Tumor necrosis factor-α (TNF-α), interferon-α (IFN-α) and interferon-γ (IFN-γ) receptors on human normal and scleroderma dermal fibroblasts in vitro

Journal of Dermatological Science, 3 (1992) 82-90 0

82

1992 Elsevier

Science Publishers

B.V. All rights reserved.

0923-181 l/92/$05.00

DESC 00130

Tumor necrosis factor-a (TNF-a), interferon-a (IFN-a) and interferon-y (IFN-JJ) receptors on human normal and scleroderma dermal fibroblasts in vitro Brian Berman 1,2 and Juana Wietzerbin 3 ‘Dermatology

Department, University of California, Davis School

of Medicine,Davis, CA, U.S.A., 2Dermatology Service, V.A.

Medical Center, Martinez, CA. U.S.A. and 31NSERM Unit 196, Paris, France

(Received

15 August 1991; accepted

Key words: TNF; Interferon;

27 November

Receptor;

1991)

Scleroderma

Abstract

Interferons alpha and gamma (IFN-a, IFN-g) and tumor necrosis factor alpha (TNF-a) exert different regulatory effects on the proliferation and biosynthetic activities of human dermal fibroblasts. Inasmuch as these cytokines bind to specific receptors in order to exert their activities, the expression of IFN-a, IFN-y and TNF-a receptors on fibroblasts from human adult normal and scleroderma skin cultured in vitro were quantitated. Adsorption was detected by incubating confluent normal and scleroderma fibroblasts with various concentrations of [ ‘2SI]cytokine. Replicate experiments revealed 19,742 + 2057 (Kd = 1.15 x 10 ~’ M) TNF-r receptors per normal dermal fibroblast and 15,006 f 75 (Kd = 6.75 x lo- ” M) TNF-a receptors to its receptor on normal and scleroderma fibroblasts revealed 130- and per scleroderma fibroblast. Cross-linking ‘?-TNF-a lOO-kDa TNF-receptor complexes. Although no quantitative or qualitative differences were detected between these two cell types with regard to receptor numbers, TNF-a affinity or receptor protein as detected by radiolabelled TNF-a, differences were detected in levels of mRNA specific for TNF-r receptors. Northern blot analysis revealed normal fibroblasts to constitutively contain mainly mRNA specific for the 55-kDa TNF receptor and indicate that they are capable of responding to TNF-a-induced up-regulation of mRNA specific for the 75 kDa TNF receptor. Scleroderma fibroblasts, however, constitutively contain mRNA for both TNF receptors and fail to respond to TNF-a up-regulation of the message for the 75-kDa receptor for TNF. Normal fibroblasts expressed 3320 f 1781 IFN-a2a and 17,563 k 1575 IFN-y receptors with scleroderma fibroblasts expressing 3076 and 16.316 IFN-r2a and IFN-;t receptors, respectively.

Introduction Fibroblasts are responsible for the production and maintenance of the connective-tissue matrix Correspondence to: Brian Berman, Dermatology Service (190). V.A. Medical Center, Martinez, CA 94553, U.S.A. Abbreviations: DME: Dulbecco’s modified Eagle’s medium; PBS: phosphate buffered saline; dpm: disintegrations per minute; FCS: fetal calf serum (heat inactivated); hu-r: human-recombinant; IFN: interferon; TNF: tumor necrosis factor; K,: dissociation constant; TNF-R: TNF-receptor.

of all organs including the dermis of the skin. Following either traumatic or pathological tissue injury, resident fibroblasts undergo a metabolic activation and exhibit increased growth and synthetic rates required to effect fibrotic repair of injured tissue. While evidence of the presence of fibroblast-stimulatory cytokines in situ is limited, numerous in vitro studies demonstrate that a variety of cytokines [ l-81, including tumor necrosis factors r (TNF-r) and -/3 (TNF-P) [9,10] can stimulate the growth, directional migration,

83

and/or connective-tissue matrix component synthesis of cultured fibroblasts. Cultured fibroblasts derived from the early phases of traumatic wound tissue or from sites of pathological fibrosis, such as keloidal or sclerodermatous skin, display activated phenotypes characterized by increased production of the connective-tissue matrix components: collagen [ 1 l-141, glycosaminoglycans [ 11,151, and tibronectin [ 141, and retain their activated phenotypes for many generations in vitro [ 1l-13,15]. Interferons (IFN)-or, -/3and -y inhibit normal dermal tibroblast growth and collagen production [ 16-231 and stimulate fibroblast production of the collagen degrading enzyme, collagenase [ 241. Moreover, short-term exposure of cultured sclerodermal fibroblasts to IFNs results in suppression of their abnormally elevated collagen production and in induction of a reduced/normalized-collagen-producing phenotype which is retained in the absence of IFNs [ 17,231. Since TNF-a activates libroblasts and IFNs may play a major role in terminating the fibrotic response by persistently deactivating tibroblast functions, and as these cytokines bind to specific receptors in order to exert their activities, the in vitro expression of TNF-r, IFN-cr and IFN-preceptors on human adult normal and scleroderma dermal libroblasts were quantitated and compared. Materials and Methods

Recombinant Escherichia c&-derived human interferon-#%‘a (Hoffmann-LaRoche, Basel, Switzerland) and interferon-y(Roussel-Uclaf, France) of 2 x lox specific activities and had 5 x 10’ U/mg protein, respectively. Interferon antiviral activity was assayed on WISH cells by inhibition of cytopathic effects caused by vesicular stomatitis virus [25]. Human recombinant tumor necrosis factor-r (spec. act. : 5 x 10’ U/mg protein) was produced by Genentech, Inc. (South San Francisco, CA) and provided by Boehringer Ingelheim (Vienna, Austria). TNF-c( cytotoxic activity was assayed on L929 cells in the presence

of actinomycin ously [ 261.

D (1 pg/ml) as described

previ-

Fibroblast cultures For comparative studies between normal and scleroderma-derived libroblasts, skin specimens were taken from the clinically involved skin of two scleroderma patients and from the identical anatomical site on two normal subjects in the course of elective cosmetic surgery. Fibroblasts were grown to confluency in Dulbecco’s modified Eagle’s medium (DME) containing penicillin (100 U/ml), streptomycin (100 pg/ml), 2 mM glutamine and 10 9, heat-inactivated single lot endotoxin-free fetal calf serum (FCS) (M.A. Bioproducts), and subcultured after trypsinization. Fibroblasts were studied between the 3rd and 5th passages and maintained their normal or hyperbiosynthetic, scleroderma, phenotypes [lo]. Labellirlg of c>qtokines Iodination of cytokines was carried out by the chloramine-T procedure as described for IFN-y [27]. In brief, 10 pg pure human-recombinant (hu-r) TNF-r, IFN-x2a or IFN-g in 100 ~1 phosphate buffered saline (PBS) were allowed to react with 2 mCi “‘I (carrier-free NalZ51 200 mCi/ml) of 10 ~1 chloramine-T after the addition (0.5 mg/ml H,O). The reaction was terminated after 60 s at room temperature by the addition of 10 ~1 sodium bisultite (1 mg/ml H?O). Iodinated cytokine was separated from free iodine by gelfiltration on a pre-calibrated column of Sephadex G-25 (Pharmacia, Uppsala, Sweden) equilibrated with PBS containing l”,b FCS. After dilution in DME containing 15 9; FCS, aliquots (0.5 pg/ml) did not have significant loss of biological activity, were stored at - 70°C and were of a specific radioactivity ranging from 446 to 600 cpm/fm. Qtokine binding assay Adsorption isotherms were obtained by incubation of 2-3 x lo4 human normal and scleroderma fibroblasts at 4 “C for 2 h with various concentrations of ‘251-hu-r TNF-c(, IFN-a2a or IFN-7 in 96-well microtiter plates containing 5 y0

x4

endotoxin-free FCS. At least triplicate cultures were treated with each concentration, After incubation, the cells were washed four times in the cold with PBS containing 0.1 mM CaCl, and 0.1 mM MgCl, supplemented with 1% FCS. Cell-associated radioactivity was measured in a gamma-counter. Non-specific binding was determined in parallel, simultaneous experiments in the presence of a 50-fold excess of unlabelled cytokine in at least triplicate cultures for each concentration of [ ‘251]cytokine tested. Non-specific binding, which never exceeded 15 % of total counts at saturation, was subtracted from total counts for determining specific binding. The dissociation constants (Kd) of the TNF-a receptor on human normal and scleroderma fibroblasts were determined by Scatchard plot analysis. The number of cytokine receptors per cell (2 SD) were calculated at receptor-saturating concentrations of [ 1251]cytokine. Receptor cross-linking Jz51-TNF-a was chemically cross-linked to its receptor on human normal and scleroderma libroblasts as previously described [28 J. Triplicate confluent normal (2 x 106) and scleroderma (1 x 106) libroblast cultures were washed with PBS containing 0.1 mM CaCl, and 0.1 mM MgCl, supplemented with 1 y0 FCS and incubated for 2 h at 4 “C with 2 nM ‘251-TNF-a in the presence or absence of a lOO-fold excess of unlabelled TNF-r. The cell monolayers were then washed 4 times with cold PBS containing 0.1 mM CaCI, and 0.1 mM MgCl, supplemented with 1% FCS, and once with Ca2+, Mg2+, FCS-free PBS. Disuccinimidyl suberate (1 mM in dimethylsulfoxide) was added to normal and scleroderma fibroblast monolayers, incubated with 12’1TNF-a in the presence or absence of an excess of unlabelled TNF-a, for 15 min at 4 “C, and the reaction was halted with the addition of 10 mM glycine. Monolayers of normal and scleroderma tibroblasts incubated with ‘251-TNF-a were handled identically as outlined above but without

the addition of the cross-linking compound, disuccinimidyl suberate. Cells were lysed with 1% Nonidet P40 and 10 PM PMSF in PBS. Following centrifugation at 10,000 X g for 10 min, supernatants were analyzed by SDS-PAGE (7.5-15x) according to Laemlli [29]. Dried gels were autoradiographed with Kodak XAR-5 film using DuPont Cronex lightening and intensifying screens. RNA isolation and mRNA detection Fibroblasts were cultured in 300-cc flasks to 90 % confluency as described above. Monolayers were washed with PBS and then incubated at 37 “C for 5 h in the presence and absence of TNF-c( (IO4 U/ml). Tissue culture media were then removed, cells washed with PBS and total cellular RNA was extracted by the guanidinium isothiocyanate/cesium chloride method, as previously described [ 301. The amount of total RNA, after determination of concentration by optical density at 260 nm, was equalized for rRNA [ 3 11, and lo-20 pg of total RNA were glyoxylated and fractionated on 1.1% agarose gel in 10 mM phosphate buffer, pH 7. Following electrophoresis, separated mRNAs were analyzed by Northern blotting after transfer to genescreen membranes according to Thomas [ 321. Prior to transfer of the RNA from the gel to the genescreen membranes, gels prepared in parallel were permeated with ethidium bromide, visualized with ultraviolet light and photographed. All RNA samples displayed the characteristic bands of ribosomal RNA at 18s and 28s without evidence of RNA digestion products. The filters were then hybridized using stringent conditions with a 2.1-kb EcoRI fragment of cDNA for the 55-kDa TNF-a receptor [33] and with a cDNA probe for the 75-kDa TNF-or receptor, labelled with [ a-‘2P]dATP and [a-32P]dCTP by the random priming method. The membranes were then subjected to autoradiography with Hypertilm (Amersham) using a DuPont Cronex intensifying screen at - 70 “C.

85

Results TNF-cr receptors on human normal and scleroderma jibroblasts

Replicate experiments revealed 19,742 t 2057 high affinity (K, = 1.15 x 10 - 9 M) TNF-a recepnormal dermal fibroblast and tors per 15,006 + 75 high affinity (K, = 6.75 x lo- lo M) TNF-a receptors per scleroderma Iibroblast (Fig. 1). No significant differences were detected at cell passages 3,4 and 5 in the number of normal fibroblast TNF-a receptors (21,999, 17,970 and 19,257, respectively) or of scleroderma TNF-a receptors at passage 4 (14,953) or 5 (15,060). Similar TNF-r binding studies detected 16,900 TNF-cr

6 25 .10-10

1 25 .x)-9

receptors per normal fibroblast of an additional line of normal Iibroblasts, and 12,748 TNF-a receptors per scleroderma fibroblast of an additional line of scleroderma fibroblasts. Chemically cross-linking ‘251-TNF-a to its receptors on normal or scleroderma fibroblasts revealed two major TNF-a/receptor complexes, 130 and 100 kDa, suggesting a receptor component of 79 kDa and 49 kDa, respectively, if trimeric (51 kDa) TNF is bound (Fig. 2A). Smaller TNF-or/receptor complexes of 88 and 66 kDa were also detected, as were the monomeric (17 kDa), and covalently cross-linked dimeric (34 kDa) and trimeric (51 kDa) forms of TNF (formed while non-covalently bound to a

25 .10-g

125I.TNF

concentration (M)

1251~TNF

concenualion W)

Fig. 1. TNF-a binding to human normal and scleroderma fibroblasts. Adsorption isotherms were obtained by incubation of 2-3 x IO” human normal (A) and scleroderma (B) fibroblasts at 4 “C for 2 h with various concentrations of ‘Z51-hu r-TNF-x as described in Materials and Methods. Non-specific binding (m) was subtracted from total binding (0) to yield specific TNF-a binding (0).

86

k'J= kD=

34

17

Fig. 2. Densitometric scans of autoradiograph of SDS-PAGE analysis of ‘*‘I-TNF-a/TNF receptors. Human normal and scleroderma libroblasts were incubated for 2 h at 4 “C with 2 nM “51-TNF-~ in the absence (A) and presence (C) of a lOO-fold excess of unlabelled TNF-r and then cross-linked to its receptor with disuccinimidyl suberate under conditions described in Materials and Methods, Analysis of normal and scleroderma libroblasts incubated with “‘1-TNF-a without subsequent cross-linking are shown in B.

TNF receptor molecule and subsequently released). The specificity of the ‘251-TNF-or receptor bands was confirmed by their no longer being detected following either competitive binding with unlabelled TNF-cr (Fig. 2C) or when the crosslinking agent was not added (Fig. 2B). Although no quantitative or qualitative differences were detected between these two cell types with regard to receptor numbers, TNF-a affinity or receptor protein as detected by radiolabelled TNF-a, differences were detected in levels of mRNA specific for TNF-a receptors. Northern blot detection of constitutive and TNF-a-induced levels of mRNAs for TNF-a receptors type I (TNF-Rl; 2.5 kb) and 2 (TNF-R2; 4.0 kb) isolated from normal and scleroderma libroblasts is shown in Fig. 3. Exposure of normal libroblasts to TNF-a for only 5 h increased TNF-R2-specific mRNA from low constitutive levels resulting in a reduction in the TNF-Rl : TNF-R2 ratio. In contrast, scleroderma fibroblasts, not exposed to

TNF-X, contained TNF-R2 mRNA which greatly exceeded that found in non-induced, normal tibroblasts. Furthermore, scleroderma fibroblasts failed to respond to TNF-a exposure, maintaining a constant ratio of TNF-Rl : TNF-R2 mRNA. IFN-a receptors on human normal and scleroderma Jibroblasts

Normal libroblasts expressed 3320 + 1781 IFN-a2a receptors while scleroderma tibroblasts expressed 3076. No significant differences but progressively fewer normal fibroblast IFN-a receptors were detected at cell passages 3,4 and 5 (5333, 2684 and 1944, respectively). IFN-;l receptors on human normal and scleroderma Jibroblasts

Normal fibroblasts expressed 17,563 + 1575 IFN-y receptors and scleroderma fibroblasts expressed 16,3 16 IFN-y receptors. No significant differences but progressively slightly fewer normal

87

SCLERODERM

NORW

5 HR CONT TNF CONT TNF kB 4.0

I Nl--HZ

2.5

82.62 81.10 % Rl 17.38 18.90 96 R2 4.75 4.48 Rl : R2

TNF-R 1

93.21 6.78 13.73

90.09 9.90 9.88

Fig. 3. Autoradiograph of Northern blot for TNF-a-receptor 1 and receptor 2 specific mRNA of normal and scleroderma fibroblasts. Confluent cultures of human normal and scleroderma dermal fibroblasts, incubated 5 h in the presence and absence of TNF-r (10” U/ml). As described in Methods, mRNA isolated from the cultured tibroblast and transferred to a membrane was hybridized for 16 h with “P-labelled cDNAs to mRNAs of human TNF-LYreceptors types 1 (75 kDa) and 2 (55 kDa). Percentages and ratios of TNF-RI and TNF-R2 mRNA were calculated from densitometric analysis.

libroblast IFN-y receptors were detected at cell passages 3, 4 and 5 (19,103, 17,632 and 15,954, respectively). Discussion

The initial step for cytokines to exert their effects is the binding to specific plasma membrane receptors on target cells. Employing ‘251-tagged purified TNF-a has allowed investigators to detect receptors displaying high affinity for TNF-a on various cells, including murine L929 fibroblasts and neonatal human foreskin (FS-4) fibroblasts [34]. We have been successful in scaling down to cellular requirement for detecting specific cytokine binding from 1x106 to 20-30,000 cells per sample and have modified the technique to perform binding studies on adherent cells. These technical changes will allow for the detection of specific cytokine binding to a limited number of cells of other adherent primary cell

cultures. The 2- to 3-fold greater numbers of TNF-a binding sites we detected per cell compared to the 7500 sites on FS-4 cells may reflect a difference between neonatal and adult fibroblasts or the anatomical source. Although we did not detect a significant difference in the number of TNF-a receptors on normal fibroblasts over 3-5 cell passages, the effect on the number of such sites of extensive passaging/cell division of diploid cells in vitro is unknown. Dermal fibroblasts of low passage derived from another patient’s sclerodermatous skin (data not shown) bound TNF-LY to a similar degree as those studied in this report. We have detected a TNF-a/receptor complex of 100 kDa on normal and scleroderma libroblasts, as has been previously reported on human cells [35,36] and may represent the 55-kDa receptor for TNF. Our detecting TNF-a/ receptor complex of 130 kDa, in addition to the 100 kDa complex, is consistent with the murine system in which cross-linking studies revealed 2 polypeptides (95 and 75 kDa) apparently capable of binding murine TNF [ 371 and probably represents the 75 kDa receptor for human TNF. Normal and scleroderma dermal tibroblasts did not differ in the numbers of expressed high affinity TNF-x receptors or in their numbers of IFN-z and IFN-;J receptors. The scleroderma fibroblasts studied maintained a super-normal collagen and glycosaminoglycan producing phenotype at the passages used, therefore the similarity with normal fibroblasts with respect to expression of cytokine receptors was not due to clonal selection of normal fibroblasts from the scleroderma skin biopsy. IFN-x (and its subspecies) and IFN-fi share a common receptor which is distinct from the receptor for IFN-y, with TNF-a and TNF-/3 also sharing a common receptor [reviewed in 381. Our detecting the expression of these three receptors with their different specificities is consistent with our previously reporting similar effects on fibroblast biosynthetic activities by IFN-a and IFN-fi and by TNF-a and TNF-P. The effects differed for each cytokine class and differed from those of IFN-y [ 9,11,24]. We appreciate that the presence

88

of specific receptors is a requirement for cytokine activity, however it is accepted that neither does the presence of receptors invoke target cell sensitivity to interferon or tumor necrosis factor, nor is there always a correlation between the degree of receptor expression and sensitivity. The correlation of receptor affinity for its specific cytokine may correlate with target cell sensitivity better than the absolute number of receptors. Differential sensitivity to cytokines by libroblasts derived from patients with scleroderma when compared to normal fibroblasts has been reported. The relative insensitivity of scleroderma fibroblasts to the mitogenic effect of platelet derived growth factor (PDGF) is probably not due to reduced numbers of receptors for PDGF, as PDGF failed to affect collagen synthesis of scleroderma fibroblasts differently from normal controls [ 391. It is interesting therefore that scleroderma tibroblasts failed to respond to TNF-a up-regulation of the message for the 75kDa receptor for TNF as did normal fibroblasts. It appears that scleroderma fibroblasts constitutively maintain elevated levels of mRNA for the 75-kDa TNF receptor, possibly due to their having been activated in vivo, perhaps by TNF-X, to maximally express TNF-R2 mRNA. It is presently unknown whether other ‘activated’ fibroblasts isolated from fibrotic areas, such as keloids, also elaborate increased levels of mRNA for TNF-R2 or whether this finding is specific for scleroderma fibroblasts. It is the 75kDa, and not the 55-7-kDa, receptor for TNF which is involved in TNF-or and TNF-P-induced proliferation of human B-lymphocytes [40]. Recently it has been reported that the two different receptors for TNF-a mediate distinct cellular responses, with TNF-Rl signalling cytotoxicity and manganous superoxide dismutase responses and TNF-R2 directing stimulation of proliferation [ 4 11. It remains to be determined whether elevated mRNA for the 75-kDa receptor for TNF impacts on the functioning of scleroderma cells or whether it is simply a marker of the activated state of these cells. Transforming growth factor-p (TGF-/?) enhances glycosaminoglycan synthesis by normal

dermal [ 91 and scleroderma libroblasts [ 421. Progressive systemic sclerosis (scleroderma) fibroblasts have been reported to be stimulated to a greater extent by TGF-j? to synthesize glycosaminoglycans than normal control fibroblasts [ 421. Interestingly, fibroblasts derived from indurated scleroderma skin expressed 2.5-times the number of TGF-fl receptors present on normal control fibroblasts, however the relative affinities of TGF-Preceptors on these cells were not assessed [ 431. Acknowledgements

We thank Dr. Louis A. Tartaglia who performed the Northern hybridization for TNF-Rl and -R2 at Genentech, Inc. This work was supported in part by The Department of Veterans Affairs Research Service, The Rothschild Foundation and INSERM. References Sporn MB, Roberts AB. Waketield LM, de Crombruggle B: Some recent advances in the chemistry and biology of transforming growth factor-beta. J Cell Biol 105: 1039-1045, 1987. Deuel TF, Huang JS: Platelet-drived growth factor: Structure, function and roles in normal and transformed cells. J Clin Invest 74: 669-676, 1984. Oka Y, Orth DN: Human plasma epidermal growth factor/beta-urogastrone is associated with blood platelets. J Clin Invest 72: 249-259, 1983. Castor CW, Miller JW, Walz DA: Structural and biological characteristics of connective tissue activating peptide (CTAP-III), a major human platelet-derived growth factor. Proc Nat1 Acad Sci USA 80: 765-769, 1983. Postlethwaite AE, Snyderman R, Kang AH: The chemotactic attraction of human fibroblasts to a lymphocytederived factor. J Exp Med 114: 1188-1203, 1976. Myers SL, Castor CW: Connective tissue activation XV: Stimulation ofglycosaminoglycan and DNA synthesis by a polymorphonuclear leucocyte factor. Arthritis Rheum 23: 556-563, 1980. Lobb RR, Fett JW: Purification of two distinct growth factors from bovine neural tissue by heparin affinity chromatograph. Biochemistry 23: 6295-6299, 1984. Clemmons DR. Van Wyk JJ: Somatomedin-C and platelet-derived growth factor stimulate human tibroblast replication. J Cell Physiol 106: 361-367, 1981.

89 9 Duncan PR, Berman B: Differential regulation of collagen, glycosaminoglycan Iibronectin and collagenase activity production in cultured human adult dermal fibroblasts by interleukin l-alpha and beta and tumor necrosis factor-alpha and beta. J Invest Dermatol 92: 699-706, 1989. 10 Bitterman PB, Wewers MD, Rennard SI, Adelberg S, Crystal RG: Modulation of alveolar macrophage-driven fibroblast proliferation by alternate macrophage mediators. J Clin Invest 77: 700-708, 1986. 11 Berman B, Duncan MR: Short-term keloid treatment in vivo with human interferon-alpha,, results in a selective and persistent normalization of keloidal Iibroblast collagen, glycosaminoglycan and cohagenase production in vitro. J Am Acad Dermatol 231: 694-706, 1989. 12 Uitto J, Bauer EA, Eisen AZ: Scleroderma: Increased biosynthesis of triple-helical type I and type III procollagem associated with unaltered expression of collagenase by skin hbroblast in culture. J Clin Invest 64: 921-930, 1979. 13 Abergel PR. Pizzurro D, Meeker CA, Lask G, Matsuoka LY, Minor RR, Chu ML, Uitto J: Biochemical composition of the connective tissue in keloids and analysis of collagen metabolism in keloid tibroblast cultures. J Invest Dermatol 84: 384-390, 1985. 14 Fleischmajer R, Perlish JS, Krieg T, Timpl R: Variability in collagen and Iibronectin synthesis by scleroderma tibroblast in primary culture. J Invest Dermatol 76: 400-403, 1981. 15 Buckingham RB, Prince RK, Rodnan GP: Progressive systemic sclerosis (PSS, scleroderma) dermal hbroblasts synthesize increased amounts of glycosaminoglycan. J Lab Clin Med 101: 659-669, 1983. 16 Duncan MR, Berman B: Gamma-interferon is the lymphokine and beta-interferon the monokine responsible for inhibition of tibroblast collagen production and late but not early tibroblast proliferation. J Exp Med 162: 516-527, 1985. 17 Duncan MR, Berman B: Persistence of a reducedcollagen-producing phenotype in cultured scleroderma hbroblasts after short-term exposure to interferons. J Clin Invest 79: 1318-1324, 1987. 18 Postlethwaite AE, Smith GN, Mainardi CL, Seyer JM, Kang AH: Lymphocyte modulation of tibroblast function in vitro: stimulation and inhibition of collagen production by different effector molecules. J Immunol 132: 2470-2477, 1984. 19 Jimenez SA, McArthur WA, Rosenbloom J: Inhibition of collagen synthesis by mononuclear cell supernates. J Exp Med 150: 1421-1431, 1980. 20 Korn JA, Halushka PV, Leroy EC: Mononuclear cell modulation of connective function: suppression of tibroblast growth by stimulation of endogenous prostaglandin production. J Clin Invest 65: 543-554, 1980. 21 Duncan MR, Perlish JS, Fleischmajer R: Lymphokine/

22

23

24

25

26

27

28

29

30

31

32

33

34

35

monokine inhibition of fibroblast proliferation and collagen production: role in progressive systemic sclerosis (PSS). J Invest Dermatol 83: 377-384, 1984. Jimenez SA, Freundlich B, Rosenbloom J: Selective inhibition ofdiploid tibroblast collagen synthesis by interferons. J Clin Invest 74: 1112-l 116, 1984. Rosenbloom J, Feldman G, Freundlich B, Jimenez SA: Inhibition of excessive scleroderma fibroblast collagen synthesis by recombinant gamma interferon. Arthritis Rheum 29: 851-856, 1986. Duncan MR, Berman B: Differential regulation of glycasminoglycan, Iibronectin and collagenase production in cultured human dermal fibroblasts by interferon+alpha, beta and gamma. Arch Dermatol Res 281: 11-18, 1989. Vaquero C, Sanceau J, Weissenbach J, Beranger F, Falcoff R: Regulation of human gamma-interferon and beta-interferon gene expression in PHA-activated lymphocytes. J Interferon Res 6: 161-170, 1986. Sugarman B, Aggarwal BB, Hars PE, Figari SS, Palldino MA, Shepard MS: Recombinant tumor necrosis factor effect on proliferation of normal and transformed cells in vitro. Science 230: 943, 1985. Berman B, Ducan MR: Pentoxifylline inhibits normal human fibroblasts in vitro proliferation, collagen, glycosaminoglycan and tibronectin production and increases collagenase activity. J Invest Dermatol 92: 605-610, 1989. Wietzerbin J, Gaudelet C, Aguet M, Falcoff E: Binding and cross-linking of recombinant mouse interferon gamma to receptors in mouse leukemic L1210 cells: interferon gamma internalization and receptor down-regulation. J Immunol 136: 2451-2455, 1986. Laemlh, UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685, 1970. Vaquero C, Sanceau J, Catinot L, Andreu G, Falcoff E, Falcoff R: Translation of mRNA from phytohaemagglutinin-stimulated human lymphocytes: characterization of interferon mRNAs. J Interferon Res 2: 217-232, 1982. Thomson CB, Challoner PB, Neiman PE, Groudine M: Levels of c-myc oncogene mRNA are invariant throughout the cell cycle. Nature 314: 363-366, 1985. Thomas PR: Hybridization of denatured RNA and small DNA fragments transferred to nitrocellulose. Proc Nat1 Acad Sci USA 77: 5201-5205, 1980. Schall, TJ, Lewis, M, Koller, KJ, Lee, A, Rice, GC, Wong GHW, Gatanga, T, Granger, GA, Lentz, R, Raab, H, Kohr, WJ, Goeddel, DV: Molecular cloning and expression of a receptor for human tumor necrosis factor. Cell 61: 361-370, 1990. Tsujimoto M, Yip YK, Vilcek J: Tumor necrosis factor: specific binding and internalization in sensitive and resistant cells. Proc Nat1 Acad Sci USA 82: 7626, 1985. Tsujimoto M, Feinman R, Kohase M, Vilcek J: Charac-

90

36

37

38

39

40

terization and affinity crosslinking of receptors for tumor necrosis factor on human cells. Arch Biochem Biophys 249: 563, 1986. Stauber GB, Aiyer RA, Aggarwal BB: Human tumor necrosis factor-cc receptor. J Biol Chem 263: 19908-19912. 1988. Ku11 FC, Jacobs S, Cuatrecasas P: Cellular receptor for ‘2SI-labeled tumor necrosis factor: specific binding, affinity labeling, and relationship to sensitivity. Proc Nat1 Acad Sci USA 82: 5756, 1985. De Maeyer E, De Maeyer-Guignard J: lnterferons and Other Regulatory Cytokines. John Wiley and Sons, New York, 1988, pp. 67-90. Le Roy EC, Mercurio S, Sherer GK: Replication and phenotypic expression of control and scleroderma human fibroblasts: responses to growth factors. Proc Nat1 Acad Sci USA 79: 1286-1290, 1982. Erikstein BK, Smeland EB, Blomhoff HK, Funderud S,

Pryelz K. Lesslauer W, Espevik T: Independent regulation of 55 kD and 75 kD TNF receptors during activation of human peripheral blood B lymphocytes. Eur J lmmunol 21: 1033-1037, 1991. 41 Tartaglia LA, Weber RF, Figari IS, Reynolds C, Palladino MA, Goeddel DV: The two different receptors for tumor necrosis factor mediate distinct cellular responses. Proc Nat1 Acad Sci USA 88: 9292-9296, 1991. 42 Falanga V, Tiegs SL, Alstadt SP, Roberts AB, Sporn MB: Transforming growth factor-beta: selective increase in glycosaminoglycan synthesis by cultures of tibroblasts from patients with progressive systemic sclerosis. J Invest Dermatol 92: 301-303, 1989. 43 Julien JM, Falanga V: Plasma and fibroblast receptor levels of transforming growth factor-beta in progressive systemic sclerosis. J Invest Dermatol 90: 572, 1988.