and activation-regulated chemokine (LARC)/macrophage ...

6 downloads 0 Views 377KB Size Report
Oct 5, 2000 - and Osamu Yoshie. Departments of Bacteriology, 1Dermatology and 4Surgery, Kinki University School of Medicine,. Osaka-Sayama 589-8511, ...
International Immunology, Vol. 13, No. 1, pp. 95–103

© 2001 The Japanese Society for Immunology

Inducible expression of a CC chemokine liver- and activation-regulated chemokine (LARC)/macrophage inflammatory protein (MIP)-3α/CCL20 by epidermal keratinocytes and its role in atopic dermatitis Takashi Nakayama, Ryuichi Fujisawa, Hidekazu Yamada1, Tatsuya Horikawa2, Hiroshi Kawasaki3, Kunio Hieshima, Dai Izawa, Satoru Fujiie4, Tadashi Tezuka1 and Osamu Yoshie Departments of Bacteriology, 1Dermatology and 4Surgery, Kinki University School of Medicine, Osaka-Sayama 589-8511, Japan 2Department of Dermatology, Kobe University School of Medicine, Kobe 650-0011, Japan 3Institute of Medical Science, University of Tokyo, Tokyo 108-8639, Japan Keywords: chemokine, chemokine receptor, dendritic cells, IL-1, Langerhans cells, skin, tumor necrosis factor

Abstract Liver- and activation-regulated chemokine (LARC)/macrophage inflammatory protein (MIP)-3α/ CCL20 is a CC chemokine which is constitutively expressed by follicle-associated epithelial cells in the mucosa, and attracts cells expressing CCR6 such as immature dendritic cells and α4β7high intestine-seeking memory T cells. Here, we examine LARC/CCL20 expression in the skin. LARC/CCL20 mRNA and protein were induced in primary human keratinocytes upon stimulation with proinflammatory cytokines such as IL-1α and tumor necrosis factor (TNF)-α. In mice, intradermal injection of IL-1α and TNF-α rapidly induced a local accumulation of transcripts for LARC/CCL20 and its receptor CCR6 with a lag of several hours in the latter. In humans, immunostaining of LARC/CCL20 was weak if any in normal skin tissues but strongly augmented in lesional skin tissues with atopic dermatitis. Furthermore, massive infiltration of cells with markers such as CD1a, CD3 or HLA-DR was present in atopic skin lesions. Many infiltrating cells were also found to be CCR6⍣ by a newly generated monoclonal anti-CCR6. However, Langerhans cells residing within the epidermis were hardly stained by anti-CCR6 in normal and atopic skin tissues. Furthermore, plasma levels of LARC/CCL20 were found to be elevated in patients with atopic dermatitis. Collectively, our results suggest that epidermal keratinocytes produce LARC/CCL20 upon stimulation with proinflammatory cytokines such as IL-1α and TNF-α, and attract CCR6expressing immature dendritic cells and memory/effector T cells into the dermis of inflamed skin such as atopic dermatitis. LARC/CCL20 may not, however, play a major role in homeostatic migration of Langerhans cells into the skin. Introduction Chemokines are a group of small (8–14 kDa), structurally related, mostly basic, heparin-binding cytokines that play pivotal roles in the host defense mechanisms by recruiting various types of leukocytes (1). In humans, over 40 members have been identified. Based on the arrangement of

the N-terminal conserved cysteine residues, chemokines are further divided into four subfamilies, i.e. CC, CXC, C and CX3C. The biological effects of chemokines are mediated by a group of seven transmembrane G protein-coupled receptors. At present, 17 chemokine receptors have been

Correspondence to: O. Yoshie Transmitting editor: M. Miyasaka

Received 23 June 2000, accepted 5 October 2000

96 Production of LARC/CCL20 by epidermal keratinocytes identified (2,3). Currently, it is useful to categorize chemokines into two major functional groups, i.e. inflammatory and immune/homeostatic chemokines. Inflammatory chemokines mainly recruit neutrophils, monocytes and eosinophils. These chemokines play major roles in acute and chronic inflammatory responses. On the other hand, immune/homeostatic chemokines mainly attract lymphocytes and dendritic cells. These chemokines are likely to be involved in genesis, homeostasis and function of the immune system (2,3). Liver- and activation-regulated chemokine (LARC) (4), also designated as macrophage inflammatory protein (MIP)-3α (5) or Exodus (6) and recently listed as CCL20 in the systematic classification of chemokine ligands (2), is a CC chemokine which was originally shown to be chemotactic for lymphocytes but not for monocytes (4). By Northern blot analysis using commercially available multi-tissue blot filters, LARC/CCL20 was found to be expressed in tissues such as liver and lung (4,6). In normal mice, however, constitutive expression of LARC/CCL20 was seen primarily in intestine and colon (7). Furthermore, follicle-associated epithelial (FAE) cells were found to express LARC/CCL20 transcripts in both mice and humans (7,8). In mice, injection of lipopolysaccharide (LPS) transiently up-regulates LARC/CCL20 mRNA in FAE and other intestinal epithelial cells (7). We identified an orphan receptor GPR-CY4 as a specific functional receptor for LARC/CCL20 and termed it CCR6 following the set rule of chemokine receptor nomenclature (9). Separately, the same receptor was cloned from dendritic cells derived from lung tissues or those differentiated in vitro from cord blood CD34⫹ precursor cells in the presence of granulocyte macrophage colony stimulating factor (GM-CSF) and tumor necrosis factor (TNF)-α (10,11). Even though dendritic cells differentiated from peripheral blood monocytes by GM-CSF and IL-4 failed to express CCR6, those cultured in the presence of GM-CSF, IL-4 and TGF-β1 were found to express CCR6 (12). It is now accepted that immature dendritic cells express CCR6 and respond to LARC/CCL20 in chemotaxis (8,13). Upon antigen loading or stimulation with proinflammatory cytokines, however, dendritic cells start maturation, and rapidly down-regulate CCR6 and up-regulate CCR7 (8,13). This probably enables maturing dendritic cells to be guided into draining lymph nodes by secondary lymphoid tissue chemokine (CCL21), a ligand for CCR7 (8,13). Beside immature dendritic cells, CCR6 was also shown to be expressed on B cells, most α4β7high intestineseeking memory T cells and a fraction of cutaneous lymphocyte-associated antigen (CLA)⫹ skin-seeking memory T cells (14). Collectively, it is conceivable that intestinal epithelial cells, especially FAE cells covering Peyer’s patches, produce LARC/CCL20 to attract CCR6-expressing immature dendritic cells and α4β7high memory T cells toward the mucosal surface. Besides the mucosal tissues, LARC/CCL20 may have a similar cell-recruitment role in the skin. Recently, Charbonnier et al. reported that the epidermal layer of clinically normal human skin was immunologically reactive for LARC/CCL20 while skin-residing Langerhans cells expressed CCR6 transcripts (15). The authors speculated that epidermal keratinocytes constitutively produce LARC/CCL20 and Langerhans cells expressing CCR6 homeostatically migrate into the epidermis by the guidance of LARC/CCL20 (15). Furthermore, Yang et al. reported that β-defensins, a group of small

(3.5–4.5 kDa) cationic antimicrobial peptides produced by epidermal keratinocytes upon proinflammatory stimulations, were capable of attracting immature dendritic cells and memory T cells through CCR6 (16). In the present study, we have demonstrated that human primary epidermal keratinocytes in culture are induced to express LARC/CCL20 mRNA and protein upon stimulation with proinflammatory cytokines such as IL-1α and TNF-α. Induction of LARC/CCL20 by these cytokines was also demonstrated in the mouse skin in vivo. In humans, we detected strong immunological staining of LARC/CCL20 in epidermal cells from skin lesions of atopic dermatitis patients. Furthermore, by using a newly generated mAb against CCR6, we could detect massive infiltration of CCR6-expressing cells in the dermis of one atopic patient. However, we hardly detected staining of epidermal Langerhans cells by anti-CCR6 in normal or atopic skin tissues. Finally, we detected elevated plasma levels of LARC/CCL20 in patients with atopic dermatitis. Thus, our results suggest that LARC/CCL20 expression in the epidermis is more related to proinflammatory responses of the skin than to constitutive migration of Langerhans cells into the epidermal layer. Methods Primary human keratinocyte culture Human neonatal dermal keratinocytes (n ⫽ 3) and keratinocyte basal medium with growth supplements were purchased from Clonetics (Walkersville, MD). Keratinocytes were seeded into six-well-type tissue culture plates (Falcon) and fed with fresh medium every 2 days until confluence. After aspiration of culture medium, cells were supplied with 3 ml of fresh culture medium and treated for 1–3 days without or with LPS from Escherichia coli (Sigma, St Louis, MO), IL-1α (PeproTech, Rocky Hill, NJ), TNF-α (PeproTech), IFN-γ (PeproTech) and IL-4 (PeproTech), either singly or in combination. After that, culture supernatants were examined with ELISA and cell pellets were lysed for RNA preparation. In some experiments, dexamethasone (Sigma) or FK506 (Fujisawa Pharmaceutical, Osaka, Japan) was also included in culture medium during treatment with cytokines. RT-PCR analysis Total RNA was prepared from cultured keratinocytes or skin biopsies by using Trizol reagent (Gibco/BRL, Gaithersburg, MD). RNA was further purified by using RNeasy (Qiagen, Hilden, Germany). Total RNA (1 µg) was reverse transcribed using oligo(dT)18 primer and SuperScript II reverse transcriptase (Gibco/BRL). Resulting first-strand DNA (20 ng total RNA equivalent) and original total RNA (20 ng) were amplified in a final volume of 20 µl containing 10 pmol of each primer and 1 U of Ex-Taq polymerase (TaKaRa Shuzo, Kyoto, Japan). The primers used were: ⫹5⬘-TACTCCACCTCTGCGGCGAATCAGAA-3⬘ and –5⬘-GTGAAACCTCCAACCCCAGCAAGGTT-3⬘ for human LARC/CCL20; ⫹5⬘-CCTGGGGGAATATTCTGGTGGTGA-3⬘ and –5⬘-CATCGCTGCCTTGGGTGTTGTAT-3⬘ for human CCR6; ⫹5⬘-TGCTCTTCCTTGCTTTGGCATGGGTA-3⬘ and –5⬘-TCTGTGCAGTGATGTGCAGGTGAAGC-3⬘ for mouse LARC/CCL20; ⫹5⬘-GGAC-

Production of LARC/CCL20 by epidermal keratinocytes 97 CTGGAGCTGTTCTTTGGGTT-3⬘ and –5⬘-GTTCCGGTGTAGAGG-CGAGGACTT-3⬘ for mouse CCR6; ⫹5⬘-GCCAAGGTCATCCATGACAACTTTGG-3⬘ and –5⬘-GCCTGCTTCACCACCTTCTTGATGTC-3⬘ for human and mouse G3PDH. Amplification conditions were denaturation at 94°C for 30 s (5 min for the first cycle), annealing at 60°C for 30 s and extension at 72°C for 30 s (5 min for the last cycle) for 33 cycles for human LARC/CCL20, 35 cycles for human CCR6, 32 cycles for mouse LARC/CCL20 and CCR6, and 27 cycles for human and mouse G3PDH. Amplification products (10 µl each) were subjected to electrophoresis on 2% agarose and stained with ethidium bromide. ELISA LARC/CCL20 was measured by a sandwich-type ELISA using mouse anti-human MIP-3α/CCL20 mAb (67310.111) (Genzyme/Techne, Cambridge, MA) as capturing antibody and biotinylated goat anti-human MIP-3α/CCL20 polyclonal antibody (R & D Systems, Minneapolis, MN) as detecting antibody. For standardization of the assay, serially diluted recombinant MIP-3α/CCL20 (PeproTech) was included on each ELISA plate. The detection range of the present ELISA was typically between 20 pg/ml and 1 ng/ml.

conditions for 1 week before experiments. Mice were treated with depilatory cream (Kanebo, Tokyo, Japan) on the back to remove hair. After 24 h, mice were anesthetized by inhalation of diethyl ether and injected intradermally with 100 µl of PBS alone or PBS containing 1 ng of IL-1α and 5 ng of TNF-α. After various time points, mice were sacrificed by cervical dislocation and skin tissues of 6 mm diameter were taken from injected sites. RT-PCR analysis was carried out using total RNA as described above. Subjects A total of 42 outpatients diagnosed as having atopic dermatitis and age-matched healthy adults working in the medical school (n ⫽ 13) were examined. Diagnosis and severity of atopic dermatitis were assessed according to the criteria of Hannifin and Rajka (18) and Rajka and Langeland (19) respectively. Fresh plasma samples were obtained by using EDTA as anti-coagulant as described previously (20). Skin biopsies were carried out from several donors as described previously (21). Informed consents were obtained from all the donors and the study was approved by the Kinki University Hospital Ethics Review Board. Immunohistochemistry

Chemotaxis assay This was carried out using murine L1.2 cells stably expressing human CCR6 as described previously (9). Recombinant human MIP-3α/CCL20 was purchased from PeproTech. In brief, 2.5 ⫻105 cells suspended in 25 µl of phenol redfree RPMI 1640 containing 1 mg/ml of BSA (Sigma) and 20 mM HEPES, pH 7.4, were applied into upper wells of a CHEMOTX chemotaxis chamber of 5 µm pore size (Neuro Probe, Gaithersburg, MD). Test samples such as recombinant MIP-3α/CCL20 or culture supernatants preincubated without or with anti-MIP-3α/CCL20 mAb (67310.111) for 30 min were applied into lower wells in a volume of 30 µl. After 4 h at 37°C, cells migrated into lower wells were lysed with 0.1% Triton X-100 and measured by using PicoGreen doublestranded DNA quantitation reagent (Molecular Probes, Eugene, OR). Cell migration was expressed by percent input cells. All assays were performed in triplicate. Generation of anti-CCR6 mAb This was carried out as described previously (17). In brief, female BALB/c mice were i.p. immunized with 107 murine L1.2 cells stably expressing human CCR6 6 times over a 3 month period. Three days after final immunization, mice were killed and spleen cells were fused with P3xAg8. Supernatants of growing hybridomas were screened by differential staining of CCR6-expressing L1.2 cells and parental L1.2 cells. Positive hybridomas were cloned by limiting dilution cultures. One anti-CCR6 mAb used in the present study, clone #125 (IgG1,κ), was confirmed to be specific for CCR6 by staining a panel of murine L1.2 cells expressing CCR1–9. The generation and characterization of anti-CCR6 mAb will be described elsewhere (Kawasaki et al.). Mice Female BALB/c mice (6 weeks old) were purchased from Kurea (Osaka, Japan) and kept in specific pathogen-free

Tissue samples were immersed in Tissue-Tek (Lab Products, Naperville, IL) and snap-frozen in liquid nitrogen. Sections of 4.5 µm thickness were fixed with acetone, treated with 3% H2O2 for 5 min, washed with Tris-buffered saline, pH 7.4, and blocked with 5% normal goat serum in Tris-buffered saline. Sections were incubated with 2 µg/ml of each mAb at room temperature for 60 min. The following mAb were used: antiCD1a (Leu 6) (Becton Dickinson, Mountain View, CA), antiCD3 (Leu 4) (Beckton Dickinson) and anti-HLA-DR (DK22) (Dako, Carpinteria, CA). After washing with Tris-buffered saline, sections were incubated with biotinylated rabbit antimouse antibody (Dako). After washing with Tris-buffered saline, sections were incubated with horseradish peroxidaseconjugated streptavidin (Dako). The sections were treated with a mixture of 3-amino-9-ethyl carbazole and H2O2 in 0.05 M acetate buffer, pH 5.0, counterstained with Mayer’s hematoxlin and mounted with Aquamout (MUTO Pure Chemicals, Tokyo, Japan). Results Induction of LARC/CCL20 mRNA in primary keratinocytes Primary human neonatal keratinocytes were kept in tissue culture until confluence and treated for 24 h without or with LPS, IL-1α, TNF-α, IFN-γ and IL-4, either alone or in various combinations. Expression of LARC/CCL20 mRNA was analyzed by RT-PCR. Representative results using primary keratinocytes from three different donors are shown in Fig. 1(A). Untreated keratinocytes expressed LARC/CCL20 at low levels (Fig. 1, lane 3). Singly, IL-1α (Fig. 1A, lane 5) and TNF-α (Fig. 1A, lane 7) clearly induced LARC/CCL20 mRNA in keratinocytes. In contrast, LPS (Fig. 1A, lane 4), IFN-γ (Fig. 1A, lane 6) or IL-4 (Fig. 1A, lane 8) were hardly effective. In combination, IL-1α ⫹ TNF-α (Fig. 1A, lane 10) most strongly induced LARC/CCL20 mRNA. This was followed

98 Production of LARC/CCL20 by epidermal keratinocytes

Fig. 1. Inducible expression of LARC/CCL20 by primary human epidermal keratinocytes. Total RNA was prepared from keratinocytes treated without or with various cytokines for 24 h. RT-PCR analysis was carried out for LARC/CCL20 and G3PDH as described in Methods. (A) Effects of various cytokines. 1, RNA(–); 2, phytohemagglutinin-stimulated peripheral blood mononuclear cells (positive control); 3, untreated keratinocytes; 4, keratinocytes treated with 30 ng/ml of LPS; 5, keratinocytes treated with 10 ng/ml of IL-1α; 6, keratinocytes treated with 100 ng/ml of IFN-γ; 7, keratinocytes treated with 50 ng/ml of TNF-α; 8, keratinocytes treated with 10 ng/ml of IL-4; 9, keratinocytes treated with 10 ng/ml of IL-1α ⫹ 100 ng/ml of IFN-γ; 10, keratinocytes treated with 10 ng/ml of IL-1α ⫹ 50 ng/ml of TNF-α; 11, keratinocytes treated with 10 ng/ml of IL-1α ⫹10 ng/ml of IL-4; 12, keratinocytes treated with 50 ng/ml of TNF-α ⫹ 100 ng/ml of IFN-γ; 13, keratinocytes treated with 50 ng/ml of TNF-α ⫹ 10 ng/ml of IL-4; 14, keratinocytes treated with 100 ng/ml of IFN-γ ⫹ 10 ng/ml of IL-4. (B) Dose–response effect of IL-1α. 1, RNA(–); 2, phytohemagglutinin-treated peripheral blood mono- nuclear cells (positive control); 3, untreated keratinocytes; 4, keratinocytes treated with 0.01 ng/ml of IL-1α; 5, keratinocytes treated with 0.1 ng/ml of IL-1α; 6, keratinocytes treated with 1 ng/ml of IL1α; 7, keratinocytes treated with 10 ng/ml of IL-1α; 8, keratinocytes treated with 100 ng/ml of IL-1α. Representative results from three independent experiments are shown.

by IL-1α ⫹ IFN-γ (Fig. 1A, lane 9) and TNF-α ⫹ IFN-γ (Fig. 1A, lane 12). Similar results were obtained by using IL-1β (data not shown). Thus, proinflammatory cytokines such as IL-1 and TNF-α were capable of inducing LARC/CCL20 mRNA in primary epidermal keratinocytes, while IFN-γ (a Th1-type cytokine) but not IL-4 (a Th2-type cytokine) had augmenting effects. When different concentrations of IL-1α were tested, the maximal induction was observed at 10 ng/ml or more (Fig. 1B). Production of LARC/CCL20 protein by primary keratinocytes We next examined production and secretion of LARC/CCL20 protein by primary keratinocytes. Both polyclonal and monoclonal anti-LARC/CCL20 clearly stained the cytoplasm of keratinocytes treated with IL-1α and TNF-α for 1–3 days (data not shown). To demonstrate secretion of LARC/CCL20 by stimulated keratinocytes, we developed a sandwich-type ELISA for LARC/CCL20 using monoclonal anti-LARC/CCL20 as catching antibody and polyclonal anti-LARC/CCL20 as detection antibody. As shown in Fig. 2, untreated keratinocytes did not secrete LARC/CCL20 at levels detectable by

Fig. 2. Secretion of LARC/CCL20 by primary human epidermal keratinocytes treated with various cytokines. Keratinocytes were treated with 10 ng/ml of IL-1α, 50 ng/ml of TNF-α, 100 ng/ml of IFN-γ and 10 ng/ml of IL-4, either singly or in combination as indicated. After 3 days, LARC/CCL20 secreted in culture supernatants was measured by ELISA as described in Methods. All assays were performed in triplicate to obtain mean ⫾ SD. Representative results from three independent experiments are shown.

our ELISA (⬎20 pg/ml) even by day 3. Upon treatment with various cytokines singly or in combination, the amounts of LARC/CCL20 secreted in culture supernatants mostly paralleled the levels of induction of LARC/CCL20 mRNA as demonstrated in Fig. 1. Thus, singly, IL-1α was most effective in inducing LARC/CCL20 secretion by keratinocytes. In contrast to more efficient induction of LARC/CCL20 mRNA by TNF-α than by IFN-γ, however, TNF-α was less potent in inducing LARC/CCL20 secretion than IFN-γ. IFN-γ may be more efficient in inducing production and/or secretion of LARC/CCL20 than TNF-α. In combination, IL-1α and TNF-α or IFN-γ synergistically induced secretion of LARC/CCL20.

Production of LARC/CCL20 by epidermal keratinocytes 99

Fig. 3. Chemotactic activity of LARC/CCL20 secreted by primary human epidermal keratinocytes. Culture supernatants of keratinocytes treated without or with 10 ng/ml of IL-1α and 50 ng/ml of TNF-α for 3 days were tested for chemotactic activity on CCR6expressing L1.2 cells. Controls included fresh keratinocyte culture medium supplemented without or with 10 ng/ml of IL-1α and 50 ng/ml of TNF-α, and chemotaxis assays using parental L1.2 cells. 1, Fresh keratinocyte medium only; 2, 0.1 nM recombinant LARC/ CCL20; 3, 1 nM recombinant LARC/CCL20; 4, 10 nM recombinant LARC/CCL20; 5, 100 nM recombinant LARC/CCL20; 6, culture supernatant of untreated keratinocytes; 7, culture supernatant from keratinocytes treated with IL-1α and TNF-α; 8, anti-LARC/CCL20 mAb ⫹ culture supernatant from keratinocytes treated with IL-1α and TNFα; 9, fresh keratinocyte culture medium supplemented with IL-1α and TNF-α; 10, culture supernatant from keratinocyte treated with IL-1α and TNF-α tested on parental L1.2 cells.

On the other hand, the combined effects of TNF-α and IFN-γ were mostly additive. We further examined whether dexamethasone and FK506, the drugs effective on atopic dermatitis (22), were capable of suppressing LARC/CCL20 production by keratinocytes treated with IL-1α. The levels of LARC/CCL20 secreted by keratinocytes treated with 10 ng/ml of IL-1α for 3 days was not affected by the presence of dexamethasone even at 10–6 M or FK506 even at 2 µg/ml (data not shown). RT-PCR analysis also demonstrated that induction of LARC/CCL20 mRNA in keratinocytes by IL-1α was not suppressed by either of these drugs (data not shown). Thus, the signaling events involved in LARC/CCL20 induction in keratinocytes by IL-1α appears to be mostly independent of transcription factors such as AP-1, NF-κB and NF-AT, the known targets of these anti-inflammatory drugs (23). Chemotactic activity of LARC/CCL20 secreted by primary keratinocytes We next tested whether LARC/CCL20 protein secreted by keratinocytes was biologically active by chemotaxis assays using murine L1.2 cells stably expressing CCR6, the unique receptor for LARC/CCL20 (9–11). This experiment was also intended to examine a possible involvement of β-defensins, which have been reported to be produced by epidermal keratinocytes upon stimulation with proinflammatory cytokines such as TNF-α and to be chemotactic through CCR6 (16). As shown in Fig. 3, L1.2 cells expressing human CCR6 migrated to recombinant LARC/CCL20 with a typical bellshaped dose–response curve and a peak response at 10–100 nM of LARC/CCL20. They also migrated to the culture supernatant of keratinocytes treated with IL-1α ⫹ TNF-α for

Fig. 4. Induction of LARC/CCL20 in mouse skin by IL-1α and TNF-α. Female BALB/c mice were treated with depilatory cream and the hair on the back was removed. After 24 h, mice were injected intradermally in the back with 100 µl of PBS containing 1 ng of IL-1α and 5 ng of TNF-α or PBS alone. Total RNA was prepared from skin biopsies (6 mm diameter) taken from the injection sites. RT-PCR analysis was carried out for LARC/CCL20, CCR6 and G3PDH as described in Methods. 1, RNA(–); 2, mouse intestine (positive control); 3 and 4, untreated skin biopsies; 5 and 6, skin biopsies at 30 min after injection with IL-1α ⫹ TNF-α; 7 and 8, skin biopsies at 1 h after injection with IL-1α ⫹ TNF-α; 9 and 10, skin biopsies at 2 h after injection with IL-1α ⫹ TNF-α; 11 and 12, skin biopsies at 3 h after injection with IL-1α ⫹ TNF-α; 13 and 14, skin biopsies at 4 h after injection with IL-1α ⫹ TNF-α; 15 and 16, skin biopsies at 8 h after injection with IL-1α ⫹ TNF-α. Representative results from two independent experiments are shown.

3 days. Chemotactic responses to the culture supernatant were completely blocked by anti-LARC/CCL20. No such responses were seen to fresh culture medium supplemented with IL-1α and TNF-α or by using parental L1.2 cells as responding cells. These results confirmed that LARC/CCL20 secreted by keratinocytes treated with IL-1α ⫹ TNF-α was biologically active and represented a major chemotactic factor for CCR6-expressing L1.2 cells in the culture supernatant. However, considering the fact that the concentration of LARC/ CCL20 in the culture supernatant was ~0.2 nM but migration of CCR6-expressing L1.2 cells to the culture supernatant was much more vigorous than to recombinant LARC/CCL20 at 1 nM, a synergistic factor(s) might be present in the culture supernatant. Induction of LARC/CCL20 in mouse skin by IL-1α and TNF-α To test the inducibility of LARC/CCL20 in the skin, we treated BALB/c mice with intradermal administration of IL-1α and TNF-α. After various time points, skin tissues at the injection sites were punched out and RT-PCR analysis for murine LARC/CCL20 and CCR6 was carried out. As shown in Fig. 4, mLARC/CCL20 mRNA was strongly induced with a peak at 1–3 h after injection of IL-1α and TNF-α. No such induction was seen by intradermal injection of PBS alone (data not shown). Furthermore, induction of mLARC/CCL20 mRNA was accompanied by local accumulation of mCCR6 mRNA with a lag of a few hours. Thus, IL-1α and TNF-α indeed induced mLARC/CCL20 expression in the skin of mice, which was most likely to be followed by accumulation of mCCR6expressing cells. Unavailability of anti-mLARC/CCL20 or antimCCR6, however, precluded us from further demonstrating production of mLARC/CCL20 protein by epidermal keratinocytes in situ upon treatment with IL-1α ⫹ TNF-α or local infiltration of mCCR6-expressing cells in the murine skin.

100 Production of LARC/CCL20 by epidermal keratinocytes Augmented expression of LARC/CCL20 and accumulation of CCR6-expressing cells in human atopic skin lesions It is known that epidermal keratinocytes produce IL-1α and TNF-α in inflammatory skin lesions such as atopic dermatitis (24). Therefore, we examined expression of LARC/CCL20 and CCR6 mRNA in normal skin tissues and atopic skin tissues by RT-PCR. As shown in Fig. 5, the contents of LARC/CCL20 and CCR6 mRNA were strongly augmented in atopic skin lesions (Fig. 2, lanes 6–8) in comparison with normal skin tissues (Fig. 2, lanes 3–5). To further examine LARC/CCL20 production and infiltration of CCR6-expressing cells in atopic skin lesions, we carried

Fig. 5. RT-PCR analysis of expression of LARC/CCL20 and CCR6 in human skin. Total RNA was prepared from peripheral blood mononuclear cells treated with phytohemagglutinin for 24 h and from skin biopsies from three normal donors and three atopic dermatitis patients. RT-PCR analysis was carried out for LARC/CCL20, CCR6 and G3PDH as described in Methods. 1, RNA(–); 2, phytohemagglutinintreated peripheral blood mononuclear cells (positive control); 3–5, skin biopsies from normal donors; 6–8, skin biopsies from atopic patients.

out immunohistochemical staining of LARC/CCL20 and CCR6 in normal skin tissues from patients undergoing mammectomy (n ⫽ 2) and lichenified plaques from atopic patients (n ⫽ 7). The representative results are shown in Fig. 6. In normal skin tissues (n ⫽ 2), epidermal keratinocytes were hardly stained by monoclonal anti-LARC/CCL20 and only sporadic cells within the dermis were positive for LARC/CCL20 (Fig. 6A). Furthermore, a newly generated monoclonal anti-CCR6 hardly stained any cells present in normal skin tissues including the epidermis where numerous Langerhans cells were present (Fig. 6B). On the other hand, we observed staining of epidermal keratinocytes by anti-LARC/CCL20 in lichenified skin lesions from atopic patients. We noticed two types of staining. In some skin samples (n ⫽ 4), clusters of epidermal keratinocytes at the basal layer, which were mostly those facing the dermal papillae, were focally stained by anti-LARC/ CCL20 (Fig. 6C). In other skin samples (n ⫽ 3), epidermal layers were more widely stained by anti-LARC/CCL20 (Fig. 6D). The former staining pattern might indicate that keratinocytes at the basal layer were induced to express LARC/CCL20 by factors produced by cells infiltrating the dermal papillae. In the case of anti-CCR6 staining, we only observed inconclusive staining in lesional skin tissues of most atopic patients (n ⫽ 5). This was probably due to the low sensitivity of this monoclonal anti-CCR6. We were, however, able to demonstrate CCR6⫹ cells infiltrating in the dermis and especially in dermal papillae in some atopic skin tissues (n ⫽ 2) (Fig. 6E and F). In these samples, we also confirmed that CCR6⫹ cells were mostly co-localized with cells positive for CD1a (Fig. 6G), CD3 (Fig. 6H) or HLA-DR (Fig. 6I). Thus, even though double-staining experiments were not done,

Fig. 6. Immunohistochemical analysis of human skin tissues. (A) Normal skin stained with anti-LARC/CCL20 mAb; (B) normal skin stained with anti-CCR6 mAb; (C) atopic skin stained with anti-LARC/CCL20 mAb; (D) atopic skin stained with anti-LARC/CCL20 mAb; (E) atopic skin stained with anti-CCR6 mAb; (F) atopic skin stained with anti-CCR6 mAb at high magnification (allow heads indicate positive cells); (G) atopic skin stained with anti-CD1a; (H) atopic skin stained with anti-CD3; (I) atopic skin stained with anti-HLA-DR.

Production of LARC/CCL20 by epidermal keratinocytes 101 CCR6-expressing cells were likely to be infiltrating dendritic cells and/or T cells. It is also notable that hardly any intraepidermal Langerhans cells were stained by anti-CCR6 (Fig. 6B and F) even though there were numerous CD1a⫹ Langerhans cells in the epidermal layer (Fig. 6G) (26). Collectively, epidermal keratinocytes in the skin indeed produce LARC/ CCL20 in inflammatory conditions such as atopic dermatitis, which in turn may induce dermal infiltration of CCR6expressing cells such as immature dendritic cells and CLA⫹ memory T cells (9–11,14). On the other hand, LARC/CCL20 and CCR6 may not play a major role in the homeostatic migration of Langerhans cells into the skin. Examination of plasma contents of LARC/CCL20 We examined LARC/CCL20 contents in plasma samples from normal donors (n ⫽ 13) and atopic patients (n ⫽ 42) by using the specific ELISA for LARC/CCL20. As shown in Fig. 7, among healthy adult donors, only two donors contained LARC/ CCL20 at levels above the detection limit (⬎20 pg/ml). Among atopic patients, however, many samples were found to contain LARC/CCL20 at high levels. Furthermore, patients with severer disease conditions tended to have high levels of LARC/CCL20 in the plasma. These results support that LARC/CCL20 has an important pathogenic role in atopic dermatitis.

Fig. 7. Plasma contents of LARC/CCL20 in normal donors and atopic patients. Plasma contents of LARC/CCL20 were determined by specific ELISA as described in Methods. All assays were performed in duplicate. Statistical significance was determined by unpaired Student’s t-test. *P ⬍ 0.05.

Discussion Even though expression of LARC/CCL20 was originally reported in tissues such as liver and lung (4,6), subsequent studies have shown that LARC/CCL20 is normally expressed by mucosal epithelial cells both in humans and mice (7,8). The intestinal immune system is the largest part of the immune system, and consists of organized lymphoid tissues such as Peyer’s patches and scattered cell components such as lymphocytes and plasma cells in the lamina propria and intraepithelial lymphocytes enriched for γδ type T cells (26). In intestinal tissues of mice, FAE cells were found to constitutively express LARC/CCL20 while other epithelial cells were also capable of transiently expressing LARC/CCL20 upon treatment of animals with LPS (7). We also observed that cells expressing CCR6 mRNA, the only known receptor for LARC/ CCL20 (9–11), were present in Peyer’s patches and localized especially in the region just beneath the FAE (unpublished results). The FAE is a single layer of cuboidal epithelial cells and contains a unique population of specialized epithelial cells termed M cells which uptake and transport luminal antigens into the underlying lymphoid tissues (26). The development of intestinal lymphoid tissues is dependent on the presence of normal bacterial flora which produce LPS (27) and the underlying lymphoid tissues also promote the development of FAE (28). Furthermore, surface expression of CCR6 has been demonstrated in immature dendritic cells, B cells and most α4β7high intestine-seeking memory T cells (10– 12,14). Recently, Cook et al. generated mice with targeted disruption of CCR6 and found that myeloid-derived dendritic cells were selectively absent from the subepithelial dome of Peyer’s patches in these mice (29). Collectively, these results strongly suggest that under the influence of bacterial flora and other factors derived from underlying lymphoid tissues, FAE constitutively express LARC/CCL20, and attract immature dendritic cells, B cells and α4β7high memory T cells toward the mucosal surface. Thus, LARC/CCL20 is likely to play an important role in the development and function of organized mucosal lymphoid tissues. The skin is also a large part of the immune system. As the case of intestinal mucosa, LARC/CCL20 may play an important role in the recruitment of immature dendritic cells and T cells into the skin. This idea is supported by the finding that a substantial proportion of CLA⫹ memory T cells also expresses CCR6 (14). However, there are crucial anatomical and functional differences between skin and intestinal tissues. The skin epidermal layer primarily provides a strong mechanical barrier against invading pathogens. Immature dendritic cells called Langerhans cells constitutively migrate into the epidermal layer and function as sentinel cells ready to uptake invading microbes and foreign antigens (25). There are no organized lymphoid tissues in the dermal layer and antigen-loaded dendritic cells migrate into regional lymph nodes to initiate immune responses. Furthermore, the ligand– receptor pairs rather unique in the skin such as thymus and activation-regulated chemokine/CCL17–CCR4 and cutaneous T cell-attracting chemokine/IL-11 receptor α locus chemokine/ CCL27–CCR10 may play important roles in the emigration of CLA⫹ skin-seeking memory T cells and other cells into the skin (30–33).

102 Production of LARC/CCL20 by epidermal keratinocytes Recently, Charbonnier et al. demonstrated that the epidermal keratinocytes from clinically normal skin was immunologically strongly positive for LARC/CCL20 while epidermis-derived Langerhans cells as well as in vitro differentiated CD34⫹ hematopoietic precursor cell-derived Langerhans-like cells expressed CCR6 (15). From these observations, the authors have proposed a scenario that epidermal keratinocytes attract CCR6-expressing Langerhans cells into the epidermal layer by constitutively producing LARC/CCL20 (15). Our present results, however, may lead to a different scenario. We have demonstrated that primary keratinocytes in culture are capable of producing LARC/ CCL20 in large quantities only upon stimulation with proinflammatory cytokines such as IL-1α and TNF-α (Figs 1 and 2). In vivo, LARC/CCL20 was rapidly induced in the skin of mice by intradermal injection of IL-1α and TNF-α, and this induction was followed by an increase in CCR6 mRNA, suggesting local accumulation of CCR6-expressing cells (Fig. 4). In humans, epidermal keratinocytes were positive for LARC/CCL20 only in lesional skin tissues from atopic patients but not in normal skin tissues (Figs 5 and 6). Consistently, a newly generated anti-CCR6 mAb stained a large number of cells infiltrating in the dermis of lesional skin from at least some atopic patient but hardly stained any Langerhans cells in the epidermis of normal or atopic skin tissues (Fig. 6). After completion of this work, Homey et al. (34) also reported marked up-regulation of LARC/MIP-3α/CCL20 and CCR6 in psoriatic skin lesions where lesional epidermal keratinocytes expressed LARC/MIP-3α/CCL20 while skin-infiltrating CLA⫹ T cells were likely to express CCR6. However, they also did not observe any staining of Langerhans cells in the epidermis by anti-CCR6. Collectively, production of LARC/CCL20 by epidermal keratinocytes appears to be closely related to proinflammatory responses of the skin. This most probably induces infiltration of CCR6-expressing cells such as immature dermal dendritic cells and CLA⫹ memory T cells in the dermis of inflamed skin tissues. On the other hand, LARC/CCL20 may not be the major chemokine involved in homeostatic migration of Langerhans cells into the epidermis. The cause of discrepancy between the results of Carbonnier et al. (15) and those of the present study is not clear at present, but different antibodies (their polyclonal antibodies versus our mAb) and staining conditions may be responsible. Thus, our present results do not exclude expression of LARC/ CCL20 at low levels by normal epidermal keratinocytes and expression of CCR6 by Langerhans cells at low levels or in a form not detectable by our anti-CCR6 mAb. Nevertheless, their RT-PCR evidence for CCR6 mRNA expression by Langerhans cells purified ex vivo is not entirely convincing because the weak signals might be partly due to contaminating keratinocytes which were also noted to express CCR6 at levels detectable by RT-PCR (15 and our unpublished results). Dendritic cells are known to be highly heterogeneous in terms of cell lineage, tissue localization, maturation stage, function and surface antigens (25). For example, typical Langerhans cells are defined as CD1a⫹E-cadherin⫹Birbeck granule⫹CD11b–CD36–factor XIIIa– while dermal dendritic cells are characterized as CD1a⫹E-cadherin–Birbeck granule– CD11b⫹CD36⫹factor XIIIa⫹ (25). While dermal dendritic cells are likely to be attracted by LARC/CCL20, Langerhans cells

may employ a chemokine(s) different from LARC/CCL20 for trafficking into the skin. In conclusion, the function of LARC/CCL20 in the skin appears to be more related to inflammatory responses than to homeostatic cell trafficking. Such roles of LARC/CCL20 are in fact quite consistent with its highly transient mode of expression in producer cells and tissues upon induction with proinflammatory stimuli (4,33). The presence of multiple mRNA destabilization signals in the 3⬘ non-coding region of LARC/ CCL20 transcripts (4,31) probably accounts for its highly transient expression. Together with the finding that βdefensins, the small cationic antimicrobial peptides produced by epidermal keratinocytes upon microbial invasion or by stimulation with TNF-α are also potent agonists for CCR6 (16), the LARC/CCL20–CCR6 system may play important roles in the host defense against invading pathogens in the skin as well as in the pathogenesis of chronic inflammatory skin diseases such as atopic dermatitis and psoriasis. Acknowledgements We thank Kinuyo Aisu for her excellent technical assistance. This work was supported in part by grants-in-aid from the Ministry of Education, Science and Culture of Japan.

Abbreviations CCL CLA FAE GM-CSF LARC LPS MIP TNF

CC chemokine ligand cutaneous lymphocyte-associated antigen follicle-associated epithelium granulocyte macrophage colony stimulating factor liver- and activation-regulated chemokine lipopolysaccharide macrophage inflammatory protein tumor necrosis factor

References 1 Rollins, B. 1997. Chemokines. Blood 90:909. 2 Zlotnik, O. and Yoshie, O. 2000. Chemokines: a new classification system and their role in immunity. Immunity 12:121. 3 Yoshie, O. 2000. Immune chemokines and their receptors: the key elements in the genesis, homeostasis and function of the immune system. Semin. Immunopathol. 22:250 4 Hieshima, K., Imai, T., Opdenakker, G., Van Damme, J., Kusuda, J., Tei, H., Sakaki, Y., Takatsuki, K., Miura, R., Yoshie, O. and Nomiyama, H. 1997. Molecular cloning of a novel human CC chemokine liver and activation-regulated chemokine (LARC) expressed in liver. J. Biol. Chem. 272:5846. 5 Ross, D. L., Vicari, A. P., Franz-Bacon, K., McClanahan, T. K. and Zlotnik, A. 1997. Identification through bioinformatics of two new macrophage proinflammatory human chemokines MIP-3α and MIP-3β. J. Immunol. 158:1033. 6 Hromas, R., Gray, P. W., Chantry, D., Godiska, R., Krathwohl, M., Fife, K., Bell, G. I., Takeda, J., Aronica, S., Gordon, M., Cooper, S., Broxmeyer, H. E. and Klemsz, M. J. 1997. Cloning and characterization of exodus, a novel β-chemokine. Blood 89:3315. 7 Tanaka, Y., Imai, T., Baba, M., Ishikawa, I., Uehira, M., Nomiyama, H. and Yoshie, O. 1999. Selective expression of liver and activation-regulated chemokine (LARC) in intestinal epithelium in mice and humans. Eur. J. Immunol. 29:633. 8 Dieu, M.-C., Vanberliet, B., Vicari, A., Bridon, J.-M., Oldham, E., Ait-Yahia, S., Briere, F., Zlotnik, A., Lebecque, S. and Caux, C. 1998. Selective recruitment of immature and mature dendritic cells by distinct chemokines expressed in different anatomic sites. J. Exp. Med. 188:373. 9 Baba, M., Imai, T., Nishimura, M., Kakizaki, M., Takagi, S. Hieshima, K. Nomiyama, H. and Yoshie, O. 1997. Identification of

Production of LARC/CCL20 by epidermal keratinocytes 103

10

11

12

13

14

15

16

17

18 19 20

21

CCR6, the specific receptor for a novel lymphocyte-directed CC chemokine LARC. J. Biol. Chem. 272:14893. Power, C. A., Church, D. J., Meyer, A., Alouani, S., Proudfoot, A. E. I., Clark-Lewis, I., Sozzani, S., Mantovani, A. and Wells, T. N. C. 1997. Cloning and characterization of a specific receptor for the novel CC chemokine MIP-3α from lung dendritic cells. J. Exp. Med. 186:825. Greaves, D. R., Wang, W., Dairaghi, D. J., Dieu, M. C., de Saint-Vis, B., Franz- Bacon, K., Rossi, D., Caux, C., McClanahan, T., Gordon, S., Zlotnik, A. and Schall, T. J. 1997. CCR6, a CC chemokine receptor that interacts with macrophage inflammatory protein 3α and is highly expressed in human dendritic cells. J. Exp. Med. 186:837. Yang, D., Howard, O. M. Z., Chen, Q. and Oppenheim, J. J. 1999. Immature dendritic cells generated from monocytes in the presence of TGF-β1 express functional C–C chemokine receptor 6. J. Immunol. 163:1737. Sallusto, F., Schaerli, P., Loetscher, P., Schaniel, C., Lenig, D., Mackay, C. R., Qin, S. and Lanzavecchia, A. 1998. Rapid and coordinated switch in chemokine receptor expression during dendritic cell maturation. Eur. J. Immunol. 28:2760. Liao, F., Rabin, R. L., Smith, C. S., Sharma, G., Nutman, T. B. and Farber, J. M. 1999. CC-chemokine receptor 6 is expressed on diverse memory subsets of T cells and determines responsiveness to macrophage inflammatory protein 3α. J. Immunol. 162:186. Charbonnier, A.-S., Kohrgruber, N., Kriehuber, E., Stingl, G., Rot, A. and Maurer, D. 1999. Macrophage inflammatory protein 3α is involved in the constitutive trafficking of epidermal Langerhans cells. J. Exp. Med. 190:1755. Yang, D., Chertov, O., Bykovskaia, S. N., Chen, Q., Buffo, M. J., Shogan, J., Anderson, M., Schoeder, J. M., Wang, J. M., Howard, O. M., Z. and Oppenheim, J. J. 1999. β-Defensins: Linking innate and adaptive immunity through dendritic and T cell CCR6. Science 286:525. Sato, K., Kawasaki, H., Nagayama, H., Serizawa, R., Ikeda, J., Morimoto, C. Yasunaga, K., Yamaji, N., Tadokoro, K., Juji, T. and Takahashi, T. A. 1999. CC chemokine receptors, CCR-1 and CCR3, are potentially involved in antigen-presenting cell function of human peripheral blood monocyte-derived dendritic cells. Blood 93:34. Haniffin, J. and Rajka, G. 1980. Diagnostic features of atopic dermatitis. Acta Derm. Venerol. 92:44. Rajka, G. and Langeland, T. 1989. Grading of the severity of atopic dermatitis. Acta Derm. Venereol. 144:13. Morita, A., Shimosako, K., Kikuoka, S., Taniguchi, Y., Kitaura, M., Sasakura, K., Tamaki, M., Tuji, T., Teraoka, H., Yoshie, O., Nakajima, T. and Hirai, K. 1999. Development of a sensitive enzyme-linked immunosorbent assay for eotaxin and measurement of its levels in human blood. J. Immunol. Methods 226:159. Yamada, H., Izutani, R., Chihara, J., Yudate, T., Matsukura, M. and Tezuka, T. 1996. RANTES mRNA expression in skin and colon of patients with atopic dermatitis. Int. Arch. Allergy Immunol. 111:19.

22 Fleischer, A. B., Jr. 1999. Treatment of atopic dermatitis: role of tacrolimus ointment a topical noncorticosteroidal therapy. J. Allergy Clin. Immunol. 104:S126. 23 Paliogianni, F., Raptis, A., Ahuja, S. S., Najjar, S. M. and Boumpas, D. T. (1993) Negative transcriptional regulation of human interleukin 2 (IL-2) gene by glucocorticoids through interference with nuclear transcription factors AP-1 and NF- AT. J. Clin. Invest. 91:1481. 24 Ansel, J., Perry, P., Brown, J., Damm, D., Phan, T., Hart, C., Luger, T. and Hefeneider, S. 1990. Cytokine modulation of keratinocyte cytokines. J. Invest. Dermatol. 94:101S. 25 Hart, D. N. J. 1997. Dendritic cells: unique leukocyte populations which control the primary immune response. Blood 90:3245. 26 Mowat, A. M. and Viney, J. L. 1997. The anatomical basis of intestinal immunity. Immunol. Rev. 156:145. 27 Smith, M. W., James, P. S. and Tivey, D. R. 1987. M cell numbers increase after transfer of SPF mice to a normal animal housing environment. Am. J. Pathol. 128:385. 28 Kerneis, S., Bogdanova, A., Kraehenbuhl, J.-P. and Pringault, E. 1997. Conversion by Peyer’s patch lymphocytes of human enterocytes into M cells that transport bacteria. Science 277:949. 29 Cook, D. N., Prosser, D. M., Forster, R., Zhang, J., Kuklin, N. A., Abbondanzo, S. J., Niu, X.-D., Chen, S.-C., Manfra, D. J., Wiekowski, M. T., Sullivan, L. M., Smith, S. R., Greenberg, H. B., Narula, S. K., Lipp, M. and Lira, S. A. 2000. CCR6 mediates dendritic cell localization, lymphocyte homeostasis, and immune responses in mucosal tissue. Immunity 12:495. 30 Campbell, J. J., Haraldsen, G., Pan, J., Rottman, J., Qin, S., Ponath, P., Andrew, D. P., Warnke, R., Ruffing, N., Kassam, N., Wu, L. and Butcher, E. C. 1999. The chemokine receptor CCR4 in vascular recognition by cutaneous but not intestinal memory T cells. Nature 400:776. 31 Morales, J., Homey, B., Vicari, A. P., Hudak, S., Oldham, E., Hedrick, J., Orozco, R., Copeland, N. G., Jenkins, N. A., McEvoy, L. M. and Zlotnik, A. 1999. CTACK, a skin-associated chemokine that preferentially attracts skin-homing memory T cells. Proc. Natl Acad. Sci. USA 96:14470. 32 Ishikawa-Mochizuki, I., Kitaura, M., Baba, M., Nakayama, T., Izawa, D., Imai, T., Yamada, H., Hieshima, K., Suzuki, R., Nomiyama, H. and Yoshie, O. 1999. Molecular cloning of a novel CC chemokine, interleukin-11 receptor α-locus chemokine (ILC), which is located on chromosome 9p13 and a potential homologue of a CC chemokine encoded by molluscum contagious virus. FEBS Lett. 460:544. 33 Utans-Schneitz, U., Lorez., H., Klinkert, W. E. F., da Silva, J. and Lesslauer, W. 1998. A novel rat CC chemokine, identified by targeted differential display, is upregulated in brain inflammation. J. Neuroimmunol. 92:179. 34 Homey, B., Dieu-Nosjean, M. C., Wiesenborn, A., Massacrier, C., Pin, J. J., Oldham, E., Catron, D., Buchanan, M. E., Muller, A., deWaal Malefyt, R., Deng, G., Orozco, R., Ruzicka, T., Lehmann, P., Lebecque, S., Caux, C. and Zlotnik, A. 2000. Up-regulation of macrophage inflammatory protein-3 alpha/CCL20 and CC chemokine receptor 6 in psoriasis. J. Immunol. 164:6621.