Portal de Periódicos da CAPES

Reduced Hyaluronan in Keloid Tissue and Cultured Keloid Fibroblasts

2000; Elsevier BV; Volume: 114; Issue: 5 Linguagem: Inglês

10.1046/j.1523-1747.2000.00950.x

1523-1747

Ludger J.M. Meyer, Barbara M. Egbert, Svetlana Shuster, Robert Stern, Shirley B. Russell, James Russell, Joel S. Trupin,

Dermatologic Treatments and Research

Extracellular matrix hyaluronan is prominent during wound healing, appearing at elevated levels early in the repair process. It is prevalent throughout the course of fetal wound healing, which is scar-free, but decreases late in adult wound repair, that is often marked by scarring. To determine whether aberrant hyaluronan metabolism is associated with the excessive scarring that characterizes keloids, cultured fibroblasts derived from keloids and from the dermis of normal human skin and scar were compared. Levels of hyaluronan in 48 h conditioned media of keloid-derived cultures were significantly lower than in cultures of normal skin and scar fibroblasts. Profiles of hyaluronan polymer size were comparable in these two cell types, suggesting that excessive hyaluronan degradation was not involved. Hydrocortisone decreased hyaluronan levels approximately 70% in the conditioned media of both keloid and normal fibroblasts. Diminished hyaluronan accumulation in keloid-derived cells compared with normal fibroblasts was also observed in an in vitro wound healing model. Histolocalization of hyaluronan in keloids, normal skin, and scar samples confirmed the biochemical observations that the dermis of keloids, which comprises most of the scar tissue, contained markedly diminished levels of hyaluronan. Alterations in hyaluronan in the epidermis overlying keloids, however, were also observed. A modest increase in hyaluronan staining intensity was observed in the epidermis of keloids, as well as changes in the patterns of distribution within the epidermis, compared with that in normal skin and scar. Increased hyaluronan was present in the granular and spinous layers of the keloid epidermis Abnormalities are present apparently in both the overlying epidermis as well as in the dermis of keloids. Aberrations in signaling between keloid stroma and keloid epidermis may underlie abnormalities that contribute to the excessive fibrosis characteristic of these lesions. Extracellular matrix hyaluronan is prominent during wound healing, appearing at elevated levels early in the repair process. It is prevalent throughout the course of fetal wound healing, which is scar-free, but decreases late in adult wound repair, that is often marked by scarring. To determine whether aberrant hyaluronan metabolism is associated with the excessive scarring that characterizes keloids, cultured fibroblasts derived from keloids and from the dermis of normal human skin and scar were compared. Levels of hyaluronan in 48 h conditioned media of keloid-derived cultures were significantly lower than in cultures of normal skin and scar fibroblasts. Profiles of hyaluronan polymer size were comparable in these two cell types, suggesting that excessive hyaluronan degradation was not involved. Hydrocortisone decreased hyaluronan levels approximately 70% in the conditioned media of both keloid and normal fibroblasts. Diminished hyaluronan accumulation in keloid-derived cells compared with normal fibroblasts was also observed in an in vitro wound healing model. Histolocalization of hyaluronan in keloids, normal skin, and scar samples confirmed the biochemical observations that the dermis of keloids, which comprises most of the scar tissue, contained markedly diminished levels of hyaluronan. Alterations in hyaluronan in the epidermis overlying keloids, however, were also observed. A modest increase in hyaluronan staining intensity was observed in the epidermis of keloids, as well as changes in the patterns of distribution within the epidermis, compared with that in normal skin and scar. Increased hyaluronan was present in the granular and spinous layers of the keloid epidermis Abnormalities are present apparently in both the overlying epidermis as well as in the dermis of keloids. Aberrations in signaling between keloid stroma and keloid epidermis may underlie abnormalities that contribute to the excessive fibrosis characteristic of these lesions. hyaluronan, hyaluronic acid HA-binding protein phosphate-buffered saline, Ca2+- and Mg2+-free Keloids are benign dermal tumors that form as a result of an inherited defect in wound healing (Murray et al., 1981Murray J.C. Pollack S.V. Pinnell S.R. Keloids: a review.J Am Acad Dermatol. 1981; 4: 461-470Abstract Full Text PDF PubMed Scopus (172) Google Scholar;Datubo-Brown, 1990Datubo-Brown D.D. Keloids: a review of the literature.Br J Plast Surg. 1990; 43: 70-77Abstract Full Text PDF PubMed Scopus (177) Google Scholar;Ehrlich et al., 1994Ehrlich H.P. Desmoulière A. Diegelmann R.F. et al.Morphologic and immunochemical differences between keloid and hypertrophic scar.Am J Pathol. 1994; 145: 105-113PubMed Google Scholar). Keloids are associated with the deposition of excessive collagenous scar tissue in response to even minor trauma. The distortion, contractures, and cosmetic problems associated with keloids present a formidable clinical challenge. Keloid tissue exhibits an elevated rate of collagen synthesis and contains higher levels of fibronectin (Kischer and Hendrix, 1983Kischer C.W. Hendrix M.J. Fibronectin (FN) in hypertrophic scars and keloids.Cell Tissue Res. 1983; 231: 29-37PubMed Google Scholar) and proteoglycans (Bazin et al., 1973Bazin S. Nicoletis C. Delaunay A. Intercellular matrix of hypertrophic scars and keloids.in: Kulonen E. Pikkarainen J. Biology of Fibroblast. Academic Press, New York1973: 571-578Google Scholar) compared with normal mature scars. Keloid-derived fibroblasts have been reported to synthesize collagen (Diegelmann et al., 1979Diegelmann R.F. Cohen I.K. McCoy B.J. Growth kinetics and collagen synthesis of normal skin, normal scar and keloid fibroblasts in vitro.J Cell Physiol. 1979; 98: 341-346Crossref PubMed Scopus (172) Google Scholar;Abergel et al., 1985Abergel R.P. Pizzurro D. Meeker C.A. et al.Biochemical composition of the connective tissue in keloids and analysis of collagen metabolism in keloid fibroblast cultures.J Invest Dermatol. 1985; 84: 384-390Crossref PubMed Scopus (220) Google Scholar) and fibronectin (Babu et al., 1989Babu M. Diegelmann R. Oliver N. Fibronectin is overproduced by keloid fibroblasts during abnormal wound healing.Mol Cell Biol. 1989; 9: 1642-1650Crossref PubMed Scopus (131) Google Scholar) at higher rates than do fibroblasts from normal skin or scars when cultured in unsupplemented medium. Although we (Russell et al., 1978Russell J.D. Russell S.B. Trupin K.M. Differential effects of hydrocortisone on both growth and collagen metabolism of human fibroblasts from normal and keloid tissue.J Cell Physiol. 1978; 97: 221-229Crossref PubMed Scopus (70) Google Scholar,Russell et al., 1982Russell J.D. Russell S.B. Trupin K.M. Fibroblast heterogeneity in glucocorticoid regulation of collagen metabolism: genetic or epigenetic?.In Vitro. 1982; 18: 557-564Crossref PubMed Scopus (13) Google Scholar;Trupin et al., 1983Trupin J.S. Russell S.B. Russell J.D. Variation in prolyl hydroxylase activity of keloid-derived and normal human fibroblasts in response to hydrocortisone and ascorbic acid.Collagen Relat Res. 1983; 3: 13-23Crossref PubMed Google Scholar) and colleagues (Ala-Kokko et al., 1987Ala-Kokko L. Rintala A. Savolainen E.R. Collagen gene expression in keloids: analysis of collagen metabolism and type I III, IV, and V procollagen mRNAs in keloid tissue and keloid fibroblast cultures.J Invest Dermatol. 1987; 89: 238-244Abstract Full Text PDF PubMed Google Scholar) observed no differences under standard culture conditions, we have reported that hydrocortisone decreases the rate of collagen synthesis (Russell et al., 1978Russell J.D. Russell S.B. Trupin K.M. Differential effects of hydrocortisone on both growth and collagen metabolism of human fibroblasts from normal and keloid tissue.J Cell Physiol. 1978; 97: 221-229Crossref PubMed Scopus (70) Google Scholar,Russell et al., 1982Russell J.D. Russell S.B. Trupin K.M. Fibroblast heterogeneity in glucocorticoid regulation of collagen metabolism: genetic or epigenetic?.In Vitro. 1982; 18: 557-564Crossref PubMed Scopus (13) Google Scholar;Trupin et al., 1983Trupin J.S. Russell S.B. Russell J.D. Variation in prolyl hydroxylase activity of keloid-derived and normal human fibroblasts in response to hydrocortisone and ascorbic acid.Collagen Relat Res. 1983; 3: 13-23Crossref PubMed Google Scholar) and the amounts of types I, III, and V collagen mRNA in normal but not in keloid cells (Russell et al., 1989Russell S.B. Trupin J.S. Myers J.C. Broquist A.H. Smith J.C. Myles M.E. Russell J.D. Differential glucocorticoid regulation of collagen mRNAs in human dermal fibroblasts. Keloid-derived and fetal fibroblasts are refractory to down-regulation.J Biol Chem. 1989; 264: 13730-13735Abstract Full Text PDF PubMed Google Scholar). We have also reported that keloid cells produce more elastin than normal cells in unsupplemented medium, and that hydrocortisone decreases the levels of elastin and elastin mRNA in normal but not keloid cells (Russell et al., 1995Russell S.B. Trupin J.S. Kennedy R.Z. Russell J.D. Davidson J.M. Glucocorticoid regulation of elastin synthesis in human fibroblasts: down-regulation in fibroblasts from normal dermis but not from keloids.J Invest Dermatol. 1995; 104: 241-245Crossref PubMed Scopus (46) Google Scholar). Hydrocortisone stimulates the growth of normal but not keloid fibroblasts (Russell et al., 1978Russell J.D. Russell S.B. Trupin K.M. Differential effects of hydrocortisone on both growth and collagen metabolism of human fibroblasts from normal and keloid tissue.J Cell Physiol. 1978; 97: 221-229Crossref PubMed Scopus (70) Google Scholar), and transforming growth factor-β stimulates the response of keloid but not normal cells to other growth factors (Russell et al., 1988Russell S.R. Trupin K.M. Rodríguez-Eaton, Russell J.D. Trupin J.S. Reduced growth factor requirement of keloid derived fibroblasts may account for tumor growth.Proc Natl Acad Sci USA. 1988; 85: 587-591Crossref PubMed Scopus (113) Google Scholar). Lastly, keloid fibroblasts show a reduced dependency on serum growth factors for proliferation relative to normal adult dermal fibroblasts (Russell et al., 1988Russell S.R. Trupin K.M. Rodríguez-Eaton, Russell J.D. Trupin J.S. Reduced growth factor requirement of keloid derived fibroblasts may account for tumor growth.Proc Natl Acad Sci USA. 1988; 85: 587-591Crossref PubMed Scopus (113) Google Scholar). These patterns of abnormal regulation of growth and matrix synthesis in cultured fibroblasts from keloids appear to reflect the in vivo phenotype and strongly support the use of keloid fibroblasts in vitro as a model system to study wound healing. In fetal wound healing, which is scar-free, hyaluronan (HA) remains elevated for a considerable period of time (DePalma et al., 1989DePalma R.I. Krummel T.M. Durham 3rd., L.A. Michna B.A. Thomas B.L. Nelson J.M. Diegelmann R.F. Characterization and quantitation of wound matrix in the fetal rabbit.Matrix. 1989; 9: 224-231Crossref PubMed Scopus (115) Google Scholar;Mast et al., 1991Mast B.A. Flood L.C. Haynes J.H. DePalma R.L. Cohen I.K. Diegelmann R.F. Krummel T.M. Hyaluronic acid is a major component of the matrix of fetal rabbit skin and wounds; implications for healing by regeneration.Matrix. 1991; 11: 63-68Crossref PubMed Scopus (86) Google Scholar;West et al., 1997West D.C. Shaw D.M. Lorenz P. Adzick N.S. Longaker M.T. Fibrotic healing of adult and late gestational fetal wounds correlates with increased hyaluronidase activity and removal of hyaluronan.Int J Biochem Cell Biol. 1997; 29: 201-210Crossref PubMed Scopus (103) Google Scholar;Longaker et al., 1991Longaker M.T. Chiu E. Adzick N.S. Stern M. Harrison M. Stern R. Studies in fetal wound healing. V. Prolonged presence of hyaluronic acid in fetal wound fluid.Ann Surg. 1991; 213: 292-296Crossref PubMed Scopus (270) Google Scholar;Stern et al., 1991Stern M. Longaker M.T. Stern R. Hyaluronic acid and its modulation in fetal and adult wound healing.in: Adzick N.S. Longaker M.T. Fetal Wound Healing. Elsevier, New York1991: 189-198Google Scholar). This is in marked contrast to wound healing in the adult, often associated with excessive scarring. HA may promote scarless healing by inhibiting fetal platelet function, including aggregation and cytokine release (Olutoye et al., 1996Olutoye O.O. Yager D.R. Cohen I.K. Diegelmann R.F. Lower cytokine by fetal porcine platelets: a possible explanation for reduced inflammation after fetal wounding.J Pediatr Surg. 1996; 31: 91-95Abstract Full Text PDF PubMed Scopus (74) Google Scholar). In adult wound repair, HA is found early on, but then falls rapidly (DePalma et al., 1989DePalma R.I. Krummel T.M. Durham 3rd., L.A. Michna B.A. Thomas B.L. Nelson J.M. Diegelmann R.F. Characterization and quantitation of wound matrix in the fetal rabbit.Matrix. 1989; 9: 224-231Crossref PubMed Scopus (115) Google Scholar;Longaker et al., 1991Longaker M.T. Chiu E. Adzick N.S. Stern M. Harrison M. Stern R. Studies in fetal wound healing. V. Prolonged presence of hyaluronic acid in fetal wound fluid.Ann Surg. 1991; 213: 292-296Crossref PubMed Scopus (270) Google Scholar). Postulating that aberrations in HA may also occur in the process of keloid formation, we examined various parameters of HA metabolism, and compared them with those occurring in scars and in normal skin, utilizing both cultured fibroblasts and surgical pathology specimens. Tissue sources of keloid and normal adult dermal fibroblasts are summarized in Table 1. Six specimens were obtained from keloid tissue, and six from nonkeloid adult skin or scar. The age range of tissue donors was 18–42 y. With one exception, a 27 y old African-American male (strain 103), all subjects were African-American females. All normal cultures were derived from normal scars, except 103, which was from normal skin. The methods of isolation, propagation, and freezing of fibroblasts have been reported (Russell et al., 1978Russell J.D. Russell S.B. Trupin K.M. Differential effects of hydrocortisone on both growth and collagen metabolism of human fibroblasts from normal and keloid tissue.J Cell Physiol. 1978; 97: 221-229Crossref PubMed Scopus (70) Google Scholar,Russell et al., 1982Russell J.D. Russell S.B. Trupin K.M. Fibroblast heterogeneity in glucocorticoid regulation of collagen metabolism: genetic or epigenetic?.In Vitro. 1982; 18: 557-564Crossref PubMed Scopus (13) Google Scholar). Cells were propagated at 37°C in an atmosphere of air and CO2 adjusted to maintain pH at 7.4 under 100% relative humidity.Table 1Clinical origin of cell culturesCulture no.OriginSexAge of patientLocation21ScarF29Abdomen58Old scarF42Abdomen103SkinM27Unknown131ScarF22Abdomen132ScarFUnknownUnknown170ScarF42Unknown33KeloidF18Earlobe50KeloidF15Earlobe124KeloidFAdolescentEarlobe125KeloidFAdolescentEarlobe134KeloidFUnknownUnknown145KeloidF31Back Open table in a new tab Clinical specimens of keloids and excisions from scar tissue were archival cases obtained from Surgical Pathology, Department of Pathology, School of Medicine, University of California, San Francisco. Slides were cut from formalin-fixed and glutaraldehyde-fixed, paraffin-embedded tissue blocks, as processed routinely in dermatopathology. Scratch experiments, an in vitro model for wound repair (Todaro et al., 1965Todaro G.J. Lazar G.K. Green H. The initiation of cell division in a contact-inhibited mammalian cell line.J Cell Comp Physiol. 1965; 66: 325-333Crossref Scopus (522) Google Scholar;Longaker et al., 1989Longaker M.T. Harrison M.R. Langer J.C. Crombleholme T.C. Verrier E.D. Spendlove R. Stern R. Studies in fetal wound healing: II. A fetal environment accelerates fibroblast migration in vitro.J Pediatr Surg. 1989; 24: 793-798Abstract Full Text PDF PubMed Scopus (20) Google Scholar), were performed as follows. Cultures were grown to 80% confluency in 75 cm2 flasks. Six uniform vertical and horizontal scratches at intervals of approximately 2 mm, were made on the cell monolayers with the sterilized tip of a glass pipette. After washing twice with the appropriate media, the monolayers were incubated for various lengths of time with either serum-containing or serum-free medium and aliquots removed for HA assay. Photographic documentation of cellular repair was performed at each experimental time point (Nikon 46008 microscope, Nikon M35S 55396 camera, Nikon AFM objective, Nikon 46152 condenser; Technical Instruments, San Francisco, CA). HA was assayed by an enzyme-linked immunosorbent assay-like microtiter technique based on a biotinylated HA-binding protein (HABP), which enables the determination of HA at nanogram levels (Fosang et al., 1990Fosang A.J. Hey N.J. Carney S.L. Hardingham T.E. Matrix. 1990; 10: 306-313Crossref PubMed Scopus (105) Google Scholar). HABP was isolated from a tryptic digest of bovine nasal cartilage, as described (Tengblad, 1979Tengblad A. Affinity chromatography on immobilized hyaluronate and its application to the isolation of hyaluronate binding proteins from cartilage.Biochem Biophys Acta. 1979; 578: 281-289Crossref PubMed Scopus (218) Google Scholar). In short, homogenized cartilage was placed in 4 M guanidine-HCl and 0.5 M NaAc at pH 5.8 overnight. The supernatant was separated, dialyzed against dH2O and 1 mM Tris-base, and lyophilized. After trypsin digestion and dialysis against 4 M guanidine-HCl, the extract was placed on an HA-Sepharose column. After washing with 1 M and 3 M NaCl, the protein was eluted and dialyzed against 0.1 M Na carbonate buffer, pH 8.5. After a final incubation with biotin, the reaction was terminated with 1 M glycine. Costar multiwell plates were precoated with 1.84 mg sulfo-N-hydroxysulfosuccinimide, 2.0 mg HA and 12.3 mg ethyl-dimethylaminopropyl-carbodiimide in 10 ml dH2O. After washing and incubating the plates overnight in a buffer solution containing 116.9 mg NaCl and 10 mg MgSO4 in 1 ml phosphate-buffered saline that was free of Ca2+ and Mg2+ (PBS-CMF) (buffer A), 300 ml blocking reagent was added to each well, and incubated in a humidified environment at 37°C for 30 min. The blocking reagent was constituted from 155 mg of dry milk powder in 1 ml PBS-CMF diluted to 30 ml with PBS just prior to use. Following this incubation, the plates were washed with buffer A with the addition of 0.05% Tween 20 (buffer B). One hundred microliters of this solution was added to each well, consisting of equal parts of sample and 1:100 diluted HABP solution, incubated at 37°C for 1 h, washed again with buffer B, and then ortho-phenylenediamine (2 mg in 4 ml citrate-PO4 buffer, pH 5.3) and 30% H2O2 (2 ml per 4 ml) were added to the wells (100 μl each) After developing for 20–30 min in the dark, the color was read at 492 nm (Titertek Multiskan Plus MK II, Labsystems, Helsinki, Finland). Positive and negative controls and a standard curve from 40 ng to 3 μg per ml HA (Sigma, St Louis, MO) were run with each plate. Each experimental point was determined in triplicate. In all cases where HA accumulation was measured in serum-containing medium, the values were corrected for the concentration of HA in fetal bovine serum. To determine the size distribution of HA polymers, Sepharose CL-4B column chromatography (Pharmacia AB, Uppsala, Sweden) was used (Henrich and Hawkes, 1989Henrich C.J. Hawkes S.P. Molecular weight dependence of hyaluronic acid produced during oncogenic transformation.Cancer Biochem Biophys. 1989; 10: 257-267PubMed Google Scholar), with a column (42 × 1 cm i.d.) of 35 ml bed volume, in buffer containing 50 mM Tris and 0.15 M NaCl, pH 8.0. Under continuous pressure (Gilson Minipuls 2 pump, Gilson Medical Electronics, Middleton, WI), 80 fractions (1 ml each) were collected on a Gilson Microfractionator. HA levels in each fraction were determined using the microtiter assay. Tissue pieces were placed in saline and immediately transferred to a solution of 2% formalin and 0.5% glutaraldehyde in 0.1 M PBS at pH 7.35, as described byHellström et al., 1990Hellström S. Tengblad A. Johansson C. Hedlund U. Axelsson E. An improved technique for hyaluronan histochemistry using microwave irradiation.Histochem J. 1990; 22: 677-682Crossref PubMed Scopus (59) Google Scholar. Specimens were processed in the usual manner for paraffin embedding. Serial sections of 4 μm thickness were mounted on polylysine-coated glass slides for hematoxylin and eosin, as well as for HA staining. To document specificity of HA staining, 1 ml of hyaluronidase buffer (0.1 M Na formate, 0.15 M NaCl, 0.1 mg bovine serum albumin per ml, 0.1% Triton X-100, pH 3.7) was combined with 100 turbidity reducing units of Streptomyces hyaluronidase (CalBiochem, Behring Diagnostics, La Jolla, CA). Approximately 0.3 ml of this solution was placed on deparaffinized control slides and incubated overnight in a humidified environment at 37°C. Biotinylated HABP was diluted 1:100 with HABP buffer solution (0.25 M Na phosphate, 0.15 M NaCl, 0.3 M guanidine-HCl, 0.08% bovine serum albumin, 0.02% NaN3, pH 7.0), and combined 1:1 with a solution of 0.1 mg per ml HA (Sigma). Hyaluronidase control slides and experimental slides were incubated in 3% bovine serum albumin at 37°C for 30 min. After rinsing with PBS-CMF, 0.3 ml of diluted HABP solution was added to each slide. At the same time, 0.3 ml of the HA and HABP control solution were added to each HA and HABP control slide. Slides were incubated overnight at 4°C. A solution of 98 ml methanol with 2 ml of 30% H2O2 was prepared for each well, containing 10 slides. Slides were rinsed with PBS-CMF and transferred to staining wells containing this solution for a 3 min incubation at room temperature. Immediately following, slides were rinsed with PBS-CMF, and incubated with an avidin–biotin complex (Vectastain ABC Kit, Peroxidase PK 4000, Vector Laboratories, Burlingame, CA) for 30 min at room temperature. After rinsing again with PBS-CMF, slides were incubated for 5 min at room temperature with 3,3′-diaminobenzidine (Peroxidase Substrate Kit, DAB SK-4100, Vector Laboratories), and rinsed with dH2O. Slides were counterstained for 15 min in 0.25% methylgreen, rinsed in dH2O, dipped in 70%, 95%, and 100% methanol, and finally in xylene. Photographs were taken on an Olympus Vanox AHBT3 microscope (Olympus, Woodbury, NY) with an integrated Olympus C-35AD-4 camera on Kodak Gold Plus 100 film (Eastman Kodak, Rochester, NY). The amount of HA accumulated by confluent cultures during a 48 h period was measured in cells derived from keloids and from normal human dermis and scars. Initially, levels of HA in cell layers and in conditioned media were determined separately (Figure 1). In a comparison of two keloid and two normal cultures, small differences were seen in the cell layers, which contained only about 25% of the total HA accumulated. In the culture media, however, the amount of HA produced by normal cells was approximately 4-fold higher than that accumulated by keloid fibroblasts. All subsequent determinations of HA were performed on conditioned media only. The specimens shown in Figure 1 were from keloid cultures numbers 33 and 125, and from normal scar specimens, from cultures numbers 21 and 131. HA accumulation was then assayed in the culture medium of all six keloid and the six normal strains grown shown in Table 1. As can be observed in Table 2, the mean value of accumulated HA was 4-fold higher in the culture medium of fibroblasts derived from a normal scar compared with those derived from keloids. The mean values were 2.60 μg HA per 105 cells for the normal fibroblast strains, and only 0.67 μg HA per 105 cells for the keloid fibroblast strains. Only the culture from keloid specimen number 145 showed overlap with the range from normal scar cultures. This was a keloid derived from the back and described by the pathologists as an ''old and unusual'' lesion.Table 2Keloid fibroblasts accumulate less HA than normal fibroblasts and hydrocortisone reduces HA accumulation in both cell typesaFibroblasts from six normal and six keloid strains were plated at 2.5 × 104 cells per 35 mm dish and fed daily with F10 + 10% fetal bovine serum alone or supplemented with 1.5 μm hydrocortisone. On day 9 the confluent cultures were fed fresh medium and incubated for 48 h. HA accumulated in the media during this period was assayed using the enzyme-linked immunosorbent assay-like assay described in Materials and Methods. Average HA accumulation was significantly higher in normal than in keloid strains in the absence (p < 0.01) and presence (p < 0.025) of hydrocortisone. Hydrocortisone caused a significant decrease in HA accumulation in both normal (p < 0.01) and keloid (p < 0.05) cellsHyaluronan (μg per 105 cells)StrainControlHydrocortisoneHC/ControlNormal 214.940.450.09 581.220.360.30 1032.470.800.32 1312.741.080.39 1323.100.360.12 1701.140.510.45 Mean ± SEM2.60 ± 0.570.59 ± 0.120.28 ± 0.06Keloid 330.140.100.71 500.110.010.09 1240.740.200.27 1250.640.300.47 1340.200.020.10 1452.190.710.32 Mean ± SEM0.67 ± 0.320.22 ± 0.110.33 ± 0.10a Fibroblasts from six normal and six keloid strains were plated at 2.5 × 104 cells per 35 mm dish and fed daily with F10 + 10% fetal bovine serum alone or supplemented with 1.5 μm hydrocortisone. On day 9 the confluent cultures were fed fresh medium and incubated for 48 h. HA accumulated in the media during this period was assayed using the enzyme-linked immunosorbent assay-like assay described in Materials and Methods. Average HA accumulation was significantly higher in normal than in keloid strains in the absence (p < 0.01) and presence (p < 0.025) of hydrocortisone. Hydrocortisone caused a significant decrease in HA accumulation in both normal (p < 0.01) and keloid (p < 0.05) cells Open table in a new tab HA accumulation was then assayed in all six keloid and six normal strains grown with 1.5 μM hydrocortisone (Table 2). Growth of fibroblasts in hydrocortisone reduced the amount of HA in the conditioned media to the same extent, approximately 70% in both keloid and normal cultures. But the mean levels of HA accumulated in normal scar fibroblasts continued to be greater than that in keloid fibroblast cultures, 0.59 and 0.22 μg per 105 cells, respectively. Hydro- cortisone suppressed levels of HA accumulation, but did not have a differential effect on the two classes of fibroblasts. The time course of HA accumulation was examined in intact and scratched monolayers of one normal strain, number 131 (Figure 2a) and one keloid strain, number 125 (Figure 2b) incubated in the presence and absence of 10% fetal bovine serum. Four replicate T75 flasks of each cell strain were grown to 80% confluence and two of the flasks from each strain were scratched as described in Materials and Methods. At 6, 12, 18, 24, and 48 h after rinsing and refeeding with fresh media, aliquots of culture media were removed for assay. In the intact cultures, very little HA secretion was detected in the absence of serum. The normal scar-derived and keloid-derived fibroblasts secreted approximately 0.2 and 0.1 pg of HA per cell, respectively, over the 48 h period. In the presence of 10% fetal bovine serum, HA secretion increased rapidly over the first 12 h and more gradually thereafter. At all time points, the normal culture secreted more HA than the keloid strain. At 48 h the average level of HA secretion was 7.4 pg per normal cell and 3.8 pg per keloid cell. HA secreted in scratched serum-fed cultures is shown in the top lines of Figure 2(a,b). A stimulation was observed in each type of culture, with the level of HA secretion reaching 12.5 pg per cell in the normal culture, and 5.2 pg per cell, in the keloid culture. Compared with the intact monolayers, scratching increased HA secretion approximately 66% in the normal strain and 34% in the keloid strain. In the absence of serum, the ''scratched'' monolayers of normal and keloid cells continued to secrete very low levels of HA into the culture medium. The size distribution of HA polymers in the conditioned medium of cell cultures was examined using CL-4B column chromatography (Figure 3). No major differences were observed. A clearly defined single peak in the excluded volume, the high molecular weight range (fractions 12–14), was present in all cultures. Representative HA chromatography profiles of material from the two cell types are shown in Figure 3. Finally, we stained for tissue distribution of HA in paraffin-embedded histologic sections of normal skin, normal scars, and keloids. Steadily decreasing intensity of staining was observed in the dermis of the specimens in the aforementioned order. The pattern of distribution in dermis of these different tissues did not vary significantly. The diminished intensity of HA staining in the keloid specimens supported our biochemical observations, that HA accumulation in the medium of keloid-derived fibroblast cultures was much less than that in normal scar-derived fibroblasts. Representative staining patterns of normal scar and keloid tissue are demonstrated in Figure 4(a,b), respectively. A control staining pattern in which the HA-binding peptide had been preincubated with HA is shown in Figure 4(c). No staining is observed. All of the surgical pathology specimens described in Table 1 were examined, and the results were consistent throughout, with much less staining intensity for HA observed in the dermis of all keloid specimens. Examination of the epidermis overlying the keloid and normal scar indicated additional differences existed between these tissues. At higher power, it was observed that intense staining for HA was present in the granular and spinous layers of the epidermis overlying the keloid specimen predominantly between keratinocytes. Staining of the basal layer was also observed, although less intense than the granular and spinous layer (Figure 4e). This is in marked contrast with th