Expression of interleukin-18 receptor in fibroblast-like synoviocytes

See Supplementary materials.

We investigated a broad panel of synovial-membrane-derived FLS (FLS^SM) cultures from 11 patients with RA who fulfilled the criteria of the American College of Rheumatology [15] and were undergoing synovectomy and from 3 patients with OA who underwent joint replacement, and of synovial-fluid-derived FLS (FLS^SF) cultures from 9 patients with RA and 4 patients with spondyloarthropathy (SpA) [16]. Tissues were minced and digested with collagenase for 2–4 hours; FLS^SF were obtained by culture of the total synovial fluid. Supernatants and nonadherent cells were removed after 24 hours, and adherent cells were further cultured to confluence in a 37°C humidified, 5% atmosphere CO₂ in Ham's F10 medium supplemented with 10% fetal calf serum, 100 U/ml penicillin, 100 μg/ml streptomycin and 2 mM glutamine. All FLS cultures were shown to be negative for CD14 and CD86 and positive for fibroblast marker Thy-1 before being used in experiments between passages 2 and 6. Stimulation experiments were performed in 1 × 10⁵ FLS/ml for 48 hours unless otherwise stated. All biopsies and synovial fluids were obtained after the patients had given their informed consent. For comparison, additional experiments were performed in the U937 monocytic cell line.

FLS (5 × 10⁵) from two patients with RA were each cultured in duplicate, either in media without additional stimuli or exposed to IL-1β, IL-2, IL-12, IL-15, IL-18, IFN-γ, tumor necrosis factor (TNF)-α, or IL-12 + IL-18. Cells were removed after 24 hours for analysis by IL-18Rα (35 cycles) and β-actin RT-PCR. RT-PCR kinetics and IL-18Rα western blotting experiments were performed if even one of the four experiments gave a positive result (increase of ODQ ≥ 100% by cytokine stimulation).

FLS (1 × 10⁴ cells) were cultured in 200 μl medium for 3, 5, or 7 days. Cell proliferation was determined by colorimetry with the tetrazolium/formazan assay [17]. Light absorption of the proliferation-dependent color product was determined with a photometer (λ = 550 nm) in the cell-culture supernatants.

IL-18 stimulated cells and unstimulated cells were gently removed from the culture dishes by brief trypsin application, washed, and diluted in PBS containing 1% fetal calf serum. Cells were stained by monoclonal antibodies against ICAM-1 (CD54) and vascular cell adhesion molecule (VCAM)-1 (CD106), respectively, and analyzed by flow cytometry (FACScan, Becton Dickinson, Franklin Lane, NJ, USA).

Collagenase (MMP-1) and stromelysin were measured in the cell-culture supernatants of IL-18-stimulated FLS and control cultures using commercial ELISA kits from Amersham Pharmacia Biotech (Freiburg, Germany). Granulocyte/macrophage-colony-stimulating factor (GM-CSF) – a FLS-derived growth factor that is also present and that is upregulated by IL-18 in cultured osteoblasts [18] – IL-6, and the chemokine IL-8 were assayed by ELISA (from, respectively, Amersham Pharmacia, Roche Diagnostics, Mannheim, Germany, and R&D Systems, Wiesbaden, Germany) after 72 hours (GM-CSF) or 24 hours (IL-6 and IL-8). Nitric oxide production was measured from its stable product, nitrite, using the Griess reaction in cell-culture supernatants after 48 and 72 hours [19]. The release of prostaglandin E₂ was determined by ELISA (Biotrend, Cologne, Germany) in 18-hour FLS cultures.

IL-18 signaling with IκB-α phosphorylation and degradation, representing a key step of NF-κB activation [20], was assessed in IL-18Rα⁺β⁺, IL-18Rα⁺β^-, and IL-18Rα^-β^- FLS from two patients with RA and one with OA and in U937 cells. For details see Supplementary material.

Statistics were calculated for the entire group of RA-FLS^SM and RA-FLS^SF or for an indicated number of these cell cultures, by the Mann-Whitney U test for 2-tailed or untailed groups, respectively.

Expression of IL-18R was detectable by semiquantitative RT-PCR in RA-FLS^SM (mean ± SEM of ODQs after 35 cycles: 10.2 ± 5.2%), RA-FLS^SF (ODQ 10.4 ± 2.1%), and SpA-FLS^SF (ODQ 4.0 ± 1.7%), but not in OA-FLS^SM (ODQ 0.7 ± 0.4%, P < 0.05). IL-18Rβ mRNA was shown by RT-PCR in RA-FLS^SM (ODQ: 35 ± 12%), RA-FLS^SF (ODQ: 9.8 ± 3.7%), and SpA-FLS^SF (ODQ: 9.0 ± 4.4%), but, again, not in OA-FLS^SM cultures (ODQ: 0.5 ± 0.5%, P < 0.05) (Fig. 1). The simultaneous presence of RT-PCR products of both IL18Rα and IL-18Rβ was clear in 5 of 20 RA-FLS cultures but in none of the cultures derived from OA or SpA. When PCR was performed under the same conditions with U937 cells, both IL-18R chains were easily detectable.

Western blotting experiments in FLS reflected the results obtained on IL-18Rα, showing detectable IL-18Rα protein in three of four RA-FLS cultures but in neither of the OA-FLS cultures examined. The molecular weights found corresponded to approximately 55 and 70 kDa in all positive experiments (Fig. 2). Cross-linking experiments with a biotinylated IL-18-GST protein (44 kDa) on FLS revealed an intense 100-kDa protein complex and two additional, less intense high-molecular-weight bands corresponding to 150–200 kDa (see Fig. 2).

Expression of IL-18Rα in RA-FLS cultures was not detectable upon stimulation with IL-1β, IL-2, IL-12, IL-15, IL-18, IL-12 + IL-18, or TNF-α. In contrast, we found a shift to fewer PCR cycles needed to detect a specific IL-18Rα RT-PCR product, but no increase of IL-18R protein, on IFN-γ stimulation in FLS (n = 6, P < 0.05; Fig. 3).

FLS were exposed to IL-18 in a large number of experiments. We observed some ICAM-1 induction by IL-18 (Δ mean fluorescence intensity (MFI) ≥ 20%) in 5 of 33 FLS cultures (Fig. 4a), but unresponsiveness (Δ MFI <20%) in most of the FLS cultures; in contrast, IL-18 strongly induced ICAM-1 expression in U937 cells (average Δ MFI +159%) (Fig. 4b). In addition, stimulation of FLS cultures with IL-18 neither increased cell proliferation nor upregulated any of the following in the supernatants: VCAM-1 expression, or the concentration of MMP-1, MMP-3, GM-CSF, prostaglandin E₂, or NO. Nitrite levels were close to the detection limit in both groups, and even lower in IL-18-stimulated FLS than in control cultures. For details, see Table 1, and for results in U937 cells and effects in FLS caused by other stimuli, see Supplementary material.

IκB-α was assayed in an IL-18Rα⁺β⁺, an IL-18Rα⁺β^-, and an IL-18Rα^-β^- FLS culture, on the basis of the RT-PCR results. Rapid Ser32 IκB-α phosphorylation and degradation was observed in the cell-protein preparations from IL-18Rα⁺β⁺ FLS (Fig. 5) and U937 cells, but not in FLS cultures, in which the mRNA level of one or both IL-18R chains was below the RT-PCR detection limit.

As is known from lymphocytes and various transformed human cell lines, including fibroblasts, the presence of both IL-18R chains is an important factor in their cellular response limiting the action of IL-18 [14]. In our experiments, many of the RA-FLS cultures studied expressed at least one IL-18R chain, but the PCR analyses considered with the results of the IκB-α assays showed that only 5 of 20 RA-FLS long-term cultures, and none of the OA or SpA FLS cultures, expressed both receptor chains in amounts sufficient for functionality of the receptor complex. For suggested ligand binding of IL-18Rα/β chains, see Supplementary material.

FLS are a heterogeneous cell population [8,21], whose characteristics may even depend on the preparation technique used [22], and we found remarkable differences of IL-18R expression among the FLS cultures as one correlate of their diversity. IκB-α activation was demonstrable only in IL-18Rα⁺β⁺ RA-FLS and in the U937 monocyte cell line. For induced expression of IL-18Rα in FLS, see Supplementary material. Preliminary experiments had shown the expression of an IL-18-binding protein in FLS cultures. Presence of this IL-18 decoy receptor had to be considered as possibly accounting for the missing IL-18 bioactivity in our experiments, but the mRNA levels found for this decoy protein in the FLS cultures elicited no demonstrable IL-18-binding protein immunoreactivity in other FLS culture supernatants (data not shown). Considered in combination with the cross-linking data, this finding shows that IL-18-binding protein is therefore not likely to account for IL-18 resistance of FLS in long-term culture. The lack of IL-18 response in most FLS cultures appears to be based on a rather low or missing constitutive expression of IL-18Rα or IL-18Rβ. Despite relevant IL-18R expression and IL-18 signaling in some FLS cultures, all the cultures we studied were, in contrast to U937 cells, more or less refractory to IL-18 in respect of a postulated upregulation of any disease relevant molecule investigated here. Methodological influences have to be considered in analyzing the results of fluorescence-activated cell sorting (FACS) in FLS [22], but the uniformity of results showing IL-18 resistance for many target molecules in almost all long-term FLS cultures suggests additional, as-yet-unknown IL-18-inhibiting post receptor events in these cells similar to the situation in transformed human fibroblasts [14].

U937 cells, with a constitutive IL-18R expression similar to that found in IL-18α⁺β⁺ FLS, effectively responded to IL-18 with high levels of ICAM-1 expression. Primary SF macrophages are, furthermore, known to be responsive to IL-18 as shown by enhanced TNF-α production [1], and these cells with a capacity for immediate IL-18 response, together with IL-18R⁺ T cells, thus seem responsible for the proinflammatory IL-18 effects in arthritis.