Article Text

Download PDFPDF

Concise report
Abatacept modulates proinflammatory macrophage responses upon cytokine-activated T cell and Toll-like receptor ligand stimulation
  1. M H Wenink1,
  2. K C M Santegoets1,
  3. A M Platt2,
  4. W B van den Berg1,
  5. P L C M van Riel1,
  6. P Garside2,
  7. T R D J Radstake1,
  8. I B McInnes2
  1. 1Department of Rheumatology, Nijmegen Center of Infection, Inflammation and Immunity (N4i) and Nijmegen Center for Molecular Life Science (NCMLS), Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands
  2. 2Division of Immunology, Infection and Inflammation, Glasgow Biomedical Research Centre, University of Glasgow, Glasgow, UK
  1. Correspondence to Dr T R D J Radstake, Department of Rheumatology, Nijmegen Center of Infection, Inflammation and Immunity (N4i) and Nijmegen Center for Molecular Life Science (NCMLS), Radboud University Nijmegen Medical Centre, Geert Grooteplein 8, 6500 HB Nijmegen, The Netherlands; t.radstake{at}reuma.umcn.nl

Abstract

Objectives We investigated whether Abatacept might reduce proinflammatory cytokine production by macrophages upon contact with cytokine activated T cells and/or stimulation with TLR ligands.

Methods Macrophages and cytokine stimulated T cells (Tck) were added together in the presence of Abatacept or a control Ig, with or without TLR ligands. The production of cytokines was determined by luminex.

Results Abatacept reduced Tck-induced production of TNFa by macrophages. Tck and TLR ligands synergistically induced the production of proinflammatory cytokines by macrophages, especially IL-12p70. The production of IL-12p70 coincided with the production of IFNg, which were both reduced in the presence of Abatacept.

Conclusions Tck induce the production of TNFa by macrophages and facilitate the highly increased production of proinflammatory cytokines in the presence of TLR ligands. Abatacept was shown to potently suppress these pathways suggesting that its role may extend beyond antigen specific T cell mediated effector function.

View Full Text

Statistics from Altmetric.com

Introduction

One of the major pathways underlying the pathological process in rheumatoid arthritis (RA) is the aberrant production of proinflammatory cytokines by macrophages. RA synovium is highly enriched with infiltrating CD68+CD163– macrophages, and their decrease upon treatment is correlated with changes in disease activity.1 Monocyte-derived CD163 macrophages differentiated in vitro in the presence of granulocyte–macrophage colony-stimulating factor (mϕ-1) exhib it the same characteristics as these infiltrating macrophages.2

One crucial pathway mediating synovial macrophage activation is that driven by a direct interaction with activated T cells. In particular, cytokine-activated T cells recapitulate the functional properties of synovial T cells4. This depends on the cell–cell contact between memory CD4+ CD45RO+ T cells and macrophages via CD69, CD18 and CD49d and was independent of the production of soluble mediators.3 CD4+CD45RO+ T cells are the main infiltrating cells in the pannus, underlining their pivotal role in the pathogenesis of RA. Brennan et al4 demonstrated that human peripheral blood lymphocytes, when cultured in the presence of interleukin (IL) 2, IL-6 and tumour necrosis factor α (TNFα) (cytokine-stimulated T cells (Tck)), display the same characteristics as activated synovial T cells.

Another mechanism whereby cytokine production is induced in macrophages is the ligation of Toll-like receptors (TLRs). Many endogenous ligands for TLRs have been found in the arthritic joint, such as the small heat shock protein B8, which activates TLR4, and self-RNA and self-DNA, which more than likely activate cells in the joint via TLR3 and TLR8, respectively.5,,8 Combinations of TLR ligands with T-cell-derived signals such as CD40 ligand or interferon γ (IFNγ) have been demonstrated to be potent in the induction of proinflammatory cytokine production by macrophages. Since Tck strongly resemble the T cells present in synovitis, it is of importance to know whether these cells are indeed capable of displaying synergistic effects with TLR ligands, thereby contributing to the deranged cytokine profile found in the RA synovial cavity.9

Abatacept (CTLA4 with an Fc tail) is a new agent in the armamentarium against RA. It blocks the instruction of T cells by antigen-presenting cells such as dendritic cells (DCs) and macrophages by binding to CD80 and CD86 on the antigen-presenting cells, thereby preventing binding to CD28 on the T cell, which is necessary for full T cell activation. We recently demonstrated in a murine system that abatacept also modulates follicular helper T cell maturation.10 Memory CD4+CD45RO+ T cells possess significant intracellular reservoirs of CTLA4, which, upon stimulation, are transported readily to the cell surface. Activation of mouse DCs via CD80/86 by CTLA4-expressing T cells resulted in the production of IFNγ by the DCs.11 It was recently suggested that CTLA4-Ig, via a direct effect on macrophages, reduces the ability of macrophages to produce cytokines upon stimulation with concanavalin-A-activated Jurkat T cells.12 We therefore investigated whether abatacept could influence the RA disease course by diminishing the proinflammatory cytokine production by macrophages induced upon contact with cytokine-activated T cells and TLR ligands.

Materials and methods

Culture of monocyte-derived type I macrophages and Tck

Peripheral blood mononuclear cells were isolated from buffy coats in healthy blood donors by using density-gradient centrifugation over Histopaque (GE Healthcare UK). The local medical ethics committee approved the study protocol. Monocytes and CD4+ T cells were obtained using CD14 or CD4 microbeads, respectively, and AutoMACS (Miltenyi Biotec: Bergisch Gladbach, Germany). Mϕ-1 were generated by culturing isolated monocytes in the presence of granulocyte–macrophage colony-stimulating factor (800 U/ml) for 6 days. Mϕ-1 were cultured in six-well plates at a concentration of 5×105 in 2 ml of culture medium (RPMI 1640, 10% fetal calf serum, penicillin/streptomycin and l-glutamine; Gibco, Grand Island, New York, USA). Fresh culture medium (1 ml) with the same supplements was added on day 3, after which mϕ-1 were harvested on day 6. In parallel, CD4+ T cells from the same donor were cultured in a complete medium with recombinant human IL-2 (25 ng/ml), IL-6 (100 ng/ml) and TNFα (25 ng/ml) (all from R&D: Minneapolis, Minnesota, USA) at a density of 2×106/ml for 6 days, after which they were also harvested.

Reagents

Abatacept was provided by Bristol-Myers Squibb. Chi L6 is a chimeric fusion protein consisting of the V region of murine L6 antigen, combined with a human IgG1 C region; it was used as a control fusion protein (control Ig) in these studies (Bristol-Myers Squibb: Uxbridge, UK). The proteins were used at a final concentration of 50 μg/ml.

Stimulation of monocyte-derived mϕ-1

Harvested day 6 mϕ-1 were washed, counted and plated at a concentration of 5×104 cells/well in 96-well culture plates. For cell contact assay after extensive washing, 2×105 cytokine-activated T cells were added to mϕ-1 in the absence or in the presence of abatacept or the control protein at a concentration of 50 μg/ml. In some experiments, Tck fixed in 4% paraformaldehyde were used. After 24 h, the supernatants were collected for analysis of cytokine levels. For TLR stimulation, TLR agonists were added to mϕ-1 either in the presence or in the absence of Tck and/or abatacept/control protein and, after 24 h, the supernatants were harvested. lipopolysaccharide was used at a concentration of 1 ng/ml (Escherichia coli 0111:B4; Sigma-Aldrich: St Louis, MO, USA), and R848 was used at a concentration of 1 μg/ml (Invivogen: San Diego, CA).

Measurement of cytokines in culture supernatants

Levels of IFNγ, TNFα, IL-12p70, IL-6, IL-10, IL-13 and IL-17 in the supernatants were measured using commercially available kits (BioSource International, Camarillo, California, USA) according to the manufacturer's instructions. Cytokine levels were measured and analysed with the Bio-Plex system (Bio-Rad: Hemel Hempstead, UK).

Statistical analysis

Differences were analysed with Student t test to compare two stimulations. p Values less than 0.05 were considered significant.

Results

After 6 days of differentiation, syngeneic mϕ-1 and Tck cells were harvested, counted and replated together in 96-well plates for cell contact assays. Mϕ-1 readily produced TNFα upon exposure to Tck, while IL-6 and IL-10 were virtually not released (figure 1A and data not shown). This effect was at least partly mediated by cell–cell contact, as paraformaldehyde-fixed Tck, which are incapable of secreting cytokines, displayed stimulatory capacity, albeit lower than that observed in live cell cocultures (figure 1B). Non-cytokine-stimulated CD4+ T cells did not induce TNFα production by mϕ-1 (data not shown). Next, we evaluated whether abatacept might interfere with the production of TNFα by mϕ-1 upon contact with Tck. Blocking of CD80/86 by abatacept resulted in a markedly decreased production of TNFα, uncovering a novel manner by which abatacept might interfere in the RA disease process (figure 1C). To additionally study the potential effect of Tck on TLR-mediated stimulation of mϕ-1, we used ligands for TLR4 and TLR7/8, together with Tck, to determine whether they would cooperate in the production of proinflammatory cytokines by mϕ-1. Synergistic release of TNFα, especially of IL-6, was observed when mϕ-1 were exposed to a TLR4 or TLR7/8 ligand together with Tck, while non-cytokine-stimulated CD4+ T cells did not induce any effect (figure 2A,B and data not shown). Abatacept did not significantly reduce the production of TNFα or IL-6 under these circumstances (data not shown).

Figure 1

Mϕ-1 readily produce tumour necrosis factor α (TNFα) upon exposure to cytokine-stimulated T cells (Tck), which are inhibited by abatacept. (A) Mϕ-1 were stimulated with Tck for 24 h, after which the supernatants were measured for the presence of TNFα. Data are from triplicate wells and are representative of five individual experiments with similar results. Bars are mean and SD. (B) Mϕ-1 were stimulated with Tck or Tck fixed in paraformaldehyde for 24 h, after which the supernatants were measured for the presence of TNFα. Data are from triplicate wells and are representative of three individual experiments with similar results. Bars are mean and SD. (C) Mϕ-1 were stimulated with Tck with medium, a control Ig (20 μg/ml) or abatacept (20 μg/ml) for 24 h, after which the supernatants were measured for the presence of TNFα. Data are from five individual experiments. Bars are mean and SEM. *p<0.05 compared to control Ig.

Figure 2

The combination of cytokine-stimulated T cells (Tck) and Toll-like receptor (TLR) ligands induces a clearly increased production of tumour necrosis factor α (TNFα) and interleukin (IL) 6 by mϕ-1. Mϕ-1 were unstimulated or stimulated with Tck or lipopolysaccharide (A), R848 (B) or a combination of both. After 24 h, the supernatants were collected and measured for their TNFα (left panels) and IL-6 (right panels) content. Data are from five individual experiments. Bars are mean and SEM. *p<0.05 compared to stimulation with TLR ligands alone.

The IL-12p70/IFNγ axis appears to play an important role in the pathogenesis of RA13 14; therefore, we aimed to determine whether abatacept might additionally modulate this pathway. Stimulation of mϕ-1 with either TLR ligands or Tck alone induced the production of marginal levels of IL-12p70 and IFNγ. However, stimulation with either TLR4 or TLR7/8 ligands, in combination with Tck, resulted in a clear release of IL-12p70 and IFNγ (figure 3A,B and data not shown). We additionally measured IL-13 and IL-17 (distinct Th2 and Th17 cytokines, respectively) and found that these cytokines were not produced (IL-17) or were produced at very low levels (IL-13) (data not shown). Blockade of CD80/86 by abatacept resulted in a significant reduction in the production of IL-12p70 and IFNγ (figure 3C,D). Abatacept had no effect on cytokine production induced by TLR ligands alone, providing evidence that the effect of abatacept is not due to inhibitory signalling through CD80/86 but is due to the blockade of Tck/mϕ-1 interactions (data not shown).

Figure 3

The combination of cytokine-stimulated T cells (Tck) and Toll-like receptor ligands licenses Tck/mϕ-1 cocultures to produce interleukin (IL) 12p70 and interferon γ (IFNγ), dampened by the presence of abatacept. Mϕ-1 were unstimulated or stimulated with Tck, lipopolysaccharide (LPS) or a combination of both. After 24 h, the supernatants were collected and measured for their (A) IL-12p70 and (B) IFNγ content. Data are from four individual experiments. Bars are mean and SEM. Mϕ-1 were stimulated with Tck and LPS in the presence of the control Ig (grey bars) or abatacept (black bars). After 24 h, the supernatants were collected and measured for their IL-12p70 (C) and IFNγ (D) content. Data are from four individual experiments. Bars are mean and SEM. *p<0.05 compared to control Ig.

Discussion

Abatacept is a valuable new drug in the treatment of RA, but its mode of action remains to be fully elucidated. We provide evidence that, in addition to inhibiting T cell activation and reducing the migration of T cells into B cell follicles,10 abatacept inhibits the Tck-induced production of proinflammatory cytokines by inflammatory CD68+CD163– macrophages.

Various pathways have been implicated in the pathogenesis of the dysregulated proinflammatory cytokine environment in RA. We demonstrate herein that the combination of TLR ligands and cytokine-activated T cells results in the synergistic release of TNFα and IL-6 and, importantly, licenses macrophages for the production of IL-12p70.

In a recent study, it was demonstrated that abatacept reduces the inflammation of the synovium without disrupting cellular constituents in patients unresponsive to anti-TNFα treatment. Interestingly, especially the expression of IFNγ was reduced by 52%.14 We provide a mechanistic explanation for this observation. We observed that abatacept reduces the production of IL-12p70 and IFNγ by Tck/mϕ-1 cocultures activated by a TLR ligand by approximately 50%, whereas the production of TNFα and IL-6 was unaffected. Th1 cells and natural killer T cells are capable of producing copious amounts of IFNγ, and IL-12p70 is crucial in both the differentiation of Th1 cells and the induction of IFNγ production by natural killer T cells.15,,17 The IL-12p70/IFNγ axis was recently demonstrated, using the K/BxN serum transfer model and IL-12p35−/− mice, to promote antibody-induced joint inflammation, underscoring the role played by IL-12p70 in arthritis.13 The production of IFNγ by macrophage/Tck cocultures upon activation with TLR ligands might explain our results regarding the highly increased release of IL-12p70. IFNγ is known to potently increase the release of IL-12p70 by macrophages upon TLR stimulation, while IL-12p70 might, in its turn, be important in the induction of IFNγ by Tck. This might lead to a vicious circle that increases the production of both IL-12p70 and IFNγ, which can be prevented by the presence of abatacept.

The question remains on how abatacept is capable of inhibiting the release of cytokines by proinflammatory macrophages upon activation by Tck. Abatacept might directly activate certain pathways in macrophages via CD80/86, resulting in the inhibition of cytokine release upon Tck stimulation, as was suggested by Cutolo et al.12 They demonstrated, using a concanavalin-A-activated Jurkat T cell line, that abatacept was capable of reducing the (immunocytochemically) detectable levels of IL-6 and TNFα in RA synovial tissue macrophages, underscoring the clinical relevance of the results described in this paper. A direct effect of abatacept on the macrophages would, however, be a rather specific inhibitory effect since no difference in cytokine release was observed upon TLR stimulation. Activation of CD80/86 by CTLA4-expressing T cells was demonstrated to induce IFNγ production by mouse DCs.11 It thus appears likely that the activation of CD80/86 by CD28 or CTLA4 expressed by Tck plays a role in the cytokine production by the macrophages, which is inhibited by abatacept. In addition, steric hindrance might play a role by inhibiting full contact in the immunological synapse between necessary components, leading to a full instruction of the macrophages by Tck. Whereas we did not add an antibody that specifically binds a receptor known to play a role in this process, Brennan et al4 did use a blocking antibody against the β2 integrin lymphocyte function-associated antigen 1 in experiments with Tck and macrophages. This had no effect on the level of TNFα produced, substantiating that a role for steric hindrance might be minimal.

Another point of interest was that we observed that stimulation with Tck, together with TLR ligands, hampered the ability of abatacept to interfere with the production of TNFα and IL-6. This appears to implicate that TLR signalling is capable of over-ruling some of the proinflammatory signalling cascades induced by Tck in mϕ-1, leading to a full-blown cytokine release despite the presence of abatacept.

In conclusion, we demonstrated that abatacept is capable of interfering with the production of TNFα by mϕ-1 induced by contact with cytokine-activated T cells. Moreover, in the presence of TLR ligands, which are present and plentiful in the arthritic joint, activated T cell/mϕ-1 cocultures produced highly increased levels of TNFα, IL-6 and even IL-12p70 and IFNγ. Abatacept was shown to potently interfere with the production of IL-12p70 and IFNγ under these circumstances.

Acknowledgments

TRDJR was sponsored by a VIDI laureate from the Dutch Association of Research (NWO). MHW was funded by an AGIKO grant from the Radboud University Nijmegen Medical Centre and in part within the framework of project D1-101 of Top Institute Pharma.

References

View Abstract

Footnotes

  • Competing interests None.

  • Ethics approval The study was approved by the Radboud University Nijmegen Medical Centre medical ethics committee and the University of Glasgow medical ethics committee.

  • Provenance and peer review Not commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.