Up-Regulation of Macrophage Inflammatory Protein-3α/CCL20 and CC Chemokine Receptor 6 in Psoriasis

Patients

Six millimeter punch biopsies were taken, after obtaining informed consent, from either lesional (n = 10) and nonlesional (n = 5) skin of psoriatic patients or from normal (n = 5) healthy individuals. Skin samples were immediately frozen in liquid nitrogen and stored at −80°C. In addition, 50 ml of heparinized blood was drawn from either psoriatic patients (n = 15) in lesional phases of the disease or healthy donors (n = 3), and PBMC were prepared using standard protocols. The psoriatic patients used in this study had been untreated for at least 3 wk. This study was approved by the local ethics committees.

Quantitative real-time PCR (TaqMan) analysis of MIP-3α/CCL20 and CCR6 mRNA expression

Skin biopsies were homogenized in liquid nitrogen using a Mikro-Dismembrator U (Braun Biotech, San Diego, CA), and RNA was extracted with RNAzol according to the manufacturer’s protocol (Tel-Test, Friendsville, TX). Four micrograms of RNA was treated with DNase I (Roche, Mannheim, Germany) and reverse transcribed with oligo(dT)_14–18 (Life Technologies, Gaithersburg, MD) and random hexamer primers (Promega, Madison, WI) using standard protocols. cDNA was diluted to a final concentration of 5 ng/μl. Ten microliters of cDNA was amplified in the presence of 12.5 μl of TaqMan universal master mix (Perkin-Elmer, Foster City, CA), 0.625 μl of gene-specific TaqMan probe, 0.5 μl of gene-specific forward and reverse primers, and 0.5 μl of water. As an internal positive control, 0.125 μl of 18S RNA-specific TaqMan probe and 0.125 μl of 18S RNA-specific forward and reverse primers were added to each reaction. Specific primers and probes for CCL20, CCR6, and the other chemokine receptors measured were obtained from Perkin-Elmer. Gene-specific probes used FAM as reporter, whereas probes for the internal positive control (18S RNA) were associated with either the JOE or VIC reporter. Samples underwent the following stages: stage 1, 50°C for 2 min; stage 2, 95°C for 10 min; and stage 3, 95°C for 15 s followed by 60°C for 1 min. Stage 3 was repeated 40 times. Gene-specific PCR products were measured by means of an ABI PRISM 7700 Sequence Detection System (Perkin-Elmer) continuously during 40 cycles. The specificity of primer probe combination was confirmed in cross-reactivity studies performed against plasmids of all known chemokine receptors (CCR1-CCR9, CXCR1-CXCR5, XCR1, CX3CR1) and the following panel of chemokines: MIP-1α/CCL3, MIP-1β/CCL4, MIP-1δ/CCL15, MIP-3β/CCL19, 6Ckine/CCL21, IP-10/CXCL10, MIG/CXCL9, I-309/CCL1, I-TAC/CXCL11, HCC-1/CCL14, HCC-4/CCL16, GRO-α/β/CXCL1/2, ENA78/CXCL5, eotaxin/CCL11, eotaxin-2/CCL24, DC-CK1/CCL18, BCA-1/CXCL13, fractalkine/CX3CL1, stromal-derived factor (SDF)-1α/CXCL12, RANTES/CCL5, PF4/CXCL4, macrophage-derived chemokine/CCL22, lymphotactin/XCL1, IL-8/CXCL8, thymus and activation-regulated chemokine/CCL17, thymus-expressed chemokine/CCL25, MCP-1/CCL2, MCP-2/CCL8, MCP-3/CCL7, and MCP-4/CCL13. Target gene expression was normalized between different samples based on the values of the expression of the internal positive control. Human cDNA libraries used in this study were generated as described previously (17, 23, 24).

Cell culture

Human primary epidermal keratinocytes, dermal fibroblasts, melanocytes and dermal microvascular endothelial cells were purchased from Clonetics (San Diego, CA) and cultured in keratinocyte, fibroblast, melanocyte, or endothelial cell growth medium (Clonetics) as previously described (25). Cells were treated with TNF-α (10 ng/ml)/IL-1β (5 ng/ml), IFN-γ (20 ng/ml), IL-4 (50 ng/ml), and IL-17 (100 ng/ml; R&D Systems, Minneapolis, MN) or were left untreated. The epidermal γδ T cell line, 7-17, was provided by Richard Boismenu (The Scripps Institute, La Jolla, CA) and cultured as described previously (26). Epidermal γδ T cells were cultured with Con A, TNF-α (10 ng/ml)/IL-1β (5 ng/ml), or medium alone. Supernatants as well as cells were harvested after 6 or 18 h. Generation of dendritic cells either from cord blood CD34⁺ hemopoietic progenitor cells or from peripheral blood monocytes was performed as described previously (19). Immature dendritic cells from CD34⁺ hemopoietic progenitor cells or monocyte-derived dendritic cells were activated for 3–24 h in the presence of CD40 ligand (CD40L)-transfected L cells (one per five dendritic cells) as described previously (27). Cells and supernatants were harvested 3, 12, 24, and 48 h after CD40L stimulation. RNA was extracted from cells as described above.

Flow cytometry and cell sorting

To analyze chemokine receptor expression of skin-homing T cell subsets, CLA⁺/CD45RO⁺/CD4⁺ cells were sorted from PBMCs isolated from two different donor pools comprising buffy coats (70 ml) from three individual donors, each using a FACS Vantage (Becton Dickinson, San Jose, CA) and the following mAbs: FITC-conjugated anti-CLA (HECA4522) mAb, PE-conjugated anti-CD45RO (UCHL1), and APC-conjugated anti-CD4 (RPA-T4; PharMingen, San Diego, CA). The purity of the cells was detected as >99.5%. Subsequently, RNA was extracted and reverse transcribed as described above.

In separate experiments the expression of CCR6 was analyzed on memory T cell subsets using the following Abs: FITC-conjugated anti-CLA (HECA4522) mAb (PharMingen), FITC-conjugated anti-CD45RO (UCHL1) mAb (PharMingen), APC-conjugated anti-CD8 (RPA-T8) mAb (PharMingen), Cy-Chrome-conjugated anti-CD4 (RPA-T4) mAb (PharMingen), anti-CCR6-PE conjugated (53103.111) mAb (R&D Systems), and mouse IgG2b-PE-conjugated (R&D Systems). Briefly, 10⁶ PBMCs were stained with anti-CD4, anti-CD8, anti-CLA, or anti-CCR6 mAb or isotype and analyzed using a FACSCalibur and CellQuest software (Becton Dickinson).

Chemotaxis

PBMCs from either normal or psoriatic donors were incubated for 1 h at 37°C. Nonadherent cells were harvested and used for Transwell chemotaxis assays. Enriched lymphocytes in DMEM (pH 6.9)/1% BSA were added to the top chamber of 3-μm pore size polycarbonate Transwell culture insert (Costar, Cambridge, MA) and incubated with the indicated concentrations of recombinant hMIP-3α/CCL20 (R&D Systems) in the bottom chamber for 3 h. The number of migrating cells was determined by flow cytometry with anti-CLA (HECA-452) and anti-CD3 mAbs (PharMingen). A known number of 15-mm microsphere beads (Bangs Laboratories, Fishers, IN) was added to each sample before analysis to determine the absolute number of migrating cells. Chemotaxis indices were calculated as the ratio of the number of cells migrating toward CCL20 divided by the number of migrating cells in the negative control.

In situ hybridization

In situ hybridization was performed as previously described (19). Coupled primers were used for amplifying by RT-PCR the majority of the open reading frame of the CCL20 gene. +77/MIP-3α/CCL20 (5′-TTGCTCCTGGCTGCTTTG-3′, forward primer) and −425/MIP-3α/CCL20 (5′-ACCCTCCATGATGATGTGCAAG-3′) were used with an annealing temperature of 62°C. Then, PCR products were cloned into pCRII TA cloning vector (Invitogen, Leek, The Netherlands), RNA probes were synthesized using Sp6 and T7 RNA polymerases (Roche) and radiolabeled with [³⁵S]UTP (Amersham, Aylesbury, U.K.). Sense and antisense ³⁵S-labeled probes of CCL20 were obtained by run-off transcription of the 367-bp fragment and then partially degraded by alkaline hydrolysis for 20 min at 60°C. Six-micron cryostat sections were prepared on charged electrostatic slides (SuperFrost/Plus, Polylabo, Strasbourg, France) and fixed with cold acetone for 20 min and with 4% paraformaldehyde for 20 min, followed by 0.1 M triethanolamine/0.25% acetic anhydride. The sections were hybridized overnight at 50°C (2–3 × 10⁶ cpm/slide), RNase A treated, and exposed for 40 days. After development, the sections were stained with hematoxylin.

Immunohistochemistry

Frozen 6-μm tissue sections were fixed in acetone and paraformaldehyde before the immunostaining. To block the nonspecific binding of avidin, biotin system components, or endogenous peroxidase activity, sections were pretreated with avidin D and biotin solutions (Blocking kit, Vector, Biosys, Compiegne, France) for 10 min each step and with PBS containing 0.3% hydrogen peroxide (Sigma, St. Louis, MO) for 15 min at room temperature. After brief washing in PBS, the sections were incubated with blocking serum (2% normal rabbit serum) for at least 30 min before adding both primary Abs. Sections were double stained simultaneously with anti-hMIP-3α/CCL20 goat polyclonal Ab (IgG isotype, R&D Systems) and anti-hCD3 mouse mAb (Leu 4, IgG1 isotype, Becton Dickinson, Mountain View, CA) for 1 h at room temperature in a humid atmosphere. The binding of goat IgG was detected using biotinylated rabbit anti-goat IgG followed by streptavidin-peroxidase (both included in the Vectastain ABC kit: goat IgG PK-4005, Vector), and the binding of mouse IgG1 was detected by rabbit alkaline phosphatase-labeled anti-mouse IgG (D0314, Dako, Glostrup, Denmark) at the same time at room temperature in a humid atmosphere. The peroxidase and alkaline phosphatase activities were revealed using 3-amino-9-ethylcarbazole substrate (SK-4200, Vector) and alkaline phosphatase substrate III (SK-5300, Vector) for 5–10 min at room temperature, respectively. Negative controls were established by adding nonspecific isotype controls as primary Abs.

Generation of mouse mAbs against hMIP-3α/CCL20 and development of an hMIP-3α/CCL20 ELISA

Inbred BALB/c mice were immunized with three successive i.p. injections of CFA (Sigma, St. Louis, MO) or IFA or without Freund’s adjuvant, respectively, with 50 ng of purified hCCL20 obtained from supernatants of hCCL20 transient-transfected COP5 cells. Spleens were removed for fusion 3 days after a final i.v. injection of hCCL20. Hybridization was conducted using the nonsecreting myeloma cell line SP2/0-Ag8 with polyethylene glycol 1000 (Sigma). Human CCL20 transient-transfected COP5 cells were cultured for 2 days in 96-well plates and fixed in acetone. Then, hybridoma supernatants were harvested after 6 days and incubated for 30 min on fixed hCCL20 transient-transfected COP5 cells. Ab binding was then revealed with peroxidase-conjugated sheep anti-mouse IgG (Biosys) at a 1/200 dilution in PBS for 30 min at 37°C. Positive hybridomas were cloned by limiting dilution and expanded using a high density culture system (Integra cell line CL1000, Integra Biosciences, Paris, France). After sodium sulfate precipitation, the mAbs were purified by anion exchange chromatography on a HyperD column and peroxidase-labeled (Sepracor, Villeneuve, France). An ELISA was set up using one of the hCCL20 mAbs, 319F6, as a capture mAb and a peroxidase-coupled mouse anti-hCCL20 mAb to reveal the captured hCCL20. The assay proved to be specific for hCCL20 with a sensitivity of 0.2 ng/ml.

Analysis of hMIP-3α/CCL20 bioactivity by calcium mobilization assay

A cell line expressing the human CCR6 chemokine receptor was provided by Chuan Chu Chou (SPRI, Kennilworth, NJ). Briefly, the CCR6 open reading frame was cloned into the pME18sneo eukaryotic expression vector and transfected into the murine B cell line, BAF/3, by electroporation. Stable transfectants were isolated by selection in medium containing 1 mg/ml G418. CCR6 expression was confirmed using calcium signaling, ligand binding analysis with recombinant human CCL20 (R&D Systems), and immunohistochemistry with anti-CCR6 (53103.111) mAb (R&D Systems). The average number of binding sites per cell was 220,000. To measure the biological activity of the CCL20 produced by keratinocytes, fibroblasts, or endothelial cells, supernatants from these cell cultures were concentrated 20-fold using Centriplus concentrators with a cutoff of 3 kDa (Amicon, Beverly, MA). The calcium response to supernatants from these resting or activated cells was measured using standard protocols (18). Briefly, the BAF/3 parental and CCR6 transfectant were loaded for 60 min at 37°C with 3 μM Indo-1A (Molecular Probes, Eugene, OR). Cells were washed and resuspended in HBSS (Life Technologies/BRL, Grand Island, NY) to a final concentration of 10⁷ cells/ml. Calcium mobilization was measured using a spectrophotometer (Photon Technology International, South Brunswick, NJ) with excitation at 350 nm and dual simultaneous recording of fluorescence emission at 400 and 490 nm. Relative intracellular calcium levels are expressed as the 400 nm/490 nm emission ratio. Experiments were performed at 37°C with constant mixing in cuvettes containing 10⁶ cells in 2 ml of HBSS with 1 mM CaCl_2. To demonstrate the specificity of CCL20-induced calcium mobilization, neutralization studies were performed using a blocking mouse anti-human CCL20 (IgG1; 67310.111) mAb (R&D Systems) or isotype control (IgG1; Sigma).

CCL20 and its receptor CCR6 are significantly up-regulated in lesional psoriatic vs nonlesional or normal skin

After cloning and initial characterization of mouse and human CCL20 and identification of its receptor, CCR6, we were interested in investigating the potential role of this ligand-receptor pair in human diseases (17, 18). To this end, we undertook a systematic screening of human tissue cDNA libraries with CCL20- and CCR6-specific TaqMan probe and primers. Our cDNA library panel included various libraries derived from human autoimmune disease samples. This initial screening showed that CCL20 was expressed >100 times higher in a cDNA library derived from lesional psoriatic skin (568,000 fg/50 ng of cDNA) compared with normal skin (5,530 fg/50 ng of cDNA). Moreover, TaqMan analyses showed abundant CCR6 message (864 fg/50 ng of cDNA) in the cDNA library generated from psoriatic skin. In contrast, CCR6 was undetectable in a cDNA library derived from normal skin. We then sought to analyze other chemokine receptors that have been reported to be up-regulated in psoriasis in our cDNA libraries from normal and psoriatic skin. To this end, we studied the expression of the IL-8R and confirmed previous reports (28, 29). CXCR1 and CXCR2 were constitutively expressed in the cDNA library derived from normal skin and markedly up-regulated in the psoriatic skin cDNA library (data not shown). Thus, our cDNA libraries derived from normal or psoriatic skin provided representative tools to study gene expression. These initial observations prompted us to investigate a possible role for CCL20 and CCR6 in the pathogenesis of psoriasis in more detail. We initially sought to validate these observations in more patient samples.

Consecutive quantitative real-time PCR analyses of individual cDNAs derived from lesional (n = 10), nonlesional psoriatic (n = 5), or normal (n = 5) skin confirmed that both CCL20 and CCR6 were significantly up-regulated in lesional psoriatic vs nonlesional or normal skin (p < 0.005; Fig. 1⇓, A and B). An average 7- and 4-fold induction of CCL20 and CCR6 could be detected, respectively. The differences in CCL20 expression levels probably reflect interindividual differences of patients or various clinical stages of the disease. We did observe, however, that differences in CCL20 expression generally correlated with disease severity. Notably, although CCL20 expression was variable within lesional psoriatic skin, direct comparison with nonlesional psoriatic skin consistently showed that CCL20 mRNA was up-regulated in inflamed skin (Fig. 1⇓A).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Quantitative TaqMan PCR analysis of MIP-3α/CCL20 and CCR6 expression in psoriasis. (A and B) Pattern of CCR6 and CCL20 expression in normal (n = 5) and nonlesional (n = 5) psoriatic skin vs lesional (n = 10) psoriatic skin. Values are expressed as femtograms of target gene in 50 ng of total cDNA. Measurements of individual samples and the mean ± SD are shown. A Student t test was performed. ∗, p < 0.005. C, Level of CCR6 expression in PBMCs derived from either psoriatic patients in lesional phases of the disease (n = 10) or healthy donors (n = 5). Values are expressed as femtograms of target gene in 50 ng of total cDNA (mean ± SD). Student t test was performed. ∗∗, p < 0.001.

To look more precisely at the distribution of CCL20, we performed in situ hybridization (Fig. 2⇓) and immunohistochemistry (Fig. 3⇓) using specific probes and polyclonal Abs directed against CCL20. In situ hybridization with a CCL20-specific probe showed marked expression of this chemokine in the basal as well as suprabasal layers of the epidermis of lesional psoriatic skin (Fig. 2⇓, A–C). Two different expression pattern of CCL20 mRNA could be observed. First, in more chronic lesions CCL20 mRNA expression was predominantly located within the stratum granulosum of the epidermis (Fig. 2⇓, A and C). In contrast, more active and severe psoriatic lesions showed CCL20-specific hybridization signals within the stratum spinosum and partly in the stratum basale of the epidermis (Fig. 2⇓B). Sense (vs antisense) controls (Fig. 2⇓D) as well as in situ hybridization with normal (Fig. 2⇓F) and nonlesional psoriatic (Fig. 2⇓E) skin confirmed the specific detection of CCL20 hybridization in lesional psoriatic skin.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Expression of MIP-3α/CCL20 mRNA within the epidermis of lesional psoriatic skin. Detection of CCL20 by in situ hybridization. In lesional psoriatic skin of two different donors (A and B), suprabasal layers of the epidermis yielded a strong and specific signal for CCL20 in a multifocal hybridization pattern (A and B, darkfield illumination; C, corresponding lightfield illumination of donor 1). Control studies performed with sense riboprobes were uniformly negative (D). Furthermore, nonlesional psoriatic (E) as well as skin normal (F) showed no hybridization with CCL20-specific riboprobes. E, epidermis; D, dermis. Original magnification: A, B, C, and F, ×100; D and E, ×200).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Immunohistochemistry of MIP-3α/CCL20 expression within the epidermis of lesional psoriatic skin. In agreement with results using in situ hybridization, anti-hCCL20-specific (red) polyclonal Abs showed a marked and specific staining within the basal and suprabasal layers of lesional psoriatic skin of three different donors (A, B, and H). Focal accumulation of T cells (anti-CD3, blue) in the papillary dermis of lesional psoriatic skin was adjacent to foci of CCL20-expressing epidermal cells (red; B, D, and F). D and F, Higher magnifications of the inset area in B. Moreover, CCL20-expressing keratinocytes colocalize with CD3⁺ intraepidermal T lymphocytes in lesional psoriatic skin (D and F). In contrast, nonlesional psoriatic skin (C) as well as normal skin (E) showed no CCL20-specific staining. Staining with isotype controls (G) demonstrated the specificity of CCL20 staining. E, epidermis; D, dermis; P, papillary dermis. Original magnification: A, ×200; B, ×100; C, ×200; D, ×400; E, ×200; F–H, ×1000.

Furthermore, immunohistochemical staining of lesional psoriatic skin confirmed the focal up-regulation of CCL20 protein within basal and suprabasal layers of the epidermis (Fig. 3⇑, A, B, D, F, and H). In contrast, nonlesional psoriatic (Fig. 3⇑C) and normal skin (Fig. 3⇑E) showed no specific staining for CCL20. Staining with isotype controls confirmed the specificity of CCL20 detection (Fig. 3⇑G). Importantly, there was a very good correlation between the detection of CCL20 mRNA by in situ hybridization and CCL20 protein by immunohistochemistry (see, for example, Fig. 2⇑, A and B, and Fig. 3⇑, A and B), confirming the two different expression patterns and the specificity of CCL20 detection in lesional psoriatic skin. Furthermore, CCL20 staining could be completely blocked by preincubation of anti-CCL20 Abs with recombinant CCL20 (data not shown).

Attempts to detect CCR6 by immunohistochemistry for CCR6 were inconclusive due to the low sensitivity of the Ab. Therefore, we performed double stainings for CCL20 and CD3 to study localization of CCL20-expressing cells and T cells (Fig. 3⇑, B, D, and F). Immunohistochemistry of lesional psoriatic skin showed that the focal accumulation of T cells in the papillary dermis of lesional psoriatic skin was directly adjacent to foci of CCL20-expressing epidermal cells (Fig. 3⇑, B and D). Furthermore, CCL20-expressing keratinocytes within the epidermis colocalized with intraepidermal CD3⁺ T cells (Fig. 3⇑, D and F). Moreover, Fig. 3⇑H confirms that CCL20 is strongly expressed within the cytoplasma of keratinocytes of lesional psoriatic skin.

Skin homing CLA⁺ T cells of psoriatic patients respond to lower concentrations of CCL20 and show higher chemotactic responses compared with those of normal donors

Clinically, it is well known that infections may trigger psoriatic episodes, and recently, it has been suggested that superantigens may play a role in T cell activation during the pathogenesis of psoriasis (30, 31, 32). However, very little is known about chemokine receptor expression on PBMCs of normal healthy donors vs psoriatic patients. Interestingly, CCR6 was also significantly up-regulated in PBMCs derived from psoriasis patients (n = 10) vs PBMC from healthy donors (n = 5; p < 0.001). PBMCs from psoriatic patients expressed, on the average, 4-fold higher levels of CCR6 mRNA compared with PBMCs from healthy donors (Fig. 1⇑C). The CLA⁺ T cell subset represents a skin-associated population of memory T cells that migrates preferentially to normal and chronically inflamed cutaneous sites (33).

In subsequent experiments we focused on the chemokine receptor profile of pathologically relevant skin-homing CLA⁺ memory T cells. Flow cytometric analyses revealed that CCR6 was expressed at high levels on the surface of skin-homing CLA⁺ T cells of normal donors (Fig. 4⇓). Moreover, CCR6 was predominantly expressed on the CD4⁺ subset of CLA⁺ T cells (Fig. 4⇓, C and D). The latter observation may account for the therapeutic effect of anti-hCD4 Abs in the treatment of psoriasis (34, 35, 36).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

CCR6 expression on skin-homing CLA⁺ T cells. A, CCR6 is highly expressed on skin-homing CLA⁺ memory T cells. PBMCs were stained with mAb directed against CD4, CD8, CLA, and CCR6. Percentages indicate the relative number of double-positive cells analyzed in total PBMCs. B, CD4 and CD8 gates. C and D, CCR6 is exclusively expressed by the CD4 subset of skin-homing CLA⁺ memory T cells. Dot plots show the expression of CCR6 on gated CD4⁺ or CD8⁺ skin-homing T cells. Percentages indicate the relative number of CLA⁺/CCR6⁺ cells within the CD4 or CD8 subpopulations of PBMCs. E, Chemokine receptor expression in sorted CLA⁺/CD4⁺/CD45RO⁺ memory T cells. Sorted cells from two different donor pools comprising cells from three different donors in each pool were analyzed for chemokine receptors (CCR2, CCR3, CCR5, CXCR1, CXCR2, CXCR3) that are either known or are likely to be involved in skin inflammation and were compared with the expression of CCR6. Values are expressed as femtograms of target gene in 50 ng of total cDNA.

In agreement with flow cytometric analyses, TaqMan analyses on sorted CLA⁺ memory T cells from normal donors also showed CCR6 expression at high levels. In fact, the expression of CCR6 was the highest detected among chemokine receptors, especially compared with receptors of other chemokines (IL-8/CXCL8, GRO-α/CXCL1, IP-10/CXCL10, Mig/CXCL9, MCP-1/CCL2, RANTES/CCL5) that have been reported to be associated with psoriasis (29, 37, 38, 39, 40, 41). CCR6 expression was 100 to >1000 times higher than that of CXCR1, CXCR2, CXCR3, CCR2, CCR3, and CCR5 on this skin-homing subset of memory T cells (Fig. 4⇑E).

The next question we addressed was whether differences of CCR6 mRNA expression in psoriatic vs normal PBMCs could be confirmed at the protein level and further characterized using flow cytometric analyses of T lymphocyte subpopulations (Table I⇓). Within the T cell compartment, CCR6 is preferentially expressed on CD4⁺ T cells in both normal donors (n = 5) and psoriatic patients (n = 3). In total CD4⁺ T cells CCR6 expression was lower on psoriatic T cells compared with those from normal individuals. However, within the subpopulation of skin-homing CLA⁺ T cells, CCR6 expression was higher on psoriatic than on normal lymphocytes. Furthermore, we noticed that more CLA⁺ T cells could be detected within PBMCs of psoratic vs normal donors (Table I⇓). This observation parallels recent findings of Davison et al. (42).

The latter observations had functional implications, as psoriatic CLA⁺ T cells responded in Transwell chemotaxis assays to lower concentrations of CCL20 and showed higher chemotactic responses than CLA⁺ T cells from normal donors (Fig. 5⇓). CCL20 significantly attracted CLA⁺ skin-homing T cells from both normal and psoriatic donors. Overall, 2–21% of the starting population of CD3⁺/CLA⁺ lymphocytes migrated toward CCL20.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

MIP-3α/CCL20-induced chemotaxis of skin-homing CLA⁺ T cells in normal and psoriatic donors. Enriched lymphocytes from either normal (A) or psoriatic donors (B) were analyzed in Transwell chemotaxis assays for their response to hCCL20. The number of CD3⁺/CLA⁺ migrating cells was determined, and chemotaxis indices were calculated from triplicate measurements. Representative data for one of three normal or psoriatic donors are shown.

TNF-α, IL-1β, IFN-γ, IL-17, and CD40L regulate the expression of CCL20 in cellular constituents of the skin

The pattern of CCL20 expression within the epidermis suggested that keratinocytes may be a major source of CCL20 in the skin. To further investigate the cellular origin of CCL20 within the skin and to gain insights into its regulation, we cultured human primary keratinocytes, melanocytes, and dermal fibroblasts with TNF-α/IL-1β, IFN-γ, IL-4, IL-17, or medium alone as a control. Furthermore, we determined whether cultured epidermal γδ T cells, resting or stimulated with either TNF-α/IL-1β or Con A, may express CCL20. Human primary dermal microvascular endothelial cells were also cultured in the presence or the absence of TNF-α/IL-1β. We used TNF-α and IL-1β stimulation, since these proinflammatory cytokines are known to be up-regulated during inflammatory conditions and in lesional psoriatic skin (43, 44, 45, 46, 47). Furthermore, we were interested in the effects of Th cell-derived cytokines such as IL-4, IFN-γ, and IL-17 on cellular constituents of the skin. TaqMan analyses showed that only keratinocytes and dermal microvascular endothelial cells constitutively express low levels of CCL20, and that TNF-α/IL-1β can induce CCL20 mRNA expression in both keratinocytes and dermal microvascular endothelial cells (Fig. 6⇓A). In addition, TNF-α/IL-1β stimulation of dermal microvascular endothelial cells induced strong up-regulation of CCL20 expression (Fig. 6⇓A). Activation of both CD34⁺ hemopoietic progenitor cell- or monocyte-derived dendritic cells with CD40L also induced CCL20-specific transcripts (Fig. 6⇓A). Interestingly, primary melanocytes also showed significant expression of CCL20 following TNF-α/IL-1β stimulation. Moreover, CCL20 expression could be markedly induced in these cells by IFN-γ or IL-4 stimulation (Fig. 4⇑A). Keratinocytes showed a weak (2- to 4-fold) up-regulation of CCL20 mRNA after activation with either IFN-γ or IL-4 (Fig. 4⇑A). In contrast, resting or stimulated epidermal γδ T cells failed to express CCL20 mRNA (Fig. 6⇓) under any conditions. Similar results were observed in keratinocytes (n = 4), melanocytes (n = 2), CD34⁺ hemopoietic progenitor cell-derived dendritic cells (n = 2), monocyte-derived dendritic cells (n = 2), dermal microvascular endothelial cells (n = 2), and dermal fibroblasts (n = 2) of different donors. Interestingly, keratinocytes, fibroblasts, melanocytes, or epidermal γδ T cells were never observed to express significant levels of CCR6 mRNA (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

A, MIP-3α/CCL20 mRNA expression in cellular constituents of the skin. Quantitative TaqMan PCR analysis of CCL20 expression in cDNA obtained from cultured human primary keratinocytes, dermal fibroblasts, melanocytes, dermal microvascular endothelial cells, and epidermal γδ T cells (7 ,8 ,9 ,10 ,11 ,12 ,13 ,14 ,15 ,16 ,17 ) treated with medium alone or with TNF-α/IL-1β, IFN-γ, or IL-4 for 6 or 18 h. Analysis of MIP-3α/CCL20 expression in dendritic cells derived from CD34⁺ hemopoietic progenitor cells or monocytes that were cocultured with or without CD40L. Values are expressed as femtograms of target gene in 50 ng of total cDNA. Representative data from single donors are shown. B, Human primary keratinocytes produce biologically active CCL20. Concentrated supernatant harvested from resting and stimulated keratinocytes (KC) as well as the medium controls were assayed for intracellular calcium mobilization in CCR6-transfected BAF/3 cells. The parental BAF/3 cells showed no responses to any of the supernatants; however, calcium mobilization could be induced by hSDF-1α because BAF/3 cells express endogenous CXCR4. Despite the desensitization of endogenous CXCR4, supernatants from keratinocytes activated with TNF-α/IL-1β could induce marked Ca²⁺ mobilization responses in CCR6 transfectants. Treatment with anti-MIP-3-α/CCL20 mAb completely neutralized supernatant-induced calcium mobilization in CCR6-transfected BAF/3 cells; however, the isotype control showed no effect. Similar results were obtained for dermal fibroblasts and dermal microvascular endothelial cells (data not shown). The addition of concentrated culture medium (with or without cytokines: TNF-α/IL-1β, IL-4, IFN-γ) did not produce any intracellular Ca²⁺ mobilization in parental or CCR6-transfected cells.

Analyses of CCL20 protein expression by ELISA confirmed that activated keratinocytes, dermal microvascular endothelial cells, dermal fibroblasts, and monocyte-derived dendritic cells are potent producers of this chemokine. Supernatants from cultured primary human keratinocytes, dermal microvascular endothelial cells, and dermal fibroblasts activated with TNF-α and IL-1β showed a marked induction of CCL20 protein, while resting cells showed little or no production of CCL20 (Table II⇓). These levels of hCCL20 production are in the range of biological activity reported for prokaryote-derived recombinant hCCL20. Low levels of CCL20 protein were also detected after stimulation of keratinocytes with IFN-γ or IL-17; however, additional TNF-α stimulation showed synergistic effects and markedly enhanced CCL20 protein production (Table II⇓).

Supernatants of monocyte-derived dendritic cells showed significant production of CCL20 protein (1.14–8.76 ng/ml) 12–48 h following CD40L stimulation (Table II⇑). CCL20 protein expression generally confirmed the data obtained using real-time quantitative PCR (TaqMan). However, supernatants from CD34⁺ hemopoietic progenitor cells stimulated with CD40L did not contain detectable CCL20 protein despite a marked induction of CCL20 mRNA expression. This may be due to the high level of CCR6 expression by these cells, suggesting that they may be binding and internalizing CCL20.

We then sought to determine whether the CCL20 protein detected in these supernatants was biologically active. To this end, we tested supernatants of keratinocytes, fibroblasts, or endothelial cells, either resting or following stimulation with TNF-α/IL-1β, IFN-γ, or IL-4, in a calcium signaling assay using CCR6-transfected BAF/3 cells. These cells are known to express endogenous CXCR4. Therefore, to obtain a CCR6-specific assay we blocked the endogenous CXCR4 with human SDF-1α before testing the supernatants for CCL20 activity (Fig. 4⇑B). In agreement with the ELISA data, supernatants from keratinocytes (Fig. 4⇑B), dermal fibroblasts (data not shown), and dermal microvascular endothelial cells (data not shown) stimulated with TNF-α/IL-1β induced significant calcium mobilization responses in CCR6-transfected BAF/3 cells, but not in the parental untransfected BAF/3 cells (Fig. 6⇑B). However, the parental BAF/3 cell line showed the expected calcium mobilization response due to the triggering of endogenous CXCR4 by SDF-1α/CXCL12 (Fig. 6⇑B). Furthermore, treatment with anti-CCL20 mAb completely neutralized supernatant-induced calcium mobilization in CCR6-transfected BAF/3 cells; however, isotype control showed no effect (Fig. 6⇑B). Concentrated medium with or without cytokine (TNF-α/IL-1β, IL-4, IFN-γ) addition did not produce any intracellular Ca²⁺ mobilization in parental or CCR6-transfected cells (data not shown). These results confirm the production of bioactive CCL20 protein by keratinocytes, dermal fibroblasts, and dermal microvascular endothelial cells initially detected at the mRNA level by quantitative PCR (Fig. 6⇑A).