A proinflammatory role for IL-18 in rheumatoid arthritis

Abstract

IL-18 is a novel cytokine with pleiotropic activities critical to the development of T-helper 1 (Th1) responses. We detected IL-18 mRNA and protein within rheumatoid arthritis (RA) synovial tissues in significantly higher levels than in osteoarthritis controls. Similarly, IL-18 receptor expression was detected on synovial lymphocytes and macrophages. Together with IL-12 or IL-15, IL-18 induced significant IFN-gamma production by synovial tissues in vitro. IL-18 independently promoted GM-CSF and nitric oxide production, and it induced significant TNF-alpha synthesis by CD14(+) macrophages in synovial cultures; the latter effect was potentiated by IL-12 or IL-15. TNF-alpha and IFN-gamma synthesis was suppressed by IL-10 and TGF-beta. IL-18 production in primary synovial cultures and purified synovial fibroblasts was, in turn, upregulated by TNF-alpha and IL-1beta, suggesting that monokine expression can feed back to promote Th1 cell development in synovial membrane. Finally, IL-18 administration to collagen/incomplete Freund's adjuvant-immunized DBA/1 mice facilitated the development of an erosive, inflammatory arthritis, suggesting that IL-18 can be proinflammatory in vivo. Together, these data indicate that synergistic combinations of IL-18, IL-12, and IL-15 may be of importance in sustaining both Th1 responses and monokine production in RA.

Figure 1

(a) IL-18 mRNA was detected in RA and OA synovial membranes by RT-PCR (representative of 20 RA and 10 OA tissues). (b) Significantly higher levels of IL-18 were detected in RA SF than in sera (18 matched samples assayed after removal of rheumatoid factor).

Figure 2

(a) IL-18 was detected immunochemically (brown) in synovial membrane using anti–IL-18 antibody (D3B6). (b) CD3 localization (brown) in parallel sections indicated IL-18 expression in T cell–rich aggregates. Double staining with anti–IL-18 (blue/black) and (c) anti-CD3 (brown) or (d) anti-CD68 (brown) antibodies, without nuclear counterstain, localized IL-18 mainly to CD68⁺ macrophages. (Black arrow, double positive; white arrow, IL-18 single positive).

Figure 3

(a) IL-18 receptor mRNA expression was detected by RT-PCR in 10 of 10 RA synovial membranes (5 representative data are shown in lanes 1–5). LPS-stimulated THP-1 cells served as positive control. (b) Pooled data (mean ± SEM) to characterize IL-18R⁺ cells in RA synovial fluid mononuclear cells (n = 8 patients).

Figure 4

Recombinant IL-18 (100 ng/mL), IL-12 (5 ng/mL), and/or IL-15 (100 ng/mL) were added to SF (a) and synovial membrane (b) cultures for 48 hours in the presence (empty boxes) or absence of IL-10 (20 ng/mL)/TGF-β (5 ng/mL) (filled boxes), and IFN-γ production was determined by ELISA. Data (mean ± SEM) are pooled from 10 individual RA patient samples.

Figure 5

(a) TNF-α production in cultures, identical to those described in Figure 4, was determined by ELISA. Data (mean ± SEM) are pooled from 10 individual RA patients, each cultured in triplicate. (b) Enhanced intracellular TNF-α expression in CD14⁺ monocytes within synovial fluid mononuclear cells measured by FACS after addition of 100 ng/mL IL-18. Data are representative of 10 similar experiments. Isotype-matched control antibody staining was less than 1%.

Figure 6

(a) IL-18 from synovial membrane cultures representative of 4 positive cultures: 2 were negative. (b) Upregulation of IL-18 mRNA in SFB cell lines (3rd passage, n = 10 RA patients) by TNF-α and IL-1β in vitro. Immunohistological localization in SFBs using D3B6 mAb showed that IL-18 was present after (c), but not before (d), stimulation with TNF-α and IL-1β.