Abstract

   Peri-prosthesis osteolysis (PPO) represents the most severe
   complication of total joint arthroplasty (TJA) surgery and imposes the
   primary cause of prosthesis failure and subsequent revision surgery.
   Antiresorptive therapies are usually prescribed to treat PPO,
   especially for elderly people. Nevertheless, the efficacy of
   anti-osteoporotic medications remains constrained. Recent therapeutic
   strategies to promote periprosthetic osseointegration by restoring
   osteoblast function are considered more effective approaches. However,
   the precise mechanism underlying the inhibition of osteogenesis
   triggered by wear particles remains enigmatic. Herein, we demonstrate
   that wear particles inhibit osteoblast function by inducing ferroptosis
   to sabotage extracellular mineralization and arouse periprosthetic
   osteolysis. The suppression of ferroptosis could significantly rescue
   osteogenesis thus alleviating PPO. Furthermore, Glutathione Peroxidase
   4 (GPX4) has been identified as a key target in regulating osteoblastic
   ferroptosis. By utilizing virtual screening techniques, we have
   successfully conducted a comprehensive screening of a natural compound
   known as Urolithin A (UA), which exhibits remarkable inhibition of
   osteoblastic ferroptosis while simultaneously promoting the process of
   osteogenesis through its precise targeting mechanism on GPX4.
   Meanwhile, UA improves the osteolytic conditions significantly in vivo
   even when the adjunction of titanium (Ti) nanoparticles. This strategy
   has great potential in treating peri-prosthesis osteolysis and
   potentially broadens the scope of clinical therapy.

Graphical Abstract

   [46]graphic file with name 12951_2024_3049_Figa_HTML.jpg

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s12951-024-03049-4.

   Keywords: Peri-prosthesis osteolysis, Titanium nanoparticles,
   Osteoblastogenesis, Ferroptosis, GPX4, Urolithin A

Introduction

   Total joint arthroplasty (TJA) is a surgical strategy widely used to
   treat degenerative, post-traumatic severe arthritis of the hip and knee
   [[47]1]. However, the wear particle-induced aseptic loosening around
   the prosthesis and subsequent osteolysis can lead to TJA failure and
   surgical revision [[48]2, [49]3]. Nanoparticles including titanium(Ti),
   chromium and polyethylene generated by prosthetic wear accumulate at
   the bone-implant interface and are engulfed by immune cells to induce
   an inflammatory response, thus leading to the further promotion of
   osteoclastic bone resorption and inhibition of osteoblast-mediated
   extracellular matrix mineralization as well as the eventual occurrence
   of PPO [[50]4]. Based on the underlying pathological mechanisms of PPO,
   endeavors have been undertaken to employ bone resorption inhibitors for
   the treatment of PPO [[51]5]. Despite some reduction in osteoclast
   activation and bone resorption, the outcomes remain unsatisfactory with
   frequent occurrence of side effects. Targeting osteoblastic function
   could alleviate the pathological process of PPO and extend the life of
   a prosthesis by improving the specific mechanism due to the direct
   detrimental impact of wear particles on osteoblasts [[52]6]. As a
   consequence, it is necessary to further explore the mechanism by which
   wear particles affect osteoblasts and extracellular matrix
   mineralization for the development of novel and effective therapies.

   Ferroptosis, a novel form of the pathway that regulates cell death, is
   characterized by iron-mediated accumulation of lipid peroxidation
   [[53]7]. Previous studies have shown that ferroptosis in osteoblasts
   could be induced by a variety of pathological conditions, which is
   manifested by a decrease in the osteogenic differentiation of
   mesenchymal stem cells, as well as the decreased cell activity of
   mature osteoblasts and the damaged function of extracellular matrix
   mineralization [[54]8, [55]9]. As a result of downregulation of
   Runt-related transcription factor 2 (Runx2) expression, ferroptosis
   inhibits osteogenic commitment and differentiation in mesenchymal stem
   cells and pre-osteoblasts, leading to a decrease in alkaline
   phosphatase and osteocalcin [[56]10]. Ferroptosis could also suppress
   osteoblast growth and impair osteoblast differentiation by
   downregulating PI3K/AKT/FOX3a/DUSP14 signaling pathway [[57]11,
   [58]12]. Moreover, Xu et al. discovered that the presence of CoCrMo
   nanoparticles facilitated osteoblastic ferroptosis, thereby
   exacerbating the pathological progression of aseptic loosening
   [[59]13]. However, the underlying mechanisms and optimal targets for
   intervention warrant further intricate investigation.

   Recent studies have clearly demonstrated the significant role of System
   Xc^−/GSH/GPX4 in regulating ferroptosis. GPX4 utilizes reduced
   glutathione (GSH) to efficiently convert phospholipid hydroperoxides
   into lipid alcohols, effectively preventing ferroptosis [[60]14].
   Inhibiting the activity of system Xc^− (cystine/glutamate transporter)
   and GPX4 induces ferroptosis, highlighting the crucial role of GPX4 in
   regulating this process [[61]15]. Additionally, studies have shown that
   the NRF2/System Xc^−/GPX4 pathway can be activated to counteract
   ferroptosis and promote bone regeneration when there is an excess of
   reactive oxygen species (ROS) and lipid peroxidation [[62]16, [63]17].
   However, the role of GPX4 in PPO and related mechanisms remain unclear.
   PPO is characterized by elevated levels of ROS and reduced osteogenic
   function, suggesting that GPX4 and ferroptosis may play a critical role
   in related pathological processes. Therefore, targeting GPX4 may be a
   potential therapeutic strategy for PPO.

   In the current investigation, we found that osteoblastic ferroptosis
   constituted a pivotal factor contributing to the pathogenesis of PPO.
   The induction of osteoblastic ferroptosis by Ti nanoparticles and the
   subsequent reduction in osteogenic function observed in clinical PPO
   samples, an osteolysis mouse model, and during osteoblast
   differentiation were all correlated with a significant decrease in GPX4
   expression. The specific inhibitor of ferroptosis, known as
   ferrostatin-1 (Fer-1), exerts a suppressive effect on osteoblast
   susceptibility to ferroptosis by elevating the expression of GPX4,
   thereby rescuing their osteogenic function and effectively halting the
   progression of peri-prosthesis osteolysis. Taking into account the
   instability of Fer-1 and its implications for further clinical
   application, as well as the crucial anti-ferroptosis role of GPX4, we
   identified a small molecule urolithin A (UA) through virtual screening
   and molecular docking targeting GPX4, which possesses stable
   physicochemical properties, minimal toxicity, antioxidant and
   rejuvenating properties, significantly improving osteogenic function
   [[64]18, [65]19]. Mechanistically, the protective effects of UA against
   osteoblastic ferroptosis rely on the activation of GPX4 to mitigate Ti
   nanoparticle-induced osteoblastic ferroptosis. In vivo, UA exerts a
   therapeutic impact on peri-prosthesis osteolysis induced by Ti
   nanoparticles through its potent stimulation of bone formation. In
   conclusion, this study provides a new insight into Ti
   nanoparticle-induced peri-prosthesis osteolysis and a novel therapeutic
   approach for PPO.

Method

Drugs and reagents

   The acquisition of Ti nanoparticles was made from the Nanjing Emperor
   Nano Materials Company. Fer-1, UA, N-acetylcysteine (NAC)and Erastin
   were available from MedChem Express (MCE, USA). Minimum essential
   medium α (α-MEM, for culturing bone marrow-derived mesenchymal stem
   cells (BMSCs) and MC3T3-E1 cells). Fetal bovine serum (FBS) was
   purchased from VivaCell (China). Phosphate buffer saline (PBS) was
   purchased from Solarbio (China). RIPA lysis buffer and bovine serum
   albumin (BSA) were procured from Beyotime (China) for experimental use.

Cell culture and viability assay

   The MC3T3-E1 cell line was procured from the Cell Bank affiliated with
   the Chinese Academy of Sciences. BMSCs were collected from 6-week-old
   SD rats, as previously described [[66]20]. The BMSCs and MC3T3-E1 cells
   were cultured in α-MEM supplemented with 10% FBS at a temperature of
   37 °C under a CO[2] concentration of 5%. Osteogenesis of BMSCs or
   MC3T3-E1 cells was induced with Osteogenic medium (Sigma, USA). To
   mimic the osteolytic microenvironment, the cells were exposed to a
   medium supplemented with 5 µg/cm^2 or 10 µg/cm^2 Ti nanoparticles,
   which was refreshed every three days. The cells were subsequently
   exposed to 5µmol/L and 10 µmol/L of UA or 5µmol/L Fer-1 for treatment.

   Cell viability in MC3T3-E1 cells and BMSCs survivability were tested
   using CCK8 assay kit (Yeasen, China) according to the manufacture’s
   instruction. BMSCs and MC3T3-E1 cells were seeded at a density of
   4 × 10^3 cells per well in 96-well plates. In order to evaluate the
   cytotoxicity of Ti nanoparticles, Fer-1, UA, and Erastin on these
   cells, different concentrations were added and incubated. The MC3T3-E1
   cells and BMSCs inhibition were plotted using GraphPad Prism 8.

Alkaline phosphatase (ALP) assay

   BMSCs were seeded at a density of 1 × 10^4 cells per well in 24-well
   plates and cultured in osteogenic medium for a duration of 7 days with
   indicated treatment. The cells were washed three times with PBS prior
   to being fixed in 4% paraformaldehyde at a temperature of 4°C for a
   duration of 20 min. After undergoing three times of PBS washing, the
   cells were subjected to incubation in a staining solution comprising
   5-bromo-4-chloro-3-indolyl-phosphate/nitro blue tetrazolium (BCIP/NBT,
   Beyotime, China) for a duration of 30 min at room temperature under
   light-free conditions. Subsequently, the absorbance was measured at
   530 nm.

Alizarin red S (ARS) staining

   A mineralization evaluation was conducted using the alizarin red
   staining assay. 1 × 10^4 BMSCs were seeded in each well in a 24-well
   plate and subjected to a 21-day culture in an osteogenic medium. After
   undergoing three rounds of PBS washing, BMSCs were meticulously fixed
   in a 4% polyformaldehyde solution for a duration of 15 min.
   Subsequently, the cells were delicately incubated with alizarin red
   solution (Yeasen, China) at room temperature for an additional period
   of 15 min. Then BMSCs were washed 3 times with PBS before photographing
   under a microscope. The calcium nodules were dissolved in a solution
   containing 10% monohydrate of 1-Hexadecylpyridinium chloride, resulting
   in the dissolution of the alizarin red complex. Absorbance was recorded
   at 600 nm and analyzed using Image J software (National Institutes of
   Health, USA).

Total RNA isolation and real-time PCR

   Total RNA was isolated from cells and tissues using TRIzol reagent
   (Beyotime, China) and NanoDrop (Thermo Fisher Scientific, USA) was
   utilized for RNA concentration measurement. Then, total RNA reverse
   transcribed with PrimeScriptTM RT Master Mix (Yeasen, China). Real-time
   PCR was performed in a CFX96™ thermal cycler (Bio-Rad Laboratories).
   The 2^−ΔΔCq formula was performed to quantify the gene expression.
   Human and Mouse GAPDH gene was used as endogenous control for sample
   normalization. Primers are listed in Table [67]S2 and [68]S3.

Western blotting

   Proteins in samples were obtained by lysing with RIPA buffer (Beyotime,
   China). The BCA protein assay (Beyotime, China) was utilized to
   determine the protein concentrations. Afterwards, a careful transfer of
   the proteins onto a polyvinylidene fluoride membrane was conducted. The
   membranes were sealed with TBS containing 5% skimmed milk (Yeasen,
   China) for 1 h and incubated in primary antibody solution at 4 °C
   overnight. The following primary antibodies were put into use: COL1-a1
   (ab260043, 1:1000), Runx2 (ab23981, 1:1000), Osterix (ab209484,
   1:1000), GPX4 (ab125066, 1:1000), SLC7A11(ab307601, 1:1000), COX2
   (ab1798001:1000), 4HNE (ab46545, 1:1000) all obtained from Abcam. After
   incubating with HRP-conjugated secondary antibodies (Beyotime, China)
   for 1 h at ambient temperature. Then, the ECL kit was utilized to
   observe the membranes. Protein quantification was performed using Image
   J.

RNA sequencing

   Whole-transcriptome sequencing was performed using MC3T3-E1 cells
   induced by Ti nanoparticles (10 µg/cm^2) in osteogenic medium for 72 h
   and Relevant RNA was acquired using the conventional TRIzol extraction
   technique. Samples were sequenced using Illumina HiSeq 4000 instrument.
   We conducted comprehensive Gene Ontology (GO) and Kyoto Encyclopedia of
   Genes and Genomes (KEGG) pathway analyses to delve deeper into the
   subject matter. To further enhance our investigation, we employed
   OmicStudio tools for gene set enrichment analysis (GSEA) and
   bioinformatics analysis.

Transmission electron microscope

   Transmission Electron Microscopy (TEM) was employed to examine changes
   in mitochondrial morphology. After the specified treatment, cells were
   subjected to fixation using a 2.5% solution of glutaraldehyde (Electron
   Microscope Grade) (Solarbio, China) for a duration of 16 h at a
   temperature of 4 °C. After fixation, the cells underwent dehydration,
   embedding, sectioning, and staining in a meticulous sequence. Finally,
   the TEM samples were elegantly visualized utilizing the capabilities of
   a state-of-the-art Hitachi Model H-7650 transmission electron
   microscope.

Ferroptosis-related detection

Mitochondrial membrane potential detection

   Mitochondrial membrane potential was measured by fluorescence
   microscopy of cells using Mitochondrial Membrane Potential Assay Kit
   with TMRE (Beyotime, China). Briefly, BMSCs were seeded in a 24-well
   plate at a density of 2 × 10^4 per well, cultured overnight, and
   treated as described above. The cells were then incubated with 1 TMRE
   staining solution for 30 min in the dark at 37 °C and analyzed with a
   fluorescent microscope (Carl Zeiss, German).

ROS analysis

   To measure the ROS production, we used a
   2,7-dichloro-dihydro-fluorescein diacetate (DCFH-DA) staining kit
   (Beyotime, China). Essentially, MC3T3-E1 cells were incubated for
   30 min with DCFH-DA in FBS-free media in different groups in the dark
   for 20 min at 37 °C. The cells were meticulously fixed and captured
   through the lens of a fluorescence microscope, courtesy of Carl Zeiss
   (Germany).

Lipid peroxidation assay and GSH assays

   The assessment of cellular lipid peroxidation was conducted using the
   MDA assay kit (Beyotime, China) to measure the concentration of
   malondialdehyde (MDA). MC3T3-E1 cells were treated as described above.
   After indicated treatment, the cells were thoroughly rinsed with
   phosphate-buffered saline (PBS) and subsequently lysed using a lysis
   buffer solution. The supernatant was collected and utilized for the
   detection of MDA after centrifuge. The MDA content is carried out in
   strict accordance with the manufacturer’s guidelines. A microplate
   reader was utilized to assess the absorbance of cellular MDA at a
   wavelength of 532 nm. The relative concentration of GSH in cell lysates
   was quantified using a glutathione assay kit (Sigma, USA) following the
   manufacturer’s instructions.

FerroOrange staining

   Intracellular Fe^2+ level was analyzed by FerroOrange staining.
   1 × 10^4 BMSCs per well were cultured in 24-well plates. After
   undergoing the specified treatments, osteocytes were thoroughly rinsed
   with HBSS solution and subsequently exposed to 1 µM FerroOrange probes
   (DOJINDO, Japan) for a duration of 30 min. The resulting stained cells
   were then captured using fluorescence microscopy (Carl Zeiss, Germany).

Iron measurement

   To evaluate the Ti nanoparticle-induced ferroptosis, cellular iron was
   detected using the Iron Assay Kit (Solarbio, China) according to the
   manufacturer’s instructions. Briefly, 4 × 10^5 MC3T3-E1 cells per well
   were cultured in 6-well plates. After the specified treatments,
   mineralized osteocytes were collected and lysed in a buffer
   specifically designed for iron analysis. The resulting cell lysate was
   then subjected to centrifugation at 8000 × g for a duration of 10 min.
   Furthermore, the liquid portion above the sediment was utilized to
   determine the concentration of iron present. The experimental protocol
   strictly adhered to the guidelines provided by the manufacturers. A
   microplate reader was employed to measure absorbance at a wavelength of
   510 nm.

Immunofluorescence staining

   BMSCs were seeded on coverslips in 24-well plates in α-MEM for 24 h and
   then cultured in osteogenic differentiation medium with different
   interventions. Cells were seeded in 24-well plates at a density of
   2 × 10^4 per well. After undergoing various treatment conditions, the
   BMSCs were subsequently exposed to 4% paraformaldehyde (PFA, NCM
   Biotech, China) at ambient temperature for a duration spanning from 20
   to 30 min. The cells were subsequently treated with 0.5% Triton X-100
   for a duration of 15 min to induce permeabilization. After being
   obstructed by a 1% BSA solution enriched with bovine serum albumin for
   a duration of 60 min, the cells were subjected to an overnight
   incubation at a chilly temperature of 4°C in the presence of the
   primary antibody. The OCN dilution ratio used for immunofluorescence
   staining was 1:400 (abcam, USA). Subsequently, the cells underwent
   three washes lasting for 10 min each using PBS. To facilitate
   visualization, a green-fluorescent Alexa Fluor 488 rabbit anti-mouse
   IgG (abcam, USA) (manufactured by Invitrogen) was utilized. After
   additional washing with PBS and subsequent staining with DAPI for a
   duration of 10 min, the cells were observed using a fluorescence
   microscope produced by Carl Zeiss (Germany).

Cellular thermal shift assay (CETSA)

   As previously mentioned, the interaction between Urolithin A and GPX4
   was investigated using CETSA [[69]21]. For the CETSA, the cell lysate
   was treated with DMSO or Urolithin A (10 µM) on a shaker at 4 °C for
   3 h. The mixtures were then aliquoted into separate tubes and heated
   for 3 min at various temperatures. Following cooling, the cell lysates
   were centrifuged at 12,000 rpm for 15 min at 4 °C, and the supernatants
   were collected for subsequent Western blot analysis.

Scanning electron microscope

   The morphology of Ti nanoparticles was observed with a FEI Quanta 600
   field-emission scanning electron microscope. The acceleration voltage
   of 5 kV was used for the imaging. Samples for scanning electron
   microscopy were prepared from Ti nanoparticles powders after drying
   overnight and gold plating. The particle size distribution and average
   particle size of Ti nanoparticles were counted using image J.

Bone tissue from clinical samples

   Bone tissues were procured from patients who underwent Total hip
   arthroplasty (THA). One group presented with developmental dysplasia of
   the hip (DDH) and underwent primary total joint arthroplasty (PTJA).
   The other group who underwent revised total joint arthroplasty (RTJA)
   for aseptic loosening. All patients underwent surgeries at the First
   Affiliated Hospital of Soochow University (Table [70]S1). This study
   has received approval from the esteemed Ethics Committee of the First
   Affiliated Hospital of Soochow University (2023107).

Animals and osteolysis induced by wear particles

   Ti nanoparticle-induced mice calvarial osteolysis were established
   according to the literature and were generated as later described
   [[71]22]. All animal experiments were approved by the Ethics Committee
   of Soochow University (SUDA20221009A06). According to the random number
   table, the 8-week-old female C57BL/6 mice were randomly divided into
   five groups: Control group, Ti group (surgery), Fer-1 group
   (Ti + 0.1 mg/kg Fer-1), Low-UA group (Ti + 10 mg/kg UA), and High-UA
   group (Ti + 20 mg/kg UA). Doses of UA and Fer-1 have no physiological
   toxicity in vivo [[72]23, [73]24]. The size of Ti nanoparticles ranged
   from 25.39 to 121.67 nm (Fig [74]S1A and B). After surgical anesthesia,
   the mouse head was disinfected with 75% ethanol and covered with a
   sterile towel. A sterile scalpel was used to cut the skin of the mouse
   cranial roof to expose the cranium and peel the fascia of the cranium
   surface. According to previous studies, 40 mg of Ti nanoparticles (40
   µL) were meticulously dispersed onto the calvariae surfaces, followed
   by the delicate application of sutures [[75]25]. Penicillin was
   administered 3 days after surgery to prevent infection. After surgery,
   all C57BL/6 mice were fed for 2 weeks in a specific, asymptomatic
   environment at room temperature and conventional humidity. The Fer-1
   group received a fixed dose of Fer-1 by intraperitoneal injection daily
   after surgery. The UA group received a fixed dose of UA by gavage daily
   after surgery. The control group received physiological saline. All
   mice were sacrificed 14 days after the surgical procedure, and their
   crania were collected and then preserved in a solution containing 4%
   paraformaldehyde.

Micro-computed tomography (micro-CT) analysis

   Cranial surface erosion was measured using micro-CT (SkyScan1176,
   Belgium). According to the manufacturer’s instructions, the scanning
   parameters for achieving isometric resolution are configured at 9 μm,
   while the X-ray energies are precisely set to 80 kV and 100 mA. 3D
   image reconstruction and quantitative analysis using SkyScan software.
   An equal area (3 mm in diameter) of area of interest (ROI) was found in
   the center of each cranium. We utilized SkyScan software to assess bone
   mineral density (BMD, mg/cm^3), bone volume (BV, mm^3), bone
   volume/tissue volume (BV/TV), and pores within the region of interest.

Histological and immunohistochemical analysis

   After undergoing decalcification for a period of 28 days in a solution
   containing 10% ethylene diamine tetraacetic acid (EDTA) obtained from
   Sigma (USA), the crania were embedded in molten paraffin. Afterwards,
   the tissue samples were cut into 6 μm sections and processed following
   the guidelines provided by the manufacturer for staining with
   hematoxylin and eosin (H&E). The resulting visuals were captured using
   a Carl Zeiss microscope (Germany).

   The assessment of dynamic bone formation was conducted using the
   calcein double labeling assay. In brief, a solution of calcein (Sigma,
   USA) was administered intraperitoneally in phosphate buffered saline
   (PBS) at a dosage of 10 mg/kg on the 1st and 8th days prior before
   sacrifice. The fluorescence microscope (Germany) was utilized to
   capture images of undecalcified bone sections labeled with calcein, and
   the calculation of mineral apposition rate (MAR) was performed using
   image J.

   Immunohistochemistry (IHC) staining was employed to detect the
   expression levels of GPX4. In brief, tissue sections were
   deparaffinized and rehydrated using a gradient method for antigen
   retrieval. Subsequently, GPX4 primary antibody (ab125066) from Abcam
   (USA) was applied to the tissue sections and incubated overnight at
   4 °C. Following that, corresponding secondary antibodies were used for
   blocking the sections for 30 min. The DAB Kit provided by Beyotime
   (China) was used to initiate the chromogenic reaction. Analysis was
   conducted on randomly selected areas throughout the entire field of
   view. Finally, fluorescence microscopy (Zeiss, Germany) was utilized to
   observe the tissue sections.

   Immunofluorescence staining was performed to ascertain the presence of
   GPX4, in accordance with established protocols and procedures. In
   brief, tissue sections were subjected to dewaxing, antigen retrieval,
   and blocking with 2% BSA for a duration of 60 min. Subsequently,
   primary antibodies targeting GPX4 (ab125066) from Abcam (USA) were
   added and incubated overnight at a temperature of 4 °C. Following this
   step, the corresponding secondary fluorescent antibodies 488 (Abcam,
   USA) were applied onto the sections in darkness for an hour.
   Afterwards, DAPI was utilized to counterstain the nuclei for a period
   of 10 min. Finally, observation of the tissue sections was performed
   using a fluorescence microscope manufactured by Zeiss (Germany).

Quantification and statistical analysis

   The data represent the mean ± standard deviation (SD). The statistical
   analysis was conducted using GraphPad Prism version 8.0 software. A
   Student’s t-test was employed to ascertain the presence of
   statistically significant disparities between two groups. The
   significant difference among multiple groups was accessed by one-way
   ANOVA where appropriate (*P < 0.05; **P < 0.01).

Results

Ti nanoparticles suppressed osteogenic differentiation

   In order to investigate the impact of wear particles on the suppression
   of osteoblasts, we have opted to utilize Ti nanoparticles as an
   intervention in osteoblasts to replicate the complex microenvironment
   of osteolysis in vitro (Fig. [76]S2A). The results of ALP staining and
   ARS staining demonstrated a significant decrease in the osteogenic
   activity and extracellular mineralization when treated with both the
   5 µg/cm^2 and 10 µg/cm^2 Ti nanoparticles (Fig. [77]1A-D). The
   expression level of osteogenic genes was found to be significantly
   reduced in both the low- and high-dose groups exposed to Ti
   nanoparticles (Fig. [78]1E). Similar results were also confirmed by
   western blot (Fig. [79]1F and G). These results showed that the
   presence of Ti nanoparticles had a significant inhibitory effect on
   osteogenic differentiation and the mineralization of the extracellular
   matrix.

Fig. 1.

   [80]Fig. 1
   [81]Open in a new tab

   Ti nanoparticles suppressed osteogenic differentiation. (A)-(B)
   Representative images of ALP staining and ARS staining. (scale bar
   = 100 μm). (C)-(D) Quantitative analysis of relative ALP activity and
   ARS recovery ratio. (E) The mRNA levels of osteoblast-specific genes
   with the intervention of Ti nanoparticles. (F) Protein levels of
   COL1-a1, Runx2 and Osterix. (G) Quantification of osteogenesis-related
   markers shown in (F). Data are mean ± SD, n = 3; *P < 0.05, **P < 0.01
   vs. Control. ns indicates not significant

Ferroptosis was involved in osteoblasts with the intervention of Ti
nanoparticles

   In order to investigate the underlying mechanism through which Ti
   nanoparticles exert detrimental effects on osteoblasts, we performed
   RNA sequencing to identify genes that displayed differential expression
   between the control and Ti groups. Total mRNA was extracted from
   osteoblasts following exposure to 10 µg/cm^2 Ti nanoparticles for a
   duration of 72 h. The Illumina HiSeq 2500 platform was employed to
   conduct RNA-seq analysis and generate mRNA profiles (Fig. [82]2A and
   Fig. [83]S3A). The volcano plot revealed that a total of 1762 genes
   were upregulated, while 1008 genes were downregulated after Ti
   nanoparticles treatment (Fig. [84]2B). The aforementioned findings were
   represented graphically in the form of a heat map, and then GO
   categories were used to annotate the 2770 differentially expressed
   genes (DEGs) (Fig. [85]2C, D and Fig. [86]S3D). The GO enrichment
   analysis demonstrated a significant enrichment of pathways associated
   with the transport and homeostasis of iron ions, which may potentially
   be linked to the occurrence of ferroptosis [[87]26]. In addition,
   pathway enrichment analysis was conducted using the Kyoto Encyclopedia
   of Genes and Genomes (KEGG) on the dataset obtained from RNA
   sequencing. Interestingly, the genes that exhibited significant
   differential expression led to a notable alteration in the ferroptosis
   pathway, as depicted in Fig. [88]2E and Figure [89]S3B. The Ti group
   exhibited a significant enhancement and stimulation of the ferroptosis
   pathway, which was verified through gene set enrichment analysis (GSEA)
   (Fig. [90]2F). Additionally, the induction of ferroptosis resulted in a
   notable decrease in the level of GPX4 and the expression of
   osteogenesis-related genes, including Col1-a1, Alp, Runx2, and Sp7.
   This confirmation was supported by the visually striking heat map and
   cluster analysis conducted on the gene set linked to ferroptosis and
   differentially expressed genes (DEGs) (Fig. [91]2G and Fig. [92]S3C)
   [[93]27]. Likewise, we observed that Ti nanoparticles suppressed
   osteoblast function to a degree comparable to the classical inducer of
   ferroptosis called erastin (Fig. [94]S4A-C). In summary, this RNA-seq
   analysis demonstrated that the inclusion of Ti nanoparticles has the
   capacity to elicit a reduction in osteogenic activity by triggering the
   onset of ferroptosis.

Fig. 2.

   [95]Fig. 2
   [96]Open in a new tab

   RNA sequencing of osteoblasts under the intervention of Ti
   nanoparticles. (A) A schematic diagram depicting the experimental
   design and protocol for sample preparation in RNA sequencing. (B) The
   volcano plot revealed differentially expressed genes. (C) Heat map
   depicts genes that were differentially expressed between the Control
   and Ti groups. Red: upregulated genes. Blue: downregulated genes. (D)
   Significant pathway of GO Enrichment analysis. (E) KEGG pathway
   enrichment analysis of the differentially expressed genes. (F) GSEA
   enrichment plot for the ferroptosis pathway based on RNA sequencing.
   (G) The heat map revealed the genes significantly regulated in relation
   to osteogenesis and ferroptosis

   Subsequently, the impact of Ti nanoparticles on ferroptosis were
   investigated in osteoblasts. The TEM analysis demonstrated a reduction
   or absence of mitochondrial ridges and disruption of the outer
   mitochondrial membrane (OMM) in osteoblasts after a 72-hour
   intervention with Ti nanoparticles, which serve as a critical indicator
   of ferroptosis (Fig. [97]3A). Simultaneously, the inclusion of Ti
   nanoparticles led to a notable elevation in ROS levels and a decline in
   the mitochondrial membrane potential, indicating an impairment in
   mitochondrial functionality (Fig. [98]3B). Ti nanoparticles stimulation
   resulted in a sharp increase of FerroOrange in the fluorescence
   intensity, which suggested that iron homeostasis was disrupted in
   osteoblasts, thus inducing ferroptosis (Fig. [99]3B). Likewise, the
   generation of MDA and the concentration of total iron (Fe) exhibited
   the identical increased trends (Fig. [100]3 Cand D). The results of
   western blot and RT-PCR indicated that with the increase of the
   concentration of Ti nanoparticles, the protein levels and gene levels
   of critical regulators significantly changed (Fig. [101]3E-G).
   Consistently, Ti nanoparticles significantly suppressed the content of
   GSH and the subsequent impact on the activity of GPX4, which was
   consistent with the results of clinical samples (Fig. [102]S5A). Taken
   together, the obtained detection results provide conclusive evidence
   supporting the correlation between Ti nanoparticles and osteoblastic
   ferroptosis.

Fig. 3.

   [103]Fig. 3
   [104]Open in a new tab

   Ti nanoparticles promote osteoblastic ferroptosis. (A) TEM images of
   osteoblasts after treatment with Ti nanoparticles (5 and 10 µg/cm^2).
   Scale bar = 5 μm (upper) and 500 nm (lower). (B) Representative images
   of Ti nanoparticle-induced ROS generation (scale bar = 100 μm),
   Mitochondrial membrane potential (scale bar: 50 μm) and FerroOrange
   staining (scale bar = 50 μm) after Ti nanoparticles treatment. (C) The
   level of MDA. (D) The content of cellular total iron. (E) The mRNA
   levels of ferroptosis-related genes. (F) Protein levels of 4HNE, COX2,
   SLC7A11 and GPX4 after the intervention of Ti nanoparticles. (G)
   Quantification of ferroptosis-related markers in (F). Data are
   mean ± SD, n = 3; *P < 0.05, **P < 0.01 vs. Control. ns indicates not
   significant

Osteoblastic ferroptosis was involved in peri-prosthesis osteolysis

   In order to investigate the potential relationship between PPO and
   ferroptosis, bone tissue samples were collected from the RTJA group,
   while bone tissue from the PTJA group was utilized as a control (Fig.
   [105]S6A). In the RTJA group, a notable decrease in the expression of
   markers associated with bone formation, such as COL1-a1, Runx2, and
   Osterix was observed when compared to the PTJA group (Fig. [106]4A and
   B). In the RT-PCR assay, a significant reduction in bone formation
   capacity was observed subsequent to the loosening of the prosthesis
   (Fig. [107]4E). The findings of our study aligned with prior research
   indicating a decline in bone formation capacity when a prosthesis
   becomes loose [[108]28]. Further investigations have uncovered notable
   alterations in the expression levels of crucial ferroptosis-related
   proteins, such as GPX4 and SLC7A11, in the RTJA group compared to the
   PTJA group. Specifically, the expression levels of GPX4 and SLC7A11
   were significantly downregulated, while COX2 expression was observed to
   be increased (Fig. [109]4C and D). The progressive modification in the
   expression of genes linked to ferroptosis offers additional
   confirmation of its potential role in inhibiting the osteogenic
   function during peri-prosthesis osteolysis (Fig. [110]4F). Meanwhile,
   the results of immunofluorescence staining demonstrated a significant
   decrease in the number of GPX4-positive cells observed on the surface
   of bone tissue in the RTJA group compared to the PTJA group. This
   finding suggests that GPX4 may play a critical role in the suppression
   of ferroptosis induced by peri-prosthetic osteolysis (Fig. [111]S6B).
   Additionally, significant increases in MDA levels, a byproduct of lipid
   oxidation, and total iron concentrations were observed in the RTJA
   group. These findings suggest the presence of iron accumulation and
   lipid peroxidation in the bone tissue during peri-prosthetic
   osteolysis, as depicted in Fig. [112]4G and H. The assessment of the
   content of GSH, which acts as the reducing substrate for GPX4 activity,
   was also conducted. The results revealed a significant decrease in the
   levels of GSH in the RTJA group compared to the PTJA group
   (Fig. [113]4I). Taken together, the aforementioned findings suggest
   that osteogenic ferroptosis plays a crucial role in the suppression of
   osteogenic function induced by wear particles.

Fig. 4.

   [114]Fig. 4
   [115]Open in a new tab

   Osteoblastic ferroptosis is involved in peri-prosthesis osteolysis. (A)
   Protein levels of COL-a1, Runx2 and Osterix in the PTJA and RTJA group.
   (B) Relative protein quantification of osteogenic markers in (A). (C)
   Protein levels of COX2, SLC7A11 and GPX4. (D) Relative protein
   quantification of ferroptosis-related markers in (C). (E) The mRNA
   levels of osteoblast-specific genes. (F) The mRNA levels of
   ferroptosis- related genes. (G) The level of MDA. (H) The content of
   total iron. (I) The content of GSH. Data are mean ± SD, n = 3;
   *P < 0.05, **P < 0.01 vs. RTJA. ns indicates not significant

Suppression of ferroptosis promoted osteogenic differentiation and alleviated
PPO

   Given the significant role of ferroptosis in the suppression of
   osteogenesis and the promotion of PPO, it is postulated that the
   inhibition of ferroptosis could potentially restore osteogenesis and
   mitigate PPO. Fer-1 was identified as an early synthetic
   radical-trapping antioxidant (RTA) with documented efficacy in
   inhibiting ferroptosis so that it was employed for the subsequent study
   [[116]29]. The CCK8 analysis was utilized to assess the impact of Fer-1
   concentration on osteoblast proliferation and viability (Fig.
   [117]S7A). Fer-1 was applied to osteogenic differentiated MC3T3-E1
   cells with 10 g/cm^2 Ti nanoparticles for 3 days. The western blot
   results demonstrated that Fer-1 treatment effectively inhibited the
   degradation of SLC7A11 and GPX4 and decreased the expression of COX2,
   indicating the inhibition of osteoblastic ferroptosis (Fig. [118]5A and
   B). Hereafter, ALP staining and ARS staining were used to verify the
   effects of Fer-1 on osteogenic function. As anticipated, Fer-1
   effectively restored the inhibited osteogenic cell function caused by
   Ti nanoparticles (Fig. [119]5C-F). The expression level of OCN was
   further investigated through immunofluorescence staining. Treatment
   with Ti (10 µg/cm^2) significantly reduced the average fluorescence
   intensity of OCN. However, Fer-1 exhibited a notable upregulation in
   OCN expression (Fig. [120]5G). In the meanwhile, the osteogenic-related
   markers were further detected by western blot and RT-PCR, and
   osteogenic-related markers such as COL1-a1, Runx2, and Osterix were
   significantly upregulated after the treatment of Fer-1 (Fig.
   [121]S7B-D). In conclusion, the application of Fer-1 demonstrated a
   significant ability to restore osteoblast differentiation by
   effectively suppressing ferroptosis in an in vitro setting.

Fig. 5.

   [122]Fig. 5
   [123]Open in a new tab

   Suppression of osteoblastic ferroptosis promoted osteogenic
   differentiation. (A) Protein levels of COX2, SLC7A11 and GPX4 after Ti
   nanoparticles (10 µg/cm^2) and Fer-1 (5µmol/L) treatment. (B)
   Quantification of ferroptosis-related markers in (A). (C)-(D)
   Representative images of ALP staining and ARS staining. (scale bar
   = 100 μm). (E)-(F) Quantitative analysis of relative ALP activity and
   ARS recovery ratio. (G) Illustrative images of immunofluorescence
   staining; green (OCN), red (F-actin) and blue (nuclei). (scale bar
   = 25 μm). Data are mean ± SD, n = 3; *P < 0.05, **P < 0.01 vs. Ti
   group. ns indicates not significant

   To provide additional clarity on the influence of Fer-1 on
   peri-prosthesis osteolysis, we established a calvarial osteolysis model
   by Ti nanoparticles (40 mg per mouse) and treated with Fer-1
   (0.1 mg/kg). As shown by the micro-CT and 3D reconstructed images in
   Fig. [124]6A, the Ti group induced an extensive fibrous layer
   containing a high concentration of inflammatory cells and Fer-1
   treatment significantly alleviated fibrous hyperplasia and bone
   resorption. Compared to the Ti group, the Fer-1 group showed higher
   bone morphometric parameters including BMD, BV as well as BV/TV
   (increased approximately 36.5%, 69.6% and 40.6%, respectively),
   indicating better osteogenesis after the Fer-1 treatment
   (Fig. [125]6B-D). A noticeable decrease in the number of pores was
   observed in the Fer-1 group, which was statistically significant
   (Fig. [126]6E). H&E staining confirmed the mitigation of osteolysis
   induced by Ti nanoparticles following the effective administration of
   Fer-1 (Fig. [127]6F). Additionally, the application of Fer-1 led to
   enhanced bone formation and a higher rate of mineral apposition, as
   demonstrated by the calcein double labeling (Fig. [128]6G). The changes
   were further confirmed by quantitatively analyzing the average
   thickness of mineral apposition (Fig. [129]S8A). Concomitantly, the
   Fer-1 group showed the increased number of GPX4-positive cells in the
   area of the osteolysis than the Ti group, indicating the suppression of
   osteoblastic ferroptosis (Fig. [130]6H and Fig. [131]S8B). The H&E
   staining of the major organs showed no obvious morphological
   abnormalities after Fer-1 treatment as shown in Fig. [132]S8C. These
   encouraging results indicate that the Fer-1 therapy is effective in
   alleviating the Ti nanoparticles induced osteolysis in vivo.

Fig. 6.

   [133]Fig. 6
   [134]Open in a new tab

   Fer-1 mitigated Ti nanoparticle-induced osteolysis. The concentration
   of Ti nanoparticles and Fer-1 in vivo were 40 mg per mouse and
   0.1 mg/kg, respectively. (A) Exhibited 3D reconstruction images of
   Micro CT. (B) BMD (mg/cm^3). (C) BV (mm^3). (D) BV/TV (%). (E) Number
   of pores. (F) Representative images of H&E staining. (scale bar
   = 100 μm and 25 μm). (G) Calcein double labelling. (scale bar = 25 μm).
   (H) Immunohistochemical analysis of GPX4 protein. (scale bar = 100 μm
   and 25 μm). Data are mean ± SD, n = 5; *P < 0.05, **P < 0.01 vs. Ti
   group. ns indicates not significant

Identification of natural small molecules targeting GPX4 to suppress
ferroptosis

   Our study has revealed the significant role of GPX4 and ferroptosis in
   inhibiting bone formation caused by Ti nanoparticles. Additionally,
   Fer-1 showed significant effectiveness in suppressing osteoblastic
   ferroptosis, which reduced the osteolytic effects caused by Ti
   nanoparticles and promoted osteogenesis. However, despite its
   widespread use in experiments, the clinical application of Fer-1 has
   been limited due to its instability. This includes issues such as
   hydrophobicity, inadequate pharmacokinetics, and unknown toxicity with
   prolonged in vivo use [[135]30, [136]31]. In recent years, there has
   been a significant increase in the demand for pharmaceuticals, which
   has led to a greater emphasis on the use and potential of natural
   products in drug research [[137]32]. Natural products offer several
   advantages, including enhanced stability, a wide range of
   physiologically active components, and the potential to avoid adverse
   effects associated with conventional chemical synthesis medications
   [[138]33]. The active site of the GPX4 protein was identified using the
   SiteMap module within the Schrödinger software package. Relevant drugs
   were then screened in the natural product database using a model of the
   GPX4 protein binding (Fig. [139]7A) [[140]34, [141]35]. Subsequently,
   we utilized the Canvas 1.1 program to filter out pan-assay interference
   compounds (PAINS) from the broad screening of interfering compounds. We
   employed the QikProp 3.2 program within the Schrödinger software
   package to calculate the pharmacokinetic properties of the compounds
   selected. Additionally, molecules obtained through various
   pharmacophore screenings and docking were merged, duplicate molecules
   were removed, and the remaining small molecules were subjected to
   structural clustering analysis using FCFP_6 fingerprints. Finally, the
   optimal docking conditions using the Glide force field were selected,
   conducting docking studies with the GPX4 protein as the receptor. An
   active pocket was defined with a size of 10 Å × 10 Å × 10 Å. We
   conducted a screening of Urolithin A, a natural product, taking into
   account its physicochemical properties, docking scores, multiple
   interactions, preservation challenges, and economic factors. This
   compound is a natural metabolite of ellagitannin, a group of compounds
   found in pomegranate, as well as in other fruits and nuts. UA has been
   shown to have properties that can reverse mitochondrial and muscle
   aging, as shown in Fig. [142]7B [[143]36]. The validation of the GPX4
   protein and UA was conducted using network pharmacology and molecular
   docking techniques. The appropriate conformation and spatial
   arrangement of UA demonstrated the steric complementarity between UA
   and the binding site of GPX4. The docking score between UA and GPX4 is
   -6.614. The binding mode revealed that UA binds the pocket of GPX4 by
   4 H-bond bindings: ASP-21 (2.2 Å), PHE-100 (3.1 Å), VAL-98 (3.2 Å),
   LYS-90 (2.1 Å) (Fig. [144]7C). All of these interactions were
   beneficial for stabilizing the binding complexes. To validate the
   virtual screening and molecular docking results, we conducted cellular
   thermal shift assays (CETSA). As shown in Fig. [145]7D and E, compared
   to DMSO alone, GPX4 showed markedly enhanced thermal stability
   following treatment with UA. In conclusion, these findings suggest that
   UA is capable of effectively targeting GPX4 to exert its functions.

Fig. 7.

   [146]Fig. 7
   [147]Open in a new tab

   Identification of natural small molecules targeting GPX4 to suppress
   ferroptosis. (A) The active site of the GPX4 protein. (B) Screening by
   GPX4 protein to obtain a 2D map of urolithin A small molecules and
   forces of action. (C) The binding mode of GPX4 with UA. H-bond
   (yellow): ASP-21 (2.2 Å), PHE-100 (3.1 Å), VAL-98 (3.2 Å), LYS-90 (2.1
   Å). (D) Cellular thermal shift assay between UA and GPX4. (E) The
   levels of GPX4 protein under different temperatures in the presence of
   UA and DMSO treatment. (F) Representative TEM images of osteoblasts
   after UA treatment. Scale bar = 5 μm (upper) and 500 nm (lower). (G)
   The content of MDA. (H) Representative images of Ti
   nanoparticles-induced ROS generation (scale bar = 100 μm),
   Mitochondrial membrane potential (scale bar = 50 μm) and FerroOrange
   staining (scale bar = 50 μm) after UA treatment. (I) The content of
   cellular total iron. (J) Protein levels of COX2, SLC7A11, GPX4. (K) The
   mRNA levels of ferroptosis-related genes. Data are mean ± SD, n = 3;
   *P < 0.05, **P < 0.01 vs. Ti group. ns indicates not significant

   The cytotoxicity and proliferative toxicity of UA were initially
   assessed using the CCK-8 assay. The results indicated that the
   concentration of 20 µmol/L UA did not exhibit significant cytotoxic
   effects, which was used in the following study to investigate the
   effects of UA on osteoblastic ferroptosis and osteogenesis (Fig.
   [148]S9A). Mitochondrial ultrastructure examined by TEM revealed that
   the mitochondrial morphology in the UA group showed a significant
   improvement compared to the group exposed to Ti nanoparticles,
   specifically characterized by a reduction in mitochondrial
   fragmentation (Fig. [149]7F). Similarly, UA treatment effectively
   suppressed the generation of MDA induced by Ti nanoparticles, exerting
   a potent inhibitory effect on ferroptosis (Fig. [150]7G). Furthermore,
   the staining of MMP, ROS, and FerroOrange was performed to provide
   additional evidence of the impact of UA on osteoblastic ferroptosis.
   These results indicated that UA treatment resulted in an elevation of
   MMP levels, while concurrently decreasing the production of Fe^2+ and
   ROS, which is comparable to the role of the ROS inhibitor NAC
   (Fig. [151]7H and Fig. [152]S9D). Similar patterns were also noted in
   the assessment of the overall iron content within the cells
   (Fig. [153]7I). The expression levels of ferroptosis-related markers
   were also examined following the treatment of UA. Western blot assays
   revealed that UA not only promoted protein level of GPX4 and SLC7A11,
   but also reduced the expression level of COX2 (Fig. [154]7J and
   [155]S9B). The mRNA levels of Cox2, Slc7a11 and Acsl4 and the
   quantification of GSH demonstrated the consistent expression as
   expected (Fig. [156]7K and Fig. [157]S9C). Likewise, we observed that
   UA suppressed osteoblastic ferroptosis following intervention with Ti
   nanoparticles to a degree comparable to the ROS inhibitor NAC (Fig.
   [158]S9E-G). Additionally, Erastin, the classical ferroptosis inducer,
   induced osteoblastic ferroptosis was also inhibited by UA treatment
   (Fig. [159]S10).

   To provide additional evidence regarding that UA inhibited ferroptosis
   rely on the upregulation of GPX4, specific siRNA was employed to
   effectively downregulate GPX4 expression (Fig. [160]8A). As expected,
   the protective effects of UA on Ti nanoparticles induced ferroptosis
   are markedly attenuated by GPX4 silence (Fig. [161]8B-D). The results
   showed that UA-activated GPX4 suppressed ferroptosis.

Fig. 8.

   [162]Fig. 8
   [163]Open in a new tab

   Silencing GPX4 reversed UA inhibition of ferroptosis. (A) The mRNA
   level of Gpx4 after GPX4 silence. (B) The mRNA levels of
   ferroptosis-specific genes after GPX4 silence in each group. (C)
   Protein levels of COX2, SLC7A11 and GPX4 after GPX4 silence. (D)
   Quantification of ferroptosis-related markers shown in (C). Data are
   mean ± SD, n = 3; *P < 0.05, **P < 0.01 vs. Ti + siGPX4 group. ns
   indicates not significant

UA promoted osteoblast differentiation and attenuated Ti nanoparticle-induced
osteolysis by the activation of GPX4

   Based on the above results, we further evaluated the effect of UA on Ti
   nanoparticle-induced inhibition of osteogenic function. The ALP and ARS
   staining results provide direct evidence that UA had a significant
   impact on the restoration of osteoblast activity impaired by Ti
   nanoparticles, as well as the promotion of extracellular matrix
   mineralization (Fig. [164]9A-D). Subsequently, the expression of
   osteogenic indicators at the protein and gene levels was detected in
   the condition with or without UA treatment. We found that UA treatment
   significantly promoted the expression of osteogenic
   differentiation-related proteins, including COL1-a1, Runx2, and Osterix
   and the relative gene expression in a dose-dependent manner
   (Fig. [165]9E-G). Meanwhile, the results of cell immunofluorescence
   staining further supported the promoting effect of UA on osteogenesis
   (Fig. [166]9H). However, these positive effects were impaired by GPX4
   silence (Fig. [167]9I-L). Taken together, the above results
   demonstrated that UA could alleviate the inhibition of osteogenesis
   induced by Ti nanoparticles, and this effect was dependent on the
   activation of GPX4.

Fig. 9.

   [168]Fig. 9
   [169]Open in a new tab

   UA promoted osteoblast differentiation by the activation of GPX4.
   (A)-(B) Representative images of ALP and ARS staining during Ti
   nanoparticles and UA treatment. (scale bar = 100 μm). (C)-(D)
   Quantitative analysis of ALP activity and ARS recovery ratio. (E) The
   mRNA levels of osteogenesis-related gene after UA treatment. (F)
   Protein levels of COL-a1, Runx2 and Osterix. (G) Quantification of
   osteogenic markers shown in (F). (H) Cellular immunofluorescence of
   OCN. (scale bar = 25 μm). Data are mean ± SD, n = 3; *P < 0.05,
   **P < 0.01 vs. Ti group. ns indicates not significant. (I)-(J)
   Representative images of ALP and ARS staining after GPX4 silence.
   (scale bar = 100 μm). (K)-(L) Quantitative analysis of ALP activity and
   ARS recovery ratio. Data are mean ± SD, n = 3; *P < 0.05, **P < 0.01
   vs. Ti + UA + siGPX4 group. ns indicates not significant

   In view of the encouraging bone-promoting effect of UA in vitro, we
   then explored whether UA could reduce Ti nanoparticle-induced
   osteolysis. As shown by the micro-CT scanning and 3D reconstruction
   results in Fig. [170]10A, the bone destruction of osteolysis improved
   significantly after the UA treatment in a dose-dependent manner. The
   bone morphometric parameters observed in the UA group exhibit
   superiority over those of the Ti group (Fig. [171]10B-E). In addition
   to micro-CT, histological analysis serves to validate the effects of UA
   in mitigating osteolysis. As depicted in Fig. [172]10F, an extensive
   fibrous layer containing a high concentration of inflammatory cells is
   observed in the Ti group. In contrast, the UA group exhibited a
   significant increase in bone mass and a notable decrease in fibrous
   hyperplasia and bone destruction (Fig. [173]10F). The dynamic
   osteogenesis was found to be higher than that in the Ti group, and it
   was even comparable to the rate observed in the Control group
   (Fig. [174]10G and Fig. [175]S11A). On the other hand, Ti nanoparticle
   significantly reduced the level of GPX4 in bone destruction areas,
   while UA treatment effectively upregulated the expression of GPX4 to
   resist ferroptosis for osteoblasts (Fig. [176]10H and Fig. [177]S11B).
   The histological examination of major organs in the experimental mice
   revealed no significant morphological abnormalities among the different
   groups (Fig. [178]S11C). These results provide evidence that the UA
   treatment reduces further bone destruction induced by Ti nanoparticle
   via the suppression of ferroptosis and promotion of osteogenesis.

Fig. 10.

   [179]Fig. 10
   [180]Open in a new tab

   UA attenuated Ti nanoparticle-induced osteolysis. (A) Representative 3D
   reconstruction images of micro-CT. (B) BMD (mg/cm^3). (C) BV (mm^3).
   (D) BV/TV (%). (E) Number of pores. (F) Representative images of H&E
   staining. (scale bar = 100 μm and 50 μm). (G) Calcein double labelling.
   (scale bar = 25 μm). (H) Immunohistochemical analysis of GPX4 protein.
   (scale bar = 100 μm and 50 μm). Data are mean ± SD, n = 5; *P < 0.05,
   **P < 0.01 vs. Ti group. ns indicates not significant

Discussion

   Total joint arthroplasty is widely recognized as the most advanced
   surgical treatment for trauma, degenerative arthritis, and other severe
   joint conditions [[181]37]. However, the presence of wear particles,
   such as Ti nanoparticles, can lead to peri-prosthesis osteolysis and
   subsequent aseptic loosening, which can cause total joint arthroplasty
   to fail. This not only creates a significant economic burden but also
   has social implications [[182]38, [183]39]. Previous studies have
   primarily focused on examining the effects of wear particles on
   osteoclast stimulation. However, these studies have overlooked the
   important role of osteoblasts in bone formation [[184]40, [185]41].
   Ameliorating the inhibitory effects of wear particles on osteogenic
   differentiation may represent a crucial factor for the therapeutic
   potential of PPO. The study found that the presence of Ti
   nanoparticle-induced osteoblastic ferroptosis by the inhibition of
   GPX4. Fer-1 reversed this process and restored osteogenic function to
   protect against bone loss in vitro and vivo. Considering the
   limitations of Fer-1 in the clinic, a natural compound known as UA was
   screened using network pharmacology for its specific targeting of GPX4.
   Subsequently, it was further confirmed that UA alleviated Ti
   nanoparticle-induced osteoblastic ferroptosis and promoted osteoblast
   differentiation, but this effect was inhibited by silencing GPX4. In
   addition, UA effectively alleviated osteolysis-induced bone loss in the
   mouse model. This study validated the feasibility of osteolysis therapy
   via the suppression of osteoblastic ferroptosis and the results provide
   insights into drug development for future clinical applications.

   Ferroptosis is a form of regulated cell death that is mediated by iron,
   which occurs when lipid peroxides accumulate on cellular membranes,
   resulting in a toxic cellular environment [[186]42]. Several studies
   have suggested a potential link between ferroptosis and various
   diseases, including cancer, neurodegenerative diseases, acute kidney
   injury, and ischemia reperfusion [[187]43–[188]47]. Recent studies have
   shown that ferroptosis has an impact on orthopedic diseases, including
   osteoporosis and PPO, which leads to a decrease in osteogenic function
   and impairs the ability to form bone [[189]13, [190]48]. However,
   further investigation is necessary to understand the underlying
   mechanisms and identify the optimal targets for clinical applications.
   Our study analysed experimental data from clinical specimens and in
   vitro experiments, revealing a strong correlation between stimulation
   of osteoblastic ferroptosis by Ti nanoparticles and the development of
   periprosthetic osteolysis. RNA Seq analysis revealed that wear
   particles downregulated the expression of GPX4 and osteogenesis-related
   genes in osteoblasts, resulting in ferroptosis and a decrease in bone
   formation. Our findings demonstrate the effectiveness of Fer-1, a
   classic iron death inhibitor, in countering the induction of
   ferroptosis by Ti nanoparticles and promoting bone formation both in
   vivo and in vitro. The study provides evidence that wear particles
   induce osteoblast ferroptosis in PPO. The presence of Ti nanoparticles
   in peri-prosthetic osteolysis has been found to trigger osteoblastic
   ferroptosis, which reduces the ability of bone formation.

   The small molecule Fer-1 provides cellular protection against damage
   caused by lipid peroxides. It achieves this by effectively capturing
   free radicals and stabilizing lipid peroxides, which in turn prevents
   further oxidative reactions [[191]49]. Additionally, Fer-1 has been
   found to regulate intracellular iron ion metabolism. This compound
   effectively inhibits the accumulation and release of iron ions,
   preventing excessive free iron ions from participating in oxidative
   reactions and reducing the production of oxidative stress [[192]50]. In
   their research on ferroptosis in orthopedic diseases, Xu et al. found
   that Fer-1 can inhibit osteoblastic ferroptosis and improve
   peri-prosthetic osteolysis induced by CoCrMo nanoparticles [[193]13]. A
   recent study has shown that Fer-1 has the ability to slow down the
   progression of osteoporosis by inhibiting ferroptosis in osteoblasts
   and suppressing senescence markers associated with osteoblasts. This
   helps to prevent age-related osteoporosis [[194]16]. Our study has
   confirmed the important role of Fer-1 in promoting osteogenesis and
   inhibiting ferroptosis both in vitro and in vivo. However, despite the
   overwhelming evidence of its usefulness, Fer-1 has been limited in its
   clinical application because it is unstable in vivo [[195]30, [196]51].

   We discovered a natural product called Urolithin A through virtual
   screening of the GPX4 protein, which is known as a key target of
   ferroptosis. UA, a component of the ellagitannin family, has shown its
   potential to improve metabolic health in humans and rodents. It has
   also been found to increase the lifespan of worms by supporting
   mitophagy and promoting optimal mitochondrial function [[197]52,
   [198]53]. Previous studies have shown that UA has a protective effect
   against ferroptosis in acute lung injury. It achieves this by
   increasing the expression of important proteins such as GPX4 and
   SLC7A11, as well as activating the Keap1-Nrf2/HO-1 signaling pathway
   [[199]54]. UA also demonstrates significant therapeutic potential in
   the prevention of diabetes mellitus (DM)-related Alzheimer’s disease
   (AD) pathology, which can be realized by inhibiting TGM2-mediated MAM
   assembly and maintaining mitochondrial calcium homeostasis as well as
   reactive oxygen species (ROS) stability [[200]18, [201]55].
   Furthermore, under the potential of a functional polymer-lipid drug
   delivery system, this lays the groundwork for further investigation
   into the immunotherapeutic potential of UA for chemotherapy-induced
   acute kidney injury in cancer patients [[202]56–[203]58]. After
   examining the role of UA and GPX4 proteins in ferroptosis, we
   discovered that UA specifically binds to the GPX4 protein through
   molecular docking. Our study demonstrated the effectiveness of UA in
   enhancing the expression and activity of GPX4, as evidenced by western
   blot and RT-PCR results. This highlights the significant potential of
   UA.

   In recent years, UA has emerged as a crucial factor in the management
   of various orthopedic diseases. Our previous study showed that UA
   inhibits osteoclastogenesis and postmenopausal osteoporosis by
   suppressing inflammation and NF-κB-mediated pyroptotic pathways
   [[204]23]. Recent studies have found that UA effectively inhibits the
   harmful effects of inflammation-induced bone loss and specifically
   targets the PI3K/Akt/NF-κB pathway. This provides relief for
   individuals with osteoarthritis [[205]59]. In this study, we found that
   UA effectively suppressed ferroptosis in osteoblasts while also
   enhancing their osteogenic function in vitro. Additionally, our in vivo
   experiments provided evidence that UA promotes GPX4 expression and
   reduces periprosthetic osteolysis by suppressing ferroptosis. The
   compound demonstrates a therapeutic effect similar to Fer-1, making it
   a promising treatment option for PPO.

Conclusion

   Osteoblastic ferroptosis was found to contribute to the pathogenesis of
   PPO. The specific inhibitor of ferroptosis, Fer-1 promoted the
   osteoblast susceptibility to ferroptosis by elevating the expression of
   GPX4 and rescue their osteogenic function thus halting the progression
   of peri-prosthesis osteolysis. A small molecule, UA, targeting GPX4 was
   found via network pharmacology and was confirmed to mitigate Ti
   nanoparticle-induced osteoblastic ferroptosis and impaired
   osteogenesis. In animal model, UA exerted a therapeutic impact on
   osteolysis induced by Ti nanoparticles. In conclusion, this study
   provides a new insight into osteolysis and a novel therapeutic approach
   for PPO.

Electronic supplementary material

   Below is the link to the electronic supplementary material.
   [206]Supplementary Material 1^ (7.3MB, docx)

Acknowledgements