Published online before print February 15, 2008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Interdisciplinary Group of Molecular Immunopathology, Dermatology/Medical Immunology, University Hospital Charité, Berlin, Germany;
Institute of Virology, University Hospital Charité, Berlin, Germany;
Institute of Anatomy, University Hospital Charité, Berlin, Germany;
Department of Trauma and Reconstructive Surgery, University Hospital Charité, Berlin, Germany;
|| Global Drug Discovery; Bayer Schering Pharma, Berlin, Germany;
¶ Department of Dermatology, University Hospital Charité, Berlin, Germany; and
# Institute of Medical Immunology, University Hospital Charité, Berlin, Germany
1Correspondence: Interdisciplinary Group of Molecular Immunopathology, Dermatology/Medical Immunology, Campus-Charité Mitte, University Hospital Charité, Charitéplatz 1, D-10117 Berlin, Germany. E-mail: kerstin.wolk{at}charite.de
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) or dendritic cells (DCs) strongly changed their production capacity of these cytokines. Maturation of DCs with bacterial stimuli induced high expression of IL-28/IL-29 and IL-20. Simulated T cell interaction and inflammatory cytokines induced IL-29 and IL-20 in maturing DCs, respectively. Compared with monocytes, DCs expressed only minimal IL-19 levels and no IL-24. The differentiation of monocytes into M reduced their IL-19 and terminated their IL-20, IL-24, and IL-29 production capacity. Like monocytes, neither M nor DCs expressed IL-22 or IL-26. The importance of maturing DCs as a source of IL-28/IL-29 was supported by the much higher mRNA levels of these mediators in maturing DCs compared with those in CMV-infected fibroblasts, and the presence of IL-28 in lymph nodes but not in liver of lipopolysaccharide-injected mice. IL-19, IL-20, IL-22, IL-24, and IL-26 do not seem to affect M or DCs as deduced from the lack of corresponding receptor chains. The significance of IL-20 and IL-28/IL-29 coexpression in maturing DCs may lie in the broadly amplified innate immunity in neighboring tissue cells like keratinocytes. In fact, IL-20 induced the expression of antimicrobial proteins, whereas IL-28/IL-29 enhanced the expression of toll-like receptors (TLRs) and the response to TLR ligands. However, the strongest response to TLR2 and TLR3 activation showed keratinocytes in the simultaneous presence of IL-20 and IL-29.
Key Words: interferon-lambda • cytomegalovirus • chondrocytes • TLR2 • TLR3 • legumain
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, IL-28β, and IL-29, were identified [2
, 3
]. Because of structural and functional aspects (see below), the latter form a theoretical bridge between the IL-10 family and the type I IFNs, and this may be a reason for the overdue official unification of all of these mediators in the IL-10–IFN family (also called class II cytokines).
All members of the IL-10–IFN family are secreted proteins of less than 200 amino acids in length. Despite their rather limited primary structure similarity, these mediators have a related -helical secondary structure. In fact, they are composed of six or seven -helices. These helices are arranged in an antiparallel conformation, resulting either in monomeric, bundle-like proteins (e.g., IL-19, IL-22, and type I IFNs) or, with a 90° angle between the first four and the last two helices, in intertwining V-shaped dimers (e.g., IL-10, IFN-
) [4
5
6
7
8
].
The IL-10–IFN family members exert their biological effects via trans-cytoplasmic membrane receptor complexes each composed of an R1 type receptor chain and an R2 type receptor chain [4
5
6
7
8
]. The receptor chains are related by their extracellular moieties and belong to the cytokine receptor family class 2, which additionally comprises the IL-22 binding protein, tissue factor, and some signaling-incompetent and soluble splice forms of these membrane-associated receptors. With the exception of the four fibronectin type III domains in IFN-R1, the extracellular moieties of the receptor chains for the IL-10–IFN family contain two of such fibronectin type III domains. The binding of cytokines to their receptor complexes induces a signal transduction mainly through JAK–STAT pathways. According to the current convention, the R1 receptor chains are those with the longer intracellular moiety capable of binding STAT molecules; however, this understanding does not fit the type I IFN receptor complex discovered first. Interestingly, some of the IL-10–IFN family members share receptor chains 
(Fig. 6A)
[4
5
6
7
8
]. In fact, the IL-10R2 chain is part of the receptor complexes for IL-10, IL-22, IL-26, IL-28, IL-28β, and IL-29. Furthermore, the IL-20R2 chain forms a receptor complex with IL-20R1 to mediate effects of IL-19, IL-20, and IL-24, as well as a receptor complex with IL-22R1 to mediate effects of only IL-20 and IL-24. This last point simultaneously indicates that some cytokines are able to bind to more than one receptor complex and that not only single receptor chains, but even whole receptor complexes are shared among different family members. IL-28, IL-28β, and IL-29 also share a receptor complex, which is composed of IL-28R1 and IL-10R2.
 View larger version (28K): [in a new window] |
Figure 1. Expression of the novel IL-10–IFN family members during the maturation of DCs. Immature DCs, obtained from peripheral blood monocytes as described in the Materials and Methods section, were exposed or not to the following maturation stimuli for 2, 6, and 18 h: TNF- / IL-1β (A), CD40L/IFN- (B), LPS (C). Analysis of the expression of IL-10–IFN family members was performed by qPCR. Expression data are given from two experiments (2, 6, and 18 h; mean ± range) or 4 independent experiments (6 h; means±SE) as relative to the house-keeping gene (HPRT) expression.
|
 View larger version (20K): [in a new window] |
Figure 2. Expression of the novel IL-10–IFN family members in resting and stimulated monocytes, M, and mature DCs. Monocytes, M, and mature DCs were obtained as described in the Materials and Methods section and stimulated or not (control) with LPS for 6 h. Analysis of the expression of IL-10–IFN family members was done by qPCR. Expression data are given from 3 independent experiments (mean±SE). As control cells, CD4-positive T cells isolated from the peripheral blood of healthy donors and stimulated by immobilized anti-CD3 and anti-CD28 mAbs for 6 h, as described previously [11
], were analyzed. T cell expression data are given from 1 experiment. All data are given as relative to the house-keeping gene (HPRT) expression. Mo, monocytes; M, macrophages; mDC, mature DCs; Tc, CD4-positive T cells.
|
 View larger version (19K): [in a new window] |
Figure 3. Secretion of IL-29 by monocytic cells and summary of the mRNA expression data of IL-10–IFN family members in monocytic cells. (A) Immature DCs, obtained from peripheral blood monocytes, as described in the Materials and Methods section, were exposed to the indicated maturation stimuli for 6 h. The IL-29 concentration in the culture supernatants was analyzed by ELISA. Data are given from three (mean±SE) experiments. (B) Monocytes, M, and mature DCs obtained as described in the Materials and Methods section were stimulated or not (control) with LPS for 6 h. The IL-29 concentration in the culture supernatants was analyzed by ELISA. Data are given from three (mean±SE) experiments. Mo, monocytes; M, macrophages; mDC, mature DCs. (C) Table summarizing data from Figures 1
and 2
.
|
 View larger version (7K): [in a new window] |
Figure 4. In contrast to DCs, virus-infected cells do not generally show high IL-28 and IL-29 expression. Human embryonic lung fibroblasts were infected or not (controls) with human CMV and analyzed 2, 6, 18, 42, and 66 h afterward for the expression of IL-28 and IL-29, as well as for IFN-β1 by qPCR. Expression data are given from three independent experiments (mean±SE) as relative to the house-keeping gene (HPRT) expression.
|
 View larger version (14K): [in a new window] |
Figure 5. Expression of IL-28 and IFN-β1 in lymph node and liver in mice. Mice were injected intraperitoneally with LPS or PBS. Directly before (0 h) and 1, 3, 6, 24, 48, and 72 h after application, lymph nodes and liver samples were collected and analyzed by qPCR. Expression data from 3 mice per time point are given, relative to the house-keeping gene (HPRT) expression (mean±SE).
|
 View larger version (27K): [in a new window] |
Figure 6. Expression of the receptor chains for the novel IL-10–IFN family members in monocytes, M, immature DCs, and mature DCs. (A) Known combinations of receptor chains in receptor complexes enabling the cellular effects of the IL-10–IFN family members [4
, 5
, 7
, 8
]. (B) Monocytes were isolated from the peripheral blood of healthy donors and analyzed immediately or used for the generation of M or immature DCs as described in the Materials and Methods section. Immature DCs were recultured with TNF-/IL-1β (generation of mature DCs) or not (retaining immature DCs) for 2 days. Analysis of the expression of the R1 and R2 chains of the receptor complexes was done by qPCR. Expression data are given from six independent experiments (mean±SE). Significance of the differences was tested by the Wilcoxon matched-pairs signed-rank test (one-sided; *, P<0.05). As control cells, primary human chondrocytes and keratinocytes were analyzed. For these cells, expression data from one experiment are given. All data are given as relative to the house-keeping gene (HPRT) expression. Mo, monocytes; M, macrophages; iDC, immature DCs; mDC, mature DCs; Ch, chondrocytes; K, keratinocytes.
|
is exclusively produced by T, NK, and NK-T cells [5
]. As a prototypical T1 cytokine, IFN- activates the cellular immunity by promoting the antigen-presenting capacity of professional and nonprofessional antigen-presenting cells. Moreover, it increases the inflammatory potential and intracellular pathogen killing of these cells. As the activity of IL-28, IL-28β, and IL-29 in cellular viral defense mimics functional aspects of the type I IFNs, these mediators are also referred to as type III IFNs or IFN-
s. Like type I IFNs, they protect cells from the virus’s cytolytic effects by inducing the cellular expression of antiviral proteins and increase cellular MHC I expression necessary for the action of cytotoxic T cells [2
, 3
, 28
]. If present in combination with type I IFNs, they are also able to inhibit the proliferation of CD4+ T cells [29
]. Very recently, the Gallagher group described that the presence of IL-28/29 during T-cell activation increases the T1/T2 cytokine ratio [30
] and that these cytokines induced monokines and some chemokines in peripheral blood mononuclear cells (PBMCs) [31
, 32
].
As we and others have demonstrated in earlier studies, monocytes are important producers of IL-19, IL-20, and IL-24 upon cellular activation [10
, 11
, 33
] but do not seem to be targets of IL-19, IL-20, IL-22, IL-24, or IL-26 due to the limited expression of the respective R1 receptor chains [11
, 12
, 21
, 34
]. Monocytes are not terminally differentiated cells. After their 
48-h-long stay in blood, they can populate the tissues and differentiate into either macrophages (M) or dendritic cells (DCs). This study aimed to illuminate the role of M and DCs in the biology of the novel IL-10–IFN family members.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) or 10 ng/ml GM-CSF/10 ng/ml IL-4 (generation of immature DCs). Regarding the DC culture, cells were then recultured with 10 ng/ml TNF-/10 ng/ml IL-1β/1 µg/ml CD40L, 10 ng/ml IFN-, or 100 ng/ml Escherichia coli 0127:B8 LPS (Sigma-Aldrich, Deisenhofen, Germany) (for the induction of maturation process) or without further stimuli (retaining immature DCs) in the continuous presence of GM-CSF/IL-4 for 2, 6, and 18 h. Additionally, DCs fully matured for 48 h by 10 ng/ml TNF-/10 ng/ml IL-1β were generated. Stimulation of monocytes, M, and mature DCs was performed using 100 ng/ml LPS for 6 h. For the quantitative analysis of intracellular phospho-STAT1 and phospho-STAT3, total PBMCs were precultured for 1.5 h and afterward stimulated or not (0 ng/ml) with 1, 10, 100, and 1,000 ng/ml of IL-29 and IFN-β1 for 20 min. Primary human chondrocytes were isolated from the hyaline cartilage of the knee joint and cultured for 18 h as described previously [36 ]. Human fetal lung fibroblasts, maintained as described previously [37 ], were infected with CMV strain AD169 or left without infection. Adsorption was allowed for 1 h at 37°C. Cells were analyzed at 0, 2, 6, 18, 42, and 66 h after the adsorption period. CD4-positive T cells were isolated from PBMCs by negative selection using the MACS system (Miltenyi Biotec) and stimulated or not by immobilized anti-CD3 and anti-CD28 monoclonal antibodies for 6 h as described previously [11 ].
Primary human keratinocytes were obtained from Promocell (Heidelberg, Germany) and Cascade Biologics (Karlsruhe, Germany) and cultured in KGM medium (Lonza, Verviers, Belgium). For investigating the effects of IL-20, IL-29, or the combination thereof, cells were precultured for 1 to 2 days and were exposed or not (control) to these cytokines. For Western blot analysis of STAT phosphorylation, stimulation was done either with 40 ng/ml of each cytokine for 20, 40, and 60 min (kinetic measurement) or with 16, 40, and 100 ng/ml of each cytokine for 20 min (dose response analysis). For qPCR analysis, stimulation was done with 40 ng/ml of each cytokine for 24 h. Moreover, keratinocytes, stimulated or not (control) with 40 ng/ml IL-20, IL-29, or the combination thereof for 24 h, were either washed or not, and (S)-[2,3-Bis(palmitoyloxy)-(2-RS)-propyl]-N-palmitoyl-(R)-Cys-(S)-Ser-(S)-Lys4-OH (Pam3CSK4) (final concentration 10 µg/ml) (Axxora, Lörrach, Germany), or polyinosinic–polycytidylic acid [poly(I:C)] (final concentration 10 µg/ml) (Sigma-Aldrich), or medium, was added for another 24 h. All cytokines and CD40L used in this study were obtained from R&D Systems (Wiesbaden-Nordenstadt, Germany).
Mice
Fourteen-week old male BALB/c mice were intraperitoneally injected with 5 µg/g body weight LPS from Escherichia coli (0111:B4, Sigma, Taufkirchen, Germany) or a corresponding volume of PBS (controls). Before (0 h) or 1, 3, 6, 24, 48, and 72 h post-LPS application, mice were killed, whereby lymph nodes and liver samples were harvested and snap frozen for qPCR analysis. This study was approved by the regional authorities in Berlin for provisions on labor, health, and technical safety.
Flow cytometric analysis
To assess the purity of isolated monocytes and CD4-positive T cells, cells were stained with the following fluorescence-labeled mAb clones: SK7 (anti-CD3) (BD Biosciences, Heidelberg, Germany), 13B8.2 (anti-CD4), B9.11 (anti-CD8), RMO52 (anti-CD14), 3G8 (anti-CD16), J4.119 (anti-CD19), N901 (anti-CD56), NC1 (anti-CD57) (Coulter Immunotech, Hamburg, Germany). To confirm the development of M, as well as immature and mature DCs, we additionally used the mAb clones: BL6 (anti-CD1a), HB15a (anti-CD83) (Coulter Immunotech), L307.4 (anti-CD80), 2331 (anti-CD86), and L243 (anti-HLA-DR) (BD Biosciences). Purity of isolated monocytes was 90 to 96%. In line with the literature data, M exhibited high expression of CD14 and CD16 and elevated CD86 and HLA-DR levels compared with monocytes. The differentiation of monocytes into DCs was accompanied by clearly increased expression of CD80, CD1a, and HLA-DR, and their additional maturation led to the de novo expression of CD83 and a further elevation in CD80, CD86, and HLA-DR expression (data not shown).
For the detection of intracellular phospho-STAT1 and phospho-STAT3 in monocytes, stimulated PBMCs were fixed (BD cytofix, BD Biosciences), permeabilized (BD Perm Buffer III, BD Biosciences), and stained with the following fluorescence-labeled monoclonal antibody clones, corresponding to the suppliers instructions: 4a (anti-phospho-STAT1), 4 (anti-phospho-STAT3) (both from Cell Signaling Technology, Beverly, MA, USA), SK7 (anti-CD3) (BD Biosciences), and RMO52 (anti-CD14) (Coulter Immunotech).
The analyses were performed by means of a FACS Calibur instrument and Cellquest software (BD Biosciences).
Quantitative mRNA analysis
Before RNA isolation, mouse samples were homogenized in Invisorb lysing solution (Invitek, Berlin, Germany) during thawing by means of Ultraturrax tissue homogenizer (Jahnke and Kunkel, Staufen, Germany), and treated afterward with 4 mg/ml proteinase K for 1 h (Macherey-Nagel, Düren, Germany). Isolation of total cellular RNA from mouse tissue and human cells was performed using Invisorb RNA kit II (Invitek). The mRNA was reverse transcribed as described previously [34
, 38
]. Quantitative PCR analyses were done using the ABI Prism 7700 Sequence Detection System (Applied Biosystems, Weiterstadt, Germany). For the detection of IL-10, IL-19, IL-20, IL-22, IL-24, IL-26, IL-10R1, IL-10R2, IL-20R1, IL-20R2, IL-22R1, β-defensin 2, legumain, cathepsin H, cathepsin L, and the house-keeping gene hypoxanthine phosphoribosyl-transferase 1 (HPRT), previously established detection systems with amplification efficiencies of 100% using exon–exon boundary spanning FAM/TAMRA double-labeled probes were used as described. For the detection of all other expressions, systems purchased from Applied Biosystems were used together with the respective matching system for HPRT.
ELISA
Human IL-29 was quantified using the DuoSet ELISA development kit from R&D Systems. Detection of human IL-8 and TNF- was accomplished using the Immulite system (DPC Biermann, Bad Nauheim, Germany).
Western Blot analysis
Cell lysis, protein electrophoresis, and Western blot analysis were performed as described previously [39
]. Blotted samples were incubated with polyclonal antibodies against phospho-STAT1 (Tyr 701), phospho-STAT3 (Tyr 705), phospho-STAT5 (Tyr 694), and total STAT3 (all from Cell Signaling Technology), followed by incubation with peroxidase-conjugated AffiniPure F(abī)2 goat anti-rabbit IgG (H and L) secondary Ab fragment (Dianova, Hamburg, Germany) and enhanced chemiluminescence detection (Amersham Pharmacia Biotech, Freiburg, Germany).
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
/IL-1β, 2) soluble CD40L/IFN- as an analog to T cell-caused maturation of DCs, or 3) the bacterial stimulus LPS. IL-19, IL-20, IL-22, IL-24, IL-26, IL-28, and IL-29 mRNA levels were analyzed using real-time PCR analysis on reverse transcribed mRNA (quantitative PCR, qPCR) 2, 6, and 18 h after initiation of the maturation process and correspondingly, in immature DCs. As a comparison, the expression of IL-10 and IFN-β1 was also characterized in these cells. As Fig. 1A
1B
1C
demonstrate, IL-20, IL-28, immature DCs expressed IL-10 and IL-19 mRNA. Maturation of DCs increased IL-10 and IL-19 expression and induced the expression of IL-29, and IFN-β1, but not of IL-22, IL-24, or IL-26. Interestingly, the different DC maturation stimuli caused different cytokine expression patterns. In fact, TNF-/IL-1β and LPS, but not CD40L/IFN-, induced IL-10 and IL-20 expression. Furthermore, CD40L/IFN- and LPS, but not TNF-/IL-1β, led to a clear IL-29 and IFN-β1 expression, whereby LPS provoked much higher levels of IL-29 and IFN-β1 than did CD40L/IFN-. Regarding IL-19, expression levels were similar during TNF- / IL-1β-, LPS-, and CD40L/IFN--induced DC maturation.
Afterward, we questioned whether mature DCs and M would express the novel IL-10–IFN family members after stimulation with LPS. These cells, as well as the original monocytes, were stimulated with LPS or left unstimulated (controls) for 6 h. CD4-positive T cells, activated for 6 h by immobilized anti-CD3 and anti-CD28 monoclonal antibodies, were included in these investigations as controls. As shown in Fig. 2
, the differentiation of monocytes into M or mature (TNF-/IL-1β-caused) DCs reduced (IL-19) or even appeared to terminate (IL-20 and IL-24) their capacity to express IL-19, IL-20, and IL-24 mRNA after LPS stimulation. IL-19 expression in activated M at the chosen time point was fifteen times lower than that in stimulated mature DCs and more than 400 times lower than that in activated monocytes. Regarding IL-29, mature DCs expressed high mRNA levels of this cytokine after LPS stimulation compared with moderate levels in monocytes (Fig. 2)
. However, the IL-29 mRNA levels in activated mature DCs were still approximately 10 times lower than that expressed during LPS-initiated maturation of immature DCs (Fig. 1)
. Interestingly, the IL-29 expression in activated mature DCs was not associated with any IL-28 expression, indicating different regulations of the expression of these genes in these cells. As for IL-20 and IL-24, the IL-29 expression was completely abrogated by the differentiation of monocytes into M. As in monocytes, no expression of IL-22 or IL-26 could be detected in either of the monocyte-derived cell populations, but they were clearly expressed in the stimulated T cells (Fig. 2)
.
To verify the high IL-29 mRNA expression in DCs, we measured the secretion of this cytokine. Indeed, we found very high IL-29 protein levels in the supernatant of DCs maturing via LPS (Fig. 3A
). In line with the mRNA data, this IL-29 production was much more pronounced than that of already mature (TNF-/IL-1β-caused) DCs stimulated with LPS (Fig. 3B)
.
Taken together, the cytokine expression data suggest that the monocyte differentiation into M abrogates most of their capacity to produce the novel IL-10–IFN family members, whereas monocyte-derived DCs are sources of IL-19 and IL-20, as well as, most importantly, IL-28 and IL-29 (Fig. 3C)
. IL-28/β and IL-29 are primarily known for their expression in potentially any cell type after viral infection. To compare the IL-28 and IL-29 levels in DCs with that of unrelated virus-infected cells, we infected human embryonic lung fibroblasts with human CMV and analyzed them 2, 6, 18, 42, and 66 h afterward for the expression of these cytokines, as well as for IFN-β1. Surprisingly, and in contrast to high IFN-β1 expression already seen 2 h after the beginning of the infection, CMV-infected cells showed only minimal IL-29 expression and no IL-28 expression at all (Fig. 4
). This indicates that virus infection is not always accompanied by high IL-28/IL-29 expression and that DCs may be very important cellular sources for these mediators in vivo also independent of virus infection.
In the next step, we investigated the IL-28 expression in mice injected with LPS or PBS (control) (of note, there is no murine IL-29, making IL-28 to be the only IL-28R1 ligand in this species). Liver, as the organ containing the most M of the body, and lymph nodes, as the organ where maturing DCs migrate to, were taken before (0 h) and 1, 3, 6, 24, 48, and 72 h after injection. Expression of IL-28 and, as a comparison, IFN-β1 was analyzed by qPCR. As shown in Fig. 5
, mRNA of both cytokines were expressed in lymph nodes already 1 h after LPS application, although IFN-β1 expression was much higher and more prolonged. In line with the lacking IL-28/29 expression and the low IFN-β1 expression in human M in vitro (Fig. 2)
, there was no IL-28 expression and low levels of IFN-β1 in the murine liver (Fig. 5)
.
In the second part of our study, we asked what the consequences of the cytokine expression by monocyte-derived cells might be, especially by maturing DCs. Since DCs, dependent on their differentiation/maturation state, appear to be the most important cellular sources for these three cytokines, we focused here on IL-20 and IL-28/29. In general, cytokines secreted by maturing DCs may act on 1) the DCs themselves, 2) the peripheral tissues where the DCs receive the maturation stimulus, and/or 3) the specific T cells that the DCs interact with (e.g., after their arrival in the T cell area of the tissue-draining lymph node).
In the first section of the second part of our study, we investigated whether DCs and M themselves can be targets of the novel IL-10–IFN family members. Therefore, the monocytes, M, immature DCs, and mature DCs were analyzed for the mRNA expression of the R1 (IL-10R1, IL-20R1, IL-22R1, IL-28R1) and the R2 (IL-10R2, IL-20R2) receptor chains by qPCR. Additionally, the mRNA expression of the type I IFN receptor chains, IFN-R2c and IFN-R1, was investigated. As a comparison, primary human chondrocytes obtained from the hyaline cartilage of the knee joint, as well as primary human keratinocytes were included in the study. As shown in Fig. 6B
, M, immature DCs, and mature DCs principally showed an expression pattern similar to that of monocytes, but different from chondrocytes and keratinocytes. In fact, like monocytes, M and DC populations showed high expression of IL-10R1, IL-10R2, IFN-R2c, and IFN-R1, and low levels of IL-20R2 and IL-28R1. It must be noted that because of the absent expression of any known partner chain of IL-20R2 (IL-20R1 or IL-22R1), the function of IL-20R2 per se as well as any functional consequences of a differential expression of this chain in these cells remain unknown. It is also interesting to note that the expression of IL-28R1 in the tested monocytic cell populations was generally 100 times lower than that of the IL-10 and the type I IFN receptor chains. That should mean that if assuming equal translation efficiencies for IL-10R1 and IL-28R1, as few as approximately one to five IL-28R1 protein molecules are expected on the monocyte cell surface ([40
] and our unpublished results). We therefore questioned whether this low IL-28R1 expression would be sufficient to render these cells sensitive to IL-28/IL-29. In three independent experiments, we found that the stimulation of monocytes using increasing concentrations of IL-29 (up to 1000 ng/ml) actually demonstrated only slight, if any, activation of signal transduction as assessed by flow-cytometric analysis of intracellular phospho-STAT1 and phospho-STAT3, whereas these cells clearly responded to stimulation with already 1 ng/ml of IFN-β (data not shown). These data indicate that (at least resting) monocytes cannot be sufficiently influenced by IL-28/29.
The differentiation of monocytes into DCs slightly decreased the expression of IL-10R1, IL-10R2, IFN-R2c, IFN-R1, and IL-28R1, whereas no difference could be detected between monocytes and M and between immature and mature DCs regarding these expressions. These results suggested that DCs have a lower sensitivity of these cells to IL-10 and type I IFNs compared with monocytes or M. In line with this suggestion, our preliminary experiments showed that IL-10 regulated the MHC class II expression, as well as the LPS-induced TNF- production in DCs with less efficiency than in monocytes and M. Regarding IL-20R1 and IL-22R1, the continuous absence of these chains suggests a lack of M and DC sensitivity to IL-19, IL-20, IL-22, IL-24, and IL-26, as it is the case in monocytes [12
, 21
]. In contrast to the monocytic cells, expression of IL-20R1 and IL-22R1 was clearly detected both in chondrocytes and in keratinocytes (Fig. 6B)
.
A few manuscripts and Fig. 6
have well documented that some tissue cells, but not immune cells, are important targets of novel IL-10–IFN family members [2
, 3
, 6
, 7
, 19
]. Therefore, we then characterized the effects of IL-20 and IL-29 on primary human keratinocytes. First, we investigated the signal transduction of these cytokines using Western blot analysis in kinetic and dose-response studies. As showed in Fig. 7
, treatment of human primary keratinocytes with IL-20 and IL-29 clearly increased the levels of phospho-STAT3 and of phospho-STAT1, respectively. None of these cytokines caused an activation of STAT5. Very interestingly, we found that the signal intensity for IL-20-induced phospho-STAT3 was clearly increased in the presence of IL-29 (Fig. 7)
. Second, we questioned which effects IL-20 and IL-29 do exert on keratinocytes. We speculated that in the case of a cutaneous infection, cytokines secreted locally by maturing DCs may enhance the innate immunity of the skin, may strengthen the cutaneous regeneration, and/or may help themselves to cross the extracellular matrix on their way to the regional lymph nodes. IL-20 is already known to induce antimicrobial proteins e.g., β-defensin 2 [41
] in keratinocytes; however, no effect on keratinocytes is known for IL-29. We therefore addressed the role of IL-29 in the epidermis and, using qPCR, screened for possible effects of this cytokine regarding the keratinocyte differentiation (expression of desmocollin 1, kallikrein 7, late cornified envelope 1B), the innate immunity of these cells (expression of β-defensins, psoriasin, cathelicidin, TLR 2, 3, and 4), the chemokine production (IL-8, CXCL1, CCL27), and the expression of proteases that degrade the extracellular matrix enabling DC emigration (matrix metalloproteinase 9). Moreover, we analyzed the expression of cystein proteinases, such as legumain, cathepsin L, cathepsin V, and their inhibitor cystatin M/E that were recently suggested to also play a role in epidermal cornification and desquamation [42
]. Using a 24-h stimulation period, we indeed found an up-regulation of TLR2 and TLR3, and a minimal increase of legumain expression by IL-29 (Fig. 8A
) and IL-28 (data not shown). None of the other molecules tested were influenced in their expression by IL-28/IL-29. IL-20 clearly induced the expression of psoriasin in keratinocytes (Fig. 8A)
. We then investigated whether the IL-29-induced increase of TLR2 and TLR3 expression would be functionally relevant. Primary human keratinocytes were stimulated or not (control) with IL-20, IL-29, or the combination thereof for 24 h. Then, cells were extensively washed, and control medium, Pam3CSK4 (ligand for TLR2), or poly(I:C) (ligand for TLR3) was added for another 24 h. In line with the observed IL-29-induced up-regulation of TLR2 and TLR3 (Fig. 8A)
, we found clearly increased sensitivity of IL-29 pretreated keratinocytes to the respective TLR ligands, as assessed by IL-8 secretion (Fig. 8B)
. Interestingly, if the cytokines were not removed from the cultures prior to TLR stimulation (Fig. 8C)
, IL-20, which alone had no influence on keratinocyte TLR expression or IL-8 secretion, amplified the TLR ligand induced IL-8 production, particularly in the IL-29 pretreated culture. This surprising observation was based on five independent experiments using cells from different donors and was significant. We then analyzed further parameters in this culture system (TLR stimulation in the presence of IL-20 and IL-29), namely, the secretion of TNF- protein and the cellular mRNA expression of β-defensin 2. Similar to the IL-8 secretion, the production of both mediators was highest if IL-29 and IL-20 were simultaneously present in the culture (Fig. 8D
and 8E)
.
 View larger version (39K): [in a new window] |
Figure 7. IL-20 and IL-29 induce STAT phosphorylation in keratinocytes. Primary human keratinocytes were stimulated or not (control) with IL-20, IL-29, or combinations thereof and cellular tyrosine phosphorylation of STAT1, 3, and 5, and total STAT3 levels were assessed by Western blot analysis. (A) Cytokine stimulation was done with 40 ng/ml of each cytokine for 20, 40, and 60 min. (B) Cytokine stimulation was done with 16, 40, and 100 ng/ml of each cytokine for 20 min. Data from 1 out of 3 independent experiments are shown.
|
 View larger version (23K): [in a new window] |
Figure 8. Influence of IL-20 and IL-29, simultaneously produced by DCs during LPS-induced maturation, on keratinocytes. Primary human keratinocytes were stimulated or not (control) with IL-20, IL-29, or combinations thereof for 24 h. (A) Gene expression of psoriasin, TLR2, TLR3, and legumain was analyzed by qPCR. Expression data are given from 3 independent experiments (mean±SE) as relative to the house-keeping gene (HPRT) expression. (B–E) After this stimulation period, cells were either washed (B) or not (C–E), and control medium, Pam3CSK4 (TLR2 ligand, TLR2L), or poly(I:C) (TLR3 ligand, TLR3L) was added for further 24 h. (B, C) Supernatants were analyzed for IL-8 content by the Immulite system. Data from 5 independent experiments are given as the mean ± SE. Significance of the differences was tested by the Wilcoxon matched-pairs signed-rank test (*, P<0.05). (D) Supernatants were analyzed for TNF- content by the Immulite system. Data from 3 independent experiments are given as the mean ± SE. (E) Cells were lysed, and analysis of cellular β-defensin 2 expression was performed by qPCR. Data are given from 3 independent experiments as relative to the house-keeping gene (HPRT) expression as the mean ± SE.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and DCs are tissue-resident cells. Whereas M are important cells with inflammatory and scavenger functions, DCs are the most specialized antigen-presenting cells. DCs are necessary, particularly for the initiation of the adaptive immune response against pathogens that invade the host for the first time. In their immature stage, these cells take up pathogens that they encounter in the tissue. Their activation leads to their maturation and migration from the tissues into the local lymph nodes, where they may activate circulating specific T cells even in their naive stage and cause the generation of effector T cells. Cytokines are key regulators of M and DC functions. In fact, the inflammatory and scavenger capacities of M are influenced by the cytokine milieu. Furthermore, the type of effector T cells induced by DCs is dependent on the cytokines present during DC maturation and DC–T cell interaction. For instance, the presence of IFN- has been suggested to favor the DCs’ ability to generate type I T cells [43
, 44
].
We and other groups could already show in earlier studies that activated monocytes expressed IL-10, IL-19, IL-20, IL-24, and IFN-β1 [5
, 9
10
11
, 33
]. We described here for the first time that activated monocytes also expressed IL-29 (Fig. 2)
. Additionally, our current study demonstrates that the differentiation of monocytes into M or DCs strongly modifies their capacity to express these cytokines. Regarding the M, these cells showed increased (IL-10), clearly reduced (IL-19), or even lacking (IL-20, IL-24, IL-28, IL-29) capacity to express IL-10, IL-19, IL-20, IL-24, IL-28, and IL-29. This statement is based on our analysis limited to 6 h (Fig. 2)
and 18 h (data not shown) LPS stimulation of these cells. Only minimal LPS-induced (3 h) IL-29 expression in M was reported by Siren et al. [45
]. Unfortunately, Siren and co-workers did not include monocytes in their study as a comparison [45
].
Regarding DCs, our data suggest that apart from their capacity to express large amounts of IFN-β1, these cells are important sources of IL-20, IL-28, and IL-29, whereas their capacity to express other novel family members was either reduced (IL-19) or terminated (IL-24) compared with monocytes (Fig. 1
and 2)
. Interestingly, the cytokine expression was most pronounced during DC maturation, and the cytokine profile varied depending on the maturation modus. In fact, maturation by bacterial LPS led to coexpression of IL-20, IL-28, and IL-29. However, when maturation of DCs was initiated by inflammatory cytokines, clear IL-20 expression was observed while IL-28 and IL-29 expression was lacking. Vice versa, we did not detect any IL-20 expression when maturation was induced using stimuli naturally delivered via T cell interaction (CD40L/IFN-), although IL-29 was expressed in this condition. The expression of IL-29 in monocytic cells seems to be regulated similarly to that of IFN-β1. Like IFN-β1, IL-29 observed during LPS-induced DC maturation was expressed to a much higher extent than in DCs maturing with CD40L/IFN-. The LPS-induced IL-29 mRNA expression was clearly accompanied by high IL-29 secretion. Moreover, IL-29 was expressed in mature DCs after LPS stimulation with levels higher than that of LPS-stimulated monocytes. mRNA expression of IL-29 in LPS-stimulated monocyte-derived immature DCs was also reported by Coccia et al. [46
]. Similar to our study, transient IL-28 expression was observed simultaneously.
The relevance of our in vitro results regarding the IL-28/29 expression in DCs and M during LPS-induced maturation and stimulation was supported by experiments with LPS-injected mice showing early IL-28 and IFN-β1 expression in lymph nodes and lacking IL-28 and lower IFN-β1 expression in the liver (Fig. 5)
. This differential pattern may reflect the myeloid cell composition in both organs with the liver predominantly consisting of M and the lymph nodes where DCs migrate to after receiving a maturing signal in the peripheral tissue.
It should also be noted that, in contrast to the high IL-28/29 expression levels in LPS-stimulated DCs, only minimal IL-28/29 expression could be observed in CMV-infected fibroblasts in our study (Fig. 4) . Interestingly, IFN-β1 expression was high in both LPS-stimulated DCs and CMV-infected fibroblasts. This demonstrates a different expression regulation and a nonredundant role of IL-28/29 and type I IFNs in the viral defense.
We have previously shown that IL-24 is highly expressed cutaneously during skin inflammation in humans and mice [12
]. The lacking capacity of both DCs and M to produce IL-24, as demonstrated by the present study, suggests that local, nonimmune tissue cells, such as keratinocytes, are the main sources of this cytokine in the inflamed tissue. This assumption is in line with our recent data showing expression of IL-24 by in vitro activated human keratinocytes [12
].
This study also emphasizes that, like monocytes themselves, neither M nor DCs seem to express IL-22 or IL-26. Additional investigations from our group also excluded any expression of these cytokines in nonimmune tissue cells ([12
, 21
] and our unpublished observations). These data support the assumption that IL-22 and IL-26 cannot be produced by any cells other than T, NK, and probably NK-T cells.
Additionally, we show here that the expression pattern of the receptors for the novel IL-10–IFN family members was similar in monocytes, M, and DCs (Fig. 6)
. Compared with monocytes, slightly decreased levels of IL-10R1 and IL-10R2, found in DCs but not in M, corresponded with our preliminary data showing reduced sensitivity of DCs to IL-10. Furthermore, significantly lower IL-28R1 expression was found in DCs compared with that in monocytes and M. This last fact does not align with the previous data by Mennechet et al. describing an increase of IL-28R1 expression and IL-28/29 sensitivity during the monocyte differentiation into DCs [47
]. Additional investigations are necessary to explain this discrepancy.
It should also be noted, that monocytes, M, and DCs expressed IL-20R2 and that this chain expression was observed in our study to be increased in immature DCs. However, none of the currently known partner chains of IL-20R2 was present in the cells. Further studies focusing on the discovery of a possible combination of IL-20R2 with an additional R1 chain (e.g. IL-10R1, IL-28R1, or a so far unknown R1 chain) would be the first step in understanding the increased IL-20R2 expression in DCs.
The IL-20, IL-28, and IL-29 effects on keratinocytes demonstrated in this report in connection with the facts already known about these mediators prompt speculations about the biological relevance of the expression of these mediators by maturing DCs (depending on the nature of the activation stimulus). Cytokines produced during DC maturation may act on these cells themselves, on tissue cells in the organ where the DCs receive the maturation stimulus and/or on cells in the draining lymph node where the DCs migrate to. We demonstrated here that IL-20 could not act on DCs because of the lack of corresponding R1 type receptor chains. It is already known that IL-20 cannot act on T cells either [11
, 16
]. Therefore, the production of IL-20 during the DC maturation induced by either a bacterial (LPS) or an inflammatory stimulus (TNF-/IL-1β) suggests effects of this cytokine on the tissue cells of the organs where the immature DCs are located at the time of a bacterial infection. This is in line with data demonstrating induction of psoriasin and other antibacterial proteins (e.g., β-defensin-2) by IL-20 in keratinocytes (Fig. 8
and [41
]), which would contribute to the clearing of the cutaneous infection. The Schroeder group recently found, that psoriasin confers upon human skin the most important protection against gram-negative bacteria, especially Escherichia coli [48
]. β-defensin-2 also acts primarily against gram-negative bacteria. Interestingly, IL-20 was not expressed in CD40L/IFN--exposed immature DCs, and in contrast to IL-28 and IL-29, IL-20 did not induce the expression of TLR2 (receptor, e.g., for bacterial peptidoglycans and lipoteichoic acid, fungal zymosan, and for several herpes viruses [49
, 50
]) and of TLR3 (receptor for double-stranded viral RNA [49
]) in keratinocytes. Moreover, the expression of IL-28 and IL-29 after LPS- or CD40L/IFN--induced maturation suggests a role of these cytokines apart from the well-established antiviral responses, in which these cytokines are induced by the virus-infected cells and primarily act in an autocrine manner. The application of CD40L/IFN- mimics the stimulatory interaction of DCs with T cells. As mentioned above, the Gallagher group described how the presence of IL-28/IL-29 during T cell activation increases the T1/T2 cytokine ratio [30
]. Here, we demonstrated that IL-28/29 clearly up-regulated the expression of TLR2 and TLR3 in primary keratinocytes, as well as increased the sensitivity toward respective TLR ligands in these cells, as demonstrated by elevated secretion of IL-8 and TNF- and increased expression of β-defensin 2 (Fig. 8)
. These data suggest an elevated recognition of and defense against gram-positive bacteria, fungi, as well as herpes and dsRNA viruses. Of note, no TLR4 expression was detected in keratinocytes under any condition tested (data not shown).
These facts give the impression that after their emigration (e.g., from the skin), maturing DCs leave behind enhanced innate immunity. If this is the case, the extent of the enhanced innate immunity of tissue cells is the consequence of the cooperative (simultaneous) action of IL-20 and IL-28/IL-29. It appears that IL-28/29 and IL-20 particularly activate distinct arms of innate immunity, antiviral vs. anti-microbial, in keratinocytes, the cells that are the primary barrier for both viral and bacterial infections. However, our data showing that IL-29 increased 1) the TLR2 expression and 2) the TLR2/3 stimulation-dependent β-defensin 2 expression, may suggest that IL-28/29 can also play a role in bacterial defense. Interestingly, IL-20 and IL-28/IL-29 can even work together. In fact, the effect on TLR stimulation was more pronounced when keratinocytes were simultaneously activated with both cytokines instead with only one.
Moreover, there seems to be further cooperative effects of IL-20 and IL-28/IL-29 on keratinocytes, as observed by the additive down-regulating effect of both cytokines on the expression of cathepsin H in these cells (data not shown). The molecular mechanisms behind the cooperative actions of IL-20 and IL-29 regarding the different effects are unknown. However, the increased signal intensity for IL-20-induced phospho-STAT3 in the presence of IL-29 (Fig. 7) could be seen as a hint for the mechanism behind the amplified expression induced by both cytokines of at least β-defensin 2, in whose gene promoter region the binding side for STAT3 can be found [21 ].
Apart from the inducing effect on TLR2 and TLR3 expression, IL-29 slightly up-regulated the expression of legumain, a cystein proteinase that was recently suggested to be associated with excessive epidermal cornification in inchq mice [42 ].
In summary, we demonstrated here that DCs may produce IL-20, IL-28, and IL-29 during the maturation process depending on their maturation modus. In the event of cutaneous bacterial infection, these cytokines may cooperate to modify the function of neighboring tissue cells.
Received August 7, 2007; revised December 21, 2007; accepted December 26, 2007.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. Li, X. Liu, Y. Zhou, and S. B. Su Interferon-{lambda}s: the modulators of antivirus, antitumor, and immune responses J. Leukoc. Biol., July 1, 2009; 86(1): 23 - 32. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Siegemund, N. Schutze, S. Schulz, K. Wolk, K. Nasilowska, R. K. Straubinger, R. Sabat, and G. Alber Differential IL-23 requirement for IL-22 and IL-17A production during innate immunity against Salmonella enterica serovar Enteritidis Int. Immunol., May 1, 2009; 21(5): 555 - 565. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Wahl, W. Muller, F. Leithauser, G. Adler, F. Oswald, J. Reimann, R. Schirmbeck, A. Seier, J. M. Weiss, B. Prochnow, et al. IL-20 Receptor 2 Signaling Down-Regulates Antigen-Specific T Cell Responses J. Immunol., January 15, 2009; 182(2): 802 - 810. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
