First we investigated whether several mediators that are known to be important in RA were able to enhance CCL18 production by MoDC. In line with previous studies, unstimulated immature DC (iDC) produced significant amounts of CCL18 19. Interestingly, incubation with TNF-α, IL-1β, IL-13, IL-15, IL-17, IL-18 and IFN-γ did not stimulate CCL18 secretion when added to day 6 iDC (n = 6). In contrast, the anti-inflammatory IL-10 strongly induced CCL18 production by iDC (p = 0.03) (figure 1a). Next we tested whether factors well known to induce maturation or T cell mimicking could induce CCL18 production. These experiments demonstrated that LPS, CD40L and RANKL did not enhance CCL18 production (n = 3) (figure 1b). Recent studies demonstrated that other TLR pathways than TLR4 are all capable of inducing DC maturation, but have different effects on cytokine production 293031. However, stimulation of TLR2 (pam₃cys), TLR3 (poly (i:c)), TLR4 (LPS) or TLR7/8 (R848) did not sort any effect on CCL18 secretion (n = 6) (figure 1c), whereas they did elicit a potent cytokine response 31. Since IL-13 is more abundantly present in RA than IL-4 and since some conflicting results have been published on CCL18 production induced by LPS when DC were cultured with IL-13 vs. IL-4, we compared these culture methods (n = 6). In both the IL-4 and IL-13 cultures, IL-10 strongly induced CCL18 (p = 0.03 for both IL-4 and IL-13 culture), while LPS again did not (figure 2). In addition, IL-10 in combination with LPS was not significantly different from IL-10 alone (figure 2). Also co-stimulation with LPS and the cytokines tested (as in figure 1) did not sort any effect on CCL18 secretion (data not shown).

MoDC and alternatively activated MΦ (AaMΦ) 3233 are known to produce CCL18. Both these cell types originate from CD14+ monocytes and depend on IL-4 or IL-13 (in combination with GM-CSF for MoDC) for their differentiation. To determine whether CCL18 secretion by myeloid cells is dependent on these cytokines, monocytes were freshly isolated and stimulated with GM-CSF, IL-4, IL-13 and IL-10 alone or in combinations (n = 6). Even after 6 days, unstimulated and GM-CSF treated monocytes/macrophages did not secrete CCL18, whereas both IL-4 and IL-13 stimulation resulted in a clear secretion of CCL18, which is in line with previous findings on AaMΦ 16. Interestingly, stimulation with IL-10 alone only had a minimal effect on CCL18 production by these monocytes/macrophages. When IL-10 was provided together with IL-4 or IL-13, this resulted in 3- and 2-fold increase in CCL18 secretion respectively (figure 3). Interestingly, already in low concentrations, IL-10 had its synergistic effect with IL-4 (figure 4a). In order to rule out effects of adherence, we cultured CD14+ monocytes/macrophages for three days in teflon bags 34 and in rotation discs 35. The morphology of these cells was comparable with freshly isolated monocytes according to their forward/side scatter pattern (data not shown). In both cultures, IL-4 did still induce CCL18 production in the same way as the cultures in 24-wells plates (figure 4b). As a proof of principle, we next tested whether the synergy between IL-4/IL-13 and IL-10 could already be observed after only 24 hours. Intriguingly, we could indeed observe a clear CCL18 secretion after 24 hours upon stimulation of freshly isolated monocytes with IL-4/IL-13 and IL-10, whereas stimulation with IL-4, IL-13 or IL-10 alone did only result in a minor or even undetectable CCL18 secretion (n = 3) (figure 4c). Since IL-10 appeared to synergize with IL-4 and IL-13, we investigated whether these cytokines could up-regulate each other's receptors, possibly resulting in enhanced signaling. This was not the case; IL-10 did not up regulate either the IL-4/IL-13 common receptor IL-4Rα1 or the specific IL-13Rα2. Furthermore, IL-4 had no effects on IL-10Rα (data not shown).

We and others demonstrated CCL18 expression in RA ST in the lining and the peri-vascular regions 1020. In figure 5, we show a high CCL18 expression in RA synovial tissue (figure 5a,b,c), which was preferentially located in both the synovial lining layer and the peri-vascular regions. Intriguingly, CCL18 was also expressed in control synovial tissue, although not as abundant as in RA ST (figure 5d,e,f). Notably, some parts of the sections were even negative for CCL18, which is in sharp contrast with RA. In order to explain the abundant CCL18 expression in RA, we tested whether incubation with RA SF could induce CCL18 production on monocytes/macrophages. Since RA SF itself contains CCL18 1820, we cultured freshly isolated monocytes for 3 days in the presence of SF, washed the cells and cultured on for another 3 days in the absence of RA SF (n = 6). Firstly, this pre-incubation with RA SF resulted in marked CCL18 production (mean 676 (± 151) pg/ml) (figure 6b). Secondly, culture of freshly isolated monocytes in the presence of RA SF, resulted in a 9- and 10-fold increase in CCL18 secretion upon stimulation with IL4/IL-13 respectively and a 22-fold increase compared with IL-10 alone (figure 6a). Intriguingly, this synergistic effect with IL-4, IL-13 and IL-10 could still be observed after 3 days of culture in the complete absence of RA SF (figure 6b), indicating that the cell does not require a continuous stimulation in order to secrete CCL18. To exclude intrinsic differences between monocytes/macrophages from RA patients and controls may contribute to the effects on CCL18 secretion, we tested whether monocytes/macrophages from RA patients (n = 3) responded differently to combinations of IL-4, IL-13, IL-10 and SF. No difference in the CCL18 secretion pattern was observed between monocytes/macrophages of healthy controls and RA patients upon these stimuli (data not shown), ruling out intrinsic differences in monocytes in RA that affect CCL18 secretion.

Since the synergy caused by RA SF appeared similar to the synergy between cytokines we already observed (figure 4), we tested whether IL-10 and/or IL-13, both present in RA SF, were responsible for this phenomenon by blocking these cytokines with neutralizing antibodies. The potency of these antibodies was first tested by determining their ability to inhibit the synergistic CCL18 secretion upon stimulation with a combination of IL-10 and IL-4/IL-13. Addition of anti-IL-10 resulted in a 73% inhibition of the synergy between IL-4 and IL-10 and anti IL-13 completely abrogated the synergistic effect of IL-13 with IL-10 (figure 6c). Unexpectedly however, blockade of IL-10 or IL-13 in SF in the presence of IL-4 or IL-13 and IL-10 respectively did not inhibit the synergy between these cytokines and SF (figure 6a&b), suggesting that SF contains a yet unidentified factor that triggers CCL18 secretion.

For cell culture experiments, 50 ml peripheral blood was taken from healthy volunteers and RA patients after receiving informed consent in 10 ml lithium heparine (Vacutainer, USA) tubes. Synovial biopsies from RA patients were taken with small needle arthroscopy (Storz, Tutlingen, Germany). Synovial fluid from RA patients was obtained during arthroscopy. For our experiments in which monocytes were stimulated with SF, a pool of SF from 10 different RA patients was used. Synovial samples from healthy controls were taken during scheduled arthroscopic procedures by orthopedic surgeons in patients with traumatic knee injuries. The Nijmegen medical ethics committee (MEC) approved these studies.

For stimulation of iDC, we used 20 ng/ml recombinant (rh) IL-1β, rhTNF-α, rhIL-10, rhIL-13, rhIL-15, rhIL-17, rhIL18, 10 ng/ml IFN-γ (all R&D systems, Minneapolis, USA), or 20 ng/ml RANKL and CD40L (Pepro Tech, Rocky Hill, USA). DCs were cultured with 500 U/ml IL-4 and 800 U/ml GM-CSF. The same IL-4 concentration was used for monocyte stimulations. For Toll-like receptor stimulation, 10 μg/ml pam₃cys (TLR2), 25 μg/ml poly (i:c) (TLR3), 2 μg/ml lipopolysacharide (LPS) (TLR4), or 1 μg/ml R848 (TLR7/8) was used 31. Blockade of IL-10 (Ebioscience, San Diego, USA) and blockade of IL-13 (Diaclone, Becanson, France) was achieved with a 1000× excess of a neutralizing antibody. For FACS analysis, we used mouse-anti human antibodies against CD14, (Dako, Glostrup, Denmark), CD83 (Beckman Coulter, Mijdrecht, The Netherlands), IL-4Rα (Santa Cruz, California USA), IL-13Rα II (R&D systems, Minneapolis, USA) and IL-10Rα (R&D systems, Minneapolis, USA) or mouse-isotype control (goat IgG for IL-13RαII). For ELISA, mouse anti-human and biotynilated goat anti-human CCL18 were used as capture and detection antibody (R&D systems, Minneapolis, USA). A standard curve was made with rhCCL18 (R&D systems, Minneapolis, USA). Immuno histochemistry for CCL18 was performed with AZN-CK18B 18 as a primary antibody.

MoDC were cultured using essentially the same protocol as described previously 1340. In brief, peripheral blood mononuclear cells were isolated from venous blood by density gradient centrifugation over Ficoll-Hypaque (Amersham Biosciences, Roosendaal, The Netherlands). The interphase was collected and washed with citrated phosphate buffered saline, and the cells were allowed to adhere for 1 hour at 37°C in RPMI-1640 (Life Technologies, Breda, The Netherlands) supplemented with 2% human serum in culture plates (Costar, Badhoeverdorp, The Netherlands). Adherent cells were cultured in RPMI-1640 Dutch modification supplemented with 10% fetal calf serum L-glutamine (Life Technologies, Breda, The Netherlands) and antibiotic-antimycotic agents (Life Technologies, Breda, The Netherlands) (culture medium) in the presence of IL-4 (500 U/ml; Strathmann Biotech, Hamburg Germany) and granulocyte monocyte-colony stimulating factor (GM-CSF) (800 U/ml; R&D systems, Minneapolis USA) for 6 days. Fresh culture medium with the same supplements was added at day 3, and then iDC were harvested at day 6. To generate mature DC, immature DC were re-suspended in a concentration of 0,5 × 10⁶/ml in fresh IL-4 and GM-CSF containing culture medium. Immature DC were then stimulated with cytokines or maturation stimuli in the presence of IL-4 and GM-CSF. DC were harvested after another 48 hours of culture. For CCL18 measurements in supernatant of cells stimulated with TLR ligands, aliquoted culture supernatant from previous experiments was used 31.

For the culture of monocytes/macrophages, CD14+ cells were isolated with magnetic cell sorting and separation (MACS). In brief, mononuclear cells were labelled with anti CD14 microbeads (Miltenyi Biotec, Amsterdam, the Netherlands) and incubated for 30 minutes at 4°C. CD14 positive cells were then separated from the other cells using a MACS column (Miltenyi Biotec, Amsterdam, the Netherlands) according to the manufacturers instructions. CD14+ cells were cultured in a concentration of 0,5 × 10⁶ cells/ml in culture medium for up to 6 days. Where appropriate, fresh culture medium was added on day 3. After 6 days, supernatant was collected for ELISA and cells were prepared for FACS analysis. In some additional experiments, monocytes/macrophages were cultured for three days in the same concentration and in the same media in teflon bags 34 or in rotation discs (Cellon, Luxembourg) 35 to prevent adherence of the cells. In experiments in which monocytes/macrophages were stimulated with RA SF, the cells were cultured for three days in the presence of 100 μl RA SF. Cells were then harvested and re-suspended in fresh culture medium without SF, but with the cytokines that were present in the first three days. Anti-IL-10 or anti-IL-13 neutralizing antibodies were only present during the first three days.

For immuno histochemistry, frozen ST was cut into 7 μm sections and mounted on slides, air-dried, and stored at -80°C. Before staining, the cryosections were air-dried, fixed in acetone for 10 min and air-dried again. The sections were then stained with 5 μg/ml mouse anti human CCL18 or isotype control at 37°C for 1 hour at room temperature (RT) and washed in PBS. Endogenous peroxidase was blocked with 0,3% H₂O₂/methanol. After another wash-step, the sections were incubated with a biotin-conjugated horse anti-mouse antibody at RT for 30 min. Next, the samples were washed and incubated with avidin-biotin-HRP complex (Vector, Burlingham, UK) at RT for 20 min. Next, the section were stained with diaminodenzidine (DAB) (Sigma, Zwijndrecht, the Netherlands). Finally, sections were then counterstained with hematoxylin, rehydrated and mounted in to allow microscopic evaluation of the samples.

The phenotype of cells was characterized by using flow cytometry techniques (FACS). For this aim, cells were harvested and collected by means of centrifugation and further processed on melting ice. Cells were diluted in buffer solution (PBS with 1% bovine albumine, pH 7,4) in a concentration of 1.10⁶ cells/ml and plated in v-shaped 96 wells plates (1.10⁵ cells per plate). Cells were labeled with monoclonal mouse- or goat anti human antibodies or mouse-isotype control (goat IgG for IL-13RαII) and incubated at a temperature of 4°C for 45 minutes. Cells were then washed and labeled with goat-anti-mouse (or rabbit anti-goat when appropriate) FITC (Zymed Laboratories, South San Francisco, USA) as a secondary antibody. After another 30 minute incubation at 4°C, cells were again washed, diluted in buffer solution and transferred into FACS tubes. Cells were gated according to their forward and side scatters and fluorescence was measured with a FACSCalibur^® (Becton-Dickinson, San Jose, USA) and Cellquest^® software.

For the detection of CCL18 protein levels of CCL18 in supernatant, a sandwich ELISA was performed as described previously 1850. In brief, maxisorb ELISA plates (Nunc, Roskilde, Denmark) were coated overnight with 50 μl/well 1 μg/ml capture antibody in PBS. Next, the plates were washed 3 times with PBS and blocked with 300 μl 1% Bovine Albumin (Sigma, Zwijndrecht, the Netherlands) in PBS for a minimum of 1 h at RT. After washing 3 times with ELISA wash buffer (PBS containing 0.05% Tween-20), the plates were incubated with 50 μl/well of serial dilutions of the sample for 2 hrs at RT. Serial dilutions of rhCCL18 were used to obtain a standard curve. After washing 3 times with ELISA wash buffer, the plates were incubated with 50 μl/well of 0.05 μg/ml secondary antibody at RT for 1 hr. Thereafter, the plates were washed 3 times with ELISA wash buffer, and incubated with 50 μl/well of streptavidin conjugated to Poly-Horse Radish Peroxidase (CLB, Amsterdam) for 20 minutes at RT. After washing 3 times with ELISA wash buffer, the presence of HRP was detected using 50 μl/well 3,3',5,5-tetramethylbenzidine (TMB) (Biomerieux, Marcy l'Etoile, France) diluted in peroxide buffer (UP) (Biomerieux, Marcy l'Etoile, France). The reaction was stopped with 50 μl/well 2,5M H₂SO₄. Absorbance was measured at 450 nm using a Magellan Tracker V4.XX (Tecan Austria GMBH). As an internal control for inter-assay variability, a sample of pooled normal human serum (n = 300) was taken along in all assays. The maximal accepted inter-assay variability is 10%. The detection limit of the ELISA is 100 pg/ml.

CCL18 production levels by monocyte derived cells upon different stimulations were compared with a Wilcoxon Signed Rank test. P-values < 0,05 were considered significant.