Hyaluronic acid metabolism in keloid fibroblasts

Hyaluronic Samuel

Acid Metabolism

M. Alaish,

Dorne R. Yager, Robert Richmond,

0 Hyaluronic acid (HA), an important component of the tissue extracellular matrix, is a ubiquitous glycosaminoglycan (GAG) that forms a pericellular coat on the surface of cells. It has been speculated that this pericellular HA boundary may localize cytokines, such as transforming growth factor+l, which is known to stimulate collagen production. The purpose of this study was to examine the role of HA and its cell surface receptor (CD44], an active participant in HA degradation, as they relate to keloid formation. Dermal excisions from both normal patients (n = 13) and keloid patients (n = 13) were analyzed for HA content using an alcian blue staining technique. Fibroblast cell cultures were used to quantitate HA synthesis and CD44 receptor density. Histological analyses showed a greater HA content in keloid tissue compared with normal dermal tissue. In agreement with this observation, keloid fibroblasts were found to synthesize significantly more HA than normal dermal fibroblasts (2469 * 483 cpm versus 1122 -C 256 cpm, P = .OZ). Treatment of keloid fibroblasts with triamcinolone acetonide reduced the level of HA synthesis to that of normal fibroblasts (1560 + 477 cpm versus 1293 -1- 264 cpm, P = .6). However, there was no significant difference in HA receptor density on keloid cells compared with normal skin fibroblasts. Therefore, the increased HA deposits found in keloids are attributable to increased synthesis rather than to decreased degradation mediated by the CD44 receptor. Copyright o 1995 by W. B. Saunders Company INDEX WORDS: Hyaluronic acid, keloid, triamcinolone, fibroblast, wound healing.

CD44

receptor,

K

ELOIDS are characterized by excessive connective tissue matrix, predominantly caused by an increased accumulation of col1agen.l Another extracellular matrix component, hyaluronic acid (HA), is a ubiquitous glycosaminoglycan (GAG) and is thought to be an important mediator in normal tissue repair. In keloids, the role of HA and its cell surface receptor, an active participant in HA degradation, has not been defined. Therefore, this study was designed to test the hypothesis that an alteration in HA metabolism is responsible for the excessive collagen accumulation found in keloids. MATERIALS

AND METHODS

Normal patients were defined as those patients with no wound healing pathology noted on history, physical examination, or dermal biopsy, whereas keloid patients were defined as those patients with a history, physical examination, and dermal biopsy consistent with a clinical diagnosis of keloid formation. All patients were black and ranged from 18 to 24 years of age. Informed consent was obtained from all patients, and this study was approved by the institutional review board (IRB #9408-9K). Dermal excisions from both normal patients (n = 13 patients) and keloid patients (n = 13) were fixed in 4% formalin and embedded in paraffin for light microscopy. Alcian blue staining was performed Journalof

fedtatnc

Surgery,

Vol30,

No 7 (July),

1995:

pp 949-952

in Keloid Fibroblasts F. Diegelmann,

and I. Kelman

Cohen

Virginia

to show the GAG content in the tissue sections. Pretreatment with bovine testicular hyalurohidase, an enzyme specific for dermal HA, followed by alcian blue staining was used to determine the relative amount of HA in the tissue sections. The sections were examined and scored by a pathologist on a scale of 0 to 4+ (absent to abundant) using a blinded technique. Dermal excisions from both normal patients (n = 5) and keloid patients (n = 5) were explanted to establish fibroblast cell cultures to be used for (1) quantitation of HA synthesis, (2) CD44 Western blot analysis, and (3) CD44.flow cytometry analysis. All cell lines were maintained in Dulbecco’s modification of Eagle’s medium (DMEM), which was supplemented with 10% fetal calf serum, penicillin, and streptomycin (100 pg/mL each) and amphotericin B (Fungizone) (0.25 FgimL). All HA synthesis experiments were done using fourth passage cells plated at a standard density of 3 x lo4 cells per 30 mm culture well and treated with or without triamcinolone acetonide (0.5 PM) for 48 hours. Preconfluent keloid and normal cells were pulse-labelled with 6-3H-glucosamine (40 Ci/mmol) for 24 hours. The medium was collected and pooled with two washes of the cell layer. After protease digestion to isolate GAGS, triplicate samples were treated with or without Streptomyces hyaluronlyticus hyaluronidase (100 units/ml) and precipitated by addition of cetylperidinium chloride. Precipitates were collected on filters, and HA synthesis was determined by scintillation counting. Counts were normalized with respect to total DNA content in each sample using a fluorometric DNA assay.*Using the Bradford protein assay,the counts were also normalized for total protein concentration in each sample. Statistical analysis was performed using an unpaired Student’s t test, with a P value of less than .05 considered significant. For the Western blot analysis, confluent normal and keloid fibroblasts (fourth passage cells plated at a standard density of 5 x lo5 cells per 100 mm culture well) were rinsed with DMEM, scraped off the plate, placed in gel electrophoresis buffer, and sonicated. Debris was removed by centrifugation. The proteins present in the fibroblast lysate were separated on a 7.5% acrylamide gel and then transferred to nitrocellulose paper for Western blot analysis. The primary antibody used was a monoclonal antibody to a human CD44 receptor (clone BU52; The Binding Site, Ltd, San Diego, CA), and the secondary antibody used was a horseradish-peroxidase-labelled antibody to murine immunoglobulin G (IgG) (Amersham, Inc, Amersham, UK). Five different normal and keloid fibroblast cell lines were analyzed. The blots were developed by chemiluminescence, and autoradiographs of the blots were assessed by densitometry to quantitate the relative

From the Wound Healing Center, Division of Plastic and Reconstmctive Sueev, Medical College of VirginialI&ginia Commonwealth University, Richmond, VA. Presented at the 1994 Annual Meeting of the Section on Surgey of the American Academy of Pediatrics, Dallas, Texas, October 21-23, 1994. Supported by National Institutes of Health Grant GM-20298 and the Plastic Surgery Educational Foundation. Address reprint requests to Samuel M. Alaish, MD, Depaltment of Surgery Medical College of Virginia, Richmond, VA 23298-0117. Copyright o 1995 by W.B. Saunders Company 0022-346819513007-0009$03.00/O 949

950

ALAISH

amounts of the HA receptor. Using the aforementioned protein and DNA assays, the amount of the HA receptor was then normalized for total protein and total DNA in each sample. Statistical analysis was performed with an unpaired Student’s t test. For the flow cytometry analysis, fourth passage normal and keloid fibroblasts were plated at a standard density of 5 x lo5 cells per 100 mm culture plate and were analyzed by flow cytometry. At confluence, the cells were rinsed with DMEM, lifted off the plates with 5 mM EDTA and placed in phosphate-buffered saline with 0.1% bovine serum albumin (PBS-BSA). The cells were incubated with the BUZZ antibody for 1 hour, washed with PBS-BSA incubated with a fluorescein-labelled antibody to murine IgG (FITC conjugate; The Binding Site, Ltd), washed with PBS-BSA, and then placed in a single-cell suspension with PBS-BSA. Normal and keloid cells incubated without the BU52 or FITC antibodies, with BU.52 but without FITC antibody, and without BU52 but with FITC antibody served as control groups. Approximately 20,000 cells of each group were run through the Fluorescence Activated Cell Sorter (FACS). Relative fluorescence was representative of HA receptor density. A Fisher’s exact test was performed on the data to determine significance. RESULTS

Histological analysis showed there to be significantly more GAGS in the keloid tissue sections than in the normal skin tissue sections. In addition, this difference seemed to be largely caused by an increased HA content in keloid tissue compared with normal dermal tissue. Keloid fibrobiasts synthesized significantly more HA than normal fibroblasts (2469 + 483 cpm versus 1122 f 256 cpm, P = .02). Treatment of keloid fibroblasts with triamcinolone acetonide reduced the level of HA synthesis to that of normal fibroblasts (1560 + 477 cpm versus 1293 ? 264 cpm, P = .6). We conclude that HA synthesis is elevated in keloid

ET AL

fibroblasts as compared with normal fibroblasts and that this increase can be normalized with the administration of triamcinolone acetonide. Results of the Western blot analysis showed that the HA receptor was expressed on both the adult and keloid fibroblasts and was the expected size of 85 kd. Using the BU52 antibody, five different normal and keloid cell lines were tested. After normalizing for total DNA and also total protein, the amount of CD44 on kelaid fibroblasts was expressed as a ratio relative to the amount of CD44 on normal skin fibroblasts. The mean density of CD44 on keloid fibroblasts was not significantly different from that of the normal fibroblasts. Flow cytometry analysis was used to corroborate the Western blot analysis. Typical light scatter and fluorescence histograms are shown in Fig 1. Analysis of another set of cell lines again showed no significant difference (P = 1.0) between the fluorescence of the keloid fibroblasts and the normal skin fibroblasts. DISCUSSION

In conditions of excessive tissue fibrosis, the abundant extracellular matrices are comprised of supranorma1 levels of both collagen and GAGS including HA.3 HA is known to form a pericellular coat on the surface of cells.4 This pericellular boundary of HA is thought to be mediated by cell surface receptors.5Jj Transforming growth factor-l31 (TGF-l31) is known to stimulate collagen production,7,8 and it is speculated that HA could play a role in maintaining TGF-l31 around the cell’s microenvironment9 This is anala-

NORMAL ELI52 (CD44) % POSITIVE CELLS 98%

3

FORWARD

SCATTER

KELOID I

I

94%

--,

FORWARD

SCATTER

Fig 1. Typical light scatter (left) and fluorescence histograms (right) of normal skin and keloid fibroblasts. The upper graphs represent FACS analysis of normal fibroblasts. The lower graphs represent FACS analysis of keloid fibroblasts. Note that the percentage of positive cells (relative fluorescence) is not significantly different between the two groups (P = .28).

HA METABOLISM

IN KELOID

FIBROBLASTS

951

gous to the relationship between another GAG, heparin, and fibroblast growth factors (FGFs), This GAG potentiates the bioactivities of the FGFs by stabilizing the tertiary structure as well as preventing the proteolytic degradation of the FGFs.lO Heparin has been found to restore biological activity to human recombinant acidic FGF and to increase the affinity of this FGF for its receptor.lOJ1 Similarly, the interaction of HA and TGF+l may result in a stable complex that resists degradation by enzymes and thereby maintains the bioactivities of TGF+l. Intralesional steroid injections have been shown to cause keloid regression in vivo mainly by decreasing their collagen content 12;the exact mechanisms responsible for this effect remain unclear. Steroids may reduce plasma protease inhibitors, thus allowing collagenase to degrade keloid collagen.13 However, if the HA pericellular boundary not only localizes cytokines such as TGF+l but also stabilizes the plasma protease inhibitor-collagenase complex, then collagen production would be increased and degradation would be inhibited. In this study, keloid fibroblasts treated with triamcinolone acetonide showed decreased synthesis of HA compared with cells treated with vehicle alone. As the extracellular HA content diminishes, so may the role of the HA pericellular boundary. This theory might then explain why the normalization in HA synthesis by keloid fibroblasts exposed to triamcinolone acetonide in vitro correlates with the decrease in collagen content found in vivo. Understanding the implications of these results on HA metabolism requires analysis of the degradation of HA. The HA receptor is an active participant in the process of HA degradation.14 After binding, the HA receptor-ligand complex is internalized and trans-

ported to a lysosome where the HA is digested by acid hydrolases. The receptor is then recycled through a coated-pit pathway.l5 In adult human cells, the HA receptor has been well characterized. A member of the CD44 family of cell surface glycoproteins, the human HA receptor is a transmembrane glycoprotein weighing approximately 85 kd that binds to HA in the extracellular matrix.16J7 One of the hypotheses set forth in this study was that keloid cells had a decreased cell surface density of the HA receptor compared with normal dermal cells, thereby attributing the increased HA content in keloid tissue to both heightened HA synthesis and defective HA degradation. However, the results of both the Western blot and flow cytometry analyses showed that there was no significant difference in HA receptor density between normal and keloid cells. Consequently, increased HA synthesis appears to account for the increased HA content noted histologically in keloid tissue compared with normal dermal tissue. Children have a greater tendency to form and reform keloids than do adults, and sometimes the keloid tendency diminishes after puberty.l* The explanation for this and many other aspects of keloid formation remain elusive. Continued efforts to unravel the biochemical mechanisms involved in keloid formation may then lead to improved management and new treatments for the many other conditions dominated by abnormal healing in both children and adults. ACKNOWLEDGMENT The authors thank Hui Xiu Liang, MD, for her assistance in the histological analyses.

REFERENCES 1. Diegelmann RF, Cohen IK, McCoy BJ: Growth kinetics and collagen synthesis of normal skin, normal scar and keloid fibroblasts in vitro. J Cell Physio198:341-346,1979 2. Labarca C, Paigen K: A simple, rapid, and sensitive DNA assayprocedure. Anal Biochem 102:344-352,198O 3. Bertolami CN, Berg S, Freymiller EG: Glycosaminoglycan processing during tissue repair: Degradation of hyaluronic acid, in Adzick NS, Longaker MT (eds): Fetal Wound Healing. New York, Elsevier Science Publishing Co, 1992, pp 215-226 4. Underhill CB, Toole BP: Binding of hyaluronate to the surfaces of cultured cells. Cell Biol82:475-484, 1979 5. Yoneda M, Suzuki S, Kimata K: Hyaluronic acid associated with the surfaces of cultured fibroblasts is linked to a serumderived SS-kDa protein. J Biol Chem 265:5247-5257,199O 6. McGary CT, Raja RH, Weigel PH: Endocytosis of hyaluronic acid by rat liver endothelial cells. Evidence for receptor recycling. Biochem J 257875884,1989 7. Roberts AB, Sporn MB, Assoian RK, et al: Transforming growth factor type-beta: Rapid induction of fibrosis and angiogenesis in vivo and stimulation of collagen formation in vitro. Proc Nat1 Acad Sci U S A 83:4167-4171,1986

8. Peltonen J, Hsiao LL, Jaakkola S, et al: Activation of collagen gene expression in keloids: Co-localization of types I and VI collagen and transforming growth factor 81 mRNA. J Invest Dermatol97:240-248,1991 9. Mast BA, Diegelmann RF, Krummel TM, et al: Hyaluronic acid modulates proliferation, collagen and protein synthesis of cultured fetal fibroblasts. Matrix 13:441-446,1993 10. Burgess WH, Maciag T: The heparin-binding (fibroblast) growth factor family of proteins. Annu Rev Biochem 58:575-606.1989 11. Schreiber AB, Kenney J, Kowalski WJ, et al: Interaction of endothelial cell growth factor with heparin: Characterization by receptor and antibody recognition. Proc Nat1 Acad Sci U S A 82:6138-6142,1985 12. Ketchum LD, Smith J, Robinson DW, et al: Treatment of hypertrophic scars, keloids, and scar contracture by triamcinolone acetonide. Plast Reconstr Surg 38:209-218,1966 13. Diegelmann RF, Bryant CP, Cohen IK: Tissue alphaglobulins in keloid formation. Plast Reconstr Surg 59:418-423,1977 14. Culty M, Nguyen HA, Underhill CB: The hyaluronate receptor (CD44) participates in the uptake and degradation of hyaluronan. J Cell Bio1116:1055-1062,1992

952

ALAISH

15. McGary CT, Raja RH, Weigel PH: Endocytosis of hyaluranic acid by rat liver endothehal cells-Evidence for receptor cycling. Biochem J 257:875-884,1989 16. Alho AM, Underhill CB: The hyaluronate receptor is preferentially expressed on proliferating epithelial cells. J Cell Biol 108:1557-1565,1989

ET AL

17. Culty M, Miyake K, Kincade PW, et al: The hyaluronate receptor is a member of the CD44 (H-CAM) family of cell surface glycoproteins. J Cell Biol111:2765-2774,199O 18. Lindsay WK: Neoplasms, in Ravitch MM, Welch KJ, Benson CD, et al (eds): Pediatric Surgery. Chicago, Year Book Medical, 1979, pp 1485-1494, chap 133

Discussion P, Dillon (Hershey, PA): I’d like to thank the authors very much for the opportunity to discuss this paper. Again it presents some very elegant work from the laboratories that we have gotten accustomed to producing such results. I’d like to ask several questions, and would preface them with the fact that this laboratory showed the presence of hyaluronic acid in fetal wounds as being quite high. That has been one of the major differences between the environment of an adult wound, a chronic wound, and a fetal wound. Indeed that has been carried forth in other experiments now in which the addition of hyaluronic acid has been suppressive in collagen formation, inflammation, and decreased scarring. Could you address the differences in that philosophy and what you’re seeing now with a keloid wound? Because it’s the very keloid wound that one had hoped that hyaluronic acid might decrease. Secondly, you are comparing hypertrophic fibroblast to normal skin. And I’m wondering if that’s a fair comparison. Should you have been looking at normal wounded fibroblasts in a nonkeloid form to look at their metabolism of hyaluronic acid or their synthesis of hyaluronic acid? Also, have you looked at whether hyaluronic acid alone will suppress the synthesis of these hypertrophic fibroblasts? And finally, one word of caution regarding your conclusions. Using CD44 receptor density as being equal and then stating that it must be a change in the ‘metabolism may be somewhat flawed. There are at least five to seven isotypes of CD44. I’m wondering if

the use of hyaluronic acid labeled with fluorescein to actually look at the metabolism in the breakdown would be a next step in that experiment. S.M. Alaish (response):At first glance it would seem contradictory when we compare the keloid model presented today with the knowledge we have about fetal wounds. But I think I can explain it looking at two parts of this. First, in a paper presented here last year, we found that the HA receptor was four times greater in fetal cells compared with adult cells, implying a greater degradation, a greater turnover of HA and with whatever internal mechanisms that would implicate. Secondly, we found the TGF+l levels in fetal serum are less than in the adult, both total cytokine levels and active cytokine levels. In work done by Mike Longaker in the last issue of Wound Repair and Regeneration, he found that the fetal tissue had much less active TGF-l31 than the adult. So I think a combination of the decreased cytokine levels and an increased HA receptor density in the fetus can explain those differences. I think that to compare normal wound fibroblasts instead of using normal skin fibroblasts would be interesting. But we did not do that. Finally, to comment on the many types of CD44. In work done last year we used three different antibodies to CD44 and showed that the receptor density was unchanged. I didn’t present the results here today, but we used three different antibodies initially in this study and found insignificant differences in HA receptor density.