsand pericellular staining in relation to cells in the subepithelial region(arrow-heads)

Fig. 3—Immunohistochemical localization of TGF-â1 in a sectionfrom an asthmatic subject: immunostaining of blood vessel walls andof adjacent connective tissue

J. Pathol. 186: 410–415 (1998)


subjects. In a recent report, Vignola et al.17 alsodescribed the presence of extracellular immunoreactiveTGF-â in bronchial biopsy specimens from asthmaticsubjects. On the other hand, Magnan et al.18,19 did notreport that extracellular TGF-â staining was a feature inlung tissue from either asthmatic or control subjects,although bronchial epithelial cells were immunostained,especially in specimens from non-asthmatic subjects.The differing results from these various studies mayrelate to differences in the subject populations examinedor in the epitopes recognized by the different antibodies


used, as well as to the principal focus of interest withinthe individual studies.

Based on their ability to synthesize TGF-â1 in vitro,most cellular constituents of the airway wall have thepotential to contribute to expression of this cytokine inthe airways, including bronchial epithelial cells,20 vascu-lar endothelial cells,21 macrophages,22 lymphocytes,23

fibroblasts,24 mast cells,25 and, principally in asthmatics,eosinophils.26 However, with the exception of bronchialepithelial cells, the present findings provide little infor-mation about the cellular origin of the immunoreactive

Fig. 4—Co-localization of (a) immunoreactive TGF-â1 and (b) immunoreactive decorin in adjacent sectionsfrom an asthmatic subject

J. Pathol. 186: 410–415 (1998)

414 A. E. REDINGTON ET AL.

TGF-â1 detected, as staining was predominantly extra-cellular and matrix-associated. TGF-â1 is synthesizedand stored within cells as a high-molecular-weight latentcomplex.27 The relative paucity of intracellular stainingin the present study may, therefore, have occurredbecause of masking of the immunoreactive epitopesrecognized by MAb TB21.

The anti-TGF-â1 MAb TB21 used in the presentstudy was raised against TGF-â1 isolated and purifiedfrom human platelets. It is known to recognize bothmonomeric and dimeric forms of mature human TGF-â1, but the question of whether it exhibits cross-reactivity with TGF-â2 or TGF-â3 has not beenaddressed (Serotec, personal communication). It istherefore possible that, in addition to TGF-â1, otherTGF-â isoforms may contribute to the overall immuno-staining detected in the present study. In rodent lung,studies using isoform-specific antibodies have indicatedthat all three TGF-â isoforms are present in proximalconducting airways and that they exhibit very similarpatterns of immunostaining.12,13

TGF-â1 has profound effects in relation to cellularproliferation and differentiation, integrity of the ECM,and immune function. The critical importance of con-trolled expression of this cytokine is illustrated by theeffects of targeted inactivation28 or overexpression29 ofthe TGF-â1 gene in mice, both of these disruptionsresulting in perinatal lethality. One mechanism bywhich the overactivity of TGF-â may be preventedis rapid removal of the free growth factor by sequestra-tion in the ECM. In vitro, TGF-â has been reported tobind to a number of ECM components includingfibronectin,30 thrombospondin,31 the basement mem-brane proteins collagen IV and laminin,32 and twoother small chondroitin–dermatan sulphate proteo-glycans, biglycan and fibromodulin.33 However, in thepresent study attention has been focused on theproteoglycan decorin, because of the ability of thismolecule to neutralize the biological activity ofTGF-â both in vitro8 and in vivo.9 In agreement withprevious studies of decorin expression in human tis-sues,34,35 immunoreactive decorin was localized to thesubepithelial connective tissue. The co-localization ofimmunoreactive TGF-â1 and decorin at these sites sug-gests that matrix-associated TGF-â is bound, at least inpart, to decorin and allows speculation that this inter-action provides a reservoir of TGF-â1 which may bereleased in an active form in response to appropriatestimuli. In the context of asthma, one stimulus ofrelevance could be the release of mast cell heparin,which is known to bind TGF-â136 and may thereforehave the potential to elute the growth factor from itsmatrix-bound state.

In conclusion, we have shown that immunoreactiveTGF-â1 is present in the airway wall in both atopicasthmatic and normal subjects. Immunoreactivity wasprincipally extracellular and there was co-localizationwith the small binding proteoglycan decorin. This inter-action may provide a reservoir of TGF-â1 which can bereleased in an active form in response to appropriatestimuli. More generally, this finding draws attention tothe potential importance of cytokine–matrix interactions


as factors regulating, either positively or negatively, theproperties of cytokines.

REFERENCES1. Redington AE, Howarth PH. Airway wall remodelling in asthma. Thorax

1997; 52: 310–312.2. Roche WR, Beasley R, Williams JH, Holgate ST. Subepithelial fibrosis in

the bronchi of asthmatics. Lancet 1989; i: 520–524.3. Chetta A, Foresi A, Del Donno M, et al. Bronchial responsiveness to

distilled water and methacholine and its relationship to inflammation andremodeling of the airways in asthma. Am J Respir Crit Care Med 1996; 153:910–917.

4. Chetta A, Foresi A, Del Donno M, Bertorelli G, Pesci A, Olivieri D.Airways remodeling is a distinctive feature of asthma and is related toseverity. Chest 1997; 111: 852–857.

5. Massagué J The transforming growth factor-â family. Annu Rev Cell Biol1990; 6: 597–641.

6. Redington AE, Madden J, Frew AJ, et al. Transforming growth factor-beta1 in asthma: measurement in bronchoalveolar lavage fluid. Am J Respir CritCare Med 1997; 156: 642–647.

7. Wakefield LM, Smith DM, Masui T, Harris CC, Sporn MB. Distributionand modulation of the cellular receptor for transforming growth factor-beta. J Cell Biol 1987; 105: 965–975.

8. Yamaguchi Y, Mann DM, Ruoslahti E. Negative regulation of transform-ing growth factor-â by the proteoglycan decorin. Nature 1990; 346: 281–284.

9. Border WA, Noble NA, Yamamoto T, et al. Natural inhibitor of trans-forming growth factor-â against scarring in experimental kidney disease.Nature 1992; 360: 361–364.

10. Djukanovic R, Wilson JW, Lai CKW, Holgate ST, Howarth PH. The safetyaspects of fiberoptic bronchoscopy, bronchoalveolar lavage, and endo-bronchial biopsy in asthma. Am Rev Respir Dis 1991; 143: 772–777.

11. Britten KM, Howarth PH, Roche WR. Immunohistochemistry on resinsections: a comparison of resin embedding techniques for small mucosalbiopsies. Biotech Histochem 1993; 68: 271–280.

12. Pelton RW, Johnson MD, Perkett EA, Gold LI, Moses HL. Expression oftransforming growth factor-â1, -â2, and -â3 mRNA and protein in themurine lung. Am J Respir Cell Mol Biol 1991; 5: 522–530.

13. Santana A, Saxena B, Noble NA, Gold LI, Marshall BC. Increasedexpression of transforming growth factor â isoforms (â1, â2, â3) inbleomycin-induced pulmonary fibrosis. Am J Respir Cell Mol Biol 1995; 13:34–44.

14. Flanders KC, Thompson NL, Cissel DS, et al. Transforming growthfactor-â1: histochemical localization with antibodies to different epitopes.J Cell Biol 1989; 108: 653–660.

15. Thompson NL, Flanders KC, Smith JM, Ellingsworth LR, Roberts AB,Sporn MB. Expression of transforming growth factor-â1 in specific cells andtissues of adult and neonatal mice. J Cell Biol 1989; 108: 661–669.

16. Aubert J-D, Dalal BI, Bai TR, Roberts CR, Hayashi S, Hogg JC.Transforming growth factor â1 gene expression in human airways. Thorax1994; 49: 225–232.

17. Vignola AM, Chanez P, Chiappara G, et al. Transforming growth factor-âexpression in mucosal biopsies in asthma and chronic bronchitis. Am JRespir Crit Care Med 1997; 156: 591–599.

18. Magnan A, Frachon I, Rain B, et al. Transforming growth factor â innormal human lung: preferential location in bronchial epithelial cells.Thorax 1994; 49: 789–792.

19. Magnan A, Retornaz F, Tsicopoulos A, et al. Altered compartmentalizationof transforming growth factor-â in asthmatic airways. Clin Exp Allergy1997; 27: 389–395.

20. Sacco O, Romberger D, Rizzino A, Beckmann JD, Rennard SI, SpurzemJR. Spontaneous production of transforming growth factor-â2 by primarycultures of bronchial epithelial cells: effects on cell behavior in vitro. J ClinInvest 1992; 90: 1379–1385.

21. Phan SH, Gharaee-Kermani M, Wolber F, Ryan US. Stimulation of ratendothelial cell transforming growth factor-â production by bleomycin.J Clin Invest 1991; 87: 148–154.

22. Assoian RK, Fleurdelys BE, Stevenson HC, Miller PJ, Matdes DK, RainesEW. Expression and secretion of type â transforming growth factor byactivated human macrophages. Proc Natl Acad Sci USA 1987; 84: 6020–6024.

23. Kehrl JH, Wakefield LM, Roberts AB, et al. Production of transforminggrowth factor â by human T lymphocytes and its potential role in theregulation of T cell growth. J Exp Med 1986; 163: 1037–1050.

24. Kelley J, Fabisiak JP, Hawes K, Absher M. Cytokine signalling in the lung:transforming growth factor-â secretion by lung fibroblasts. Am J Physiol1991; 260: L123–L128.

25. Pennington DW, Lopez AR, Thomas PS, Peck C, Gold WM. Dogmastocytoma cells produce transforming growth factor beta1. J Clin Invest1992; 90: 35–41.

26. Wong DTW, Elovic A, Matossian K, et al. Eosinophils from patients withblood eosinophilia express transforming growth factor â1. Blood 1991; 78:2702–2707.

J. Pathol. 186: 410–415 (1998)


27. Wakefield LM, Smith DM, Flanders KC, Sporn MB. Latent transforminggrowth factor-â from human platelets: a high molecular weight complexcontaining precursor sequences. J Biol Chem 1988; 263: 7646–7654.

28. Shull MM, Ormsby I, Kier AB, et al. Targeted disruption of the mousetransforming growth factor-â1 gene results in multifocal inflammatorydisease. Nature 1992; 359: 693–699.

29. Zhou L, Dey CR, Wert SE, Whitsett JA. Arrested lung morphogenesis intransgenic mice bearing an SP-C-TGF-â1 chimeric gene. Dev Biol 1996; 175:227–238.

30. Fava RA, McClure DB. Fibronectin-associated transforming growth factor.J Cell Physiol 1987; 131: 184–189.

31. Murphy-Ullrich JE, Schultz-Cherry S, Höök M. Transforming growthfactor-â complexes with thrombospondin. Mol Biol Cell 1992; 3: 181–188.

32. Paralkar VM, Vukicevic S, Reddi AH. Transforming growth factor â type 1binds to collagen IV of basement membrane matrix: implications fordevelopment. Dev Biol 1991; 143: 303–308.


33. Hildebrand A, Romaris M, Rasmussen LM, et al. Interaction of the smallinterstitial proteoglycans biglycan, decorin and fibromodulin with trans-forming growth factor â. Biochem J 1994; 302: 527–534.

34. Voss B, Glössl J, Cully Z, Kresse H. Immunocytochemical investigation onthe distribution of small chondroitin sulfate–dermatan sulfate proteoglycanin the human. J Histochem Cytochem 1986; 34: 1013–1019.

35. Bianco P, Fisher LW, Young MF, Termine JD, Robey PG. Expression andlocalization of the two small proteoglycans biglycan and decorin in devel-oping human skeletal and non-skeletal tissues. J Histochem Cytochem 1990;38: 1549–1563.

36. McCaffrey TA, Falcone DJ, Du B. Transforming growth factor-â1 is aheparin-binding protein: identification of putative heparin-binding regionsand isolation of heparins with varying affinity for TGF-â1. J Cell Physiol1992; 152: 430–440.

J. Pathol. 186: 410–415 (1998)

co-localization of immunoreactive transforming growth factor-beta1 and decorin in bronchial biopsies...

Documents