Expression of toll-like receptor 2 on CD16+ blood monocytes and synovial tissue macrophages in rheumatoid arthritis

Authors


Abstract

Objective

CD16 (IgG Fcγ receptor type IIIA [FcγRIIIA])–expressing CD14+ monocytes express high levels of Toll-like receptor 2 (TLR-2) and are able to efficiently produce proinflammatory cytokines such as tumor necrosis factor α (TNFα). To understand the role of CD16 and TLR-2 in monocyte and macrophage activation in rheumatoid arthritis (RA), we investigated the expression of TLR-2 on CD16+ blood monocytes and synovial tissue macrophages and the effect of CD16 and TLR-2 activation on cytokine production.

Methods

The expression of CD14, CD16, TLR-2, and TLR-4 on blood monocytes was measured by flow cytometric analysis. CD16 and TLR-2 expression in RA synovial tissue was detected by 2-color immunofluorescence labeling. CD16+ mature monocytes were prepared by incubating blood monocytes in plastic plates for 24 hours. These adhered monocytes were stimulated with lipoteichoic acid (LTA), anti-FcγRIII antibody, and Hsp60 for 5 hours, and culture supernatants were measured for various cytokines by immunoassay. The activation of NF-κB was detected by electrophoretic mobility shift assay.

Results

The frequency of CD16+ cells in all blood monocytes was significantly increased in patients with RA compared with healthy controls. TLR-2 was expressed at higher levels on CD16+ monocytes than on CD16− monocytes, while TLR-4 was expressed similarly on both monocytes. In RA synovial tissue, CD16+/TLR-2+ cells were distributed mainly in the lining layer. TLR-2 expression on monocytes was enhanced by macrophage colony-stimulating factor (M-CSF) and interleukin-10 (IL-10), but was reduced by transforming growth factor β1, while CD16 expression was inducible by these cytokines. Adhered monocytes (∼50% CD16+) produced TNFα, IL-1β, IL-6, IL-8, IL-12 p40, IL-1 receptor antagonist, and IL-10 after LTA stimulation. This cytokine response was inhibited significantly by anti–TLR-2 antibody and partly by anti–TLR-4 antibody. Anti-FcγRIII antibody stimulation markedly enhanced the LTA-induced TNFα response. Hsp60 could stimulate TNFα production by adhered monocytes, which was inhibited similarly by anti–TLR-2 antibody and anti–TLR-4 antibody. NF-κB activation in adhered monocytes was induced by LTA, but this NF-κB activity was not augmented by anti-FcγRIII antibody stimulation.

Conclusion

These results suggest that CD16+ monocytes and synovial tissue macrophages with high TLR-2 expression may be induced by M-CSF and IL-10, and their production of TNFα could be simulated by endogenous TLR ligands such as Hsp60 and FcγRIIIA ligation by small immune complexes in RA joints.

The Toll-like receptors (TLRs) have been identified as a new group of receptors, by means of which cells of the innate immune system, including macrophages and dendritic cells, recognize microbial pathogens (1–3). The engagement of TLRs by pathogen-derived ligands rapidly initiates a signaling pathway that activates the transcription factor NF-κB and the MAPK cascade, leading to the induction of proinflammatory cytokines and costimulatory molecules (4). Ten different proteins have so far been defined as members of the human TLR family. The motif pathogen-associated molecular patterns such as lipopolysaccharide (LPS) are recognized by TLR-4 (5), and those such as lipoteichoic acid (LTA) and peptidoglycan are recognized mainly by TLR-2 (6). In addition, it has recently been shown that endogenous molecules such as Hsp60 and fragmentation products of fibronectin and hyaluronic acid can trigger an inflammatory response via TLR-2 and TLR-4 (7–10). Because these molecules are released from stressed or necrotic cells or after enzymatic degradation of endogenous macromolecules, it was proposed that these TLRs might function in the biologic system as surveillance receptors that monitor tissues for disease states (11).

Rheumatoid arthritis (RA) is a chronic inflammatory disease that primarily affects multiple synovial joints. Synovial tissue macrophages contribute to both chronic inflammation and joint destruction principally through the elaboration of proinflammatory cytokines, including tumor necrosis factor α (TNFα), interleukin-1 (IL-1), IL-8, and granulocyte–macrophage colony-stimulating factor (GM-CSF) (12). In particular, TNFα is now recognized as a master cytokine that governs the disease process by activating macrophages and other cell types to induce numerous proinflammatory mediators (13). The significance of this cytokine has been well proven by the clinical efficacy of its blockade in RA patients with active disease (14).

In association with joint inflammation, circulating monocytes become activated before entry into the joint in RA (12). We previously demonstrated that CD16+ monocytes with low CD14 expression are increased in patients with active RA (15). CD14+,CD16+ cells were defined as a subset of monocytes that account for ∼10% of all monocytes in healthy individuals (16). The CD16+ monocytes have been termed the proinflammatory type, because compared with the classic CD14++,CD16− monocytes, these cells express higher levels of class II major histocompatibility complex (MHC) molecules, adhesion molecules, and chemokine receptors and are capable of more efficiently producing proinflammatory cytokines, but not the antiinflammatory cytokine IL-10 (17). Consistent with this concept, CD16+ monocytes are expanded rapidly in severe infection, and cytokines such as M-CSF, IL-10, and transforming growth factor β (TGFβ) have been implicated in the development of the CD16+ phenotype (17).

The CD16 cell surface molecule induced on monocytes and macrophages is the IgG Fcγ receptor type IIIA (FcγRIIIA), which plays a dominant role in the inflammatory responses triggered by IgG-containing small immune complexes (18). It has been demonstrated that ligation of FcγRIIIA but not of FcγRI or FcγRII, using their specific antibodies, can induce TNFα and IL-1α production from human macrophages (19). In this regard, CD16+ macrophages have been found most densely in the synovial lining layer (20), where TNFα and IL-1 are produced abundantly (13). Thus, the induction and activation of CD16 in macrophages appear to be important in the pathogenesis of RA.

Recent studies have shown that CD16+ monocytes express higher levels of TLR-2 and produce larger amounts of TNFα in response to the TLR-2–specific bacterial lipopeptide when compared with CD16− monocytes, suggesting the relevance of TLR-2 to CD16+ monocyte activation (21). To learn the role of CD16 and TLR-2 in monocyte and macrophage activation in RA, we investigated the expression of TLR-2 on monocytes and synovial tissue macrophages in patients with RA; TLR-2 induction on monocytes activated by M-CSF, IL-10, and TGFβ1; the effect of TLR-2 and FcγRIIIA activation on production of various cytokines and NF-κB activation in CD16+ mature monocytes; and the effect of Hsp60 stimulation on the production of TNFα.

PATIENTS AND METHODS

Patients and samples.

The total patient population consisted of 40 patients with RA (31 women and 9 men; mean ± SD age 53.5 ± 10.1 years) diagnosed according to the revised 1987 criteria of the American College of Rheumatology (formally, the American Rheumatism Association) (22). Most patients were receiving nonsteroidal antiinflammatory drugs, prednisolone (≤7.5 mg/day), and disease-modifying antirheumatic drugs. The mean (±SD) values for clinical parameters in the study patients (n = 35) were as follows: erythrocyte sedimentation rate 59 ± 36 mm/hour, serum C-reactive protein level 18.9 ± 25.6 mg/liter, and IgM class rheumatoid factor (RF) titer 144 ± 158 units/ml. Twenty healthy volunteers (16 women and 4 men) matched for age (50.3 ± 8.5 years) served as controls. Synovial tissue samples were obtained from patients with RA, at the time of surgical treatment. All patients gave informed consent.

Flow cytometry.

Peripheral blood mononuclear cells (PBMCs) were prepared from heparinized blood samples by centrifugation over Ficoll-Hypaque density gradients (Pharmacia, Uppsala, Sweden). Cells were washed well with RPMI 1640 medium (Life Technologies, Gaithersburg, MD), and were resuspended in phosphate buffered saline (PBS) with 1% heat-inactivated fetal calf serum (FCS; Life Technologies). Cell surface expression of CD14, CD16, CD64, TLR-2, and TLR-4 was analyzed by cell surface staining and flow cytometric analysis, as described previously (15). PBMC suspensions were incubated with biotinylated mouse anti–TLR-2 monoclonal antibody (mAb) (TL2.1; eBioscience, San Diego CA), anti–TLR-4 mAb (HTA125; eBioscience), or isotype-matched control mAb, followed by incubation with R-phycoerythrin (R-PE)/cyanin 5.1–conjugated streptavidin (Dako, Kyoto, Japan). Cells were washed and then incubated with fluorescein isothiocyanate (FITC)–conjugated anti-CD14 mAb (Leu M3; Becton Dickinson, Franklin Lakes, NJ) or anti-CD64 mAb (Immunotech, Marseilles, France) and PE-conjugated anti-CD16 mAb (Leu11; Becton Dickinson). After washing, cells were resuspended in 1% FCS/PBS. Analysis was performed on a FACScan flow cytometer (Becton Dickinson), and the monocytes were specifically analyzed by selective gating based on parameters of forward and side light scatter.

Two-color immunofluorescence labeling.

Cryostat sections (4 μm) from RA synovial tissues were fixed in acetone and blocked with 10% goat serum for 30 minutes. Double immunofluorescence was performed by serially incubating sections with 10 μg/ml of mouse IgG1 anti-CD16 mAb (3G8; Immunotech) or anti-CD3 mAb (UCTH1; Dako, Glostrup, Denmark), mouse IgG2a anti–TLR-2 mAb (TL2.1; eBioscience), and isotype control mAb at 4°C overnight, followed by incubation with rhodamine-conjugated goat anti-mouse IgG1 mAb (Southern Biotechnology Associates, Birmingham, AL) and FITC-conjugated goat anti-mouse IgG2a mAb (Jackson ImmunoResearch, West Grove, PA) for 30 minutes at room temperature. The double immunofluorescence of sections was examined with an LSM510 inverted laser-scanning confocal microscope (Zeiss, Jena, Germany) and illuminated with 488 nm and 568 nm of light. Images decorated with FITC and rhodamine were recorded simultaneously through separate optical detectors with a 530-nm band-pass filter and a 590-nm long-pass filter, respectively. Pairs of images were superimposed for colocalization analysis.

Isolation and culture of blood monocytes.

PBMCs were resuspended at a density of 5 × 106 cells in culture medium (RPMI 1640 medium supplemented with 25 mM HEPES, 2 mML-glutamine, 2% nonessential amino acids, 100 units/ml penicillin, and 100 μg/ml streptomycin) with 10% FCS. The cell suspensions were incubated in 6-well plates (Corning, Corning, NY) at 37°C for 2 hours in a humidified atmosphere containing 5% CO2. After nonadherent cells were removed, adherent monocytes were gently harvested using a rubber spatula. The purity of monocytes was determined by flow cytometric analysis of CD14 and CD3 expression and was found to be >95% in both patients with RA and healthy individuals. The monocyte suspensions were incubated at a density of 1 × 106 cells/ml in 10% FCS/culture medium in 6-well plates for 24 hours. Culture medium was removed, and adhered monocytes were then stimulated for 5 hours in fresh 10% FCS/culture medium with 1 μg/ml of LTA (Sigma, St. Louis, MO), 30 μg/ml of anti-FcγRIII mAb (3G8; Immunotech), LTA plus anti-FcγRIII mAb, or LTA plus isotype control mAb. Culture supernatants were collected and cell-free samples were stored at −30°C until the cytokine assay.

To examine the effect of Hsp60 on TNFα production, adhered monocytes were stimulated for 5 hours with 1 μg/ml of Hsp60 (Stressgen, Victoria, British Columbia, Canada) in the presence of 10 μg/ml of polymyxin B (Sigma), and culture supernatants were measured for TNFα concentrations.

For blocking experiments, adhered monocytes were incubated with 10% FCS/culture medium with anti–TLR-2 mAb (TL2.1; eBioscience), anti–TLR-4 mAb (HTA125; eBioscience), both mAb, or isotype control mAb for 30 minutes before LTA or Hsp60 stimulation.

To examine the effects of M-CSF, IL-10, and TGFβ1 on CD16 and TLR-2 expression on monocytes, aliquots of fresh PBMC suspensions were evaluated for CD16 and TLR-2 expression, and the remaining cell suspensions were incubated in polypropylene tubes (Becton Dickinson) at a density of 2 × 106 cells/ml in 1% FCS/culture medium with or without 10 ng/ml of M-CSF (Otsuka Pharmaceutical, Tokushima, Japan), IL-10 (Becton Dickinson), or TGFβ1 (R&D Systems, Minneapolis, MN). Cells were harvested 24 hours later, and expression of CD16 and TLR-2 was determined by flow cytometry.

Immunoassays for cytokines.

Concentrations of IL-1β, IL-6, IL-8, and IL-10 were measured by cytometric beads array (CBA) with a series of anticytokine mAb–coated beads and PE-conjugated anti-cytokine mAb followed by flow cytometric analysis, using the CBA kit (Becton Dickinson) and CBA software (Becton Dickinson). Levels of TNFα, IL-12 p40, and IL-1 receptor antagonist (IL-1Ra) were measured in duplicate by the quantitative sandwich enzyme-linked immunosorbent assay (ELISA) using cytokine-specific capture and biotinylated detection mAb and recombinant cytokine proteins (all from Becton Dickinson), according to the manufacturer's protocol. The detection limits for each of the cytokines were as follows: 20 pg/ml for TNFα, IL-1β, IL-6, IL-8, IL-10, and IL-12 p40, and 40 pg/ml for IL-1Ra.

Electrophoretic mobility shift assay (EMSA) for NF-κB.

To examine the effects of LTA and FcγRIII ligation on NF-κB activation, adhered mature monocytes were incubated at a density of 1 × 106 cells/ml in 1% FCS/culture medium with 1 μg/ml of LTA, 30 μg/ml of anti-FcγRIII mAb, or LTA and anti-FcγRIII mAb for 30 minutes. Cells were collected, and nuclear proteins were prepared by the mini extraction procedure, as previously described (23). Double-stranded oligonucleotides for NF-κB were labeled with α32P-dATP (Amersham, Little Chalfont, UK) with Klenow fragment (Takara, Tokyo, Japan). The sequences were as follows: 5′- GATCCGAGGGGACTTTCCGCTGGGGACTTTCCAGG-3′ GATCCCTGGAAAGTCCCCAGCGGAAAGTCCCCTAG. Nuclear extracts were incubated with 32P-labeled probe in binding buffer for 30 minutes. Samples were electrophoresed on 5% polyacrylamide gel, followed by autoradiography.

To verify the specificity of NF-κB protein binding, unlabeled NF-κB consensus oligonucleotides and 2.0 μg of rabbit anti–NF-κB p65 polyclonal antibody (H-286; Santa Cruz Biotechnology, Santa Cruz, CA) were added to the binding reaction and incubated on ice for 30 minutes prior to addition of the probe.

Statistical analysis.

Samples with values below the detection limit for the assay were regarded as negative and assigned a value of zero. Data were expressed as the mean ± SEM of the number of samples evaluated. The statistical significance of differences between 2 groups was determined by the Mann-Whitney U test. P values less than 0.05 were considered significant. The correlation coefficient was obtained by Spearman's rank correlation test.

RESULTS

High expression of TLR-2 on CD14+,CD16+ blood monocytes in RA.

The cell surface expression of CD14,CD16 (FcγRIIIA) and CD64 (FcγRI) on peripheral blood monocytes from patients with RA and healthy controls was determined by flow cytometric analysis. Figure 1A shows representative staining patterns of CD14+,CD16+ and CD64 expression in the monocyte population from a patient with RA. Monocyte subpopulations could be defined as CD14++,CD16− and CD14+,CD16+ cells, and the intensity of CD64 expression was decreased in CD16+ monocytes. The mean frequency of CD16+ cells in all CD14-expressing blood monocytes was significantly increased in patients with RA as compared with healthy controls (Table 1). The CD16+ monocyte frequency was found to correlate with the serum C-reactive protein (CRP) level (n = 35 patients; r = 0.46, P < 0.05) and the erythrocyte sedimentation rate (ESR) (r = 0.52, P < 0.001), which are clinical indicators of disease activity, but not with age, disease duration, or the drug therapy used. Thus, CD16+ blood monocytes were increased in patients with RA according to disease severity, confirming our previous observation (15).

Figure 1.

A, Flow cytometric determination of CD14+,CD16+ blood monocytes and CD64 expression on CD16+ monocytes. Peripheral blood mononuclear cells (PBMCs) from patients with rheumatoid arthritis (RA) and healthy controls were stained with fluorescein isothiocyanate (FITC)–conjugated anti-CD14 antibody or anti-CD64 antibody and phycoerythrin (PE)-conjugated anti-CD16 antibody. Flow cytometric analysis of cell surface expression of CD14, CD16, and CD64 was performed by gating on monocytes according to light scatter profile. In this RA patient, CD14+,CD16+ monocytes appearing in the upper right gate account for 24.5% of all CD14-expressing cells, and CD64 expression is decreased on CD16+ monocytes. Similar staining patterns of CD16 and CD64 on monocytes were found in other RA patients and controls. B, Histographic patterns of Toll-like receptor 2 (TLR-2) and TLR-4 expression on CD16+ and CD16− monocytes. PBMCs from RA patients were stained with biotinylated anti–TLR-2 antibody and anti–TLR-4 antibody, followed by incubation with R-PE/cyanin 5.1–conjugated streptavidin. Cells were then stained with FITC-conjugated anti-CD14 antibody and PE-conjugated anti-CD16 antibody, and analyzed as described above.

Table 1. Frequency of CD14+,CD16+ monocytes and intensity of TLR-2 and TLR-4 expression on CD16+ and CD16− monocytes in RA patients and healthy controls*
 RA patients (n = 40)Controls (n = 20)
  • *

    Values are the mean ± SEM. Peripheral blood mononuclear cells from patients with rheumatoid arthritis (RA) and healthy controls were prepared. The frequency of CD14+,CD16+ monocytes in all CD14-expressing monocytes and the intensity of Toll-like receptor 2 (TLR-2) and TLR-4 on CD16+ and CD16− monocytes were determined by flow cytometric analysis. Intensity was expressed as the difference between the mean fluorescence intensity (MFI) of staining with anti–TLR-2 antibody or anti–TLR-4 antibody and isotype-matched control antibody.

  • P < 0.01 versus controls.

  • P < 0.001 versus controls.

CD14+,CD16+ monocytes, %15.7 ± 1.29.9 ± 0.5
TLR-2 MFI, units  
 Total monocytes450 ± 38433 ± 44
 CD16+ monocytes649 ± 48600 ± 53
 CD16− monocytes421 ± 38412 ± 44
TLR-4 MFI, units  
 Total monocytes109 ± 1323 ± 5
 CD16+ monocytes95 ± 1020 ± 3
 CD16− monocytes112 ± 1424 ± 5

We next compared the intensity of TLR-2 and TLR-4 expression on CD16+ and CD16− monocytes. Figure 1B shows representative histographic patterns of TLR-2 and TLR-4 expression on CD16+ and CD16− monocytes from a patient with RA. The mean fluorescence intensity of TLR-2 on CD16+ monocytes was significantly higher than that of CD16− monocytes in patients with RA (P < 0.005), as well as in healthy controls (P < 0.005) (Table 1). In contrast, the intensity of TLR-4 was similar in both monocyte subsets from RA patients and healthy controls, although the level of TLR-4 expression was greater in RA monocytes. In addition, the intensity of TLR-2 expression on all monocytes, but not the intensity of TLR-4 expression, was found to correlate with the CRP level (Figure 2). These results suggest that TLR-2 expression may be enhanced during CD16+ monocyte maturation in vivo in association with disease activity, while TLR-4 expression remains constant.

Figure 2.

Correlation between the intensity of Toll-like receptor 2 (TLR-2) expression on blood monocytes and serum C-reactive protein (CRP) levels in patients with rheumatoid arthritis. The intensity of TLR-2 expression on CD14-expressing blood monocytes was determined by flow cytometric analysis and expressed as the mean fluorescence intensity (MFI). The correlation coefficient was analyzed by Spearman's rank correlation test.

TLR-2 expression in CD16+ synovial lining macrophages.

To determine the expression of TLR-2 on CD16+ macrophages at the site of inflammation, synovial tissues from 5 patients with RA were analyzed by 2-color immunofluorescence labeling using anti-CD16 antibody and anti–TLR-2 antibody. Figure 3 shows representative staining patterns of CD16 and TLR-2 expression in the synovial tissue. The majority of CD16+ cells were localized to the lining layer, while TLR-2+ cells were more widely distributed throughout the tissue, excluding lymphocyte infiltrates. Colocalization analysis revealed that the synovial lining layer contained a large number of cells intensely stained for both CD16 and TLR-2, but these cells were sparsely distributed in the sublining layer. The lack of TLR-2 expression on T cells was confirmed by double labeling with anti-CD3 antibody and anti–TLR-2 antibody. These results indicate that CD16+ synovial lining macrophages also express TLR-2 in RA.

Figure 3.

Expression of CD16 and Toll-like receptor 2 (TLR-2) in rheumatoid arthritis (RA) synovial tissue. Synovial tissue sections from RA patients were stained with anti-CD16 antibody or anti-CD3 antibody and anti–TLR-2 antibody, followed by incubation with isotype-specific rhodamine- and fluorescein isothiocyanate–conjugated secondary antibodies. Two-color immunofluorescence confocal images were obtained for CD16 and TLR-2 expression (red and green staining, respectively). The 2 images were superimposed, and double-positive cells are shown in yellow. Similar staining patterns were obtained in additional analyses with 4 synovial tissue samples from different patients.

Effects of M-CSF, IL-10, and TGFβ on TLR-2 expression on monocytes.

Because M-CSF, IL-10, and TGFβ1 have been shown to induce CD16+ monocytes (15, 17), we examined the effects of these cytokines on TLR-2 expression in blood monocytes. PBMCs of healthy controls were incubated with or without M-CSF, IL-10, or TGFβ1 for 24 hours, and the expression of surface TLR-2 and CD16 in fresh and cultured monocytes was determined by flow cytometric analysis. As shown in Figure 4, of these CD16-inducing cytokines, M-CSF and IL-10 clearly enhanced the intensity of TLR-2 expression, and IL-10 was likely more potent. However, TGFβ1 reduced TLR-2 expression. These results suggest that M-CSF, IL-10, and TGFβ1 have differential roles in the induction of TLR-2 during CD16+ macrophage maturation.

Figure 4.

Induction of CD16 and Toll-like receptor 2 (TLR-2) expression on blood monocytes by macrophage colony-stimulating factor (M-CSF), interleukin-10 (IL-10), and transforming growth factor β1 (TGFβ1). Peripheral blood mononuclear cells from healthy controls (2 × 106 cells/ml in 1% fetal calf serum/culture medium) were incubated with or without 10 ng/ml of IL-10, M-CSF, or TGFβ1 in polypropylene tubes. Cells were harvested 24 hours later, and the expression of surface CD16 and TLR-2 was determined by flow cytometric analysis. Similar results were obtained in additional experiments with 2 different healthy controls. MFI = mean fluorescence intensity.

Cytokine production by CD16+ mature monocytes after LTA stimulation.

To examine the effect of TLR-2 activation on cytokine production by CD16+ monocytes, the CD16+ cell–enriched mature monocyte population was prepared by incubating blood monocytes in plastic plates for 24 hours. Figure 5 shows representative histographic patterns of CD16 expression on fresh blood and adhered monocytes from a patient with RA and a healthy individual. The frequency of CD16+ monocytes was markedly increased after 24-hour adherence in both RA patients (n = 6) (from 15.5 ± 2.4% to 45.9 ± 4.4%) and healthy controls (n = 5) (from 9.0 ± 1.5% to 50.1 ± 7.9%). These adhered monocytes were stimulated with or without LTA for 5 hours, and culture supernatants were measured for various cytokines by immunoassay. Mature monocytes secreted significant amounts of TNFα, IL-1β, IL-6, IL-8, IL-12 p40, IL-1Ra, and IL-10 when stimulated with LTA (Table 2). There was no significant difference between RA and control monocytes in terms of the levels and the pattern of cytokine production.

Figure 5.

Induction of CD16 expression on adhered monocytes. Purified monocytes from 6 patients with rheumatoid arthritis (RA) and 5 healthy controls (HC) (1 × 106 cells/ml in 10% fetal calf serum/culture medium) were incubated in plastic plates, and adhered cells were harvested 24 hours later. The expression of surface CD16 on these mature monocytes was determined by flow cytometric analysis. Similar results were obtained in additional experiments with 5 different RA patients and 4 healthy controls. Ab = antibody.

Table 2. Effects of LTA and anti-FcγRIII antibody stimulation on cytokine production by adhered monocytes*
 MediumAnti-FcγRIII antibodyLTAAnti-FcγRIII antibody + LTA
  • *

    Values are the mean ± SEM. CD16+ mature monocytes were prepared by incubating monocytes from 6 healthy controls and 5 patients with rheumatoid arthritis (RA) (1 × 106 cells/ml in 10% fetal calf serum/culture medium) in plastic plates for 24 hours. Adhered monocytes were then stimulated with or without lipoteichoic acid (LTA) (1 μg/ml), anti-Fcγ receptor III (anti-FcγRIII) antibody (30 μg/ml), or a combination of LTA and anti-FcγRIII antibody for 5 hours. Culture supernatants were measured for concentrations of tumor necrosis factor α (TNFα), interleukin-1β (IL-1β), IL-6, IL-8, IL-12 p40, IL-1 receptor antagonist (IL-1Ra), and IL-10 by immunoassay. Isotype-matched control antibody for anti-FcγRIII antibody had no significant effect on cytokine production (data not shown).

RA monocytes    
 TNFα, pg/mlNot detected8 ± 3297 ± 51952 ± 225
 IL-1β, pg/ml657 ± 40716 ± 551,357 ± 2531,565 ± 321
 IL-6, ng/ml1.04 ± 0.451.07 ± 0.5010.24 ± 3.6510.42 ± 3.65
 IL-8, ng/ml7.9 ± 2.37.9 ± 1.928.2 ± 7.828.3 ± 8.0
 IL-12 p40, pg/ml14.4 ± 9.220.2 ± 6.331.0 ± 10.765.6 ± 9.1
 IL-1Ra, pg/ml1,770 ± 4951,800 ± 5062,390 ± 5382,130 ± 517
 IL-10, pg/mlNot detectedNot detected34 ± 622 ± 5
Control monocytes    
 TNFα, pg/mlNot detected7 ± 4194 ± 75776 ± 258
 IL-1β, pg/ml758 ± 99998 ± 941,653 ± 2691,757 ± 256
 IL-6, ng/ml0.98 ± 0.191.53 ± 0.249.80 ± 2.5712.05 ± 2.88
 IL-8, ng/ml16.2 ± 3.319.9 ± 4.029.2 ± 7.431.5 ± 6.9
 IL-12 p40, pg/ml9.0 ± 5.819.5 ± 9.150.3 ± 22.073.8 ± 26.9
 IL-1Ra, pg/ml1,824 ± 1841,837 ± 1692,088 ± 1251,985 ± 63
 IL-10, pg/mlNot detectedNot detected58 ± 1343 ± 8

To learn to what extent TLR-2 was responsible for this LTA-induced cytokine response, adhered monocytes from healthy controls were incubated with or without anti–TLR-2 antibody, anti–TLR-4 antibody, both antibodies, or control antibody for 30 minutes before LTA stimulation. As shown in Figure 6, LTA-induced TNFα production (with control antibody) was inhibited by 48.8% with anti–TLR-2 antibody treatment, by 21.1% with anti–TLR-4 antibody, and by 57.9% with both antibodies. These results indicate that the LTA-induced response is mediated mainly via the TLR-2 receptor, although TLR-4 is partly involved in LTA stimulation.

Figure 6.

Inhibition by anti–Toll-like receptor 2 (anti–TLR-2) antibody and anti–TLR-4 antibody of lipoteichoic acid (LTA)–induced tumor necrosis factor α (TNFα) production by adhered monocytes. Adhered monocytes from 4 healthy controls (1 × 106 cells/ml in 10% fetal calf serum/culture medium), prepared after 24-hour incubation in plastic plates, were treated with or without TLR-2 antibody (10 μg/ml), anti–TLR-4 antibody (10 μg/ml), both antibodies, or isotype-matched control antibody for 30 minutes. Cells were then stimulated with or without 1 μg/ml of LTA for 5 hours. Concentrations of TNFα in culture supernatants were measured in duplicate by enzyme-linked immunosorbent assay. Values are the mean and SEM. ND = not detected. ∗ = P < 0.05 versus control antibody (Ab).

Cytokine production by CD16+ mature monocytes after stimulation with LTA and FcγRIIIA ligation.

Previous studies have shown that the ligation of FcγRIIIA receptors by anti-FcγRIII antibody induces TNFα and IL-1α production from monocyte-derived macrophages (19). Consistent with this observation, anti-FcγRIII antibody stimulation was found to dose-dependently induce TNFα production by adhered monocytes (data not shown). The anti-FcγRIII antibody–induced cytokine response was limited at 30 μg/ml, but the addition of this concentration of anti-FcγRIII antibody could markedly augment TNFα production by LTA-stimulated adhered monocytes (Table 2). Anti-FcγRIII antibody stimulation also increased LTA-induced production of IL-1β, IL-6, IL-8, and IL-12p40, but the combined effects were mostly additive. In contrast, anti-FcγRIII antibody reduced the production of IL-1Ra and IL-10, which are known as antiinflammatory molecules in RA (13). Therefore, FcγRIII activation appears to shift the TLR-mediated monocyte response toward a more proinflammatory type.

TNFα production by CD16+ mature monocytes after Hsp60 stimulation.

It has been shown that autologous Hsp60, which are released from stressed or necrotic cells, can induce the activation of proinflammatory genes in macrophage-lineage cells via TLR-2 and TLR-4 (7, 8). Consistent with these findings, human Hsp60 was able to stimulate TNFα production from adhered mature monocytes in a dose-dependent manner (∼10 μg/ml) in the presence of polymyxin B (data not shown).

To know the importance of TLR-2 and TLR-4 in the Hsp60-induced TNFα response, adhered monocytes were incubated with or without anti–TLR-2 antibody, anti–TLR-4 antibody, both antibodies, or control antibody for 30 minutes before Hsp60 stimulation. As shown in Figure 7, Hsp60-stimulated TNFα production (with control antibody) in patients with RA and healthy controls was inhibited by 43.4% and 38.6%, respectively, with anti–TLR-2 antibody treatment, by 55.1% and 42.1%, respectively, with anti–TLR-4 antibody, and by 63.4% and 58.6%, respectively, with both antibodies. Thus, the Hsp60-induced response appears to be dependent on both TLR-2 and TLR-4 expression.

Figure 7.

TNFα production by adhered monocytes after stimulation with Hsp60 and its inhibition by anti–TLR-2 antibody and anti–TLR-4 antibody. Monocytes from 3 patients with rheumatoid arthritis (RA) and 3 healthy controls (HC) (1 × 106 cells/ml in 10% fetal calf serum/culture medium) were incubated in plastic plates for 24 hours, followed by 30-minute treatment with or without anti–TLR-2 antibody, anti–TLR-4 antibody, both antibodies, or isotype-matched control antibody. Adhered monocytes were then stimulated with or without Hsp60 (1 μg/ml) in the presence of polymyxin B (10 μg/ml) for 5 hours. TNFα production was measured by enzyme-linked immunosorbent assay. Values are the mean and SEM. ∗ = P < 0.05 versus control Ab (see Figure 6 for other definitions).

NF-κB activation by LTA but not FcγRIIIA ligation in CD16+ mature monocytes.

The activation of TLRs by their ligands induces proinflammatory mediators such as TNFα, by triggering the intracellular signaling pathway leading to NF-κB activation (3, 4). To determine whether the combined effect of LTA and anti-FcγRIII antibody stimulation on TNFα production was attributable to increased NF-κB activation, we examined by EMSA the levels of NF-κB DNA binding activity in adhered monocytes when stimulated with LTA, anti-FcγRIII antibody, or LTA and anti-FcγRIII antibody for 30 minutes. As shown in Figure 8, NF-κB activation was significantly induced by LTA, but not by anti-FcγRIII antibody. This LTA-induced NF-κB activity was not augmented in the presence of anti-FcγRIII antibody. Therefore, it is unlikely that NF-κB–activating signaling is involved in the FcγRIIIA ligation–induced macrophage activation.

Figure 8.

Activation of NF-κB in lipoteichoic acid (LTA) and anti-Fcγ receptor type III (anti-FcγRIII) antibody–stimulated adhered monocytes. After 24-hour incubation, adhered monocytes from healthy controls (1 × 106 cells/ml in 1% fetal calf serum/culture medium) were stimulated with or without LTA (1 μg/ml), anti-FcγRIII antibody (30 μg/ml), or a combination of LTA and anti-FcγRIII antibody for 30 minutes. Nuclear extracts were prepared by the mini extraction procedure, and electrophoretic mobility shift assay was performed, as described in Patients and Methods. The specificity of NF-κB protein binding was verified by inhibition experiments with unlabeled NF-κB consensus oligonucleotides (competitor) and anti–NF-κB p65 antibody. Similar results were obtained in additional experiments with 2 different healthy controls.

DISCUSSION

The CD14+,CD16+ monocytes express higher levels of class II MHC molecules, adhesion molecules, and chemokine receptors and are able to more efficiently produce proinflammatory cytokines, as compared with the classic CD14++,CD16− monocytes (17). Thus, this minor population is considered to comprise proinflammatory monocytes that can rapidly migrate to the site of inflammation and readily differentiate into tissue macrophages. We previously demonstrated that CD16+ monocytes are expanded in patients with active RA in association with a cytokine spillover from the inflamed joints (15). In the present study, we found that CD16+ monocytes more strongly express TLR-2 than do CD16− monocytes, in RA patients as well as healthy controls (21), and that CD16+ synovial lining macrophages also express TLR-2. M-CSF and IL-10 may be crucial in TLR-2 induction during CD16+ monocyte and macrophage maturation. Furthermore, our studies using adhered mature monocytes suggest that TNFα production is synergistically augmented after costimulation with FcγRIIIA ligation and TLR-2 agonist, and Hsp60 is a possible endogenous ligand of TLR-2.

Cell surface expression of the CD16 molecule is closely related to the maturation and activation of macrophage-lineage cells. There are 3 different types of the macrophage FcγRs, designated FcγRI (CD64), a high-affinity receptor for monomeric IgG, and FcγRII (CD32) and FcγRIIIA (CD16), 2 low-affinity receptors (18, 24). These FcγRs can all mediate effector functions such as phagocytosis, antibody-dependent cytotoxicity, and the release of inflammatory mediators, but FcγRIII is particularly important in the inflammatory responses initiated by small immune complexes (18). CD16 is highly inducible on activated macrophages, while CD64 and CD32 are constitutively expressed on monocytes and macrophages.

In RA synovial tissue, CD32 is homogeneously expressed throughout the tissue, but the expression of CD16 and CD64 shows reciprocal distribution (20). CD64+,CD14+ cells are abundant in the perivascular zone of the sublining layer, but CD16+ cells with weaker CD14 expression are mostly distributed in the lining layer. Lining CD16+ cells also significantly express CD68, class II MHC molecules, and intracellular adhesion molecules and produce TNFα and IL-1 (12). They are thus believed to represent the highly activated phenotype of synovial macrophages. It has been suggested that the ligation of FcγRIIIA induced by small IgG RF–containing immune complexes may be involved in proinflammatory cytokine production by tissue macrophages (19). Furthermore, CD16+ monocytes and synovial tissue macrophages are major producers of human cartilage gp39, one of the autoantigens present in the joint, and its expression in the lining layer correlates well with joint destruction (25). Therefore, CD16+ macrophages appear to contribute to both the chronic inflammation and the local autoimmune response in RA.

Coexpression of CD16 and TLR-2 in mature monocytes and synovial lining macrophages indicates that TLR-2 induction may be associated with the maturation of CD16+ macrophage-lineage cells. We found that TLR-2 expression on monocytes was enhanced by M-CSF and IL-10 but was reduced by TGFβ1, although CD16 was inducible by these cytokines. IL-10 is well known as a negative regulator for monocytes and macrophages, but this cytokine has been shown to augment the growth and differentiation of monocytes in cooperation with M-CSF (26). Both cytokines are produced in the inflamed synovial tissue (12) and are detectable at high levels in plasma and synovial fluid from patients with active RA (15). These findings suggest that overproduction of M-CSF and IL-10 at the site of disease may be responsible for the concomitant induction of TLR-2 in CD16+ monocytes and synovial tissue macrophages.

CD16+ mature monocytes, prepared after 24-hour adherence in plastic, produced various cytokines after LTA stimulation, including TNFα, IL-1β, IL-6, IL-8, IL-12 p40, IL-1Ra, and IL-10. LTA stimulation was found to be mediated mainly via TLR-2, although TLR-4 was partly involved in this response. Thus, TLR-2 agonist is a strong inducer of many cytokines in macrophages. In addition, FcγRIIIA ligation could markedly augment the LTA-induced TNFα response but decrease the IL-10 response. TNFα is the key cytokine that induces the proinflammatory cytokine cascade in RA (13), while IL-10 is the powerful inhibitor of proinflammatory cytokine synthesis (27). Therefore, the combined effect of TLR-2 and FcγRIIIA stimulation appears to be to induce a proinflammatory response in macrophages.

The mechanism of the synergistic effect of LTA and FcγRIIIA ligation on TNFα production remains to be determined. TLR-2 (as well as TLR-4) induces the activation of NF-κB and MAPK cascades through MyD88, an essential adaptor protein for their signaling, leading to the synthesis of proinflammatory cytokines such as TNFα. The transcription factor NF-κB plays a prominent role in TNFα gene induction (28). The MAPK cascade, including ERKs, JNKs, and p38 MAPK, regulates both the transcription and translation of the TNFα gene. In particular, activation of the p38 MAPK signal transduction pathway enhances the translational efficacy of TNFα messenger RNA (29, 30). Conversely, FcγRIIIA stimulation is unlikely to depend on NF-κB activation, because FcγRIIIA ligation neither induced NF-κB activity nor augmented LTA-induced NF-κB activity.

The FcγRIIIA protein associates with homodimers of the common γ signaling chain, which contains a stimulatory immunoreceptor tyrosine-based activation motif (ITAM) (24, 31). In macrophages, signaling events triggered by FcγRIIIA ligation are initiated by tyrosine phosphorylation of the ITAM by the Src kinase family, with subsequent recruitment of the Syk kinase family to the phophorylated ITAM, which induces its downstream signaling intracellular protein kinase cascades, finally leading to the activation of ERKs, JNKs, and p38 MAP kinase (18). The activation of these MAPK cascades may be responsible for FcγRIIIA-mediated cytokine induction. Therefore, we speculate that the synergistic effect of LTA and FcγRIIIA ligation may be caused by the augmented MAPK cascade activity that is induced by simultaneous activation of TLR and FcγRIIIA signaling.

The existence of TLR-2 ligands in RA joints has not yet been established. Analysis of RA synovial tissues with molecular microbiologic techniques and immunohistochemistry detected the presence of bacterial cell wall constituents as well as bacterial DNA (32), suggesting the existence of pathogen-derived ligands. Furthermore, recent evidence indicates that various endogenous molecules can activate the TLR signaling pathway. Necrotic cells, but not apoptotic cells, induce NF-κB activation and inflammatory gene expression in macrophages, and this response is totally dependent on TLR-2 expression (33). Autologous Hsp60, the mitochondria molecule released from stressed and necrotic cells, elicits a proinflammatory response in innate immune cells, and this response is mediated via TLR-2 and TLR-4 (7, 8). Because human mitochondrial Hsp60 protein is widely expressed throughout RA synovial tissue (34), we examined its ability to stimulate TNFα production by adhered monocytes. The results indicate that human Hsp60 peptide is a potent stimulus for CD16+ mature monocytes, and both TLR-2 and TLR-4 are involved in the Hsp60-induced TNFα response.

In conclusion, CD16+ blood monocytes and synovial tissue macrophages with high TLR-2 expression are increased in active RA, presumably due to overexpression of M-CSF and IL-10 in the inflamed joint. Activation of macrophages via both TLR-2 and FcγRIIIA receptors can induce the TNFα-driven proinflammatory cytokine cascade. The existence of TLR-2 ligands in RA joints remains unclear, but autologous Hsp60 protein may be a possible endogenous ligand of TLR-2.

Acknowledgements

The authors thank Drs. H. Inoue and K. Nishida (Okayama University) for providing clinical samples.

Ancillary