• tumour necrosis factor α
• lipopolysaccharide
• CpG-ODN
• rheumatoid arthritis

The relationship between infections and autoimmune inflammation is of paramount importance in all chronic inflammatory diseases.¹ The crucial step is the interaction between the Toll-like receptor family and bacterial components, which can lead to activation of phagocytic and dendritic cells.^2,3 Whether bacterial components may indeed reactivate autoimmune inflammation is at present still unclear. However, it is well known that oligonucleotides containing unmethylated CpG sequences (CpG-ODN), while stimulating bacterial DNA, might have immunostimulatory activity.³ We have recently presented data suggesting that moderate infections such as upper-lower respiratory or urinary tracts infections may reactivate rheumatoid arthritis (RA) in patients receiving etanercept.⁴

There is evidence in experimental animals^2,3 that Toll-like receptor 4 (activated by lipolysaccharide (LPS)) can induce a cytokine response, whereas Toll-like receptor 9, activated by CpG-ODN, induces mononuclear cell activation in a much more restricted manner. Especially in humans, B cells seem to be the more sensible to bacterial DNA activation, with antibody and autoantibody production.

In this study we were interested to determine whether and how CpG-ODN might stimulate mononuclear cells from patients with RA, using experimental conditions that tried to reproduce the physiological environment at the synovial level (similar percentage of T cells and monocytes).⁵

Five consecutive patients with RA entered the study. All patients gave their written informed consent. All patients had active-aggressive disease as defined by the high number of swollen joints (median 13, range 10–18) and by high C reactive protein values. The disease duration ranged between 2 and 14 years and their disease activity score ranged between 4.7 and 7.5, with a median value of 5.1. All the patients were analysed after a one month period of pharmacological wash out and four were still taking low doses of steroids.

Peripheral blood mononuclear cells were isolated by Ficoll-Paque (Amersham Pharmacia Biotech) density gradient centrifugation and seeded at the 0.25×10⁶/ml in 96 well plastic culture plates overnight at 37°C/5% CO₂ in RPMI 1640 containing 10% fetal bovine serum (Euroclone, Celbio), 100 U/ml of penicillin, and 100 μg/ml streptomycin. Non-adherent cells were removed by repeated washings with phosphate buffered saline. Flow cytometric analysis of the adherent cells using the Simultest (anti-leucocyte/CD45FITC and anti-Leu-M3/CD14PE, Becton and Dickinson) gave the following results: monocytes 44.8 (SEM 2.1)%, lymphocytes 38.2 (1.65)%, granulocytes 12.9 (3.8)%.

Cells were stimulated with 10 ng/ml LPS (from Salmonella typhimurium, L2262, Sigma) and/or 30 μg/ml CpG-ODN 2059 (5′-TCGTCGTTTTGTCGTTTTGTCGTT-3′) or its negative control GpC-ODN 2077 (5′-GCTAGCTTTAGAGCTTTAGAGCTT-3′) for 2, 4, 8, and 24 hours following previously reported procedures.⁶ Tumour necrosis factor α (TNFα) production was evaluated on the supernatants by ultrasensitive (range 0.5–32 pg/ml) enzyme linked immunosorbent assays (ELISAs; Biosource) when analysing ODN effect and by a less sensitive ELISA kit (Euroclone, Celbio) when analysing LPS alone or LPS plus the ODN effect. We reported the results as pg/ml of TNFα per 10⁶ total mononuclear cells initially seeded. All assays were performed in triplicate and the mean value (SE) was taken.

Without stimulation (not treated) we found no significant TNFα production over the entire period of observation, indicating a lack of any significant adhesiveness induced activation on the cells. CpG-ODN as well as its negative control, GpC-ODN, produced only a slight perturbation in TNFα secretion, indicating a non-specific oligonucleotide linked effect (fig 1A). In contrast, we observed a reproducible intense response to LPS incubation, which was not significantly altered in the presence of CpG-ODN (fig 1B). Moreover, again no difference was seen between CpG and GpC-ODN effects, thus indicating that, in our experimental conditions, mononuclear cells do not show any evidence of a specific activation induced by bacterial-like DNA.

Recently, however, it has been shown that dendritic cells may increase CpG-ODN response in monocytes.⁷ Therefore it may well be that the presence of dendritic cells, in vivo as well as in vitro, is essential for bacterial DNA activation of mononuclear cells.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Mononuclear cells from five patients with RA were stimulated with unmethylated CpG-ODN (30 μg/ml), unmethylated GpC-ODN (30 μg/ml), LPS (10 ng/ml), LPS (10 ng/ml) plus unmethylated CpG-ODN (30 μg/ml), or LPS (10 ng/ml) plus unmethylated GpC-ODN (30 μg/ml) and, at the indicated times, TNFα production (pg/ml/10⁶ cells) was monitored by ELISA (Euroclone, Celbio). Results are shown as means (SE). TNFα kinetic production was evaluated also with cell culture medium alone.