SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

To investigate the effects of interleukin-17A (IL-17A) on osteoclastogenesis in vitro.

Methods

Bone marrow cells (BMCs) were isolated from the excised tibia and femora of wild-type C57BL/6J mice, and osteoblasts were obtained by sequential digestion of the calvariae of ddY, C57BL/6J, and granulocyte–macrophage colony-stimulating factor–knockout (GM-CSF−/−) mice. Monocultures of BMCs or cocultures of BMCs and osteoblasts were supplemented with or without 1,25-dihydroxyvitamin D3 (1,25[OH]2D3), recombinant human macrophage colony-stimulating factor (M-CSF), RANKL, and IL-17A. After 5–6 days, the cultures were fixed with 4% paraformaldehyde and subsequently stained for the osteoclast marker enzyme tartrate-resistant acid phosphatase (TRAP). Osteoprotegerin (OPG) and GM-CSF expression were measured by enzyme-linked immunosorbent assay, and transcripts for RANK and RANKL were detected by real-time polymerase chain reaction.

Results

In both culture systems, IL-17A alone did not affect the development of osteoclasts. However, the addition of IL-17A plus 1,25(OH)2D3 to cocultures inhibited early osteoclast development within the first 3 days of culture and induced release of GM-CSF into the culture supernatants. Furthermore, in cocultures of GM-CSF−/− mouse osteoblasts and wild-type mouse BMCs, IL-17A did not affect osteoclast development, corroborating the role of GM-CSF as the mediator of the observed inhibition of osteoclastogenesis by IL-17A.

Conclusion

These findings suggest that IL-17A interferes with the differentiation of osteoclast precursors by inducing the release of GM-CSF from osteoblasts.

Equilibrium between bone formation and bone resorption is crucial for the maintenance of skeletal integrity in humans (1). In inflammatory diseases affecting bone, this balance between bone formation and resorption is shifted toward resorption. Rheumatoid arthritis (RA) is an example of a chronic inflammatory autoimmune disease with unknown etiology that primarily affects the human skeleton (2). Proinflammatory cytokines such as interleukin-1 (IL-1) and tumor necrosis factor α (TNFα) trigger multiple inflammatory pathways that lead to synovitis, hyperplasia of the synovial membrane that ends in cartilage, and bone destruction (3).

Osteoclasts are terminally differentiated cells formed by the fusion of mononuclear progenitor cells belonging to the monocyte/macrophage lineage (1). Macrophage colony-stimulating factor (M-CSF) and RANKL are 2 essential molecules produced by osteoblasts/stromal cells that enable differentiation and subsequently fusion of osteoclast precursors. In vitro, M-CSF is constitutively expressed by stromal cells (4, 5). RANKL is presented on the surface of osteoblasts in vitro and in vivo. The formation of osteoclasts involves a cascade of complex events that occur following binding of RANKL to its receptor RANK on osteoclast precursor cells (OPCs). Activation of RANK initiates signaling by the adapter molecule TNF receptor–associated factor 6. Subsequently, several signaling pathways lead to sequential up-regulation of the transcription factors NF-κB, c-Fos, Fra-1, and NF-ATc1 (6–9). The steroid hormone 1,25-dihydroxyvitamin D3 (1,25[OH]2D3) takes part in calcium and phosphorous homeostasis in vivo (10). In vitro, the addition of 1,25(OH)2D3 to cocultures of osteoblasts and bone marrow cells (BMCs) leads to the formation of tartrate-resistant acid phosphatase (TRAP)– positive osteoclasts by up-regulating RANKL (11).

Certain cells of the immune system develop within the confined environs of the bone marrow cavity. The bone marrow cavity allows close physical proximity and interactions between these immune cells and various bone cell lineages (12). Thereby, in inflammatory conditions, activation of cells of the immune system modulates the bone microenvironment by releasing cytokines that affect the bone cell lineages (13, 14). It is known that activated T cells secrete cytokines such as soluble RANKL (sRANKL), TNFα, IL-6, and IL-17A (2, 15, 16). These cytokines support enhanced progenitor cell proliferation and differentiation, increased cell survival via initiation of antiapoptotic pathways, and enlargement of the marrow osteoclast precursor pool and subsequent activation of bone-resorbing multinucleated giant osteoclasts in situ (17–20).

Simultaneously, activated T cells produce a repertoire of cytokines such as IL-4, IL-10, IL-13, interferon-γ (IFNγ), and granulocyte–macrophage colony-stimulating factor (GM-CSF) that play a role in inhibiting the differentiation of OPCs into multinucleated osteoclasts (21). In RA, both antiinflammatory and proinflammatory cytokines coexist in the microenvironment of inflamed joints. The imbalance in favor of proinflammatory and pro-osteoclastogenic cytokines such as TNFα and IL-1 stimulates osteoclast differentiation and leads to the activation of resorption that finally results in juxtaarticular bone resorption and joint destruction (3).

IL-17A production and secretion have been attributed to a distinct subset of CD4+ helper T cells (22, 23). Recently, it was demonstrated that not only CD4+ α/β T cells but also CD8+ α/β T cells, natural killer cells, and γ/δ T cells as well as macrophages and neutrophils are capable of producing IL-17A. IL-17A through IL-17F are 6 known members of the IL-17 family, of which IL-17A is by far the best characterized (24, 25).

IL-17A levels have been shown to be significantly increased in the synovial fluid of patients with RA, and collagen-induced arthritis (CIA) was markedly suppressed in IL-17−/− mice (26). IL-17 was shown to induce osteoclastogenesis in cocultures of murine osteoblast lineage cells and BMCs by up-regulating RANKL and prostaglandin E2 (PGE2) production by osteoblasts in vitro (20). These findings led to the supposition that IL-17A plays an important role in inflammation-associated osteoclastogenesis both in vivo and in vitro.

In the present study, the role of IL-17A in osteoclast development in vitro was further investigated. The results demonstrated that IL-17A had no direct effect on osteoclast precursors in BMCs. However, in cocultures of murine BMCs with primary murine osteoblasts, various concentrations (0.1–50 ng/ml) of IL-17A suppressed osteoclastogenesis by inducing the production of a soluble inhibitor of osteoclast development. We identified this soluble inhibitor as osteoblast-derived GM-CSF and demonstrated that in cocultures of GM-CSF−/− mouse osteoblasts and wild-type (WT) mouse BMCs, the inhibitory effect of IL-17A on osteoclast formation was lost.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Mice.

Wild-type C57BL/6 mice, GM-CSF−/− mice (kindly provided by Dr. K. L. Rudolph and Dr. A. Gompf, Ulm University, Germany), and ddY mice were bred and housed in the central animal facility of the Department of Clinical Research, University of Bern, in compliance with the Swiss and US National Institutes of Health guidelines for care and use of experimental animals. Use of the animals in the experiments of this study was approved by the State Committee for the Control of Animal Experiments (permit no. 13/07 to WH).

Cell isolation and culture.

Primary murine osteoblasts were isolated from 1–2-day-old mice by sequential collagenase digestion. Briefly, 25 calvariae were digested for 5 × 20 minutes in Hanks' balanced salt solution (HBSS; Sigma) containing 3 mg/ml collagenase II. Cells (106) were placed in 75-cm2 tissue culture flasks along with cell culture medium (α-minimum essential medium containing 10% fetal bovine serum [Inotech] and penicillin/streptomycin [100 units/ml and 100 μg/ml, respectively; Gibco-BRL Life Technologies]). Cells were allowed to grow for 4 days and then were harvested. Aliquots (106 cells/ml) were stored in liquid nitrogen until used. Before the experiments, an aliquot of osteoblasts was thawed and allowed to expand in cell culture medium for 4 days.

Osteoclasts were developed in vitro either in cultures of BMCs alone supplemented with M-CSF and RANKL or in cocultures of osteoblasts and BMCs. BMCs were isolated from 6–8-week-old male C57BL/6J mice by flushing the bone marrow from the excised tibia and femora with HBSS. After centrifugation at 1,200 revolutions per minute for 10 minutes at 4°C, the cell pellet was resuspended in cell culture medium. BMCs were grown in BD Falcon 96-well plates (Fisher Scientific) at a density of 8 × 103 in 0.1 ml cell culture medium, supplemented with 30 ng/ml M-CSF (kindly provided by Chiron) and 0, 1, and 10 ng/ml recombinant human sRANKL (PeproTech). The medium was changed after 3 days of culture. The numbers of newly formed osteoclasts were determined from day 4 to day 6.

In cocultures of primary osteoblasts and BMCs, 4 × 103 osteoblasts and 6 × 104 BMCs were grown in BD Falcon 48-well plates (Fisher Scientific) in cell culture medium supplemented with 1,25(OH)2D3 (Hoffmann-La Roche). The medium was changed after 3 days. In the experiments in which IL-17A (PeproTech) was added to the cultures, the cytokine was added throughout at a concentration of 50 ng/ml, unless stated otherwise.

Determination of osteoclast number.

To visualize osteoclasts, the cell cultures were stained for the marker enzyme TRAP, using a commercially available kit (Sigma). TRAP-positive cells with ≥3 nuclei were counted as multinucleated osteoclast-like cells. Before staining, the cells were fixed with 4% paraformaldehyde (Merck) in phosphate buffered saline for 10 minutes. Subsequently, cells were washed with distilled water 3 times. Plates were allowed to dry at room temperature overnight. The TRAP enzyme present in the cells was stained by adding substrate for TRAP and an appropriate reaction buffer with acidic pH (0.3 mg/ml diazotized Fast Garnet GBC, 2.5M acetate solution, 0.67M tartrate, and 12.5 mg/ml naphthol; Acid Phosphatase, Leukocyte Kit [Sigma-Aldrich]) for 5 minutes. The solution was discarded, and the plates were washed with distilled water 3 times and allowed to dry overnight before the TRAP-stained cells were counted.

Conditioned medium.

To investigate whether the cocultures, when treated with IL-17A, released any soluble modulators of cell development and/or activity into the cell culture supernatants, cocultures with 4 × 103 osteoblasts and 6 × 104 BMCs were seeded on 48-well plates in 0.2 ml cell culture medium supplemented with 1,25(OH)2D3 and 1,25(OH)2D3 plus IL-17A (50 ng/ml), respectively. Cell culture supernatants were collected on day 3. The effect of conditioned medium on the formation of TRAP-positive osteoclasts was tested in cultures of BMCs grown in the presence of M-CSF (30 ng/ml) and RANKL (10 ng/ml), replacing 1% and 5%, respectively, of the cell culture medium with conditioned medium. The numbers of newly formed osteoclasts were counted from day 4 to day 6.

Quantitative reverse transcription–polymerase chain reaction (PCR).

To determine the levels of transcripts encoding RANKL and RANK in osteoblasts and BMCs, cells were seeded in 24-well plates (1 × 104 osteoblasts and 1.5 × 105 BMCs) and grown for 3 days. Total RNA was isolated using an RNeasy Mini Kit (Qiagen), according to the manufacturer's instructions. Total RNA was reverse transcribed using Reverse Transcriptase M-MuLV (Roche Diagnostics). The reaction mixes were incubated for 2 minutes at 50°C followed by 10 minutes at 95°C. Thereafter, 45 cycles of 15 seconds at 95°C and 1 minute at 60°C each were performed. PCR was performed with Assays-on-Demand (Applied Biosystems) on an ABI 7500 Prism system. The transcript levels were normalized to β-glucuronidase (Mm00446953_m1), and the reactions were performed with TaqMan Fast Universal Master Mix. For quantitative PCR, the following Assays-on-Demand were used: RANKL/Tnfsf11 (Mm00441908), RANK/ Tnfsf11a (Mm00437135_m1), IFNγ (Mm00801778_m1), IL-4 (Mm00445259_m1), IL-10 (Mm00439616_m1), and IL-13 (Mm00434206_g1).

Osteoprotegerin (OPG) and GM-CSF measurements.

OPG protein levels were measured in the supernatants of cocultures of primary osteoblasts from ddY mice and WT mouse BMCs, which were treated with 10−8M 1,25(OH)2D3 with or without IL-17A (50 ng/ml), using a DuoSet ELISA (enzyme-linked immunosorbent assay) Development System (R&D Systems). GM-CSF protein levels were measured in the supernatants of cocultures of primary osteoblasts from WT, GM-CSF−/−, and ddY mice and WT mouse BMCs that were treated with 10−8M 1,25(OH)2D3 with or without IL-17A (50 ng/ml), using a BD OptEIA Mouse GM-CSF ELISA set (BD Biosciences).

Statistical analysis.

Differences in osteoclast numbers and in OPG and GM-CSF protein levels were evaluated by nonparametric t-tests using GraphPad Prism version 5 for Windows (www.graphpad.com).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

IL-17A inhibits osteoclast formation in vitro.

To assess the effect of IL-17A on osteoclast formation, cocultures of osteoblasts and BMCs were performed. IL-17A was added to the cocultures at different concentrations (0, 0.1, 1, 10, and 50 ng/ml), either alone or in combination with 10−8M 1,25(OH)2D3. The addition of 1,25(OH)2D3 alone allowed osteoclasts to develop, while treatment with IL-17A at any concentration in the absence of 1,25(OH)2D3 did not induce osteoclast formation. The addition of increasing concentrations of IL-17A (0.1–50 ng/ml) to the cocultures along with 1,25(OH)2D3 led to a dose-dependent inhibition of osteoclast formation (range 11–90%) (Figure 1A).

thumbnail image

Figure 1. A, Effect of interleukin-17A (IL-17A) on the development of osteoclasts (OCLs) in cocultures of osteoblasts and bone marrow cells (BMCs). Treatment of the cocultures with IL-17A alone had no effect on the formation of osteoclasts. The addition of IL-17A to 1,25-dihydroxyvitamin D3 (1,25[OH]2D3)–treated cocultures resulted in a dose-dependent inhibitory effect on 1,25(OH)2D3-mediated osteoclast formation (open bars). The solid bar represents control culture stimulated with 1,25(OH)2D3. The experiments were repeated 3 times with similar results. Bars show the mean ± SD of quadruplicate cultures. ∗ = P < 0.01; ∗∗∗ = P < 0.0001 versus control. B, Effect of IL-17A on osteoclast development in cultures of BMCs stimulated with macrophage colony-stimulating factor (M-CSF) and RANKL. Treatment with IL-17A alone had no effect on osteoclast formation in BMCs. M-CSF was used at a constant dose of 30 ng/ml throughout the cultures. The addition of IL-17A to cultures treated with only M-CSF had no effect on the formation of osteoclasts. The addition of IL-17A (10 ng/ml and 50 ng/ml) to cultures treated with both M-CSF and RANKL (1 ng/ml) did not have any effect on osteoclast development. Bars show the mean ± SD of triplicate cultures. TRAP = tartrate-resistant acid phosphatase.

Download figure to PowerPoint

To discern whether IL-17A exerts its inhibitory effect directly on BMCs or via osteoblasts, cultures of BMCs were supplemented with M-CSF (30 ng/ml) and RANKL (0, 1, and 10 ng/ml). Adding RANKL at 10 ng/ml to M-CSF–treated cultures induced the development of multinuclear TRAP-positive cells (mean ± SD 186.3 ± 16.25) (positive control). The addition of IL-17A (10 ng/ml and 50 ng/ml) to the cultures along with RANKL did not affect osteoclast formation (157 ± 14.73 and 175.33 ± 35.11, respectively) (Figure 1B).

IL-17A regulates RANKL and RANK gene expression but does not regulate OPG synthesis in cocultures.

The RANKL/OPG system is critical for the formation of osteoclasts. Therefore, we assessed whether the inhibitory effect of IL-17A on 1,25(OH)2D3-mediated osteoclastogenesis was attributable to alterations in the OPG and RANKL protein and transcript levels, respectively. We measured transcript levels of RANK and RANKL in cell lysates of cocultures and OPG protein levels in the supernatants of cocultures stimulated with 1,25(OH)2D3 with or without IL-17A (50 ng/ml) on day 3.

The addition of 1,25(OH)2D3 to the cocultures of BMCs and osteoblasts led to a 3.5-fold increase in the levels of transcripts encoding RANKL compared with unstimulated control cultures (Figure 2A). Simultaneously, OPG protein levels were significantly reduced compared with unstimulated controls (mean ± SD 236.58 ± 51.74 pg/ml and 1,130.5 ± 112.5 pg/ml, respectively) when 1,25(OH)2D3 was added to the cocultures (Figure 2C). The addition of IL-17A to cocultures up-regulated RANKL transcript levels by a factor of 5. OPG protein levels were not affected by treatment of the cultures with IL-17A (872.25 ± 107.89 pg/ml). The addition of IL-17A (50 ng/ml) concomitantly with 1,25(OH)2D3 resulted in a 5.8-fold up-regulation in RANKL transcript levels and a decrease in OPG protein levels (9.86 ± 8.2 pg/ml versus 1,130.5 ± 112.5 pg/ml in unstimulated controls; P < 0.01), allowing the conditions to be pro-osteoclastogenic. Relative RANK transcript levels were up-regulated 1.8-fold in cocultures that contained 1,25(OH)2D3 and were down-regulated 1.8-fold in cocultures that contained both IL-17A (50 ng/ml) and 1,25(OH)2D3 as compared with untreated controls (Figure 2B).

thumbnail image

Figure 2. Effect of IL-17A on RANK/RANKL gene expression and osteoprotegerin (OPG) synthesis in cocultures. Primary osteoblasts from murine calvariae were treated with IL-17A (50 ng/ml) and 1,25(OH)2D3. After 3 days, the levels of transcripts encoding RANKL (A) and RANK (B) were determined by quantitative real-time polymerase chain reaction. OPG protein levels in the culture supernatants were quantified by enzyme-linked immunosorbent assay (C). Treatment with 1,25(OH)2D3 alone up-regulated transcripts encoding RANKL and RANK. The addition of IL-17A up-regulated RANKL transcript levels (A) and abolished the increase in RANK levels observed with 1,25(OH)2D3 alone (B). OPG protein accumulated during the culture period, and treatment of the cocultures with 1,25(OH)2D3 resulted in a significant decrease in OPG protein production. IL-17A had no effect on OPG protein levels. Bars show the mean ± SD of 3 wells from 1 representative experiment. ∗ = P < 0.01 versus untreated control. Rel. = relative (see Figure 1 for other definitions).

Download figure to PowerPoint

IL-17A– and 1,25(OH)2D3-treated cocultures of osteoblasts and BMCs release a soluble factor that inhibits osteoclast development.

To discern whether the IL-17A–mediated inhibition of osteoclast formation requires cell–cell contact or is the result of soluble product(s) that are released into the supernatants, conditioned medium was collected from cocultures of osteoblasts and BMCs treated with IL-17A (50 ng/ml) and 1,25(OH)2D3 on day 3. The conditioned medium collected was subsequently added to cultures of BMCs. Cultures supplemented with M-CSF (30 ng/ml) and RANKL (10 ng/ml) without the addition of any conditioned medium gave rise to osteoclasts. The number of osteoclasts that formed on day 5 was counted. Culture medium that was supplemented with 1% and 5% conditioned medium showed 25% (P < 0.01) and 90.8% (P < 0.0001) inhibition of osteoclast formation, respectively (Figure 3). Culture medium supplemented with M-CSF and RANKL along with control conditioned medium did not affect osteoclast formation (data not shown).

thumbnail image

Figure 3. IL-17A–induced secretion of soluble inhibitors of osteoclast formation. Conditioned medium (CM) was collected on day 3 from cocultures of osteoblasts and BMCs stimulated with 1,25 (OH)2D3 and IL-17A (0.50 ng/ml). Conditioned medium (1% and 5%) was added to the cultures that were stimulated with M-CSF (30 ng/ml) and RANKL (10 ng/ml). After 5 days of culture, the TRAP-positive mononuclear cells were counted. Bars show the mean ± SD of triplicate cultures. ∗ = P < 0.01; ∗∗∗ = P < 0.0001 versus control. See Figure 1 for other definitions.

Download figure to PowerPoint

IL-17A affects early differentiation of osteoclast precursors in cocultures by irreversible down-regulation of RANK transcript levels.

In order to assess whether IL-17A affects 1,25(OH)2D3-mediated osteoclast formation during the early stages (day 0 to day 3) or the late stages (day 4 to day 6) of osteoclast development, the cocultures were supplemented with exogenous IL-17A (50 ng/ml) either in the initial phase (days 0–3) or in the late phase of the experiment (days 4–6). Controls included cocultures supplemented with 1,25(OH)2D3 alone and cocultures supplemented with both IL-17A (50 ng/ml) and 1,25(OH)2D3 for the duration of the culture period. The addition of IL-17A (50 ng/ml) in the initial phase completely inhibited 1,25(OH)2D3-mediated development of TRAP-positive osteoclast-like cells (P < 0.0001). The addition of IL-17A (50 ng/ml) in the late phase did not have any effect on osteoclast development (Figure 4A).

thumbnail image

Figure 4. Effect of IL-17A on early differentiation of osteoclast precursors in cocultures. A, Control cocultures (solid bars) were treated with 1,25(OH)2D3 with or without IL-17A (50 ng/ml) from day 0 to day 6. The addition of IL-17A to cocultures during the early phase (days 0–3) inhibited 1,25(OH)2D3-mediated development of TRAP-positive mononuclear cells. The addition of IL-17A to the cocultures during the late phase (days 4–6) had no effect on the development of TRAP-positive cells. ∗∗∗ = P < 0.0001 versus control. B, Cocultures of primary murine osteoblasts and BMCs were stimulated with 1,25(OH)2D3 with or without IL-17A on days 3, 4, and 6. Cocultures treated with 1,25(OH)2D3 and IL-17A on days 0–3 were resupplemented with IL-17A for an additional 24–72 hours after medium change or were not resupplemented with IL-17A after medium change. For mRNA extraction, the cocultures were lysed at 24 hours and 72 hours. Unstimulated cultures and cultures treated with 1,25(OH)2D3 with or without IL-17A throughout and were not subjected to withdrawal or addition of either IL-17A or 1,25(OH)2D3 served as controls. Withdrawal of IL-17A after 3 days had no visible effect on transcript levels. Bars show the mean ± SD of triplicate cultures from 1 representative experiment. Rel. = relative (see Figure 1 for other definitions).

Download figure to PowerPoint

It was observed that the addition of IL-17A to cocultures supplemented with 1,25(OH)2D3 blocked the up-regulation of transcripts encoding RANK messenger RNA (mRNA) (Figure 2B). To assess whether the observed effect of IL-17A on transcripts encoding RANK in BMCs of such cocultures could be reversed subsequent to the removal of IL-17A, the cocultures that were treated with IL-17A and 1,25(OH)2D3 from day 0 to day 3 next received only 1,25(OH)2D3 for a further 24–72 hours; thus, the length of culture was adequate to allow changes in transcript levels. Thereafter, the cocultures were stopped and lysed for mRNA extraction. Cocultures that were treated with IL-17A and 1,25(OH)2D3 showed 1.9-fold lower levels of transcripts encoding RANK mRNA compared with controls that received 1,25(OH)2D3 alone, regardless of when the cocultures were stopped (Figure 4B). However, the cocultures that were stimulated with IL-17A and 1,25(OH)2D3 for 0–3 days followed by discontinuation of any IL-17A treatment for a further 24–72 hours showed unmodulated RANK transcript levels.

IL-17A induces an increase in GM-CSF expression in cocultures.

To assess whether IL-17A stimulates known inhibitors of osteoclast development, transcript levels of IFNγ, GM-CSF, IL-4, IL-10, and IL-13 were determined in cocultures stimulated with 1,25(OH)2D3 with or without IL-17A (50 ng/ml). Transcripts encoding IFNγ, IL-4, IL-10, and IL-13 did not show any significant modulation in any condition in cocultures of primary murine osteoblasts and BMCs (data not shown). In contrast, transcript levels of GM-CSF were up-regulated 130-fold in cocultures stimulated with 1,25(OH)2D3 plus IL-17A (50 ng/ml) (Figure 5A) compared with unstimulated controls. In order to confirm whether the soluble inhibitor secreted into the supernatants could be GM-CSF, GM-CSF protein levels were measured by ELISA. Neither 1,25(OH)2D3 nor IL-17A alone stimulated production of GM-CSF protein. The addition of IL-17A to 1,25(OH)2D3, however, led to the production of GM-CSF protein in cell culture supernatants after day 3 (mean ± SD 110 ± 5.64 pg/ml [range 102.39–121.11]; P < 0.0001) (Figure 5B).

thumbnail image

Figure 5. Up-regulation of transcript and protein levels of granulocyte–macrophage colony-stimulating factor (GM-CSF) in cocultures treated with IL-17A and 1,25(OH)2D3. Cocultures of primary osteoblasts from ddY mice and wild-type mouse BMCs were treated with 1,25(OH)2D3 with or without IL-17A. After 3 days, transcript levels of GM-CSF were measured by quantitative real-time polymerase chain reaction (A), and protein levels were quantified by enzyme-linked immunosorbent assay (B). GM-CSF transcript levels were increased in cocultures treated with IL-17A and 1,25(OH)2D3 (A). Neither 1,25(OH)2D3 alone nor IL-17A alone induced release of GM-CSF protein, but the combination increased GM-CSF protein expression after day 3. Bars show the mean ± SD of 3 wells from 1 representative experiment. ∗∗∗ = P < 0.0001 versus control. Rel. = relative (see Figure 1 for other definitions).

Download figure to PowerPoint

IL-17A fails to inhibit osteoclastogenesis in cocultures with GM-CSF−/− mouse osteoblasts.

To confirm whether GM-CSF was the soluble mediator stimulated by IL-17A and was responsible for the inhibition of 1,25(OH)2D3-mediated osteoclast formation, we repeated the above-described experiments in cocultures by using primary osteoblasts from GM-CSF−/− mice and BMCs from WT C57BL/6J mice. TRAP-positive cells were counted in cocultures of osteoblasts and BMCs obtained from WT C57BL/6J mice (Figure 6A) and compared with osteoclast numbers obtained on day 5 in cocultures of osteoblasts from GM-CSF−/− mice and WT mouse BMCs (Figure 6B) stimulated with 1,25(OH)2D3 and TNFα with or without IL-17A (50 ng/ml). In the cocultures that were treated with 1,25(OH)2D3 and 0.5 ng/ml TNFα, the osteoclast numbers were comparable (mean ± SD 311 ± 5.29 versus 266.6 ± 12.5). In WT mouse cocultures that were treated with 1,25(OH)2D3, 0.5 ng/ml TNFα, and 50 ng/ml IL-17A, no osteoclast formation was observed. Under these same conditions, IL-17A could no longer abrogate 1,25(OH)2D3-mediated osteoclastogenesis in cocultures of GM-CSF−/− mouse osteoblasts and WT mouse BMCs (320.6 ± 17.38) compared with WT mouse cocultures (P < 0.0001).

thumbnail image

Figure 6. Effect of IL-17A on osteoclastogenesis in cocultures of primary osteoblasts from granulocyte–macrophage colony-stimulating factor–knockout (GM-CSF−/−) mice and BMCs from wild-type (WT) C57BL/6J mice. TRAP-positive cells obtained after 5 days in cocultures of WT mouse osteoblasts and BMCs were counted (A) and compared with the numbers in cocultures of GM-CSF−/− mouse osteoblasts and WT mouse BMCs (B). In both sets of experiments, the cultures were stimulated with 1,25(OH)2D3 and tumor necrosis factor α (TNFα) with or without IL-17A. A, The addition of IL-17A completely abrogated 1,25(OH)2D3- and TNFα-mediated osteoclast formation in WT mouse cocultures. B, When GM-CSF−/− mouse osteoblasts were used in cocultures along with BMCs from C57BL/6 mice, IL-17A no longer abrogated 1,25(OH)2D3- and TNFα-mediated osteoclast formation. Bars show the mean ± SD of triplicate cultures. ∗∗∗ = P < 0.0001 versus WT mouse cocultures. See Figure 1 for other definitions.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

In the present study, we showed that IL-17A exerts an inhibitory effect on 1,25(OH)2D3-mediated osteoclast formation. However, no direct role of IL-17A on osteoclast development from osteoclast precursors in BMCs was observed. We identified GM-CSF as the soluble inhibitory mediator that is released by primary murine osteoblasts in response to IL-17A. Experiments with GM-CSF−/− mouse osteoblasts and WT mouse BMCs confirmed the role of GM-CSF as the sole mediator responsible for osteoclast inhibition.

Osteoclastogenesis is a multistep process in which the RANK/RANKL/OPG system plays a pivotal role (27). Additionally, several cytokines and growth factors regulate osteoclast development by exerting stimulatory or inhibitory effects, depending on the microenvironment and model systems used. In RA, cells from the innate and adaptive immune systems infiltrate the inflamed joint. These immune cells release a host of proinflammatory cytokines such as sRANKL, IL-1, IL-6, TNFα, and IL-17 that play a role either directly by supporting juxtaarticular bone loss or indirectly via accessory cells (2).

IL-17 is a proinflammatory cytokine that stimulates human macrophages to produce TNFα and IL-1 (28). Laan et al (29) showed that IL-17 alone induced GM-CSF release in transformed human bronchial epithelial cells. Furthermore, TNFα enhanced the IL-17–induced expression of GM-CSF. Numasaki et al (30) showed that IL-17 alone failed to induce GM-CSF release in lung microvascular endothelial cells but markedly enhanced IL-1β– and TNFα-induced expression of GM-CSF. In contrast, 10−8M 1,25(OH)2D3 has been shown to inhibit GM-CSF release in human peripheral blood mononuclear cells (31). Although IL-17–mediated GM-CSF release by stromal cells has been observed, there have been few reports on the induction of GM-CSF release by IL-17A in osteoblasts. Here, it is shown that IL-17A alone did not stimulate GM-CSF production, but that IL-17A in combination with 1,25(OH)2D3 stimulated GM-CSF release in primary murine osteoblasts.

The specific action of IL-17A on osteoblasts in cocultures with BMCs was seen in the presence of 1,25(OH)2D3. In this regard, it is important to note that NF-κB binding sites and vitamin D–responsive elements have been identified in the GM-CSF promoter region (32, 33). Recently, downstream activation of NF-κB by IL-17A (34) and GM-CSF release (29) have been demonstrated to be important in the expression of protein in activated B cells and human umbilical vein endothelial cells, respectively. In our experiments, the observed effect could be attributable to the concurrent action of IL-17A and 1,25(OH)2D3 on GM-CSF promoter– or 1,25(OH)2D3-induced differentiation of osteoblasts that rendered bone-forming cells more sensitive to IL-17A. Interestingly, human peripheral blood monocytes have been shown to express receptors for IL-17 (35). However, in cultures of BMCs alone, the presence of IL-17A did not affect M-CSF– and RANKL-mediated osteoclast development.

GM-CSF is a proinflammatory cytokine that is released in a paracrine manner from T cells, macrophages, fibroblasts, and endothelial cells (36, 37) and has been studied extensively for its role in osteoclastogenesis. GM-CSF inhibits differentiation of osteoclasts in vitro. A recent study by Atanga et al showed that TNFα-mediated GM-CSF release by primary murine osteoblasts prevented differentiation of M-CSF–dependent OPCs by impeding an increase in the surface expression of RANK that is essential for osteoclast development (38).

Interestingly, other known inhibitors of osteoclastogenesis such as IL-4, IL-10, IL-13, and IFNγ remained unmodulated in the culture condition in which up-regulation of GM-CSF was observed. Our results contrast with those previously reported by Kotake et al (20). Those investigators observed abrogation of an IL-17A–mediated increase in osteoclast numbers in cocultures of primary murine osteoblast lineage cells and bone marrow cells by indomethacin, implicating PGE2 as an essential factor inducing differentiation of osteoclast precursors within the BMCs into TRAP-positive osteoclast-like cells. Simultaneously, IL-17A–mediated up-regulation of RANKL mRNA was also shown to be responsible for the observed increase in osteoclast development.

In our study, the addition of IL-17A alone induced up-regulation of transcripts encoding RANKL. IL-17A alone also increased the production of PGE2; furthermore, 1,25(OH)2D3 enhanced this effect of IL-17A by markedly up-regulating PGE2 production (data not shown). IL-17A, however, did not affect levels of OPG. Previously, Shen et al also did not observe an effect of IL-17A on OPG levels in MC3T3-E1 cells (39). This might explain the inability of IL-17A to stimulate osteoclast formation in cocultures of osteoblasts and BMCs alone. The addition of 1,25(OH)2D3 to cocultures has been shown to promote osteoclastogenesis by up-regulating RANKL and down-regulating OPG. Our results demonstrate that in cocultures that were stimulated with IL-17A and 1,25(OH)2D3, IL-17A induced the osteoblasts to release a cytokine that inhibits osteoclast formation and overrides the osteoclastogenic environment generated by 1,25(OH)2D3.

In murine models of arthritis, IL-17A has been shown to play an active role in the disease process. In IL-17−/− mice, CIA was markedly suppressed (26), and administration of anti–IL-17A antibodies to mice was shown to delay the progression of CIA (40). However, among ovariectomized mice, IL-17RA−/− littermates showed more bone loss due to estrogen deficiency than WT mice (41). The current in vivo studies portray a controversial picture of IL-17 in bone loss mechanisms, and the role of IL-17 appears to be dependent on the models adopted for investigations.

Mechanisms that inhibit the process of osteoclastogenesis are numerous and mainly affect RANK–RANKL signaling or M-CSF signaling. In mice, deficiency of RANK (42), RANKL (43), and M-CSF (44) results in osteopetrosis. The RANKL:OPG ratio has been known to be a major determinant of osteoclast development (45). In our study, no limitation of RANKL or augmented OPG levels could be associated with the observed inhibition of osteoclast formation. However, the levels of transcripts encoding RANK were low in cocultures treated with IL-17A and 1,25(OH)2D3. This observation could be attributable to a block in the differentiation process induced by GM-CSF, as seen by other investigators (38), that is present in the microenvironment of OPCs. It is known that GM-CSF inhibits differentiation of OPCs and down-regulates the activator protein 1 complex (Fra-1 and NF-ATc1) and retains the cells in an undifferentiated state (38).

Both IL-17A and GM-CSF are known to be important mediators of inflammation, and their abrogation has been shown to have beneficial effects in murine models of arthritis (46). In these models, the presence of both IL-17A and GM-CSF can be concurrently seen with pronounced osteoclastogenesis, suggesting a contradiction of the data presented herein. Based on the current observations, we postulate a model wherein IL-17A–induced GM-CSF might be able to increase the pool of monocytes in peripheral blood (47). This increase in the pool of osteoclast precursors induced by GM-CSF might then increase the migration of OPCs into bone, where, in the osteoclastogenic inflammatory microenvironment, these cells differentiate into mature bone-resorbing osteoclasts.

In summary, the present data demonstrate that GM-CSF plays a role in IL-17A–mediated inhibition of osteoclastogenesis. The observed abrogation is a result of a block in the up-regulation of RANK mRNA levels that is essential for osteoclast development. This effect of GM-CSF blocks differentiation of early osteoclast precursors into mature osteoclasts by retaining the cells in an undifferentiated state (38). However, at a later point in time and within a suitable microenvironment, such undifferentiated macrophages might adopt an alternative differentiation pathway.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Seitz had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Balani, Aeberli, Hofstetter, Seitz.

Acquisition of data. Balani.

Analysis and interpretation of data. Balani, Aeberli, Hofstetter, Seitz.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

We greatly appreciate the excellent technical support provided throughout this study by senior research assistants Silvia Dolder (Group for Bone Biology & Orthopaedic Research, Department of Clinical Research, University of Bern) and Richard Kamgang (Department of Rheumatology, Clinical Immunology & Allergology, University Hospital, Bern, Switzerland).

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES