Search

In malignant melanoma, tumor-infiltrating lymphocytes are frequently reactive with melanosomal antigens. Achieving complete remissions by peptide therapy is frequently hampered by metastases evading immune recognition. The tumor microenvironment seems to favor reduced expression of target antigens by melanoma cells. Among candidate factors, interferon-gamma (IFN-gamma) (10(2) to 10(3) U/ml) suppressed expression of antigens MART-1, TRP-1, and gp100 by M14 melanoma cells as shown by immunohistology and fluorescence-activated cell sorting analysis, reducing MART-1 expression by >65%. Northern blot analysis revealed that reduced expression was regulated at the transcriptional level, demonstrating a 79% reduction in MART-1 transcript abundance after 32 hours of IFN-gamma treatment. To evaluate consequences of IFN-gamma exposure for immune recognition, MART-1-responsive T cells were reacted with pretreated HLA-matched melanoma cells. Cytotoxicity was reduced up to 78% by IFN-gamma pretreatment, and was restored by addition of MART-1 peptide AAGIGILTV for 2 hours. Examination of melanoma lesions by quantitative reverse transcriptase-polymerase chain reaction revealed up to 188-fold more abundant IFN-gamma transcripts when compared to control skin. Laser capture microdissection and immunohistology localized most IFN-gamma-producing T cells to the tumor stroma. Reduced MART-1 expression was frequently observed in adjacent tumor cells. Consequently, IFN-gamma may enhance inflammatory responses yet hamper effective recognition of melanoma cells.

An important step in tumorigenesis involves loss of sensitivity to various apoptotic signals by malignant cells, imbuing them with an enhanced survival phenotype. NF-kappaB also regulates epidermal thickness, susceptibility to apoptosis, and tumor formation in skin. Keratinocytes were examined for their susceptibility to apoptosis using cytokines produced during an immunologic response to tumor antigens, i.e., interferon-gamma and/or tumor necrosis factor-alpha (TNF-alpha). The role for NF-kappaB in this response was examined using a retroviral vector containing a degradation-resistant form of IkappaBalpha. Whereas interferon-gamma and TNF-alpha either alone or in combination did not induce apoptosis in keratinocytes, after infection with the retrovirus to block NF-kappaB activation they became susceptible to TNF-alpha but not Fas-induced apoptosis. Moreover, when keratinocytes with repressed NF-kappaB activity were simultaneously treated with interferon-gamma, there was a synergistic induction of apoptosis by TNF-alpha that was dependent on FADD, tumor-necrosis-factor-related apoptosis-inducing ligand (TRAIL), and caspase activation. Molecular abnormalities accompanying repressed NF-kappaB activity included failure to induce TNF-RII receptor together with enhanced levels of TRAIL death receptor 4. The ability of interferon-gamma when combined with TNF-alpha to mediate keratinocyte apoptosis included induction of TRAIL coupled with diminished capacity of keratinocytes with repressed NF-kappaB activity to increase the TRAIL decoy receptor-1, as well as lower levels of several NF-kappaB-dependent antiapoptotic proteins accompanied by enhanced caspase 8 levels. These results indicate that interferon-gamma and TNF-alpha synergistically induce keratinocyte apoptosis when concomitant induction of NF-kappaB is blocked. Participants in the apoptotic response mediated by NF-kappaB, besides cell-survival proteins, include modulation of TRAIL and both death and decoy receptors. Thus, not only does NF-kappaB signaling influence the intrinsic survival pathway for keratinocytes in normal skin, but it may also play a role in determining the apoptotic response to cytokines generated during an immune response via TRAIL produced by the keratinocytes themselves.

Summary: Psoriasis is a chronic skin disease with an immunocytic infiltrate, including activated T lymphocytes, producing multiple cytokines that can influence the phenotype of epidermal keratinocytes. In these studies we examined the effect of the cytokines interferon-gamma and interleukin-13 or interleukin-4 on keratinocytes, alone and in combination, on surface levels of HLA-DR, intercellular adhesion molecule 1, and CDw60, as well as the transcription factors STAT1, STAT6, and BCL-6. As CDw60 is an acetylated form of the GD3 ganglioside and may function as a T cell costimulatory molecule, the modulation of CDw60 expression by keratinocytes in psoriatic lesions was highlighted to gain insight into potentially important T cell-keratinocyte interactions. Interferon-gamma was observed to block the interleukin-4- or interleukin-13-mediated induction of CDw60 on cultured keratinocytes, but not induction of the transcription factor STAT6. Interleukin-13 and interleukin-4 were unable to block interferon-gamma-mediated induction of STAT1 or BCL-6, however, or the upregulation of intercellular adhesion molecule 1 and HLA-DR. In psoriatic plaques, CDw60 was not consistently detected on keratinocytes in acute lesions, but was detected predominantly on basal layer keratinocytes in chronic lesions. In addition we found that BCL-6 levels were increased in psoriatic lesions; in acute lesions BCL-6 was primarily localized in the basal layer keratinocytes, whereas in chronic plaques nuclear BCL-6 was predominantly expressed by keratinocytes in the suprabasal cell layers. These studies highlight the complex modulation of the keratinocyte phenotype by immunocyte-derived cytokines, in which induction of CDw60 involving interleukin-4, or interleukin-13 was antagonized by interferon-gamma. We suggest in psoriatic plaques that the presence or absence of CDw60 expression by keratinocytes may reflect the dynamic interplay between Th-1-type cytokines such as interferon-gamma and Th-2-type cytokines such as interleukin-4 and interleukin-13. The ability of interferon-gamma to induce the transcription repressor BCL-6 may also contribute to the overall immunologic events in skin, including suppression of the intermediates in the synthetic pathway leading to expression of the T cell costimulatory ganglioside CDw60.

The MHC class I-like protein CD1d is a nonpolymorphic molecule that plays a central role in development and activation of a subset of T cells that coexpress receptors used by NK cells (NK-T cells). Recently, T cells bearing NK receptors were identified in acute and chronic lesions of psoriasis. To determine whether NK-T cells could interact with epidermal cells, we examined the pattern of expression of CD1d in normal skin, psoriasis, and related skin disorders, using a panel of CD1d-specific mAbs. CD1d was expressed by keratinocytes in normal skin, although expression was at a relatively low level and was generally confined to upper level keratinocytes immediately beneath the lipid-rich stratum corneum. In contrast, there was overexpression of CD1d in chronic, active psoriatic plaques. CD1d could be rapidly induced on keratinocytes in normal skin by physical trauma that disrupted barrier function or by application of a potent contact-sensitizing agent. Keratinocytes displayed enhanced CD1d following exposure to IFN-gamma. Combining CD1d-positive keratinocytes with human NK-T cell clones resulted in clustering of NK-T cells, and while no significant proliferation ensued, NK-T cells became activated to produce large amounts of IFN-gamma. We conclude that CD1d can be expressed in a functionally active form by keratinocytes and is up-regulated in psoriasis and other inflammatory dermatoses. The ability of IFN-gamma to enhance keratinocyte CD1d expression and the subsequent ability of CD1d-positive keratinocytes to activate NK-T cells to produce IFN-gamma, could provide a mechanism that contributes to the pathogenesis of psoriasis and other skin disorders.

The CD6 glycoprotein is expressed by T lymphocytes and is hypothesized to interact with one or more ligands expressed on antigen presenting cells (APCs). We show that CD6 mediates binding of the transformed CD4+ T cell line Hut 78 to gamma-interferon activated keratinocytes (KCs). A recombinant CD6-Ig fusion protein has been reported to bind to a CD6 ligand ALCAM, but this is the first demonstration that cell-cell adhesion of human T lymphocytes can be CD6 dependent. The known CD6 ligand ALCAM (CD166) is expressed on cultured KCs but does not appear to mediate KC-Hut 78 binding, suggesting the existence of additional CD6 ligands expressed on KCs. In functional studies using autologous KCs as APCs for tetanus toxoid specific T cell clones, KCs +/- gamma-interferon are unable to stimulate autologous T cells with recall antigen. Therefore interaction of T cell CD6 with CD6 ligands on KCs does not provide sufficient co-stimulation of primed T cells to support responses to nominal antigen.

By using superantigens, we have found previously that keratinocytes activated by IFN-gamma could serve as accessory cells, providing costimulatory signals needed to induce T cell proliferation. Here, we compared the profile of cytokines produced by T cells stimulated in the presence of activated keratinocytes with the response seen using professional APCs. When keratinocytes are used as accessory cells there is a specific defect in T cell IFN-gamma production, whereas IL-2 and IL-4 are induced at levels comparable with those seen when professional APCs are used as accessory cells. Because keratinocytes express BB-1, a CD28-ligand distinct from B7-1 or B7-2 (which are found on professional APCs), we examined the possibility that the defect in IFN-gamma production might be a result of nonproductive CD28 engagement. However, even when the CD28 pathway is directly activated by a stimulatory mAb, there is no induction of IFN-gamma production in keratinocyte-supported cultures. In these same cultures IL-2 production is increased 10-fold, thus demonstrating a specific deficiency in the induction of IFN-gamma rather than a failure to respond to CD28 stimulation. Analysis by reverse transcriptase-PCR and ELISA for the inducible p40 chain of IL-12 reveals that keratinocytes produce little if any messenger RNA and no protein for IL-12 p40 compared with professional APCs. Addition of rIL-12 to keratinocyte-supported cultures restores IFN-gamma levels to those seen when professional APCs are present. Finally, when T cells are restimulated and analyzed at later time points (10 to 14 days) we find a refinement in cytokine profiles: T cells stimulated in the presence of professional APCs produced the Th1 cytokines IL-2 and IFN-gamma, whereas T cells stimulated in the presence of activated keratinocytes produced only the Th2 cytokine IL-4. The specific ability of keratinocytes to induce a Th2 response seems most closely linked to their absence of IL-12 production, and may be important in the maintenance of peripheral tolerance to self-Ags or in the immune response to exogenous Ags, pathogens, or haptens encountered in skin.

The link between the epidermal keratinocytes of the skin and the activated T lymphocytes of the immune system is mediated by a variety of cytokines, including gamma interferon (IFN-gamma). We studied the influence of keratinocyte mitogens such as transforming growth factor-alpha (TGF-alpha), epidermal growth factor (EGF), and somatomedin-C (SM-C) on the ligand binding of 32P-labeled IFN-gamma to cultured keratinocytes derived from normal appearing adult human skin. Keratinocytes placed in a medium devoid of mitogens become growth arrested, and these quiescent cells expressed 2.4 times (28,900 versus 12,200 sites/cell) as many high affinity IFN-gamma receptors (Kd = 0.22 nM) compared to keratinocytes which were actively growing in medium containing TGF-alpha (25 ng/ml) or EGF (10 ng/ml). The reduction in IFN-gamma receptor sites by TGF-alpha/EGF was mitogen specific, as adding SM-C (500 ng/ml) did not have any effect on ligand binding, although it similarly stimulated keratinocyte growth. The reduction in IFN-gamma receptors was time dependent, occurring primarily after 24-48 hours of change in tissue culture conditions. The reduction in the number of high affinity IFN-gamma receptors by TGF-alpha/EGF had immunobiological consequences, because quiescent keratinocytes in basal medium had an increased expression of HLA-DR and intercellular adhesion molecule-1 (ICAM-1) induced by IFN-gamma, compared to actively growing TGF-alpha/EGF treated keratinocytes. These results suggest that rapidly proliferating keratinocytes exposed to TGF-alpha/EGF but not SM-C are capable of altering their response to IFN-gamma by decreasing their number of cell surface high affinity receptors for IFN-gamma.

Psoriatic skin lesions contain HLA-DR positive T lymphocytes, and other activation antigens, which suggest that the T cells may be producing lymphokines. Gamma interferon is produced by activated T cells, and its presence in psoriasis has been inferred by the lesional keratinocyte expression of 3 gamma interferon-inducible proteins i.e. HLA-DR, intercellular adhesion molecule-1, and gamma-IP-10. To determine whether gamma interferon is being produced directly in psoriatic lesions, punch biopsies of normal and diseased skin were separated into epidermal sheets and dermal fragments. Total cellular RNA was isolated from each epidermal and dermal compartment, and reverse transcribed followed by amplification of the resultant DNA by polymerase chain reaction. The amplification process involved the use of 5' and 3' primers for gamma interferon, and tumor necrosis factor-alpha, with beta-actin serving as a control. Gamma interferon mRNA, but not tumor necrosis factor alpha mRNA, was detectable in 4 of 5 psoriatic epidermal specimens. Neither mRNA was detectable in any normal skin dermal/epidermal specimens. Gamma interferon mRNA was also detectable in a single psoriatic dermal specimen. If reverse transcriptase was omitted, no polymerase chain reaction products were detected, indicating that the fragments detected were not derived from contaminating genomic DNA. These results indicate that gamma interferon mRNA can be extracted and successfully detected from human psoriatic lesional skin biopsies, using polymerase chain reaction technology. This molecular approach can easily be expanded to measure many other cytokines in both epidermal and dermal locations. The detection of gamma interferon in this clinical setting may be of particular pathophysiological significance because injection of gamma interferon has been reported to induce psoriatic lesions.

Monocytes accumulate in the epidermis and along the dermo-epidermal junction in several different inflammatory skin diseases. To determine whether human epidermal keratinocytes elaborate a specific chemotaxin responsible for the accumulation of monocytes at these anatomic sites, monocyte chemotactic activity in conditioned 16-h cultured keratinocyte supernatants were assayed using human peripheral blood monocytes as the target cell. Dilutional analysis revealed directed monocyte migration in IFN-gamma-treated (100 U/ml) keratinocyte supernatants (80% maximal FMLP response) which was 10-fold more than IFN-gamma itself or untreated keratinocyte activity alone. Gel filtration chromatography revealed that this activity eluted just ahead of a 12.5-kDa molecular mass marker. Blocking studies demonstrated that a rabbit polyclonal antibody to monocyte chemotaxis and activating factor (MCAF) inhibited all monocyte chemotaxis by greater than 80%. Keratinocytes were metabolically labeled with 35S-cysteine/methionine, and after 16 h incubation the supernatants immunoprecipitated with the same anti-MCAF antibody. MCAF was detected as a protein doublet of 12 and 9 kDa only in IFN-gamma-treated (100 U/ml) keratinocyte supernatants. Incubation with IFN-gamma and TNF-alpha (250 U/ml) in combination resulted in increased production of MCAF protein. By Northern blot analysis, MCAF mRNA was constitutively expressed in keratinocytes and upregulated only in the presence of IFN-gamma. TNF-alpha, IL-1 beta, transforming growth factor-beta and phorbol esters had no positive or negative influence on MCAF mRNA. These studies demonstrate that biologically active MCAF is elaborated by human epidermal keratinocytes upon activation by IFN-gamma, a cytokine also required for the induction of adherence between monocytes and keratinocytes. Keratinocyte-derived MCAF is likely to be important in the regulation of cutaneous monocyte trafficking and may also be responsible for the recruitment of Langerhans cells and dermal dendrocytes, which share many phenotypic features with monocytes/macrophages, to their anatomic locations in skin.

Adhesion of lymphocytes to target cells via certain cell surface molecules is important in cytotoxic T lymphocyte-mediated immune reactions. The binding of lymphocyte function-associated (LFA) antigens 1 and 2, with their respective ligands, intercellular adhesion molecule-1 (ICAM-1) and LFA-3, which are expressed on the surface of nonlymphoid cells, has been shown to be critical for lymphocyte adhesion. To determine whether basal cell carcinomas (BCCs) can escape immunodetection as a result of the inability of cytotoxic T lymphocytes to bind tumor cells, the expression of adhesion molecules on numerous BCCs, before and after exposure to interferon-gamma (IFN-gamma), was examined. Ninety-three percent of 30 freshly excised invasive BCCs did not express ICAM-1 and 73% of 11 BCCs did not express LFA-3. However, the normal-appearing basal keratinocytes in epidermis overlying nests of BCC, did express ICAM-1, particularly when a marked LFA-1+ and LFA-2+ dermal lymphocytic infiltrate was present. After BCC tissue was incubated in vitro with IFN-gamma the expression of ICAM-1 was induced on 85% of tumors studied. Thus tumor cells did not possess an absolute inability to express adhesion molecules; rather the constitutive absence of such molecules may be due to insufficient in vivo cytokine levels necessary to induce expression or a barrier preventing cytokines from reaching and interacting with tumor cells. We conclude that the absence of ICAM-1 and LFA-3 adhesion molecules is a mechanism by which BCCs can avoid immunosurveillance.

T lymphocytes and mononuclear cells preferentially accumulate in the epidermis in inflammatory skin disease. To determine the role of keratinocytes in both the chemotaxis and adhesion of these cells to the epidermis, cultured keratinocytes were incubated with IFN-gamma and tumor necrosis factor-alpha (TNF-alpha), and mRNA detected and quantitated for IL-8, monocyte chemotaxis and activating factor, and intercellular adhesion molecule-1. Whereas induction of these mRNAs was either absent, or relatively weak and transient, to either IFN-gamma or TNF-alpha alone, when administered in combination there was a dramatic increase and persistence in the induction of all three genes. Pretreatment of the keratinocytes with cycloheximide failed to eliminate transcription, implying that all three are primary response genes. Transforming growth factor-beta, which modulates other keratinocyte functions (not related to adhesion or chemotaxis of inflammatory cells) failed to induce any of the genes. These novel findings potentially explain the selective recruitment of T cells and monocytes observed in inflammatory skin disease, because IFN-gamma and TNF-alpha can co-ordinately regulate keratinocyte-derived chemoattractants and adhesion molecule production.

Because keratinocytes (KCs) express HLA-DR in a wide variety of skin diseases in which mononuclear leukocytes are observed in close apposition to KCs (i.e., graft-versus-host disease), and since gamma interferon (IFN-gamma) induces HLA-DR expression on KCs, we asked whether IFN-gamma treatment of KCs would influence the adherence of mononuclear leukocytes. When allogeneic peripheral blood mononuclear leukocytes (PBML) and a Leu-3+ T cell clone were coincubated with IFN-gamma-treated KCs (300 U/ml, 3 days), there was a marked increase in binding compared with nontreated KCs. Similar binding results were obtained using a cutaneous squamous carcinoma cell line (SCL-1) after IFN-gamma treatment. The IFN effect was relatively specific for IFN-gamma, as neither IFN-alpha nor -beta had any effect. Tumor necrosis factor exposure (500 U/ml, 3 days) increased the binding of the Leu-3+ T cell clone to both KCs and SCL-1 cells. Neutrophils displayed a less marked (but statistically significant) increase in binding to IFN-gamma-treated KCs. Using the Leu-3+ cell clone and SCL-1 cells, detailed kinetic analysis of the effect of IFN-gamma on binding was performed. The increased adherence between the cells began to appear after only 7 hours of treatment with r-IFN-gamma (300 U/ml) and reached a plateau at 48 hours, with significantly enhanced binding continuing for at least 48 hours after removal of IFN-gamma. The mechanism of binding was explored by preincubation of the PBML/Leu-3+ T cells with anti-LFA-1 (lymphocyte function-associated antigen) antibody (0.6-6.0 micrograms/ml), which totally inhibited the binding with no effect by anti-LFA-2 or -3 or class I or II antibodies despite documented binding of these antibodies to the cells. These results suggest that, after exposure to IFN-gamma, the ability of KCs to bind mononuclear leukocytes is strongly enhanced, and this adherence may be important in leukocyte trafficking in the skin as well as contributing to altered KC-leukocyte interaction, which may be of fundamental importance in a variety of skin disease.

With indirect immunofluorescence techniques we demonstrated that recombinant gamma-interferon induced the expression of the class II antigens HLA-DR and HLA-DQ as well as intercellular adhesion molecule-I (ICAM-I) on normal, cultured human keratinocytes grown in low-calcium, serum-free medium. Each antigen displayed a distinctive cellular staining pattern. HLA-DR was strongly localized to perinuclear zones with intense cell surface expression; HLA-DQ displayed a perinuclear accentuation, but with minimal cell surface staining, and ICAM-I was strongly expressed in a diffuse cytoplasmic pattern with intense cell surface expression. Keratinocytes grown in medium supplemented with 10% fetal calf serum underwent differentiation, with a diminished expression of all three antigens as compared to those grown in low-calcium, serum-free medium. These results confirm that gamma interferon can differentially regulate HLA-DR and HLA-DQ expression; that there are probably different biochemical metabolic pathways by which these three molecules are expressed on keratinocytes, and that the expression is also a function of the degree of keratinocyte differentiation. The strong cell surface expression of ICAM-I is suggested to be of major importance as the recognition molecule, by which T cells bind to gamma interferon exposed keratinocytes, and suggests an integral role for this molecule in epidermal lymphocyte trafficking.

When 2 mm punch biopsy specimens of normal skin are incubated with gamma interferon (IFN-gamma), the keratinocytes are induced to express intercellular adhesion molecule-1 (ICAM-1). Lymphocytes and monocytes that express lymphocyte function-associated antigen-1 (LFA-1) bind to cultured keratinocytes expressing ICAM-1. We have developed an in vitro adherence assay using frozen sections of normal skin that have been incubated with IFN-gamma, and then overlaid by peripheral blood mononuclear leukocytes. Although peripheral blood mononuclear leukocytes do not bind to the epidermal keratinocytes of untreated skin sections, after exposure of the skin to IFN-gamma the peripheral blood mononuclear leukocytes prominently bind to the epidermal keratinocytes that express ICAM-1. The binding by peripheral blood mononuclear leukocytes is increased approximately twofold by activation with phorbol ester treatment. The temperature dependence and kinetics of this adherence reaction reveal no binding at 8 degrees C, good binding at 24 degrees C, and optimal binding at 37 degrees C, reaching a maximal extent by 60 minutes. The adherence reaction is blocked either by pretreating the peripheral blood mononuclear leukocytes with LFA-1 antibody or the IFN-gamma-exposed skin specimen with ICAM-1 antibody. The immunophenotypic analysis of the adherent peripheral blood mononuclear leukocytes to epidermal keratinocytes revealed that most cells activated by 12-O-tetradecanoyl-phorbol-13-acetate are T lymphocytes, with CD8+ T cells binding slightly better than CD4+ T cells, with a smaller population of monocytes. These results provide additional support for a role of LFA-1, ICAM-1, and IFN-gamma in modulating keratinocyte-lymphocyte interactions.(ABSTRACT TRUNCATED AT 250 WORDS)

In normal human epidermis, expression of HLA-DR antigen is restricted to Langerhans cells (LC) and acrosyringial epithelium. However, in diseases such as lichen planus and graft-vs.-host, HLA-DR antigen appears to be expressed by keratinocytes, although the exact source of the HLA-DR is unclear. Two possibilities are that (1) the HLA-DR is shed by neighboring immunocompetent cells, or (2) that the keratinocytes are synthesizing the antigen themselves. Recently, gamma interferon has been shown to induce HLA-DR biosynthesis and expression on human malignant melanoma cells lines and on normal vascular endothelium. We report here that pure recombinant human gamma interferon (100 units/ml) induces HLA-DR expression on 60-70% of cultured human adult keratinocytes depleted of LC within 2-4 days of culture as determined by fluorescence-activated cell sorter (FACS) analysis using monoclonal antibodies. No residual LC or lymphocytes could be detected in these cultures. This is the first demonstration of HLA-DR expression by cultured human keratinocytes. This expression may be of functional significance in antigen presentation and cell-mediated cytotoxicity involving the epidermis.

We investigated the in vivo effect of recombinant interferon-gamma (IFN-gamma) and tumour necrosis factor alpha (TNF-alpha) treatment of mice on the development of the delayed-type hypersensitivity (DTH) reaction and lichenoid tissue reaction (LTR) following the local injection of cloned autoreactive T cells. Both the DTH reaction and the LTR were significantly enhanced by pretreatment with IFN-gamma, but not with TNF-alpha. Induction of class II MHC antigens on keratinocytes was not essential for the enhancement by IFN-gamma. Administration of anti-IFN-gamma antibody reduced the DTH reaction and LTR, although complete inhibition was not observed with our treatment regimen. The ability of IFN-gamma to increase the number of the cloned T cells invading the epidermis in vivo, is in keeping with our previous observation that IFN-gamma treatment of cultured keratinocytes markedly increased the adherence reaction between T cells and keratinocytes in vitro.

Normal cultured human epidermal melanocytes and melanoma cells derived from three different malignant melanomas were examined for synthesis of extracellular matrix components before and after treatment for one day with interferon-gamma, tumor necrosis factor-alpha, or both. Treatment of the cells with either cytokine individually had minimal effects on fibronectin levels. Treatment of the cells with the two agents in combination greatly stimulated fibronectin production as indicated by biosynthetic labeling and immunoprecipitation and by enzyme-linked immunosorbent assay. Synthesis of laminin was decreased slightly by the same treatment whereas thrombospondin production was stimulated slightly. The same treatment reduced melanocyte viability slightly but significantly inhibited proliferation and altered the morphology of the melanocytes. The treated cells became flattened and polygonal in shape while the untreated cells exhibited a bipolar shape with one or more long dendritic processes. These morphologic changes were not seen in cultures treated with interferon-gamma or tumor necrosis factor-alpha individually. The effects of interferon-gamma and tumor necrosis factor-alpha on fibronectin production by epidermal melanocytes are in contrast to the effects of the same treatments on fibronectin production by epidermal keratinocytes where fibronectin production is decreased but are similar to the effects of transforming growth factor-beta on a number of other cell types in which increased synthesis of fibronectin occurs and is associated with decreased growth.

The growth of cultured human keratinocytes (KC) is inhibited by gamma interferon (IFN-gamma), cyclosporin A and transforming growth factor-beta, but not by tumor necrosis factor. When these antiproliferative molecules were added to KC they induced a concentration and time-dependent inhibition of 125I-epidermal growth factor (I-EGF) binding. These anti-proliferative molecules primarily reduced the number of binding sites by approximately 25%-50% without affecting the binding affinity. Tumor necrosis factor did not influence the ligand binding by I-EGF. In parallel with the ability of the antiproliferative molecules to inhibit I-EGF binding, there was an increase in transforming growth factor-alpha production. These results suggest that several different antiproliferative molecules may share a common mechanism to inhibit cell growth by reducing I-EGF binding to KC.

In order to investigate the biologic significance of HLA-DR expression by human keratinocytes, we have examined the possibility that DR-positive keratinocytes become alloantigen presenting cells for resting T cells in the presence of interleukin-2. Using this system, gamma interferon-treated, DR-positive keratinocytes stimulate the proliferation of allogeneic, resting T cells approximately 3-fold whereas non-gamma interferon-treated, DR-negative keratinocytes do not. Because a monoclonal antibody against recombinant gamma interferon inhibits this proliferation, the stimulation is dependent on pre-incubation with gamma interferon. By contrast, since the stimulation is not inhibited by a monoclonal antibody against HLA-DR, it is not clear that the stimulation is due to class II antigen expression by keratinocytes. To rule out that gamma interferon increases the expression of class I antigens, leading to stimulation of resting T cells on that basis, we determined whether gamma interferon treatment enhances class I antigen expression by keratinocytes. The lymphokine treated cells did not demonstrate more class I antigen expression than untreated keratinocytes. Thus, the observed stimulation of allogeneic, resting T cells by gamma interferon-treated keratinocytes in the presence of IL-2 is not due to increased class I antigen expression but is due to other cell surface antigen(s) induced by recombinant gamma interferon treatment. These results suggest that gamma interferon-exposed keratinocytes in the presence of interleukin-2 may augment the activation of resting T lymphocytes and, in this manner, may contribute to cutaneous inflammation.

Recombinant gamma interferon induces class II antigen (HLA-DR) biosynthesis and expression on normal cultured human keratinocytes. HLA-DR expression was not induced on keratinocytes by recombinant alpha or beta interferons in a similar dose range nor by Con A or PHA. HLA-DR (L243) expression, as determined by FACS analysis, was detected as early as 1-2 days after addition of r-IFN-gamma to the cultures and was maximal after 4-8 days. Keratinocytes were analyzed for expression of another class II antigen, HLA-DC (Leu-10). Little or no expression of Leu-10 (DC) was detectable on these cells although Fc receptors for the IgG1 isotype were increased. These data indicate a unique role for gamma interferon in the differential regulation of keratinocyte class II antigen biosynthesis and expression. Induction of HLA-DR on keratinocytes may be functionally important in expanding the number of antigen presenting cells in the skin for the induction of an immune response and/or targeting these keratinocytes for cytolysis.

To extend our previous observation in which the binding of resting allogeneic peripheral blood mononuclear leukocytes (PBML) to recombinant gamma-interferon (IFN-gamma)-treated keratinocytes was characterized, we examined the influence of phorbol ester activation of the PBML to both autologous and allogeneic IFN-gamma-treated keratinocytes. The activation of PBML by phorbol esters (5 to 100 ng/ml) for brief periods of time (5 min to 1 h) at 37 degrees C led to an increase in the relative percentage of adherence to IFN-gamma-treated keratinocytes from 15% for non-activated PBML to 30% for phorbol ester-treated PBML. A biologically inert phorbol ester derivative did not enhance the binding reaction. No significant binding of phorbol ester-activated PBML was observed to non-IFN-gamma-treated keratinocytes. Both reduction in temperature to 4 degrees C and preincubation of the phorbol ester-treated PBML with anti-LFA-1 monoclonal antibody, led to complete inhibition of this adherence reaction indicating a role for the LFA-1 molecule in phorbol ester-activated PBML/IFN-gamma-treated keratinocyte reactions. Immunophenotypic analysis of the adherent cell population of the phorbol ester-activated PBML to the IFN-gamma-treated keratinocytes revealed that the predominant adherent cell type was the CD8+ T-cell subset (44%) versus the CD4+ T-cell subset (33%) with 23% monocytes and no binding of B lymphocytes. These results suggest that phorbol ester-activated PBML binds twice greater than resting PBML to IFN-gamma-treated keratinocytes, and this increased adherence may further contribute to homing of activated lymphocytes to the epidermis and mononuclear cell trafficking in the skin of inflammatory dermatoses.

Recombinant human gamma interferon (r-IFN-gamma) induces the synthesis and expression of HLA-DR antigen on cultured, normal, human keratinocytes depleted of Langerhans cells. After removal of r-IFN-gamma from the culture medium of keratinocytes that are expressing HLA-DR antigen, the cells continue to express this antigen for at least 2 days. r-IFN-gamma induces, in a dose dependent fashion, the synthesis of several triton-soluble proteins with the most prominent having an apparent molecular weight of 53,000. Whereas normal keratinocytes do not express HLA-DR antigen in vivo, they do express HLA-DR in a variety of skin diseases such as lichen planus, graft-versus-host disease, and mycosis fungoides. We propose that an understanding of lymphocyte-keratinocyte interactions in vivo may be achieved by further studies of the mechanism of action of r-IFN-gamma on cultured keratinocytes and that the results may provide insight into the patho-physiology leading to a number of common inflammatory and neoplastic skin diseases.

To examine whether keratinocytes (KC) may be a cellular target for the action of cyclosporin A (CsA) in psoriasis and other skin diseases, the direct antiproliferative effect of CsA was examined on rapidly proliferating KCs grown on plastic in serum-free media. CsA at concentrations of 1-10 micrograms/ml directly inhibited KC proliferation over an 8-day incubation period. CsH, an immunologically inert stereoisomer, also inhibited KC proliferation at 5-10 micrograms/ml. When CsA (1 microgram/ml) was combined with gamma-interferon (gamma-IFN; 100 units/ml), there was an additive cytostatic effect on KC proliferation. The potentiating properties of gamma-IFN (100 U/ml) were additive when combined with CsA (1 microgram/ml) and synergistic at a higher concentration of CsA (5 micrograms/ml). The cytostatic effect of CsA (5 micrograms/ml) on KC proliferation was reversible after removal of the drug. Exposure of KCs to CsA resulted in numerous lipid-laden cytoplasmic vacuoles, which began to appear after 24 hours of CsA treatment. The fact that previous reports have documented 1-3 micrograms/ml of CsA and 100 U/ml of gamma-IFN in skin lesions of psoriasis patients suggests that our in vitro results were obtained with the use of physiologically relevant concentrations of these molecules. The direct antiproliferative effects by CsA on KCs suggest that both KCs and lymphocytes may be cellular targets in vivo for the mechanism of action of CsA in psoriasis.

Rapidly proliferating keratinocytes (KCs) maintained in low calcium, serum-free medium produce and utilize thrombospondin (TSP) as an attachment and spreading factor. To begin to understand the modulation of KC TSP metabolism, gamma interferon (IFN-gamma), a product of activated T lymphocytes, was added to KC cultures. IFN-gamma was chosen because activated T cells appear at sites of cutaneous injury. Two additional cytokines including tumor necrosis factor (TNF) and IFN-beta were also examined. IFN-gamma (600 U/ml), but not TNF (500 U/ml) or IFN-beta (10(3) U/ml), as single agents decreased KC TSP biosynthesis, secretion, and utilization as an attachment factor. IFN-gamma alone did not detectably decrease TSP mRNA levels suggesting a post-transcriptional effect in KCs. However, the combination of IFN-gamma (600 U/ml) and TNF (500 U/ml) inhibited TSP mRNA production. These results demonstrate the modulation of KC TSP metabolism and biologic activity.

Twenty-three patients with chronic plaque-type psoriasis were treated with intramuscular administration of human recombinant interferon gamma. Patients were treated with doses of 0.01 to 0.25 mg/m2 daily (five out of seven days) for four weeks, or 0.25 mg/m2 three times weekly for one week with escalation to 0.5 mg/m2 for the subsequent seven weeks. Some patients treated with the 0.25-mg/m2 dose showed improvement coincident with their therapy. Although recombinant interferon gamma may have some therapeutic activity in certain patients' psoriasis, the magnitude of this effect is at best small. This result is in contrast to interferon alfa, which has been reported to cause an exacerbation of this disease. Staining of posttreatment biopsy specimens with a monoclonal antibody against HLA-DR antigen using an immunoperoxidase technique demonstrated HLA-DR expression by keratinocytes in some of the patients treated at the higher doses. No obvious correlation was seen between clinical improvement of the psoriasis and intensity or extent of HLA-DR antigen expression by keratinocytes in the skin biopsy specimens.

Gamma interferon (IFN-gamma), a product of activated lymphocytes, influences keratinocyte proliferation and differentiation. Recombinant gamma interferon (r-IFN-gamma) was radioiodinated using the Bolton-Hunter reagent and retained 90% of its biologic activity as determined by induction of HLA-DR expression. The biochemical ligand-binding properties of iodinated IFN-gamma to cultured human keratinocytes revealed a plateau of binding at 150 min at 4 degrees C, and a single class of specific high affinity receptors (kD = 1.3 X 10(-10) M; 2200 sites/cell). The binding of human IFN-gamma, to keratinocytes was inhibited by human r-IFN-gamma, but not by either murine recombinant gamma interferon or human recombinant beta interferon (r-IFN-beta). The presence of high affinity receptor sites on human keratinocytes assures the reception of appropriate immunologic signals in lymphocyte-keratinocyte interactions.

Dermatopathologists observe mononuclear leukocytes in close apposition to keratinocytes (KCs) in graft versus host disease and in other lymphocyte-mediated skin diseases, such as lichen planus, erythema multiforme, and lupus erythematosus. Since the KCs are Class II histocompatibility antigen (HLA-DR) positive in these diseases (indicating local production of gamma interferon, IFN-gamma, by activated T-cells), we sought to determine whether IFN-gamma treatment of KCs would influence the ability of allogeneic peripheral blood mononuclear leukocytes (PBMLs) to adhere to cultured KCs in vitro. The adherence of PBMLs to KC monolayers was determined by the three following methods: (a) methanol fixation of the washed KCs (after PBML incubation), followed by hematoxylin-eosin staining and direct counting of adherent PBMLs; (b) fluorescein isothiocyanate (FITC) labeling of PBML, followed by measuring the amount of FITC-PBML bound to KCs after washing either by direct visualization with a fluorescence microscope; or by (c) quantitative fluorescence spectroscopy following lysis of the adherent cells. While untreated KCs bound allogeneic PBMLs minimally 15-120 min at 37 degrees C, pretreatment of the KCs with IFN-gamma (300 U/ml, 3 days) produced significantly increased binding of the PBMLs by approximately fivefold. By contrast, IFN-alpha and IFN-beta (10(3) U/ml) had no effect. Also, despite the induction of HLA-DR on cultured human fibroblasts, no increased binding of PBMLs after IFN-gamma treatment was observed. The selective ability of IFN-gamma to produce a marked increase in adherence between KCs and PBMLs suggests a new role for IFN-gamma in the immunobiology of the skin.

Keratinocytes from involved psoriatic plaques (PP), uninvolved, clinically symptomless skin of psoriatic patients (PN) and normal healthy skin (NN) have been cultured in a low calcium serum free system for multiple passages. In this way, the keratinocytes were removed from microenvironmental factors present in the skin. While the basal rate of proliferation of the PP, PN and NN keratinocytes was not different, the PP cells produced more transforming growth factor-alpha (TGF-alpha) than NN cells, and the antiproliferative response of PP cells to gamma interferon (IFN-gamma), a product of activated T lymphocytes, was reduced. We studied IFN-gamma because it can inhibit the proliferation of NN keratinocytes, induce their differentiation and the appearance of two immunoregulatory cell surface molecules, HLA-DR and intercellular adherence molecule-I (ICAM-I), and because in another epithelial cell system, epidermal growth factor (EGF) modulates IFN-gamma activity. The mean antiproliferative effects of IFN-gamma at 50,200, and 500 U/ml for the PP group (n = 10) was less compared to the NN group (n = 11); P less than 0.001, while the PN group (n = 5) had a less dramatic, but statistically significant, reduction in growth inhibition by IFN-gamma only at 200 and 500 U/ml compared to NN cells; P less than 0.05 and P less than 0.01, respectively. The amount of TGF-alpha produced and secreted by PP keratinocytes from five different individuals was significantly greater than by NN keratinocyte cultures. In addition, IFN-gamma induced TGF-alpha to a lesser extent in PP keratinocytes compared to NN keratinocyte cultures. Keratinocytes isolated from atopic dermatitis and Sézary syndrome patients were similar to NN keratinocytes. In contrast to its differential effects on TGF-alpha production and proliferation, IFN-gamma induced similar amounts of HLA-DR and ICAM-I on PP, PN and NN keratinocytes. Thus, for the PP keratinocytes, there was a dissociation between the antiproliferative and immunomodulatory effects of IFN-gamma. These results support our previous hypothesis that the hyperproliferation and altered differentiation of keratinocytes in psoriatic plaques is linked to an altered responsiveness of the keratinocytes to IFN-gamma. Moreover, these results provide an in vitro correlate of our in vivo observation of increased TGF-alpha levels in psoriatic plaques. A new pathophysiological model to understand psoriasis is proposed which integrates these observations involving IFN-gamma and TGF-alpha. This experimental approach also provides a system to dissect biochemical pathways of pathophysiological importance for keratinocyte hyperproliferation in psoriasis.

To extend our initial observation that recombinant gamma-interferon induced expression of class II major histocompatibility (HLA-DR) antigen on normal cultured human keratinocytes, we studied the antiproliferative effects of recombinant alpha- and gamma-interferons. Both interferons reduced the number of attached cells (dose range 10-10(3) units/ml; 7.1 X 10(-11) to 7.1 X 10(-9) M) and gamma-interferon was 100 times more potent than alpha. This effect did not require the presence of Langerhans cells. gamma-Interferon reduced total cell production during the first week without increasing the percentage of cells shed. During the second week, gamma-interferon also increased the percentage of cell shedding. Although 10(2) units/ml of gamma-interferon was maximal for HLA-DR expression by cultured keratinocytes, there was increasing reduction of attached cells between 10(2) to 10(3) units/ml. The demonstration that recombinant gamma-interferon induces HLA-DR expression and also inhibits keratinocyte proliferation in vitro at nanomolar concentration expands the growing list of normal cells whose biologic function may be influenced by this lymphokine in vivo.

To extend our observation that recombinant gamma interferon (r-IFN-gamma) induces the synthesis and expression of HLA-DR antigen we have investigated 2 major areas including the modulation of r-IFN-gamma-induced HLA-DR expression and the possible immunologic consequences of keratinocyte HLA-DR expression in vitro. The induction of keratinocyte HLA-DR expression was greater for continuous compared with pulse dosage (0.5-24 h) of r-IFN-gamma and was markedly decreased after trypsinization of attached monolayers into single cell suspensions. The r-IFN-gamma caused induction of HLA-DR and this was not influenced by either pretreatment with irradiation, PGE2, or indomethacin. Both HLA-DR+ and HLA-DR- cultured keratinocytes induced RNA synthesis and gamma interferon production by allogeneic peripheral blood mononuclear leukocytes (PBMLs) indicating mononuclear cell activation. However, this activation was not followed by significant mitogenesis and only slightly increased levels of [3H]thymidine incorporation (maximal = 5800 cpm) by the PBMLs was observed. Cultured keratinocytes apparently inhibit both lectin-driven and mixed-lymphocyte reactions by producing a soluble mediator which is not dialyzable, or inhibited by pretreatment with indomethacin or anti-alpha, -beta, -gamma interferon antibodies. These results suggest that lymphocyte-keratinocyte reactions in vitro are complex and may be mediated by a variety of cytokines, lymphokines, and prostaglandins.

Two different human malignant squamous cell lines (SCL-1 and SW-1271) and normal human foreskin fibroblasts were treated with recombinant human alpha, beta, and gamma interferons. HLA-DR expression was induced in a concentration-dependent fashion only on the SCL-1 cells treated with recombinant human gamma interferon (r-IFN-gamma) (10(2)-10(3) U/ml). No HLA-DR expression was observed with alpha or beta interferon on either malignant squamous cell line, nor with gamma interferon on SW-1271 cells. All three interferons reduced the number of malignant cells growing in culture but had no effect on the fibroblasts. There was a concentration-dependent growth-inhibitory response of the malignant cells by the interferons (dose range 1-10(3) U/ml; 7.1 X 10(-12) M to 7.1 X 10(-9) M). The SCL-1 cells were 10(2) more sensitive (based on weight) to the antiproliferative effects of gamma interferon than alpha or beta interferon. A brief (30-min) exposure of the SCL-1 cells to r-IFN-gamma (10(2) and 10(3) U/ml) produced approximately the same inhibition of cell growth as continuous exposure over a 2-week period. The SW-1271 cells were equally sensitive to alpha, beta, and gamma interferons. However, the maximal inhibitory effect on SW-1271 cells was less than that observed for the SCL-1 cells. Combining beta and gamma interferon resulted in cytotoxicity with SCL-1 cells and additive cytostatic effect on the SW-1271 cells. These additional malignant cell lines with their different sensitivities to alpha, beta, and gamma interferons may prove useful in studying the mechanisms of action of various interferons.

Recombinant gamma interferon (r-IFN-gamma) increases the adherence of peripheral blood mononuclear leukocytes (PBMLs) to cultured keratinocytes and cutaneous microvascular endothelial cells (MECs). To determine which specific type of PBMLs bound to these r-IFN-gamma treated cells, we performed immunophenotyping on the adherent PBMLs. The adherent PBMLs were detached from the r-IFN-gamma treated keratinocytes and MECs by adding EDTA, and collected by cytocentrifugation, followed by immunocytochemical staining using a panel of monoclonal antibodies. Our results reveal that the relative adherent population of PBMLs was composed of approximately 60%-70% monocytes and 18%-24% Leu 2+ T lymphocytes (T-cytotoxic/suppressor) which preferentially bound to r-IFN-gamma treated keratinocytes and MECs. There was some lesser binding by Leu 3 + lymphocytes (T-helper/inducer); approximately 8%, and no binding of B lymphocytes. Since r-IFN-gamma also induced HLA-DR expression in keratinocytes and MECs, these in vitro data suggest that r-IFN-gamma may play an important role in the immunobiology of diverse skin diseases such as graft vs host disease, lichen planus, and other inflammatory dermatoses, because the keratinocytes express HLA-DR and the predominant T-cell subset in the epidermis is Leu 2 + (over the Leu 3 + T cell) in all of these conditions. These results represent a direct attempt to explain in situ immunophenotypic mononuclear leukocyte subset distribution patterns by using r-IFN-gamma and purified cultured cells such as keratinocytes and MECs. We propose that IFN-gamma, by both increasing the adherence of PBMLs, and promoting selective binding of monocytes and Leu 2 + T lymphocytes to both keratinocytes and MECs, may be important in regulating PBML localization and recirculation in the skin.

To extend our initial observations that recombinant gamma-interferon (r gamma IFN) influences the growth and differentiation of normal cultured human keratinocytes, we studied the electron microscopic changes induced by r gamma IFN. Treatment of cultured human keratinocytes with 10(3) units/ml (7.1 nM) of r gamma IFN produced a shift toward an increasing percentage of attached cells that had a mature-type, differentiated appearance rather than a basal type of proliferating cell, as observed in control, untreated cultures. This report extends the number of cell types that can be influenced by r gamma IFN at the ultrastructural level and supports the notion that r gamma IFN can influence both growth and maturation of normal cultured human keratinocytes.

Lymphocytes bind to cultured keratinocytes that are treated with interferon gamma (IFN-gamma) and tumor necrosis factor (TNF). When the lymphocytes are preincubated with antibody to lymphocyte function associated antigen-1 (LFA-1), this adherence is inhibited. Because intercellular adhesion molecule-1 (ICAM-1) is a ligand for LFA-1, we studied the cellular expression of ICAM-1, as well as two other IFN-gamma-inducible antigens, (HLA) human lymphocyte antigens DR and DQ, in both normal and diseased skin. The modulation of these cell surface antigens by IFN-gamma and TNF with the use of short-term organ cultures of skin was compared with isolated keratinocytes grown in a conventional tissue culture system. While in normal skin, keratinocytes did not express HLA-DR, DQ, or ICAM-1, when organ cultures were supplemented with IFN-gamma, rapid induction of keratinocyte ICAM-1 expression occurred after 24 hours; HLA-DR but not DQ expression occurred after 48 hours. TNF also induced keratinocyte ICAM-1 expression (although to a lesser degree than IFN-gamma) but did not induce either keratinocyte HLA-DR or DQ expression. There was good correlation of keratinocyte expression of ICAM-1 and HLA-DR by IFN-gamma and TNF when the epidermis of the organ culture system was compared with the isolated keratinocytes grown in tissue culture. The presence of intraepidermal lymphocytes correlated extremely well with keratinocyte ICAM-1 expression but not with keratinocyte HLA-DR expression in psoriasis, atopic dermatitis, lichen planus, and mycosis fungoides. The intensity of endothelial cell expression of ICAM-1 correlated with the degree of dermal inflammation. We conclude that IFN-gamma, once produced by activated T lymphocytes in the dermis, may be of importance in lymphocyte trafficking in the epidermis by the induction of keratinocyte ICAM-1 expression. The use of the short-term organ culture system, in which there is inducible ICAM-1 expression, provides an experimental bridge between purely in vitro and in vivo investigations to further our understanding of the molecular basis for lymphocyte apposition to keratinocytes in the skin.

Human epidermal keratinocytes incubated with recombinant gamma interferon (r-IFN-gamma) show, on two-dimensional gel electrophoresis, both the appearance of new proteins and the loss of others. Among [35S]methionine-labeled proteins, which are induced in an actinomycin- or alpha-amanitin-sensitive manner, is a prominent group with an apparent relative molecular mass of 53,000 and pI of 5.3-5.8. The synthesis of these proteins continues for at least 4 days in the presence of gamma interferon (IFN-gamma). Over the concentration range tested, up to 670 pM, there is no inhibition of protein synthesis, so the appearance of these proteins cannot be explained by overall inhibition of protein synthesis. Furthermore, at 4 pM we found only minor inhibition of DNA (21%) and RNA (29%) synthesis. Half-maximal induction of the prominent 53 kD proteins occurs at an interferon concentration of 0.8-3.5 pM which may be compared with a range of 1.5-30 pM for HLA-DR induction. The same prominent proteins are also induced by type I interferons. The 53 kD protein complex appears to consist of at least 4 different proteins, one of which is phosphorylated and another one of which is not induced in fibroblasts treated with IFN-gamma. We could obtain no evidence that the proteins were related by glycosylation. The presence of these proteins provides a sensitive means of identifying keratinocytes responding to interferons. Lack of these proteins in normal epidermis indicates that interferon does not play a major role in the control of keratinocyte behavior in sound skin.