Cytokine-Induced Expression of Transforming Growth Factor-α and the Epidermal Growth Factor Receptor in Neonatal Skin Explants

Cytokine-Induced Expression of Transforming Growth Factor-a and the Epidermal Growth Factor Receptor in Neonatal Skin Explants Istvan Valyi-Nagy, Pamela J. Jensen, Steven M. Albelda, and Ulrich Rodeck The Wi star Institute of Anatomy and Biology (IVN, UR); the Department of Medicine (SMA) and the Department of Dermatology (P]J), University of Pennsylvania Medical School, Philadelphia, Pennsylvania, U.S.A.

--------------------------------------------------------------------------------------------------------Hyperproliferative diseases of the epidermal keratinocytes, such as psoriasis vulgaris, are characterized by overexpression and altered distribution of the epidermal growth factor/ transforming growth factor (EG F/TG F)-a receptor, and of TGF-a itself. It is believed that overexpression of this lignad/receptor system contributes to the hyperproliferative state ofkeratinocytes in an autocrine fashion. However, little is known about the factors that regulate expression of the EGF /TGF-a receptor, as well as expression of TGF-a in stratified epithelium. We examined modulation of the immunoreactive EGF/TGF-a receptor and TGF-a eXfression in normal neonatal foreskin explants by a variety 0 cytokines present in psoriatic lesions. Human (l1U) recombinant (r) tumor necrosis factor (TNF)-a and interferon (IFN)-y induced EGF/TGF-a receptor and TGF-a expression by kera-

tinocytes as determined by immunohistochemistry. Neut~al­ izing antibodies to TNF-a and IFN-y inhibited upregulatlO n ofEGF /TGF-a receptors and TGF-a by the respective cytokines. Interleukin (IL)-8 induced expression of TGF-a, but not of its receptor. Other cytokines (TNF-P, IFN-P, IL-l~, IL-2, IL-3, IL-S, IL-6, granulocyte/macrophage colony-sUmulating factor, and macrophage colony-stimulating factor) did not alter the expression patterns of EGF/TGF-a receptors or TGF-a in normal neonatal skin explants. These eXperiments demonstrate that specific cytokines known t? be present in psoriatic lesions can induce normal epidermIS to express TGF-a and its receptor in a pattern similar to that observed in psoriatic skin. ] Invest Dermato! 99:350-356, 1992

---------------------------------------------------------------------------------------------------------

T

mice, under the control of the human keratin K14 promoter, leads to epidermal thickening, reflecting increases in keratinocyte nU.mbers and cell hypertrophy [7]. These results indicate an autocrl ne role for TGF-a in keratinocyte growth in vivo. . Activation of the TGF-a - dependent autocrine loop in keratlll O cytes may be regulated by cytokines. Various cytokines, such ad interleukin (IL)-1 [8], tumor necrosis factor-alpha (TNF-a) [9], an interferon-gamma (IFN-y) [10,11 ]' have been shown to increase steady-state TGF-a mRNA levels in cultured keratinocytes. In ~d­ dition, IFN-y appears to modulate EGF-receptor protein expreSSIOn of cultured keratinocytes. Both increased [12] and decreased [13} EGF-receptor expression was observed after IFN-y treatment 0 keratinocytes in culture. . However, studies with cultured keratinocytes cannot mimic the potential complex interactions between cells of different types nor between keratinocytes in distinct differentiation states. We haver therefore, examined the effect of various cytokines on the pattern expression of TGF-a and the EGF /TGF-a receptor in organ cU tures of normal neonatal skin.

he epidermis is one of the few tissues that is mitotically active throughout the lifespan of the organism. Two structurally and functionally related growth factors, epidermal growth factor (EGF) and transforming growth factor-a (TGF-a), appear to contribute to the proliferation of normal human keratinocytes during embryologic development and in tissue culture [1,2]. Addition ofTGF-a to cultured keratinocytes not only stimulates growth, but also induces endogenous TGF-a production, which is likely to amplify the growth response to exogenous TGF-a [3] . Proliferative effects of both growth factors, EGF and TGF-a, are mediated by activation of the tyrosine kinase moiety of the EGF /TGF-a receptor [4,5] that is expressed on the surface of keratinocytes in vivo and in vitro [6). Inhibition of the EG F/TG F-a receptor tyrosine kinase activity results in inhibition of keratinocyte proliferation [4]. Conversely, constitutive overexpression of TGF-a in the skin of transgenic

1-

Manuscript received January 29, 1992; accepted for publication February 27, 1992. This work was supported in part by the Eleanor Naylor Dana Foundation. Reprint requests to: Dr. Ulrich Rodeck, The Wistar Institute of Anatomy and Biology, 3601 Spruce Street, Philadelphia, PA 19104. Abbreviations: EGF: epidermal growth factor hu: human IFN: interferon IL: interleukin MoAb: monoclonal antibodies r: recombinant TGF: transforming growth factor TNF: tumor necrosis factor

0022-202X/92/S05.00

MATERIALS AND METHODS ere Organ Culture Short-term organ cultures of skin samples -:v carried out as described [14,15] with minor modifications. BrJ~Y' ll neonatal foreskins, obtained from four donors, were immerse. ci MCDB 153 medium containing 0.03 mM calcium, amino aC s~ ethanolamine, phosphoethanolamine, and hydrocortisone as e scribed, but no pituitary extract or polypeptide growth factors (~r medium) [16]. After removal of fatty tissue from the dermal 51 e skin samples were cut into approximately 2-mm square pieces an placed epidermal surface up in wells of 6-well tissue culture plates.

d

d

Copyright © 1992 by The Society for Investigative Dermatology, Inc.

350

VOL. 99, NO.3

SEPTEMBER 1992

Base medium (approximately 800,u1) containing cytokines was added. The epidermal surface of explants remained uncovered by cUlture medium. Explants were incubated at 37"C in a humidified Incubator. In control cultures medium was not supplemented with cy~okines, or cytokines were mixed with neutralizing antibodies pnor to addition to cultures. Explants were harvested after 24 h, embedded in OCT compound (Tissue-Tek, Miles, Inc., Elkhart, IN) , and snap-frozen at - 70°C. Cytokines, Antibodies, and Immunohistochemistry Explants were incubated with the following cytokines at different concentrations: human (hu) recombinant (r) IL-8 (Pepro Tech Inc., Rocky Hill, NJ) at 5, 10, and 25 ng/ml; hu rTNF-O' (Genentech, San Francisco, CA) at 500, 1,000, and 10,000 U/ml; hu rIFN-y (Genentech) at 500,1,000, and 10,000 U/ml; hu rIL-IO' (Genetics Institute, Cambridge, MA) at 25, '50, and 100 U /ml; hu r granulocyte/macrophage colony-stimulating factor (Genetics Institute) at 1,000 and 10,000 U / ml; hu r macrophage colony-stimulating fac~or (Genetics Institute) at 1,000 and 10,000 U /ml; hu rIL-3 (GenetICS Institute) at 700 and 1,400 U /ml; hu rIL-5 (Genetics Institute) at 50, 100, and 500 U /ml; hu rIL-6 (Genetics Institute) at 10, 20, and 4~ U/ml (Genetics Institute); hu rTNF-p (Cetus Corp., Emoryvdl e , CA) at 1,000 and 10,000 U /ml; and hu rIFN-P (Genentech) at 1,000 and 10,000 U /m!. For blocking experiments, neutralizing monoclonal antibodies (MoAb) to TNF-O' (MoAb 154.2.1) and IFN_y (mAb B33.31), kindly provided by Dr. G. Trinchieri, were Usedsimultaneously with cytokines at 1: 200 and 1 : 400 dilutions of aSCitic fluid. . For immunohistochemical staining, unfixed, 4-6 ,urn thick frozen sections were incubated for 1 h with MoAb recognizing hu TGF-O! (Ab2; Oncogene Science, Inc. , Manhasset, NY) and the EGF /TGF-O' receptor (MoAb 425) [17] at 5 and 10 ,ug/ml, respectively. Ab2 recognizes huTGF-O! in its free and receptor-bo~nd form (communication by Oncogene Science). MoAb 425 [18] bmds specifically to a protein epitope on the external domain of the hu EGF/TGF_O! receptor. It immunoprecipitates a 170-kDa protem fr?m 35 S_labeled A 431 cells, recognizes a 32P-labeled 170-kDa protem in EGF-stimulated A 431 squamous carcinoma cells and placental membranes, and binds to the proteolytically cleaved 1 ~O-kDa external domain ofhu EGF /TGF-O! receptor [18]. It recogmzes th.e occupied as well as the unoccupied EGF /TGF-O! receptor because It Immunoprecipitates a 1251-EGF/170-kDa complex from various human cell lines and tissues [18] . Scatchard analysis demonstrated that MoAb 425 binds to high- and low-affinity EGF/TGF-c::" receptors with kinetics very similar to EGF itself [18]. No prot.ell~ other than the EGF/TGF-O! receptor is detected by MoAb 425 m Immunoprecipitation analyses of cell membrane preparations of A 431 squamous carcinoma cells. . Antibody binding was visualized using a fluorescein Isothiocyanate-Iabeled goat anti-mouse immunoglobulin G at 1: 100 dilution Qackson ImmunoResearch, West Grove, PA). Cy~okine-treated organ cultures were evaluated parallel to samples InCUbated with cytokines and antibodies. In order to demonstrate viability of the organ cultures, immu~o­ per?xidase staining of sections was performed usin~ antibodies agamst endothelial cell adhesion molecules, one of wll1ch has been preViously shown to be upregulated by cytokines in orga.n cultl~re l14]. Control and TNFO!-stimulated cultures were thus stamed with an anti - PECAM-l / endoCAM MoAb that specifically·identifies all endothelial cells in tissue sections [19,20] and with an anti-Eselectin (ELAM-l) antibody H4/18 [14], kindly provided by Dr. ~lchael Bevilacqua. The bound first antibody was detected by the ectastain ABC Elite kit (Vector Laboratories, Burlingame, CA) With AEc as chromagen. No counterstain was applied in order to Increase sensitivity. Photographs were taken with a Zeiss Photomicroscope III, using Kodak T-MAX 100 film. RESULTS Morphology and Viability of Skin Explants After 24 h of organ culture, the keratinocytes appeared intact and similar in ap-

CYTOKINE MODULATION OF TGF-a AND ITS RECEPTOR

351

pearance to normal skin after hematoxylin-eosin staining (Fig lA). To confirm cellular viability, sections were stained with antibodies that specifically recognize endothelial cell adhesion molecules. The entire vasculature of the graft was visualized using the antiPECAM-l MoAb (Fig IB), and illustrates that the endothelium was intact and expressing normal cell surface antigens. To determine if these cells could respond to cytokine stimulation, sections of organ cultures exposed to base medium (Fig le) and cultures exposed to 10,000 units of TNF-O! (Fig ID) for 24 h were stained for the presence of E-selectin. Although E-selectin was completely absent in control sections (Fig le), a small percentage of vessels in the TNF-O!-treated culture showed clear expression ofE-selectin (arrowheads in Fig ID) . This relatively weak, but significant, upregulation of E-selectin at 24 h is similar to that reported by Messadi et al [14]. Effects ofTNF-O! and IFN-y on TGF-O' and EGF /TGF-O' Receptor Expression in Skin Explants Following 24 h of incubation with MCDB 153 base medium containing 0.03 mM calcium, amino acids, ethanolamine, phosphoethanolamine, and hydrocortisone, but no cytokines, a low-grade (2+) expression of TGF-O' was detected predominantly in the intermediate and upper layers and, to a lesser degree, in the basal-suprabasal region of skin explants (Figs 2A and 4A). After incubating samples with 1000 U / ml TNF-O' for 24 h, a marked increase in TGF-O! expression (4+) was noted in the intermediate and upper skin layers (Fig 2B). Also, in a few sectors, basal and suprabasal keratinocytes showed increased staining (2+) for TGF-O' (Fig 2B). Upregulation ofTGF-O' was also observed when samples were treated with 500 U /ml of TNF-O' (not shown). Expression ofTGF-O' was somewhat reduced, as compared to untreated controls (Fig 2e), when explants were incubated simultaneously with TNF-O! and MoAb 154.2.1, a neutralizing antibody to TNF-O'. Incubation with IFN-y at 1000 U /ml, but not at 500 U /ml, similarly resulted in higher expression of TGF-O', although staining intensity was less (3+) than that observed after addition ofTNF-O! (Fig ID). Incubation with IFN-yin the presence of MoAb B33 .' 1, a neutralizing antibody to IFN-y, prevented the upregulatiO! <)f TGF-O' expression (Fig lE). Expression of the EGF/TGF-O' receptor was confined to basalsuprabasal cells in untreated control samples (Fig 3A). Following incubation with either TNF-O' (500 or 1000 U/ml) or IFN-y (1000 U /ml), keratinocytes in all layers of the epidermis were evenly stained for the EGF /TGF-O! receptor (Fig 3B,D). Similar to results described for TGF-O!, upregulation of the EGF/TGF-O! receptor could not be observed when TNF-O'- or IFN-y-induced effects were inhibited with the respective neutralizing antibody (Fig 3C, E). Furthermore, treatment of cultures with the TNF-O' antibody alone resulted in reduced staining for the EG F/TG F-O' receptor when compared to untreated controls, suggesting that endogenous TNF-O' may contribute to EGF/TGF-O' receptor expression. Effect ofIL-8 on TGF-O! and EGF /TGF-O! Receptor Expression in Skin Explants Treatment of skin explants with 5 ng/ml hu rIL-8 resulted in a marked upregulation ofTGF-O! in all layers of the epidermis (4+ versus 1 or 2+ in controls) (Fig 4A,B). Similar to results obtained with IFN-y, basal and suprabasal cells stained only weakly for TGF-O! in control cultures. No modulation of EGF/ TGF-O' receptor expression by IL-8 was observed in three consecutive experiments carried out on tissues from different donors (not shown). Effects of Other Cytokines on TGF-O' and EGF /TGF-O' Receptor Expression in Skin Explants No stimulation ofTGF-O! and EG F/TG F-O! receptor expression was detected in four consecutive experiments using material from different donors when skin explants were treated with hu rIL-IO!, IL-3, IL-5, or IL-6. Similarly, TNF-P, IFN-P, macrophage/colony-stimulating factor, and granulocyte macrophage/colony-stimulating factor were ineffective when tested in three separate experiments.

352

VALYI-NAGY ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure 1. Morphology and viability of skin explants. Skin explants were stained after 24 h of organ culture. Hematoxylin-eosin staining reveals th~t :~~ keratinocytcs are normal in appearance (A). The microvasculature of the explants were visualized by immunohistochemical staining with the endo~ e its cell-specific antibody PECAM-l (B). The viability of the cells within the explants was determined by exposing organ cultures to base medium or 10,00 f~he ofTNF-a for 24 h and then staining for the inducible endothelial cell adhesion molecule, E-selectin. No expression ofE-selectin was seen on the vessel~~ J1.t11. control cu lture (C). In contrast, clear expression of E-selectin (arrowheads) was seen in some of the vessels in the TNF-a-treated graph (D) . Bars,

VOL. 99, NO.3

SEPTEMBER 1992

CYTOKINE MODULATION OF TGF-Q AND ITS RECEPTOR

353

Figure 2. TGF-a expression of neonatal skin explant cultures treated with IFN -y or TNF-a. Explants were incubated in MCDB 153 base medium with no cytokine additive (A), or in the presence ofTNF-a (B) or IFN-y (D); on B, arrowheads point at TGF-a-positive areas in the basal-suprabasal region. C and E show inhibition of cytokine-induced TGF-a expression by neutralizing antibodies to TNF-a and IFN-y, respectively. Bars, 30 jim.

354

VALYI-NAGY ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure 3_ EGF-receptor expression of neonatal skin explant cultures treated with IFN-yorTNF-a. Explants were incubated in MCDB 153 base rnedlurn with no cytokine additive (A), or in the presence ofTNF-a (B) or IFN-Y (D). C and E show inhibition of cytokine-induced EGF-receptor expressiOn by neutralizing antibodies to TNF-a and IFN-y, respectively. Bars, 30 Jirn.

VOL. 99, NO.3

SEPTEMBER 1992

Ftgure 4. Induction of TGF-a expression in skin explants by IL-8. Exp ants were incubated inMCDB base medium alone (AJ or 111 the presence

of IL-8 (BJ. Bars, 30,um.

DISCUSSION In Contrast to earlier studies on cultured keratinocytes, Out experi-

rnents evaluated the cytokine-induced upregulation ofTGF-a and Its receptor on keratinocytes in situ. This approach takes into acCOUnt that the net result of cytokine-mediated effects may be critically dependent on the complex interaction of different cell types nd distinct differentiation states that are present in stratified skin, Ut absent in keratinocyte cu ltures. E The cytokine-induced expression pattern of immunoreactive GF /TG F-a receptors in normal neonatal skin is strikingly similar to the pattern observed in active psoriatic lesions. In stratified nor~al epithelium, expression of immunoreactive EGF /TGF-a recepo;s IS restricted to basal and immediate suprabasal keratinocytes ([. 1] and this study) . However, in active psoriatic lesions, expresSl~n of EGF receptors is observed not only in the basal but also in the sb.comeal layers of stratified epithelium [6,22], reminiscent of the ~ IqUltouS expression ofEGF /TGF-a receptors in fetal skin [21]. In 11s study, TNF-a and IFN-y treatment of skin explants produced a uniform expression of EGF /TGF-a receptors across all nucleated 1trata of the epidermis, similar to the pattern observed in psoriatic eS10 ns Or fetal epidermis.

b

CYTOKlNE MODULATION OF TGF-a AND ITS RECEPTOR

355

Expression of immunoreactive TGF-a protein has been reported to be generally enhanced, but not significantly redistributed, in active psoriatic lesions when compared to normal skin or uninvolved skin from psoriatic patients [23 - 25]. This finding is reflected by elevated TGF-a mRNA levels in psoriatic plaques [10,24]. Enhanced expression of TGF-a mRNA [24] and protein [25] was observed, particularly in psoriatic suprabasal keratinocytes. This is the same location in which we found the strongest cytokineinduced increase of immunoreactive TGF-a in normal epidermis. At variance with findings ofTutbitt et al [24], TGF-a immunostaining of basal normal keratinocytes in control explants was weak. However, treatment with IL-8 led to detection of elevated TGF-a protein levels not only in the suprabasal stratum but also in basal keratinocytes. Out results suggest that the pattern of the EGF /TGF-a receptor and TGF-a expression as observed in psoriasis is not unique to this disease, but can be induced in normal neonatal skin and, thus, may be comparable to a physiologic response to cytokine stimulation of keratinocytes. Activated immune cells such as T-Iymphocytes, monocytes/macrophages, and Langerhans cells are obvious sources for cytokines such as IFN-y and TNF-a in active psoriatic lesions. In fact, Nickoloff et al [9] recently described that TNF -a is expressed predominantly by macrophages in the papillary dermis of psoriatic lesions, and focally by epidermal Langerhans cells and keratinocytes. Suprabasal keratinocytes immediately above dermal TNF-acontaining macrophages appear to produce IL-8 , suggesting a cascade of cytokine expression involving several cell types. IFN-y is a major cytokine released by activated T lymphocytes, and T lymphocytes are found at elevated levels in psoriatic skin (26]. Interestingly, elevated levels of IFN-y have been reported in sera of patients with psoriasis [27], and patients with psoriasis injected with this cytokine to treat psoriatic arthropathy developed punctiform psoriatic foci at the injection site [28], suggesting a direct involvement of this cytokine in the pathogenesis of skin lesions in psoriasis. In contrast to TNF-a and IL-8, IFN-y has not been found to be produced by keratinocytes (for review see [29]), although IFN-y mRNA was detected in psoriatic epidermis using the sensitive reverse transcription polymerase chain reaction [30]. Taken together, these results suggest that in psoriatic skin IFN-y may be produced by an epidermal cell type other than keratinocytes. It should be noted that comparatively high levels of TNF-a (500 U /ml) and IFN-y (1000 U /ml) were needed to induce significant changes ofTGF-a and EGF /TGF-a receptor expression when compared to monolayer cultutes [11]. However, the cytokine concentrations used here are comparable to the levels ofIFN-y (100 and 200 U /ml) required to induce ICAM-l and HLA-OR on keratinocytes in situ [31]. By contrast to monolayer cultures, in the explant system access of cytokines to target cells is a critical determinant. Variables such as size of the skin explant, media composition, and modifications in technique may all influence diffusion of cytokines into the explant. In support of this notion, Griffiths et al [31] found that complete immersion of skin explants in culture medium leads to cytokine-induced expression of HLA-OR on keratinocytes, whereas partial immersion of primarily the dermal portion does not [15]. We chose to use the latter protocol based on the assumption that cytokines with possible relevance in vivo would gain access to keratinocytes by diffusion from the dermis up. Of the other cytokines included in this study, IL-l [32,33] and IL-6 [34] have been found at elevated levels or differently distributed in psoriatic lesions as compared to normal skin. Under the experimental conditions chosen in this study, these two cytokines did not affect TGF-a or TGF-a/EGF receptor expression detectably. However, the role of these cytokines in these phenomena cannot be discounted. Cytokine-induced effects could only be evaluated within a short time period (24 h) after initiation of explant cultures. After this time period, cell viability decreased dramatically. It appears possible that some cytokines have indirect effects by triggering a sequence of e~ents that require a longer time period to evolve. In support of tillS notion, IL-l has been described to stimulate the production and release of IL-8 from keratinocytes [35,36].

356

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

VALYI-NAGY ET AL

Our results indicate that the pattern ofTGF-a and EGF/TGF-a receptor expression in human epidermis can be modulated by cytokines. Cytokines known to be present in psoriatic lesions enhance the intensity and/or expand the distribution of TGF-a and its receptor in normal epidermis, causing it to resemble the psoriatic state. We hypothesize that the altered expression ofTGF-a and its receptor observed in psoriatic lesions is secondary to an effect of a previously acting cytokine, either produced by infiltrating immune cells or the keratinocytes themselves.

We are gratiful to Dr. G. Trinchierijor supplying neutralizing antibodies to IFN-y and TNF-a, to Dr. M . Bevi/aqua jar the ELAM-1 antibody, and to Dr. G.F. Murphy jar advice all the skin explant technique.

15.

16.

17.

18.

19.

REFERENCES 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

Rheinwald ]G, Green H: Epidermal growth factor and multiplication of cultured human epidermal keratinocytes. N ature 265:421-424, 1977 Willie]], Pittelkow MR, Shipley GD, Scott RE: Integrated control of growth and differentiation of normal human prokeratinocytes cultured in serum-free medium: clonal analysis, growth kinetics, and cell cycle studies.] Cell PhysioI121:31-44, 1984 Coffey RJ, Derynck R, Wilcox]N. Bringman TS, Goustin AS, Moses HL, Pittelkow MR: Production and auto-induction of transforming growth ' factor-a in human keratinocytes. Nature 328:8 17 - 820. 1987 Dvir A. Milner Y, Chomsky 0, Gilon C, Gazit A. Lev itzki A: The inhibition of EGF-dependent proliferation of keratinocytes by tyrpho~tin tyrosine kinase blockers. ] Cell Bioi 113:857 - 865, 1991 Cook PW, Pittelkow MR, Shipley GD: Growth factor-independent proliferation of normal human neonatal keratinocytes: production of autocrine- and paracrine-acting mitogenic factors.] Cell Physiol 146:277 -289, 1991 Nanney LB, Magid M, Stoscheck CM, King LE: Comparison of epidermal growth factor binding and receptor distribution in normal human epidermis and epidermal appendages. ] Invest Dermatol 83:385 - 393, 1984 Vassar R, Fuchs E: Transgenic mice provide new insights into the role ofTGF-a during epidermal development and differentiation. Genes Dev 5:714 - 727. 1991 Lee SW, Morhenn VB. I1nicka M, Eugui EM, Allison AC: Autocrine stimulation of interleukin-la and transforming growth factor-a production in human keratinocytes and its antagonism by glucocorticoids.] Invest Dermatol 97:106-110, 1991 Nickoloff BJ, Karabin GD, Barker ]N, Griffith CE, Sarma V. Mitra RS, Elder ]T, Kunkel SL, Dixit VM: Cellular localization of interleukin-8 and its inducer, tumor necrosis factor-alpha in psoriasis. Am] Pathol138:129-140, 1991 Elder ]T. Fisher G], Lindquist PB , Bennett GL, Pittelkow MR, Coffey R], Ellingsworth L, Derynck R, Voorhees JJ: Overexpression of transforming growth factor a in psoriatic epidermis. Science 243:811-814,1989 NickoloffBJ, Mitra RS, Elder JT, Fisher G], Voorhees JJ: Decreased growth inhibition by recombinant gamma interferon is associated with increased transforming growth factor-a production in keratinocytes cultured from psoriatic lesions. Br] Dermatol 121:161174,1989 Krueger JG, Krane ]F, Carter DM, Gottlieb AB: Role of growth factors, cytokines, and their receptors in the pathogenesis of psoriasis. ] Invest Dermatol 94: 135S - 140S, 1990 Nickoloff B], Mitra RS: Inhibition of 125I-epidermal growth factor binding to cultured keratinocytes by anti proliferative molecules gamma interferon, cyclosporin A, and transforming growth factor beta. ] Invest Dermatol 93:799 - 803, 1989 Messadi DV, Pober ]S, Fiers W, Gimbrone MA ]r, Murphy GF: Induction of an activation antigen on postcapillary venular endothelium in human skin organ culture.] Immunol139:1557 -1562,1987

20.

21.

22.

23.

24.

25.

26.

27.

28.

29. 30.

31.

32. 33.

34.

35.

36.

Messadi DV, Pober ]S. Murphy GF: Effects of recombinant gam ma interferon on HLA-DR and DQ expression by skin cells in shortterm organ culture. Lab Invest 58:61- 67. 1988 Boyce ST. Ham RG: Cultivation. frozen storage, and clonal growth of human epidermal keratinocytes in serum-free medium. ] Tissue Culture Methods 9:83-93,1985 Rodeck U, Herlyn M. Herlyn D, Molthoff C, Atkinson B, Varello M, Steplewski Z, Koprowski H: Tumor growth modulation by a monoclonal antibody to the epidermal growth factor receptor: immunologically mediated and effector cell independent effects. Cancer Res 47:3692-3696,1987 Murthy U,BasuA. Rodeck U, Herlyn M, RossAH. DasM: Binding of an antagonistic monoclonal antibody to an intact and fragmented EGF-receptor polypeptide. Arch Biochem Biophys 252:549- 560 , 1987 Albelda SM, Oliver P, Romer L, Buck CA: EndoCAM: A novel endothelial cell-cell adhesion molecule. J Cell BioI 110:1227 _1237,1990 Muller W A, Ratti CM, McDonel1 SL, Cohn ZA: A human endothelia~ cell-restricted externally disposed plasmalemmal protein ennch e in intercellular junctions. ] Exp Med 170:399-414, 1989 Nanney LB, Stoschek CM, King LE, Underwood RA , Holbrook Immunolocalization of epidermal growth factor receptors in norma developing human skin.] Invest Dermatol 94:742-748,1990 Nanney LB, Stoschek CM, Magid M, King LE: Altered [1251) epidermal growth factor binding and receptor distribution in pson aslS . J Invest Dermatol 86:260 - 265, 1986 . , of Gottlieb AB. Chang CK, Posnett DN, Fanelli B, T am ]P: DetectIOn transforming growth factor a in normal. malignant, and hyperproliferative human keratinocytes. J Exp Med 167:670-675, 1988 Turbitt ML, Akhurst R], White SI, MacKie RM: Localization ofel evated transforming growth factor alpha in psoriatic epidermiS, J Invest Dermatol 95:229-232, 1990 Finzi E, Harkins R. Horn T: TGF-a is widely expressed in differentiated as well as hyperproliferative skin epithelium.] Invest Dermatol 96:328-332. 1991 Baker BS, Swain AS, Fry L, Valdimarsson H : Epidermal T lympl~O~ cytes and HLA-DR expression in psoriasis. Br J Dermatoll10:55 564,1984 Gomi T , Shiohara T. Munakata T , Imanishi K. Nagashima M: In te.rleukin la, tumor necrosis factor a, and interferon y in pson aslS, Arch DermatoI127:827-830, 1991 Fierlbeck G, Rassner G, Moller C: Psoriasis induced at the injecti~~ site of recombinant interferon gamma. Arch Dermatol 126:35 355, 1990 McKenzie RC, Sauder DN: The role of keratinocyte cytokin es J~ inflammation and immunity.] Invest Dermatol 95: 105S -1 07S, 19 Barker IN, Karabin GD, StoofTJ, Sarma Vl" Dixit VM, Ni~koloffJ3!: DetectIOn of mterferon-gamma mRNA III psoriatiC epidermiS Y polymerase chain reaction. ] Dermatol Sci 2: 106 - 111, 1991 Griffiths CE, Voorhees]], Nickoloff B]: Characterization of intrace!d lular adhesion molecule-l and HLA-DR expression in normal and inflamed skin: modulation by recombinant gamma interferon an tumor necrosis factor.] Am Acad Dermatol 20:617 - 629, 1989 k' 1 Romero LI, Ikejama T, Pincus SH: In situ localization of interleu in normal and psoriatic skin. J Invest DermatoI93 :5 18-5ZZ, 1 Prens EP, B'::l1Jle K, Van Damme], Bakkus M, Brake! K, Benner~; Van Joost T: Interleukin-l and interleukin-6 in psoriasis. ] Inve DermatoI95:121S-124S.1990 Grossman RM. Krueger], Yourish D, Granelli-Piperno A, M.urfi~ DP, May LT, Kupper TS, Sehgal P, Gottlieb AB: Interleukln-. on expressed at high levels in psoriatic skin and stimulates prohfer~~71 of cultured keratinocytes. Proc Natl Acad Sci USA 86:6367 , Jeukm Larsen CG. Anderson AO, Oppenheim]], Matsushima K: Inter hil 1 or tumor necrosis factor stimulate the production of neutrop st activating protein by normal fibroblasts and keratinocytes. J Inve Dermatol92:467-471,1989 W Kristensen M, Paludan K, Larsen CG, Zachariae CO, Deleuran/ete r: Jensen PK, ]0rgensen P, Thestrup-Pedersen K: Quantltatlve an mination of IL-1a-induced IL-8 mRNA levels in cultured hum J keratinocytes, dermal fibroblasts, endothelial cells, and monocytes. Invest DermatoI97:506-510, 1991

KA;

1;89