YTHDF1 transcriptionally activated by TCF4 suppresses osteoblast ferroptosis in titanium nanoparticle-induced osteolysis by accelerating GPX4 and SLC7A11 translation Authors: Da Zhong, Wenqing Xie, Zhan Liao, Hua Liu, Lemei Zhu, Fawei Gao, Xi Li, Zhen Yin, Ganzhe Xu, Peng Chen, Jian Tian, Long Wang, Ke Yin, Kunli Chen, Chenggong Wang Published: 2025 ABSTRACT AbstractBackgroundAseptic loosening and periprosthetic osteolysis driven by wear particles are major causes of total joint arthroplasty (TJA) failure. Investigating the upstream mechanisms of osteoblast ferroptosis in periprosthetic osteolysis is essential for developing therapeutic strategies to lengthen the lifespan of artificial joints.MethodTitanium nanoparticle (TN)-induced osteolysis models were established in vivo and in vitro. Bone parameters including bone mineral density, bone volume, bone volume to tissue volume ratio, and total porosity were measured by micro-CT in mice. Commercial kits were applied to test the levels of glutathione, malondialdehyde, and iron. BODIPY 581/591 C11 staining was used to detect lipid peroxidation. ChIP and dual-luciferase reporter assay were utilized to investigate the interaction among TCF4 protein and YTHDF1 promoter region. RIP was employed to detect the relationship between YTHDF1 and GPX4 or SLC7A11 mRNAs.ResultsOur findings revealed a marked enhancement of ferroptosis in clinical synovial tissues surrounding implants and TN-induced periprosthetic osteolysis models both in vitro and in vivo. Ferroptosis inhibitors significantly attenuated TN-induced osteolysis. TCF4 and YTHDF1 were both downregulated in TN-induced osteolysis, and overexpression of them improved bone microarchitecture and alleviated osteoblast ferroptosis. TCF4 bound to the promoter region of YTHDF1 to transcriptionally upregulate its expression, which subsequently promoted translation of GPX4 and SLC7A11 mRNAs. Finally, the activation of Wnt/β-catenin pathway boosted TCF4/YTHDF1 axis, and restrained osteoblast ferroptosis in osteolysis.ConclusionTCF4 targeted the promoter region of m6A reader YTHDF1 to promote its transcription, thereby reducing osteoblast ferroptosis in TN-induced periprosthetic osteolysis by promoting the translation of GPX4 and SLC7A11.Graphical Abstract Supplementary InformationThe online version contains supplementary material available at 10.1186/s12951-025-03861-6.Keywords: Periprosthetic osteolysis, Ferroptosis, TCF4, YTHDF1, M6A reader, Titanium nanoparticles Introduction Total joint arthroplasty (TJA) is considered one of the most successful surgical interventions for the management of advanced-stage joint diseases, including osteoarthritis and rheumatoid arthritis, with more than 95% of prostheses remaining functional at ten years postoperatively [1]. However, despite advancements in surgical techniques, implantation designs, and biomaterials, aseptic loosening of artificial joints due to periprosthetic osteolysis is the major complication that severely impacts the long-term durability of implants [2]. One significant contributor to periprosthetic osteolysis is the large number of wear particles, such as titanium, ceramic, polyethylene, and cobalt chromium, generated by the joint prostheses [3]. These particles trigger a series of biological reactions in surrounding tissues, including affecting survival and differentiation of osteoblasts and inhibiting bone formation, ultimately leading to osteolysis [4]. Therefore, elucidating the mechanisms by which wear particles induce osteolysis may pave the way for the development of novel therapeutic strategies. Ferroptosis, a form of programmed cell death defined by iron-dependent lipid peroxidation and pathological accumulation of reactive oxygen species (ROS), is observed in many orthopedic conditions, such as bone defect and osteoporosis [5]. However, studies reporting osteoblast ferroptosis in titanium nanoparticle (TN)-induced osteolysis are rare, leaving it unclear whether inhibiting osteoblast ferroptosis can alleviate periprosthetic osteolysis. N6-methyladenosine (m6A) is the most prevalent RNA modification in eukaryotic messenger RNA (mRNA), influencing nearly all stages of RNA lifecycle, including transcription, translation, and degradation [6]. Previous studies have demonstrated that m6A modification can regulate bone formation, resorption, and remodeling as well as ferroptosis in different types of cells [7, 8]. However, it remains unknown whether abnormal m6A modification exists in osteolysis and whether m6A modification can regulate osteoblast ferroptosis. YTH N6-methyladenosine RNA binding protein 1 (YTHDF1) is an m6A recognition protein whose primary function is to promote mRNA translation [9]. Our preliminary data about bioinformatics prediction from ENCORI (https://rnasysu.com/encori/) suggests that YTHDF1 can bind to the mRNAs of glutathione peroxidase 4 (GPX4) and solute carrier family 7 member 11 (SLC7A11), both of which are classic markers of ferroptosis [10]. Also, SRAMP online database (http://www.cuilab.cn/m6asiteapp/old) shows both GPX4 and SLC7A11 mRNAs contain m6A modification sites. In addition, Chen et al. indicated that GPX4, through m6A and 5-methylcytosine (m5C) modifications, promoted anti-cancer immunity in colon adenocarcinoma by activating the cGAS-STING signaling pathway [11]. Methyltransferase-like 3 (METTL3) promoted lung adenocarcinoma progression by enhancing m6A modification of SLC7A11 mRNA, stabilizing its expression and inhibiting ferroptosis [12]. Furthermore, our preliminary experiments revealed that YTHDF1 expression was downregulated in TN-induced osteolysis model, a finding not previously reported. Previous studies proved that YTHDF1 played an essential role in ferroptosis inhibition. For instance, in prostate cancer, YTHDF1 impaired CD8⁺ T cell-driven anti-tumor immune response and inhibited ferroptosis by modulating the m6A-dependent regulation of programmed death-ligand 1 (PD-L1) [13]. Similarly, in nucleus pulposus cells, hypoxia-inducible factor-1α (HIF-1α)-induced YTHDF1 blocked ferroptosis by enhancing m6A-mediated SLC7A11 translation [14]. These findings highlighted YTHDF1 as a key regulator of ferroptosis and a potential therapeutic target in periprosthetic osteolysis. Based on this, we speculated that the downregulation of YTHDF1 expression in osteolysis might impair the translation of GPX4 and SLC7A11 mRNAs, leading to ferroptosis in osteoblasts. Many studies, including our preliminary data, suggest that Wnt/β-catenin pathway can boost osteoblast proliferation and differentiation [15, 16] and this signaling pathway is inhibited in osteolytic diseases [17, 18]. Additionally, extensive researches have documented that activation of Wnt/β-catenin pathway can inhibit ferroptosis in different types of diseases, including gastric cancer, melanoma, hepatocellular carcinoma, etc [19–21]. Bioinformatics prediction from JASPAR (https://jaspar.elixir.no/) indicates that transcription factor 4 (TCF4), a transcription factor in Wnt/β-catenin signaling pathway [22], can bind to the promoter region of YTHDF1. However, the relationship between TCF4 and YTHDF1, and the role of TCF4 in periprosthetic osteolysis have not been reported. Previous studies have elaborated the role of TCF4 in inhibiting ferroptosis. For instance, nuclear respiratory factor 1 (NRF1)-induced circular RNA, mmu_circ_0001388 and hsa_circ_0029470, has been shown to confer resistance to ferroptosis in ischemic acute kidney injury through the miR-193b-3p/TCF4/GPX4 axis [23]. Similarly, interferon γ (IFNγ) was a key cytokine that inhibited intestinal stem cell stemness and promoted ferroptosis in colorectal cancer stem cells via the IFN receptor 2 (IFNR2)/APC/TCF4/GPX4 axis [24]. However, the regulatory role of TCF4 in ferroptosis during periprosthetic osteolysis remains poorly understood. Therefore, we hypothesized that TCF4, activated by Wnt/β-catenin signaling pathway, enhanced the transcription of the m6A modification reader YTHDF1. This, in turn, enhanced the translation of GPX4 and SLC7A11 mRNAs, eventually inhibiting osteoblast ferroptosis to alleviate periprosthetic osteolysis. This pathway offers potential as a therapeutic target to prevent or treat periprosthetic osteolysis and improve the longevity of artificial joints. Materials and methods Clinical sample collection Patients undergoing primary joint arthroplasty and subsequent revision surgery at Xiangya Hospital of Central South University (Changsha, Hunan, China) were recruited. The clinical background of patients was shown in Table 1. X-ray imaging was performed after the initial joint replacement and prior to the revision surgery to assess the prosthesis status. The joint synovial tissues surrounding the prostheses were collected from two groups, during the initial joint arthroplasty surgery and subsequent revision procedures. These tissues were dissected and stored in liquid nitrogen for future experiments. The research protocols for human samples were approved by the Ethics Committee of Xiangya Hospital of Central South University. Informed consent was acquired from all donors. Table 1Clinical background of patientsCasesGenderAgeType of implantPreoperative diagnosisYears after implantationCase 1Female75TitaniumAseptic loosening8Case 2Male70TitaniumAseptic loosening9Case 3Male68TitaniumAseptic loosening22Case 4Male65TitaniumAseptic loosening15Case 5Male60TitaniumAseptic loosening15Case 6Female79TitaniumAseptic loosening13Case 7Female69TitaniumAseptic loosening16Case 8Female60TitaniumAseptic loosening16 Clinical background of patients Cell culture and treatment The human osteoblast cell line hFOB 1.19 and the mouse osteoblast cell line MC3T3-E1 were both obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). The hFOB 1.19 and MC3T3-E1 cells were cultured in Dulbecco’s modified Eagle medium (DMEM) or α-minimum essential medium (α-MEM) respectively, supplementing with 10% fetal bovine serum (FBS), 100 µg/mL streptomycin, and 100 U/mL penicillin. Both cell lines were maintained in an incubator with 5% CO2 at 37 °C. DMEM, α-MEM, FBS, streptomycin, and penicillin were all obtained from Gibco (Carlsbad, CA, USA). The titanium nanoparticles (TN) were supplied by Nanjing Emperor Nano Material Company (Nanjing, Jiangsu, China). The particles were baked at 180 °C for 6 h and then immersed in 75% ethanol for 96 h to remove endotoxins. Subsequently, the TN particles were suspended in sterile PBS to achieve a concentration of 200 mg/mL. Cells were treated with ferroptosis inhibitors including Ferrostatin-1 (hFOB 1.19: 10 µM; MC3T3-E1: 5 µM) and Liproxstatin-1 (hFOB 1.19: 5 µM; MC3T3-E1: 1 µM), or Wnt/β-catenin signaling pathway agonist SKL2001 (40 µM) for 2 h, then stimulated with TN particles at a density of 0.1 mg/mL for 48 h. Ferrostatin-1, Liproxstatin-1, and SKL2001 were all obtained from Sigma-Aldrich (St. Louis, MO, USA). Cell transfection Custom pLenti-CMV-MCS-mPGK-GFP-T2A-Puro lentiviral vector to overexpress or knockdown TCF4 and YTHDF1, and corresponding negative controls were ordered from GenePharma (Shanghai, China). HEK 293 T cells were co-transfected with lentiviral vectors, helper plasmids psPAX2 and pMD2.G (Addgene, Watertown, MA, USA) utiilizing Lipofectamine 3000 reagent (Invitrogen, Carlsbad, CA, USA). The culture medium was collected and filtered at 48 h post-transfection. The viral titers were tested, and the lentiviruses were used to infect osteoblasts in the presence of polybrene (5 µg/mL, Sigma-Aldrich) with multiplicity of infection of 15. Cells were further cultured for 72 h before selection with puromycin (2 µg/mL, Sigma-Aldrich). Cell counting kit-8 (CCK-8) assay Cell viability was assessed using a CCK-8 Kit (ab228554, Abcam, Cambridge, UK). Osteoblasts were seeded into a 96-well plate at a concentration of 5000 cells per well. After 48 h, 10 µL of CCK-8 solution was added into each well. The absorbance was measured after 4-hour incubation using a microplate reader (Thermo Fisher Scientific, Waltham, MA, USA). Detection of lipid peroxidation Osteoblasts treated with the method described above were seeded into 12-well plates at a density of 1 × 105 cells per well and incubated for approximately 12 h. Afterward, 5 µM of BODIPY 581/591 C11 (D3861, Invitrogen) was added into medium, followed by incubation for an additional half an hour at 37 °C. After incubation, the cells were then resuspended in 500 µL PBS and subjected to flow cytometry analysis (BD Biosciences, NJ, USA). Detection of malondialdehyde (MDA), iron, and glutathione (GSH) The levels of MDA, iron, or GSH in the cell or tissue samples were examined utilizing the Lipid Peroxidation Assay Kit (ab118970, Abcam), the Iron Assay Kit (ab83366, Abcam), or the Glutathione Assay Kit (CS0260, Sigma-Aldrich) respectively following suppliers’ instructions. RNA immunoprecipitation (RIP) assay hFOB1.19 or MC3T3-E1 cells, transfected with or without sh-YTHDF1 or sh-NC, were collected for RIP assay using the EZ-Magna RIP RNA-Binding Protein Immunoprecipitation Kit (17–701, Merck Millipore, MA, USA). The cells were lysed with RIP lysis buffer. Subsequently, the samples were incubated with either an anti-YTHDF1 antibody (1:50, ab220162, Abcam) or a non-specific IgG antibody (1:500, ab172730, Abcam) for 24 h at 4 °C. Protein-antibody complexes were captured using protein G Sepharose 4 Fast Flow suspension (Cytiva, Marlborough, MA, USA). To release the bound RNA, the beads underwent proteinase K (Sangon Biotech, Shanghai, China) digestion for 1 h. Subsequently, RNA purification was performed using Trizol reagent (15596026CN, Invitrogen), and the enrichment of GPX4 and SLC7A11 mRNAs was assessed via RT-qPCR. Chromatin immunoprecipitation (ChIP) assay ChIP method was conducted utilizing Pierce magnetic ChIP Kit (26157, Thermo Fisher Scientific). Briefly, hFOB 1.19 and MC3T3-E1 cells were cross-linked with 1% formaldehyde for 10 min, lysed using ChIP lysis buffer, and sonicated. Chromatin was immunoprecipitated using anti-TCF4 antibody (1:100, ab217668, Abcam). IgG (1:500, ab172730, Abcam) was served as negative control. qPCR analysis was performed to test enrichment of YTHDF1 promoter region in precipitated DNAs. Dual-luciferase reporter assay The experimental protocol followed a previously published study [25]. Briefly, both the YTHDF1 wild-type (WT) and mutant (MUT) promoter sequences were amplified and cloned into a pGL3-luciferase vector from Promega (Madison, WI, USA). Subsequently, hFOB 1.19 or MC3T3-E1 cells were co-transfected with WT or MUT luciferase plasmid, pRL-TK (Promega), and TCF4 overexpression plasmid or the corresponding NC plasmid, using Lipofectamine 3000 transfection reagent. After 48 h, luciferase activities were examined utilizing the Dual-Luciferase Reporter Assay Kit (E1910, Promega). Animal model Procedures for animal experiments were carried out following the protocols approved by the Institute of Animal Care and Use Committee of Xiangya Hospital of Central South University. C57BL/6J male mice aged 8–10 weeks were used and three independent sets of experiments were conducted to examine the effects of ferroptosis inhibitors and TCF4/YTHDF1 overexpression on TN-induced calvarial osteolysis. The mice were divided into following groups (n = 8 per group): sham, TN implantation, TN + Ferrostatin-1, TN + Liproxstatin-1, TN + Vector, TN + TCF4 overexpression, and TN + YTHDF1 overexpression. TN-induced calvarial osteolysis model was established based on a previously published protocol [26]. The mice were anesthetized through intraperitoneal injection of xylazine and ketamine, and the skin on the calvariae was shaved and disinfected before making a midline incision. Afterward, TN suspensions (30 mg in 30 µL) were uniformly applied to the surfaces of parietal bones, after which the surgical skin incisions were sutured. In sham-operated group, an equivalent volume of PBS was administered following the same surgical procedures. For Ferrostatin-1 or Liproxstatin-1 treatment, mice were intraperitoneally administered them at respective concentrations (1 mg/kg for Ferrostatin-1 and 10 mg/kg for Liproxstatin-1) once every two days for 14 days. For TCF4/YTHDF1 overexpression treatments, adeno-associated viruses (AAV) overexpressing TCF4 or YTHDF1 (Hanbio, Shanghai, China) in 200 µL PBS were injected via the tail vein. Fourteen days post-operation, all mice were sacrificed, and their calvariae were collected. Bone parameters, including bone mineral density (BMD), bone volume (BV), bone volume to tissue volume ratio (BV/TV), and total porosity, were quantified using CTAn software (SkyScan, Aartselaar, Belgium). Histological examinations The calvariae were fixed in 4% paraformaldehyde, embedded in paraffin, and cut into slices (4 μm-thick). After deparaffinization and rehydration, morphological assessment was conducted through hematoxylin and eosin (HE) staining (ab245880, Abcam). For immunohistochemical (IHC) staining, calvarial tissue sections underwent antigen retrieval by incubation in citrate buffer for 10 min. The samples were blocked with 3% hydrogen peroxide for 10 min, followed by a 30-minute blocking step with goat serum. The samples were then incubated with primary antibody against 4-hydroxy-2-nonenal (4-HNE, 1:250, ab48506, Abcam) for 12 h at 4 °C, following incubation with HRP-conjugated secondary antibody (1:500, ab6728, Abcam) for 1 h. Hematoxylin was used for counterstaining, and the slides were mounted before being imaged under a light microscope (Zeiss, Germany). RNA extraction and reverse transcription quantitative polymerase chain reaction (RT-qPCR) Total RNA was isolated with Trizol reagent and reverse-transcribed into complementary DNA (cDNA) with the PrimeScript™ First Strand cDNA Synthesis Kit (6110 A, Takara, Osaka, Japan) following the manufacturer’s protocol. Quantitative real-time PCR was carried out using the TaqMan® Universal PCR Master Mix (4305719, Thermo Fisher Scientific) with primers synthesized by Sangon Biotech. Relative mRNA expression levels were determined using the 2–ΔΔCt method and normalized to β-actin. Western blot Protein extraction was performed using Radio Immunoprecipitation Assay (RIPA) buffer with protease inhibitors (Beyotime, Shanghai, China) at 4 °C for half an hour. Protein concentration was examined with a Pierce BCA Protein Assay Kit (10741395, Thermo Fisher Scientific). Subsequently, 30 µg of protein was separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene fluoride (PVDF) membranes (Merck Millipore). The samples were blocked, rinsed with PBS, and incubated with primary antibodies against TCF4 (1:1500, ab217668, Abcam), YTHDF1 (1:1000, ab220162, Abcam), GPX4 (1:2000, ab125066, Abcam), SLC7A11 (1:2000, ab307601, Abcam), β-catenin (1:3000, ab32572, Abcam), and β-actin (1:5000, ab8227, Abcam). Following additional PBS washes, the samples were incubated with secondary antibody (1:5000, 31460, Invitrogen) and imaged utilizing Immobilon ECL Ultra Western HRP Substrate (WBULS0100, Merck Millipore). Statistical analysis All experiments were conducted in at least three biological replicates, and each biological replicate contained three technical replicates. The analyses were conducted using GraphPad Prism 8.0 software. Data are expressed as the mean ± standard deviation (SD). Comparisons between two groups were performed using an unpaired Student’s t-test. Comparisons among multiple groups were analyzed by one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test. A P-value of less than 0.05 was considered statistically significant. Results Ferroptosis was enhanced in TN-induced periprosthetic osteolysis in vivo and in vitro Postoperative X-rays obtained in patients following primary joint arthroplasty (primary TJA) and prior to the revision surgery (revision TJA) revealed periprosthetic osteolysis around the implants (Fig. 1A). The joint synovial tissues surrounding the implants were collected during the primary joint arthroplasty and subsequent revision surgery. Elevated levels of 4-HNE, MDA, and iron, along with the decreased GSH level, were observed in joint synovial tissues around the implants collected from revision surgery (Fig. 1B-C). Additionally, reduced protein expression of GPX4 and SLC7A11 was also detected in these tissues (Fig. 1D). The murine model of osteolysis was then established by TN implantation, and it was detected that bone mineral density (BMD), bone volume (BV), bone volume to tissue volume ratio (BV/TV) were decreased, and the total porosity was increased in TN-induced osteolysis model (Fig. 1E-F). Furthermore, HE staining further confirmed the presence of bone structure destruction of calvariae (Fig. 1G). The above findings confirmed the successful establishment of an animal osteolysis model. 4-HNE, MDA, and iron levels were increased, while GSH level was decreased in the calvariae of murine osteolysis model (Fig. 1H-I). Next, an in vitro osteolysis model was created using hFOB 1.19 and MC3T3-E1 cells. Treatment with TN significantly reduced the viability of hFOB 1.19 and MC3T3-E1 cells (Fig. S1A). Additionally, TN treatment markedly induced ferroptosis, as indicated by reduced levels of GPX4, SLC7A11, GSH (Fig. S1B-C), and increased intracellular MDA and iron levels (Fig. S1D-E). Results from BODIPY 581/591 C11 staining further confirmed the elevated lipid peroxidation level in TN-treated osteoblasts (Fig. S1F). These findings indicated that ferroptosis was enhanced in periprosthetic osteolysis. Ferroptosis was enhanced in TN-induced periprosthetic osteolysis in vivo. (A) Postoperative X-rays obtained in patients following primary joint arthroplasty (primary TJA) and prior to the revision surgery (revision TJA). (B-D) Joint synovial tissues surrounding the prostheses were collected during the primary TJA and subsequent revision surgeries. (B) IHC staining was used to detect the expression of 4-HNE within the tissues. Scale bar, 100 μm. (C) Commercial kits were used to measure the levels of GSH, iron, and MDA in clinical samples. (D) Western blot was conducted to detect the expression of GPX4 and SLC7A11 in clinical samples. n = 8 in clinical samples. (E-I) The C57BL/6J mice were divided into two groups: the sham group and the osteolysis model group. The mouse osteolysis model was established by evenly spreading 30 mg of TN particles (30 µL) on the surfaces of bilateral parietal bones, followed by closure of the surgical skin incision. Mice in sham group underwent the same surgical procedures but without received TN. (E-F) Bone mineral density (BMD), bone volume (BV), bone volume to tissue volume ratio (BV/TV), and total porosity were examined by micro-CT. (G) HE staining was used to assess bone microarchitecture of calvariae. Scale bar, 100 μm. (H) IHC staining was used to detect the expression of 4-HNE in calvarial tissues. Scale bar, 100 μm. (I) Commercial kits were used to measure the levels of GSH, iron, and MDA in calvarial tissues. 8 mice were included in each group. *P < 0.05, **P < 0.01, and ***P < 0.001 Ferroptosis inhibitors alleviated TN-induced periprosthetic osteolysis in vivo and in vitro To further verify whether ferroptosis was indeed involved in periprosthetic osteolysis, ferroptosis inhibitors including Ferrostatin-1 and Liproxstatin-1 were applied in TN-induced osteolysis models. The inhibitors effectively suppressed TN-induced ferroptosis, resulting in decreased MDA and iron levels, and elevated GSH level (Fig. 2A-C) in hFOB 1.19 and MC3T3-E1 cells. The treatment with ferroptosis inhibitors rescued the inhibitory effects of TN on GPX4 and SLC7A11 expression in osteoblasts (Fig. 2D). TN-induced lipid peroxidation was also suppressed following treatment with ferroptosis inhibitors (Fig. 2E). Furthermore, as shown in Fig. 2F, treatment with ferroptosis inhibitors significantly mitigated TN-induced decrease in the viability of hFOB 1.19 and MC3T3-E1 cells. Subsequently, a murine model was also injected with Ferrostatin-1 and Liproxstatin-1. The treatment with ferroptosis inhibitors significantly alleviated TN-induced osteolysis, as indicated by improved BMD, BV, BV/TV, and reduced total porosity (Fig. 3A-B). HE staining further confirmed the reduction of osteolysis, as demonstrated by improvement of bone microarchitecture of calvariae (Fig. 3C). Treatment with ferroptosis inhibitors significantly attenuated TN-induced increase in 4-HNE, MDA, and iron levels in calvariae of mice (Fig. 3D-E). Additionally, the inhibitory effect of TN on GSH was reversed by ferroptosis inhibitors in mice (Fig. 3E). Overall, ferroptosis inhibitors effectively attenuated TN-induced periprosthetic osteolysis both in vitro and in vivo. Ferroptosis inhibitors alleviated TN-induced periprosthetic osteolysis in vitro. Osteoblasts hFOB 1.19 and MC3T3-E1 were divided into four groups: control, TN, TN + Ferropstatin-1, and TN + Liproxstatin-1. Cells were treated with ferroptosis inhibitors including Ferrostatin-1 (hFOB 1.19: 10 µM; MC3T3-E1: 5 µM) and Liproxstatin-1 (hFOB 1.19: 5 µM; MC3T3-E1: 1 µM) for 2 h, followed by stimulated with 0.1 mg/mL TN particles for 48 h. (A-C) The contents of MDA, GSH, and iron were quantified using specific commercial kits. (D) Western blot analysis was performed to measure the expression levels of ferroptosis markers GPX4 and SLC7A11. (E) The production of lipid peroxides was detected by BODIPY 581/591 C11 staining followed by flow cytometry analysis. (F) Cell viability was assessed using the CCK-8 assay. All cell experiments were conducted in at least three biological replicates, and each biological replicate contained three technical replicates. *P < 0.05, **P < 0.01, and ***P < 0.001 Ferroptosis inhibitors ameliorated TN-induced periprosthetic osteolysis in vivo. C57BL/6J mice were divided into four groups: sham, TN, TN + Ferrostatin-1, and TN + Liproxstatin-1. Mice were intraperitoneally administered Ferrostatin-1 (1 mg/kg) or Liproxstatin-1 (10 mg/kg) once every two days for 14 days. (A-B) Bone mineral density (BMD), bone volume (BV), bone volume to tissue volume ratio (BV/TV), and total porosity were examined by micro-CT in different groups of mice. (C) HE staining was conducted to assess bone microarchitecture of calvariae. Scale bar, 100 μm. (D) IHC was used to detect 4-HNE expression in calvarial tissues. Scale bar, 100 μm. (E) Specific commercial kits were utilized to measure the contents of GSH, iron, and MDA in calvarial tissues. 8 mice were included in each group. *P < 0.05, **P < 0.01, and ***P < 0.001 m6A reader YTHDF1 promoted translation of GPX4 and SLC7A11 mRNAs in osteoblasts GPX4 and SLC7A11 mRNAs have been reported to exhibit m6A modification in other diseases [11, 12]. Therefore, we first examined the expression of a series of m6A modification readers in an in vitro osteolysis model. Among various m6A readers, YTHDF1 exhibited the most significant reduction in TN-treated osteoblasts (Fig. 4A). Both gene and protein levels of YTHDF1 were markedly reduced in TN-induced osteolysis models in vitro (Fig. 4A-B) and in vivo (Fig. 4C-D). RIP results demonstrated that YTHDF1 bound to the mRNAs of GPX4 and SLC7A11 in hFOB 1.19 and MC3T3-E1 cells (Fig. 4E). Knockdown of YTHDF1 inhibited the enrichment of GPX4 and SLC7A11 mRNAs in YTHDF1 protein (Fig. 4F). Furthermore, while YTHDF1 did not affect the gene expression of GPX4 and SLC7A11 (Fig. 4G), it positively regulated their protein levels (Fig. 4H) in hFOB 1.19 and MC3T3-E1 cells. Altogether, these results indicated that YTHDF1 enhanced the translation of GPX4 and SLC7A11 mRNAs in osteoblasts. m6A reader YTHDF1 promoted translation of GPX4 and SLC7A11 mRNAs in osteoblasts. (A) RT-qPCR analysis of the expression levels of various m6A readers in the in vitro osteolysis model. (B) Detection of YTHDF1 protein expression using western blot in TN-induced osteolysis model in vitro. (C) The gene expression of YTHDF1 was examined using RT-qPCR in TN-induced osteolysis model in vivo. (D) The protein expression of YTHDF1 was examined using western blot in TN-induced osteolysis model in vivo. (E) The interactions between YTHDF1 protein and GPX4 as well as SLC7A11 mRNAs were verified using RIP assay. (F) Investigation of the interactions between YTHDF1 protein and GPX4 as well as SLC7A11 mRNAs using RIP assay in hFOB 1.19 and MC3T3-E1 after YTHDF1 knockdown. (G-H) RT-qPCR and western blot detection of GPX4 and SLC7A11 levels in hFOB 1.19 and MC3T3-E1 after overexpression or knockdown of YTHDF1. 8 mice were included in each group of animal experiments. All cell experiments were conducted in at least three biological replicates, and each biological replicate contained three technical replicates. *P < 0.05, **P < 0.01, and ***P < 0.001 YTHDF1 overexpression attenuated ferroptosis in TN-induced osteolysis model in vitro The roles of YTHDF1 in ferroptosis within the in vitro osteolysis model were further validated by overexpressing YTHDF1. Both gene and protein levels of YTHDF1 were significantly weakened by TN treatment, whereas YTHDF1 overexpression markedly increased their levels in TN-treated osteoblasts (Fig. 5A-B). YTHDF1 overexpression restored the TN-induced suppression of GPX4 and SLC7A11 expression in hFOB 1.19 and MC3T3-E1 cells (Fig. 5C). It also increased GSH level (Fig. 5D) and reversed the elevation of intracellular iron (Fig. 5E) and MDA (Fig. 5F) levels caused by TN. Furthermore, YTHDF1 overexpression significantly reduced lipid peroxidation caused by TN treatment (Fig. 5G), and improved cell viability which had been reduced by TN treatment (Fig. 5H). Collectively, these findings suggested that YTHDF1 overexpression enhanced the translation of GPX4 and SLC7A11, thereby alleviating ferroptosis in TN-induced periprosthetic osteolysis. YTHDF1 overexpression attenuated ferroptosis in TN-induced osteolysis model in vitro. Osteoblasts hFOB 1.19 and MC3T3-E1 were divided into four groups: control, TN, TN + Vector, and TN + YTHDF1. (A-B) YTHDF1 expression was analyzed using RT-qPCR and western blot. (C) The expression levels of ferroptosis markers GPX4 and SLC7A11 were assessed using western blot. (D-F) The levels of GSH, iron, and MDA were quantified using respective commercial kits. (G) Lipid peroxidation was detected using BODIPY 581/591 C11 staining in conjunction with flow cytometry. (H) Cell viability was determined using the CCK-8 assay. All cell experiments were conducted in at least three biological replicates, and each biological replicate contained three technical replicates. *P < 0.05, **P < 0.01, and ***P < 0.001 TCF4 bound to the promoter of YTHDF1 and positively regulated its expression in osteoblasts To elucidate the upstream regulatory mechanism underlying YTHDF1 downregulation-mediated ferroptosis in osteolysis, a bioinformatics tool was used. It was predicted by JASPAR that the transcription factor TCF4 bound to the promoter region of YTHDF1, and the binding sites were shown (Fig. 6A). The relative luciferase activity was significantly increased in hFOB 1.19 and MC3T3-E1 cells that were transfected with TCF4 overexpression (OE-TCF4) vector and YTHDF1-WT (Fig. 6B). In contrast, the relative luciferase activity remained unchanged in osteoblasts transfected with YTHDF1-MUT (Fig. 6B). Furthermore, YTHDF1 promoter region was highly enriched in the anti-TCF4 group compared to the anti-IgG group (Fig. 6C). The transfection efficiency of TCF4 knockdown and overexpression was confirmed by western blot (Fig. 6D). Knockdown of TCF4 led to a significant reduction in YTHDF1 mRNA and protein expression, while TCF4 overexpression had the opposite results (Fig. 6E-F), suggesting TCF4 positively regulated YTHDF1 expression in osteoblasts. Additionally, TCF4 mRNA (Fig. 6G) and protein (Fig. 6H) levels were significantly reduced in in vitro osteolysis model. Consistently, the mRNA expression of TCF4 was also decreased in in vivo osteolysis model (Fig. 6I), indicating TN stimulation reduced TCF4 expression. Altogether, TN-induced downregulation of TCF4 led to decreased YTHDF1 expression in periprosthetic osteolysis. TCF4 bound to the promoter of YTHDF1 and positively regulated its expression in osteoblasts. (A) TCF4 transcription factor binding to the YTHDF1 promoter was predicted using the JASPAR bioinformatics tool. (B) The transcriptional activity of YTHDF1 was evaluated in TCF4 overexpressed hFOB 1.19 and MC3T3-E1 cells using dual-luciferase reporter assay. (C) ChIP assay was used to assess the binding of TCF4 protein to the YTHDF1 promoter in hFOB 1.19 and MC3T3-E1 cells. (D) TCF4 expression was assessed by western blot following overexpression and knockdown of TCF4. (E-F) YTHDF1 expression was examined by RT-qPCR and western blot following overexpression or knockdown of TCF4. (G-H) TCF4 expression was examined by RT-qPCR and western blot in the in vitro osteolysis model. (I) TCF4 expression was examined by RT-qPCR in the in vivo osteolysis model. 8 mice were included in each group of animal experiments. All cell experiments were conducted in at least three biological replicates, and each biological replicate contained three technical replicates. *P < 0.05, **P < 0.01, and ***P < 0.001 TCF4 transcriptionally upregulated YTHDF1 expression to diminish ferroptosis in osteolysis To further validate the roles of TCF4 in YTHDF1 reduction-mediated ferroptosis in osteolysis, TCF4 was overexpressed and YTHDF1 was silenced in osteoblasts treated with TN. Overexpression of TCF4 attenuated the inhibitory effect of TN on the expression of TCF4, YTHDF1, GPX4 and SLC7A11 (Fig. 7A). It also restored GSH level suppressed by TN treatment (Fig. 7B). But, YTHDF1 knockdown had no effects on TCF4 expression and reversed other effects of TCF4 overexpression (Fig. 7A-B). Additionally, TCF4 overexpression reversed the TN-induced increase in iron accumulation (Fig. 7C), MDA level (Fig. 7D), and lipid peroxidation (Fig. 7E). However, knockdown of YTHDF1 reversed the protective effects induced by TCF4 overexpression (Fig. 7C-E). TCF4 overexpression increased cell viability following TN treatment, which was reversed by YTHDF1 knockdown (Fig. 7F). Collectively, YTHDF1 depletion reversed the protective effects of TCF4 overexpression against TN-induced ferroptosis in osteoblasts. TCF4 transcriptionally activated YTHDF1 expression to diminish ferroptosis in osteolysis. Osteoblasts hFOB 1.19 and MC3T3-E1 were divided into six groups: control, TN, TN + Vector, TN + TCF4, TN + TCF4 + sh-NC, and TN + TCF4 + sh-YTHDF1. (A) The expression levels of TCF4, YTHDF1, GPX4, and SLC7A11 were analyzed using western blot in hFOB 1.19 and MC3T3-E1 cells. (B-D) The levels of GSH, iron, and MDA were quantified using respective commercial kits. (E) Lipid peroxidation was detected using BODIPY 581/591 C11 staining in conjunction with flow cytometry. (F) Cell viability was determined using the CCK-8 assay in hFOB 1.19 and MC3T3-E1 cells. All cell experiments were conducted in at least three biological replicates, and each biological replicate contained three technical replicates. *P < 0.05, **P < 0.01, and ***P < 0.001 TCF4 and YTHDF1 overexpression ameliorated TN-induced osteolysis by restraining ferroptosis in vivo The regulatory role of TCF4/YTHDF1 axis was further validated in an in vivo murine model. Overexpression of TCF4 or YTHDF1 significantly alleviated osteolysis in TN-treated mice, evidenced by improved BMD (Fig. 8A), increased BV, a higher BV/TV ratio, and reduced total porosity (Fig. 8B). HE staining results further confirmed the amelioration of bone microarchitecture of calvariae in TCF4 and YTHDF1 overexpression groups (Fig. 8C). IHC analysis of 4-HNE expression indicated a decrease in lipid peroxidation with TCF4/YTHDF1 overexpression (Fig. 8D). Additionally, TCF4 or YTHDF1 overexpression led to increased GSH level and decreased levels of iron and MDA in calvariae (Fig. 8E). Overexpression of TCF4 significantly increased the expression of TCF4 and YTHDF1 in calvariae of osteolysis model, but, YTHDF1 overexpression exerted no obvious effects on TCF4 expression (Fig. 8F). Collectively, these results suggested that TCF4 or YTHDF1 overexpression mitigated osteolysis via refraining ferroptosis in TN-treated mice. TCF4 and YTHDF1 overexpression ameliorated TN-induced osteolysis by restraining ferroptosis in vivo. C57BL/6J mice were divided into four groups: sham, TN, TN + Vector, TN + TCF4, and TN + YTHDF1 group. Adeno-associated viruses (AAV) overexpressing TCF4 or YTHDF1 in 200 µL PBS were injected via the tail vein. (A-B) Bone mineral density (BMD), bone volume (BV), bone volume to tissue volume ratio (BV/TV), and total porosity were examined by micro-CT. (C) HE staining was conducted to assess bone microarchitecture of calvariae. Scale bar, 100 μm. (D) IHC was used to detect 4-HNE expression in calvarial tissues. Scale bar, 100 μm. (E) Specific commercial kits were utilized to measure the contents of GSH, iron, and MDA levels in calvarial tissues. (F) The expression levels of TCF4 and YTHDF1 were assessed in calvarial tissues using RT-qPCR. 8 mice were included in each group. *P < 0.05, **P < 0.01, and ***P < 0.001 TCF4 upregulation mediated by Wnt/β-catenin pathway activation suppressed ferroptosis in TN-induced osteolysis model in vitro Finally, the involvement of Wnt/β-catenin pathway in ferroptosis of osteolysis was investigated. The activation of Wnt/β-catenin pathway by SKL2001 upregulated the expression levels of β-catenin, TCF4, and YTHDF1 in TN-treated osteoblasts (Fig. S2A). Additionally, activation of this pathway led to increased levels of ferroptosis markers, including GPX4 and SLC7A11 (Fig. S2B). SKL2001 treatment enhanced cell viability in TN-induced osteoblasts (Fig. S2C). Activation of Wnt/β-catenin signaling pathway by SKL2001 resulted in higher GSH level (Fig. S2D) and lower iron (Fig. S2E) and MDA (Fig. S2F) levels. Furthermore, BODIPY 581/591 C11 staining combined with flow cytometry confirmed a decrease in lipid peroxidation upon SKL2001 treatment (Fig. S2G). These results suggested that activating Wnt/β-catenin pathway mitigated TN-induced osteolysis by suppressing ferroptosis via the TCF4/YTHDF1 axis. Discussion Wear particles generated by joint prostheses often cause aseptic loosening and osteolysis, significantly reducing the lifespan of artificial joints [2]. Osteolysis primarily results from impaired osteoblast function, however, it remains unclear whether osteoblast ferroptosis plays a crucial role in wear particles-induced osteolysis. This study demonstrated that ferroptosis was markedly enhanced in clinical synovial tissues surrounding implants and TN-induced osteolysis models both in vitro and in vivo. The significantly decreased levels of TCF4 and YTHDF1 were firstly observed in osteolysis models. We also firstly revealed that TCF4 directly bound to the promoter region of YTHDF1, leading to the transcriptional upregulation of YTHDF1 in osteoblasts. Furthermore, YTHDF1 specifically bound to the mRNAs of GPX4 and SLC7A11, promoting their translation and thereby inhibiting ferroptosis in osteolysis. Finally, we found that the activation of Wnt/β-catenin pathway boosted TCF4/YTHDF1 axis. Collectively, our results emphasized the critical role of ferroptosis in periprosthetic osteolysis, suggesting Wnt/β-catenin pathway and TCF4/YTHDF1 axis might be potential therapeutic targets to mitigate ferroptosis in osteolysis. Periprosthetic osteolysis is primarily caused by impaired osteoblast function around the implants [27]. This dysfunction includes reduced bone formation, increased bone resorption, and elevated susceptibility to cell death pathways, such as ferroptosis [28]. Ferroptosis has been increasingly recognized in the initiation and progression of orthopedic diseases, such as osteoporosis, osteoarthritis, and osteolysis [29, 30]. For example, it was shown that melatonin prevented high glucose-induced osteoblast ferroptosis in type 2 diabetic osteoporosis by activating Nrf2/HO-1 signaling pathway [31]. Another study reported that metformin mitigated osteoarthritis progression by modulating chondrocyte lipid metabolism and inhibiting ferroptosis [32]. A recent study by Xu et al. documented that peri-implant osteolysis induced by CoCrMo nanoparticles was mediated through the promotion of osteoblast ferroptosis via Nrf2/ARE signaling pathway [33]. Furthermore, they indicated that blocking ferroptosis with Ferrostatin-1 and the Nrf2 activator Oltipraz improved particles-induced osteolysis [33]. TN particles are widely used in clinical prosthetic materials due to their superior biocompatibility and mechanical properties [34]. However, further studies are needed to explore the potential roles of ferroptosis in osteolysis caused by TN. In our study, we found that ferroptosis was promoted in clinical synovial tissues surrounding implants and TN-induced osteolysis models both in vitro and in vivo, characterized by elevated levels of 4-HNE, MDA, iron, alongside decreased GSH, GPX4, and SLC7A11 levels. Additionally, we observed treatment with ferroptosis inhibitors significantly suppressed the above phenomena. These results further support that ferroptosis plays a vital role in the development of TN-induced periprosthetic osteolysis. Only two published studies have demonstrated that TN induces ferroptosis in osteoblasts, contributing to periprosthetic osteolysis. Specifically, Dong et al. reported that DNA methylation-mediated repression of GPX4 led to osteoblast ferroptosis and TN-induced osteolysis [35]. Additionally, another study showed that activation of GPX4 by small molecules such as Urolithin A inhibited osteoblast ferroptosis and restored osteogenesis, significantly alleviating peri-prosthetic osteolysis [36]. However, the upstream regulatory mechanisms governing ferroptosis in this context remain largely unexplored. YTHDF1, as a specific reader of m6A modification, enhances the translation efficiency of its target mRNAs by binding to initiation factors and facilitating ribosome assembly [9]. Past literatures have shown that YTHDF1 plays a significant role in regulating both physiological and pathological processes, including osteogenesis and bone metastasis [37, 38]. YTHDF1 induced human bone marrow mesenchymal stem cell differentiation and osteogenesis by translationally regulating zinc finger protein 839 (ZNF839), which enhanced Runx2 transcriptional activity [38]. Additionally, YTHDF1 promoted bone metastasis in breast cancer by enhancing the expression of EZH2 and CDH11 [39]. However, the understanding of YTHDF1 in osteolysis remains limited. In the present study, we firstly demonstrated that YTHDF1 was downregulated in TN-induced osteolysis models, and YTHDF1 targeted GPX4 and SLC7A11 mRNAs to promote their translation. Furthermore, we showed that YTHDF1 overexpression improved bone microarchitecture and protected osteoblasts against TN-induced ferroptosis both in vitro and in vivo. To the best of our knowledge, this is the first report documenting the roles of YTHDF1 in TN-induced osteolysis. We further investigated the upstream regulatory mechanisms of YTHDF1, for the first time, identified that TCF4 was decreased in TN-induced osteolysis and it bound to the promoter region of YTHDF1 to positively regulate its expression in osteoblasts. The role of TCF4 in bone diseases was also reported in previous studies. MiR-129-5p was shown to inhibit osteoblast differentiation and bone formation by targeting TCF4 and suppressing Wnt/β-catenin pathway [40]. Additionally, the mechanism by which TCF4 binds to the promoter region of downstream genes to transcriptionally regulate their expression has been documented. For instance, the β-catenin/TCF4 complex directly bound to ZEB1 promoter to activate its transcription in colorectal carcinoma cells [41]. Similarly, Pei et al. demonstrated that β-catenin/TCF4 complex targeted the promoter of FGF21 and mediated its transcription in mouse stomach cancer cells [42]. But, we are the first team to report the relationship between TCF4 and YTHDF1 in TN-induced periprosthetic osteolysis. TCF4, a recognized co-activator of β-catenin, serves as a crucial signaling molecule in Wnt signaling pathway, which plays a critical role in osteoblast differentiation and the regulation of bone formation, and involves in regulating many osteolytic diseases [43]. It was reported that melatonin suppressed TN-induced osteolysis by activating Wnt/β-catenin pathway [44]. Another study indicated that resveratrol reduced the progression of TN-induced osteolysis through inducing Wnt/β-catenin pathway [45]. Furthermore, the reduction of the SOST gene mitigated particles-induced osteolysis by enhancing Wnt/β-catenin signaling pathway [46]. The Wnt/β-catenin signaling pathway also plays a critical role in mediating ferroptosis in different cell types. For instance, USP8 promoted hepatocellular carcinoma tumorigenesis and imparted resistance to ferroptosis by stabilizing β-catenin [47]. Similarly, Wnt/β-catenin signaling inhibited ferroptosis in gastric cancer by targeting GPX4 [19]. In agreement with these findings, we revealed that Wnt/β-catenin signaling pathway agonist SKL2001 boosted TCF4/YTHDF1 axis and attenuated TN-induced ferroptosis. Application of Wnt/β-catenin signaling pathway agonists may serve as an effective strategy to alleviate TN-induced periprosthetic osteolysis by inhibiting ferroptosis. In addition to classical ferroptosis inhibitors such as Ferrostatin-1 and Liproxstatin-1 we used, several clinically approved drugs, including Atorvastatin and Celecoxib, have shown potential in suppressing ferroptosis in various disease models [48–51] and also have been used in orthopedic diseases​​ [52–55]. These agents may exert additive or synergistic effects when combined with genetic modulation of TCF4/YTHDF1 axis, offering a promising strategy to enhance therapeutic efficacy against ferroptosis-mediated osteolysis. Although this study primarily focuses on osteoblast ferroptosis, osteoclast activation is also a critical factor in the development of periprosthetic osteolysis [56]. Increasing evidence indicates that Wnt/β-catenin pathway not only promotes osteoblast function but also suppresses osteoclast differentiation and activation [57–60]. For instance, compounds such as Imperatorin enhanced osteogenesis while inhibiting osteoclastogenesis through activating AKT/GSK3β/β-catenin pathway [57]. Furthermore, conditional deletion of ATP6AP2 in osteoclast-lineage cells led to increased osteoclastogenesis and trabecular bone loss by reducing Wnt/β-catenin signaling [58]. Given that the TCF4 is modulated by Wnt/β-catenin signaling, it is plausible that TCF4/YTHDF1/GPX4/SLC7A11 regulatory axis may also influence osteoclast differentiation and activation, thereby contributing to the overall control of periprosthetic osteolysis and aseptic loosening. This plausible mechanism calls for further investigation to clarify its mechanistic contribution and determine its clinical or translational potential in periprosthetic osteolysis. In summary, our research demonstrated that the activation of Wnt/β-catenin signaling pathway led to the upregulation of TCF4, which enhanced translation of GPX4 and SLC7A11 mRNAs and reduced ferroptosis by promoting YTHDF1 transcription in TN-induced periprosthetic osteolysis. This suggests that TCF4/YTHDF1 axis may represent a promising target for preventing TN-induced osteolysis. While the TCF4/YTHDF1 axis holds therapeutic potential, its clinical realization requires rigorous target validation, innovative delivery system, and patient-specific approaches. Collaborative efforts integrating structural biology, nanotechnology, and translational medicine will be critical to overcoming these challenges and advancing precision therapies. Supplementary Information Supplementary Material 1. Supplementary Material 2. Author contributions Da Zhong and Chenggong Wang wrote the main manuscript text and Wenqing Xie, Zhan Liao, Hua Liu, Lemei Zhu, Fawei Gao, Xi Li, Zhen Yin, Ganzhe Xu, Peng Chen, Jian Tian, Long Wang, Ke Yin and Kunli Chen prepared figures. All authors reviewed the manuscript. Funding This work was supported by the National Natural Science Foundation of China (82472609), Natural Science Foundation of Hunan Province (2023JJ30941 and 2022JJ30941), Hunan Provincial Education Commission Foundation (24B0885), Hunan Provincial Research Project on Degree and Graduate Education Reform (2025JGYB039) and Horizontal Research Funding of Central South University (JXDY-ZD-20231101). Data availability All data generated or analyzed during this study are included in this published article. Declarations Ethics approval and consent to participate The research protocols for human samples were approved by the Ethics Committee of Xiangya Hospital of Central South University (Changsha, Hunan, China). Procedures for animal experiments were conducted following the protocols approved by the Institute of Animal Care and Use Committee of Xiangya Hospital of Central South University. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Supplementary Material 2. Data Availability Statement All data generated or analyzed during this study are included in this published article.