Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

Expression of Small Extracellular Chondroitin/Dermatan Sulfate Proteoglycans Is Differentially Regulated in Human Endothelial Cells

1997; Elsevier BV; Volume: 272; Issue: 19 Linguagem: Inglês

10.1074/jbc.272.19.12730

ISSN

1083-351X

Autores

Lassi Nelimarkka, Varpu Kainulainen, Elke Schönherr, Susanna Moisander, Matti Jortikka, Mikko J. Lammi, Klaus Elenius, Markku Jalkanen, Hannu Järveläinen,

Tópico(s)

Cell Adhesion Molecules Research

Resumo

We have examined the expression of the small extracellular chondroitin/dermatan sulfate proteoglycans (CS/DS PGs), biglycan, decorin, and PG-100, which is the proteoglycan form of colony stimulating factor-1, in the human endothelial cell line EA.hy 926. We have also examined whether modulation of the phenotype of EA.hy 926 cells by tumor necrosis factor-α (TNF-α) is associated with specific changes in the synthesis of these PGs. We demonstrate that EA.hy 926 cells, when they form monolayer cultures typical of macrovascular endothelial cells, express and synthesize detectable amounts of biglycan and PG-100, but not decorin. On SDS-polyacrylamide gel electrophoresis both PGs behave like proteins of the relative molecular weight of ∼250,000. TNF-α that changed the morphology of the cells from a polygonal shape into a spindle shape and that also stimulated the detachment of the cells from culture dish, markedly decreased the net synthesis of biglycan, whereas the net synthesis of PG-100 was increased. These changes were parallel with those observed at the mRNA level of the corresponding PGs. The proportions of the different sulfated CS/DS disaccharide units of PGs were not affected by TNF-α. Several other growth factors/cytokines, such as interferon-γ, fibroblast growth factors-2 (FGF-2) and -7 (FGF-7), interleukin-1β, and transforming growth factor-β, unlike TNF-α, modulated neither the morphology nor the biglycan expression of EA.hy 926 cells under the conditions used in the experiments. However, PG-100 expression was increased also in response to FGF-2 and -7 and transforming growth factor-β. None of the above cytokines, including TNF-α, was able to induce decorin expression in the cells. Our results indicate that the regulatory elements controlling the expression of the small extracellular CS/DS PGs in human endothelial cells are different. We have examined the expression of the small extracellular chondroitin/dermatan sulfate proteoglycans (CS/DS PGs), biglycan, decorin, and PG-100, which is the proteoglycan form of colony stimulating factor-1, in the human endothelial cell line EA.hy 926. We have also examined whether modulation of the phenotype of EA.hy 926 cells by tumor necrosis factor-α (TNF-α) is associated with specific changes in the synthesis of these PGs. We demonstrate that EA.hy 926 cells, when they form monolayer cultures typical of macrovascular endothelial cells, express and synthesize detectable amounts of biglycan and PG-100, but not decorin. On SDS-polyacrylamide gel electrophoresis both PGs behave like proteins of the relative molecular weight of ∼250,000. TNF-α that changed the morphology of the cells from a polygonal shape into a spindle shape and that also stimulated the detachment of the cells from culture dish, markedly decreased the net synthesis of biglycan, whereas the net synthesis of PG-100 was increased. These changes were parallel with those observed at the mRNA level of the corresponding PGs. The proportions of the different sulfated CS/DS disaccharide units of PGs were not affected by TNF-α. Several other growth factors/cytokines, such as interferon-γ, fibroblast growth factors-2 (FGF-2) and -7 (FGF-7), interleukin-1β, and transforming growth factor-β, unlike TNF-α, modulated neither the morphology nor the biglycan expression of EA.hy 926 cells under the conditions used in the experiments. However, PG-100 expression was increased also in response to FGF-2 and -7 and transforming growth factor-β. None of the above cytokines, including TNF-α, was able to induce decorin expression in the cells. Our results indicate that the regulatory elements controlling the expression of the small extracellular CS/DS PGs in human endothelial cells are different. Proteoglycans (PGs) 1The abbreviations used are: PG(s), proteoglycan(s); BAE, bovine aortic endothelial; CPC, cetylpyridinium chloride; CS, chondroitin sulfate; CSF-1, colony stimulating factor-1; DS, dermatan sulfate; FCS, fetal calf serum; FGF, fibroblast growth factor; GAG, glycosaminoglycan; HS, heparan sulfate; HSF, human skin fibroblast; IFN-γ, interferon-γ; IL-1β, interleukin-1β; PAGE, polyacrylamide gel electrophoresis; TNF-α, tumor necrosis factor-α; TGF-β, transforming growth factor-β. 1The abbreviations used are: PG(s), proteoglycan(s); BAE, bovine aortic endothelial; CPC, cetylpyridinium chloride; CS, chondroitin sulfate; CSF-1, colony stimulating factor-1; DS, dermatan sulfate; FCS, fetal calf serum; FGF, fibroblast growth factor; GAG, glycosaminoglycan; HS, heparan sulfate; HSF, human skin fibroblast; IFN-γ, interferon-γ; IL-1β, interleukin-1β; PAGE, polyacrylamide gel electrophoresis; TNF-α, tumor necrosis factor-α; TGF-β, transforming growth factor-β. are complex macromolecules that consist of a protein core and one or more glycosaminoglycan (GAG) side chains, covalently bound to the core protein (1Kjellen L. Lindahl U. Annu. Rev. Biochem. 1991; 60: 443-475Crossref PubMed Scopus (1658) Google Scholar). Nearly 30 individual PGs with various functions residing in the extracellular matrix, on the cell surface, or inside the cells, have been identified (1Kjellen L. Lindahl U. Annu. Rev. Biochem. 1991; 60: 443-475Crossref PubMed Scopus (1658) Google Scholar, 2Elenius K. Jalkanen M. J. Cell Sci. 1994; 107: 2975-2982PubMed Google Scholar, 3Gressner A.M. Eur. J. Clin. Chem. Clin. Biochem. 1994; 32: 225-237PubMed Google Scholar, 4Iozzo R.V. Murdoch A.D. FASEB J. 1996; 10: 598-614Crossref PubMed Scopus (544) Google Scholar). Among the PGs of the extracellular matrix, four small chondroitin/dermatan sulfate (CS/DS) PG species, biglycan (5Fisher L.W. Termine J.D. Young M.F. J. Biol. Chem. 1989; 264: 4571-4576Abstract Full Text PDF PubMed Google Scholar), decorin (6Krusius T. Ruoslahti E. Proc. Natl. Acad. Sci U. S. A. 1986; 83: 7683-7687Crossref PubMed Scopus (408) Google Scholar), PG-100 (7Schwarz K. Breuer B. Kresse H. J. Biol. Chem. 1990; 265: 22023-22028Abstract Full Text PDF PubMed Google Scholar), and epiphycan (4Iozzo R.V. Murdoch A.D. FASEB J. 1996; 10: 598-614Crossref PubMed Scopus (544) Google Scholar, 8Shinomura T. Kimata K. J. Biol. Chem. 1992; 267: 1265-1270Abstract Full Text PDF PubMed Google Scholar, 9Shinomura T. Kimata K. Oike Y. Maeda N. Yano S. Suzuki S. Dev. Biol. 1984; 103: 211-220Crossref PubMed Scopus (42) Google Scholar) are currently known. Two of these PGs, biglycan and decorin, are highly homologous with each other in terms of their core protein structure (5Fisher L.W. Termine J.D. Young M.F. J. Biol. Chem. 1989; 264: 4571-4576Abstract Full Text PDF PubMed Google Scholar), and together with epiphycan, fibromodulin, and lumican they both belong to the small leucine-rich PG gene family (4Iozzo R.V. Murdoch A.D. FASEB J. 1996; 10: 598-614Crossref PubMed Scopus (544) Google Scholar). With some exceptions biglycan and decorin differ in the number of GAG chains attached to their core proteins. Biglycan is usually substituted with two GAG chains, whereas decorin typically has only one GAG chain (10Choi H.U. Johnson T.L. Pal S. Tang L.-H. Rosenberg L. Neame P.J. J. Biol. Chem. 1989; 264: 2876-2884Abstract Full Text PDF PubMed Google Scholar, 11Neame P.J. Choi H.U. Rosenberg L.C. J. Biol. Chem. 1989; 264: 8653-8661Abstract Full Text PDF PubMed Google Scholar, 12Mann D.M. Yamaguchi Y. Bourdon M.A. Ruoslahti E. J. Biol. Chem. 1990; 265: 5317-5323Abstract Full Text PDF PubMed Google Scholar). On sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) biglycan has variously been shown to migrate at positions of protein molecular weight markers of ∼200,000 or more, while decorin migrates at positions of protein molecular weight markers of 87,000–180,000 (13Rosenberg L.C. Choi H.U. Tang L.-H. Johnson T.L. Pal S. Webber C. Reiner A. Poole A.R. J. Biol. Chem. 1985; 260: 6304-6313Abstract Full Text PDF PubMed Google Scholar, 14Roughley P.J. White R.J. Biochem. J. 1989; 262: 823-827Crossref PubMed Scopus (77) Google Scholar, 15Schmidt A. Biochem. Soc. Transact. 1989; 17: 19-20Crossref PubMed Scopus (2) Google Scholar, 16Järveläinen H.T. Kinsella M.G. Wight T.N. Sandell L.J. J. Biol. Chem. 1991; 266: 23274-23281Abstract Full Text PDF PubMed Google Scholar, 17Schönherr E. Järveläinen H.T. Kinsella M.G. Sandell L.J. Wight T.N. Arterioscler. Thromb. 1993; 13: 1026-1036Crossref PubMed Google Scholar). The third small extracellular CS/DS PG species, originally isolated from the culture medium of human osteosarcoma cells and tentatively named PG-100 (7Schwarz K. Breuer B. Kresse H. J. Biol. Chem. 1990; 265: 22023-22028Abstract Full Text PDF PubMed Google Scholar), has recently been demonstrated to be identical with the macrophage colony-stimulating factor CSF-1 (18Partenheimer A. Schwarz K. Wrocklage C. Kölsch E. Kresse H. J. Immunol. 1995; 155: 5557-5565PubMed Google Scholar). PG-100 is therefore a PG form of this cytokine (19Suzu S. Ohtsuki T. Makishima M. Yanai N. Kawashima T. Nagata N. Motoyoshi K. J. Biol. Chem. 1992; 267: 16812-16815Abstract Full Text PDF PubMed Google Scholar, 20Price L.K.H. Choi H.U. Rosenberg L. Stanley E.R. J. Biol. Chem. 1992; 267: 2190-2199Abstract Full Text PDF PubMed Google Scholar), and it can be called a "part-time" PG to be differentiated from biglycan, decorin, and other PGs that usually exist only in the GAG containing form (21Ruoslahti E. J. Biol. Chem. 1989; 264: 13369-13372Abstract Full Text PDF PubMed Google Scholar). The fourth small CS/DS PG, epiphycan, has been shown to be closely related to osteoinductive factor (22Madisen L. Neubauer M. Plowman G. Rosen D. Segarini P. Dasch J. Thompson A. Ziman J. Bentz H. Purchio A.F. DNA Cell Biol. 1990; 9: 303-309Crossref PubMed Scopus (75) Google Scholar), and the expression of this PG seems to be cartilage specific (8Shinomura T. Kimata K. J. Biol. Chem. 1992; 267: 1265-1270Abstract Full Text PDF PubMed Google Scholar). The exact functions of the small CS/DS PGs are still somewhat controversial. Decorin, the most thoroughly investigated molecule in this group, has been shown to bind to collagen fibrils and to regulate fibril formation (23Scott J.E. Orford C.R. Biochem. J. 1981; 197: 213-216Crossref PubMed Scopus (330) Google Scholar, 24Vogel K.G. Paulsson M. Heinegård D. Biochem. J. 1984; 223: 587-597Crossref PubMed Scopus (696) Google Scholar, 25Scott J.E. Biochem. J. 1988; 252: 313-323Crossref PubMed Scopus (537) Google Scholar, 26Schönherr E. Hausser H. Beaven L. Kresse H. J. Biol. Chem. 1995; 270: 8877-8883Abstract Full Text Full Text PDF PubMed Scopus (180) Google Scholar). Decorin has also been found to interact with the cytokine transforming growth factor-β (TGF-β) and to be involved in cell proliferation and differentiation (27Yamaguchi Y. Ruoslahti E. Nature. 1988; 336: 244-246Crossref PubMed Scopus (206) Google Scholar, 28Yamaguchi Y. Mann D.M. Ruoslahti E. Nature. 1990; 346: 281-284Crossref PubMed Scopus (1282) Google Scholar, 29Santra M. Skorski T. Calabretta B. Lattime E.C. Iozzo R.V. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 7016-7020Crossref PubMed Scopus (205) Google Scholar). The role of biglycan in these processes is obscure, although it has been shown to bind to these same molecules (30Hildebrand A. Romaris M. Rasmussen L.M. Heinegård D. Twardzik D.R. Border W.A. Ruoslahti E. Biochem. J. 1994; 302: 527-534Crossref PubMed Scopus (849) Google Scholar, 31Kresse H. Hausser H. Schönherr E. Bittner K. Eur. J. Clin. Chem. Clin. Biochem. 1994; 32: 259-264PubMed Google Scholar, 32Schönherr E. Witsch-Prehm P. Harrach B. Robenek H. Rauterberg J. Kresse H. J. Biol. Chem. 1995; 270: 2776-2783Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar). The pericellular localization of biglycan (33Fleischmajer R. Fisher L.W. MacDonald E.D. Jacobs Jr., L. Perlish J.S. Termine J.D. J. Struct. Biol. 1991; 106: 82-90Crossref PubMed Scopus (150) Google Scholar) as well as its ability to bind to fibronectin (34Bidanset D.J. LeBaron R. Rosenberg L. Murphy-Ullrich J.E. Hook M. J. Cell Biol. 1992; 118: 1523-1531Crossref PubMed Scopus (90) Google Scholar) suggest that biglycan is likely to interfere with cell adhesion (34Bidanset D.J. LeBaron R. Rosenberg L. Murphy-Ullrich J.E. Hook M. J. Cell Biol. 1992; 118: 1523-1531Crossref PubMed Scopus (90) Google Scholar). There is also recent evidence to suggest that biglycan is associated with cell migration (35Kinsella M.G. Tsoi C.K. Järveläinen H.T. Wight T.N. J. Biol. Chem. 1997; 272: 318-325Abstract Full Text Full Text PDF PubMed Scopus (93) Google Scholar). CSF-1 is known to control the growth and differentiation of mononuclear phagocytes (36Clark S.C. Kamen R. Science. 1987; 236: 1229-1237Crossref PubMed Scopus (1181) Google Scholar) as well as the growth of several other cell types, including endothelial cells (37Detmar M. Tenorio S. Hettmannsperger U. Ruszczak Z. Orfanos C.E. J. Invest. Dermatol. 1992; 98: 147-153Crossref PubMed Scopus (105) Google Scholar). However, PG-100 exhibits less than 1% of the biological activity of mature CSF-1 (18Partenheimer A. Schwarz K. Wrocklage C. Kölsch E. Kresse H. J. Immunol. 1995; 155: 5557-5565PubMed Google Scholar) and should perhaps be considered an extracellular storage form of this cytokine (19Suzu S. Ohtsuki T. Makishima M. Yanai N. Kawashima T. Nagata N. Motoyoshi K. J. Biol. Chem. 1992; 267: 16812-16815Abstract Full Text PDF PubMed Google Scholar). We have earlier shown that the principal small extracellular CS/DS PG species synthesized by bovine aortic endothelial (BAE) cells in monolayer cultures is biglycan (16Järveläinen H.T. Kinsella M.G. Wight T.N. Sandell L.J. J. Biol. Chem. 1991; 266: 23274-23281Abstract Full Text PDF PubMed Google Scholar). This result has been shown to be true for human umbilical vein endothelial cell monolayers as well (16Järveläinen H.T. Kinsella M.G. Wight T.N. Sandell L.J. J. Biol. Chem. 1991; 266: 23274-23281Abstract Full Text PDF PubMed Google Scholar). We have also shown that BAE cell monolayer cultures do not express detectable amounts of either decorin or type I collagen (16Järveläinen H.T. Kinsella M.G. Wight T.N. Sandell L.J. J. Biol. Chem. 1991; 266: 23274-23281Abstract Full Text PDF PubMed Google Scholar). However, when BAE cells change their phenotype and form so called sprouting cultures, the synthesis of both decorin and type I collagen is initiated (38Iruela-Arispe M.L. Hasselaar P. Sage E.H. Lab. Invest. 1991; 64: 174-186PubMed Google Scholar, 39Järveläinen H.T. Iruela-Arispe M.L. Kinsella M.G. Sandell L.J. Sage E.H. Wight T.N. Exp. Cell Res. 1992; 203: 395-401Crossref PubMed Scopus (0) Google Scholar) indicating that the synthesis of decorin and type I collagen is associated with an in vitro angiogenesis phenomenon. In this study we have examined the expression of the small extracellular CS/DS PGs in the permanent human endothelial cell line EA.hy 926 (40Edgell C-J.S. McDonald C.C. Graham J.B. Proc. Natl. Acad. Sci. U. S. A. 1983; 80: 3734-3737Crossref PubMed Scopus (1343) Google Scholar). Previously this cell line has been shown to possess several characteristics typical of endothelial cells (40Edgell C-J.S. McDonald C.C. Graham J.B. Proc. Natl. Acad. Sci. U. S. A. 1983; 80: 3734-3737Crossref PubMed Scopus (1343) Google Scholar, 41Debault L.E. Esmon N.L. Esmon C.T. Edgell C-J.S. Fed. Proc. 1984; 43: 783Google Scholar, 42Bussolino F. Wang J.M. Defilippi P. Turrini F. Sanavio F. Edgell C.-J.S. Aglietta M. Arese P. Mantovani A. Nature. 1986; 337: 471-473Crossref Scopus (544) Google Scholar, 43Suggs J.E. Madden M.C. Friedman M. Edgell C.J. Blood. 1986; 68: 825-829Crossref PubMed Google Scholar, 44Emeis J.J. Edgell C.J. Blood. 1988; 71: 1669-1675Crossref PubMed Google Scholar, 45Saijonmaa O. Nyman T. Hohenthal U. Fyhrquist F. Biochem. Biophys. Res. Commun. 1991; 181: 529-536Crossref PubMed Scopus (41) Google Scholar, 46Corder R. Khan N. Harrison V.J. Biochem. Biophys. Res. Commun. 1995; 207: 355-362Crossref PubMed Scopus (32) Google Scholar). However, the production of extracellular matrix molecules, including PGs, of EA.hy 926 cells has remained unidentified. Furthermore, we have examined whether modulation of the phenotype of EA.hy 926 cells is associated with specific changes in the synthesis of the small extracellular CS/DS PGs. We demonstrate that EA.hy 926 cells express biglycan, but not decorin. We also demonstrate that these cells synthesize PG-100, the M r of which (∼250,000) is similar to that of biglycan synthesized by these same cells. In addition, the inhibitory effect of TNF-α on biglycan expression and its stimulatory effect on PG-100 expression by EA.hy 926 cells, associated with the altered morphology and adhesion of the cells, are shown. Cells of the permanent human endothelial cell line EA.hy 926 (40Edgell C-J.S. McDonald C.C. Graham J.B. Proc. Natl. Acad. Sci. U. S. A. 1983; 80: 3734-3737Crossref PubMed Scopus (1343) Google Scholar) were grown on plastic dishes (Nunc, Roskilde, Denmark) at 37 °C in 95% air + 5% CO2 in bicarbonate-buffered Dulbecco's modified Eagle's medium (Flow Laboratories, Irvine, United Kingdom) containing 100 IU/ml of penicillin (Oriola, Finland) and 50 μg/ml streptomycin (Oriola), and supplemented with 10% (v/v) of fetal calf serum (FCS, Life Technologies, Inc., Paisley, Scotland), and 1 × HAT (100 μm hypoxanthine, 0.4 μm aminopterin, 16 μm thymidine; Life Technologies, Inc.). For subcultures, the cells were detached with 0.01% (w/v) trypsin (Sigma). Human skin fibroblasts (HSFs) that were used as control cells were derived from a skin biopsy of a healthy person, and these cells were subcultured as described previously (47Näntö-Salonen K. Penttinen R. J. Inherited Metab. Dis. 1982; 5: 197-203Crossref PubMed Scopus (12) Google Scholar). The culture medium used for HSFs was the same as described above for EA.hy 926 cells with the exception that HAT was omitted. For identification of sulfated PGs in the culture media, confluent cell cultures were labeled for 24 h with 50 μCi/ml Na2[35S]O4 (Amersham, United Kingdom) in sulfate-free Dulbecco's modified Eagle's medium supplemented with 10% (v/v) FCS, whereafter the media were collected and the detached cells were removed by low speed centrifugation (2000 rpm) for 5 min. The incorporation of [35S]sulfate into newly synthesized PGs was evaluated by the cetylpyridinium chloride (CPC) precipitation assay (48Rapraeger A. Bernfield M. J. Biol. Chem. 1985; 260: 4103-4109Abstract Full Text PDF PubMed Google Scholar) before further analyses of the culture media (see below). In the experiments examining the effect of TNF-α on the synthesis of the small extracellular CS/DS PGs, the cells were grown to confluence in the serum containing culture medium. Next, the cells were kept in the serum-free culture medium for 4 h, whereafter the cells were labeled as described above in serum-free culture medium containing TNF-α at various concentrations. After removing the detached cells from the culture media by low speed centrifugation (see above), the amount of [35S]sulfate-labeled PGs in the samples was determined by the CPC precipitation assay (48Rapraeger A. Bernfield M. J. Biol. Chem. 1985; 260: 4103-4109Abstract Full Text PDF PubMed Google Scholar) before further analyses. Cell number calculations were performed with a hemocytometer. Before the calculations, the cells were fixed with a formaldehyde containing buffer solution (10%, v/v, of 37% formaldehyde in 0.085 m NaCl and 0.1 mNa2 SO4). Two aliquots of each cell suspension were taken for the cell counting. The viability of the cells in the control and TNF-α treated cultures was measured using the trypan blue exclusion test as described (49Järveläinen H.T. Pelliniemi T.T. Rönnemaa T. Scand. J. Clin Lab. Invest. 1985; 45: 223-228Crossref PubMed Scopus (14) Google Scholar). For the identification of CS/DS and heparan sulfate (HS) PGs in the culture media, equal volumes of [35S]sulfate-labeled media were ethanol-precipitated, dried, dissolved into appropriate buffers, and digested with 0.1 IU/ml of chondroitin ABC lyase from Proteus vulgaris(Seikagaku Co., Tokyo, Japan) or 6 milliunits/ml of heparitinase from Flavobacterium heparinum (Seikagaku Co.), respectively (50Jalkanen M. Nguyen H. Rapraeger A. Kurn N. Bernfield M. J. Cell Biol. 1985; 105: 3087-3096Crossref Scopus (116) Google Scholar, 51Rapraeger A. Jalkanen M. Endo E. Koda J. Bernfield M. J. Biol. Chem. 1985; 260: 11046-11052Abstract Full Text PDF PubMed Google Scholar). The immunoprecipitation of biglycan and PG-100 from the culture media with polyclonal antisera against these two PGs (see below) was performed as described previously (52Santala P. Larjava H. Nissinen L. Riikonen T. Määttä A. Heino J. J. Biol. Chem. 1994; 269: 1276-1283Abstract Full Text PDF PubMed Google Scholar). Aliquots of [35S]sulfate-labeled media corresponding to 500,000 cpm as determined by the CPC precipitation assay (48Rapraeger A. Bernfield M. J. Biol. Chem. 1985; 260: 4103-4109Abstract Full Text PDF PubMed Google Scholar) were taken for the analysis. The immunoprecipitants were run on a gradient SDS-PAGE as described (see below). Partial purification of PGs by ion exchange chromatography was performed by applying equal aliquots of [35S]sulfate-labeled media to a DEAE-Sephacel (Pharmacia, Uppsala, Sweden) column equilibrated with 0.2 m NaCl in the buffer containing 50 mm sodium acetate, 2.0 murea, 1 mm phenylmethylsulfonyl fluoride, and 0.1% Triton X-100. The column was eluted with a linear gradient of 0.2–1.0m NaCl in the same buffer as above. The amount of radioactivity in sulfated PGs in each fraction was determined by the CPC precipitation assay (48Rapraeger A. Bernfield M. J. Biol. Chem. 1985; 260: 4103-4109Abstract Full Text PDF PubMed Google Scholar). All fractions that contained [35S]sulfate-labeled PGs were pooled before further analyses. SDS-PAGE of the samples was carried out essentially as described by Laemmli (53Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (205487) Google Scholar) on a 4–12% linear gradient gel with a 3% stacking gel (16Järveläinen H.T. Kinsella M.G. Wight T.N. Sandell L.J. J. Biol. Chem. 1991; 266: 23274-23281Abstract Full Text PDF PubMed Google Scholar). The positions of the radioactive bands were visualized by fluorography of dried gels previously treated with 2,5-diphenyloxazole (Fisons, UK).14C-Methylated protein molecular weight standards (Amersham) were used to estimate the average sizes of [35S]sulfate-labeled macromolecules. [35S]Sulfate-labeled medium samples or partially purified PGs from similar medium samples were lyophilized, whereafter they were digested with chondroitin ABC lyase and heparitinase to remove all GAG chains (see above). Next, SDS-PAGE was carried out (see above) under reducing conditions using a 12.5% gel, whereafter the proteins were electroblotted to a nitrocellulose membrane (Schleicher & Schuell, Dassel, Germany). Specific bands were detected by immunostaining (18Partenheimer A. Schwarz K. Wrocklage C. Kölsch E. Kresse H. J. Immunol. 1995; 155: 5557-5565PubMed Google Scholar, 54Hausser H. Hoppe W. Rauch U. Kresse H. Biochem. J. 1989; 263: 137-142Crossref PubMed Scopus (42) Google Scholar) using polyclonal antisera against biglycan and PG-100 (see below). Biglycan antiserum was a polyclonal antiserum made in a rabbit against a synthetic peptide of human biglycan (amino acids 11–24) that was conjugated to bovine serum albumin before injections (5Fisher L.W. Termine J.D. Young M.F. J. Biol. Chem. 1989; 264: 4571-4576Abstract Full Text PDF PubMed Google Scholar). This antiserum, called LF-51, was kindly provided by Dr. L. Fisher (National Institute of Dental Research, Bethesda, MD). PG-100 antiserum was a polyclonal antiserum that was raised in a rabbit against PG-100 core protein as described (7Schwarz K. Breuer B. Kresse H. J. Biol. Chem. 1990; 265: 22023-22028Abstract Full Text PDF PubMed Google Scholar). This antiserum was kindly provided by Dr. H. Kresse (University of Münster, Münster, Germany). [35S]Sulfate-labeled medium samples from control and TNF-α stimulated EA.hy 926 cell cultures were digested with 0.1 IU/ml of chondroitin ABC lyase in a buffer containing 200 mm Tris, 60 mm sodium acetate, pH 8.0, and 5 μg of chondroitin sulfate A (Seikagaku Co.) as a carrier. The digestions were performed for 6 h at 37 °C. After ethanol precipitation overnight at 4 °C the supernatants and the wash solutions (70% ethanol) were combined, ethanol was removed by evaporation, and the sulfated disaccharides released by chondroitin ABC lyase treatment were separated by ion exchange chromatography (55Midura R.J. Salustri A. Calabro A. Yanagishita M. Hascall V.C. Glycobiology. 1994; 4: 333-342Crossref PubMed Scopus (39) Google Scholar). The average length of the GAG chains was estimated using a Sephacryl S-300 (Pharmacia) gel filtration (56Lammi M. Tammi M. Anal. Biochem. 1988; 168: 352-357Crossref PubMed Scopus (76) Google Scholar, 57Lammi M.J. Tammi M. J. Biochem. Biophys. Methods. 1991; 22: 301-310Crossref PubMed Scopus (9) Google Scholar). Total cellular RNA was isolated from the cultures using the single-step method (58Chomczynski P. Sacchi N. Anal. Biochem. 1987; 162: 156-159Crossref PubMed Scopus (62893) Google Scholar). Northern blot analyses using GeneScreen Plus membranes (DuPont NEN) were performed essentially as described previously (16Järveläinen H.T. Kinsella M.G. Wight T.N. Sandell L.J. J. Biol. Chem. 1991; 266: 23274-23281Abstract Full Text PDF PubMed Google Scholar). Random priming (Multiprime DNA Labeling System, Amersham, UK) was used for labeling of the cDNAs with 5′-[α-32P]dCTP (Amersham) to a high specific activity. For the quantitation of the hybridization signals, the membranes were probed with a rat glyceraldehyde-3-phosphate dehydrogenase cDNA (59Fort P. Marty L. Piechaczyk M. El Sabrouty S. Dani C. Jeanteur P. Blanchard J.M. Nucleic Acids Res. 1985; 13: 1431-1442Crossref PubMed Scopus (1970) Google Scholar). The following cDNA probes were used: a biglycan cDNA (5Fisher L.W. Termine J.D. Young M.F. J. Biol. Chem. 1989; 264: 4571-4576Abstract Full Text PDF PubMed Google Scholar) for the full-length human biglycan core protein; a 1338-base pair CSF-1 cDNA (18Partenheimer A. Schwarz K. Wrocklage C. Kölsch E. Kresse H. J. Immunol. 1995; 155: 5557-5565PubMed Google Scholar) for the PG-100 core protein; a decorin cDNA (6Krusius T. Ruoslahti E. Proc. Natl. Acad. Sci U. S. A. 1986; 83: 7683-7687Crossref PubMed Scopus (408) Google Scholar) for the full-length human decorin core protein; a 1.1-kilobase perlecan cDNA, called HS-1 (60Murdoch A.D. Dodge G.R. Cohen I. Tuan R.S. Iozzo R.V. J. Biol. Chem. 1992; 267: 8544-8557Abstract Full Text PDF PubMed Google Scholar); and pHCAL1U (61Mäkelä J.K. Raassina M. Virta A. Vuorio E. Nucleic Acids Res. 1988; 16: 349Crossref PubMed Scopus (78) Google Scholar) for pro-α1(I) collagen. These cDNA probes were kindly provided by Drs. Fisher (National Institute of Dental Research), Kresse (University of Münster, Münster, Germany), Krusius (The Finnish Red Cross, Helsinki, Finland), Iozzo (Thomas Jefferson University, Philadelphia, PA), and Vuorio (University of Turku, Turku, Finland), respectively. The following cytokines were used: TNF-α and fibroblast growth factor-7 (FGF-7), purchased from PeproTech; fibroblast growth factor-2 (FGF-2), purchased from PeproTech or Boehringer Mannheim; interleukin-1β (IL-1β), interferon-γ (IFN-γ), and TGF-β, all purchased from Boehringer Mannheim. TNF-α was used at concentrations of 5, 25, and 50 ng/ml. FGF-7, FGF-2, IL-1β, IFN-γ, and TGF-β were used at concentrations of 50 ng/ml, 10 ng/ml, 5 IU/ml, 1000 IU/ml, and 20 ng/ml, respectively. Student's unpaired t test was used to describe the significance of the differences between the results of the control and cytokine-treated EA.hy 926 cell cultures. SDS-PAGE of [35S]sulfate-labeled medium sample from a confluent monolayer culture of EA.hy 926 cells demonstrated that these cells synthesize and secrete into the medium two size classes of [35S]sulfate-labeled macromolecules, one that remains on the top of the separating gel (Fig.1, lane 1, band III), and another one with the M r of ∼250,000 (Fig. 1, lane 1, band II). Chondroitin ABC lyase and heparitinase digestions of [35S]sulfate-labeled medium samples prior to SDS-PAGE indicated that the macromolecules of band II are CS/DS PGs, while those of band III represent HS containing PGs (Fig. 1, lanes 2–4), e.g. perlecan (60Murdoch A.D. Dodge G.R. Cohen I. Tuan R.S. Iozzo R.V. J. Biol. Chem. 1992; 267: 8544-8557Abstract Full Text PDF PubMed Google Scholar), since these cells express abundant mRNA for this PG (data not shown). The composition of sulfated PGs in the culture medium of EA.hy 926 endothelial cell monolayers differed from that of sulfated PGs present in the culture medium of HSFs used as control cells. Besides containing PGs migrating in band II and band III (Fig. 1, lane 5), the culture medium of HSFs contained two additional size classes of sulfated PGs, one with the M r of ∼120,000 (Fig. 1, lane 5, band I) and another one that does not enter the separating gel (Fig. 1,lane 5, band IV). Both of these additional PG size classes were sensitive to chondroitin ABC lyase digestion (Fig. 1, lane 6) indicating that they represent CS/DS PGs. Based on the results of previous studies, the band I PGs (M r of ∼120,000) most probably represent decorin, while the PGs of band IV mainly represent versican (6Krusius T. Ruoslahti E. Proc. Natl. Acad. Sci U. S. A. 1986; 83: 7683-7687Crossref PubMed Scopus (408) Google Scholar, 13Rosenberg L.C. Choi H.U. Tang L.-H. Johnson T.L. Pal S. Webber C. Reiner A. Poole A.R. J. Biol. Chem. 1985; 260: 6304-6313Abstract Full Text PDF PubMed Google Scholar, 62Krusius T. Gehlsen K.R. Ruoslahti E. J. Biol. Chem. 1987; 262: 13120-13125Abstract Full Text PDF PubMed Google Scholar, 63Zimmermann D.M. Ruoslahti E. EMBO J. 1989; 8: 2975-2981Crossref PubMed Scopus (499) Google Scholar, 64V

Referência(s)