Inhibition of ATP-binding cassette transporter downregulates interleukin-1beta-mediated autocrine activation of human dermal fibroblasts.

Fibroblasts constitute an important source of cytokines during inflammatory processes in the skin. Interleukin-1 is a potent, pleiotropic cytokine that is induced in activated human dermal fibroblasts. Interleukin-1 further induces many inflammatory mediators, including the chemokine interleukin-8. As fibroblasts express both interleukin-1 and the interleukin-1 receptor complex, the cellular response may be enhanced by autocrine activation. Interleukin-1alpha and interleukin-1beta lack a signal peptide and are translocated at the plasma membrane using an alternative secretory pathway, which may involve ATP-binding cassette transporter proteins. We hypothesize that inhibition of this pathway prevents secretion of interleukin-1, thereby downregulating interleukin-1-dependent autocrine induction of interleukin-8. We used the ATP-binding cassette 1 transporter inhibitor glybenclamide, which has been previously shown to block interleukin-1beta secretion in human monocytes. Using enzyme-linked immunosorbent assay, we assessed the effect of glybenclamide on interleukin-8 production in human dermal fibroblasts. In interleukin-1beta-transfected human dermal fibroblasts, interleukin-8 was induced through an autocrine activity of interleukin-1beta. Glybenclamide disabled this activation loop and significantly reduced interleukin-8. In human dermal fibroblasts that were stimulated with tumor necrosis factor alpha to reach high interleukin-1 expression levels, glybenclamide similarly suppressed interleukin-8. In contrast, glybenclamide did not affect interleukin-8 production in cells stimulated with interleukin-1 only. Glybenclamide did not affect caspase-1 in fibroblasts, which was expressed as an inactive precursor form, irrespective of treatments with tumor necrosis factor alpha and/or glybenclamide. Using overexpressing, interleukin-1-transfected COS-1 cells, inhibition of interleukin-1alpha and interleukin-1beta secretion was directly demonstrated on Western blots. These results are consistent with glybenclamide preventing externalization of interleukin-1 and subsequent autocrine induction of interleukin-8 in human dermal fibroblasts. Acting through such a mechanism, ATP-binding cassette transporter inhibitors may downregulate inflammation locally.     

Keratinocytes express the CD146 (Muc18/S-endo) antigen in tissue culture and during inflammatory skin diseases

The CD146 (or MUC18/MEL-CAM) antigen is a cell adhesion molecule of the immunoglobulin superfamily. Besides in melanoma, expression of CD146 antigen has been demonstrated in breast epithelia and hair follicles. We studied its expression by human keratinocytes in culture as well as in neoplastic and inflammatory skin diseases. Staining of primary cultured keratinocytes revealed expression of CD146 on the cell membrane, preferentially on cell-cell contact sites. Western blot analysis of keratinocytes detected a band of approximately 113 kDa, corresponding to the CD146 protein. In contrast to primary keratinocytes, neither CD146 protein nor mRNA expression was found in the keratinocyte-derived cell lines A431 and HaCaT. Treatment of keratinocytes with the proinflammatory cytokines interleukin-1 and interleukin-6, tumor necrosis factor-alpha, and interferon-gamma, resulted in no change of CD146 expression and incubation with phorbol 12-myristate 13-acetate led to a reduction of CD146 on keratinocytes. By contrast, when culturing keratinocytes in medium devoid of growth supplements, a distinct upregulation was observed as compared with culture in fully supplemented medium. In normal human epidermis expression of the CD146 antigen was not detectable. It was strongly upregulated, however, on suprabasal keratinocytes in psoriasis, in lichen planus, in the epidermis overlying skin neoplasms, and in viral warts. In squamous cell carcinomas and basal cell carcinomas only a minority of tumor cells expressed CD146. Our findings suggest that the CD146 antigen represents an activation marker of keratinocytes and may be involved in cutaneous inflammatory tissue reaction.     

Counterregulation of interleukin-18 mRNA and protein expression during cutaneous wound repair in mice.

Recent work has suggested interleukin-18 to represent a proinflammatory cytokine that contributes to systemic and local inflammation. As the process of cutaneous wound healing crucially involves an inflammatory phase of repair, we investigated the regulation of interleukin-18 during the repair process. In non-wounded skin we observed high levels of interleukin-18 mRNA, whereas corresponding interleukin-18 protein was expressed only at low basal levels. Upon injury, we found a rapid and large induction of interleukin-18 protein expression, which is directly correlated with decreasing mRNA levels within the wound. Immunohistochemical analysis revealed different sites of expression in the wounded area, with keratinocytes as one major source of interleukin-18 production. The counterregulation of interleukin-18 mRNA and protein expression during wound repair in vivo might represent a general mechanism for interleukin-18 expressional regulation, as cytokine-stimulated keratinocytes exhibit a similar downregulation of interleukin-18 mRNA that is directly associated with increasing interleukin-18 protein levels in vitro. The rapid induction of interleukin-18 during wound healing suggests a role for interleukin-18 within the early phase of repair rather than a role in costimulation of interferon-gamma release from T cells, which are present in high numbers within the wounded area only during the late inflammatory phase of repair.     

Human skin: source of and target organ for angiotensin II

The present study examined the expression of angiotensin receptors in human skin, the potential synthesis of angiotensin II (Ang II) in this location and looked for a first insight into physiological functions. AT1 and AT2 receptors were found within the epidermis and in dermal vessel walls. The same expression pattern was found for angiotensinogen, renin and angiotensin-converting enzyme (ACE). All components could additionally be demonstrated at mRNA level in cultured primary keratinocytes, melanocytes, dermal fibroblasts and dermal microvascular endothelial cells, except for AT2 receptors in melanocytes. The ability of cutaneous cells to synthesize Ang II was proved by identifying the molecule in cultured keratinocytes. Furthermore, in artificially wounded keratinocyte monolayers, ACE-mRNA expression was rapidly increased, and enhanced ACE expression was still found in cutaneous human scars 3 months after wounding. These findings suggest that the complete renin-angiotensin system is present in human skin and plays a role in normal cutaneous homeostasis as well as in human cutaneous wound healing.     

Mast cell-fibroblast interactions: human mast cells as source and inducers of fibroblast and epithelial growth factors

As mast cells have been implicated in cutaneous repair processes, we have examined the ability of human mast cells to produce important epithelial and fibroblast growth factors or to stimulate the production of such factors in dermal fibroblasts. Isolated, highly purified human dermal mast cells and human leukemic mast cells were examined for mRNA and partly also for protein expression of these molecules as such or after preincubation with interleukin-4, stem cell factor, or with phorbol myristate acetate. In addition, mast cells were studied for their ability to induce fibroblast growth factor 2 and fibroblast growth factor 7 secretion from dermal fibroblasts. Both dermal and leukemic mast cells expressed fibroblast growth factor 2, fibroblast growth factor 7, and heparin-binding epidermal growth factor, but not hepatocyte growth factor at mRNA level, and dermal mast cells expressed fibroblast growth factor 10 in addition. At protein level, spontaneous fibroblast growth factor 2 secretion was noted that was markedly enhanced by phorbol myristate acetate, whereas no fibroblast growth factor 7 protein was detected under these conditions. Instead, human mast cell-1 supernatants induced enhanced fibroblast growth factor 7 secretion from dermal fibroblasts, with phorbol-myristate-acetate-stimulated supernatants being more effective. This effect could be reproduced with histamine and was H1-receptor mediated. Tryptase was ineffective but stimulated instead fibroblast growth factor 2 secretion from fibroblasts. These data demonstrate for the first time the ability of mast cells to express and/or secrete several growth factors of the fibroblast growth factor family as well as heparin-binding epidermal growth factor directly or indirectly via stimulation of fibroblasts, underlining the potentially pivotal role of these cells during human tissue repair and homeostasis.     

Detection of melanocortin-1 receptor antigenicity on human skin cells in culture and in situ

The proopiomelanocortin (POMC) products alpha-melanocyte stimulating hormone (alpha-MSH) and adrenocorticotropin (ACTH) bind to specific receptors known as the melanocortin (MC) receptors. There is increasing evidence that the MC receptor subtype 1 (MC-1R) is expressed in vitro by several other cutaneous cell types besides melanocytes and keratinocytes. Our knowledge on the MC-1R expression in skin, however, remains fragmentary. In order to examine the expression of MC-1R in human skin cells in vitro and In situ, we made use of a recently described antibody directed against the amino acids 2-18 of the human MC-1R. Flow cytometry analysis revealed the highest MC-1R antigenicity in normal melanocytes and keratinocytes, followed by dermal fibroblasts, microvascular endothelial cells and WM35 melanoma cells. Little or no expression was detected in KB carcinoma cells and Fs4 fibroblasts. In normal human skin, immunoreactivity for the anti-MC-1R antibody was detected in hair follicle epithelia, sebocytes, secretory and ductal epithelia of sweat glands, and periadnexal mesenchymal cells. Interfollicular epidermis was largely unreactive in adult skin as opposed to undifferentiated keratinocytes of fetal skin. Our findings form a framework within which MC-1 receptor expression can be studied in various skin diseases.     

Interleukin-17 is produced by both Th1 and Th2 lymphocytes, and modulates interferon-gamma- and interleukin-4-induced activation of human keratinocytes

Interleukin-17 is a T-cell-derived cytokine, detected in skin affected by allergic contact dermatitis and psoriasis, which regulates keratinocyte expression of adhesion molecules and chemokines. In this study, we have analyzed whether interleukin-17 production segregates with a particular T helper (Th) cell subset, and have examined the capacity of interleukin-17 to modulate the activation of keratinocytes induced by Th1 and Th2 cytokines. A panel of 80 nickel-specific CD4+ T cell clones (36 Th0, 30 Th1, and 14 Th2) was isolated from peripheral blood or lesional skin of allergic contact dermatitis patients. Significant amounts (> 50 pg per ml) of interleukin-17 were released by about 50% of activated Th0, Th1, and Th2 cells. Interleukin-17 alone and in cooperation with interleukin-4, or to a lesser extent with interferon-gamma, decreased the interleukin-1 receptor antagonist to interleukin-1alpha ratio in the supernatants as well as in cell lysates from keratinocytes. In addition, interleukin-17 stimulated the release of growth-regulated oncogene-alpha, granulocyte-macrophage colony stimulating factor, and interleukin-6, with synergistic or additive effects when used together with interferon-gamma or interleukin-4. Interleukin-17 and interleukin-4 also increased stem cell factor release, a function that was inhibited by interferon-gamma. Moreover, interleukin-17 and interleukin-4 enhanced interferon-gamma-induced expression of intercellular adhesion molecule 1, but not CD40, on keratinocytes. The constitutive expression of interleukin-17 and interferon-gamma receptors on keratinocytes was not modulated by interleukin-17, interferon-gamma, or interleukin-4, whereas the interleukin-4 receptor was significantly downregulated by interferon-gamma. As a whole, the results indicate that interleukin-17 can participate relevantly in T-cell-mediated skin immune responses by amplifying both interferon-gamma- and interleukin-4-induced activation of keratinocytes.     

Upregulation of interleukin-18 expression in mouse primary keratinocytes induced to differentiate by calcium

Interleukin-18 (IL-18) is an immunomodulatory cytokine that stimulates interferon-gamma production by T helper cells. Recently, basal keratinocytes have been shown to constitutively express IL-18, and IL-18 expression increases in the suprabasal keratinocytes in psoriatic lesions. In the study reported here we showed that in mouse epidermis, IL-18 immunoreactivity was markedly increased in the granular and cornified layers. To further investigate whether differentiated keratinocytes synthesize more IL-18, we examined the expression of mouse IL-18 in primary mouse keratinocytes induced to differentiate by calcium, an in vitro cell culture system mimicking keratinocyte differentiation in the epidermis. We demonstrated that IL-18 mRNA and protein in cultured keratinocytes were increased by calcium treatment in a time- and dose-dependent manner. The upregulation of IL-18 was associated with an increase in keratinocyte differentiation markers, and was dependent on the synthesis of new RNAs and proteins. However, the IL-18 protein in the cytoplasm was predominantly in the precursor form, and no increase in IL-18 activity was detected in the culture medium treated with calcium. Furthermore, blocking the calcium-induced keratinocyte differentiation with protein kinase C inhibitor inhibited the upregulation of IL-18 expression. These findings suggest that IL-18 is synthesized in keratinocytes mainly in the inactive precursor form, and its expression is upregulated as basal keratinocytes differentiate in the epidermis.     

Gene transfer of secreted-type modified interleukin-18 gene to B16F10 melanoma cells suppresses in vivo tumor growth through inhibition of tumor vessel formation

Interleukin-18 is a novel cytokine identified as a strong inducer of interferon-gamma. Interleukin-18 has been shown to have similar bioactivities to interleukin-12 and to have antitumor efficacy in experimental models. In this study, we investigated whether the introduction of the interleukin-18 gene to B16F10 melanoma cells can induce antitumor response or not. Before the transfection, we modified the interleukin-18 gene to enable transfected tumor cells to secrete bioactive interleukin-18, because interleukin-18 does not have a signal sequence and requires processing by the interleukin-1 converting enzyme to attain the mature form. We found that B16 melanoma cells transduced with hybrid cDNA consisting of the interferon-beta signal sequence and mature interleukin-18 sequence, but not native interleukin-18, secreted a large amount of interleukin-18 and exhibited retarded tumor growth when injected in syngeneic mice. The antitumor effect was mostly abrogated by administration of anti-interferon-gamma antibody, but was not affected by in vivo depletion of CD8+ T cells or natural killer cells. Histologic analysis revealed that vascularization was markedly reduced and that necrosis was extensively induced in interleukin-18-secreting B16F10 melanoma (B16/IL18) tissues, whereas abundant tumor vessel formation was observed in B16/IL18 tissues of interferon-gamma-neutralized mice. We also found that chemokines, interferon-inducible protein-10 and monokine induced by interferon-gamma, were produced in B16/IL18 tissues and that the expression of both chemokines was dependent on that of interferon-gamma in the tumor tissues. Further, we showed that B16 melanoma cells secreted both chemokines in response to interferon-gamma. In addition, the expression of angiogenin, an angiogenic factor of melanoma, in B16 melanoma cells was reduced by interferon-gamma treatment. These results indicate that gene transfer of secreted-type interleukin-18 to B16F10 melanoma cells is a useful method of triggering an antitumor response without any systemic adverse effects and that the antitumor efficacy is mainly mediated by antiangiogenic activity, which is possibly involved in at least two dynamic changes induced by interferon-gamma inside B16 melanoma cells: the upregulation of antiangiogenic chemokines, interferon-inducible protein-10 and monokine induced by interferon-gamma, and the downregulation of angiogenic factor, angiogenin.     

Keratinocytes inhibit expression of connective tissue growth factor in fibroblasts in vitro by an interleukin-1alpha-dependent mechanism

The wound healing process concludes with downregulation of fibroblast activity. Clinical observations suggest that the regenerating epidermis suppresses this activity. An important regulator of fibroblast activity is the fibrogenic cytokine connective tissue growth factor. We hypothesized that epidermal keratinocytes may affect fibroblast activity via this cytokine. We demonstrate keratinocyte-mediated suppression of connective tissue growth factor at both the mRNA and protein levels by around 50% or more when fibroblasts were cultured in multiwell plates with keratinocyte cultures in accompanying semipermeable cell culture inserts, or stimulated by keratinocyte-conditioned media. Both basal and transforming-growth-factor-beta1-stimulated levels of connective tissue growth factor were inhibited. A 3 h coculture period with keratinocytes was sufficient to suppress connective tissue growth factor expression by fibroblasts, but the inhibition developed over a time period of around 16 h. The putative keratinocyte-derived factor(s) responsible for these effects was found to be soluble and stable. By analyzing cytokines secreted by keratinocytes we identified interleukin-1alpha as a potent inhibitor of connective tissue growth factor mRNA expression in fibroblasts. Involvement of this cytokine in keratinocyte-mediated connective tissue growth factor suppression was confirmed by using anti-interleukin-1alpha antibodies. Tumor necrosis factor alpha or prostaglandins did not appear to be involved. In conclusion, our results indicate that interleukin-1alpha secretion by keratinocytes provides a mechanism for the downregulation of connective tissue activity during the end-stage of wound healing, when epithelia coverage has developed over the wound area.     

Human dermal fibroblast interleukin-1 receptor antagonist (IL-1ra) and interleukin-1 beta (IL-1 beta) mRNA and protein are co-stimulated by phorbol ester: implication for a homeostatic mechanism.

Although the major functions of fibroblasts are to produce extracellular matrix and to maintain a structural framework for organ systems, recent studies have demonstrated that fibroblasts are active participants in inflammatory processes by synthesizing various inflammatory mediators. In this report, we provide evidence that fibroblasts may contribute to the regulation of inflammation by the synthesis of both the intracellular form and the secretory form of interleukin-1 receptor antagonists in conjunction with interleukin-1 beta production. Indirect immunofluorescence microscopy localized interleukin-1 receptor antagonist and interleukin-1 beta proteins primarily in the fibroblast cytoplasm. Polymerase chain reaction amplification of reverse-transcribed mRNA with primers specific for the intracellular form of interleukin-1 receptor antagonist detected cDNA fragments present in both unstimulated and phorbol ester-stimulated fibroblasts, identical in molecular size to that in unstimulated keratinocytes. Amplification with primers specific for the secretory form of interleukin-1 receptor antagonist, however, detected cDNA fragments in phorbol ester-stimulated fibroblasts and phytohemagglutinin-stimulated peripheral mononuclear cells, but not in unstimulated fibroblasts or keratinocytes. The amplified fibroblast cDNA sequences for both intracellular and secretory interleukin-1 receptor antagonists were confirmed by digestion with three restriction endonucleases. By ethidium bromide visualization of amplified cDNA derived from serially diluted total cellular RNA and by Southern blot hybridization analysis of amplified cDNA, we have demonstrated that fibroblast interleukin-1 receptor antagonist mRNA and interleukin-1 beta mRNA were co-stimulated by phorbol ester. Similarly, ELISA demonstrated that fibroblast cytoplasmic interleukin-1 receptor antagonist protein and interleukin-1 beta protein were co-stimulated by phorbol ester. Our data suggests that the intracellular form of interleukin-1 receptor antagonist may be important in maintaining physiologic homeostasis in fibroblasts during interleukin-1 beta induction and release.     

Suppression of vitamin D receptor and induction of retinoid X receptor alpha expression during squamous differentiation of cultured keratinocytes

To gain more insight in the role of the vitamin D system in epidermal differentiation, we studied the expression of the vitamin D receptor and its heterodimeric partner retinoid X receptor alpha in cultured normal human keratinocytes during squamous differentiation, as triggered by different approaches. Northern and western blot analysis allowed us to investigate mRNA and protein levels of these nuclear receptors and of markers for growth control (c-myc, cyclin D1, p21WAF1) and differentiation (keratinocyte transglutaminase, small proline rich proteins). Growing cells to postconfluence was a potent stimulus for growth arrest and differentiation with concomitant suppression of vitamin D receptor and induction of retinoid X receptor alpha, at both the mRNA and the protein level. These changes could be prevented by concomitant treatment with epidermal growth factor or keratinocyte growth factor. Subjecting the cells to a calcium switch leading to stratification and differentiation lowered vitamin D receptor protein levels without affecting vitamin D receptor mRNA and induced both retinoid X receptor alpha mRNA and protein. Interferon-gamma and the phorbolester 12-O-tetradecanoyl phorbol 13-acetate, two well-known inducers of keratinocyte differentiation, both inhibited vitamin D receptor expression but only interferon-gamma induced retinoid X receptor alpha. The decreased vitamin D receptor expression was accompanied by reduced vitamin D responsiveness (as assessed by 24-hydroxylase mRNA induction) in postconfluent, high calcium, and 12-O-tetradecanoyl phorbol 13-acetate treated keratinocytes but not with interferon-gamma treatment. Taken together, our results associate vitamin D receptor expression with undifferentiated, proliferating keratinocytes, whereas retinoid X receptor alpha expression appears to be related to the differentiated phenotype. Therefore, proliferating and differentiating keratinocytes may be differentially targeted by active vitamin D metabolites.     

Dermcidin is constitutively produced by eccrine sweat glands and is not induced in epidermal cells under inflammatory skin conditions

BACKGROUND: Antimicrobial peptides (AMPs) are important effector molecules of innate immunity, protecting epithelial surfaces of multicellular organisms. In human skin two classes of AMPs-the beta-defensins and the cathelicidins-are produced by keratinocytes primarily under inflammatory conditions. In contrast, dermcidin (DCD), a recently discovered AMP with broad-spectrum activity, is expressed in eccrine sweat glands and transported via sweat to the epidermal surface. OBJECTIVES: To investigate whether DCD expression is induced under inflammatory conditions in epidermal keratinocytes. METHODS: Lesional skin of the inflammatory skin diseases atopic dermatitis, psoriasis and lichen planus was analysed by immunohistochemistry using a polyclonal anti-DCD antiserum. We also examined whether DCD RNA expression is induced in cultured human keratinocytes, fibroblasts, melanocytes and melanoma cells. RESULTS: Whereas DCD was constitutively expressed in eccrine sweat glands of all skin biopsies, we found that, independent of the type of the inflammatory skin lesion, DCD protein expression was not induced in human epidermal keratinocytes. In contrast, beta-defensin 2 was expressed in epidermal keratinocytes of inflammatory human skin, but not in keratinocytes of healthy human skin. Upon stimulation of the cultured cells with 12-O-tetradecanoyl-phorbol-13-acetate, tumour necrosis factor-alpha, lipopolysaccharide or H2O2, DCD mRNA expression was not detected in primary keratinocytes, fibroblasts and melanocytes, but was detected in MeWo and SKMEL28 melanoma cells. CONCLUSIONS: These results indicate that, unlike human cathelicidins and beta-defensins which are inducible peptides that primarily function in response to injury and inflammation, DCD is exclusively part of the constitutive innate defence of human skin. By modulating surface colonization, DCD may help to prevent local and systemic invasion of pathogens.     

Growth and phenotypic characteristics of human nevus cells in culture

Nevus cells were isolated from the three cutaneous components, epidermis, basal layer, and dermis, of nonmalignant pigmented lesions and were cultured separately in the presence or absence of the phorbol ester 12-0-tetradecanoyl phorbol-13-acetate in medium that supports the rapid proliferation of melanocytic cells. The separation procedure used provided cultures that were essentially free from normal melanocytes (dermis) or fibroblasts (epidermis). In short term culture, nevus cells of all skin compartments expressed markers associated with differentiated melanocytes, such as presence of premelanosomes and melanosomes and elevated tyrosinase levels. Nevus cells also expressed melanoma-associated antigens, such as NGF-receptor, transferrin-related p97, proteoglycan, and HLA-DR as detected with monoclonal antibodies. After several subpassages, cells showed a decreased expression of melanoma-associated antigens, decreased tyrrosinase levels, and melanosomes could no longer be detected. Morphologically, these cells were similar to fibroblasts. The disappearance of melanoma-associated cell surface antigens was concomitant with the appearance of a melanocyte-associated 145 kd protein that might serve as a marker of fibroblast-like differentiation in nevus cells and normal melanocytes. Nevus cell cultures grown in the presence of 12-0-tetradecanoyl phorbol-13-acetate maintained a stable differentiated phenotype throughout their lifespan. As reported earlier, nevus cells in culture, irrespective of the presence or absence of 12-0-tetradecanoyl phorbol-13-acetate, have a finite lifespan in vitro, grow anchorage-independent in soft agar, but do not form tumors when xenografted to nude mice. These studies demonstrate that nevus cells isolated from the epidermal, basal layer, and dermal components of lesional skin can serve as models to characterize the initial steps of tumor progression in a human cell system.     

Cytokine release and cytotoxicity in human keratinocytes and fibroblasts induced by phenols and sodium dodecyl sulfate

Phenolic compounds used in pharmaceutical and industrial products can cause irritant contact dermatitis. We studied the effects of resorcinol, phenol, 3,5-xylenol, chloroxylenol, and 4-hexyl-resorcinol on normal human epidermal keratinocytes and dermal fibroblasts for cytotoxicity and cytokine release, determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide methodology and enzyme-linked immunosorbent assay, respectively. An inverse correlation between phenol concentrations causing a 50% reduction in keratinocyte and fibroblast viability at 24 h and their octanol water-partition coefficients (i.e., hydrophobicity) was observed. 3,5-xylenol, chloroxylenol, hexyl-resorcinol, and sodium dodecyl sulfate, but not resorcinol or phenol, induced release of interleukin-1alpha from keratinocytes at cytotoxic concentrations. Variable release of tumor necrosis factor-alpha and interleukin-8 from keratinocytes occurred only at toxic threshold concentrations of the phenols or sodium dodecyl sulfate. Subtoxic concentrations of phenols or sodium dodecyl sulfate did not induce cytokine release from keratinocytes. Neither the phenols nor sodium dodecyl sulfate induced release of the chemokines interleukin-8, growth-related oncogene-alpha or monocyte chemotactic protein-1 from fibroblasts. Conditioned media from keratinocytes treated with cytotoxic concentrations of 3,5-xylenol, chloroxylenol, hexyl-resorcinol, or sodium dodecyl sulfate stimulated further release of the chemokines from fibroblasts above that obtained with control media. Rabbit anti-interleukin-1alpha serum inhibited keratinocyte-conditioned media induction of chemokine release. We have shown a structure-cytotoxicity relationship for a series of phenols as well as an association of interleukin-1alpha release with a cytotoxic effect. We demonstrated a cytokine cascade amplification step by the actions of stimulated keratinocyte media on cultured dermal fibroblasts, identifying interleukin-1alpha as the principal initiator of chemokine synthesis.