SEARCH

SEARCH BY CITATION

Keywords:

  • IL-33;
  • ST2;
  • OSTEOIMMUNOLOGY;
  • OSTEOCLAST;
  • BONE RESORPTION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Since the hematopoetic system is located within the bone marrow, it is not surprising that recent evidence has demonstrated the existence of molecular interactions between bone and immune cells. While interleukin 1 (IL-1) and IL-18, two cytokines of the IL-1 family, have been shown to regulate differentiation and activity of bone cells, the role of IL-33, another IL-1 family member, has not been addressed yet. Since we observed that the expression of IL-33 increases during osteoblast differentiation, we analyzed its possible influence on bone formation and observed that IL-33 did not affect matrix mineralization but enhanced the expression of Tnfsf11, the gene encoding RANKL. This finding led us to analyze the skeletal phenotype of Il1rl1-deficient mice, which lack the IL-33 receptor ST2. Unexpectedly, these mice displayed normal bone formation but increased bone resorption, thereby resulting in low trabecular bone mass. Since this finding suggested a negative influence of IL-33 on osteoclastogenesis, we next analyzed osteoclast differentiation from bone marrow precursor cells and observed that IL-33 completely abolished the generation of TRACP+ multinucleated osteoclasts, even in the presence of RANKL and macrophage colony-stimulating factor (M-CSF). Although our molecular studies revealed that IL-33 treatment of bone marrow cells caused a shift toward other hematopoetic lineages, we further observed a direct negative influence of IL-33 on the osteoclastogenic differentiation of RAW264.7 macrophages, where IL-33 repressed the expression of Nfatc1, which encodes one of the key transciption factors of osteoclast differentiation. Taken together, these findings have uncovered a previously unknown function of IL-33 as an inhibitor of bone resorption. © 2011 American Society for Bone and Mineral Research.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Osteoimmunology is an emerging area of research that focuses on the interactions between bone and immune cells and whose existence is supported by many arguments.1 First, since hematopoetic precursor cells are located within the bone marrow, they are in close proximity to bone-forming osteoblasts and bone-resorbing osteoclasts. Second, in contrast to osteoblasts, which derive from mesenchymal progenitor cells, osteoclasts are of hematopoetic origin and develop by fusion of myeloid precursor cells.2 And third, there is a body of evidence showing that the activity of bone cells can be modulated by cytokines derived from immune cells.3–6 This includes the cytokine receptor activator of NF-κB ligand (RANKL), the major positive regulator of osteoclastogenesis, which is produced not only by osteoblasts but also by activated T cells.7 Based on these observations, the interactions between bone and immune cells have been considered to be of critical importance for skeletal homeostasis. In addition, inflammatory processes can trigger increased bone turnover and bone destruction, thereby explaining some of the pathologies observed in disorders such as rheumatoid arthritis and inflammatory bowel disease.8, 9 Thus it is of paramount clinical importance to identify specific cytokines that regulate the activity of osteoblasts and osteoclasts because these molecules and, even more, their receptors could represent excellent targets for the treatment of bone-loss disorders.

One of the first identified cytokines mediating an interaction of activated immune cells and bone-resorbing osteoclasts was interleukin 1β (IL-1β), which belongs to the IL-1 cytokine family, whose members (IL-1α, IL-1β, IL-18, and IL-33) are known primarily for their proinflammatory roles.10–12 In fact, purification and subsequent sequence analysis of an osteoclast-activating factor produced by stimulated peripheral blood mononuclear cells identified IL-1β as a candidate whose regulatory effect on osteoclastogenesis then was confirmed in vitro and in vivo.10, 13–17 In addition, IL-1 has been suggested to reduce bone formation, although the physiologic relevance of this function remains to be established.16, 18 Importantly, IL-18, another member of the IL-1 family, does not influence bone cells in the same way. In fact, while osteoclast formation is inhibited by IL-18, both in vitro and in vivo, IL-18 has a mitogenic effect on osteoblasts and is required for the bone-anabolic action of parathyroid hormone.19–23 The most recently identified member of the IL-1 cytokine family is IL-33, which is expressed in many tissues, in contrast to its receptor ST2, which is expressed primarily by TH2 lymphocytes and has been shown to regulate type 2 immune responses.24–26 Although there is recent evidence suggesting an important role of IL-33 and ST2 in the development of rheumatoid arthritis, a possible influence of IL-33 on bone cells has not been analyzed so far.27

Here we show that the expression of Il33 increases in the course of primary osteoblast differentiation, which led us to address the question of whether IL-33 would affect differentiation and function of bone cells. Using molecular analysis and functional assays, we were able to demonstrate that IL-33 only has no significant influence on the mineralization of calvarial osteoblast cultures, whereas even low concentrations of IL-33 completely abolished osteoclast formation from bone marrow precursor cells and from the macrophage cell line RAW264.7. Most important, however, we observed that mice lacking the IL-33 receptor ST2 display low trabecular bone mass caused by increased bone resorption, thus suggesting a physiologic function of IL-33 as an inhibitor of osteoclastogenesis, which may be relevant for the treatment of bone-loss disorders.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Primary osteoblasts

Primary murine osteoblasts were isolated from calvaria of newborn mice, as described previously.28 At 80% confluency, cells were differentiated by adding β-glycerophosphate and ascorbic acid to a final concentration of 10 mM and 50 µg/mL, respectively. To determine the effect of IL-33 on primary osteoblast differentiation and mineralization, cells were grown in the presence of recombinant IL-33 (R&D Systems, Minneapolis, MN, USA) from the day of seeding. Von Kossa staining of mineralized matrix was performed as described previously.28 For alizarin red staining, cells were fixed with 90% ethanol, washed twice with water, and incubated with 40 mM alizarin red staining solution (pH 4.2) for 10 minutes at room temperature, followed by five subsequent washes with water. To quantify matrix mineralization, the stained cultures were incubated with 10% acetic acid for 30 minutes before heating for 10 minutes at 85°C. After removing the cellular remnants by centrifugation, 400 µL of supernatant was neutralized with 150 µL of 10% ammonium hydroxide, and 150 µL of each sample were used to determine absorption at 405 nm. To analyze the immediate effects of IL-33, cells were serum-starved over night on day 10 of differentiation and then treated with IL-33 for 15 minutes or for 6 days, respectively.

Expression analysis

RNA was isolated using RNeasy Mini Kit (Qiagen, Valencia, CA, USA), and DNase digestion was performed according to manufacturer's instructions. Concentration and quality of RNA were determined using a NanoDrop ND-1000 system (NanoDrop Technology, Wilmington, DE, USA). For the genome-wide expression analysis, 5 µg of RNA were used for first-strand cDNA synthesis. Synthesis of biotinylated cRNA was carried out using the IVT Labeling Kit (Affymetrix, Santa Clara, CA, USA). For Gene Chip hybridization, the fragmented cRNA was incubated in hybridization solution at 45°C for 16 hours before the Gene Chips (MG $#) 2.0, Affymetrix) were washed using the Affymetrix Fluidics Station 450. For RT-PCR expression analysis, 1 µg of RNA was reversed transcribed using SuperScriptIII (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. Gene-specific primers were used to amplify IL-33 (5'-TGT CCT GCA AGT CAA TCA GGC-3' and 5'-GTG CCA GGA AGA TTC TTG CAT T-3'), Il1rl1 (5'-TGT TAT CAG AAG CCC CAA CTT G-3' and 5'-CTT GGC TCT TGG AGA GCT TTG-3'), Ear11 (5'-TCC GGC CAG TCA TTA TTC CA-3' and 5'-AGA GGA CTC GGT TCT TCC CAA C-3'), Ccl24 (5'-CGA GTG GTT AGC TAC CAG TTG G-3' and 5'-GCG TCT CTG GAC AGC AAA CTT-3'), Arg1 (5'-AAA GGC CGA TTC ACC TGA GCT-3' and 5'-GTT GAG TTC CGA AGC AGC CA -3'), Chi3l4 (5'-AGA TCT TTG ATG CCA CCC AGG-3' and 5'-GCC TTG CAA CTT GCA CTG TGT A-3'), Oscar (5'-GAA CTG CTG GTA ACG GAT CAG C-3' and 5'-TGT GTG CCG ATC AAA AGG AAG-3'), Nfatc1 (5'-TGC AGC TAC ACG GTT ACT TGG A-3' and 5'-CGA TGA TGG CTC GCA TGT TA-3'), Gapdh (5'-GAC ATC AAG AAG GTG GTG AAG CAG-3' and 5'-CTC CTG TTA TTA TGG GGG TCT GG-3'), and B2m (5'-TGC TAT CCA GAA AAC CCC TCA A-3' and 5'-TGC TTA ACT CTG CAG GCG TAT G-3'), respectively. For quantitative PCR analysis, predesigned TaqMan Assays (Applied BioSystems, Inc., Foster City, CA, USA) and Taqman Gene Expression Mastermix (Applied BioSystems) were used. Reactions were performed on a StepOnePlus system (Applied BioSystems). Gapdh expression was used as internal control.

Western blotting

Cells were washed with cold PBS and lysed with cold RIPA buffer (1% NP40, 1% sodium desoxycholate, 0.1% sodium dodecylsulfate, 150 mM sodium chloride, 2 mM EDTA, 10 mM sodium phosphate) containing a protease and phosphatase inhibitor cocktail (Roche, Basel, Switzerland). After 20 minutes of incubation on ice, lysates were centrifuged for 15 minutes at 4°C, and supernatants were stored at −80°C until further use. Protein concentrations were determined using BioRad (Munich, Germany) protein assay according to manufacturer's instructions. For Western blotting, equal amounts of protein were subjected to SDS-PAGE and then transferred to Hybond PVDF membranes (Amersham Biosciences, Freiburg, Germany). Membranes were blocked for 1 hour using blocking buffer (TBS containing 0.1% Tween-20 and 5% nonfat dry milk) and incubated with primary antibody at a dilution of 1:1000 at 4°C overnight. The antibodies were directed against p-p42/44 (No. 9101, Cell Signaling, Danvers, MA, USA), p-Stat3 (No. 9134, Cell Signaling), and β-actin (MAB1501, Millipore, Billerica, MA, USA).

Skeletal analysis

Il1rl1-deficient mice have been described previously.25 Genotyping from heterozygous offspring was performed by PCR using primers amplifying the wild-type (5'-TTG GCT TCT TTT AAT AGG CCC-3' and 5'-TGT TGA AGC CAA GAG CTT ACC-3') and mutant (5'-CTA TCA GGA CAT AGC GTT GGC TAC C-3' and 5'-TGT TGA AGC CAA GAG CTT ACC-3') alleles, respectively. Before their skeletal analysis, all mice (genetic background C57Bl/6) received two injections of calcein (9 and 2 days before euthanization). After their initial analysis by contact X-ray (Faxitron X-ray Corp., Lincolnshire, IL, USA), vertebral bodies L2 to L5 and one tibia from each animal were dehydrated and embedded nondecalcified into methyl methycrylate for sectioning. Sections were stained either with toluidine blue or by the von Kossa/van Gieson procedure, as described previously.28 Static and cellular histomorphometry was carried out on toluidine blue–stained sections using the OsteoMeasure system (Osteometrics, Decatur, GA, USA) following the guidelines of the American Society of Bone and Mineral Research.29 Dynamic histomorphometry for determination of the bone-formation rate was performed on two consecutive nonstained 12-µm sections. Tartrate-resistant acid phosphatase (TRACP) activity staining was performed on decalcified sections using napthol AS-MX phosphate (Sigma, St Louis, MO, USA) and Fast Red Violet LB salt (Sigma) in 40 mM acetate buffer (pH 5). The trabecular bone volume and the cortical thickness of femurs were quantified by micro–computed tomographic (µCT) scanning using a µCT40 (Scanco Medical, Bruettisellen, Switzerland).

Osteoclast differentiation

For osteoclastogenesis, the bone marrow was flushed out of the femurs with α modified essential medium (α-MEM) containing 10% fetal bovine serum (FBS). Cells were plated at a density of 5 × 106 cells per milliliter, and the adherent cells were cultured in α-MEM containing 10% FBS and 10 nM 1,25-dihydroxyvitamin D3 (Sigma). Recombinant IL-33 was added from the day of seeding until the day of analysis, with the exception of the experiments shown in Fig. 5C. Beginning on day 3 after seeding, macrophage colony-stimulating factor (M-CSF) and RANKL (both from Peprotech, Hamburg, Germany) were added to final concentrations of 20 and 40 ng/mL, respectively, with the exception of the experiment shown in Fig. 4C, where osteoclastogenic differentiation was analyzed in the absence of recombinant M-CSF and RANKL. Purification of Cd11b+ cells was performed using anti-CD11b microbeads and MiniMACS columns (No. 130-049-601, Miltenyi Biotec Inc., Berisch Gladbach, Germany). In brief, 2 × 108 bone marrow cells were resuspended in 1.8 mL of PBS (including 0.5% bovine serum albumin and 2 mM EDTA) before 200 µl of anti-CD11b microbeads was added. After incubation for 20 minutes, cells were washed and finally resuspended in 1 mL of PBS before the cell suspension was applied onto the column. After extensive washing steps and removal of the CD11b population, Cd11b+ cells were released, and an aliquot of the cell suspension was used to confirm the purity (>97%) of the separated cell populations via fluorescence-activated cell-sorting (FACS) analysis (data not shown). Cd11b+ cells were plated at an initial density of 1.5 × 106 cells per milliliter and cultured in the presence of M-CSF (20 ng/mL) and various concentrations of IL-33 starting on the day of seeding. In contrast, RANKL (40 ng/mL) was added starting 3 days thereafter. RAW264.7 cells were obtained from ATCC (Wesel, Germany) and cultured in DMEM + GlutaMAXX (No. 31966-021, Gibco, Darmstadt, Germany) containing 10% fetal calf serum (FCS). For osteoclast differentiation, cells were plated at a density of 105 cells per milliliter, with RANKL (50 ng/mL) and IL-33 (various concentrations) being added immediately. The formation of TRACP+ multinucleated osteoclasts was quantified 3 days after plating.

Osteoclastogenesis assays

For TRACP staining, the medium was removed, and the cells were washed twice with PBS, followed by fixation with cold methanol for 5 minutes. After two subsequent washes with water, cells were dried for 2 minutes and then stained as described using Naphtol ASMX-Phosphate (Sigma) as a substrate. The number of osteoclasts then was determined by counting all TRACP+ multinuclear cells in one well. For resorption assays, bone marrow cells were seeded onto dentin slices of 1 mm thickness. After 10 and 20 days of differentiation, the dentin slices were incubated in sodium hypochloride for 2 minutes, and residual cells were removed with a cell scraper. To visualize the resorbed areas, the slices were dipped three times in 0.2% toluidine blue staining solution, as described previously.30

FACS analysis

Analyses of surface marker expression via flow cytometry were conducted after blocking unspecific binding sites with COHN fraction II (Sigma) using the following fluorescently labeled primary antibodies: CCR3/CD193 (CCR3-phycoerythrin conjugate; R&D Systems), CD206/MMR (CD206-Alexa 647 conjugate, BioLegend, Uithoorn, The Netherlands), Ly-6G (Ly-6G-phycoerythrin, BD Biosciences, San Jose, CA, USA), CD23 (CD23-fluorescein; BD Biosciences), CD11b (CD11b-phycoerythrin, CALTAG Laboratories, Buckingham, UK), and ST2 (ST2-fluorescein, MD Biosciences, St. Paul, MN, USA). Additionally, appropriately matched and fluorescently conjugated isotype controls were used. Flow cytometry was performed with a FACSCalibur flow cytometer (BD Biosciences). Data were processed with CellQuest-Pro software (BD Biosciences).

Statistical analysis

Data are presented as means ± SD. Statistical analysis was performed using unpaired two-tailed Student's t tests. Asterisks indicate statistically significant differences (*p < .05; **p < .005; ***p < .0005).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Il33 is expressed by differentiated osteoblasts

To identify differentially expressed genes during osteoblast differentiation, we cultured primary osteoblasts derived from newborn mouse calvariae and isolated RNA before the addition of ascorbate and β-glycerophosphate (day 0), as well as 5 or 25 days thereafter (day 5 and day 25, respectively). This RNA then was subjected to a genome-wide expression analysis using Affymetrix Gene Chip hybridization.31, 32 Here we found that Il33 expression was markedly induced between days 5 and 25 of differentiation, whereas the expression of the other IL-1 family members was either undetectable (Il1a and Il1b) or not regulated (Il18) (Fig. 1A). Given the potential importance of these results, we went on to perform quantitative RT-PCR and confirmed that the expression of Il33 increases gradually during the course of differentiation, whereas the Il1rl1 gene, encoding the IL-33 receptor ST2, also was expressed, albeit at higher levels at the earlier stages of differentiation (Fig. 1B). Since these results suggested a potential relevance of IL-33 in the regulation of bone remodeling, we analyzed the expression of Il33 and Il1rl1 in different tissues, including bone, and in cultured osteoclasts and osteoblasts using RT-PCR. Here we confirmed the expression of Il33 in differentiated osteoblasts and a number of other tissues (Fig. 1C). In contrast, Il1rl1 expression was restricted to only a few tissues, including vertebrae, femurs, and calvaria, thus suggesting a role in the regulation of bone cell differentiation and function.

thumbnail image

Figure 1. Expression of Il33 and Il1rl1 in osteoblasts. (A) Affymetrix signal intensities for the expression of Il33, Il1a, Il1b, and Il18 in osteoblasts on day 0, day 5, and day 25 of differentiation. (B) Quantitative RT-PCR for the expression of Il33 and Il1rl1 at different stages of primary osteoblast differentiation. Bars represent mean ± SD from three independent experiments and indicate the fold expression compared with day 0. (C) RT-PCR expression analysis for Il33 and Il1rl1 in different tissues and primary bone cells (Ocl = osteoclasts; Obl = osteoblasts) at different stages of differentiation. Gapdh expression was used as a control for cDNA quality.

Download figure to PowerPoint

IL-33 stimulation of osteoblasts induces the expression of factors promoting bone resorption

To address the question of whether the expression of Il33 and Il1rl1 in osteoblasts might be functionally relevant, we first differentiated primary osteoblasts for 10 days with increasing concentrations of recombinant IL-33. Using von Kossa and alizarin red staining, we observed that IL-33 treatment did not affect matrix mineralization (Fig. 2A), and the same was observed when the cultures were differentiated for 20 days (Fig. 2B). We next performed a genome-wide expression analysis in order to identify genes being regulated following a stimulation with 100 ng/mL of IL-33 for 6 hours. Here we found that the expression of several genes with a known influence on bone formation and matrix mineralization was not affected by IL-33 administration (Fig. 2C). In contrast, we observed a marked induction in the expression of the cytokine receptor gene Il13ra2, which has been shown to regulate collagen production in the lung,33 and of Lcn2, encoding the secreted molecule lipocalin-2, for which diverse functions have been reported34 (Fig. 2D). In addition, IL-33 administration also caused an induction of several cytokine-encoding genes that positively influence osteoclastogenesis, such as Csf2, Il6, and Il1.16, 35, 36 Most important, however, the expression of Tnfsf11, the gene encoding RANKL, also was higher in IL-33-treated cells, whereas the expression of Tnfrsf11b, encoding the RANKL antagonist osteoprotegerin (OPG),37 was repressed by IL-33 (Fig. 2D).

thumbnail image

Figure 2. Influence of IL-33 on mineralization and gene expression in osteoblasts. (A, B) Von Kossa (top) and alizarin red (bottom) staining of osteoblasts that were differentiated for 10 (A) or 20 days (B) in the presence of the indicated IL-33 concentrations. The quantification of the alizarin red incorporation is given below. Bars represent mean ± SD (n = 8). (C) Affymetrix signal intensities and signal log ratios (SLRs) for the expression of osteoblast marker genes in the absence or presence of 100 ng/mL of IL-33 (incubation for 6 hours). (D) SLRs of selected genes being regulated by IL-33 administration in the same experiment. (E) Quantitative RT-PCR for the expression of Csf2, Tnfsf11, and Tnfrsf11b following a 6-hour administration of the indicated concentrations of IL-33. Values represent mean ± SD from three independent experiments. The dotted lines indicate the mean expression levels in the absence of IL-33.

Download figure to PowerPoint

Given the potential relevance of these latter findings, we confirmed them by quantitative RT-PCR using three independently isolated osteoblast cultures stimulated for 6 hours with increasing concentrations of IL-33. Here we found a dose-dependent increase in the expression of Csf2 and Tnfsf11 and a dose-dependent decrease in the expression of Tnfrsf11b, thus suggesting an indirect influence of IL-33 on osteoclastogenesis (Fig. 2E). Since we have shown previously that Tnfsf11 expression in osteoblasts can be induced in a Stat3-dependent manner,38 we next treated cultures on day 10 of differentiation with IL-33 for 15 minutes and observed the expected phosporylation not only of p42/4439 but also of Stat3 (Fig. 3A). Since this experiment provided further evidence for an immediate effect of exogenously added IL-33 on osteoblasts, we next analyzed whether the induction of Csf2 and Tnfsf11 expression by IL-33 requires the presence of the IL-33 receptor ST2, which is encoded by the Il1rl1 gene. Therefore, we isolated primary calvarial osteoblasts from wild-type and Il1rl1-deficient mice and stimulated them with IL-33 on day 10 of differentiation. Using quantitative RT-PCR, we were able to demonstrate that the induction of both Csf2 and Tnfsf11 by IL-33 is fully abrogated in Il1rl1-deficient cultures (Fig. 3B), which led us to address the question of whether Il1rl1−/− mice would display a skeletal phenotype associated with decreased bone resorption.

thumbnail image

Figure 3. Low trabecular bone mass in Il1rl1-deficient mice. (A) Western blot analysis of p42/44 and Stat3 phosphorylation following stimulation of primary osteoblasts (day 10 of differentiation) with the indicated IL-33 concentrations for 15 minutes. Probing with β-actin was used as a loading control. (B) Quantitative RT-PCR for the expression of Csf2 and Tnfsf11 following a 6-hour administration of IL-33 (100 ng/mL) to wild-type (+/+) and Il1rl1-deficient (−/−) primary osteoblasts on day 10 of differentiation. Values represent mean ± SD from three independent experiments. (C) Von Kossa/van Gieson staining of vertebral body sections from 6-week-old wild-type and Il1rl1-deficient mice. Quantification of the trabecular bone volume per tissue volume (BV/TV) is given below. (D) Von Kossa/van Gieson staining of tibia sections and quantification of the BV/TV in the same mice. (E) µCT scanning and quantification of the trabecular bone volume in the distal femurs. (F) Cross-sectional µCT scanning and quantification of the cortical thickness (C.Th.) of the femurs. All values represent mean ± SD (n = 4 mice). Asterisks indicate statistically significant differences compared with wild-type littermates.

Download figure to PowerPoint

Low trabecular bone mass and increased osteoclast numbers in Il1rl1-deficient mice

Using nondecalcified histology of spine sections, we unexpectedly found that the trabecular bone volume of 6-week-old Il1rl1−/− mice was decreased by more than 40% compared with wild-type littermates (Fig. 3C). The same observation was made in sections from the tibia, thereby demonstrating that the loss of ST2 function in mice causes a systemic low-bone-mass phenotype (Fig. 3D). To address the question of whether this is also true for cortical bone, we further performed µCT analysis of the femurs, where we again observed a significant reduction of the trabecular bone volume (Fig. 3E), whereas the cortical thickness was not changed compared with wild-type littermates (Fig. 3F).

To analyze the cellular basis of the observed osteopenia, we first performed dynamic histomorphometry to determine the bone-formation rate. Here we did not observe a difference between wild-type and Il1rl1−/− littermates, and the same was the case regarding the numbers of bone-forming osteoblasts (Fig. 4A). In contrast, TRACP activity staining of osteoclasts followed by histomorphometric quantification revealed that the number of osteoclasts and the surface covered by them were increased by more than 40% in Il1rl1−/− mice (Fig. 4B). Taken together, these results suggested that IL-33 has no physiologic relevance as a regulator of bone formation but rather a direct negative influence on osteoclastogenesis.

thumbnail image

Figure 4. Increased osteoclastogenesis in Il1rl1-deficient mice. (A) Fluorescent micrographs of endosteal bone surfaces showing the same distance between calcein labeling fronts in wild-type and Il1rl1-deficient mice. The histomorphometric quantification of the osteoblast number per bone perimeter (ObN/BPm) and the bone-formation rate per bone surface (BFR/BS), which were determined at trabecular bone surfaces, is shown below. (B) TRACP acitivity staining reveals a higher number of bone-resorbing osteoclasts at trabecular bone surfaces of Il1rl1-deficient mice. The histomorphometric quantification of the osteoclast number per bone perimeter (OcN/BPm) and the osteoclast surface per bone surface (OcS/BS) is shown below. (C) TRACP activity staining demonstrating the absence of multinucleated osteoclasts in bone marrow cultures differentiated in the presence of 10 nM 1,25-dihydroxyvitamin D3 and at least 5 ng/mL of IL-33 for 10 days.

Download figure to PowerPoint

RANKL-independent inhibition of osteoclastogenesis by IL-33

To address this possibility, we next isolated bone marrow cells from wild-type mice and differentiated them for 10 days, only in the presence of 1,25-dihydroxyvitamin D3. Using TRACP activity staining, we were able to detect multinucleated osteoclasts in these cultures, but there were no such cells when the differentiation was performed in the presence of IL-33 (Fig. 4C), thereby confirming the antiosteoclastogenic function of IL-33 deduced from the analysis of Il1rl1−/− mice. We next differentiated the bone marrow cells in the presence of M-CSF and RANKL, which were added to the culture medium beginning on day 3 after seeding. Since the cells were plated onto dentin slices, we were able to quantify their resorptive activity after 10 and 20 days and observed a marked negative influence of IL-33, thereby demonstrating a RANKL-independent antiosteoclastogenic effect of IL-33 (Fig. 5A).

thumbnail image

Figure 5. RANKL-independent inhibition of osteoclastogenesis by IL-33. (A) Toluidine blue staining of dentin slices (on days 10 and 20 of differentation) reveals a complete abolishment of resorptive activity by IL-33 when bone marrow cells were differentiated in the presence of 20 ng/mL of M-CSF and 40 ng/mL of RANKL. Quantification of the resorbed area is given below. (B) TRACP activity staining (on day 6 of differentiation) of multinucleated osteoclasts in bone marrow cultures differentiated in the presence of 20 ng/mL of M-CSF, 40 ng/mL of RANKL, and various concentrations of IL-33. The numbers of TRACP+ multinucleated cells (TRACP+ MNCs) are given below. (C) Quantification of TRACP+ multinucleated cells following differentiation of bone marrow in the presence of the indicated IL-33 concentrations. IL-33 was added either for the whole course of the experiment (left), from day 1 to day 4 (center), or from day 4 to day 7 (right), as indicated above. All values represent mean ± SD of three independent experiments.

Download figure to PowerPoint

To assess the concentration-dependent influence of IL-33 more precisely, we next differentiated the bone marrow cells for 6 days in the presence of lower IL-33 concentrations before we quantified the number of TRACP+ multinuclated cells in the cultures. Here we observed that IL-33 concentrations below 0.1 ng/mL did not affect osteoclastogensis significantly, whereas concentrations above 0.5 ng/mL nearly abolished the generation of TRACP+ multinucleated cells (Fig. 5B). To address the question of whether IL-33 exerts its negative function during the initial or the later stages of differentiation, we added IL-33 either for the whole course of differentiation (days 1 to 6) or only from day 1 to day 3 and from day 3 to day 6, respectively. When we quantified the number of TRACP+ multinucleated cells in these cultures, we observed that their generation was not abolished when IL-33 was present from day 3 to day 6 of differentiation, thus suggesting that the antiosteoclastogenic effect of IL-33 is mediated at the level of osteoclast precursor cells (Fig. 5C).

IL-33 promotes the generation of eosinophils, basophils, and M2 macrophages at the expense of osteoclasts

To understand the underlying mechanism of the antiosteoclastogenic effect of IL-33, we next performed Affymetrix Gene Chip hybridization and compared bone marrow cells differentiated into osteoclasts for 6 days in either the absence or presence of IL-33. Here we found that most of the genes that are regulated by IL-33 treatment of primary calvarial osteoblasts also were differentially expressed in a similar way in bone marrow cultures with two exceptions, Tnfsf11 and Tnfrsf11b (Fig. 6A). In fact, while Tnfrsf11b (encoding OPG) was not expressed under either culture condition (data not shown), the expression of Tnfsf11 (encoding RANKL) was markedly lower in cultures being differentiated in the presence of IL-33. However, since the antiosteoclastogenic effect of IL-33 also was observed in the presence of exogenous RANKL, the most interesting result from this Gene Chip comparison was that the cultures differentiated in the presence of IL-33 were characterized by an increased expression of genes known as markers for eosinophils (Ear11 and Ccl24) and M2 macrophages (Arg1 and Chi3l4). In addition, the expression of osteoclastogenesis markers (Oscar, Nfatc1, and Calcr) was reduced when the cultures were differentiated in the presence of IL-33, thereby confirming the antiosteoclastogenic effect of IL-33 at the molecular level (Fig. 6B).

thumbnail image

Figure 6. IL-33 promotes the generation of specific myeloid cell types at the expense of osteoclasts. (A) Results of an Affymetrix Gene Chip hybridization comparing osteoclast cultures differentiated in the absence or presence of 100 ng/mL of IL-33 for 6 days. Shown are the SLRs for the same genes that were found to be regulated by IL-33 in osteoblasts. (B) SLRs of selected genes being regulated by IL-33 administration in the same experiment. (C) RT-PCR expression analysis for Ear11, Ccl24, Arg1, Chi3l4, Oscar, and Nfatc1 in cultures differentiated in the absence or presence of IL-33. B2m expression was monitored as a nonregulated control. (D) FACS analysis of osteoclast cultures differentiated in the absence or presence of 100 ng/mL of IL-33 using the indicated antibodies. (E) Quantification of the cells positive for the indicated surface antigens and CD11b. (F) Quantification of the total cell number and the number of ST2+ cells in cultures differentiated in the absence or presence of 100 ng/mL of IL-33. Bars represent mean ± SD of three independent experiments.

Download figure to PowerPoint

Since we were able to confirm these observations by RT-PCR using independently isolated samples (Fig. 6C), we went on to perform FACS analysis in order to quantify the effects of IL-33 on the differentiation of myeloid progenitor cells (Fig. 6D). Using antibodies known to be expressed by eosinophil (anti-CCR3), neutrophil (anti-Ly-6G), basophil (anti-CD23) or M2 macrophage (anti-CD206) populations, we observed that IL-33 significantly increased the relative number of all these myeloid cell types, with the exception of neutrophils, and that 60% of the cells cultured in the presence of IL-33 for 5 days were positive for CD206 (Fig. 6E). In addition, we observed that the total cell number was significantly higher in these cultures and that IL-33 not only caused an increase of its own expression (Fig. 6B) but also enhanced the expression of its receptor ST2 (Fig. 6F).

Based on this result, we next analyzed the influence of IL-33 on the osteoclastogenic differentiation of bone marrow cells derived from Il1rl1−/− mice. Here we did not observe a significant difference regarding the number of TRACP+ multinucleated cells between wild-type and Il1rl1−/− cultures in the absence of IL-33. However, even at a concentration of 100 ng/mL, the antiosteoclastogenic action of IL-33 was fully abrogated in Il1rl1−/− cultures (Fig. 7A). This was confirmed by analyzing the expression of Arg1 and Oscar, where we found that the regulatory influence of IL-33 was blunted in the absence of ST2 (Fig. 7B). Since the interaction of IL-33 and ST2 is known primarily to increase the production of TH2 cytokines, and since it has been shown that the RANKL-independent inhibition of osteoclast formation by the TH2 cytokines IL-4 and IL-13 requires the presence of Stat6,40 we have further analyzed the influence of IL-33 on the osteoclastogenic differentiation of bone marrow cells from Stat6-deficient mice.41 Here we found that IL-33 fully abolished the generation of TRACP+ multinucleated cells, thereby suggesting that the antiosteoclastogenic effect of IL-33 is not the consequence of an increased production of IL-4 and IL-13 (Fig. 7C).

thumbnail image

Figure 7. The antiosteoclastogenic effect of IL-33 requires the presence of ST2. (A) Quantification of TRACP+ multinucleated cells following a 6-day differentiation of bone marrow cultures from wild-type (+/+) and Il1rl1-deficient (−/−) mice in the absence or presence of 100 ng/mL of IL-33. (B) RT-PCR expression analysis for Arg1 and Oscar in cultures differentiated in the same way. B2m expression was monitored as a nonregulated control. (C) Quantification of TRACP+ multinucleated cells following a 6-day differentiation of bone marrow cultures from Stat6-deficient mice in the absence or presence of various concentrations of IL-33. (D) TRACP activity staining of multinucleated osteoclasts in CD11b-purified bone marrow cultures differentiated in the presence of various concentrations of IL-33 for 6 days. The quantification of TRACP+ multinucleated cells found in cultures from wild-type and Il1rl1-deficent mice is given below. Bars represent mean ± SD of two independent experiments performed in triplicate (n = 6). (E) RT-PCR expression analysis for Il1rl1, Arg1, Oscar, and Nfatc1 in wild-type cultures differentiated for 3 or 5 days in the absence or presence of 100 ng/mL of IL-33. (F) RT-PCR expression analysis for Arg1, Oscar, and Nfatc1 in Il1rl1-deficient cultures differentiated in the same way.

Download figure to PowerPoint

An antiosteoclastogenic effect of IL-33 is observed in the absence of bone marrow stromal cells and in RAW264.7 macrophage cultures

To analyze whether the antiosteoclastogenic effect of IL-33 is mediated indirectly through bone marrow stromal cells, we next separated the stromal cell and leukocyte populations by CD11b immunoaffinity. When we differentiated the CD11b+ population into osteoclasts, again by adding RANKL and M-CSF, we observed that IL-33 inhibited osteoclastogenesis by more than 80% and that this inhibition was abrogated when the same experiments were performed with cells derived from Il1rl1−/− mice (Fig. 7D). Interestingly, however, IL-33 administration did not fully diminish the differentiation into TRACP+ multinucleated cells from wild-type CD11b+ precursor cells, thereby suggesting at least a minor influence of bone marrow stromal cells, which is further underscored by the observation that the number of TRACP+ multinucleated cells also was higher in the absence of IL-33 when compared with the mixed populations analyzed previously. Although these experiments suggested that the antiosteoclastogenic effect of IL-33 is largely independent of bone marrow stromal cells, they did not rule out the possibility that it is mediated indirectly through the enhanced generation of other myleoid cell types, such as M2 macrophages. In this regard, it is important to state that the induction of Arg1 expression still was observed following IL-33 administration to CD11b+ cells from wild-type mice (Fig. 7E) but not from Il1rl1−/− mice (Fig. 7F).

Therefore, to demonstrate a direct effect of IL-33 on osteoclastogenesis, we finally used the macrophage cell line RAW264.7, which can be differentiated into multinucleated osteoclasts by adding RANKL.42 Here we observed that IL-33 nearly abolished osteoclastogenesis (Fig. 8A), which again was accompanied by decreased expression of Oscar and Nfatc1 (Fig. 8B) but not by induction of Arg1. Based on these findings, we reasoned that a molecular analysis of RAW264.7 cells is the preferable tool to determine which relevant genes might be affected by IL-33 treatment at the transcriptional level. Therefore, we performed another Affymetrix Gene Chip hybridization in which we compared RAW264.7 cells treated with RANKL alone for 24 hours or treated with RANKL and IL-33 for the same time. While we observed expression of neither the eosinophil markers Ear11 and Ccl24 nor the M2 macrophage markers Arg1 and Chi3l4 in both samples (data not shown), we found that the expression of several genes associated with osteoclastogenesis was repressed in the presence of IL-33 (Fig. 8C). These included Atp6v0d2, encoding a V-ATPase subunit required for osteoclast fusion43; Nfatc1, encoding one of the key transcription factors inducing osteoclast-specific gene expression44; and the osteoclast differentiation markers Ctsk (encoding cathepsin K) and Acp5 (encoding TRACP). In contrast, while the expression of Fos, encoding another relevant transcription factor for osteoclastogenesis, was only moderately reduced, the expression levels of Csfr1 (encoding the M-CSF receptor FMS), Tnfrsf11a (encoding the RANKL receptor RANK), and Nfkb1 (encoding the p50 subunit of NF-κB) were virtually unaffected by IL-33 administration. This was further confirmed by quantitative RT-PCR, where we observed a marked repression of Nfatc1 expression in RAW264.7 cells treated with IL-33, but no immediate effect of IL-33 on the expression of Tnfrsf11a (Fig. 8D).

thumbnail image

Figure 8. IL-33 mediates a direct antiosteoclastogenic effect on RAW264.7 macrophages. (A) TRACP activity staining of multinucleated osteoclasts in RAW264.7 cultures differentiated in the presence of 50 ng/mL of RANKL and various concentrations of IL-33 for 3 days. The quantification of TRACP+ multinucleated cells per visual field (VF) is given below. Values represent mean ± SD of two independent experiments performed in triplicate (n = 6). (B) RT-PCR expression analysis for Il1rl1, Oscar, and Nfatc1 in RAW264.7 cultures differentiated for 1, 2, or 3 days in the absence or presence of 100 ng/mL of IL-33. (C) Affymetrix signal intensities and SLRs for the expression of osteoclastogenesis-associated genes following a 24-hour differentiation in the presence of 100 ng/mL of IL-33. (D) Quantitative RT-PCR for the expression of Nfatc1, Fos, and Tnfrsf11a in RAW264.7 cultures differentiated for 1, 2, or 3 days in the absence or presence of 100 ng/mL of IL-33. Bars represent mean ± SD of two independent experiments performed in triplicate (n = 6). The dotted red lines indicate the mean expression levels in the absence of IL-33.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

IL-33 was only recently discovered as a new member of the IL-1 family that serves as a ligand for the receptor ST2.26 Although its biologic activities are just beginning to be understood, in vivo studies have suggested a role of IL-33 in stimulating the production of TH2 cytokines such as IL-4. This is also underscored by the finding that Il1rl1-deficient mice, lacking functional ST2, display impaired TH2-mediated immune responses in a pulmonary granuloma model induced with Schistosoma mansoni eggs.25 In addition, IL-33 has been shown to promote the production of proinflammatory cytokines by mast cells,39, 45 and together with the finding that Il1rl1−/− mice displayed an attenuated response to collagen-induced arthritis, these results suggested a critical role of IL-33 in the pathogenesis of rheumatoid arthritis.27 Other studies have highlighted the potential relevance of ST2 in the cardiovascular system because the impairment of cardiac functions following aortic constriction was more pronounced in Il1rl1−/− mice than in their wild-type littermates.46, 47 In this article, we were able to demonstrate another previously unknown function of the IL-33/ST2 system, namely, an inhibitory effect on osteoclastogenesis. This is underscored not only by in vitro experiments demonstrating a negative influence of IL-33 on the formation of osteoclasts from bone marrow precursor cells but also by the analysis of Il1rl1−/− mice.

Our study was initiated based on the finding that Il33 expression increases in the course of primary osteoblast differentiation, which led us to perform in vitro experiments to analyze the effects of IL-33 on osteoblasts. Here we observed that IL-33 treatment did not affect the production of mineralized matrix but led to an increased expression of specific cytokine-encoding genes, including Tnfsf11. Since this induction of gene expression was not observed in cultures lacking ST2, we went on to analyze the skeletal phenotype of Il1rl1−/− mice. Here we did not observe a significant difference in terms of osteoblast numbers and bone-formation rate, thereby demonstrating that IL-33 does not influence bone formation in vivo. Moreover, since we observed a marked increase in osteoclastogenesis in Il1rl1−/− mice, it appears that a ST2-dependent stimulation of RANKL production by osteoblasts is not able to counteract the direct inhibitory influence of IL-33 on osteoclast formation, which was confirmed subsequently in a series of in vitro experiments.

Here we found that IL-33, at concentrations above 0.5 ng/mL, completely abolished the generation of functional TRACP+ multinucleated osteoclasts from bone marrow precursor cells, even in the presence of RANKL (40 ng/mL) and M-CSF (20 ng/mL). By performing the same experiments with cells from Il1rl1−/− mice, we were further able to demonstrate that the antiosteoclastogenic effect of IL-33 requires the presence of ST2. This was not necessarily expected because there are several indications for a nuclear function of pro-IL-3348–50 and because the administration of recombinant IL-33 also stimulated endogenous Il33 expression. In contrast, the antiosteoclastogenic effect of IL-33 still was observed when we performed the same experiments with cells from Stat6−/− mice, which is important because the TH2 cytokines IL-4 and IL-13, whose expression is induced by IL-33, have been shown to mediate a RANKL-independent antiosteoclastogenic effect in a Stat6-dependent manner.5, 40

To understand the molecular mechanisms underlying the observed effects of IL-33 on the differentiation of bone marrow precursor cells, we first analyzed gene expression in cultures that had been differentiated for 6 days in either the absence or the presence of IL-33. Here we observed that IL-33 promoted the generation of specific myeloid cell populations at the expense of osteoclasts. This finding, which was confirmed by FACS analysis, is in agreement with recently published studies by others demonstrating a positive influence of IL-33 on the generation of eosinophils, basophils, and alternatively activated macrophages.51–53 Since these results raised the possibility that the antiosteoclastogenic effect of IL-33 is mediated indirectly through other cytokines expressed by these cell types, we finally used the macrophage cell line RAW264.7 to demonstrate a direct negative effect of IL-33 on the generation of multinucleated osteoclasts, which is best explained by the observed negative influence on Nfatc1 expression. In fact, Nfatc1 is known as a key transcription factor of osteoclastogenesis regulating the expression of specific genes such as Ctsk, Acp5, Atp6v0d2, and Tm7sf4, the latter one encoding the transmembrane protein DC-STAMP, which is essential for the fusion of osteoclast precursor cells.44, 54–57 Since all these genes were found to be expressed at lower rates in RAW264.7 cells treated with IL-33, we believe that IL-33 exerts its antiosteoclastogenic effect, at least in part, by reducing Nfatc1-dependent gene expression in mononuclear osteoclast precursor cells, thereby impairing their fusion capacity.

Regardless of the underlying mechanism by which IL-33 inhibits osteoclastogenesis, however, we believe that the most important finding reported in our article is the increased osteoclastogenesis of Il1rl1−/− mice. This phenotype is especially remarkable because it develops spontaneously, whereas all the other functions reported for ST2 so far had to be uncovered by challenging the Il1rl1−/− mice with specific stimuli. This implies that regulation of bone resorption is one of the major functions of the IL-33/ST2 interaction and that activating this system might be one therapeutical approach for the treatment of bone-loss disorders.

Disclosures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

All the authors state that they have no conflicts of interest.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

This study was supported by the Focus Program SPP1468 “Osteoimmunology—IMMUNOBONE—A program to unravel the mutual interactions between the immune system and bone” of the Deutsche Forschungsgemeinschaft.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References