S100A9 Induces RANKL Expression in Osteocytes Utilizing IL‐6/sIL‐6R Combination and gp130 Signaling Pathways Authors: Eijiro Sakamoto, Tetsuya Yoshimoto, Rie Kido, Mikihito Kajiya, Hiromichi Yumoto Published: 2026 ABSTRACT ABSTRACTCalprotectin, a heterodimer of S100A8 and S100A9, is predominantly expressed in neutrophils, monocytes, and chondrocytes. Clinical studies demonstrate that elevated circulating calprotectin levels correlate with diseases followed by inflammatory bone loss, including rheumatoid arthritis, osteoarthritis, and periodontitis. We previously reported that S100A9 directly increased RANKL and IL‐6 expression in an osteocyte‐like cell line. While RANKL is well‐established as an essential factor for the formation and activation of osteoclasts, the role of S100A9‐induced IL‐6 in bone resorption remains unclear. On the other hand, osteocytes require sIL‐6R to activate the gp130 signaling pathway. Therefore, we hypothesized that S100A9‐induced IL‐6 cooperates with soluble IL‐6 receptor (sIL‐6R) to increase RANKL through gp130‐mediated signaling. Here, we show that S100A9 increases the bone resorption area and TRAP‐positive cell number in the murine calvaria. In addition, Rankl and Il6 mRNA levels in osteocytes were upregulated by S100A9 treatment. In vitro results demonstrated that S100A9 increased not only Rankl and Il6 expression in osteocytes, but also sIL‐6R secretion in macrophages. Since osteocytes strongly express gp130, the IL‐6 complexed with sIL‐6R enhanced gp130 downstream signaling, including JAK1 and STAT3 activation in osteocytes. Blockade of either Il6 or gp130 suppressed S100A‐induced effects. Notably, gp130 antagonist treatment inhibited S100A9‐induced bone resorption in vivo. Collectively, these findings suggest that S100A9 promotes RANKL‐mediated osteoclastogenesis through IL‐6/sIL‐6/gp130 signaling in osteocytes.Keywords: calprotectin, interleukin‐6, macrophage, osteocyte, receptor activator of nuclear factor‐κB ligand, S100A9, soluble interleukin‐6 receptor Introduction Calprotectin is a heteromeric complex of S100A8 and S100A9, which are calcium‐binding proteins and are predominantly expressed by neutrophils, monocytes/macrophages, and epithelial cells [1, 2]. Previous studies have demonstrated that circulating calprotectin levels are elevated in patients with various inflammatory diseases, including inflammatory bowel disease, rheumatoid arthritis (RA), and Crohn's disease [3, 4]. Calprotectin has been reported to regulate the inflammatory response by increasing the expression of pro‐inflammatory cytokines such as interleukin (IL)‐6, IL‐8, and tumor necrosis factor (TNF)‐α across multiple cell types [5, 6]. Calprotectin binds to toll‐like receptor 4 (TLR4) and receptor for advanced glycation end‐products (RAGE) and subsequently activates downstream signaling cascades, including mitogen‐activated protein kinase (MAPK), phosphoinositide 3‐kinase (PI3K), and nuclear factor kappa‐light chain enhancer of activated B cells (NF‐κB) [7]. Additionally, calprotectin increases expression of intercellular adhesion molecule (ICAM)‐1 and vascular cell adhesion protein (VCAM)‐1 [8], resulting in further amplification of the inflammatory response. According to these findings, calprotectin modulates the early stage of the inflammatory response and is known as one of the damage‐associated molecular patterns (DAMPs) [9]. Osteocytes are the most abundant cell component in bone tissue and compose 90%–95% of all bone cells in adult bone [10]. Osteocytes differentiate from mesenchymal stem cells through osteoblasts and are embedded within the mineralized tissue and bone matrix. The embedded osteocytes extend dendrites to either the bone surface or the vascular space and communicate with the other cell types [11]. Osteoblast‐osteocyte lineage cells express specific proteins during the differentiation process. Identified osteocyte markers in the early stage include phosphate‐regulating neutral endopeptidase on chromosome X (PHEX), dentin matrix protein 1 (DMP1), and E11/gp38. In the late stage, mature osteocytes regulate skeletal remodeling by expressing receptor activator of nuclear factor‐κB ligand (RANKL), which is an essential cytokine for osteoclast formation, and sclerostin, which acts as a negative regulator for osteoblasts [12]. RANKL and its soluble decoy receptor, osteoprotegerin (OPG), are widely studied in osteoclast biology, and osteocytes have been reported as a major source of RANKL [13]. Transmission of the RANKL signal from osteocytes is essential to differentiate osteoclast precursors into mature osteoclasts through RANK, a RANKL receptor, and activation of nuclear factor of activated T‐cells, cytoplasmic 1 (NFATc1) signaling [14]. Osteocyte‐derived OPG is also important for the function of RANKL in osteocytic control of osteoclast formation [15]. Therefore, osteocytes regulate the balance of RANKL and OPG expressions, resulting in bone resorption through osteoclast formation. Local and systemic inflammation profoundly affect osteocyte function and RANKL expression [16]. Multiple inflammatory cytokines, including IL‐1β and TNF‐α, directly upregulate Rankl expression while simultaneously suppressing Opg expression in osteocytes, thereby promoting osteoclastic bone resorption [17, 18, 19]. Moreover, our previous study demonstrated that S100A9, but not S100A8, directly upregulates Rankl production in osteocyte‐like cells through TLR4 and RAGE signaling pathways [20]. These findings suggest that osteocyte‐derived RANKL plays an important role in inflammatory bone destruction; however, it is still unclear how S100A9 is involved in local bone resorption. Osteocytes are also known to express pro‐inflammatory cytokines, including TNF‐α, IL‐6, IL‐1β, and IL‐8 [21]. Notably, we observed that S100A9 treatment upregulates Il6 expression in osteocytes by approximately 180‐fold, despite osteocytes expressing very low levels of Il6 under normal conditions [20]. This substantial induction suggests that S100A9‐derived IL‐6 may have autocrine or paracrine effects on osteocytes themselves. IL‐6 canonical signaling occurs through the classical IL‐6 receptor (IL‐6R) and glycoprotein 130 (gp130), which together activate intracellular signaling cascades including Janus kinase (JAK) and signal transducer and activator of transcription (STAT) pathways [22]. While osteoblasts express functional IL‐6R/gp130 complexes [23], a few studies described that osteocytes express markedly low levels of IL‐6R under basal conditions [24]. Therefore, responsiveness of osteocytes to IL‐6 depends upon trans‐signaling, a process mediated by soluble IL‐6 receptor (sIL‐6R). sIL‐6R is generated through proteolytic cleavage from the cell surface of immune and inflammatory cells, particularly lymphocytes and macrophages, and is significantly increased following calprotectin stimulation of macrophages [25]. This mechanism suggests a cell–cell cooperation model in which osteocyte‐derived IL‐6 functions in trans‐signaling via macrophage‐derived sIL‐6R. In this study, we investigated the effect of S100A9 on bone resorption through RANKL production in osteocytes and subjected them to osteocyte‐macrophage crosstalk mediating an IL‐6/sIL‐6 axis. We found that injection of S100A9 induced bone resorption and increased mRNA levels of Rankl and Il6 in a mouse model. Furthermore, S100A9 also increased sIL‐6R secretion in macrophages, and the combination of osteocyte‐derived IL‐6 and macrophage‐derived sIL‐6R induced higher expression levels of RANKL, compared to S100A9 alone. Finally, we clarified that treatment with a gp130 inhibitor blocked S100A9‐induced bone resorption in a mouse model. Our results have important findings that are involved in the novel function of S100A9 in bone metabolism and osteocyte‐macrophage crosstalk through the IL‐6/sIL‐6R axis. Materials and Methods Recombinant S100A9 Mouse recombinant S100A9 protein was obtained from USBiological Life Sciences (MA, USA). The stock solution was prepared by dissolving it in Dulbecco's phosphate‐buffered saline (DPBS, pH 7.4) and stored at −80°C until use. To confirm whether the stock solution contains only S100A9, the solution was separated on a 14% acrylamide Native‐PAGE gel at a constant 22 mA. The gel was then fixed with a methanol and acetic acid solution for 20 min. The separated proteins were visualized with Coomassie Brilliant Blue. S100A9 (14–16 kDa) and S100A9 homodimer (28–32 kDa) were observed, but not S100A8 (8 kDa) and S100A8/A9 heterodimer (22–24 kDa) (Figure S1), indicating that the S100A9 solution contains both S100A9 monomer and homodimer, but no contamination of S100A8. Mice All animal experiments were performed according to the animal protocol approved by the Nagasaki University Animal Care and Use Committee (approval No. 2310161904). C57BL/6J male mice were obtained from CLEA Japan (Shizuoka, Japan) and were acclimated for at least 1 week to the Nagasaki University animal facility. The mice were maintained on a 12‐h light–dark cycle at a constant temperature of 22°C and fed CLEA Rodent Diet CA‐1 (CLEA Japan). Daily health checks and body weight measurements were performed during the in vivo experiments. In Vivo Animal Experiments The effect of S100A9 on bone resorption was assessed using the in vivo calvarial injection model according to our previous report [26]. In a dose experiment, weight‐matched 10‐week‐old male mice were treated with control vehicle (DPBS) or 6.25, 12.5, 25, and 50 nM mouse recombinant S100A9. 5 mg/mL Escherichia coli ‐derived lipopolysaccharide ( E. coli LPS, InvivoGen, CA, USA) was used as a positive control. S100A9 and E. coli LPS were administered by calvarial injection, with 20 μL of reagents injected onto the murine calvaria using an insulin syringe with a 30‐gauge needle (NIPRO, Osaka, Japan). Injection was performed every other day for three times. On day seven, all the mice were euthanized by carbon dioxide treatment followed by cervical dislocation, and the calvariae were dissected from the mice. The dissected calvariae were fixed with 4% paraformaldehyde (PFA) for μCT and histomorphometric analysis, and snap‐frozen in liquid nitrogen for gene expression analysis. To investigate the role of the IL‐6 pathway in the effects of S100A9, sc144 (10 mg/kg, TargetMol Chemicals Inc., MA, USA), a gp130 inhibitor, was given to the mice. Weight‐matched 10‐week‐old male C57BL/6J mice were divided into 4 groups: vehicle control (DPBS/dimethyl sulfoxide, DMSO), S100A9 (50 nM S100A9/DMSO), sc144 (DPBS/sc144), and S100A9/sc144 (50 nM S100A9/sc144). Mice were pre‐treated with either sc144 or DMSO (as a vehicle control) by intraperitoneal injection (i.p. injection) for 3 days before the first calvarial injection. Calvarial injection of S100A9 and E. coli LPS was performed as described earlier, and daily i.p. injection was continued during the experimental period. On Day 10, mice were sacrificed, and calvarial tissues were harvested for μCT and histomorphometric analysis. Micro‐Computed Tomography (μCT) Analysis Analysis of the bone microarchitecture was performed by μCT. The harvested calvariae were fixed with 4% PFA with gentle shaking for 24 h. The PFA‐fixed calvariae were scanned using Skyscan 1272 (Bruker, MA, USA) with a 0.5 mm aluminum filter at 70 kV and 142 mA X‐ray source. Scanned data were reconstructed using NRecon software (Bruker) with identical reconstruction parameters: dynamic image range was 0–0.12, smoothing levels were 2, ring artifact reduction levels were 5, and beam‐hardening correction was 20%. The bone resorption area in a 4 × 4 mm square, centered around the point of intersection between the sagittal and coronal sutures, was measured using ImageJ software (version 1.5.2, National Institute of Health, USA). Bone volume (BV) and tissue volume (TV) in the same area were analyzed, and BV/TV was calculated by CTAnalyser software (Bruker). Three‐dimensional (3D) images were created by CTVox software (Bruker). Histomorphometric Analysis The fixed calvariae were decalcified with EDTA solution (0.5 M, pH 7.2) for 10 days and embedded in paraffin. The embedded calvariae were sectioned using a Leica SM2010R microtome (Leica, Wetzlar, Germany) at 5 μm slices and used for hematoxylin and eosin (H‐E) staining to investigate the microstructure of bone. To analyze osteoclast formation in the calvariae, tartrate‐resistant acid phosphatase (TRAP) staining was performed according to the described method below. Ex Vivo Macrophage Culture To investigate the interaction between osteocytes and macrophages through an S100A9 axis, an ex vivo bone marrow macrophage (BMM) culture model was used. The hindlimb bones were dissected from 10‐week‐old male C57BL/6J mice and were scraped in DPBS using a surgical scarper to remove the soft tissues, such as muscle and connective tissue. Both sides of the epiphysis were then cut, and bone marrow was flushed out using a 27‐gauge needle and syringe with DPBS. To isolate BMMs, flushed marrow was treated with ice‐cold RBC lysis buffer (eBioscience, CA, USA) to remove the red blood cells and incubated in α‐MEM supplemented with 10% fetal bovine serum (FBS) and antibiotics at 37°C and 5% CO2 for 3 h. After incubation, non‐adherent cells were harvested and cultured with 25 ng/mL murine recombinant macrophage colony‐stimulating factor (M‐CSF, R&D Systems, MN, USA) in a 100 mm dish for 2 days to differentiate them into osteoclast precursors. Osteoclast precursors were seeded on a 48‐well plate (IWAKI, Shizuoka, Japan) at 2 × 104 cells/well cell density and cultured for 24 h. To investigate the effect of S100A9 and the S100A8/A9 heterodimer on protein expression in macrophages, cells were treated with 10 nM S100A9 or 10 nM S100A8/A9 heterodimer (BioLegends CA, USA) for 24 h, and then supernatant and cell lysate were obtained for the following experiments. Ex Vivo Osteocyte Culture Primary osteocytes were isolated according to the previous report [27]. Briefly, the cleaned bones were longitudinally cut into two pieces and sequentially digested using 300 active U/mL collagenase (FUJIFILM‐Wako, Osaka, Japan) and 5 mM Ethylenediaminetetraacetic acid (EDTA, pH 7.4, FUJIFILM‐Wako). All digestion steps were performed in a CO2 incubator at 37°C for 25 min on an orbital shaker at 170 rpm and were repeated nine times. The fifth to ninth cell suspensions were combined and used as an osteocyte‐enriched cell population. A specific osteocyte marker, Sost (encoding gene of sclerostin), expression levels in the osteocyte‐enriched cell population were determined according to our previous report [26]. We confirmed that this population expressed higher levels of Sost compared to the osteoblast‐enriched cell population (Figure S2). The isolated cells were seeded on a 12‐well plate coated with Type I rat tail collagen (Corning, NY, USA) at a density of 4.5 × 103 cells/cm2 and cultured in α‐MEM supplemented with 10% FBS. In an ex vivo experiment using conditioned media, BMMs were treated with S100A9 or the S100A8/A9 heterodimer for 24 h, and the supernatant was collected from the culture. Then, osteocytes were cultured with 1% or 10% conditioned media for 24 h. To examine the direct effect of S100A9 and the S100A8/A9 heterodimer on Il6 gene expression, osteocytes were treated with S100A9, and mRNA isolation and RT‐qPCR were performed using the following method. Ex Vivo Osteocyte/Macrophage Co‐Culture System To further investigate communication between osteocytes and macrophages, we co‐cultured primary osteocytes with BMMs. First, primary osteocytes were seeded on a 48‐well plate at 2 × 103 cells/well and cultured with αMEM supplemented with 10% FBS and 1% penicillin/streptomycin in a CO2 incubator for 24 h. Then, M‐CSF‐treated BMMs were detached using 0.25% trypsin and 0.02% EDTA, and the cell suspension was seeded onto the primary osteocyte culture at 2 × 104 cells/well. Only M‐CSF‐treated BMMs were also cultured as a negative control. Two types of cells were cultured for 24 h, and then 10 nM of S100A9 was added to the culture. Culture media was replaced with fresh media supplemented with S100A9 every 48 h. On Days 5–6 after setting up the co‐culture system, osteoclast differentiation was determined by TRAP staining. TRAP Staining Osteoclast formation was assessed by TRAP staining in the in vivo and ex vivo experiments. TRAP‐positive cells were determined using a TRAP staining kit (Sigma) according to the manufacturer's protocol. In the mouse calvaria, TRAP‐positive cells were counted using a 300 × 300 μm grid in ImageJ software. In the osteocyte/macrophage co‐culture system, all of the TRAP‐positive cells with more than 3 nuclei in each well were manually counted 3 times/well, and the averages were calculated as the TRAP (+)‐MNC number. In Vitro Cell Culture MLO‐Y4, a murine osteocyte‐like cell line, was cultured according to the original report [28] and used for in vitro experiments. In brief, MLO‐Y4 cells were seeded on a 75 cm2 culture flask (Corning), coated with rat tail‐derived Type I collagen (Corning), and cultured with α‐MEM containing 5% FBS (Cytiva, Tokyo, Japan), 5% calf serum (CS, Cytiva), and antibiotics (100 U/mL penicillin and 50 μg/mL streptomycin) at 37°C and 5% CO2. Cells were maintained under 60%–70% confluency. To investigate mechanisms of RANKL expression in S100A9‐treated osteocytes, MLO‐Y4 cells were seeded on a 12‐well plate coated with rat tail Type I collagen at 3 × 104 cells/cm2 cell density. Cells were treated with 50 nM S100A9 with or without recombinant soluble IL‐6 receptor (sIL‐6R, R&D Systems) for 1 or 24 h, and then total mRNA and cell lysate were harvested. To investigate the signaling pathways utilized by S100A9‐induced IL‐6, MLO‐Y4 cells were pre‐treated with 50 nM sc144 for 1 h, and then S100A9 and sIL‐6R were given to the cells. Small Interfering RNA (siRNA) Transfection To investigate S100A9‐induced IL‐6 signaling, MLO‐Y4 cells were transfected with siRNA for IL‐6 using Lipofectamine RNAiMax transfection reagent (Thermo Fisher, MA, USA). MLO‐Y4 cells were seeded on a collagen‐coated 24‐well plate at 3 × 104 cells/cm2 cell density, and the cells were cultured with FBS‐reduced α‐MEM (2.5% FBS and 2.5% CS). When cells reached 40% confluency, the mixture of 10 nM IL‐6‐specific siRNA (siIL‐6, QIAGEN, Hilden, Germany) or scrambled control siRNA (siControl, Santa Cruz, CA, USA) with Lipofectamin RNAiMax transfection reagent was added to the culture, and the cells were incubated in a CO2 incubator for 24 h. Transfected MLO‐Y4 cells were then treated with 50 nM S100A9 and 50 nM sIL‐6R for 2 or 24 h following protein and mRNA isolation. RNA Isolation and Reverse Transcription Quantitative Polymerase Chain Reaction (RT‐qPCR) Total RNA of minced calvariae, cultured primary macrophages, osteocytes, and MLO‐Y4 cells was isolated using RNAiso plus (Takara Bio, Shiga, Japan) and purified using a Nucleo Spin RNA kit (Takara Bio) according to the manufacturer's instructions. The concentration of RNA was determined by NanoDrop Lite (Thermo Fisher), and 100–1000 ng RNA was treated with DNase I (Thermo Fisher) and used for reverse transcription by ReverTra Ace kit (TOYOBO, Osaka, Japan) according to the manufacturer's protocol. RT‐qPCR was performed using 25 ng of template cDNA with SYBR Green master mix (ThunderBird qPCR mix, TOYOBO) on a Stratagene MX3000p (Agilent Technologies, CA, USA). All the data were normalized to Actb, as the housekeeping gene, and relative gene expression levels were determined using the 2−∆∆Ct method. Protein Extraction and Western Blotting Cell lysates of MLO‐Y4 cells, primary osteocytes, and BMMs were extracted using RIPA buffer (Santa Cruz) supplemented with a protease inhibitor cocktail and phosphatase inhibitor (both are provided by Santa Cruz). Cultured cells were washed with ice‐cold DPBS three times and lysed in ice‐cold RIPA buffer. All procedures were performed on ice to avoid protein degradation. Collected protein concentration was determined using the BCA protein assay kit (FUJIFILM Wako), and 10 μg of protein was used for western blotting. Culture media from BMMs were collected into 1.5 mL microtubes and centrifuged at 10000 rpm, 4°C for 10 min to remove cellular components. The supernatant was treated with a protease inhibitor cocktail and used to detect secreted protein from BMMs. The protein samples were loaded onto a 10% acrylamide gel and separated by SDS‐PAGE at a constant 20 mA. The separated proteins were transferred to a PVDF membrane (Millipore Sigma, MA, USA) at a constant 50 V, and then the membrane was incubated with 5% BSA or skim milk in Tris‐buffered saline with 0.1% tween20 (TBS/T Buffer) for 1 h. Primary antibodies for IL‐6R (1/1000 dilution, abcam, Cat# ab128008, RRID: AB_11142611), gp130 (1:1000 dilution, BioLegend, Cat# 149401, RRID: AB_2565294), phospho‐JAK1 (1:1000 dilution, Cell Signaling Technology, MA, USA, Cat#3331, RRID: AB_2799703), JAK1 (1:1000 dilution, Cell Signaling Technology Cat# 3332, RRID: AB_2128499), phospho‐JAK2 (1:1000 dilution, abcam, Cambridge, UK, Cat# 1477–1, RRID: AB_562153), JAK2 (1:1000 dilution, abcam Cat# 2863‐1, RRID: AB_2128534), phospho‐STAT3 (1:000 dilution, BioLegend, Cat# 651003, RRID: AB_2571891), STAT3 (1:000 dilution, BioLegend Cat# 691502, RRID: AB_2632650), and β‐actin (1/2000 dilution, Cell Signaling Technology Cat# 4967, RRID: AB_330288) along with horseradish peroxidase (HRP)‐conjugated secondary antibodies (Cell Signaling Technology Cat# 7074, RRID: AB_2099233). Immunoreactive bands were visualized using Clarity Western ECL substrate (Bio‐Rad, CA, USA) and ChemiDoc Touch MP system (Bio‐Rad). In semi‐quantitative analysis, the density and intensity of each band were measured using ImageJ software (NIH). The targeted protein expression levels were normalized by β‐actin or non‐phosphorylated proteins, and then fold changes were calculated using the ratio of the control to the experimental groups. Statistical Analysis The data were shown as mean ± standard deviation (SD). The Shapiro–Wilk test was used to determine normality of the data, and then the unpaired t‐test was performed for pairwise comparisons of normally distributed data. For statistical comparison of two or more groups, one‐way or two‐way ANOVA with the Tukey post hoc test was used. p‐value < 0.05 was considered significant. All statistical analyses were performed using the GraphPad PRISM 8.1.2 software (GraphPad Software, MA, USA). Results S100A9 Injection Increases Bone Resorption via Osteoclast Formation in Mouse Calvaria We previously reported that S100A9, but not S100A8, induces RANKL expression in cultured osteocytes [20]. Therefore, we first examined whether S100A9 administration induces osteoclast formation and bone resorption on murine calvaria. μCT analysis revealed that the S100A9‐injected calvarial tissue exhibited a significant increase in osteolysis dose‐dependently compared to DPBS‐injected calvaria at day seven post injection, with bone loss comparable to that observed with E. coli LPS treatment used as a positive control (Figure 1A,B). Next, we performed histological analysis to focus on the effects of S100A9 on bone structure and osteoclast formation. Roughened bone surfaces with disrupted mineral organization were observed in the S100A9‐treated calvaria, whereas intact bone structure was shown in the control (Figure 1C). Consistent with increased bone destruction, TRAP staining revealed a marked increase in osteoclast numbers in S100A9‐injected calvaria compared to the control (Figure 1D,E). Since our previous study demonstrated that S100A9 induced RANKL expression in osteocytes [20], mRNA levels in calvarial tissue were determined by RT‐qPCR. S100A9 administration significantly upregulated Rankl mRNA expression (encoded by the Tnfsf11 gene), whereas Opg mRNA levels (Tnfrsf11b) remained largely unchanged relative to controls. Consequently, the Rankl/Opg ratio, a key parameter reflecting osteoclastogenic potential, was substantially increased in S100A9‐treated bone tissue (Figure 1F). Previously, we also reported that S100A9 induces IL‐6 expression, but not the other inflammatory cytokines, such as IL‐1β and TNF‐α, in a cultured osteocyte‐like cell line [20]. Therefore, Il6 mRNA levels were examined, and Il6 mRNA expression was also significantly upregulated in S100A9‐injected calvaria (Figure 1G). Collectively, these in vivo results demonstrate that S100A9 administration promotes bone resorption through coordinated upregulation of Rankl and Il6 expression in osteocytes, thereby facilitating osteoclast formation and bone destruction. (A) μCT images of mouse calvaria injected with control (DPBS), S100A9 (6.25, 12.5, 25, 50 nM), and E.coli LPS (5 mg/mL, as positive control) for 7 days. Injection of the S100A9 dose‐dependently increased the pit‐like structure. (B) Quantitative analysis of microstructure in mouse calvaria treated with control, S100A9, and E.coli LPS. A significant increase in the bone resorption area was observed in 12.5, 25, and 50 nM S100A9‐treated calvaria. The bone resorption area was increased, but BV/TV was decreased by S100A9 treatment. Data are presented as mean ± SD (n = 4), and p values (*p < 0.05, ***p < 0.001) are calculated by One‐way ANOVA with Tukey's multiple comparison test. (C) H‐E staining of the mouse calvaria treated with control vehicle and S100A9 (50 nM) for 7 days. Destructed bone tissue and a rough bone surface were observed in the S100A9‐treated group. (D) TRAP staining of the mouse calvaria treated with control vehicle and S100A9 (50 nM) for 7 days. Large‐sized and TRAP‐positive cells (black arrows) were observed on the surface of bone resorption in the S100A9 group. (E) Quantitative results of (D). Data are presented as mean ± SD (n = 4), and p values (***p < 0.001) are calculated by unpaired t‐test. The number of osteoclast‐like cells was increased by S100A9 treatment. (F, G) RT‐qPCR analysis for Rankl, Opg, and Il6 mRNA expression in the calvaria treated with control vehicle and S100A9 (50 nM) for 7 days. Rankl/Opg ratio was calculated using fold changes of Rankl and Opg. S100A9 increased Rankl and Il6 mRNA expression levels. Data are presented as mean ± SD (n = 4–6), and p values (*p < 0.05) are calculated by unpaired t‐test. S100A9 Increases RANKL Expression in Osteocytes Through the IL‐6/sIL‐6R Axis We have shown that local injection of S100A9 increased TRAP‐positive osteoclast‐like cells and bone resorption. However, S100A9 is heterodimerized with S100A8 under normal conditions. Therefore, we next investigated the difference in the effects of S100A9 and the S100A8/A9 heterodimer and the molecular mechanisms by which S100A9‐induced RANKL and IL‐6 affect bone resorption. Since macrophages secrete sIL‐6R in response to calprotectin [29], we hypothesized that S100A9‐treated macrophages produce sIL‐6R, which is involved in RANKL expression in osteocytes. We treated primary osteocytes with conditioned media from S100A9‐stimulated BMMs or S100A8/A9‐stimulated BMMs and assessed Rankl mRNA levels. While 1% conditioned media did not affect Rankl expression, 10% conditioned media increased Rankl mRNA by 2‐fold (Figure 2A), indicating that macrophage‐derived factors are essential for enhancing osteocytic RANKL production. In contrast, S100A8/A9‐treated conditioned media did not increase Rankl expression (Figure 2B). To identify the responsible factor, we focused on sIL‐6R, which has been previously identified as a calprotectin‐induced product of macrophages [29]. Western blot analysis confirmed that S100A8/A9 slightly increased sIL‐6R, but S100A9 treatment enhanced sIL‐6R secretion into the supernatant of BMM cultures, with levels substantially elevated compared to control cultures (Figure 2C,D). Furthermore, S100A9 treatment significantly increased Il6 mRNA in both in vitro and ex vivo cultured osteocytes, but the S100A8/A9 heterodimer had practically no effect on Il6 expression (Figure 2E). These findings suggest that the coordinated production of IL‐6 by osteocytes and sIL‐6R by macrophages in response to S100A9, but not the S100A8/A9 heterodimer, creates a favorable microenvironment for enhanced RANKL induction. (A) RT‐qPCR analysis for Rankl mRNA in primary osteocytes treated with conditioned media from S100A9‐treated macrophages. S100A9‐treated conditioned media increased Rankl mRNA in osteocytes. Data are presented as mean ± SD (n = 4), and p values (**p < 0.01) are calculated by One‐way ANOVA with Tukey's multiple comparison test. (B) RT‐qPCR analysis for Rankl mRNA in primary osteocytes treated with conditioned media from S100A8/A9 heterodimer‐treated macrophages. S100A8/A9 treated conditioned media did not significantly change Rankl expression. (C) Western blotting for sIL‐6R in conditioned media from S100A9‐ or S100A8/A9 heterodimer‐treated macrophages. S100A9 increased sIL‐6R secretion in the macrophage culture. (D) Quantitative analysis of (C). sIL‐6R levels in the supernatant were normalized by β‐Actin levels in cell lysate. Secreted sIL‐6R levels were significantly increased by S100A9, but not by the S100A8/A9 heterodimer. Data are presented as mean ± SD (n = 4), and p values (**p < 0.01) are calculated by unpaired t‐test. (E) RT‐qPCR for Il6 mRNA in S100A9‐treated MLO‐Y4 (in vitro) and primary osteocytes (ex vivo). S100A9 increased Il6 mRNA in both MLO‐Y4 and primary osteocytes. Data are presented as mean ± SD (n = 3–4), and p values (**p < 0.01, ***p < 0.001) are calculated by unpaired t‐test. (F) Western blotting for gp130 in MLO‐Y4 (in vitro) and primary osteocytes (ex vivo) without S100A9 treatment. Both MLO‐Y4 and primary osteocytes express gp130. (G) RT‐qPCR analysis for Rankl mRNA and (H) quantification of RANKL protein levels in MLO‐Y4 cells treated with S100A9, sIL‐6R, and the combination of S100A9 and sIL‐6R. S100A9 combined with sIL‐6R showed higher levels of RANKL expression. Data are presented as mean ± SD (n = 3), and p values (***p < 0.001) are calculated by Two‐way ANOVA with Tukey's multiple comparison test. It has been reported that gp130, but not IL‐6 receptor, is essential to activate IL‐6‐mediating signaling in late osteoblasts and osteocytes [30]. Consistent with this, very low levels of IL‐6R were observed in primary osteocytes compared to osteoblasts (Figure S3). To clarify how the IL‐6/gp130 axis mediates RANKL induction in osteocytes, we examined gp130 protein expression in osteocytes in vitro and ex vivo culture systems by western blotting. Similar to primary osteocytes, MLO‐Y4 cells, a mouse osteocyte‐like cell line, express gp130 (Figure 2F). Therefore, to investigate whether S100A9‐induced IL‐6 and sIL‐6R promote RANKL expression in osteocytes, MLO‐Y4 cells were treated with S100A9 and recombinant sIL‐6R. S100A9 alone significantly increased RANKL production by 2‐fold, whereas sIL‐6R alone did not alter RANKL mRNA or protein levels. Interestingly, when cells were treated with both S100A9 and sIL‐6R, RANKL expression was higher than in the S100A9‐only group at both the mRNA and protein levels (Figure 2G,H). Collectively, these results suggest that S100A9‐induced IL‐6 in osteocytes synergizes with macrophage‐derived sIL‐6R to activate gp130‐mediated signaling, thereby enhancing RANKL expression in osteocytes. This mechanism elucidates the crucial role of osteocyte‐macrophage crosstalk in inflammatory bone resorption. S100A9‐Induced IL‐6 Activates JAK1/STAT3 Signaling in Osteocytes Gp130 serves as the common signal‐transducing receptor for the IL‐6 cytokine family, with downstream signaling predominantly mediated through the JAK/STAT cascade [31]. Therefore, we sought to investigate whether S100A9‐induced IL‐6, combined with sIL‐6R, activates JAK/STAT signaling, and compare this reaction with the direct effects of S100A9‐independent IL‐6 and sIL‐6R in osteocytes. MLO‐Y4 cells were treated with S100A9 and recombinant sIL‐6R, and cells were harvested after treatment at 0.5, 1, and 2 h. S100A9 and sIL‐6R treatment enhanced JAK1 phosphorylation at 1 h and persisted through 2 h post‐treatment. However, JAK2 phosphorylation was not observed at any time point, indicating selective activation of the JAK1 isoform. Since we have demonstrated that S100A9 upregulates RANKL expression in osteocytes through STAT3 activation [20], we next focus on phosphorylation in STAT3. STAT3 phosphorylation was detected 2 h after S100A9 plus sIL‐6R treatment (Figure 3A,B). Western blotting for downstream molecules of gp130 in MLO‐Y4 cells treated with either the combination of S100A9 and sIL‐6R or the combination of IL‐6 and sIL‐6R for 0.5, 1, and 2 h. (A) Western blotting for p‐JAK1, JAK1, p‐JAK2, JAK2, p‐STAT3, and STAT3 in S100A9 and sIL‐6R‐treated MLO‐Y4 cells. (B) Quantitative analysis of (A). S100A9 and sIL‐6R significantly increased phosphorylation levels of JAK1 and STAT3, but not JAK2. Data are presented as mean ± SD (n = 3), and p values (ns: No significance, *p < 0.05, ***p < 0.001 vs. 0 h) are calculated by One‐way ANOVA with Dunnett's multiple comparison test. (C) Western blotting for p‐JAK1, JAK1, p‐JAK2, JAK2, p‐STAT3, and STAT3 in IL‐6 and sIL‐6R‐treated MLO‐Y4 cells. (D) Quantitative analysis of (C). Treatment with IL‐6R and sIL‐6R demonstrated similar effects on the phosphorylation of JAK1 and STAT3, but at distinct time points. Data are presented as mean ± SD (n = 3), and p values (ns: Not significant, *p < 0.05, **p < 0.01 vs. 0 h) are calculated by One‐way ANOVA with Dunnett's multiple comparison test. Both treatments enhanced phosphorylation of JAK1 and STAT3, but at different time points. On the other hand, treatment with IL‐6 and sIL‐6R enhanced JAK1 phosphorylation within 0.5 h and sustained through 2 h, and no phosphorylation of JAK2 was observed. In addition. direct IL‐6/sIL‐6R treatment induced STAT3 phosphorylation more rapidly, within 0.5 h (Figure 3C,D). These temporal differences suggest that S100A9‐induced IL‐6 achieves functionally equivalent JAK1/STAT3 activation as direct IL‐6/sIL‐6R stimulation, albeit with slightly delayed kinetics. These findings demonstrate that S100A9‐induced IL‐6, in combination with macrophage‐derived sIL‐6R, activates canonical gp130/JAK1/STAT3 signaling in osteocytes. S100A9 Utilizes IL‐6 and gp130‐Dependent Signaling to Further Increase RANKL To establish the functional necessity of S100A9‐induced IL‐6 in promoting RANKL expression, we performed IL‐6 loss‐of‐function studies using siRNA. Transfection of MLO‐Y4 cells with siIL‐6 significantly suppressed Il6 mRNA expression under S100A9 and sIL‐6R treatment conditions compared to siControl (Figure 4A). In addition, IL‐6 knockdown substantially attenuated phosphorylation of both JAK1 and STAT3 in S100A9 and sIL‐6R‐treated cells, demonstrating that IL‐6 is essential for activation of this signaling cascade (Figure 4B,C). Since activated JAK/STAT signaling is known to induce RANKL expression in MLO‐Y4 [24], we next examined Rankl mRNA levels in siIL‐6‐transfected MLO‐Y4. Rankl expression was significantly decreased in siIL‐6‐transfected cells under S100A9 and sIL‐6R treatment compared to siControl‐transfected cells (Figure 4D). To determine whether the IL‐6/sIL‐6R mediated enhancement of RANKL expression proceeds through gp130 signaling, MLO‐Y4 cells were pre‐treated with sc144, a selective gp130 antagonist, followed by treatment with S100A9 and sIL‐6R. Pre‐treatment with sc144 effectively blocked S100A9‐ and sIL‐6R‐induced Rankl expression in osteocytes (Figure 4E). These results suggest that S100A9‐induced IL‐6 plays an essential role in amplifying RANKL expression in osteocytes through gp130/JAK1/STAT3‐dependent signaling. (A) RT‐qPCR for Il6 mRNA in MLO‐Y4 cells transfected with either siControl or siIL‐6. Transfection with siIL‐6 blocked S100A9 and sIL‐6R‐induced upregulation of Il6 mRNA. Data are presented as mean ± SD (n = 3), and p values (**p < 0.01, ***p < 0.001) are calculated by One‐way ANOVA with Tukey's multiple comparison test. (B) Western blotting for downstream molecules of gp130. NT means “No treatment”. (C) Quantitative analysis of (B). Il6 knockdown blocked S100A9 and sIL‐6R‐induced phosphorylation of JAK1 and STAT3. Data are presented as mean ± SD (n = 3), and p values (*p < 0.05, **p < 0.01, ***p < 0.001) are calculated by One‐way ANOVA with Tukey's multiple comparison test. NT means “No treatment”. (D and E) RT‐qPCR for Rankl mRNA in (D) siIL‐6‐transfected MLO‐Y4 and (E) sc144‐treated MLO‐Y4 cells. S100A9‐ and sIL‐6R‐induced Rankl mRNA levels were blocked by siIL‐6 knockdown and sc144 treatment. Data are presented as mean ± SD (n = 4), and p values (*p < 0.05, ***p < 0.001) are calculated by One‐way ANOVA (C) or Two‐way ANOVA (D) with Tukey's multiple comparison test. S100A9 Upregulates Osteoclastogenesis in the Macrophage/Osteocyte Co‐Culture System To evaluate the functional significance of S100A9‐induced IL‐6 and sIL‐6R signaling in promoting bone resorption, we examined osteoclast differentiation in an integrated macrophage/osteocyte co‐culture system. First, only macrophages without osteocytes were treated with 1, 10, and 20 nM S100A9 to confirm whether S100A9 directly regulates osteoclast formation. All concentrations of S100A9 did not affect osteoclast formation. Next, macrophages were co‐cultured with osteocytes and were treated with S100A9. In the vehicle control group (DPBS), only a few osteoclast precursors, but no mature osteoclasts, were observed. In contrast, osteoclast‐like cells appeared in the S100A9‐treated groups. At 1 nM S100A9, fusion of TRAP‐positive cells was observed, indicating early osteoclast differentiation. Treatment with 10 nM S100A9 markedly increased the number of mature, multinucleated osteoclasts characterized by 3–4 nuclei and prominent actin rings. However, further dose escalation to 20 nM S100A9 resulted in reduced osteoclast formation compared to the 10 nM dose, suggesting a biphasic dose–response relationship (Figure 5A). Quantification of TRAP‐positive multinucleated cells confirmed these observations, with both 10 and 20 nM S100A9 significantly increasing osteoclast‐like cell numbers relative to control, although the 20 nM group exhibited significantly fewer osteoclasts compared to the 10 nM group. In contrast, TRAP‐positive cells were not detected in the macrophage monoculture treated with S100A9 (Figure 5B). These in vitro findings suggest that S100A9 promotes osteoclastogenesis in the macrophage/osteoclast co‐culture system, likely through IL‐6 and sIL‐6R‐mediated enhancement of RANKL signaling. (A) TRAP staining images of the macrophage monoculture and osteocyte/macrophage co‐culture system treated with S100A9, and (B) counting data of TRAP‐positive multinuclear cell number. S100A9 did not induce osteoclastogenesis in the macrophage culture without osteocytes, but 10 and 20 nM S100A9 increased osteoclast‐like cells in the osteocyte/macrophage co‐culture. The scales indicate 100 μm. Data are presented as mean ± SD (n = 3), and p values (***p < 0.001) are calculated by one‐way ANOVA with Tukey's multiple comparison test. Blocking gp130 Protects Bone Against S100A9‐Induced Bone Resorption To determine whether pharmacological inhibition of gp130 could prevent S100A9‐induced bone resorption in vivo, mice were pre‐treated with 10 mg/kg sc144 or vehicle control (DMSO), followed by injection of S100A9 on the calvaria. In μCT‐derived 3D images, sc144 treatment alone did not affect bone formation and resorption. In contrast, S100A9 markedly increased bone resorption pits, while sc144 pretreatment significantly reduced S100A9‐induced bone loss (Figure 6A). Quantitative morphometric analysis confirmed that sc144 significantly decreased the resorbed bone surface area in S100A9‐injected calvaria. Similarly, S100A9‐induced reduction in BV/TV was partially restored by sc144 administration (Figure 6B). (A) μCT images of calvaria injected with vehicle control (DPBS) or S100A9 in sc144 pre‐treated mice. Sc144 treatment decreased S100A9‐induced bone resorption. (B) Quantitative analysis of microstructure in sc144 and S100A9‐treated calvaria. Sc144 blocked the S100A9‐induced increase in bone resorption area and decrease in BV/TV. Data are presented as mean ± SD (n = 4), and p values (**p < 0.01, ***p < 0.001) are calculated by Two‐way ANOVA with Tukey's multiple comparison test. (C) H‐E staining of calvaria injected with control (DPBS) or S100A9 in sc144 pre‐treated mice. S100A9‐induced bone tissue destruction was decreased by pre‐treatment with sc144. (D) TRAP staining of calvaria injected with control (DPBS) or S100A9 in sc144 pre‐treated mice. S100A9‐induced appearance of TRAP‐positive cells (black arrows) was decreased by pre‐treatment with sc144. (E) Quantitative results of (D). The S100A9‐induced increase in TRAP‐positive cell number was blocked by pre‐treatment with sc144. Data are presented as mean ± SD (n = 4), and p values (ns: No significant, **p < 0.01, ***p < 0.001) are calculated by Two‐way ANOVA with Tukey's multiple comparison test. (F, G) RT‐qPCR analysis of Rankl, Opg, and Il6 mRNA expression in calvaria treated with S100A9 and sc144. S100A9‐induced upregulation of Rankl and Il6 mRNA was blocked by sc144. Data are presented as mean ± SD (n = 5), and p values (ns: No significant, *p < 0.05, **p < 0.01, ***p < 0.001) are calculated by Two‐way ANOVA with Tukey's multiple comparison test. Histological analysis further characterized the protective effects of gp130 antagonism. Treatment of sc144 alone did not alter bone morphology, but the bone surface in the sc144/S100A9 group appeared smoother compared with the S100A9‐only group (Figure 6C). The number of TRAP‐positive osteoclasts in the sc144 alone group was comparable to that in the vehicle group, whereas S100A9‐induced osteoclast formation was clearly suppressed by sc144 administration (Figure 6D,E), demonstrating that gp130 signaling is required for S100A9‐mediated osteoclast formation. To investigate the molecular mechanisms underlying gp130 antagonist‐mediated bone protection, we analyzed gene expression in calvarial tissue by RT‐qPCR. S100A9‐induced upregulation of Rankl was significantly decreased by sc144, consistent with our in vitro results. Opg expression remained unchanged, and the Rankl/Opg ratio showed a similar pattern to Rankl (Figure 6F). Interestingly, sc144 did not affect Il6 levels, although S100A9 upregulated Il6 by 23‐fold (Figure 6G), indicating that sc144 blocked RANKL expression without changing Il6 expression. These results suggest that gp130 signaling plays a crucial role in S100A9‐induced bone resorption and that blocking gp130 by sc144 effectively protects bone from S100A9‐mediated osteolytic activity. Discussion Calprotectin, a hetero‐complex of S100A8 and S100A9, is a key molecule in inducing an immune response in the early stage of inflammation. Although these S100 proteins normally form heterodimers, other forms, including monomers, trimers, and tetramers, can be detected in the circulation [32]. The individual levels of S100A8, A9, and S100A8/A9 heterodimer in the circulation have been reported by several studies. In the human study, S100A9 and S100A8/A9 levels were slightly higher than S100A8, and these levels could be different between males and females [33, 34]. A previous study showed higher secretion of IL‐1β, TNF‐α, and IL‐6 in S100A9‐treated human fibroblast‐like cells than in S100A8‐treated cells [35]. Similarly, another study reported that S100A8 upregulated IL‐6 expression approximately twice as strongly as S100A9 in the human keratinocyte‐like cell culture [36]. Furthermore, we previously discovered that S100A9, but not S100A8, significantly increased RANKL and IL‐6 expression in osteocyte‐like cells [20]. These findings suggest that S100A8 and S100A9 have independent roles in the inflammatory response. In this study, we obtained commercial S100A9, which contains S100A9 monomer and homodimer. Although S100A9 monomer exists under normal conditions, IL‐1β and bacterial lipopolysaccharide promote S100A9 homodimerization [37]. Therefore, our results demonstrated the role of S100A9‐monomer/homodimer in inflammatory bone resorption and whether blocking S100A9‐induced IL‐6 signaling improves bone resorption. We have shown that injection of recombinant S100A9 protein into the mouse calvaria increased bone resorption in a dose‐dependent manner, and an increase in TRAP‐positive cell number was observed in the bone resorption area. Importantly, when we sought mRNA expression levels in calvaria injected with S100A9, not only the Rankl/Opg ratio, but also Il6 mRNA levels were significantly increased. These results are consistent with our previous observation from MLO‐Y4‐A2 cells in vitro [20], but the role of inflammatory cytokines produced by osteocytes is less clear. Osteocytes require sIL‐6R to activate IL‐6 signaling, resulting in upregulation of RANKL expression [24]. It has also been reported that IL‐6 in combination with sIL‐6R increased the number of TRAP‐positive cells and expression levels of nuclear factor of activated T‐cells, cytoplasmic 1 (NFATc1), and c‐Fos in BMM culture [38, 39]. These findings suggest that S100A9‐induced IL‐6 did not directly regulate RANKL expression in osteocytes and osteoclast formation, and sIL‐6R was involved in an increase in TRAP‐positive cell number. S100A8 is expressed by neutrophils and keratocytes, and injected S100A9 might form an S100A8/A9 heterodimer in vivo. Therefore, we next investigate the specific effects of S100A9 on RANKL expression through the IL‐6/sIL‐6R axis using an in vitro culture system. Calprotectin has been reported to increase sIL‐6R but does not affect IL‐6 secretion levels in human macrophage‐like THP‐1 cells [29]. To examine whether S100A9 and the S100A8/A9 heterodimer upregulate sIL‐6R in macrophages and its effects on Rankl mRNA expression in osteocytes, primary osteocytes were cultured with conditioned media harvested from BMMs. Adding S100A9‐treated conditioned media significantly increased Rankl expression in osteocytes, and a dramatic increase in sIL‐6R was observed in the supernatant of S100A9‐treated macrophages. Although the S100A8/A9 heterodimer slightly increased sIL‐6R in macrophages, as in the previous study, conditioned media did not increase Rankl expression in osteocytes. These results strongly suggest that the effect of S100A9 monomer/homodimer on sIL‐6R expression is stronger than that of the S100A8/A9 heterodimer in macrophages, and S100A9‐induced sIL‐6R in macrophages has an important role in the regulation of Rankl mRNA in osteocytes. In fact, cellular membrane‐bound IL‐6R expression was much lower in osteocytes than in osteoblasts, but high levels of gp130 expression were observed in osteocytes. Moreover, Il6 mRNA levels in both primary osteocytes and MLO‐Y4 cells were upregulated by S100A9, but not by S100A8/A9 heterodimer treatment, indicating that S100A9 can induce IL‐6 expression in osteocytes, regardless of primary cells and immortalized cell lines. Further results showed that S100A9 and sIL‐6R treatment increased Rankl mRNA, whose levels were greater than those of S100A9 treatment alone, indicating that S100A9‐induced IL‐6 cooperated with sIL‐6R to regulate Rankl mRNA in osteocytes. Although osteocytes exist deep within the bone matrix, they communicate with the other osteocytes and cell types through the bone space and narrow channels, called the lacunar‐canalicular system [40]. Since the diameter of canaliculi is approximately 250–300 nm and is connected to the blood vessels running through bone tissue, osteocytes can release and receive soluble proteins [41, 42]. This unique structure of bone may explain the crosstalk between osteocytes and macrophages through IL‐6 and sIL‐6R. Compared to the treatment of IL‐6 combined with sIL‐6R, S100A9 and sIL‐6 treatment showed very similar activation of the intracellular signaling. IL‐6 and sIL‐6R obviously enhanced phosphorylation of JAK1 and STAT3, but not JAK2, 3 h after treatment. Although the same molecules were phosphorylated in the S100A9 and sIL‐6R‐treated group, phosphorylation of JAK1 and STAT3 was observed at the later time point. This difference in the timing of phosphorylation suggests that S100A9 induces IL‐6 secretion, and then the combined S100A9‐induced IL‐6 with sIL‐6R phosphorylates JAK1/STAT3 signaling in osteocytes. Previous studies in osteocytes reported that the combination of IL‐6 and sIL‐6R enhanced phosphorylation of JAK2/STAT3 at a 24‐h treatment [24], but 30‐min treatment was also shown to phosphorylate STAT3 [43]. Since we demonstrated that downregulation of Il6 levels by siRNA blocked S100A9 and sIL‐6R‐induced JAK1 and STAT3 phosphorylation, STAT3 has an important role in IL‐6 signaling in osteocytes. Moreover, inhibition of gp130 by sc144 blocked an increase in Rankl mRNA in S100A9 and sIL‐6R‐treated MLO‐Y4 cells. These findings suggest that S100A9‐induced IL‐6, combined with sIL‐6R, utilizes gp130/JAK/STAT3 signaling to further increase Rankl expression in osteocytes; however, the time‐dependent effects on JAK1 or JAK2 phosphorylation are not yet clear. Osteocytes have been well known to regulate osteoclast formation through expression of RANKL, OPG, and macrophage colony‐stimulating factor (M‐CSF) [44, 45]. To investigate the direct crosstalk between macrophages and osteocytes, we co‐cultured two types of cells and treated them with 1, 10, and 20 nM S100A9. The number of TRAP‐positive and multinuclear cells was significantly increased in the 10 nM group, suggesting that S100A9 positively regulates osteoclast formation due to expression of RANKL in osteocytes. In contrast, mature osteoclasts were not observed in the S100A9‐treated macrophage culture without osteocytes, indicating that communication with osteocytes is essential for osteoclastogenesis. Since we have shown that S100A9 increases IL‐6 secretion in osteocytes and sIL‐6R in macrophages, an IL‐6/sIL‐6R/gp130 axis can explain this increase in osteoclast‐like cells. In addition, the combination of IL‐6 and sIL‐6R upregulates M‐CSF expression in the bone cells [46], indicating that S100A9 may increase not only RANKL but also M‐CSF expression to regulate osteoclast formation through an IL‐6/sIL‐6R/gp130 signaling. We finally demonstrated that pre‐treatment with sc144, an antagonist for gp130, blocked S100A9‐induced bone resorption area, osteoclast formation, and Rankl expression in osteocytes. sc144 was reported to bind specifically to gp130 and has been used to investigate the gp130‐dependent signaling and its effects [47]. The recent study demonstrated that treatment with sc144 blocked IL‐6‐induced phosphorylation of STAT3 and skin inflammation [48]. Our results, in combination with these reports, suggest that pre‐treatment with sc144 blocked RANKL‐induced osteoclast formation due to the inhibition of the gp130‐dependent signaling pathways. The most striking in vivo result is that treatment with sc144 did not affect S100A9‐induced Il6 mRNA, indicating that this decrease in osteoclast number on calvaria was not due to secreted IL‐6 by osteocytes. On the other hand, the combination of IL‐6 and sIL‐6R directly promotes osteoclast formation without a RANKL‐dependent mechanism, and the osteoclast precursor‐specific gp130‐deleted mice exhibit narrower marrow space in femora [49, 50]. Since sc144 also antagonizes gp130 expressed on macrophages [51], treatment with sc144 may inhibit not only the RANKL‐dependent mechanism but also the direct effect of S100A9‐induced IL‐6 on osteoclast formation. While we demonstrated a novel function of S100A9 on inflammatory bone resorption through the regulation of IL‐6/sIL‐6 expression, there are limitations to the study. Although we have shown the roles of S100A9 monomer/homodimer in the early stage of inflammatory bone resorption in this study, S100A9 is heterodimerized with S100A8 (S100A8/A9 heterodimer, calprotectin) under normal conditions. In fact, inflammation‐associated factors including bacterial lipopolysaccharide and IL‐1β are reported to increase not only the S100A8/A9 heterodimer but also the S100A9 homodimer [37]. Further investigations using inflammatory bone resorption disease model animals, such as RA and periodontitis, are required to elucidate the distribution of the S100A8/A9 heterodimer and the S100A9 homodimer. The second limitation is the lack of a specific function of gp130 on S100A9‐induced bone resorption. An antagonist for gp130, sc144, was injected to investigate IL‐6/sIL‐6 signaling in this study, but administration of sc144 may have unclear side effects. These side effects may directly inhibit sIL‐6R expression in macrophages and osteoclast formation, resulting in osteoclastogenesis being blocked by sc144 in the mouse bone and co‐culture system. Further study will investigate whether S100A9‐induced bone resorption and osteoclast formation are still or not observed in the osteocyte‐specific gp130 deleted mice. In summary, we demonstrated that a component of calprotectin, S100A9, upregulated secretion of IL‐6 in osteocytes and sIL‐6R in macrophages, and this combination induced osteoclast formation through gp130/JAK1/STAT3 signaling and RANKL production in osteocytes. S100A9 may be a novel therapeutic target protein to prevent and improve the local inflammatory bone destruction due to RA, OA, and periodontitis. Author Contributions E.S.: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, writing – review and editing, writing – original draft. T.Y.: data curation, formal analysis, investigation, methodology, writing – review and editing, writing – original draft. R.K.: formal analysis, investigation, writing – review and editing. M.K.: methodology, supervision, writing – review and editing. H.Y.: resources, supervision, writing – review and editing. Funding This work was supported by MEXT | JSPS | Japan Society for the Promotion of Science (JSPS), 19K19000. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Figure S1: Native‐PAGE assay of recombinant S100A9 solution dissolved in D‐PBS. Both monomer and homodimer were observed in S100A9 solution. Figure S2: RT‐qPCR analysis of Sost in osteoblast‐enriched cell population and osteocyte‐enriched cell population obtained by sequential digestion. The later cell suspension highly expressed Sost mRNA. Data are presented as mean ± SD (n = 5 and 7), and p value (**p < 0.01) was calculated by unpaired t‐test. Figure S3: Western blotting for IL‐6R in murine primary osteoblasts and osteocytes. Expression levels of IL‐6R in osteocytes were lower than in osteoblasts. Acknowledgments Financial support was provided by the Japan Society for the Promotion of Science (No. 19 K19000 to ES). We appreciate Prof. Toshihisa Komori (Nagasaki University, Department of molecular bone biology) for providing mouse osteocyte‐like MLO‐Y4. Data Availability Statement The data that support the findings of this study are openly available in DRYAD at DOI: 10.5061/dryad.nzs7h456p. References Associated Data Supplementary Materials Figure S1: Native‐PAGE assay of recombinant S100A9 solution dissolved in D‐PBS. Both monomer and homodimer were observed in S100A9 solution. Figure S2: RT‐qPCR analysis of Sost in osteoblast‐enriched cell population and osteocyte‐enriched cell population obtained by sequential digestion. The later cell suspension highly expressed Sost mRNA. Data are presented as mean ± SD (n = 5 and 7), and p value (**p < 0.01) was calculated by unpaired t‐test. Figure S3: Western blotting for IL‐6R in murine primary osteoblasts and osteocytes. Expression levels of IL‐6R in osteocytes were lower than in osteoblasts. Data Availability Statement The data that support the findings of this study are openly available in DRYAD at DOI: 10.5061/dryad.nzs7h456p.