The histone demethylase KDM5C aggravates titanium particle-induced osteolysis by promoting macrophage inflammation and osteoclastogenesis Authors: Zeming Lei, Liangzi Yin, Bin Zhang, Ziyuan Wang, Yansheng Wang, Lunhao Bai Published: 2026 ABSTRACT Wear debris-induced periprosthetic osteolysis (PPO) is a major complication following total joint arthroplasty, leading to implant failure. Macrophage inflammatory response and osteoclastogenesis play important roles in bone destruction and resorption in PPO. KDM5C (lysine-specific demethylase 5 C), a H3K4 demethylase, functions as a transcriptional repressor in various biological processes, including immune response, energy metabolism, and neurodevelopment. However, its role in PPO pathogenesis remains unclear. In this study, we observed elevated KDM5C expression in macrophages exposed to Titanium (Ti) particles in vitro. Genetic inhibition of KDM5C attenuated Ti particle-induced upregulation of pro-inflammatory factors and reduced the release of inflammatory cytokines. Moreover, KDM5C knockdown suppressed osteoclast differentiation, as evidenced by the decrease in osteoclast number and osteoclast-specific marker expression. Mechanistically, KDM5C was found to repress NFE2L2 transcription by decreasing H3K4 trimethylation (H3K4me3) at its promoter. Inhibition of NFE2L2 abolished the protective effects of KDM5C knockdown on macrophage inflammation and osteoclast differentiation. The calvarial osteolytic model in vivo also demonstrated that silencing KDM5C alleviated Ti particle-induced bone destruction, inflammation, and osteoclastogenesis. Collectively, these findings suggest that KDM5C promotes macrophage inflammatory response and osteoclastogenesis by repressing NFE2L2 transcription through decreasing H3K4me3 levels. It may offer a novel insight into the potential therapeutic target for wear particle-induced osteolysis.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1007/s00018-026-06309-8. Introduction Total joint arthroplasty (TJA) is a widely used surgical approach for treating joint diseases, effectively relieving patients’ pain and improving their quality of life [1]. However, wear debris may be generated from prosthetic implants and fixed Kirschner wires (K-wires) due to the friction from repetitive, including titanium (Ti), chromium, and polyethylene [2, 3]. The wear particles often result in periprosthetic osteolysis (PPO) and aseptic loosening, ultimately leading to TJA failure [4, 5]. Therefore, developing effective strategies to prevent and treat PPO is of significant clinical importance. It is well established that wear particles derived from prosthetic implants and K-wires recruit macrophages to the surrounding tissue, triggering a local inflammatory response [6, 7]. These macrophages secrete a series of proinflammatory cytokines and chemokines, which in turn directly or indirectly promote osteoclast differentiation, activation, and maturation [8, 9]. This cellular cascade leads to progressive bone destruction and PPO. Rao et al. have shown that wear debris activates macrophages towards proinflammatory phenotypes and stimulates osteoclastogenesis [10]. Inhibiting macrophage activation and osteoclast function protects against Ti-induced osteolytic bone loss [11, 12]. Thus, it is critical to investigate the mechanisms by which wear particles contribute to PPO pathogenesis. Histone modification is the key event implicated in the imbalance of bone homeostasis and macrophage inflammatory response [13, 14]. In macrophages, H3K9me2 (histone H3 lysine 9 dimethylation), catalyzed by the histone methyltransferase G9a, reduces the nuclear accumulation of NFATC1 to repress osteoclastogenesis [15]. Conversely, the histone methyltransferase ASH1L induces H3K4 trimethylation (H3K4me3) to activate STAC2 transcription, thereby inhibiting inflammation and osteoclast differentiation [16]. The histone demethylase KDM5C (lysine-specific demethylase 5 C), also known as JARID1C or SMCX, is a demethylase that negatively regulates transcription activation by removing H3K4me3 marks [17]. Previous studies have reported that KDM5C plays important roles in regulating neurodevelopment, carcinogenesis, and immune response [18–20]. Drongitis and colleagues reveal elevated IL-6 expression in glioblastoma multiforme subtypes with high KDM5C levels, suggesting a positive link between KDM5C and proinflammatory signaling [21]. In breast cancer cells, KDM5C knockdown reduced the expression of inflammatory factors, such as IL-6, IL-1 A, and CXCL2 [22]. Similarly, Ngwa et al. suggest that the conditional knockout of KDM5C decreased TNF-α levels in ischemic brains [23]. KDM5C silencing promotes a shift in microglia from a proinflammatory to an anti-inflammatory phenotype [24]. In the context of bone homeostasis, KDM5C deficiency in hematopoietic or myeloid lineages has been shown to increase bone mass and impair osteoclastogenesis in mice [25]. Nevertheless, the potential role of KDM5C in PPO remains unclear. In this work, we aimed to explore the role of KDM5C and underlying mechanisms in Ti-induced macrophage inflammatory response and osteoclast differentiation in vitro and in vivo. Materials and methods Ethics statement for animal experiments All animal experiments were performed according to the Guide for the Care and Use of Laboratory Animals. Shenyang Medical College’s ethical committee has approved the animal procedures (Approval No. SYYXY2024081301). Preparation for Ti particles Ti particles were purchased from ThermoFisher (Cat No. 000681) and prepared according to a previous study [26]. Ti particles were sterilized for 6 h at 180 °C to remove endotoxins. The endotoxin concentration was determined using a Limulus test kit (Cat No. EC32545, Bioendo). The endotoxin level less than 0.1 EU/ml was defined as endotoxin-free. Ti particles were suspended in PBS for the in vivo and in vitro experiments. Primary BMDMs isolation, Ti particle stimulation, and osteoclast differentiation Primary bone marrow-derived macrophages (BMDMs) were isolated from the femur and tibiae of C57BL/6 mice as previously described [27]. In Brief, bone marrow cells were harvested from the marrow cavity and centrifuged for 5 min at 150 g. After the cultivation in α-Minimum Essential Medium (α-MEM; Cat No. M3402, Solarbio) containing 10% FBS overnight, non-adherent cells were collected and cultured in fresh α-MEM supplemented with 10% FBS and 10 ng/ml macrophage-colony stimulating factor (M-CSF; Cat No. HY-P7085, MedChemExpress) for 72 h. Then, the adherent BMDMs were collected and identified using F4/80 (Cat No. E-AB-F0995D, Elabscience) and CD11b (Cat No. F41011b01, MultiSciences) antibody by flow cytometry (Cat No. Novocyte, Agilent). BMDMs were treated with 1.2 mg/ml Ti particles for 48 h according to the previous study [28], and subsequently cultured in osteoclastic differentiation medium containing 30 ng/ml M-CSF and 50 ng/ml RANKL (Cat No. HY-P73388, MedChemExpress) for 5 days. Lentiviral construction and infection For lentiviral packaging, HEK293T cells were transfected with the shuttle plasmids pLVX-shRNA1 (Cat No. BR004, Fenghui) with KDM5C shRNA-1, KDM5C shRNA-2, and NFE2L2 shRNA, the packaging plasmids pSPAX2 (Cat No. BR036, Fenghui), and the envelope plasmids pMD2.G (Cat No. BR037, Fenghui) using Lipofectamine 3000 (Cat No. L3000015, Invitrogen). After 72 h, cell supernatants were collected, filtered, and used as viral stocks. For lentivirus infection, BMDMs were incubated with the lentivirus for 48 h and were used for further experiments. Mouse calvarial osteolysis The murine calvarial osteolysis model using Ti particles was established according to previous studies [12, 29]. Briefly, 7-week-old male C57BL/6 mice (20 ± 2 g) were acclimated for one week and then prepared for further experiments. After anesthesia, the skin on the calvarium was shaved, sterilized, and a 0.5 cm incision was made along the midsagittal line. Then, the cranial periosteum was removed, and 30 mg Ti particle suspensions in 100 µL PBS were spread on the surface of the calvaria. The Sham mice underwent the same surgical procedures without Ti particles. The incisions were sutured layer by layer. Mice were sacrificed 2 weeks after the operation, and calvarial samples were collected for Micro-CT scanning, histological analysis, immunoblotting, and real-time quantitative PCR (RT-qPCR). For lentiviral infection, 100 µL lentiviral vectors (LV-shKDM5C/LV-shNC) at 1 × 108 TU/ml were prepared and injected into an absorbable gelatin sponge (0.5 × 0.5 × 0.5 cm3). Then, the sponge was placed on the calvarial surface for 3 weeks before the operation. Micro-CT Calvarial tissues were detected using a Micro-CT imaging system (Quantum GX, PE) with X-ray at 70 kV, 60 µA, and a field of view (FOV) of 25 mm. To reduce metal artifacts, Ti particles on the surface of calvariae were removed before scanning. The three-dimensional (3D) images of the calvariae were reconstructed using the Quantum GX software. Then, the bone mineral density (BMD) and bone volume/tissue volume (BV/TV) were calculated using the Caliper Analyze software. Enzyme-linked immunosorbent assay (ELISA) For the determination of inflammatory cytokines, commercial ELISA kits (FineTest) were used to detect the concentrations of TNF-α (Cat No. EM0183), IL-1β (Cat No. EM0109), and IL-6 (Cat No. EM0121) in cell supernatants. F-actin ring formation assay To measure the formation of the F-actin ring, cells were fixed in 4% paraformaldehyde after osteoclast induction. Then cells were incubated with rhodamine-labeled phalloidin (Cat No. CA1610, Solarbio), followed by counterstaining with DAPI. The photographs were taken by a fluorescent microscope (Cat No. BX53, Olympus). Histological and immunohistochemical staining Calvarial samples were decalcified in 10% EDTA for 3 weeks, embedded in paraffin, and sliced into 5-µm sections for H&E, TRAP, and double-immunofluorescence staining. Cells were fixed in 4% paraformaldehyde and washed in PBS for TRAP and immunofluorescence staining. For histopathological analysis, section slides were stained with hematoxylin (Cat No. H8070, Solarbio) and eosin (Cat No. A600190, Sangon) (H&E). TRAP (tartrate-resistant acid phosphatase) staining was performed to evaluate osteoclast differentiation using the Tartrate-Resistant Acid Phosphatase (TRAP) Stain Kit (Cat No. G1050, Servicebio) according to the manuscript’s protocols. Cells with more than 3 nuclei were considered to be osteoclasts. For immunofluorescence (IF) staining in vitro, cells were stained with a primary antibody against KDM5C (Cat No. Orb1539369, Biorbyt) overnight at 4 °C, and incubated with the anti-mouse IgG antibody conjugated Alexa Fluor 488 (Cat No. 4408, CST) at room temperature for 1 h. For double staining of calvarial specimens, sections were incubated with the primary antibodies against F4/80 (Cat No. 28463-1-AP, Proteintech) and KDM5C (Cat No. Orb1539369, Biorbyt) overnight at 4 °C, followed by labeling with the secondary antibodies Alexa Fluor 555-conjugated anti-rabbit IgG (Cat No. 4413, CST) and Alexa Fluor 488 anti-mouse IgG (Cat No. 4408, CST) at room temperature for 90 min. Cell nuclei were counterstained using DAPI (Cat No. D106471, Aladdin). Images were captured by a microscope (Cat No. BX53, Olympus). Chromatin immunoprecipitation (ChIP) ChIP assay was performed using the ChIP assay kit (Cat No. P2078, Beyotime) as described by the manufacturer. In short, BMDMs were incubated with 1% formaldehyde for 10 min at 37 °C, and washed in cold PBS. After resuspension in the SDS lysis buffer, cells were sonicated to produce chromatin fragments. Cell lysates were immunoprecipitated with the primary antibody against H3K4me3 (Cat No. A22146, Abclonal), KDM5C (Cat No. 14426-1-AP, Proteintech), or IgG (Cat No. A7028, Beyotime) control overnight at 4 °C, followed by incubation with Protein A + G Agarose/Salmon Sperm DNA beads. Finally, cross-links were removed, and DNA was purified and prepared for RT-qPCR. Primers used in the ChIP assay were shown as follows: ChIP-NFE2L2-F AGTACGTGTAAAGGAACCCTG, and ChIP-NFE2L2-R GCTGCCACTCCCTGATT. Real-time quantitative PCR (RT-qPCR) Total RNA from cells or calvarial samples was extracted using TRIpure lysates (Cat No. RP1001, BioTeke), quantified using the ultraviolet spectrophotometer (Cat No. NanoDrop One, ThermoFisher), and then reverse-transcribed into cDNA samples using All-in-One First-Strand SuperMix (Cat No. MD80101, Magen Biotech). RT-qPCR was performed using SYBR Green PCR reagent (Cat No. SR4110, Solarbio) and 2×Fast Taq plus PCR master mix (Cat No. BL1014, Biosharp) via the Pangaea 3 instrument (Aperbio). The relative mRNA expression was normalized to the housekeeping gene GAPDH. The following primers were utilized in this work: TNF-α-F ACAGAAAGCATGATCCGCGA; TNF-α-R TTGCTACGACGTGGGCTAC; IL-1β-F TTCCCATTAGACAACTGC; IL-1β-R GATTCTTTCCTTTGAGGC; IL-6-F TAACAGATAAGCTGGAGTC; IL-6-R TAGGTTTGCCGAGTAGA; NFATC1-F CCACTCCACCCACTTCT; NFATC1-R CTTCGCCCACTGATACG; Casthepsin K-F AGGGTCCCAGACTCCAT; Casthepsin K-R GTCAATGCCTCCGTTCT; c-Fos-F CGAAGGGAACGGAATAAG; c-Fos-R CTCTGGGAAGCCAAGGTC; ACP5-F CGGCTACTTGCGGTTTCAC; ACP5-R TCCTTGGGAGGCTGGTCTT; CEMIP-F AGCACTTCCTCCACCTT; CEMIP-R CCACATCTTGAACCCATT; NFE2L2-F ACAGTGCTCCTATGCGTGAA; NFE2L2-R TCTGGGCGGCGACTTTAT; GAPDH-F GGCAAATTCAACGGCACAGT; GAPDH-R GGGTCTCGCTCCTGGAAGAT. Immunoblotting Whole proteins were extracted using the RIPA buffer (Cat No. PR20001, Proteintech) with protease inhibitor (Cat No. PR20032, Proteintech). Protein samples were quantified by a BCA assay kit (Cat No. PK10026, Proteintech), and an equal volume of protein was separated on SDS-PAGE. Subsequently, proteins were transferred to PVDF membranes (Cat No. LC2005, ThermoFisher) and blocked in 5% skim milk for 2 h at room temperature. Protein bands were incubated with primary antibodies against KDM5C (Cat No. orb1539369, Biorbyt), iNOS (Cat No. 22226-1-AP, Proteintech), COX-2 (Cat No. 27308-1-AP, Proteintech), NFE2L2 (Cat No. 80593-1-AP, Proteintech), CEMIP (Cat No. 21129-1-AP, Proteintech), and GAPDH (Cat No. 60004-1-Ig, Proteintech) overnight at 4 °C, and then incubated with appropriate secondary goat anti-mouse antibody (Cat No. SA00001-1, Proteintech) or goat anti-rabbit antibody (Cat No. SA00001-2, Proteintech) for 1 h at room temperature. Blots were developed using the Ultrasensitive ECL detection kit (Cat No. PK10003, Proteintech), and optical density values were analyzed via the Gel-Pro-Analyzer software. Statistics All statistics were expressed by mean ± SD, and analyzed using GraphPad Prism software. For multiple comparisons, one-way ANOVA followed by Tukey’s post hoc test was used to assess the statistical significance. P values less than 0.05 were set as statistically significant. Results The histone demethylase KDM5C is highly expressed in Ti particle-induced macrophages Macrophages are one of the predominant types of immune cells that control bone homeostasis. As a consequence, the differentially expressed genes (DEGs) in control and Ti particle-treated macrophages were identified based on the GSE153558 dataset and presented in a Volcano plot (Fig. 1A). The threshold was set as |log2FC|>1 and adjp value < 0.05. Gene ontology (GO) enrichment analysis of both the upregulated and down-regulated genes in Ti-treated macrophages revealed multiple signalings related to inflammation, immunity response, and bone homeostasis, such as “regulation of inflammatory response”, “osteoclast differentiation”, and “regulation of bone resorption” (Fig. 1B). We found that the inflammatory response and mineral absorption signaling in Ti particle-treated macrophages were activated by GSEA analysis (Fig. 1C). Considering that “epigenetic regulation of gene expression” and “regulation of DNA-binding transcription factor activity” were enriched in Ti-induced macrophages (Fig. 1B), we performed an integrative analysis about all DEGs in GSE153558, transcription factors (TFs) and cofactors from Animal TFDB, together with osteoclastogenesis-related genes from GeneCards (Fig. 1D). A total of 14 overlapped genes were identified, and the expression profiles were shown in the heatmap (Fig. 1E). Among these candidate genes, the histone demethylase KDM5C was upregulated (log2FC = 2.65, adj p < 0.01) in Ti-treated macrophages, and selected for further investigation. Fig. 1The histone demethylase KDM5C is highly expressed in Ti particle-induced macrophages. (A) Volcano plot of the differentially expressed genes (DEGs) in RAW264.7 cells comparing control and Ti particle treatments from the GSE153558 dataset (n = 3). (B) Scatter plot representing the enriched gene ontology (GO) term of upregulated genes (left, blue color) and downregulated genes (right, purple color). (C) GSEA analysis for the hallmark Inflammatory Response and Mineral Absorption signatures. (D) Venn diagram showing the overlapped genes from all DEGs in GSE153558, transcriptional factors (TFs) and cofactors from the Animal TFDB website, and osteoclastogenesis-related genes from the GeneCards website. (E) Heatmap representing the expression profiles of the overlapped genes in GSE153558. (F) Experimental diagram for the isolation of macrophages from the bone marrow of mice, Ti particle stimulation, and examinations. (G) Flow cytometry analysis for the identification of isolated BMDMs. (H) Immunoblots for KDM5C protein expression in Ti particle-induced BMDMs (n = 3). (I) Immunofluorescent images of KDM5C in BMDMs post Ti particle treatment for 48 h (n = 3) The histone demethylase KDM5C is highly expressed in Ti particle-induced macrophages. (A) Volcano plot of the differentially expressed genes (DEGs) in RAW264.7 cells comparing control and Ti particle treatments from the GSE153558 dataset (n = 3). (B) Scatter plot representing the enriched gene ontology (GO) term of upregulated genes (left, blue color) and downregulated genes (right, purple color). (C) GSEA analysis for the hallmark Inflammatory Response and Mineral Absorption signatures. (D) Venn diagram showing the overlapped genes from all DEGs in GSE153558, transcriptional factors (TFs) and cofactors from the Animal TFDB website, and osteoclastogenesis-related genes from the GeneCards website. (E) Heatmap representing the expression profiles of the overlapped genes in GSE153558. (F) Experimental diagram for the isolation of macrophages from the bone marrow of mice, Ti particle stimulation, and examinations. (G) Flow cytometry analysis for the identification of isolated BMDMs. (H) Immunoblots for KDM5C protein expression in Ti particle-induced BMDMs (n = 3). (I) Immunofluorescent images of KDM5C in BMDMs post Ti particle treatment for 48 h (n = 3) To verify the role of KDM5C in macrophages upon Ti stimulation, primary mouse bone marrow-derived macrophages (BMDMs) were isolated, and the purity (> 95%) was identified by flow cytometry (Fig. 1F- G). We found that KDM5C protein levels were significantly increased by Ti particles (1.2 mg/ml) in a time-dependent manner (Fig. 1H). Furthermore, after 48 h and 72 h of Ti particle treatment, similarly upregulated levels of KDM5C were demonstrated (relative fold change from 3.57 to 3.59). Thus, BMDMs stimulated with 1.2 mg/ml for 48 h were used in further experiments. IF staining also demonstrated a dramatic upregulation of KDM5C in Ti particle-induced BMDMs (Fig. 1I). KDM5C knockdown attenuates Ti particle-induced macrophage inflammatory response The lentivirus carrying shRNAs against KDM5C was prepared to inhibit KDM5C expression. The two lentiviruses targeting KDM5C significantly reduced its protein levels (Fig. 2A). We then investigated the effects of KDM5C on Ti particle-induced macrophage inflammation using the KDM5C knockdown lentivirus (Fig. 2B). Immunoblotting analysis demonstrated that high expression levels of KDM5C in Ti-treated BMDMs were suppressed by KDM5C silencing (Fig. 2C). In the previous study, we suggested that CEMIP (Cell migration inducing hyaluronidase 1) promoted Ti-induced M1 macrophage polarization [30]. Herein, we confirmed that the high mRNA and protein levels of CEMIP induced by Ti particles were decreased following KDM5C knockdown in BMDMs (Fig. 2D). The protein levels of iNOS and COX-2 in Ti-treated BMDMs were significantly reduced by LV-shKDM5C (Fig. 2E). As well, KDM5C inhibition effectively suppressed TNF-α, IL-6, and IL-1β secretion (Fig. 2F-H). These data reveal that KDM5C may promote macrophage inflammatory response induced by wear particles. Fig. 2KDM5C knockdown attenuates Ti particle-induced macrophage inflammatory response. (A) Experimental design using lentivirus targeting NC or KDM5C infection for BMDMs. Immunoblotting analysis for KDM5C expression (n = 3). (B) Experimental diagram for macrophage inflammation. BMDMs were infected with shNC/shKDM5C lentivirus for 48 h, followed by 1.2 mg/ml Ti particles. Forty-eight hours later, whole cell lysates and supernatants were collected. (C) KDM5C protein level was determined using Immunoblotting (n = 3). (D) RT-qPCR and Immunoblotting analysis examined expression levels of CEMIP mRNA and protein. (E) Immunoblots for the expression levels of inflammatory factors (iNOS and COX-2) (n = 3). (F-H) ELISA assay for the contents of TNF-α, IL-6, and IL-1β in cell supernatants (n = 3) KDM5C knockdown attenuates Ti particle-induced macrophage inflammatory response. (A) Experimental design using lentivirus targeting NC or KDM5C infection for BMDMs. Immunoblotting analysis for KDM5C expression (n = 3). (B) Experimental diagram for macrophage inflammation. BMDMs were infected with shNC/shKDM5C lentivirus for 48 h, followed by 1.2 mg/ml Ti particles. Forty-eight hours later, whole cell lysates and supernatants were collected. (C) KDM5C protein level was determined using Immunoblotting (n = 3). (D) RT-qPCR and Immunoblotting analysis examined expression levels of CEMIP mRNA and protein. (E) Immunoblots for the expression levels of inflammatory factors (iNOS and COX-2) (n = 3). (F-H) ELISA assay for the contents of TNF-α, IL-6, and IL-1β in cell supernatants (n = 3) KDM5C knockdown suppresses Ti particle-induced osteoclast differentiation We further explored whether KDM5C is involved in the role of Ti particle-induced osteoclastogenesis (Fig. 3A). As shown in Fig. 3B-C, TRAP staining demonstrated that the increased number of TRAP-positive osteoclasts induced by Ti particles was significantly reduced following KDM5C knockdown. Phalloidin staining results suggested that KDM5C knockdown inhibited the formation of F-actin rings in Ti particle-treated cells (Fig. 3D-E). As well, the significant upregulation of osteoclast differentiation markers by Ti stimulation, including NFATC1, Cathepsin K, c-FOS, and ACP5, was decreased by KDM5C knockdown (Fig. 3F-I). These findings indicate that KDM5C is a positive regulator of Ti particle-induced osteoclast differentiation. Fig. 3KDM5C knockdown suppresses Ti particle-induced osteoclast differentiation. (A) Experimental protocols for osteoclastogenesis. BMDMs were infected with shNC/shKDM5C lentivirus for 48 h, followed by 1.2 mg/ml Ti particles. Forty-eight hours later, cells were treated with 30 ng/ml M-CSF together with 50 ng/ml RANKL for another 5 days to induce osteoclast differentiation. (B-C) Representative images showing the osteoclast differentiation by TRAP staining. Quantitative results showing the number of TRAP-positive osteoclasts (n = 3). (D-E) Representative images showing the F-actin ring formation using Phalloidin staining. The ratio of F-actin ring area was quantified (n = 3). (F-I) Expression levels of NFATC1, Cathepsin K, c-FOS, and ACP5 mRNA were examined by RT-qPCR (n = 3) KDM5C knockdown suppresses Ti particle-induced osteoclast differentiation. (A) Experimental protocols for osteoclastogenesis. BMDMs were infected with shNC/shKDM5C lentivirus for 48 h, followed by 1.2 mg/ml Ti particles. Forty-eight hours later, cells were treated with 30 ng/ml M-CSF together with 50 ng/ml RANKL for another 5 days to induce osteoclast differentiation. (B-C) Representative images showing the osteoclast differentiation by TRAP staining. Quantitative results showing the number of TRAP-positive osteoclasts (n = 3). (D-E) Representative images showing the F-actin ring formation using Phalloidin staining. The ratio of F-actin ring area was quantified (n = 3). (F-I) Expression levels of NFATC1, Cathepsin K, c-FOS, and ACP5 mRNA were examined by RT-qPCR (n = 3) KDM5C represses NFE2L2 transcription via its H3K4 demethylase activity Considering that KDM5C functions as an H3K4 demethylase, we analyzed the enrichment of H3K4me3 in BMDMs using ChIP-Atlas to explore the molecular mechanism by which KDM5C regulates macrophage inflammation and osteoclastogenesis. Integrative genomics viewer (IGV) analysis of multiple datasets using ChIP-Atlas showed high H3K4me3 levels enriched in BMDMs (Fig. 4A, upper panel). NFE2L2 (nuclear factor, erythroid 2-like 2), also named NRF2, is a famous antioxidant and anti-inflammatory factor, and has been reported to protect against wear debris-induced osteolysis [31]. We also noticed the enrichment of H3K4me3 on the NFE2L2 promoter in BMDMs by ChIP-Atlas (Fig. 4A, lower panel). Loss of KDM5C significantly restored the downregulated mRNA expression of NFE2L2 in Ti particle-treated BMDMs (Fig. 4B). Thus, the ChIP with RT-qPCR analysis was performed to determine the occupancy of H3K4me3 and KDM5C at the NFE2L2 promoter. The binding chromatin position and primer sequences for ChIP assay were presented in Fig. 4C. We demonstrated that both H3K4me3 and KDM5C were significantly enriched at the NFE2L2 promoter compared to the IgG control in Ti-treated BMDMs (Fig. 4D-E). Following KDM5C knockdown, the modification level of H3K4me3 was increased, but the enrichment of KDM5C was reduced at the promoter region of NFE2L2 (Fig. 4D-E). These data suggest that KDM5C represses NFE2L2 transcription by directly reducing H3K4me3 levels. Fig. 4KDM5C represses NFE2L2 transcription via its H3K4 demethylase activity. (A) Integrative genomics viewer (IGV) analysis showing H3K4me3 peaks overall and at the NFE2L2 promoter in BMDMs based on SRX9311730, SRX957614, SRX951934, SRX9311731, SRX012920, and SRX9311727 datasets using ChIP-Atlas. (B) RT-qPCR analysis for NFE2L2 mRNA levels in BMDMs (n = 3). (C) Diagram for the NFE2L2 gene and primers for CHIP with RT-qPCR. (D) CHIP with RT-qPCR assay showing the enrichment of H4K3me3 at the NFE2L2 promoter regions in BMDMs (n = 3). (E) CHIP with RT-qPCR assay showing the enrichment of KDM5C at the NFE2L2 promoter regions in BMDMs (n = 3) KDM5C represses NFE2L2 transcription via its H3K4 demethylase activity. (A) Integrative genomics viewer (IGV) analysis showing H3K4me3 peaks overall and at the NFE2L2 promoter in BMDMs based on SRX9311730, SRX957614, SRX951934, SRX9311731, SRX012920, and SRX9311727 datasets using ChIP-Atlas. (B) RT-qPCR analysis for NFE2L2 mRNA levels in BMDMs (n = 3). (C) Diagram for the NFE2L2 gene and primers for CHIP with RT-qPCR. (D) CHIP with RT-qPCR assay showing the enrichment of H4K3me3 at the NFE2L2 promoter regions in BMDMs (n = 3). (E) CHIP with RT-qPCR assay showing the enrichment of KDM5C at the NFE2L2 promoter regions in BMDMs (n = 3) KDM5C potentiates Ti particle-induced macrophage inflammation and osteoclastogenesis by decreasing NFE2L2 expression Then, we knocked NFE2L2 down using lentiviral vectors in BMDMs and observed a significant decrease in NFE2L2 protein (Fig. 5A). To further clarify whether NFE2L2 mediates KDM5C’s role in macrophage inflammation, BMDMs were co-infected with LV-shKDM5C and LV-shNFE2L2 and stimulated with Ti particle suspensions (Fig. 5B). Results suggested that NFE2L2 downregulation reversed the increased NFE2L2 mRNA levels, as well as the decreased CEMIP mRNA and protein expressions in Ti-treated BMDMs following KDM5C knockdown (Fig. 5C-D). The low concentrations of inflammatory cytokines in cell supernatants caused by KDM5C silencing were upregulated by knockdown of NFE2L2 (Fig. 5E-G). In addition, we explored the role of NFE2L2 in KDM5C-regulated osteoclast differentiation (Fig. 5H). As shown in Fig. 5I-J, the reduction of TRAP-positive osteoclasts by KDM5C inhibition was restored by NFE2L2 knockdown. Similarly, knockdown of NFE2L2 reversed the decreased expression of the osteoclast-specific marker NFATC1 induced by KDM5C inhibition (Fig. 5K). Altogether, these data reveal that KDM5C downregulates NFE2L2 transcription levels via its H3K4 demethylase activity to enhance Ti particle-induced macrophage inflammation and osteoclast differentiation. Fig. 5KDM5C potentiates Ti particle-induced macrophage inflammation and osteoclastogenesis by decreasing NFE2L2 expression. (A) Experimental design using shNC and shNFE2L2 lentivirus infection in BMDMs. Immunoblotting analysis for NFE2L2 protein levels (n = 3). (B) Experimental diagram for macrophage inflammation. BMDMs were infected with shNC/shKDM5C/shNFE2L2 lentivirus for 48 h, followed by 1.2 mg/ml Ti particles. Forty-eight hours later, whole cell lysates and supernatants were collected. (C) RT-qPCR was used to detect NFE2L2 mRNA levels (n = 3). (D) RT-qPCR and Immunoblotting analysis for both CEMIP mRNA and protein expressions (n = 3). (E-G) ELISA assay for the contents of TNF-α, IL-6, and IL-1β in cell supernatants (n = 3). (H) Experimental protocols for osteoclastogenesis. BMDMs were infected with shNC/shKDM5C lentivirus for 48 h, followed by 1.2 mg/ml Ti particles. Forty-eight hours later, cells were treated with 30 ng/ml M-CSF together with 50 ng/ml RANKL for another 5 days to induce osteoclast differentiation. (I-J) Representative images showing the osteoclast differentiation by TRAP staining. Quantitative results showing the number of TRAP-positive osteoclasts (n = 3). (K) The NFATC1 mRNA level was examined by RT-qPCR (n = 3) KDM5C potentiates Ti particle-induced macrophage inflammation and osteoclastogenesis by decreasing NFE2L2 expression. (A) Experimental design using shNC and shNFE2L2 lentivirus infection in BMDMs. Immunoblotting analysis for NFE2L2 protein levels (n = 3). (B) Experimental diagram for macrophage inflammation. BMDMs were infected with shNC/shKDM5C/shNFE2L2 lentivirus for 48 h, followed by 1.2 mg/ml Ti particles. Forty-eight hours later, whole cell lysates and supernatants were collected. (C) RT-qPCR was used to detect NFE2L2 mRNA levels (n = 3). (D) RT-qPCR and Immunoblotting analysis for both CEMIP mRNA and protein expressions (n = 3). (E-G) ELISA assay for the contents of TNF-α, IL-6, and IL-1β in cell supernatants (n = 3). (H) Experimental protocols for osteoclastogenesis. BMDMs were infected with shNC/shKDM5C lentivirus for 48 h, followed by 1.2 mg/ml Ti particles. Forty-eight hours later, cells were treated with 30 ng/ml M-CSF together with 50 ng/ml RANKL for another 5 days to induce osteoclast differentiation. (I-J) Representative images showing the osteoclast differentiation by TRAP staining. Quantitative results showing the number of TRAP-positive osteoclasts (n = 3). (K) The NFATC1 mRNA level was examined by RT-qPCR (n = 3) KDM5C knockdown mitigates Ti particle-induced mouse calvarial osteolysis Finally, a Ti particle-induced calvarial osteolysis mouse model was established to investigate the role of KDM5C in vivo (Fig. 6A). We confirmed an obvious high KDM5C expression in calvarial tissues with Ti particle treatment (Fig. 6B). Double-immunofluorescent staining suggested that Ti particle-induced high expression of KDM5C is mainly expressed in F4/80-labeled macrophages in calvarial tissues (Fig. 6C). Thus, loss-of-function experiments with lentivirus were performed in vivo. We confirmed that knockdown of KDM5C effectively inhibited its expression in the macrophages of calvarial tissues (Supplementary Fig. 1). The 3D reconstruction of calvariae showed evident osteolysis and a decrease in BMD and BV/TV in Ti particle-treated mice using Micro-CT scanning, which was relieved by KDM5C inhibition using the lentivirus (Fig. 6D). Histological analysis and TRAP staining suggested that KDM5C downregulation mitigated Ti particle-induced calvarial erosion and osteoclast number (Fig. 6E-F). The increased levels of osteoclast markers, including NFATC1, Cathepsin K, c-FOS, and ACP5, in calvarial tissues of Ti particle-treated mice were suppressed by KDM5C depletion (Fig. 6G-J). In addition, we found that KDM5C knockdown caused a significant reduction of TNF-α, IL-6, and IL-1β mRNA levels (Fig. 6K-M). Altogether, the in vivo results confirm that KDM5C promotes wear particle-induced osteolysis by inhibiting macrophage inflammation and osteoclastogenesis. Fig. 6KDM5C knockdown mitigates Ti particle-induced mouse calvarial osteolysis. (A) Experimental design using the Ti particle-induced mouse calvarial osteolysis model. (B) Immunoblots for KDM5C protein expression in calvarial tissues (n = 6). (C) Immunofluorescence representative images for expression and localization of KDM5C and F4/80 (n = 6). (D) Micro-CT images of calvarial tissues and bone parameters (n = 6). The bone resorption areas were pointed with red arrows. (E) Representative H&E images showing the bone damage of calvarial tissues (n = 6). Red arrows indicated the bone corrosion. (F) Representative TRAP images representing osteoclastogenesis of calvarial tissues (n = 6). Red arrows indicated the osteoclasts. (G-J) RT-qPCR assay examined NFATC1, Cathepsin K, c-FOS, and ACP5 mRNA levels in calvarial tissues (n = 6). (K-M) Expression levels of TNF-α, IL-6, and IL-1β were measured by RT-qPCR (n = 6) KDM5C knockdown mitigates Ti particle-induced mouse calvarial osteolysis. (A) Experimental design using the Ti particle-induced mouse calvarial osteolysis model. (B) Immunoblots for KDM5C protein expression in calvarial tissues (n = 6). (C) Immunofluorescence representative images for expression and localization of KDM5C and F4/80 (n = 6). (D) Micro-CT images of calvarial tissues and bone parameters (n = 6). The bone resorption areas were pointed with red arrows. (E) Representative H&E images showing the bone damage of calvarial tissues (n = 6). Red arrows indicated the bone corrosion. (F) Representative TRAP images representing osteoclastogenesis of calvarial tissues (n = 6). Red arrows indicated the osteoclasts. (G-J) RT-qPCR assay examined NFATC1, Cathepsin K, c-FOS, and ACP5 mRNA levels in calvarial tissues (n = 6). (K-M) Expression levels of TNF-α, IL-6, and IL-1β were measured by RT-qPCR (n = 6) Discussion TJA is the primary surgical intervention for end-stage bone and joint diseases. However, wear particles generated from implants often trigger PPO and subsequently TJA failure [32, 33]. Identification of the molecular mechanisms and novel therapeutic targets for PPO is urgently needed. Herein, we observed a significant increase in KDM5C expression after Ti particle treatment in calvarial tissues and macrophages. Genetic inhibition of KDM5C attenuated macrophage activation and osteoclast differentiation both in vivo and in vitro, suggesting that KDM5C might be a potential target for preventing or treating PPO. During the development of wear particle-induced PPO, multiple cell types, including immune cells, osteoclasts, osteoblasts, fibroblasts, and mesenchymal stem cells, contribute to chronic aseptic inflammation and disrupt the balance between bone resorption and formation [34, 35]. Among these, macrophages, the key immune cells, function as the first line of defense against wear particles and play a central role in regulating inflammatory response [36]. Wear particles trigger macrophage activation, subsequently causing inflammatory cascades. The inflammatory cytokines recruit more macrophages to the site of injury, promoting osteoclast activation and maturation in PPO [6, 7]. Considering the important role of macrophages in osteolysis, we made a great focus on the underlying mechanisms of macrophage inflammatory response. The DEG expression profiles induced by Ti particles in macrophages from the GSE153558 dataset, and we found that the histone demethylase KDM5C was upregulated. To our knowledge, the number of wear particles generated varies significantly among patients with PPO. Previous studies suggest that the concentration of wear particles in periprosthetic tissues ranges from 107 to 1011 particles/g [37, 38]. Based on the reported relationship that 0.1 mg/ml corresponds to 2 × 106 particles/ml [39], the mass concentration of wear particles in local periprosthetic tissues is calculated to be approximately 0.5 to 5000 mg/ml. Thus, the concentration of Ti particles used in this study fell within the clinically relevant range, indicating its physiological relevance for investigating cellular responses to PPO. It is known that the histone demethylase KDM5C functions as a transcriptional repressor by removing the enrichment of H3K4 methylation on promoters [40]. The pro-inflammatory effects of KDM5C have been widely reported among various organs and cell types. Xiao et al. demonstrate that KDM5C knockdown decreased inflammatory gene expression in breast cancer [22]. In ischemic brains and microglia, genetic inhibition of KDM5C reduces the pro-inflammatory (M1) marker expressions, whereas it increases the levels of the anti-inflammatory (M2) factors [23, 24]. Consistent with these findings, the present study showed that Ti particles induced a significant elevation of KDM5C along with macrophage infiltration in calvaria, and KDM5C knockdown inhibited Ti particle-induced macrophage inflammatory response. However, in plasmacytoid dendritic cells of diabetic wounds, the decreased expression of KDM5C is demonstrated to drive naive CD4+ T cells differentiated into the proinflammatory Th17 subtypes, indicating its anti-inflammatory property [41]. The distinct role of KDM5C in inflammatory responses may be attributable to variations in cell types and microenvironments. Furthermore, in the work, we demonstrated that KDM5C knockdown decreased CEMIP expression in macrophages. Our previous study suggested that the elevated CEMIP expression promoted macrophage inflammation osteoclastogenesis in Ti particle-induced osteolysis [30]. This indicated that KDM5C might function as an upstream factor of CEMIP in regulating macrophage inflammatory response. Osteoclasts and macrophages share a common progenitor and undergo divergent differentiation into their respective cell lineages. As the osteoclast precursors, macrophages may fuse and differentiate into multinucleated mature osteoclasts under RANKL and M-CSF stimulation [42]. In addition, macrophages secrete inflammatory cytokines to promote osteoclastogenesis, ultimately exacerbating osteolytic bone loss in PPO [11, 43]. We demonstrated the inhibitory effects of KDM5C silencing on osteoclast differentiation in vitro, which was confirmed in calvarial tissues induced by Ti particles, as evidenced by bone loss and erosion. The positive regulation of KDM5C on osteoclastogenesis was in accordance with what Liu et al. reported. NFE2L2 is a master regulator with antioxidant and anti-inflammatory characteristics [44, 45]. The lack of NFE2L2 enhances osteoclast differentiation and osteoclast activity, thereby increasing bone resorption [46, 47]. Dong et al. reveal that NFE2L2 is a protective regulator in Ti particle-induced calvarial osteolysis by inhibiting inflammatory response and osteoclastogenesis [31]. However, whether NFE2L2 is involved in the role of KDM5C and the underlying mechanisms in macrophage inflammatory response and osteoclastogenesis remains unclear. Evidence from ChIP-Atlas revealed the enrichment of H3K4me3 in macrophages and at the NFE2L2 promoter. Our results suggested that KDM5C directly repressed NFE2L2 transcription by decreasing H3K4me3 enrichment on NFE2L2 promoter regions. Loss of NFE2L2 reversed the inhibitory effects of KDM5C silencing on Ti particle-induced macrophage inflammatory response and osteoclast differentiation. Taken together, KDM5C might promote macrophage activation and osteoclastogenesis by inhibiting NFE2L2 transcription through epigenetic regulation. However, there are some limitations in the present work. First, Ti particles are widely used as the main stimulators to mimic osteolysis caused by failed TJA in cell and animal experiments [31, 48, 49]. The single concentration of Ti particles in vitro in this study was selected based on the well-established literature [28]. A further full dose-response analysis would provide more comprehensive evidence. Second, although Ti particles were chosen for the in vitro and in vivo experiments due to their stability and adherent properties, polyethylene debris is the major source of wear particles in clinical practice [50]. The comparison with polyethylene particles is logical and necessary. Future studies will be focused on the more clinically relevant polyethylene particles to strengthen translational values. Third, the calvarial osteolysis model is well-established to investigate the cellular and molecular mechanisms of wear particle-induced osteolysis. This model is commonly used to mimic an acute inflammatory response, which is different from the long bone due to the lack of load-bearing capacity and fluid pressure. It does not fully mimic the chronic, load-bearing environment of periprosthetic osteolysis after TJA. The Ti rods implanted into the animal femur might be more appropriate to mimic the clinical features of PPO, and further validations will be performed. In addition, small-molecule inhibitors of KDM5C, such as KDM5A-IN-1, are necessary to evaluate their effects on PPO in future studies, which may be more convincing for clinical trial applications. In conclusion, the present study demonstrates that KDM5C promotes macrophage activation and osteoclastogenesis to aggravate wear particle-induced osteolysis. The underlying mechanism of KDM5C is through repressing NFE2L2 transcription by removing H3K4me3 modification. Our findings suggest that KDM5C may be a potential therapeutic target for PPO, offering a basis for exploring related therapeutic strategies. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1