Expression and regulation of CCL18 in synovial fluid neutrophils of patients with rheumatoid arthritis

Abstract

Rheumatoid arthritis (RA) is characterized by the recruitment of leukocytes and the accumulation of inflammatory mediators within the synovial compartment. Release of the chemokine CCL18 has been widely attributed to antigen-presenting cells, including macrophages and dendritic cells. This study investigates the production of CCL18 in polymorphonuclear neutrophils (PMN), the predominant cell type recruited into synovial fluid (SF). Microarray analysis, semiquantitative and quantitative reverse transcriptase polymerase chain reaction identified SF PMN from patients with RA as a novel source for CCL18 in diseased joints. Highly upregulated expression of other chemokine genes was observed for CCL3, CXCL8 and CXCL10, whereas CCL21 was downregulated. The chemokine receptor genes were differentially expressed, with upregulation of CXCR4, CCRL2 and CCR5 and downregulation of CXCR1 and CXCR2. In cell culture experiments, expression of CCL18 mRNA in blood PMN was induced by tumor necrosis factor alpha, whereas synthesis of CCL18 protein required additional stimulation with a combination of IL-10 and vitamin D3. In comparison, recruited SF PMN from patients with RA were sensitized for CCL18 production, because IL-10 alone was sufficient to induce CCL18 release. These results suggest a release of the T cell-attracting CCL18 by PMN when recruited to diseased joints. However, its production is tightly regulated at the levels of mRNA expression and protein synthesis.

Figure 1

Microarray analysis of chemokine gene expression in synovial fluid polymorphonuclear neutrophils. RNA of SF synovial fluid polymorphonuclear neutrophils (SF PMN) from nine patients with rheumatoid arthritis (RA) was analyzed for the expression of CC chemokines (a) and CXC, C and CX3C chemokines (b). Data obtained by Gene Pix™ Analysis Software were normalized, transformed and denoted as x-fold regulation versus the expression of blood PMN from healthy donors. Bars represent the median expression between nine RNA samples. Genes with median expression ratios less than -1 or more than +1 were significantly regulated. (c) SF PMN from one representative patient with RA (no. 2) were subjected to flow cytometry with fluorescein isothiocyanate-conjugated anti-CD66b, allophycocyanin (APC)-conjugated anti-CD56, phycoerythrin (PE)-conjugated anti-CD3, anti-CD19-APC and anti-CD14-APC antibodies.

Figure 2

Semiquantitative and quantitative RT-PCR analysis of CCL18 mRNA in polymorphonuclear neutrophils. (a) Total RNA from synovial fluid polymorphonuclear neutrophils (SF PMN) of rheumatoid arthritis (RA) patients nos 1, 2, 7 and 8 and from blood PMN from healthy donor no. 2 was amplified by semiquantitative RT-PCR with primers for CCL18, CXCL10 and actin and subjected to agarose gel electrophoresis. PCR was repeated twice. (b) Total RNA from blood and SF PMN from patients with RA (n = 9) and from blood PMN from healthy donors (n = 4) were subjected to quantitative RT-PCR with primers for CCL18 and HPRT. CCL18 transcript levels are presented as relative expression ratios. Bars represent median expression between RNA samples.

Figure 3

Microarray analysis of chemokine receptor gene expression in synovial fluid polymorphonuclear neutrophils. (a) RNA of synovial fluid polymorphonuclear neutrophils (SF PMN) from nine patients with rheumatoid arthritis (RA) was analyzed for chemokine receptor expression. Data obtained by Gene Pix™ Analysis Software were normalized, transformed and denoted as x-fold regulation versus expression of blood PMN from healthy donors. Bars represent median expression between nine RNA samples. Genes with median expression ratios less than -1 or more than +1 were significantly regulated. (b) Total RNA from SF PMN from RA patients nos 1, 2, 7 and 8 and from blood PMN from healthy donor no. 2 was amplified by semiquantitative RT-PCR with primers for CCRL2, CXCR4 and actin and subjected to agarose gel electrophoresis. PCR was repeated twice.

Figure 4

Induction of CCL18 mRNA and protein in synovial fluid and blood polymorphonuclear neutrophils. (a) Synovial fluid polymorphonuclear neutrophils (SF PMN) of patients with rheumatoid arthritis (RA) and blood PMN from healthy donors and patients with RA were incubated with 20 ng/ml IL-10, 10^-7 M vitamin D₃ and 10 ng/ml TNF-α for 48 hours. Levels of CCL18 in the supernatant were determined by ELISA. Data represent the geometric mean ± SEM for three independent experiments performed in duplicate. (b) PMN from healthy donors were incubated for 24 hours with various amounts of recombinant TNF-α. Total RNA was amplified by semiquantitative RT-PCR with primers for CCL18 and actin and subjected to agarose gel electrophoresis. This result is representative of three independent experiments with PMN from three different donors. (c) Blood PMN from healthy donors were incubated in the presence or absence of SF from patients with RA for 10 hours, then washed twice and incubated for a further 48 hours with 20 ng/ml IL-10 and 10^-7 M vitamin D₃. Data represent the geometric mean ± SEM for three independent experiments performed in duplicate.

Figure 5

CCL18 induction of polymorphonuclear neutrophils co-cultured with endothelial cells requires cell–cell contact and TNF-α. (a) CCL18 levels were measured by ELISA in culture supernatants of 5 × 10⁵ peripheral blood mononuclear cells, 5 × 10⁶ polymorphonuclear neutrophils (PMN) or 5 × 10⁵ EA-hy.926 cells alone or after co-culture with EA-hy.926 cells after incubation for 48 hours. Direct cell–cell contact of PMN and EA-hy.926 cells was prevented by Boyden chambers (bc) as indicated. Data represent the geometric mean ± SEM of CCL18 measured in four independent experiments performed in duplicate. (b) CCL18 levels were measured in culture supernatants of PMN and EA-hy.926 cells alone or after co-culture of both cell types after incubation for 48 hours. Cultures were supplemented with anti-TNF-α antibodies (infliximab; 50 μg/ml) or 10 ng/ml TNF-α as indicated. Data represent the geometric mean ± SEM of three independent experiments performed in duplicate. (c) RNA was prepared from PMN and EA-hy.926 cells alone or after co-culture of both cell types after incubation for 24 hours. Cultures were supplemented with TNF-α (10 ng/ml) as indicated. RNA samples were amplified by semiquantitative RT-PCR with primers for CCL18 and actin and subjected to gel electrophoresis. This result is representative of three independent experiments with PMN from three different donors. (d) TNF-α levels were measured by ELISA in culture supernatants of PMN and EA-hy.926 cells alone and in co-cultures of PMN and EA-hy.926 cells before or after γ-irradiation (40 Gy) of one of these cell types. The irradiated cell type is marked with an asterisk. Data represent the geometric mean ± SEM for three independent experiments performed in duplicate.