Dendritic cells activated by interferon-γ/STAT1 express interleukin-31 receptor and release pro-inflammatory mediators upon interleukin-31 treatment

Abstract

Interleukin 31 (IL-31) is a T cell-derived cytokine that signals via a hetero-dimeric receptor composed of IL-31 receptor alpha (IL-31RA) and oncostatin M receptor beta (OSMRB). Although several studies have aimed to investigate IL-31-mediated effects, the biological functions of this cytokine are currently not well understood. IL-31 expression correlates with the expression of IL-4 and IL-13 and is associated with atopic dermatitis in humans, indicating that IL-31 is involved in Th2-mediated skin-inflammation.

Since dendritic cells are the main activators of Th cell responses, we posed the question of whether dendritic cells express the IL-31 receptor complex and govern immune responses triggered by IL-31. In the present study, we report that primary human CD1c⁺ as well as monocyte-derived dendritic cells significantly up-regulate the IL-31RA receptor chain upon stimulation with interferon gamma (IFN-γ). Electrophoretic mobility shift assays, ChIP assays and siRNA-based silencing assays revealed that STAT1 is the main transcription factor involved in IFN-γ-dependent IL-31RA expression. Subsequent IL-31 stimulation resulted in a dose-dependent release of pro-inflammatory mediators, including TNF-α, IL-6, CXCL8, CCL2, CCL5 and CCL22. Since these cytokines are crucially involved in skin inflammation, we hypothesize that IL-31-specific activation of dendritic cells may be part of a positive feedback loop driving the progression of inflammatory skin diseases.

INTRODUCTION

Interleukin-31 (IL-31) is a four-helix bundle cytokine closely related to the IL-6 cytokine family (1). The main sources of IL-31 are T cells, in particular activated CD4⁺ T cells such as skin-infiltrating cutaneous lymphocyte antigen (CLA)⁺ memory T cells (1-3) and T cells activated under T helper (Th) 2 skewing conditions (1).

A very recent study suggests that mast cells are an additional source of IL-31: human mast cells secrete IL-31 and other pruritogenic mediators upon stimulation with human β-defensins and cathelicidin (4).

IL-31 signals via a heterodimeric receptor complex composed of IL-31 receptor alpha (IL-31RA), previously termed gp130-like monocyte receptor (GLM-R) (5) or gp130-like receptor (GPL) (6-7), and oncostatin M receptor beta (OSMRB) (1, 8). Several in vitro studies investigating signaling pathways downstream of the receptor demonstrated activation of p38MAPK, ERK1/2 and JNK1/2, and phosphorylation of the STAT family members STAT1, STAT3 and STAT5 (1, 5, 7, 9-10).

At present, little is known about the biological consequences of IL-31 signaling. Transgenic mice overexpressing IL-31 develop a severe skin phenotype closely resembling the skin from patients with atopic dermatitis (AD) (1, 11). Analysis of skin biopsies from patients with different types of inflammatory skin diseases showed that IL-31 is overexpressed predominantly in pruritic forms of skin inflammation (12). Moreover, leukocytes from patients with AD or allergic contact dermatitis (ACD) show significantly enhanced IL-31 expression which is associated with elevated expression of IL-4 and IL-13 as compared to healthy volunteers (3, 12). On the basis of these findings, IL-31 has recently begun to be looked at as a possible mediator in the pathogenesis of Th2 cytokine-mediated inflammatory skin diseases like AD and ACD. In contrast, other published studies support a role for IL-31-induced signaling in limiting the severity of Th2-mediated inflammation in the lung and gut (13-14). For example, IL-31RA-deficient mice injected with Schistosoma mansoni eggs developed a more severe pulmonary Th2 inflammation than did wild-type (WT) animals. The results of a similar study likewise suggest a regulatory role for IL-31/IL-31R interactions in the intestine following infection with the gastrointestinal helminth Trichuris muris (14). Taken together, these data indicate that IL-31/IL-31R interactions may play an important role in limiting Th2-mediated inflammatory responses in the lung and the intestine, whereas in the skin, IL-31 action is positively correlated with inflammation.

Dendritic cells (DCs) are highly specialized, professional antigen-presenting cells (APCs) equipped with a plethora of intracellular and extracellular pattern recognition receptors. Upon activation by various stimuli, DCs undergo a maturation process that endows them with distinct DC functions, including capture and presentation of antigen, migration, co-stimulation, and the release of T cell-polarizing cytokines (15).

Since DCs are the main orchestrators of T-cell responses, we sought to address a potential contribution of DCs to the complex picture of IL-31-mediated effects. As none of the studies analyzing IL-31 receptor expression has reported on its expression on DCs, IL-31-mediated effects on DCs have remained speculative so far. Here we show that IL-31RA surface expression on DCs is enhanced in response to interferon gamma (IFN-γ). Moreover, we demonstrate that the transcription factor STAT1 is crucial for IFN-γ-dependent IL-31RA expression. Once DCs express the cognate receptor they become responsive to IL-31, and secrete substantial amounts of pro-inflammatory cytokines and chemokines upon IL-31 stimulation. Because the mediators released by IL-31-stimulated DCs are crucially involved in skin inflammation, we speculate that activation of DCs by IL-31 is part of a positive feedback loop driving the progression of inflammatory skin diseases.

METHODS

All studies involving human cells were conducted in accordance with the guidelines of the World Medical Association’s Declaration of Helsinki.

Generation of human monocyte-derived DCs and isolation of primary human blood DCs

Monocyte-derived DCs (moDCs) were generated according to standard protocols. Briefly, adherent monocytes were cultured in DC medium [RPMI 1640 (PAA, Pasching, Austria), 10% fetal calf serum (FCS; PAA, Pasching, Austria), 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, 50 μM β-mercaptoethanol (all Gibco Laboratories; Grand Island, NY)] supplemented with 50 ng/ml GM-CSF and 50 ng/ml IL-4 (generous gift from Novartis, Vienna, Austria) for 6 days. Primary CD1c⁺ DCs were isolated from CD19-depleted peripheral blood mononuclear cells (PBMCs) using a BDCA1⁺ kit (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer’s instructions. Isolated cells were phenotyped by FACS-analysis and cultured in RPMI 1640 (PAA, Pasching, Austria), 10% fetal calf serum (FCS; PAA, Pasching, Austria), 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin (all Gibco Laboratories; Grand Island, NY).

DC stimulation

For IL-31 receptor expression studies, 1 × 10⁵ CD1c⁺ DCs/ml and 1.5 × 10⁵ moDCs/ml, respectively, were plated in DC medium. Cells were stimulated with 0-25 ng/ml IFN-γ (ImmunoTools, Friesoythe, Germany), 100 ng/ml E. coli lipopolysaccharide (LPS) 055:B5 (Sigma-Aldrich, Vienna, Austria), 1 μg/ml staphylococcal enterotoxin B (SEB; Sigma-Aldrich), 50 ng/ml IL-4 (kind gift from Novartis, Vienna, Austria), 30 ng/ml thymic stromal lymphopoietin (TSLP) (ImmunoTools, Friesoythe, Germany), 5 μg/ml α-CD40 (kind gift from Novartis, Vienna, Austria), crosslinked via goat anti-mouse F(ab’) fragments (1 μg/ml) (Jackson ImmunoResearch Europe, UK) and IL-31 (R&D Systems, Biomedica, Vienna, Austria) at concentrations ranging from 0-250 ng/ml. For cytokine-expression studies, 1 × 10⁵ CD1c⁺ DCs/ml were plated in DC medium and pretreated with 10 ng/ml IFN-γ (ImmunoTools, Friesoythe, Germany) to induce IL-31R expression. After 48 hours of incubation, cells were harvested, washed once, re-plated in DC medium and stimulated with IL-31 for 24 hours.

Western blot and immunofluorescence

Western blot analysis was performed as described previously (16). In brief, cells were stimulated for the indicated times, harvested, and lysed in ice-cold NP-40 lysis buffer (Invitrogen, Lofer, Austria) according to the manufacturer’s instructions. Proteins were separated by SDS-PAGE (Invitrogen) and blotted onto nitrocellulose membrane (Bio-Rad, Vienna, Austria). All antibodies were purchased from Cell Signaling Technology (Danvers, MA) and used according to the manufacturer’s instructions. For flow cytometry, cells were harvested, washed, resuspended in phosphate-buffered saline (PBS; PAA, Pasching, Austria) containing 3% FCS (PAA, Pasching, Austria) and stained with the following antibodies: α-CD1a-FITC (Immunotools, Friesoythe, Germany), mouse IgG control (BD Biosciences, Erembodegen, Belgium), α-CD1c-FITC (eBioscience, Vienna, Austria), α-IL-31RA-PE, α-OSMRB-APC, goat IgG-PE and mouse IgG-APC (all from R&D Systems, Biomedica, Vienna, Austria).

RNA isolation and quantitative real-time RT-PCR

Total RNA from cells was isolated using TRIzol reagent (Invitrogen, Lofer, Austria) and reverse transcribed with RevertAid H Minus M-MulV reverse transcriptase (MBI Fermentas, St. Leon-Roth, Germany) according to the manufacturer’s instructions. Quantitative real-time RT-PCR (qRT-PCR) was carried out on a Rotorgene 3000 (Corbett Research, Mortlake, Australia) using iQ SYBR Green Supermix (Bio-Rad, Vienna, Austria) and the primers listed below. The large ribosomal protein P0 (RPLP0) was used as a reference. The specificity of the PCRs was checked by recording a melting curve for the PCR products. Relative mRNA expression levels were calculated using the formula x=2^−ΔCt, where Ct represents the threshold cycle of a given gene and ΔCt signifies the difference between the Ct values of the gene in question and the Ct value of the reference gene RPLP0. Induction ratios were calculated using the formula x=2^−ΔΔCt. ΔΔCt represents the difference between the ΔCt values of induced and uninduced samples.

Primer sequences are as follows: human IL-31RA sense 5′-GGCATGGAGATGATTTCAAGGATAAGCTAAACCTG-3′, anti-sense 5′-CTGGCTTCATCTGTGAAAATTTCTTGCAGAAC-3′, human OSMRB sense 5′-AAAAGGCATTGATTGTGGACAACCTAAAGCC-3′, anti-sense 5′-ATGTCAGGATAACAGGTCTCCTTGATCCACTGAC-3′, human IL-31RA long isoform (detects NM_139017, NM_001242636, NM_001242639) sense 5′-GGACCCGCCCAGAAGCCAA-3′, anti-sense 5′-CTTCTCCCTTGGTGTGCTCTGGA-3′ human IL-31RA short isoform (detects NM_001242637, NM_001242638) sense 5′-TTCTGTCTTCCTGCCCAACTTCAATA-3′, antisense 5′-TTGCATCCGTAGGTGGTCCATG-3′ human RPLP0 sense 5′-GGCACCATTGAAATCCTGAGTGATGTG-3′ anti-sense 5′-TTGCGGACACCCTCCAGGAAG-3′.

STAT1 siRNA knockdown

To avoid unintended side-effects (e.g. cellular stress responses like the PKR/interferon response) in the knockdown experiments, we used a chemically modified oligonucleotide employing Stealth RNAi^™ technology (Invitrogen) with the following sequence: Sense 5′-GGAUUGAAAGCAUCCUAGAACUCAU-3′, antisense 5′-AUGAGUUCUAGGAUGCUUUCAAUCC-3′.

Cells were transfected with Lipofectamine RNAiMax reagent (Invitrogen) according to the manufacturer’s guidelines. Briefly, 5 × 10⁵ cells were plated in antibiotics-free DC-medium and transfected with 100 pmol/well of siRNA. For moDC experiments, cells were transfected on day five of the differentiation period. Transfection efficiency was routinely >90%, as assessed by flow cytometry with fluorescent control RNA oligonucleotides (BlockIT, Invitrogen). Knockdown efficacy was analyzed four days post transfection by Western blotting.

Preparation of nuclear extracts and electrophoretic mobility shift assays (EMSA)

Nuclear extracts from IFN-γ-induced or non-induced moDCs were prepared according to the method of Andrews and Faller (17). Generation of double-stranded oligonucleotide probes and EMSAs were carried out as described previously (18-20). For competition assays, non-labeled oligonucleotide was added in 50-fold molar excess to the binding reaction 30 minutes prior to the radiolabeled probe. Super-shifting of bands was achieved by adding 500 ng/μl antibody (α-STAT1, α-STAT3 and α-STAT5, all purchased from Santa Cruz Biotechnology, Heidelberg, Germany) to the binding reaction. Sequences of the oligonucleotides are as follows (STAT consensus nucleotides underlined): site1: sense 5′-GTTATATAACATTTTCTTTTTCCTGGAATTAATAATTAT-3′, antisense 5′-GGCATAATTATTAATTCCAGGAAAAAGAAAATGTTATAT-3′, site 1 mut: sense 5′-GTTATATAACATTTTCTTTTATCTGGAATTAATAATTAT-3′, anitsense 5′-GGCATAATTATTAATTCCAGATAAAAGAAAATGTTATAT-3′, site 2 sense 5′-GGCCACCATTCTTGGAATTCTCTCTCTTCTATTTGAAATT-3′, antisense 5′-GCAAATTTCAAATAGAAGAGAGAGAATTCCAAGAATGGTG-3′, site 3 sense 5′-GGGTACAACATTTTTTCTTATTTTCTATGAAGGTTTAA-3′, antisense 5′-GAGTTAAACCTTCATAGAAAATAAGAAAAAATGTTGTA-3′.

Chromatin Immunoprecipitation (ChIP) assay

moDCs were generated as described above and cultured in DC medium at a density of 1.5 × 10⁵ cells/ml. In total, 5 × 10⁵ moDCs were stimulated with 10 ng/ml IFN-γ for 4 hours. ChIP was performed using the LowCell# ChIP kit (Diagenode, Liège, Belgium) according to the manufacturer’s instructions. Chromatin was fragmented by 15 minutes of sonication in a Biorupter sonicator (Diagenode, Liège, Belgium). Precipitation reactions contained chromatin from 3 × 10⁴ cells and 3μg of antibodies (α-STAT1, Santa Cruz Biotechnology, Heidelberg, Germany or rabbit negative control IgG, Diagenode, Liège, Belgium). Precipitation of the STAT1-binding motif was analyzed by means of qRT-PCR using iQ SYBR Green Supermix (Bio-Rad, Vienna, Austria) and the primers listed below.

IL-31RA STAT1 site 1 sense 5′-GCTGAGCCACATATTGGACTTTAA-3′, antisense 5′- TGGAGAAGGATGAGTAATCGGC-3′.

Cytokine detection

Cytokine secretion was measured by using commercially available ELISA kits for TNF-α, IL-6 (both PeproTech, Eubio, Vienna, Austria), IL-12p70 (BD-Pharmingen, Erembodegen, Belgium), CXCL8/IL-8, CCL2/MCP-1, CCL5/RANTES, CCL17/TARC and CCL22/MDC (all R&D Systems, Biomedica, Vienna, Austria).

Statistical analysis

Cytokine production and mRNA expression of selected genes was compared using a Student’s t-test. Values of p ≤ 0.05 were considered statistically significant (* p≤0.05, ** p≤0.01, *** p≤0.001).

RESULTS

Expression of IL-31RA in human DCs is induced by IFN-γ

Numerous studies have shown IL-31R to be expressed on keratinocytes (12), epithelial cells (10, 21) and on some types of APCs (1, 5, 21-22). However, none of those studies investigated IL-31R expression on DCs. To address this question, primary CD1c⁺ DCs were isolated from human blood samples. Since previous studies reported that IL-31RA expression is induced by various stimuli, including IFN-γ, LPS, and staphylococcal enterotoxin B (SEB) (1, 21-23), DCs were either left untreated or stimulated with IFN-γ, LPS or SEB. Additionally, one aliquot of cells was treated with IFN-γ and LPS simultaneously, since this combination was shown to be required to induce mRNA expression of both receptor chains (IL-31RA and OSMRB) in monocytes (1). Based on the finding that IL-31 is preferentially released by activated Th2 cells (1), we further treated DCs with IL-4, the Th2-associated cytokine TSLP (24) and α-CD40, which mimics DC activation by T cells via CD40/CD40L interactions (25). Compared to untreated cells, which show moderate expression of IL-31RA (13.8%) and OSMRB (5.8%), a strong increase in IL-31RA⁺ cells was observed upon IFN-γ treatment, whereas treatment with SEB, IL-4, TSLP and α-CD40 had hardly any effect or even reduced the expression of IL-31RA and OSMRB (Figure 1). Although in monocytes LPS was shown to be a potent inducer of OSMRB expression at the mRNA level (1), LPS or the combination of IFN-γ and LPS did not increase surface expression of OSMRB, but rather seemed to counter-regulate IL-31RA expression on CD1c⁺ DCs (Figure 1). These findings indicate that the expression of IL-31RA on primary human DCs is mainly stimulated by IFN-γ.

Figure 1.

Flow cytometry analysis of IL-31R expression in CD1c⁺ DCs. Cells were treated with IFN-γ, SEB, LPS, IL-4, TSLP and α-CD40 for 48 hours. Thereafter, cells were harvested, washed and stained with isotype-specific antibodies or antibodies directed against IL-31RA and OSMRB. Percentages of IL-31RA⁺ positive and IL-31RA⁺/OSMRB⁺ double positive cells are shown. Results are representative of 3 independent experiments.

Human CD1c⁺ DCs represent the major subset of myeloid DCs in human blood. However, they still are a rare cell population, comprising less than 1% of circulating PBMCs. Because studies exploring the molecular mechanisms of IFN-γ-induced IL-31RA expression require higher cell numbers, we analyzed whether IL-31 receptor expression on monocyte-derived DCs (moDCs) is similar to primary CD1c⁺ DCs. Therefore, moDCs and primary CD1c⁺ DCs were stimulated with IFN-γ for 48 hours and IL-31RA and OSMRB surface expression was analyzed by flow cytometry. Strong induction of IL-31RA expression was observed in primary as well as in moDCs (Figure 2A).

Figure 2.

Concentration-and time-dependent expression of IL-31RA. A Flow cytometry analysis of IL-31R expression on moDCs and primary CD1c⁺ DCs. Cells were treated with 10 ng/ml IFN-γ for 48 hours. Flow cytometry analysis was performed as described above. Data show representative results of one of 9 donors (CD1c⁺ DCs) and one of 5 donors (moDCs). B moDCs were stimulated with IFN-γ concentrations as indicated. 24 hours after IFN-γ treatment, IL-31RA mRNA expression was determined by qRT-PCR. Results show mean and standard deviation of two experiments. 48 hours after IFN-γ treatment, the percentage of IL-31RA⁺ positive cells was determined by flow cytometry. Data show representative results of one of two donors. C moDCs were stimulated with 10 ng/ml of IFN-γ for the indicated times. IL-31RA mRNA expression was determined by qRT-PCR. Results show mean and standard deviation of three experiments. The percentage of IL-31RA⁺ positive cells was determined by flow cytometry. Data show representative results of one of three donors.

To further optimize the conditions for IL-31RA expression we performed time kinetic- and dose-response studies. moDCs were treated with IFN-γ concentrations ranging from 0 to 25 ng/ml. IL-31RA mRNA expression and surface expression were analyzed after 24 h and 48 h, respectively. Although qRT-PCR revealed the highest expression of IL-31RA in cells treated with 25 ng/ml (Figure 2B, upper panel), IL-31RA surface expression peaked at a concentration of 9 ng/ml (Figure 2B, lower panel). For time kinetic studies, moDCs were treated with 10 ng/ml of IFN-γ for time periods ranging from 2 h to 72 h. As shown in Figure 2C (upper panel), IL-31RA mRNA peaks at 24 h while IL-31RA surface expression reaches a maximum after 72 h (Figure 2C, lower panel). Taken together, a maximum of IL-31RA expression was observed after treatment with ~10 ng/ml IFN-γ for 72 h.

IFN-γ enhances mRNA expression of short and long IL-31RA isoforms

It was reported that at least five different isoforms (isoform 1, NM_139017, isoform 2, NM_001242636, isoform 3, NM_001242637, isoform 4, NM_001242638, and isoform 5, NM_001242639) of IL-31RA exist; they are identical throughout the transmembrane domain, but clearly differ from each other with respect to the intracellular domain (1, 6). Screening of several melanoma, glioblastoma and tumor cell lines revealed that mainly two receptor isoforms are naturally expressed: the long receptor containing 745 residues and a short isoform of 560 amino acids (26). The shorter isoforms (isoforms 3 and 4) lack three tyrosine residues known to be critical for STAT activation (7). To analyze the expression of short and long isoforms, CD1c⁺ DCs from nine different human donors were cultured in the absence or presence of IFN-γ for 6 hours. qRT-PCR using isoform-specific primers and primers detecting all isoforms revealed that both short and long IL-31RA isoforms were upregulated by IFNγ. Yet, the effect of IFN-γ on the expression of short isoforms seems to be more pronounced (Figure 3).

Figure 3.

Analysis of IL-31RA isoforms expressed by primary CD1c⁺. DCs 6 hours after IFN-γ treatment, mRNA expression of IL-31RA isoforms was determined by qRT-PCR. Results for nine different donors are shown. Mean values of all donors are indicated by black bars (n.s. not significant, * p≤0.05, ***p≤0.001).

IFN-γ induces binding of STAT1 to specific motifs within the IL-31RA promoter

IFN-γ signals primarily through the JAK/STAT pathway. Activation of the IFN-γ receptor complex activates JAK1, which in turn phosphorylates critical tyrosine residues within the IFNGR1 chain thereby providing docking sites mainly for the transcription factor STAT1 but also for STAT3 and STAT5 (27-29). Phosphorylated STATs then form homodimers and enter the nucleus where they bind gamma-activation-sequences (GAS) and initiate or suppress transcription of IFN-γ-responsive genes (27-32).

To analyze STAT activation in human DCs, moDCs were generated, stimulated with 10 ng/ml IFN-γ for the indicated times and activation of STAT1, STAT3 and STAT5 was analyzed by Western blotting. Interestingly, not only STAT1 but also STAT3 and STAT5 were phosphorylated 30 minutes after IFN-γ stimulation, and the activation persisted for up to 16 hours (Figure 4).

Figure 4.

Analysis of IFN-γ-dependent STAT activation. moDCs were cultured in the absence (−) or presence (+) of IFN-γ (10 ng/ml) for the indicated times. Total protein was isolated and subjected to polyacrylamide gel electrophoresis. To analyze IFN-γ-dependent tyrosine-phosphorylation, specific antibodies directed against pSTAT1, pSTAT3 and pSTAT5 were used. Antibodies to total STATs were used as a control for equal loading. All results are representative of 5 independent experiments.

To investigate whether STATs contribute to IL-31RA expression by binding to the gene locus, we analyzed a 2050-bp fragment of the IL-31RA promoter for the presence of putative GAS motifs characterized by the sequence TTC(N)₃GAA (33-35). In silico analysis revealed the presence of three putative motifs located within positions −2047/−2041 (site 1), −1794/−1788 (site 2), and −1036/−1030 (site 3) relative to the translational start site (6) (Figure 5A). Interaction of the respective STATs with these putative GAS motifs within the IL-31RA promoter was assessed by EMSA. Incubation of nuclear extracts from IFN-γ-induced moDCs with the oligonucleotide harboring site 1 resulted in the formation of an IFN-γ-dependent nucleoprotein complex (Figure 5B). In contrast, no IFN-γ-induced complex was observed using probes containing the more proximal GAS motifs (site 2 or site 3). Addition of a 50-fold molar excess of unlabeled wild-type (WT) oligonucleotide (competitor) resulted in a loss of the IFN-γ-induced complex formation, whereas addition of a mutated oligonucleotide competitor, in which a mutation from TTC(N)₃GAA to TAT(N)₃GAA was introduced into the GAS consensus motif of site 1, did not block formation of the IFN-γ-induced complex (Figure 5C). In order to determine which of the IFN-γ-activated STATs forms a complex with site 1, specific antibodies directed against STAT1, STAT3 or STAT5 were added prior to the addition of radiolabeled probes. α-STAT1 antibodies specifically reduced the formation of the IFN-γ-induced nucleoprotein complex and led to the formation of a super-shifted complex (Figure 5C), whereas neither α-STAT3 nor α-STAT5 antibodies had an effect on the IFN-γ-induced complex. However, addition of α-STAT3 antibodies diminished a second nucleoprotein complex which appears independent of IFN-γ stimulation. These results suggest that IFN-γ exclusively induces the interaction of STAT1 and specific DNA motifs within the human IL-31RA promoter. To evaluate the capacity of STAT1 to bind the GAS consensus motif of site1 in living cells, we performed chromatin immunoprecipitation with STAT1-specific antibodies. Enrichment of specifically precipitated DNA containing the motif of site 1 was analyzed by qRT-PCR using specific primer pairs. Precipitation of chromatin with normal rabbit IgG serum resulted in equal amplification of IFN-γ and uninduced samples. In contrast, immunoprecipitation from IFN-γ-induced cells with α-STAT1 resulted in enrichment of the GAS-binding motif of site 1 (Figure 5D). This indicates that IFN-γ stimulates the direct interaction of STAT1 with a specific GAS motif in the IL-31RA promoter.

Figure 5.

IFN-γ induces STAT1 binding to the IL-31 RA promoter. A Schematic representation of the human IL-31RA genomic locus as described by Diveu et al. (6). Positions of the three putative STAT-binding motifs are indicated relative to the translational start site. B Nuclear extracts from untreated and IFN-γ-treated moDCs were prepared and EMSAs using radiolabeled probes harboring the sequences -2047/-2041 (site 1), -1794/-1788 (site 2), and -1036/-1030 (site 3) were performed. C For competition assays, 50-fold molar excess of unlabeled wild-type (WT) or mutated (mut) oligonucleotide was added to the binding reaction. In the supershift experiments, nuclear extracts were pre-incubated with an antibody directed against STAT1, STAT3 or STAT5. Data show representative results of one of four donors. D Chromatin immunoprecipitation assay. To demonstrate STAT1 binding to the IL-31RA promoter in living cells, chromatin fragments from uninduced and IFN-γ-treated moDCs were precipitated using α-STAT1 or an equal amount of control IgG. The data represent the amplification signal obtained by qRT-PCR for the STAT1 binding site (site 1) within the human IL-31RA promoter, expressed as a percentage of the corresponding input samples. Results show representative data of one of three independent experiments.

IFN-γ-induced IL-31RA expression is decreased in STAT1-deficient DCs

To investigate the functional role of STAT1 in IL-31RA expression, RNA interference experiments were carried out in moDCs. Four days after transfection, the silencing efficiency of the STAT1 siRNA was determined. As shown by Western blot analysis, STAT1 siRNA dramatically reduced pSTAT1 protein levels in IFN-γ-treated cells compared to cells transfected with control oligonucleotide, whereas ERK levels remained nearly unaffected (Figure 6A). qRT-PCR analysis 24 hours after IFN-γ stimulation revealed a reduction in IFN-γ-dependent IL-31RA mRNA expression in STAT1-silenced cells compared to cells transfected with control oligonucleotide (Figure 6B). Specificity of the STAT1 siRNA was controlled by transfecting STAT1 siRNA into HEK293 cells. Four days after transfection the expression of STAT1, STAT3 and STAT5 was analyzed by Western blotting (Figure 6C). Taken together, these data clearly show that the transcription factor STAT1 is critically involved in the IFN-γ-induced expression of IL-31RA.

Figure 6.

IFN-γ-induced IL-31RA mRNA expression depends on STAT1. A moDCs were transfected with control siRNA (BlockIT) or with STAT1 siRNA and stimulated with IFN-γ for 30 minutes. Total cell lysates were taken and analyzed for STAT1 phosphorylation by Western blot. ERK protein is shown as a control for equal loading. One representative blot out of five is shown. B 24 hours after IFN-γ stimulation, IL-31RA mRNA was detected by qRT-PCR. Results represent the mean and standard deviation of three independent experiments (* p≤0.05). C As a control for the specificity of STAT1 siRNA, HEK293 cells were transfected with STAT1 siRNA or control siRNA. Three days after transfection, cell lysates were analyzed for total STAT1, STAT3 and STAT5 protein by Western blot (upper panel). As a loading control, total ERK protein was detected (lower panel).

IL-31 stimulation of primary CD1c⁺ DCs expressing the IL-31R complex results in the release of pro-inflammatory cytokines

Recent literature provides evidence that IL-31 is a potent inducer of pro-inflammatory mediators in various cell types, including epithelial cells, eosinophils, colonic subepithelial myofibroblasts, PBMCs and macrophages (10, 22, 36). Whereas human bronchial epithelial cells and freshly isolated human eosinophils respond to IL-31 by secreting epidermal growth factor (EGF), vascular endothelial growth factor (VEGF), monocyte chemoattractant protein-1 (MCP-1/CCL2), IL-6, and CXCL8/IL-8 (10), PBMCs and macrophages instead respond to IL-31 and co-treatment with SEB by releasing IL-6 and IL-1β (22). To determine the effects of IL-31 receptor activation in human primary CD1c⁺ DCs, cells from 6 different donors were isolated and stimulated with IFN-γ. After 48 hours, IL-31RA and OSMRB expression was measured by flow cytometry. Only donor DCs that showed increased expression of IL-31RA (n=5) were selected for IL-31 treatment. IL-31 was added at concentrations ranging from 0 to 250 ng/ml and supernatants were harvested 24 hours post-cytokine stimulation. Quantification of cytokines and chemokines revealed a dose-dependent release of TNF-α, IL-6, CXCL8/IL-8, CCL2/MCP-1, CCL5/RANTES and CCL22/MDC. In contrast, IL-12 and CCL17/TARC were hardly detectable and showed no increase upon IL-31 stimulation (Figure 7). Taken together, these data show that IFN-γ-experienced DCs are highly responsive to IL-31 and release a number of pro-inflammatory cytokines and chemokines upon IL-31 stimulation.

Figure 7.

Cytokine release by IL-31-stimulated DCs. CD1c⁺ DCs were treated with IFN-γ for 48 hours. DCs of donors expressing the IL-31 receptor complex were stimulated with IL-31 at the indicated concentrations. After 24 hours, supernatants were collected and cytokine release was measured by ELISA. Results represent the mean and standard deviation of five donors. *p≤0.05, **p≤0.01, ***p≤0.001 compared to 0ng/ml values.

DISCUSSION

DCs are pivotal in bridging the innate and adaptive immune responses. They are the most potent type of APC and are uniquely capable of promoting the differentiation of naive Th cells into different types of effector Th cells, thereby initiating specific immune responses (37). DCs express a large repertoire of innate immune receptors and sentinel the periphery where they can recognize and respond to various microbial components. Upon exposure to microbial stimuli, DCs undergo phenotypic and functional changes, becoming immunogenic APCs and priming appropriate T-cell responses (15, 38). Besides microbial components, specific chemokines and cytokines can contribute to DC activation (24, 39-40).

The present study provides the first evidence that, under certain conditions, DCs express the heterodimeric IL-31 receptor, which allows for IL-31-dependent DC activation. Interestingly, IFN-γ acts as potent inducer of IL-31 receptor expression whereas LPS and SEB, two stimuli known to effectively prime DCs to induce Th cell responses (38, 41), do not enhance IL-31 receptor expression. In contrast, macrophages and monocytes up-regulate IL-31RA expression in response to SEB, and show enhanced IL-31RA mRNA expression upon IFN-γ stimulation (1, 21, 23). Although one group was able to show enhanced IL-31RA protein levels in IFN-γ stimulated monocytes by western blotting (23), little alterations in IL-31RA surface expression could be observed by flow cytometry (unpublished observations, 22).

To date, signaling pathways which directly influence the expression of IL-31RA expression have not been determined. Based on the predominant role of IFN-γ in stimulating IL-31 receptor expression in human DCs we investigated molecular mechanisms underlying this process. It has become apparent that other signal-transduction proteins like mitogen-activated protein kinases (MAPKs) (42), phosphatidyl-inositol-3 kinase (PI3-K) (43) and nuclear factor (NF)-kappa B (44) act in cooperation with or parallel to the canonical JAK-STAT pathway. However, the majority of the pleiotropic effects of IFN-γ – in particular the regulation of gene expression in response to IFN-γ – are mediated by the canonical JAK-STAT pathway (45). Although IFN-γ primarily induces activation of STAT1, several studies reported that STAT3 and STAT5 are also activated by IFN-γ (27-28, 46s). Analysis of STAT proteins in IFN-γ-stimulated DCs showed phosphorylation of STAT1, STAT3 and STAT5. However, supershift assays revealed that only STAT1 was able to bind to one of the putative GAS-motifs in the proximal promoter. A trend towards increased binding of STAT1 in IFN-γ stimulated DCs was further observed in ChIP assays. This indicates the direct association of STAT1 and a specific GAS motif within the IL-31RA locus. Notably, supershift assays employing STAT3-specific antibodies indicate that STAT3 may be part of a second nucleoprotein complex which occurs in an IFN-γ-independent way. Notwithstanding this possibility, the pivotal role of STAT1 in IFN-γ-induced IL-31RA expression was substantiated by the results of the STAT1-silencing experiments. Thus, this study is the first to describe a signaling pathway which is directly involved in the regulation of IL-31RA.

So far, a close relationship between IL-31 expression and skin inflammation has been demonstrated (1-3, 12). In addition, increased IL-31 levels were found in the sera of patients with allergic asthma (3). These findings indicate that IL-31 may contribute to the development of Th2-related diseases. Although Th2-related diseases such as AD are characterized by the initial activation of Th2 cytokines, Th1 cytokines such as IFN-γ and IL-12 are considered to be important players during the later (chronic) phase of the diseases (47-48). Thus, we assume that IFN-γ-dependent IL-31R expression may play a role in the manifestation of Th2-related inflammatory disorders such as AD. However, this assumption is in contrast to the observation that parasitic infection of IL-31RA-deficient mice results in excessive Th2 inflammation and that IL-31 may thus have anti-inflammatory properties (13-14).

Recently, Bilsborough and colleagues provided a possible explanation for this inconsistency. They hypothesized that the observed effects in IL-31RA−/− mice were not due to the lack of IL-31R-mediated signaling, but rather to the physical absence of IL-31RA. Since the functional IL-31 receptor is composed of IL-31RA and OSMRB, absence of IL-31RA would increase the relative amount of available OSMRB to form a heterodimeric receptor with gp130, and therefore give rise to increased OSM signaling. Thus, the authors speculated that an increase in OSM signaling, as observed in IL-31RA-deficient mice (49), may account for the enhanced Th2 inflammation in IL-31RA-deficient mice (13-14). In line with these findings, our present study provides evidence that IL-31 acts as pro-inflammatory rather than as an anti-inflammatory cytokine. The observed secretion of pro-inflammatory cytokines and chemokines is in good accordance with IL-31-induced cytokine-secretion profiles described for bronchial epithelial cells (10), colonic myofibroblasts (36), and PBMCs (22). In CD1c⁺ DCs, IL-31 concentrations needed to induce the release of specific cytokines and chemokines seem to be relatively high. These findings are in agreement with other reports, showing that similar or even higher concentrations of IL-31 are required to induce the release of inflammatory mediators (10, 22, 36, 50) and may be explained by the relatively low expression of OSMRB on the surface of moDCs and CD1c⁺ DCs.

DC-derived cytokines and chemokines deliver important signals that promote T-cell polarization and determine the type of T effector cell. Although IL-4, which is unique for the induction of Th2 responses, is not released by DCs, a number of additional DC-derived factors are known to contribute to Th2 inflammation, among them the chemokines CCL2/MCP-1, CCL17/TARC and CCL22/MDC. The observation that IL-31-treated DCs show increased CCL2/MCP-1 expression but barely detectable amounts of IL-12 is of particular interest since recent data from Del Cornò and colleagues demonstrate that CCL2/MCP-1 inhibits TLR-induced IL-12 production in moDCs and, as a consequence, decreases IFN-γ production in DC–T-cell co-cultures (51). These findings indicate that IL-31-treated DCs may promote the development of Th2- rather than Th1-mediated immune responses. In addition, the release of MDC/CCL22 further indicates a role for IL-31-stimulated DCs in Th2-mediated inflammation. Elevated levels of MDC/CCL22 and TARC/CCL17 have been detected in skin lesions from AD patients. Additionally, both chemokines were reported to promote trafficking of Th2 cells (52). However, TARC/CCL17 was not produced by IL-31-stimulated DCs. CCL5/RANTES is a chemokine that attracts T cells to inflammatory sites and plays a predominant role in allergic and inflammatory skin diseases such as AD, ACD or psoriasis (53-56). Two classical pro-inflammatory cytokines, TNF-α and CXCL8/IL-8, are released by IL-31-activated DCs, as well. TNF-α was described as one of the AD-associated cytokines that is mainly expressed in late phases of the disease (57) whereas CXCL8/IL-8 is specifically up-regulated in psoriatic skin and is responsible for the typical intra-epidermal collection of neutrophils (56, 58-59).

The release of IL-31-induced cytokines and chemokines is initiated by binding of IL-31 to IL-31RA which forms a heterodimeric receptor complex with OSMRB (1, 5-7, 60). IL-31RA, the signaling chain which first binds IL-31 (60), is expressed in several isoforms (1, 6-7, 26). The longer isoforms contain three tyrosine residues known to be critical for STAT activation (7). As a consequence, IL-31 treatment of cells expressing the long isoforms results in the induction of STAT-mediated signaling, while stimulation of cells expressing the short isoforms fails to activate STAT signaling. Notably, we showed up-regulation of short and long isoforms in IFN-γ-treated DCs, suggesting that STATs are only partially involved in the activation of DCs. This is in line with our observation that IL-31 signaling in DCs results in the release of several cytokines and chemokines, which were shown to be regulated by STAT-independent signaling mechanisms (61-67). These findings indicate that, in addition to a STAT-dependent mechanism, other signaling pathways may be involved in the activation of DCs by IL-31.

In conclusion, this study demonstrates that IFN-γ/STAT1 signaling renders DCs responsive to the Th2 cytokine IL-31 by inducing IL-31 receptor expression. Subsequent IL-31 stimulation primes DCs to release mediators which are involved in inflammatory skin diseases. An important role of Th2 as well as Th1 cytokines was shown for the immunopathogenesis of AD, one of the most common inflammatory skin disorders. While the classical Th2 cytokines IL-4 and IL-13 are implicated in the initial phase of AD, the Th1 cytokine IFN-γ is instead associated with disease chronicity (47-48). The finding that IL-31 is associated with AD (3, 12) and our observation showing the predominant role of IFN-γ in the expression of IL-31RA indicates that IL-31-treated DCs may contribute to the progression of AD in the late phase of the disease. Although distinct effects of IL-31-treated DCs on T-cell activation and T-cell differentiation remain elusive and need to be addressed in further studies, our findings suggest a positive regulatory feedback loop that might enhance inflammation in the chronic phase of pruritic skin diseases, including AD.

Abbreviations used in this article

ACD

Allergic contact dermatitis

AD

Atopic dermatitis

ChIP

Chromatin immunoprecipitation

DC

Dendritic cell

GAS

Gamma-activation sequence

IL-31RA

IL-31 receptor alpha

moDC

Monocyte-derived dendritic cell

OSMRB

Oncostatin M receptor beta

qRT-PCR

Quantitative real-time polymerase chain reaction

SEB

Staphylococcal enterotoxin B

siRNA

Small interfering RNA

TSLP

Thymic Stromal Lymphopoietin