Article Text
Abstract
Objective To investigate the role of Fcγ receptors (FcγRs) in osteoclastogenesis and osteoclast function.
Methods Bone destruction was analysed in arthritic knee joints of several FcγR-knockout mouse strains. Unfractionated bone marrow cells were differentiated in vitro towards osteoclasts in the absence or presence of immune complexes (ICs) and stimulated thereafter for 24 h with tumour necrosis factor α (TNFα) or lipopolysaccharide (LPS). In addition, mature osteoclasts were stimulated with ICs. Experiments were analysed for osteoclast formation, bone resorption and the expression of FcγRs and osteoclast markers.
Results Bone destruction was significantly increased in arthritic knee joints of FcγRIIB-deficient mice. All FcγR classes were highly expressed on osteoclast precursors. Expression of the inhibitory FcγRIIB was similar on mature osteoclasts compared to macrophages, whereas activating FcγR levels were significantly lower. IC stimulation of mature osteoclasts did not affect their number or their bone resorptive capacity. ICs significantly inhibited differentiation of unfractionated bone marrow cells towards osteoclasts, bone resorption and expression of osteoclast markers. In the presence of ICs, osteoclastogenesis of FcγRIIB−/− precursors and bone resorption remained inhibited. In contrast, ICs could not inhibit osteoclast formation or bone resorption of FcRγ-chain−/− precursors. When IC-inhibited osteoclastogenesis was followed by stimulation with TNFα or LPS, the inhibitory effects of ICs were overruled.
Conclusion Activating FcγRs mediate IC-induced inhibition of osteoclastogenesis, which might be overruled in the presence of proinflammatory mediators. This suggests that the balance of FcγR-mediated inflammation, through proinflammatory cytokine production, as well as the direct inhibitory effect of ICs on osteoclastogenesis determines the net effect on bone loss.
- Rheumatoid Arthritis
- Inflammation
- Autoimmune Diseases
- Autoantibodies
Statistics from Altmetric.com
Introduction
Rheumatoid arthritis (RA) is characterised by chronic inflammation and joint destruction. Disruption of the balance between bone formation and resorption in favour of the latter leads to functional disability.1–3 Osteoclasts are derived from haematopoietic precursors of the monocyte/macrophage lineage and differentiate into multinucleated osteoclasts under the influence of macrophage colony-stimulating factor (M-CSF) and receptor activator of NF-κB ligand (RANKL).4
In addition to M-CSF and RANKL, co-stimulatory signalling via immunoreceptor tyrosine-based activation motif (ITAM)-coupled receptors is essential for osteoclast formation and function,5–7 and is regulated by immunoreceptor tyrosine-based inhibition motif (ITIM)-mediated signalling. The balance between these pathways determines the magnitude of the response. Previous studies have demonstrated inhibition of osteoclast formation and/or activity by ITIM-bearing receptors via the recruitment and activation of the tyrosine phosphatase SHP-1,8 ,9 its upstream receptor signal regulatory protein α10 or the inositol 5-phosphatase SHIP.11 Interestingly, SHIP plays a central role in the regulation of bone destruction in the immune complex-mediated K/BxN serum transfer arthritis model12 in which Fcγ receptors (FcγRs) also play a crucial role.13 ,14
Immune complexes (ICs), composed of antibodies bound to their specific antigen, are abundantly present in the serum and joints of patients with RA and play an important role in mediating inflammation and joint destruction.15 ICs regulate immunogenic and tolerogenic responses via FcγRs, which recognise the Fc domain of IgG and are expressed on haematopoietic cells such as monocytes, macrophages and B cells. Four FcγR classes have been identified in mice. FcγRI, FcγRIII and FcγRIV are activating receptors signalling through the ITAM-bearing FcRγ-chain. The inhibitory FcγRIIB mediates its downstream signalling through ITIM phosphorylation, thereby recruiting SHIP and counteracting activating FcγRs or other ITAM-bearing receptors.16–18
IC formation is a major element of the experimental antigen-induced arthritis (AIA) model characterised by rapid development of inflammation and severe cartilage and bone erosions upon triggering with the antigen. It closely resembles the pathogenesis of human RA and, using this model, we have previously demonstrated that the activating FcγRI and FcγRIII play a crucial role in matrix metalloproteinase-mediated cartilage destruction.19 FcγRIIB is an essential inhibitory receptor in the regulation of joint destruction and inflammation, controlling the influx of mononuclear cells.19 ,20 Furthermore, FcγRIIB-deficient mice display enhanced bone destruction and osteoclast numbers. In contrast, bone erosions in wild type (WT) and FcRγ-chain−/− mice, lacking expression of all activating FcγRs, were comparable (table 1).20
Although these studies have established a role for IC-activated FcγRs in the development of RA, it is not known whether ICs can directly modulate the activity of osteoclast precursors and osteoclasts. Our objective was to investigate the role of ICs and FcγRs in osteoclast differentiation and activation.
Methods
Mice
FcγRIIB−/− mice in C57BL/6 background, FcγRI/II/III−/− and FcRγ-chain−/− mice were generated as previously described.21–23 WT C57BL/6 mice were purchased from Charles River Laboratories (Sulzfeld, Germany). Mice were housed in filter-top cages and a standard diet and water were provided ad libitum. All animal studies were approved by the local Animal Experimentation Committee.
AIA induction
Mice were immunised with 100 μg methylated BSA (mBSA) (Sigma, St Louis, Missouri, USA) emulsified in 100 μl Freund's complete adjuvant (FCA; Difco Laboratories, Detroit, Michigan, USA). Injections were divided over both flanks and footpads of the forelegs. Heat-killed Bordetella pertussis (RIVM, Bilthoven, The Netherlands) was administered intraperitoneally as an additional adjuvant. Two subcutaneous booster injections with a total of 50 μg mBSA/FCA were given in the neck region 1 week after the initial immunisation. Three weeks later, AIA was induced by injecting 60 μg mBSA in 6 μl phosphate buffered saline (PBS) directly into the knee joint, resulting in chronic arthritis.
Histology
Seven days after AIA induction the knee joints were isolated, fixed in 4% formalin, decalcified in formic acid, dehydrated and embedded in paraffin. Sections were stained with H&E to study bone erosion and scored on a scale ranging from 0 (no erosion) to 3 (complete bone loss). Osteoclast activity was visualised by immunostaining for cathepsin K (CTSK). Sections were deparaffinised, rehydrated and incubated with anti-CTSK (Abcam, Cambridge, UK) or normal rabbit IgG. Subsequently, sections were incubated with biotinylated goat anti-rabbit IgG and binding was detected using the ABC-HRP kit (Elite Kit, Vector Laboratories, Burlingame, California, USA). Peroxidase was developed with diaminobenzidine and sections were counterstained with haematoxylin. For IgG immunolocalisation, sections were stained with goat anti-mouse peroxidase, developed with diaminobenzidine and counterstained with haematoxylin.
Cell cultures
Bone marrow was isolated from femurs and tibias as described previously.24 Cells were plated into 96-well plates at a density of 1×105 cells per well in 150 μl α-Minimum Essential Medium (α-MEM) with 5% fetal calf serum and 1% Pen/Strep containing 30 ng/ml recombinant murine (rm)M-CSF (R&D Systems, Minneapolis, Minnesota, USA) for macrophage differentiation or 30 ng/ml rmM-CSF and 40 ng/ml rmRANKL (R&D Systems) for osteoclast differentiation. Additionally, cells were seeded on 650 µm thick bovine cortical bone slices. The medium was refreshed every 3 days. At indicated time points ICs (100 µg/ml), recombinant murine tumour necrosis factor α (rmTNFα, 10 ng/ml; R&D Systems) or lipopolysaccharide (LPS, 100 ng/ml; Sigma) were added. ICs were obtained by heating 10 mg/ml rabbit IgG (Sigma) at 63°C for 30 min. After heating, the solution was centrifuged and the IC concentration in the supernatant was determined by absorbance reading at 280 nm. As a control, monomeric rabbit IgG (m-IgG) without heat treatment was used.
FACS analysis
Bone marrow subsets were analysed as previously described.25 Briefly, bone marrow cells were labelled with biotinylated ER-MP12, recognising CD31. After washing, cells were labelled with FITC-conjugated ER-MP20, recognising Ly-6C, and streptavidin-PE conjugate (Biolegend, San Diego, California, USA). ER-MP12 and ER-MP20 were kindly provided by Dr P Leenen (Erasmus UMC, Rotterdam, The Netherlands). To determine FcγR expression, cells were labelled with anti-FcγRI, anti-FcγRII/III (BD PharMingen, San Diego, California, USA), anti-Ly17.2 (clone K9.361, FcγRIIB-specific), anti-FcγRIII (R&D Systems) or anti-FcγRIV (kindly provided by F Nimmerjahn, University of Erlangen-Nuremberg, Erlangen, Germany). Data acquisition was performed on a FACSCalibur and CellQuest software (BD Biosciences, San Jose, California, USA).
TRACP staining
After 6 days of culture, cells were fixed in 4% PBS-buffered paraformaldehyde and stained for TRACP activity using the leucocyte acid phosphatase kit (Sigma). Only TRACP-positive cells with three or more nuclei were considered osteoclasts. Each osteoclast was assigned to one of three following groups: 3–5, 6–10 or >10 nuclei per cell, since the number of nuclei may reflect the maturity of the osteoclast.26
Bone resorption
To determine bone resorption by osteoclasts cultured on bone, cells were removed by sonicating the bone slices in 10% ammonia. After washing, slices were incubated in a 10% saturated alum solution (KAl(SO4)2·12H2O), washed again and resorption pits were stained with Coomassie Brilliant Blue. Bone resorption was quantified using the Qwin image analysis system (Leica Imaging Systems, Rijswijk, The Netherlands).
RNA isolation and real-time quantitative PCR
RNA was isolated using TRIzol reagent and reverse transcribed into cDNA as previously described.27 ,28 Real-time QPCR was performed using the ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, California, USA) as described elsewhere.28 Primer sequences are shown in online supplementary table S1. Product specificity was confirmed by post-amplification dissociation curve analysis. Samples were normalised for GAPDH expression using the ΔCt method. Relative gene expression was calculated as 2−(ΔCt).
Statistical analysis
Differences between two groups were tested using the unpaired Student t test or Mann–Whitney U test. Multiple comparisons were tested using one-way ANOVA followed by the Tukey multiple comparison test or Kruskal-Wallis test followed by Dunn's multiple comparison test. p Values <0.05 were considered significant.
Results
Immunolocalisation of IgG at bone erosion sites in arthritic mice
Bone destruction in arthritic knee joints of FcγRIIB-deficient mice was enhanced whereas bone destruction in WT and FcRγ-chain−/− mice was comparable.20 In addition, immunolocalisation of IgG, suggestive of ICs, were abundantly present in depositions at bone erosion sites (figure 1A–E).
ICs inhibit osteoclastogenesis of unfractionated bone marrow cells but do not affect mature osteoclasts
We first determined the expression of FcγRs by bone marrow-derived osteoclasts and compared it with their expression by bone marrow-derived macrophages which are known to express all FcγR classes. Osteoclasts expressed significantly lower mRNA and protein levels of the activating FcγRI, FcγRIII and FcγRIV whereas the expression of the inhibitory FcγRIIB was comparable (figure 2A,B).
To investigate the effect of FcγR activation on osteoclast formation and function, bone marrow cells were either differentiated towards osteoclasts in the presence of ICs or precursors were differentiated into mature osteoclasts for 5 days prior to 24 h of IC stimulation. Differentiation in the presence of ICs resulted in significantly reduced osteoclast numbers containing >5 nuclei per cell whereas the number of osteoclasts containing 3–5 nuclei was unchanged (figure 3A,D). Furthermore, mRNA levels of the osteoclast markers NFATc1, DC-STAMP, calcitonin receptor, CTSK and TRACP were significantly downregulated (figure 3B). The monocyte/macrophage marker F4/80, on the other hand, was significantly upregulated (figure 3B), suggesting a shift of IC-differentiated cells towards an intermediate and macrophage-like phenotype. Functional analysis showed significantly lower bone resorption by IC-differentiated cells compared with control osteoclasts (figure 3C,D). Inhibition was also observed when BSA:anti-BSA ICs were used (see online supplementary figure S1), suggesting that this is a general IC-mediated inhibitory mechanism. Because our in vivo data strongly suggested a potential role for FcγRIIB, we continued using heat-aggregated IgGs, thought to be the preferential receptor for this type of IC.29
To address a potential inhibitory mechanism, the expression of cytokines was studied in IC-differentiated cultures. IC-differentiated cells showed reduced interleukin (IL)-1β expression (a pro-osteoclastogenic cytokine) and enhanced IL-10 and IL-12 expression (anti-osteoclastogenic cytokines).
Stimulation of mature osteoclasts with ICs did not affect their number or their size (figure 3A,D). In addition, no differences were found in the mRNA levels of osteoclast markers whereas F4/80 was significantly upregulated (figure 3B). Bone resorption was similar for non-stimulated and IC-stimulated osteoclasts (figure 3C,D).
Precursors differentiated in the presence of m-IgG and mature osteoclasts stimulated with m-IgG displayed normal osteoclast formation and function, indicating that the observed effects are specifically IC-mediated (figure 3A–C).
IC-induced inhibition of osteoclastogenesis is independent of the inhibitory FcγRIIB
Osteoclast progenitors within the bone marrow are present at various stages of maturity and, as a result, differ in their potential to differentiate into osteoclasts.30 Based on the expression of CD31 and Ly-6C, the composition of bone marrow subpopulations of WT and FcγRIIB−/− mice was shown to be comparable (figure 4A). For both genotypes, in vitro osteoclastogenesis resulted in comparable osteoclast numbers, osteoclast size, mRNA levels of osteoclast markers and bone resorption (figure 4B–D). In the presence of ICs, however, osteoclastogenesis of both genotypes resulted in significantly reduced numbers of large and intermediate size osteoclasts (6–10 and >10 nuclei/cell, respectively; figure 4B). No differences were found in the formation of small osteoclasts (3–5 nuclei/cell). Inhibition of osteoclastogenesis was reflected by significantly decreased osteoclast marker expression and reduced bone resorption levels (figure 4C,D). These results show that IC-induced inhibition of osteoclastogenesis is independent of FcγRIIB.
Activating FcγRs mediate IC-induced inhibition of osteoclastogenesis
Although activating FcγRs are scarcely present on mature osteoclasts, their expression levels on precursors are unknown. We investigated FcγR expression on osteoclast precursor subsets in the bone marrow (see online supplementary figure S2 for gating of subpopulations and histograms). Each progenitor subset expressed high levels of activating and inhibitory FcγRs (figure 5A). We hypothesised that ICs exert their inhibitory effect via activating FcγRs in the early phase of osteoclastogenesis. Bone marrow cells from FcRγ-chain−/− mice lacking expression of all activating FcγRs and WT controls were isolated and differentiated towards osteoclasts. Under non-stimulated conditions, the number and size of osteoclasts were similar for WT and FcRγ-chain−/− cells and bone resorption levels were comparable (figure 5B–D). However, differentiation in the presence of ICs resulted in reduced osteoclast numbers and decreased bone resorption for WT precursors, whereas osteoclast numbers and bone resorption levels were similar for non-stimulated and IC-stimulated FcRγ-chain−/− cultures (figure 5B–D).
Proinflammatory stimuli can overrule IC-induced inhibition of osteoclastogenesis
In contrast to the in vitro findings, osteoclast-mediated bone destruction in vivo was enhanced in FcγRIIB-deficient mice, despite the abundant presence of ICs. During arthritis, proinflammatory mediators are produced due to FcγR-mediated induction of inflammation.16 Some of them can also stimulate osteoclast formation and activation. The presence and composition of pro- and anti-osteoclastogenic mediators might therefore determine the net effect on bone loss. To mimic an inflammatory environment, WT bone marrow cells were differentiated towards osteoclasts in the presence of ICs. After 5 days of culture, TNFα or LPS were added and 24 h later cultures were analysed for osteoclast formation and bone resorption. In the presence of ICs osteoclastogenesis was inhibited, but subsequent addition of TNFα or LPS overruled the inhibitory effect of ICs. Both stimuli triggered robust osteoclast formation and bone resorption, which were significantly increased compared with IC-differentiated cells and non-stimulated osteoclasts but similar to osteoclasts stimulated with only LPS or TNFα (figure 6A,B).
Discussion
Autoantibody formation and IC deposition in articular joints elicit local immune responses and play a crucial role in the pathogenesis of RA.31 ,32 Our data demonstrate the presence of ICs at bone erosion sites during AIA and significantly enhanced bone destruction in arthritic knee joints of FcγRIIB-deficient mice, whereas FcRγ-chain−/− mice display similar bone erosion levels as WT controls.
As previously demonstrated,20 the expression of FcγRIIB on mature bone marrow-derived osteoclasts was comparable to its expression on bone marrow-derived macrophages whereas the expression of activating FcγRs was significantly lower. This suggests that mature osteoclasts are likely to respond to ICs via the inhibitory FcγRIIB. However, IC stimulation of mature osteoclasts did not affect their number or their bone resorptive capacity. This might be due to a lack of activating FcγRs, which are normally co-expressed with and required for co-crosslinking to the inhibitory FcγRIIB upon IC ligation, in order for FcγRIIB to exert an inhibitory effect.17 ,33
Although all FcγR classes are expressed on myeloid and lymphoid progenitors early in development,34 ,35 their function on osteoclast precursors has not been thoroughly investigated. Osteoclast maturity is thought to be reflected by the number of nuclei per cell, which is positively correlated to the level of bone resorption.26 We found that osteoclastogenesis in the presence of ICs strongly inhibited osteoclast formation, reflected by significantly decreased osteoclast numbers containing >5 nuclei/cell and reduced bone resorption.
It is surprising to find that activating FcγRs inhibit osteoclastogenesis of bone marrow progenitors since it is generally thought that activating FcγRs positively stimulate cellular processes. However, under specific conditions, the FcRγ-chain, as well as other ITAM-bearing adaptors, can propagate inhibitory signals via recruitment of SHP-1, representing a novel mechanism of immune regulation.36 Whether this mechanism plays a role in our study is currently being investigated.
Another explanation could be the existence of a competitive mechanism. The inhibitory receptor Ly49Q positively regulates osteoclast differentiation as a result of competition with other ITIM-bearing receptors for binding SHP-1.37 A similar competitive mechanism may apply for ITAM-bearing adaptor proteins, which could explain why our data suggest that activating FcγRs are downregulated on osteoclast precursors upon stimulation with M-CSF and RANKL. In macrophages the ITAM-bearing FcRγ-chain is associated with activating FcγRs and mediates phagocytosis and the release of cytokines and chemokines upon IC ligation.38 However, during osteoclastogenesis the FcRγ-chain is required for the cell surface expression of OSCAR, and co-stimulatory signalling of this receptor upon ligation to its as yet unidentified ligand is needed for normal osteoclast formation and function.7 ,39 This suggests that downregulation of activating FcγRs is beneficial for osteoclastogenesis by favouring OSCAR expression.
It was recently shown that distinct bone marrow-derived osteoclast progenitors differ in their potential to differentiate into osteoclasts.30 An abnormal composition of precursor subsets in the bone marrow may disturb osteoclastogenesis in vivo, and can be a confounding variable accounting for the increased bone erosions observed in arthritic knee joints of FcγRIIB−/− mice. Although not analysed in arthritic mice, analysis of freshly isolated bone marrow revealed that this composition was comparable between WT and FcγRIIB−/− mice and is therefore unlikely to explain the observed differences in arthritic bone loss in vivo.
The severe bone destruction observed in arthritic FcγRIIB-deficient animals can be explained by the loss of negative regulation of activating FcγRs when FcγRIIB is absent. Importantly, worsening of disease when FcγRIIB is absent is not only observed during arthritis but has also been demonstrated in other autoimmune diseases.18 ,40 ,41 Loss of FcγRIIB enables continuous activation of FcγR-bearing cells by ICs, leading to sustained inflammation.20 Macrophages in inflamed tissues are thought to differentiate into osteoclasts and can also accelerate bone resorption through the production of proinflammatory pro-osteoclastogenic factors like IL-1β, TNFα and RANKL.42 Since mature osteoclasts are not susceptible to IC stimulation (as shown in this study), resident osteoclasts present within the joint prior to arthritis onset can become activated by RANKL, IL-1β or TNFα.43 ,44 Additionally, activated macrophages produce chemokines, attracting more mononuclear osteoclast precursors to the inflamed joint.45 ,46 Although ICs can inhibit osteoclastogenesis of infiltrated precursor cells, this inhibition is reversible and is quite easily overruled when proinflammatory mediators are present. Furthermore, the induction of F4/80 suggests that IC-stimulated precursors acquire a more macrophage-like phenotype and therefore may be able to produce more chemokines and cytokines, creating a self-perpetuating inflammatory loop. Additionally, even though ICs are crucial mediators of disease in the AIA model, it has also been demonstrated that T cells play an important role. Blocking IL-17 during reactivation of AIA prevents joint inflammation and bone erosion,47 and Th17 cells are now recognised to function as an osteoclastogenic T cell subset.48 Together, this will result in an increase in the total number of osteoclasts within the joint which ultimately leads to severe bone destruction.
The absence of severe bone erosions in arthritic FcRγ-chain−/− mice can be explained by the absence of sustained IC-induced activation of joint cells and accompanying enhanced production of inflammatory mediators. Unlike FcγRIIB-deficient mice in which IC-mediated inflammation is strongly enhanced, inflammation in WT FcRγ-chain−/− mice is comparable,20 resulting in similar bone erosion levels. Ongoing development of severe bone erosions in FcγRI/II/III−/− mice might be caused by activation of FcγRIV, which is still present in these mice,49 and detectable in osteoclasts and macrophages (this study) whereas the absence of FcγRIIB enables the establishment of an amplifying inflammatory loop. Although the exact role of FcγRIV remains unclear, a comparison of different arthritis models and various FcγR knock-out mice clearly indicates that this receptor is important for the development of joint destruction.23
In conclusion, our study demonstrates that activating FcγRs, but not the inhibitory FcγRIIB, mediate IC-induced inhibition of osteoclastogenesis. In the presence of proinflammatory mediators such as TNFα and LPS, the inhibitory effect might be overruled. This suggests that the balance of FcγR-mediated induction of inflammation, through proinflammatory cytokine production, as well as the direct inhibitory effect of ICs on osteoclastogenesis determines the net effect on bone loss.
Acknowledgments
The authors thank Annet Sloetjes for excellent technical assistance.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online tables
- Data supplement 2 - Online figure 1
- Data supplement 3 - Online figure 2
Footnotes
-
Contributors All authors of the paper fulfill the criteria of authorship.
-
Funding This study was financially supported by the Dutch Arthritis Association (R0001188).
-
Competing interests None declared.
-
Provenance and peer review Not commissioned; externally peer reviewed.