Interleukin-37 inhibits osteoclastogenesis and alleviates inflammatory bone destruction

2.1 Reagents

Recombinant human IL-37b/IL-1F7b protein (7585-IL-025/CF), recombinant murine RANKL (462-TEC-010/CF), and macrophage colony stimulating factor (M-CSF; 416-ML-010) were from R&D Systems (Minneapolis, MN). Antibodies against p65, p-p65, inhibitor of κBα (IκBα), p-IκBα, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) were from Cell Signaling (Danvers, MA). Antibodies against nuclear factor of activated T cells 1 (NFATc1; sc-13033), myeloid differentiation factor 88 (MyD88; sc-74532), and single Ig IL-1-related receptor (SIGIRR; sc-515818) were from Santa Cruz Biotechnology (Santa Cruz, CA). Antibody against IL-37 (ab-57187) and RANKL (ab-9957) were obtained from Abcam (Cambridge, MA). Osteo Assay Surface was from Corning (Corning, NY). The cell counting kit-8 (CCK-8) was purchased from Dojindo Molecular Technologies (Dojindo, Japan). The tartrate-resistant acid phosphatase (TRAP) stain kit was from Sigma-Aldrich (St. Louis, MO). The Focal Adhesion and Actin Cytoskeleton Staining Kit was obtained from Millipore (Darmstadt, Germany). MyD88 pharmacologic inhibitor (ST 2825) was purchased from Medchemexpress (Princeton, NJ). LPSs from Escherichia coli O55:B5 were purchased from Sigma-Aldrich. Coimmunoprecipitation (Co-IP) Kit (Pierce; 88804) were purchased from Thermo Fisher Scientific Inc., (Waltham, MA).

2.2 Clinical specimens

Clinical specimens were collected from patients at Daping Hospital (Chongqing, China). The study followed the law and was approved by the Daping Hospital Clinical Ethics Committee. All specimens were collected with the informed consent of the patients. Bone and adjacent tissue specimens were removed from six patients with traumatic osteomyelitis-infected or noninfectious closed fractures. Samples were fixed with 4% paraformaldehyde (PFA) for 24 hr and processed into paraffin sections with a thickness of 5 μm for hematoxylin and eosin (H&E) staining. Immunohistochemistry for IL-37 was performed on sections using a primary antibody against IL-37 (1:200). Stained sections were observed under a microscope (DMI 6000B; Leica, Wetzlar, Germany).

2.3 Cytotoxicity and apoptosis assays

The cell counting kit-8 (CCK-8) assay was performed to measure cell proliferation, according to the manufacturer’s instructions. RAW264.7 cells were seeded onto 96-well plates at a density of 0.3 × 10⁴ cells/well. Cells were treated with different concentrations of IL-37b in the presence of RANKL (50 ng/ml) and M-CSF (50 ng/ml) for 48 and 72 hr, respectively. After completing the abovementioned procedures, 10 μl of CCK-8 working solution was added to each well and the plate was incubated at 37°C for 2 hr. The absorbance was measured at 450 nm using a spectrophotometer (Synergy H4; BioTek Instruments, Winooski, VT). We measured the apoptosis rate of cultured cells using flow cytometry (BD, Triangle, NC). Precursor cells were divided into a RANKL pretreatment group and nonpretreatment group, respectively, and were then stimulated with 100 ng/ml LPS and 50 ng/ml M-CSF for 72 hr. Annexin V-FITC/PI staining was performed to detect the apoptosis of cultured cells. This procedure was performed according to the manufacturer’s instructions (Life Technologies, Carlsbad, CA).

2.4 Cell culture and osteoclast differentiation analyses

RAW264.7 cells were purchased from the American Type Culture Collection (Rockville, MD). Before the induction experiment, mouse osteoclast precursor RAW264.7 cells were incubated in a complete medium supplemented with 10% fetal bovine serum ( Life Technologies), 1% penicillin and streptomycin (Thermo Fisher Scientific) at 37°C for 24 hr. Precursor cells were seeded onto a plate in the medium containing RANKL (50 ng/ml) and M-CSF (50 ng/ml) and stimulated with or without IL-37b (50–200 ng/ml) for 72 hr. Culture media were refreshed for 2 days. Isolated bone marrow macrophages from C57BL/6 mice were obtained using previously reported methods (Dou et al., 2016) and were cultured with M-CSF (50 ng/ml) and RANKL (50 ng/ml) for 5 days.

Cells were washed twice with phosphate-buffered saline (PBS) before an incubation with 4% paraformaldehyde for 20 min. Then, we evaluated the process of osteoclast formation by performing TRAP staining and the number of TRAP-positive multinucleated cells with three or more nuclei under an optical microscope (DMI 6000B; Leica). In addition, RAW264.7 cells were pretreated with RANKL and M-CSF for 24 hr and washed twice before being stimulated with M-CSF (50 ng/ml) or LPS (100 ng/ml) in the presence or absence of IL-37b (200 ng/ml) for 72 hr (Shinohara et al., 2015). After culture, cells were stained with TRAP and focal adhesion and actin cytoskeleton (FAK). Next, we investigated the effect of IL-37b on osteoclast differentiation under LPS-induced inflammatory conditions.

2.5 Bone resorption assay in vitro

To verify the effect of IL-37b on osteoclast bone resorption in vitro, preosteoclasts were transferred onto 96-well plates coated with hydroxyapatite (Osteo Assay Surface; Corning, NY). Cells were cultured with RANKL (50 ng/ml) and M-CSF (50 ng/ml) for 5 days in the presence or absence of IL-37b at 200 ng/ml. We evaluated pit formation using a method described in a previous report (Dou et al., 2016), and the absorption area was analyzed using the ImageJ software (NIH, Bethesda, MD).

2.6 RNA isolation and quantitative real-time polymerase chain reaction (qRT-PCR)

RAW264.7 cells were stimulated with RANKL (50 ng/ml) and M-CSF (50 ng/ml) for 24 or 72 hr. The experimental group was additionally supplemented with 100 or 200 ng/ml IL-37b. We extracted the total RNA from cultured cells using TRIzol reagent (Life Technologies). The RNA was synthesized into complementary DNAs using a reverse transcription kit  (Takara Bio, Shiga, Japan). qRT-PCR was performed using SYBR Premix Ex Taq II (Takara Bio Inc.) in a PCR detection system (Bio-Rad, Hercules, CA), and the primer sequences are shown in Table 1. The expression of the target gene was normalized to the levels of the GAPDH messenger RNA (mRNA).

2.7 Western blot and Co-IP analyses

We treated RAW264.7 cells that had been cultured with RANKL and M-CSF with 200 ng/ml IL-37b, and then, we extracted the total proteins in radioimmunoprecipitation assay (RIPA) buffer containing phosphatase inhibitors and protease inhibitors. For samples obtained during surgery, specimens were repeatedly crushed in liquid nitrogen, resulting in the production of a supernatant containing the protein sample. Lysates were separated on 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis gels and transferred to the polyvinylidene difluoride membranes. All membranes were blocked in a TBST solution (tris-buffered saline [TBS] containing 0.1% Tween) containing 5% nonfat milk for 3 hr before an overnight incubation with primary antibodies (NFATc1, GAPDH, IκBα, p-IκBα [phospho-S32], p-p65 [phospho-S536], p65, or IL-37) at 4°C. After the completion of this process, membranes were washed three times with TBST and immersed in secondary antibody at room temperature for 1.5 hr. For this experiment, GAPDH expression served as an internal control.

For the Co-IP analysis, RAW264.7 cells were treated with M-CSF (50 ng/ml), RANKL (50 ng/ml), and IL-37b (200 ng/ml) for 48 hr. Cells were lysed in RIPA buffer after washes with PBS. The supernatant was incubated with a SIGIRR antibody (1:50) or mouse IgG overnight at 4°C, followed by an incubation with protein A/G magnetic beads for 4 hr at 4°C. The complexes were washed with PBS and subjected to western blot analysis using antibodies against SIGIRR (1:2,000), MyD88 (1:2,000), and GAPDH.

2.8 Immunofluorescence staining

Precursor cells were incubated in the presence or absence of 200 ng/ml IL-37b for 1 hr before treatment with RANKL for an additional 30 min. Cells were fixed with 4% PFA for 20 min, and permeabilization and blocking was performed before the cells were incubated with the primary antibody at 4°C. Cells were washed three times with PBS and then incubated with a Cy3-conjugated secondary antibody and 4’,6-diamidino-2-phenylindole (DAPI) in the dark. The colocalization of the p65 protein and nucleus was consistent with the previous reports (Fuchsová, Novák, Kafková, & Hozák, 2002), and the gray value was analyzed using the ImageJ software (NIH, Bethesda, MD). For focal adhesion and actin cytoskeleton stainingthe blocked cells were stained with antivinculin antibody (1:300) and tetramethylrhodamine-5-isothiocyanate-conjugated phalloidin (1:300), which labels F-actin, and then incubated with an Alexa Fluor 488-conjugated secondary antibody. The nuclei were stained with DAPI before microscopic observations.

2.9 Inflammatory bone destruction model and radiographic analysis

Six-week-old C57BL/6 female mice were obtained from the animal center of the Third Military Medical University. The whole procedure was performed according to the guidelines of the Animal Care and Use Committee of the Third Military Medical University. Twelve mice were randomly allocated to three groups (n = 4 in each group): the control group (PBS-treated), an LPS-treated group, and the IL-37b-treated group, which received LPS and IL-37b (4 μg/day) treatments. All animals were anesthetized by an intraperitoneal injection of 4% chloral hydrate (5 μl/g) to reduce the level of suffering. As described in the literature (Chiang, Kyritsis, Graves, & Amar, 1999), the mice were subcutaneously injected with LPS (100 μg/day) for 7 days before sacrifice to induce inflammatory bone destruction. The calvarial bones were fixed, decalcified, and cut into 5-μm-thick sections for H&E staining and TRAP staining.

For the micro-computed tomography (micro-CT) analysis, the femur tissues were scanned with viva CT 40 instrument (Scanco Medical, Brüttisellen, Switzerland) and the bone volume per tissue volume was calculated as described in a previous study (Dou et al., 2016). The scan of the trabecular bone was performed from the growth plate. The 3D images and bone metrology analysis were obtained by standard methods (Scanco Medical; CT Evaluation Program, V5.0A).

2.10 Statistics

All procedures were repeated at least three times to verify the results. The analysis was performed using GraphPad Prism 7(GraphPad Software, La Jolla, CA). The data were represented as the mean ± SD, and the significant differences in data among groups were compared by one-way analysis of variance. Values of *P < 0.05 and **P < 0.01 were considered to be statistically significant.

3.1 High levels of the IL-37 protein accumulate in bone-infected lesions

Increased IL-37 levels have been confirmed in numerous inflammatory and allergic diseases (Quirk & Agrawal, 2014). To investigate whether inflammation caused the local production of IL-37, we examined the expression of the IL-37 protein in bones and surrounding tissues. Immunoblotting was also performed to evaluate the IL-37 expression in paraffin-embedded sections from fractures and the surrounding nonskeletal tissues. As shown in Figure 1a, IL-37 was expressed at high levels in the infected connective tissues near the fracture. In images of H&E staining, we observed significant inflammatory cell infiltration in sections from patients with bone infections, and we observed a locally high expression of IL-37 at approximately the same location using immunohistochemistry (Figure 1b). On the basis of these findings, the effects of elevated levels of the IL-37 protein on bone remodeling became the focus of our study.

3.2 IL-37 prevents inflammatory bone destruction in vivo

Although controversy exists regarding whether LPS may directly stimulate osteoclast differentiation, LPS causes inflammation, shock, and bone destruction in vivo. Therefore, we evaluated the potential correlation between the effect of IL-37 on bone loss and inflammation in the LPS-treated model. In fact, when the mice were subcutaneously injected with LPS, we observed a significant decrease in their activity and viability due to endotoxemia. The information we obtained from the sections of a 7-day treatment group showed that IL-37b effectively relieved LPS-induced localized inflammatory responses and severe bone destruction. In addition, TRAP staining revealed a significant reduction in the number of osteoclasts in the IL-37b-treated group (Figure 1c). Micro-CT was performed using isolated femurs. The bone metrology analysis revealed a significant increase in the bone volume per tissue volume and a decreased trabecular spacing following the application of IL-37b to the mouse model of LPS-induced bone loss (Figure 1d). Thus, IL-37b to be involved in the mechanism regulating inflammatory bone loss in vivo. Therefore, we further verified whether the protective effect of IL-37b on inflammatory bone destruction was achieved by regulating osteoclast differentiation in vitro.

3.3 IL-37 suppresses RANKL-mediated osteoclastogenesis and inhibits the bone-resorbing activity of osteoclasts

We performed the CCK-8 assay to confirm that the actions of IL-37b on osteoclast precursor RAW264.7 cells did not depend on cytotoxicity. On the basis of the cell proliferation data, treatment with 50–200 ng/ml IL-37b for 48 or 72 hr was not toxic and did not reduce the number of cells (Figure 2a).

To observe the direct impact of IL-37b on RANKL-induced osteoclastogenesis, precursor cells were treated with M-CSF and RANKL for 72 hr in the presence or absence of IL-37b (50–200 ng/ml) and then stained with a TRAP staining. Notably, 200 ng/ml IL-37b almost completely suppressed RANKL-induced osteoclast formation and cells were replaced by TRAP-negative cells (Figure 2b). The number of TRAP-positive cells with three or more nuclei was reduced following the addition of IL-37b (Figure 2c). Therefore, 200 ng/ml IL-37b was used in subsequent experiments. Similarly, we also examined the effects of IL-37b on osteoclast formation from bone marrow macrophages. After 5 days in culture, 200 ng/ml IL-37b inhibited RANKL-induced differentiation of  bone marrow derived macrophages (BMMs) into osteoclasts (Figure 2d). On the basis of our data, IL-37b directly suppressed osteoclastogenesis in vitro.

To analyze the role of IL-37b in osteoclastic bone resorption, we measured bone resorption in vitro and counted the number of bone lacunae. Precursor cells were seeded on bovine bone slides and were cultured on RANKL to produce osteoclasts. We performed the pit formation assay in 96-well plates (Corning Osteo Assay Surface) and found that IL-37b reduced RANKL-mediated bone degradation. Meanwhile, RANKL-induced osteoclasts formed large, deeper, and more extensive areas of bone degradation, whereas the addition of IL-37b inhibited bone resorption (Figure 2E). Thus, IL-37b negatively regulated osteoclast formation and bone resorption in vitro. Next, we selected RAW 264.7 cells as precursor cells for potential mechanistic studies.

3.4 IL-37 inhibits osteoclast differentiation by downregulating the expression of NFATc1 in a MyD88-dependent manner

The RANKL signaling pathway strongly activates NFATc1, the principal regulator of osteoclast differentiation (Lamoureux et al., 2014). As a ligand for RANKL, RANK activates the transcription factor complex AP-1, which is composed of c-Fos and c-Jun (Asagiri & Takayanagi, 2007). In our study, we examined the effect of IL-37b on the expression of the critical osteoclastogenic transcription factor NFATc1 to determine the molecular mechanism by which IL-37b activity regulates osteoclastogenesis.

IL-37b reduced the level of the NFATc1 protein level, as analyzed by Western blot analysis (Figure 3a). The qRT-PCR analysis showed that IL-37b attenuated RANKL-induced expression of the NFATc1 mRNA during osteoclastogenesis (Figure 3b). In addition, the levels of mRNA associated with osteoclast formation were downregulated by the 24 or 72 hr IL-37b treatment (Figure 3c). IL-37b suppressed the expression of dendritic cell-specific transmembrane protein, cathepsin K, MMP9, and calcitonin receptor. These genes are all closely related to osteoclast fusion and bone degradation (Dou et al., 2014). However, 100–200 ng/ml IL-37b did not produce noticeable effects on the expression of the PU.1 mRNA compared with the untreated control group (Figure 3c). IL-1 family members act directly or indirectly through MyD88 to transduce intracellular signals. Thus, the protein plays a critical role in the IL-1 signaling pathway (Picard et al., 2003). ST 2825 is a specific inhibitor of MyD88, a small peptide compound that binds to the Toll/IL-1 receptor (TIR) domain of MyD88. This compound inhibits MyD88 dimerization and prevents it from activating and transducing IL-1 signaling, thereby blocking the IL-1 signals (Loiarro et al., 2007). We examined whether IL-37, a member of the IL-1 family, also necessarily cooperate with the MyD88 protein in regulating osteoclast formation using the MyD88 inhibitor. The images of TRAP staining indicated that 10 μg/ml ST 2825 reversed the antiosteoclastogenic effects of IL-37b (Figure 3d). Immunoblotting also showed that the inhibition of MyD88 contributed to the partial restoration of the of NFATc1 protein level (Figure 3e). Co-IP revealed an interaction between SIGIRR and MyD88 when the precursor cells were treated with RANKL and IL-37b (Figure 3f). SIGIRR is one of the important intracellular proteins involved in the IL-37 anti-inflammatory pathway (Nold-Petry et al., 2015). On the basis of these results, IL-37b negatively regulates osteoclast differentiation through the inhibition of NFATc1 expression in a MyD88-dependent manner.

3.5 IL-37 regulates osteoclast differentiation through the inhibition of NF-κB signaling

As shown in the current study, IL-37 is involved in the mechanism regulating the expression of the essential transcription factor NFATc1 during osteoclast differentiation in a MyD88-dependent manner. The activation of RANKL (RANK) in osteoclast precursor cells induces NFATc1 expression in the nucleus, an activity that is inseparable from the NF-κB signaling pathway in this process. The classical NF-κB pathway is required for osteoclastogenesis, and many target genes associated with osteoclast activation are also controlled by this signaling pathway (Boyle, Simonet, & Lacey, 2003). We investigated the activation of the canonical NF-κB pathway after RANKL stimulation and further revealed the effect of IL-37b on signal transduction during osteoclastogenesis. We observed the levels of phosphorylation IKB and p65 to avoid interference because the synthesis of the IKB protein is accompanied by an increase in the differentiation of monocytes into macrophages (Strålberg et al., 2017). Although the control group of cells stimulated with RANKL exhibited increased phosphorylation of IκBα and p65, IL-37b significantly inhibited the activation of these proteins in response to the agonist (Figure 4a). NF-κB proteins are stored in the cytoplasm of nonstimulated cells, and once cells are exposed to many stimuli, such as RANKL, the proteins are rapidly transported to the nucleus (Asagiri & Takayanagi, 2007). Therefore, we examined the subcellular localization of the NF-κB p65 subunit in RAW264.7 cells using immunofluorescence staining. The nuclear localization of p65 was markedly increased in the cells treated with RANKL, but the nuclear translocation was suppressed upon the addition of IL-37b. As shown in Figure 4b, five cells were randomly selected as the targets, and the fluorescence intensities of DAPI and p65 were measured. IL-37b inhibited the entry of p65 into the nucleus. This finding was consistent with the result showing that the activity of the NF-κB signaling pathway was increased in RANKL-treated cells and decreased in the presence of IL-37b.

3.6 IL-37 negatively regulates LPS-induced osteoclastogenesis

First, we determined the association between LPS and precursor cells in vitro. We examined the effect of LPS on the proliferation of precursor cells using the CCK-8 assay (Figure 5a). LPS did not induce the formation of TRAP-positive cells in the absence of pretreatment with RANKL, with or without M-CSF (Supporting Information Figure 1A). In contrast, the addition of 100 ng/ml LPS significantly increased the apoptosis of precursor cells. Interestingly, progenitor cells pretreated with RANKL exhibited a reduced rate of apoptosis following stimulation with LPS (Supporting Information Figure 1B). By staining focal adhesions and the actin cytoskeleton, we observed the LPS-mediated differentiation of precursor cells into osteoclasts, but both the size and number of nuclei formed were significantly less than cells stimulated with RANKL alone (Figure 5b). The results of TRAP staining also confirmed that LPS-mediated osteoclastogenesis was not separable from the RANKL pretreatment (Figure 5c). Likewise, IL-37b altered this process. Moreover, IL-37b reduced the expression of the NFATc1 mRNA in precursor cells (Figure 5d). In fact, this culture was also used in some previous studies, because the production of osteoclasts is not completely avoidable as cells are exposed to RANKL in the microenvironment in vivo (Strålberg et al., 2017). In addition, we examined the expression of proinflammatory cytokines under physiological and pathological conditions because these factors have been shown to significantly promote osteoclast formation. Low levels of proinflammatory cytokines were detected following RANKL stimulation, further indicating that IL-37b regulated the osteoclast differentiation through a mechanism that was not solely dependent on its anti-inflammatory effects (Figure 5e). Therefore, the role of IL-37b in LPS-mediated osteoclastogenesis is at least mediated by the downregulation of inflammatory factors through the direct regulatory mechanism described above (Figure 6).