3D-printed manganese dioxide incorporated scaffold promotes osteogenic-angiogenic coupling for refractory bone defect by remodeling osteo-regenerative microenvironment

Abstract

The treatment of refractory bone defects is a major clinical challenge, especially in steroid-associated osteonecrosis (SAON), which is characterized by insufficient osteogenesis and angiogenesis. Herin, a microenvironment responsiveness scaffold composed of poly-L-lactide (PLLA), and manganese dioxide (MnO2) nanoparticles is designed to enhance bone regeneration by scavenging endogenous reactive oxygen species (ROS) and modulating immune microenvironment in situ. A catalase-like catalytic reaction between MnO2 and endogenous hydrogen peroxide (H2O2) generated at the bone defect area, which typically becomes acidic and ROS-rich, triggers on-demand release of oxygen and Mn2+, significantly ameliorating inflammatory response by promoting M2-type polarization of macrophages, reprograming osteoimmune microenvironment conducive to angiogenesis and osteogenesis. Furthermore, the fundamental mechanisms were explored through transcriptome sequencing analysis, revealing that PLLA/MnO2 scaffolds (PMns) promote osteogenic differentiation by upregulating the TGF-β/Smad signaling pathway in human bone marrow mesenchymal stem cells (hBMSCs). Overall, the PMns exhibit superior immunomodulatory, excellent osteogenic-angiogenic properties and promising candidates as bone graft substitutes for therapy clinical refractory bone defects.

Headnote

ABSTRACT

The treatment of refractory bone defects is a major clinical challenge, especially in steroid-associated osteonecrosis (SAON), which is characterized by insufficient osteogenesis and angiogenesis. Herin, a microenvironment responsiveness scaffold composed of poly-L-lactide (PLLA), and manganese dioxide (MnO2) nanoparticles is designed to enhance bone regeneration by scavenging endogenous reactive oxygen species (ROS) and modulating immune microenvironment in situ. A catalase-like catalytic reaction between MnO2 and endogenous hydrogen peroxide (H2O2) generated at the bone defect area, which typically becomes acidic and ROS-rich, triggers on-demand release of oxygen and Mn2+, significantly ameliorating inflammatory response by promoting M2-type polarization of macrophages, reprograming osteoimmune microenvironment conducive to angiogenesis and osteogenesis. Furthermore, the fundamental mechanisms were explored through transcriptome sequencing analysis, revealing that PLLA/MnO2 scaffolds (PMns) promote osteogenic differentiation by upregulating the TGF-β/Smad signaling pathway in human bone marrow mesenchymal stem cells (hBMSCs). Overall, the PMns exhibit superior immunomodulatory, excellent osteogenic-angiogenic properties and promising candidates as bone graft substitutes for therapy clinical refractory bone defects.

ARTICLE INFO

Keywords:

Osteogenic-angiogenic coupling

Steroid-associated osteonecrosis

Manganese dioxide

Refractory bone defect

(ProQuest: ... denotes formulae omitted.)

1. Introduction

Steroid-associated osteonecrosis (SAON) is a challenging orthopedic disease that arises following long-term and/or high-dose of corticosteroid treatment for infectious diseases (e.g., Corona Virus Disease 2019 [COVID-19], Severe Acute Respiratory Syndrome [SARS]) and rheumatoid diseases (e.g., Ankylosing Spondylitis [AS], rheumatoid arthritis) [1,2]. Recently, the morbidity of SAON has risen annually, becoming the leading cause of non-traumatic osteonecrosis [3]. Core decompression (CD) surgery can slow the progression of osteonecrosis by removing necrosis bone tissue and reducing intraosseous pressure in the early stage of SAON [4,5]. Bone defects caused by CD surgery are considered a type of refractory bone defect that requires filling to enhance the mechanical strength of the surgical site and prevent bone collapse [6]. Refractory bone defect typically refers to a bone defect that is difficult to heal or repair through conventional treatment methods due to its severity, size, or location. These defects can arise from various causes, including trauma, tumor resection, infection, or diseases like osteoporosis and ostenecrosis [7]. Due to their challenging nature, these bone defects often require specialized treatment strategies, such as autologous or allogeneic bone grafting, biomaterial scaffolds, or advanced regenerative medical technologies to promote bone regeneration and repair [8]. Autogenous bone graft is considered the gold standard after CD surgery. However, its application is limited by issues such as donor scarcity, secondary damage, and immune rejection [9]. To overcome these limitations, 3D-printed scaffolds have attracted significant attention among researchers due to its exceptional capability for personalized customization and high precision [10–12]. These scaffolds provide several advantages for facilitating bone regeneration, including appropriate mechanical strength, moderate biodegradability, excellent osteoconductivity and osteoinductivity [13]. Despite these advantages, challenges associated with pathophysiological microenvironment continue to hinder bone regeneration. In SAON, bone regeneration process is delayed due to lipid metabolism abnormalities, inadequate blood supply, intravascular coagulation, and apoptosis of osteoblasts, with the primary cause being intraosseous ischemia from vascular impairment caused by allergic vasculitis prior to steroid administration [14]. With a deeper understanding of bone regeneration, it has become increasingly clear that the combination of angiogenesis and osteogenesis in necrotic zones offer the most promising treatment for SAON-related bone defects [15].

The necrotic defect zone is characterized by hypoxia due to impaired intraosseous vasculature, which compromises cell viability and proliferation, leading to high concentrations of reactive oxygen species (ROS), primarily consisting of hydroxyl radicals (HO•), superoxide anions (O2 − ), and H2O2 [16,17]. Excessive ROS can induce apoptosis in osteogenic precursor cells and mature osteoblasts while simultaneously promoting the proliferation and activity of osteoclast progenitor cells [18]. Furthermore, the generation of osteoclast and macrophage colony-stimulating factor (M-CSF) further elevates the ROS levels, prolonging the inflammatory stage and vasculopathy [19]. Consequently, bone regeneration is entrapped in a vicious cycle characterized by an unfavorable microenvironment for new bone formation and dysregulated osteoimmune homeostasis.

Refractory bone defects trigger a complex immune responses, in which aberrant or prolonged immune responses hinder bone regeneration, primarily due to inadequate immune regulation [20,21]. After bone defects, macrophages are recruited and polarized to the M1 phenotype, secreting pro-inflammatory factors such as tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6), monocyte chemotactic protein-1 (MCP-1), and interleukin-1β (IL-1β). These cytokines trigger the impairment of adjacent bone tissue. Alternatively, macrophages can polarize to the M2 phenotype, secreting anti-inflammatory factors such as interleukin-4 (IL-4), interleukin-10 (IL-10) and transforming growth factor-beta (TGF-β), which alleviate inflammatory response and promote osteogenesis and angiogenesis [22]. Therefore, regulating M1/M2 phenotype transformation to inhibit inflammatory response in the bone defect area is an effective strategy to promote bone regeneration.

MnO2 can effectively modulate the pathological microenvironment of SAON and demonstrates a "Three-in-one" function in SAON bone repair [23]. (i) MnO2 has ROS scavenging capacity, effectively decomposing H2O2 into oxygen (O2) and H2O, which alleviates hypoxic microenvironment, protects the antioxidant enzyme system, and prevents oxidative stress damage to cells [24]. (ii) Manganese ions (Mn2+) released from MnO2 degradation act as immune stimulator and participate in various physiological processes, including immune response [25]. (iii) Mn2+ has been reported to enhance osteoblast adhesion and viability and increase the expression of osteogenic-related gene in various manganese-containing biomaterials [26]. Additionally, Mn2+ inhibits the function and formation of osteoclasts in osteoporotic rats [27]. Previous studies have shown that MnO2-based biomaterials can facilitate bone tissue repair [28,29], but there is limited knowledge on whether MnO2 regulate osteoimmune microenvironment to promote osteogenic-angiogenic coupling in SAON bone defect. Thus, the use of MnO2 in SAON bone defects may help decrease ROS levels, supply O2, and regulate the osteoimmune microenvironment to enhance osteogenesis and angiogenesis, thereby aiding in the repair of clinical refractory bone defects.

Bone graft substitutes used to fill osteonecrosis areas after CD for treating SAON require excellent bioactive properties and appropriate mechanical strengths. PLLA is promising biodegradable polymer with compressive strength (2–39 MPa) comparable to natural cancellous bone and has been approved for clinical application [30]. However, PLLA is only osteoconductive and lacks osteoinductivity properties, necessitating scaffold modification [31]. Low-temperature rapid prototyping (LT-RP) 3D printing technology precisely controls the deposition of solution-based materials point-by-point through jetting or extrusion into a low-temperature environment. This process allows for rapid freezing and solidification, accompanied by phase separation to form micropores. Subsequent freeze-drying enables solvent evaporation and scaffold solidification. Compared to other printing techniques, the printing process does not require reaction initiators and high temperature that might decrease the bioactivity and biocompatibility of the implants [32]. In this study, we used MnO2 nanoparticles as an effector particle combined with PLLA as substrate material to fabricate PMns, using LT-RP 3D printing technology. Our findings demonstrated that PMns significantly scavenged ROS and suppressed inflammatory response by regulating macrophage polarization. Recognizing the crucial roles of scavenging ROS and immunoregulation in bone generation, we implant PMns into femur condylar defect tunnel of rats with SAON to investigate their effects on osteogenesis, angiogenesis, and osteoimmune regulation. This study provides evidence that PMns exhibit excellent osteogenic-angiogenic coupling effects and osteoimmunomodulatory functions conducive to refractory bone defects repairing (Scheme 1).

2. Results

2.1. Structure and mechanical properties of PMns

We used medical-grade degradable PLLA as matrix and MnO2 nanoparticles as effector particles to fabricate PMns with varying MnO2 contents (0 % PLLA, 2.5 % P2.5Mn, 5 % P5Mn, 10 % P10Mn, 20 % P20Mn) using low-temperature deposition 3D printing technology (Fig. 1A, Fig. S1A). Micro-CT images demonstrated that the scaffolds exhibited interconnected porous structures in both horizontal and vertical orientations, with high interconnectivity between the pores (Fig. 1B–D). Additionally, we used scanning electron microscopy (SEM) to assess macrostructure and microstructure of the scaffolds, revealing macropores ranging from 400 to 600 μm and micropores from 5 to 40 μm in the transverse sections (Fig. 1E–G, Fig. S1B). SEM analysis verified the nanoscale size (30–100 nm in diameter) of the MnO2 particles (Fig. S2), while SEM and energy dispersive X-ray spectroscopy (EDS) of the scaffold cross-sections confirmed homogeneous distribution of MnO2 nanoparticles (Fig. 1H–K). The mechanical strength of the PMns was evaluated using compression test. The results showed that the compression strength of PMns (32.19 ± 0.82 MPa for P2.5Mn, 37.83 ± 2.28 MPa for P5Mn, 35.99 ± 3.00 MPa for P10Mn, 42.61 ± 2.79 MPa for P20Mn) was higher than that of PLLA (30.61 ± 0.89 MPa), and the compression modulus of PMns (1.91 ± 0.18 MPa for P2.5Mn, 2.55 ± 0.29 MPa for P5Mn, 2.78 ± 0.51 MPa for P10Mn, 2.59 ± 0.28 MPa for P20Mn) was also superior to that of PLLA (1.77 ± 0.08 MPa) scaffolds (Fig. 1L and M). The degradation characteristics of scaffolds in PBS were investigated, revealing a sustained release of manganese throughout the degradation process (0.53 ± 0.02 ppm for P10Mn after 12 weeks), with the pH of the solution maintained at approximately 7.4. Additionally, the weight loss of the different scaffolds (4.3 ± 0.33 % for P10Mn after 12 weeks) was measured at various degradation time points (Fig. 1N–P). Additionally, we investigated the capability of scaffolds to scavenge ROS. PMns can decompose H2O2, slowly generating O2 under acidic conditions (Fig. 1Q, Fig. S3). We used an H2O2 kit to test the ability of the CAT-enzyme. The results indicated that PMns exhibits peroxidase activity and can effectively scavenge H₂O₂ (Fig. S4A). To confirm the release of Mn ions from PMns under acidic conditions, we placed PMns in an acidic solution with a pH of 6.8 and measured the concentration of Mn ions. Compared to the normal PBS solution, PMns released Mn ions more efficiently under acidic conditions. The results indicate that the PMns can continuously and stably release Mn ions in the acidic microenvironment of bone defects (Fig. S4B).

2.2. Biocompatibility of PMns in vitro

The biocompatibility of the scaffolds was evaluated by detecting cell viability, proliferation, adhesion, and morphology. hBMSCs were seeded on various scaffolds for 7 days, and laser scanning confocal microscope analysis of the cytoskeleton staining (Fig. S5A) indicated that all scaffolds provided an optimal environment for cell attachment, with similar spreading of the hBMSCs. After 7 days of co-culture, the proliferation of hBMSCs in the leaching solution of different scaffolds was measured at various time points by live/dead staining (Fig. S5B) and CCK-8 cell counting method (Fig. S5C), demonstrating that the addition of varying amounts of MnO2 nanoparticles to pure PLLA scaffold did not induce toxicity in hBMSCs. These findings confirm that the PMns exhibit excellent biocompatibility and provide a favorable environment for cell growth.

2.3. PMns protect cells from oxidative damage by scavenging ROS

The antioxidant defense capacity of scaffolds was assessed by simulating an oxidative stress microenvironment with H2O2 and observing the response of hBMSCs and HUVECs on various scaffolds (Fig. 2A). After 24 hours of exposure to H2O2, live/dead staining (Fig. 2B and C, Fig. S6A) and ROS level testing (Fig. 2D and E) in hBMSCs suggested that manganese dioxide-containing scaffolds significantly reduced cell death and scavenged intracellular excessive ROS. Furthermore, flow cytometry demonstrated that composite scaffolds improved the viability of hBMSCs under oxidative stress conditions (Fig. 2F, Fig. S6B). Similarly, live/dead staining and DCFH-DA staining in HUVECs (Fig. 2G and H, Fig. S6C) were significantly lower in the PMns groups compared to the PLLA or Blank groups treated with H2O2. Meanwhile, the wound healing assay showed that HUVECs maintained normal function and exhibited better migration in the PMns groups (403.83 ± 8.49 μm for P10Mn at 24 hours) (Fig. 2I). Quantitative analysis of dead/live ratio, DCFH-DA staining and migration distance of HUVECs (Fig. 2J, K and L) demonstrated that ROS scavenging capability of P10Mn in HUVECs under oxidative stress conditions. To elucidate the mechanism of ROS scavenging by PMns, we analyzed the expression of antioxidant genes in hBMSCs and HUVECs (Fig. 2M, Fig. S6D-F). The findings showed a significant increase in the expression of antioxidant genes after cocultured with PMns, suggesting that manganese dioxide-containing scaffolds effectively reduce intracellular ROS levels and mitigate oxidative damage by up-regulating antioxidant genes such as SIRT1, SOD2, and CAT, thereby restoring normal cell function (Fig. 2N).

2.4. Pro-osteogenic properties of PMns and transcriptomic analysis in vitro

To assess the pro-osteogenic properties of PMns on hBMSCs in vitro, several assays were performed including ALP staining, ALP activity detection, ARS, and the expression of osteogenic genes. After 7 days of osteogenic induction, the PMns groups showed significantly higher ALP staining and activity compared to the PLLA, Control, and Blank groups. The results between the P10Mn and P20Mn groups were comparable (Fig. 3A and B). After 14 days, extensive calcium deposition was observed in the PMns groups, almost completely covering the surface with calcium nodules, showing the same trend as the ALP results (Fig. 3C and D). Based on the combined mechanical and osteogenic results, the P10Mn group was selected for subsequent experiments. qRT-PCR data suggested that the P10Mn group significantly increase the expression of osteogenesis-specific genes compared to the PLLA group (Fig. 3E–H).

To further explore the underlying biological mechanisms of PMns in promoting osteogenic differentiation of hBMSCs, we conducted RNASeq analysis 14 days after osteogenic induction using the Illumina NovaSeq 6000 sequencing platform. We sequenced the P10Mn group and the osteogenic induction without scaffolds as Control groups. The volcano plots of all expressed genes (Fig. 3I) revealed a cluster of differentially expressed genes, with 529 genes upregulated and 680 genes downregulated (|log2(FC)|≥1 and P ≤ 0.05). Gene Ontology (GO) annotations highlighted that several osteogenic-related genes, such as NRG1, GBP4, TCF7, COL10A1, BMP2, BMP6, were upregulated in P10Mn group (Fig. 3J), consistent with the in vitro osteogenesis experiments. Furthermore, KEGG enrichment analysis revealed that differentially expressed genes were mainly involved in cytokinecytokine receptor interaction, the MAPK signaling pathway, TNF signaling pathway, and TGF-β signaling pathway (Fig. 3K), which are involved in the regulation of osteogenesis during differentiation. Additionally, Western blot results indicated significant upregulation of TGFβRI, TGF-βRII, p-Smad2/3, Smad4, Runx2, and Osterix in the P10Mn group compared to Control group (Fig. 3L). These results indicate that P10Mn scaffold can promote osteogenic differentiation by regulating the TGF-β/Smad signaling pathway (Fig. 3M).

2.5. PMns promote angiogenesis by regulating macrophages polarization in vitro

The immunomodulatory effects were assessed by co-culturing RAW 264.7 cells with scaffolds under various conditions for 24 hours, including blank group that without LPS and scaffolds. The P10Mn group showed a marked decrease in the expression of iNOS, a marker for M1phenotype macrophages, compared to the other groups (Fig. 4A and B). There was a notable increase in the expression of CD163, a marker for M2-phenotype macrophages, following treatment with P10Mn (Fig. 4C and D). No significant differences were observed in the expression of CD163 and iNOS between the LPS group and the LPS + PLLA group. These findings indicate that PMns may facilitate the transition from M1phenotype macrophages to M2-phenotype macrophages. Besides, the regulatory effect of various scaffolds on pro-inflammatory genes in RAW 264.7 cells after LPS treatment was analyzed using qRT-PCR. The results showed high levels of TNF-α, IL-6, and iNOS in the LPS and LPS + PLLA groups, indicated M1-phenotype macrophages activation, whereas these inflammatory markers were significantly decreased in the P10Mn group (Fig. 4EG). The study also explored the influence of P10Mn on M2phynotype macrophages gene expression, revealing a marked upregulation the expression of IL-10, Arg-1, and CD206 (Fig. 4H–J).

To investigate the influence of PMns on macrophage polarization and its effect on angiogenesis, the conditioned medium (CM) derived from macrophages was used to assess the release of relevant cytokines via enzyme-linked immunosorbent assay (ELISA) after co-culture with PLLA or P10Mn scaffolds. The concentrations of the pro-inflammatory factors IL-1β and TNF-α were decreased in P10Mn group compared to Blank and PLLA groups. (Fig. 4K and L). Meanwhile, higher levels of VEGF and TGF-β were observed in the CM of macrophages co-cultured with P10Mn scaffold (Fig. 4M and N). These results further suggested that the PMns could enhance macrophages polarization to M2phenotype, which is consistent with the results of immunofluorescence, promoting angiogenesis by secreting cytokines. To verify the effect of CM on HUVECs, transwell, wound healing and tube formation assay were conducted (Fig. 4O). Crystal violet staining (Fig. 4P) and wound healing assay (Fig. 4Q) suggested that HUVECs migration was remarkably promoted in P10Mn group compared to Blank and PLLA groups. The tube formation assay showed a similar trend to cell migration. A significantly increased number of meshes with more tube-like structures was observed in P10Mn group compared to Blank and PLLA groups (Fig. 4R). Furthermore, quantitative analysis of transwell, migration distance and number of meshes confirmed the enhanced angiogenesis of HUVECs by PMns treatment (Fig. 4S,T and U).

2.6. PMns scavenge ROS at the site of defect in vivo

We utilized ROS Brite™ 700 probe to evaluate ROS scavenging capabilities of the PMns at 1, 2, 4, and 8 weeks post-implantation in mice skull defect models (Fig. S7A). In vivo bioluminescence imaging results indicated that the microenvironment of the bone defect area was in a state of oxidative stress, with ROS levels remaining high. After scaffold implantation, ROS levels in P10Mn group were observed to be lower compared to the Control and PLLA groups, exhibited a progressive decline in a time-dependent manner from 1 to 8 weeks. (Fig. S7B and C).

2.7. PMns promote neovascularization and bone regeneration in vivo

We used dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), microangiography, micro-CT and histological staining to assess regeneration of blood vessels and bone tissue in the defect area of SAON rats at different time points (Fig. 5A). Time-intensity curves from DCE-MRI were obtained for the bone defect area (Fig. 5B). Various timeintensity curve patterns were categorized into three types, with maximum enhancement indicated blood perfusion levels (Fig. S8). Furthermore, micro-CT-based angiography was employed to assess new blood vessel formation in the distal femur. The results from micro-CTbased angiography revealed enhanced neovascularization in the P10Mn group compared to the Control and PLLA groups at different time points (Fig. 5C). The Control and PLLA groups exhibited lower maximum enhancement compared to the P10Mn group (Fig. 5D). In the P10Mn group, the signal intensity curve declined more rapidly after reaching its peak compared to the Control and PLLA groups at the same time point (Fig. 5E). The osteogenic efficacy of P10Mn scaffolds was assessed in SAON rats with distal femoral condylar defects. Threedimensional micro-CT scans of the defect areas were used to assess bone regeneration at 4, 8, and 12 weeks post-surgery. The regions implanted with P10Mn scaffolds showed a significantly larger volume of new bone formation compared to the Control and PLLA groups at each evaluated time point (Fig. 5F). Additionally, two-dimensional assessments revealed higher trabecular bone thickness in the P10Mn group compared to the Control and PLLA groups (Fig. 5G). Quantitative assessments of new bone tissue formation, including bone mineral density (BMD), bone volume fraction (BV/TV), trabecular thickness (Tb.Th), and trabecular separation (Tb.Sp) in the region of interest (ROI) of surgical area, showed that BMD, BV/TV, and Tb.Th were significantly higher in the P10Mn group at different time points (Fig. 5H–K). Furthermore, a reduction in Tb.Sp in the P10Mn group indicated more dense bone formation in the defect area.

2.8. Immunomodulatory properties of PMns in vivo

Given the pivotal role of macrophages in the immune response and their association with bone regeneration, representative markers of M1type macrophages (iNOS) and M2-type macrophages (CD163) were stained in femoral slices of SAON rats from various experimental groups at different time points. In the bone defect area, consistent with in vitro immunofluorescence staining and qRT-PCR findings, there was a detectable reduction in the number of iNOS+ cells (Fig. 6A and B), while the number of CD163+ cells exhibited a significant increase (Fig. 6C and D) after treatment with P10Mn scaffolds compared to the Blank and PLLA groups. Additionally, we examined the levels of cytokines in the serum of SAON rats using ELISA at 4 and 8 weeks post-surgery. TNF-α, IL-1β, and IL-6, which are pro-inflammatory factors, were significantly decreased (Fig. 6E–G), while the anti-inflammatory factor IL-10 exhibited an opposite trend (Fig. 6H) in the P10Mn groups compared to the Blank and PLLA groups. The in vivo experiments revealed that many macrophages accumulated at the scaffold transplantation site, suggesting that macrophages were recruited to the defect area during bone injury repair. The P10Mn group promoted M2-phenotype macrophage polarization in the defect area, further demonstrating that PMns possess excellent immunomodulatory function.

2.9. Histological analysis

The formation and maturation of new bone tissue in the defect area were histologically confirmed using Goldner's trichrome staining (Fig. 7A) and H&E staining (Fig. S9A). The images show that the bone defect area was predominantly filled with adipose tissue and sparse osteocytes in the Control groups at different time points. Conversely, the P10Mn group displayed significant new bone tissue formation and bone ingrowth into the scaffold at same time intervals. Additionally, sequential fluorescence micrographs of the undecalcified sections (Fig. 7B) revealed a higher mineral apposition rate (Fig. 7C) and bone formation rate (Fig. 7D). Compared to the Control and PLLA groups, the P10Mn group exhibited superior osteogenic function in promoting new bone formation. To assess neovascularization, immunofluorescence staining of type H vessels (Fig. 7E) and vascular endothelial growth factor (VEGF) (Fig. S9B) was performed to investigate the angiogenic effect of the scaffolds in vivo. Significantly higher expression of CD31 (Fig. 7F) and EMCN (Fig. 7G) was observed in the P10Mn group compared to the Control and PLLA groups after scaffold implantation.

3. Discussion

Core decompression is a crucial early intervention for SAON, aimed to reduce intramedullary pressure by removing necrotic bone tissue and subsequently implanting bone grafts to promote osteogenesis and angiogenesis [33]. An ideal bone graft substitute for SAON should possess appropriate mechanical properties, moderate biodegradability matching the rate of new bone growth, and excellent osteoconductivity and osteoinductivity to promote the generation of blood vessels in necrotic areas. After the removal of necrotic bone tissue, a range of microenvironmental changes occurs in the defect area, including elevated levels of ROS, increased inflammation, acidic pH, and low oxygen tension, all of which hinder the bone tissue regeneration [34, 35]. Hence, strategies aimed at improving the bone defect microenvironment and suppressing inflammation may promote osteogenesis and angiogenesis in SAON bone defects.

In this study, we utilized LT-RP 3D printing technology to develop PMns, with biodegradable PLLA serving as a matrix and nano-MnO2 as effector particles. MnO2 nanomaterials possess several advantages for bone tissue regeneration in the biomedical field, including scavenging endogenous ROS, generating O2 and improving hypoxic conditions. Additionally, the degradation product Mn2+ exhibits positive effects on osteogenesis and immune regulation. PLLA is a biodegradable polymer with a degradation rate of 1–2 years, aligning with the new bone regeneration rate of SAON bone defects, and has been approved by Food and Drug Administration (FDA) for clinical application [30,36]. We successfully prepared PLLA incorporating varying percentages of MnO2 nanoparticles using LT-RP 3D printing technology at − 30 ◦C (Fig. 1A). The porosity of the scaffolds can be precisely controlled using LT-RP 3D printing technology, featuring macropores and micropores that enhance new vessels and bone tissue formation and facilitate cell adhesion, proliferation, and extracellular matrix formation [37]. The PMns efficiently eliminate ROS and release O2 in acidic conditions (Fig. 1Q and Fig. S3), mitigating tissue hypoxia in the defect area and promoting tissue regeneration in SAON bone defects [38]. The compressive strength of the scaffolds, which depends on the presence of MnO2 nanoparticles, ranges from 2 to 3.5 MPa (Fig. 1L and M), comparable to that of human trabecular bone (2–12 MPa) [39]. Consequently, the scaffold developed in this study is also theoretically well-suited for addressing human cancellous bone defects.

As indicated by previous studies, excessive ROS can damage biomacromolecules such as nucleic acids, proteins, and lipids, impairing cell function, accelerating cell apoptosis, and hampering bone regeneration processes [40,41]. Therefore, restoring redox balance and effectively achieving antioxidation are crucial for promoting bone tissue repair. Manganese (Mn) plays a crucial role in regulating ROS through multiple molecular mechanisms, primarily due to its involvement in antioxidant defense systems, such as its role in superoxide dismutase (SOD) and mitochondrial homeostasis [42]. Therefore, the incorporation of manganese in biomaterials is an effective strategy for regulating ROS. Owing to its excellent ROS scavenging property, the PMns provide a protective effect for cells under excessive ROS conditions (Fig. 2B–L). We further investigated the underlying mechanism of the PMns' antioxidant capacity. Through the results of transcriptomics (Fig. 3I and J), we found that PMns upregulate the activities of antioxidant enzymes such as SOD2. Based on this, we performed qRT-PCR (Fig. 2M and Fig. S6D-F) on antioxidant genes related to ROS generation, including SIRT1, SOD2, and CAT, and demonstrated that PMns promote the upregulation of antioxidant genes expression. Therefore, we speculate that the remarkable antioxidant capacity of PMns may be attributed of the enhancement of cellular antioxidant capacity. Furthermore, generating O2 to improve hypoxia is another advantage of using PMns at SAON bone defect site. Another pathological feature of SAON is the poor O2 supply to the bone defect site, leading to decrease in cellular function [2]. It has been reported that delivering a certain amount of O2 to the bone defect site can promote bone regeneration [28,29,43]. After bone defect occurs, the microenvironment at the damage site is initially acidic and gradually becomes alkaline as tissue repair progresses, with an initial increase in H2O2 levels at the bone defect site [44]. In vivo results indicated that PMns effectively scavenged ROS in cranial defect of mice (Fig. S7). Consequently, implanting PMns at the bone defect site facilitates decomposition of H2O2 and the release of O2, improving the hypoxic microenvironment and thereby promoting bone tissue repair.

The osteogenic efficacy of the PMns was verified by evaluating the differentiation of hBMSCs in vitro. Comparative analyses between PLLA and PMns consistently demonstrated that PMns significantly augmented the differentiation of hBMSCs into osteocytes (Fig. 3A–H). Consistent with previous studies, the incorporation of Mn into biomaterials has been shown to promote osteogenesis and facilitate the adhesion of BMSCs [45,46]. Moreover, the ability of the PMns to augment bone tissue regeneration in vivo was confirmed using a distal femoral condylar defect model in rats with SAON. Both micro-CT reconstruction and histological staining results revealed more extensive formation of new bone tissue in the PMns group compared to the Control and PLLA groups (Fig. 6F–K, Fig. 7A and B, Fig. S9A). Thus, the PMns were confirmed to improve osteogenesis both in vitro and in vivo.

To further explore the molecular biological mechanism, transcriptomic sequencing was performed on Control group and P10Mn group (Fig. 3I–K). The GO enrichment analysis and volcano plots highlighted the upregulation of Smad1 in the P10Mn group, which is significant due to its role in the TGF-β signaling pathway, crucial for osteoblast differentiation [47]. The TGF-β signaling pathway has the highest enrichment score among other signaling pathways. (Fig. 3K). The canonical TGF-β signaling pathway involves interactions with TFG-βRI and TFG-βRII, initiating a cascade of Smad phosphorylation. Smad2/3 acting as a substrate for TGF-β, recruits other family members such as Smad4, to form the Smad complex. Once activated, this complex can induce the transcription of osteogenic genes in the nucleus, such as Runx2 and Osterix [48]. Consequently, six key proteins from the canonical Smad-dependent pathway (TGF-βRI, TGF-βRII, p-Smad2/3, Smad4, Runx2, and Osterix) were selected for verification by western blotting. The results showed a significant upregulation of these proteins in the P10Mn scaffold compared to the Control group (Fig. 3L), strongly indicating that the P10Mn scaffold promotes osteogenesis through the TGF-β/Smad signaling pathway. Previous studies have identified that Mn2+ acts as a crucial co-factor in boosting the ligand affinity of integrin and induces its expression [49]. As a cell adhesion molecule, integrin promotes interactions between cells and the extracellular matrix by binding to specific proteins and serving as a connector. Elevated integrin expression in the early stages enhances cell adhesion and dispersal, maintains cell adhesion strength and promotes the adhesion and proliferation of osteoblast-like cells (MC3T3 cells), as well as the osteogenic differentiation of hBMSCs [50–52]. This enhances the integration of autologous bone or bone graft substitutes and promotes the expression of bone-related proteins, thereby increasing the bone repair rate of composite materials [53]. The downstream pathways of integrin, particularly the RhoA/ROCK and MAPK signaling pathways, are significantly activated to promote osteogenic differentiation [54,55]. There is extensive crosstalk between integrins and TGF-β; integrins binding to RGD can activate latent TGF-β, and TGF-β signaling significantly influences the expression of genes encoding integrin subunits [47, 56]. Further research has shown that manganese can enhance the bioactivity of cartilage oligomeric matrix protein by upregulating intracellular TGF-β binding sites and stimulating the synthesis of chondroitin sulfate, a key component in bone formation [57]. Thus, there may be the complex interactions between TGF-β and integrin signaling pathways in promoting osteogenesis by the PMns.

Bone tissue regeneration by implanted materials depends on a series of complex biological responses, with the inflammatory response playing a pivotal role in the regeneration process. Initially, most macrophages polarize into the M1-phenotype, exacerbating inflammation and leading to tissue fibrosis [58]. The persistence of M1-phenotype macrophages contributes to chronic inflammation, which impedes bone regeneration. Conversely, a shift to the M2-phenotype alleviates inflammation and promote angiogenesis. M2-phenotype macrophages enhance the organization of the developing vasculature by secreting VEGF and platelet-derived growth factor-BB (PDGF-BB), along with osteoinductive factors like BMP-2, BMP-4, and TGF-β, facilitating new bone tissue formation [59]. In our study, the results suggested that the P10Mn scaffold effectively modulates macrophage polarization from M1 to M2, significantly decreasing the expression of pro-inflammatory genes and significantly increasing the expression of anti-inflammatory genes (Fig. 4A–J). Furthermore, to examine the influence of macrophage regulation on angiogenesis, we analyzed the conditioned medium from scaffolds co-cultured with macrophages using ELISA assay. The findings indicated that our scaffolds not only regulate the transformation of macrophages toward the M2-phenotype but also promote the secretion of cytokines that enhance angiogenesis in vitro (Fig. 4M–U). Furthermore, immunofluorescence and Elisa results reveal that the P10Mn scaffold can regulate macrophage polarization to improve osteoimmune microenvironment in vivo (Fig. 6).

The angiogenic potential of the P10Mn scaffold was evaluated using DCE-MRI and micro-CT-based angiography in the distal femoral condylar defects of SAON rats. DCE-MRI results, reflecting the functional level of blood perfusion [60], showed a lower "washout rate" in the Control and PLLA groups compared to the P10Mn group, indicating a slower return of the injected contrast agent from the neovascularization to the bloodstream. The "maximum enhancement", representing the volume of blood perfusion, was also lower in the Control and PLLA groups compared to the P10Mn group at 2 weeks post-surgery. Over time, "maximum enhancement" gradually decreased and "washout rate" increased in the P10Mn group, suggesting a gradual restoration of vascular function. This trend was not observed in the Control and PLLA groups (Fig. 5B, D and E), highlighting the superior ability of the P10Mn scaffold to promote the regeneration of functional blood vessels. Micro-CT-based angiography results reflect the amount of neovascularization [6]. More neovascularization was noted in the P10Mn group compared to the Control and PLLA groups (Fig. 5C). Immunofluorescent staining for type H vessels and VEGF indicated more new vessel tissue formation in the P10Mn group compared to the Control and PLLA groups (Fig. 7E, Fig. S9B). Collectively, these findings underscore the exceptional capability of PMns in modulating the osteoimmune microenvironment and promoting neovascularization.

Ensuring excellent biocompatibility is crucial for the future clinical translation of the fabricated scaffolds. H&E staining of major organ slices from SAON rats at 12 weeks after PMns implantation indicated no signs of tissue necrosis, inflammation, infection, or tumors in the visceral microstructure (Fig. S10A). Blood biochemical tests demonstrated normal liver and kidney function without any complications (Fig. S10B-K).

5. Experimental section

5.1. Fabrication of PMns

The fabrication of the PMns involved dissolving medical-grade polyL-lactide (PLLA, Mw 200 kD; obtained from Shandong Institute of Medicine Instruments, China) in a 15 % (w/v) homogeneous solution of 1,4-dioxane (Shanghai Ling Feng Chemical Reagent Co., Ltd., China). Manganese dioxide powder (MnO2, size around 60–100 nm; provided by J&K Scientific, China) was added to the solution, which was vigorously stirred overnight using a magnetic stirrer to form a uniform suspension. The suspension, with various concentrations of 0 % (PLLA, pure poly-Llactide), 2.5 % (P2.5Mn, the mass ratio of PLLA to MnO2 was 97.5:2.5), 5 % (P5Mn, the mass ratio of PLLA to MnO2 was 95:5), 10 % (P10Mn, the mass ratio of PLLA to MnO2 was 90:10) and 20 % (P20Mn, the mass ratio of PLLA to MnO2 was 80:20) was used to prepare the porous scaffolds by a LT-RP 3D printing technology (CLRF-2000-II, Tsinghua University, China) at − 30 ◦C. The macroscopic and internal structures of the porous scaffolds were crafted based on a stereolithography model for 3D porous scaffold blocks. Dimensional parameters, including porosity and porous interconnectivity, for each scaffold type (cubic scaffolds measuring 20 × 20 × 20 mm3 ), were maintained consistently across different groups. The scaffolds were lyophilized using a freeze dryer (Alpha 2–4 LDplus, Marin Christ, Germany) for 72 hours to fully evaporate 1,4-dioxane solvent after fabrication. Prior to conducting cell or animal experiments, all scaffolds were sterilized through gamma-ray irradiation (JPY ION-TECH. CO., LTD, China).

5.2. Characterization of PMns

After gold sputtering, the microstructure of the scaffolds was analyzed using field emission scanning electron microscopy (SEM, Zeiss Supra55, Carl Zeiss, Germany) and energy-dispersive X-ray spectroscopy (EDS, X-Max 20, UK). A micro-CT machine (Skyscan 1176, Bruker, Belgium) was employed to scan the P10Mn scaffold, and CT analyzer was used to reconstruct images. Compression tests were conducted using mechanical testing instrument (Instron Electropuls E10000, USA). Three samples from each group were tested to determine the compressive modulus and stress strength for comparing mechanical properties. The scaffolds, shaped into 5 × 5 × 5 mm3 cubes, were degassed in a vacuum drying chamber and subsequently soaked in phosphate-buffered saline (PBS) solution at a mass-to-volume ratio of 0.1 g/mL. For the Mn ion concentration, we also evaluated the release of P10Mn in a solution with a pH of 6.8. The degradation solution was collected weekly, and the pH value and Mn concentration were determined using a pH meter (S220 Seven Compact, Mettler Toledo, China) and inductively coupled plasma optical emission spectrometry (ICP-OES, 700 Series, Agilent Technologies, USA), respectively. Furthermore, the H2O2 scavenging capacity of scaffolds was tested by immersing them in a solution containing 1 mM H2O2 and observing the generated bubbles after 30 minutes. Additionally, the O2 release profiles of P10Mn scaffolds at 100 μM and 200 μM H2O2 were measured using a dissolved oxygen meter (INESA Scientific Instrument Co., Ltd, China). To assess the CAT-enzyme activity of PMns, the scaffolds of different groups were incubated with 10 mM H2O2 for different time points. The remaining amount of H2O2 was measured using a H2O2 kit (Catalase Assay Kit, Beyotime, China), and the H2O2 clearance rate of the scaffolds of different groups was calculated.

5.3. Evaluation of biocompatibility in vitro

Human bone marrow mesenchymal stem cells (hBMSCs, ATCC PCS500-012) were co-cultured with the scaffolds. After culturing for 7 days on the scaffolds, live/dead staining was performed. Additionally, after 7 days of co-culture, cells were incubated with FITC-phalloidin (Yeasen Biotechnology, 40735ES75) and DAPI (Beyotime Biotechnology, C1005). A laser scanning confocal microscope (ZEISS, LSM900) was used to observe the morphology, nuclei (blue, 405 nm) and actin (green, 496 nm). The cytotoxicity of leaching solution from different scaffolds was assessed in accordance with GB/T16886.5–2017.

5.4. Evaluation of intracellular ROS scavenging

hBMSCs were plated in a 24-well transwell setup at a density of 4 × 104 cells per well. After 12 hours, cells in the bottom chamber were treated with 100 μM hydrogen peroxide (H2O2). Meanwhile, different scaffolds were placed in the top chamber. While a setup with an empty top chamber served as the blank group. After 24 hours, hBMSCs were treated with DCFH-DA (Reactive Oxygen Species Assay Kit, Beyotime Biotechnology) to measure ROS production. live/dead cell staining was then conducted using a Leica DMi8 inverted fluorescence microscope (Germany). Concurrently, cell apoptosis levels were evaluated using an apoptosis detection kit. In brief, hBMSCs were seeded in 12-well plates. After 12 hours, 100 μM H2O2 and various scaffolds were added and cocultured for 24 hours. Subsequently, both adherent and non-adherent cells were harvested and stained with Annexin V-FITC and propidium iodide (PI) for 30 minutes. Following filtration, the cell suspension was analyzed by flow cytometry. For HUVECs, after being seeded for 12 hours, a pipette tip was used to scratch a vertical line in the plate, and the same treatment was applied as described above. After 24 hours, the scratch was observed for healing, and the expression of antioxidant genes of HUVECs was detected.

5.5. Evaluation of osteogenesis properties in vitro

Alizarin Red S (ARS) staining: hBMSCs at a density of 4 × 104 cells per well were cultured in a 12-well plate using basal medium. After 3 days of culture, the cells were exposed to α-MEM complete medium with osteogenic induction medium and various scaffolds (with the blank group lacking osteogenic induction medium). After 14 days of differentiation, the samples were stained with 1 % Alizarin Red S (SigmaAldrich, A5533) to assess calcium nodule formation and matrix mineralization. After staining, the samples were examined under an inverted microscope. The residual Alizarin Red S was removed using 10 % hexadecyl pyridinium (Sigma-Aldrich, 17692378) and quantified using a microplate reader (Thermo Fisher Scientific, USA). ALP staining and ALP activity assay: hBMSCs were grown in the osteogenic induction medium as described. After 7 days, ALP staining was performed using the ALP kit (Beyotime Biotechnology, C3206), and optical images were captured with an inverted light microscope. Cellular ALP activity was assessed by lysing the cells with RIPA buffer (Thermo Scientific, 89900) after 7 days of osteogenic induction. Subsequently, cellular proteins were extracted, and ALP activity was measured using ALP kit (Beyotime Biotechnology, P0321).

5.6. Detection of osteogenic-related gene expression

The mRNA levels of osteogenic-specific genes were examined by qRT-PCR. Cells were harvested at different stages of osteogenic induction (7 and 14 days), and RNA was isolated using a Total RNA kit (Axygen, RNA-250). The extracted mRNA was converted to cDNA using PrimeScript™ RT Mix (Takara, RR036A). Amplification of target genes in the cDNA was conducted in a 10 μl reaction mixture containing SYBR Premix Ex TaqII (Takara, RR820B), cDNA, and specific primers (Table S1) with the Light Cycler 96 system (Roche, USA). The expression levels of mRNA were normalized to the reference gene GAPDH (Glyceraldehyde-3-Phosphate Dehydrogenase).

5.7. Detection of osteogenic-related protein expression

Osteogenic proteins were extracted from hBMSCs after 14 days of osteogenic induction using RIPA buffer. The proteins were then equally loaded and separated on gels at specified concentrations before being transferred to polyvinylidene fluoride (PVDF) membranes. Western blotting was performed following the methods described in a previous study [61]. The antibodies used included Anti-TGF beta Receptor I (ab235575), Anti-TGF beta Receptor II (ab186838), Anti-p-Smad2/3 (ab272332), Anti-Smad4 (ab110175), Runx2 (ab76956), Osterix (ab94744), and β-actin (ab8226), all sourced from Abcam.

5.8. RNA sequencing (RNA-seq) and bioinformatic analyses

hBMSCs (4 × 104 cells per well) were incubated in 12-well plates for 3 days. The existing medium was then replaced with osteoblast-inducing conditional media, and the cells were co-cultured with various scaffolds. After 14 days of differentiation, the cells were lysed using Trizol reagent (Life, 15596026), and the lysates were stored at − 80 ◦C until needed for sequencing. RNA integrity was verified using gel electrophoresis and Qubit analysis (Thermo, Waltham, MA, USA). The library preparation was prepared using the VAHTS Stranded mRNA-seq Library Prep Kit for Illumina (Vazyme, China), and sequencing was performed on the Illumina Novaseq 6000 system (provided by Genergy Biotechnology Co., Ltd. Shanghai, China). Skewer was utilized to process the raw sequencing data, and data quality was confirmed by FastQC (v0.11.2). The clean sequences were mapped to the reference genome using STAR (2.5.3a). Transcript levels were quantified as FPKM (Fragments Per Kilobase of exon per Million mapped reads) using StringTie (v1.3.1c). DEseq2 (v1.16.1) was used to identify differentially expressed genes (DEGs).

5.9. Macrophage phenotype regulation and gene expression on PMns

The material samples were initially placed in 12-well plates according to the experimental setup. Each well was inoculated with 2 × 105 RAW 264.7 cells and incubated for 12 hours. Subsequently, the cells were treated with a medium containing 100 ng/ml lipopolysaccharide (LPS, Sigma-Aldrich, USA) for 24 hours to induce polarization (the Control group did not receive LPS). Antibodies against iNOS (Abcam ab210823) and CD163 (Biorbyt, orb544646) were used to assess macrophage polarization via immunofluorescence. After 36 hours, total RNA was extracted from the cells, and mRNA was isolated as previously described

5.10. Detection of inflammatory cytokines in conditioned medium

RAW 264.7 cells at a density of 2 × 105 cells per well were seeded into 12-well plates with various scaffolds and co-cultured for 24 hours (with the blank group contained no scaffold). After 24 hours, the supernatant was collected by centrifuging at 1000 rpm for 5 minutes to obtain conditioned medium. This conditional medium was then analyzed using an Inflammatory Cytokine Assay ELISA kit (Thermo Fisher Scientific, USA).

5.11. Evaluation of angiogenesis properties

Transwell assay:5 × 104 HUVECs were seeded into the upper chamber of 24-well transwell plate, while the lower chamber was filled with conditioned medium derived from 24 hours co-culture of macrophages with various scaffolds. After 12 hours, the lower chamber was washed three times with PBS, then crystal violet staining was performed. Wound healing assay: Culture–Insertinμ–Dish (ibidi, Germany) was seeded with 5 × 104 HUVEC cells per chamber. After 24 hours, the inserts were removed, and conditioned medium was added to the chamber. Images were taken at 0 and 24 hours. Tube formation assay: Matrix gel (Abw, Mogengel Biotechnology co., Ltd) and DMEM medium were mixed in a 1:1 vol ratio and added to u-slide (ibidi, Germany) then incubated at 37 ◦C for 30 minutes. Then, 1.3 × 104 HUVEC cells were seeded on the surface, and 50 μL of different conditioned media were added to each well. After 6 hours, the tubes formed in each group were stained with calcein-AM and imaged using an inverted fluorescence microscope (Leica, DMi8, Germany). Image J software was used for quantitative analysis.

5.12. Construction of animal disease model and scaffolds implantation

Male Sprague-Dawley rats, six weeks old, were obtained from the Guangdong Medical Laboratory Animal Center (Guangzhou, China) and housed in a sterile environment at the Shenzhen Institute of Advanced Technology (Shenzhen, China). The protocols for animal experiments were approved by the Research Ethics Committee at the Shenzhen Institute of Advanced Technology. Briefly, the rats were received two intraperitoneal injections of lipopolysaccharide (LPS) at a dose of 10 mg/kg and an interval of 24 hours between each. Then three successive intramuscular injections of methylprednisolone (MPS, Pfizer, USA) every 24 hours. After two weeks osteonecrosis induction, the rats were anesthetized using isoflurane gas. Bilateral incisions were made on the skin, and the muscles were bluntly dissected to expose the femoral condyles. A tunnel with a diameter of 3.0 mm and a depth penetrating the opposite cortical bone was drilled into each using electric drill, into which specific scaffolds were placed. Three groups were formed for the study: control group with no scaffold, PLLA group with PLLA scaffolds, and P10Mn group with P10Mn scaffolds, with each group containing six subjects per time point. After the procedure, the rats were received intraperitoneal injections of penicillin for three days to prevent infection.

5.13. Detection of ROS levels in skull defect area

C57BL/6 wild type mice, aged eight weeks, were obtained from the Guangdong Medical Laboratory Animal Center (Guangzhou, China) and housed in a sterile environment at the Shenzhen Institute of Advanced Technology (Shenzhen, China). The mice were initially anesthetized with isoflurane gas. A circular, full-thickness bone defect with a diameter of 3.0 mm was then drilled into the skull. The defects were filled with various scaffolds (with one group receiving no scaffold as a control), and the incisions were sutured in layers. After surgery, the mice received intraperitoneal penicillin injections for three days to prevent infections. Bioluminescence imaging was used to monitor the ROS production at 1, 2, 4, and 8 weeks after surgery by injecting the ROS Brite™ 700 probe into the skull defect area.

5.14. Perfusion function evaluated by DCE-MRI

SAON rats were anesthetized with gas, and an indwelling needle was placed in the tail vein for observations at 2, 4, and 8 weeks following scaffold implantation. Before the injection of Gadoteric Acid Meglumine Salt (Hengrui Pharmaceuticals Co., Ltd, China), seven initial images were captured using dynamic MRI (3T UNITED IMAGING, China) in a T1-weighted sequence to delineate the region of interest (ROI) and establish baseline images. Subsequently, 0.5 ml of contrast agent and 0.5 ml of saline were injected through the indwelling needle for a continuous scanning time of 20 minutes.

... (1)

... (2)

5.15. Neovascularization analysis (microfil)

At 2, 4, and 8 weeks after implantation, the anesthetized rats were placed in a supine position. The skin and muscle were incised layer by layer. Subsequently, we dissected the abdominal aorta and the inferior vena cava, and catheters were individually inserted into each vessel. Next, 1000 ml of heparin saline (50 U/ml) was administered through the abdominal aortic catheter to heparinize the lower limb vascular system. A slow injection of 10 ml of microfil angiographic agent (Micro-fil, Flow Tech Inc., USA) was performed through the abdominal aortic catheter. The two catheters were then clamped to ensure that the lower limb filled with the contrast agent. The rats were euthanized and kept at 4 ◦C overnight. The bilateral femoral specimens were collected, fixed in 10 % formalin for 72 hours, and then decalcified with 10 % ethylenediaminetetraacetic acid (EDTA, pH 7.4). After decalcification, microCT analysis (Skyscan 1176, Bruker, Belgium) was performed on the distal femoral samples. 3D angiographic reconstructions of the samples were carried out using NRecon software and CT-Analyzer. The analysis parameters were set with a minimum grayscale threshold of 90 HU and a maximum of 255 HU.

5.16. Micro-CT 3D analysis for bone generation

The animals were anesthetized and euthanized with isoflurane gas, and their femurs were harvested at different time points. New bone formation in the defect areas was evaluated by micro-CT scanning (Skyscan 1176, Bruker, Belgium). The 3D structures of the femur samples were reconstructed using NRecon software. An area of interest (AOI) with a diameter of 3.0 mm was selected to analyze the new trabecular bone in the defect area. The binary image threshold was set from 68 to 255, and the bone mineral density (BMD) of the new trabecular bone was calibrated using a standard rat BMD sample. Bone parameters analysis was conducted using CT-Analyzer software. Additionally, CTvol software was used to reconstruct 3D models of the new trabecular and cortical bone.

5.17. Detection of cytokines in SAON rats' blood

After anesthetized, blood was collected from the rats and centrifuged at 4000 g for 20 minutes to separate the serum. Serum cytokine levels (IL-1β, MCP-1, IL-6, and TNF-α) were measured using ELISA kits according to the manufacturer's instructions. Optical densities were measured with a microplate reader (Thermo Fisher Scientific, USA), and cytokine concentrations were calculated based on standard curves.

5.18. Undecalcified histological sections

Calcein-AM (Sigma-Aldrich, 17783) was injected subcutaneously for two consecutive days, followed by a 10-day interval, after which xylenol orange (Sigma-Aldrich, 52097) was injected for two consecutive days in SAON rats. The animals were then anesthetized and euthanized, and their femurs were excised and fixed in 10 % formalin. The specimens were dehydrated through an ethanol gradient from 70 % to 100 % and embedded in methyl methacrylate (Aladdin, M109623). Each section was cut along the longitudinal axis into 8 μm thick slices and fix with 1 % gelatin. During analysis, green fluorescence from Calcein and the red fluorescence from xylenol orange were visualized using fluorescence microscope (Leica, DMi8, Germany) at excitation wavelengths of 480 nm and 560 nm, respectively. A yellow double-headed arrow was used to measure the distance newly formed bone. The bone regeneration rate was quantified using the mineral apposition rate (MAR, μm per day) and bone formation rate (BFR, %). Goldner's trichrome staining was performed according to the kit instructions (Sigma-Aldrich, HT10316).

5.19. Decalcified histological sections

The samples were decalcified in 10 % EDTA (Boster Biological Technology Co., Ltd, China) for eight weeks. After decalcification, the samples were embedded in paraffin, and the central femoral sections were cut into continuous 4 μm slices. These sections were stained with hematoxylin and eosin (H&E) from Beyotime Biotechnology (C0105). For immunofluorescence, the sections were treated with 0.3 % Triton X100 and blocked with 5 % goat serum. The sections were then incubated overnight with primary antibodies including CD31 (Bioss bs-0195R), Endomucin (Bioss bs-5884R-PE), VEGF (Invitrogen JH121), iNOS (Abcam ab210823), and CD163 (Biorbyt orb544646). After primary incubation, the sections were washed and then incubated with secondary antibodies for 1 h at room temperature. Nuclei were stained with DAPI and examined using a laser scanning confocal microscope (ZEISS, LSM900).

5.20. Statistical analysis

All statistical analyses were performed using GraphPad Prism 9 software (La Jolla, CA, USA). Data are presented as the mean ± standard deviation (SD). Comparison between two independent sample groups were carried out using the t-test. For differences among multiple groups, a one-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test was used. p＜0.05 was considered statistically significant.

CRediT authorship contribution statement

Yipei Yang: Writing – original draft, Visualization, Validation, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Zhenyu Yao: Writing – review & editing, Writing – original draft, Methodology, Investigation, Data curation. Yuanyi Sun: Validation, Methodology, Formal analysis, Data curation. Yangyi Nie: Software, Methodology, Formal analysis, Data curation. Yuanchi Zhang: Validation, Software, Methodology. Ziyue Li: Software, Data curation. Zhiheng Luo: Data curation. Wenjing Zhang: Data curation. Xiao Wang: Data curation. Yuhan Du: Data curation. Wei Zhang: Conceptualization. Ling Qin: Writing – review & editing, Supervision, Project administration. Hongxun Sang: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization. Yuxiao Lai: Writing – review & editing, Writing – original draft, Supervision, Project administration, Funding acquisition, Conceptualization.

Ethics approval and consent to participate

All animal experiments were performed in compliance with the guidelines for the care and use of laboratory animals and were approved by the Institutional Animal Care and Use Committee (IACUC) of Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences (SIAT-IACUC-20190725-YGS-ZHZX-CYQ-A0580-03). Every effort was made to minimize animal suffering and to reduce the number of animals used.

Declaration of competing interest

The authors declare that there are no conflicts of interest.

Acknowledgements

YP.Y. ZY.Y. and YY.S contributed equally to this work. This work was financially supported by the National Natural Science Foundation of China (81871767), Shenzhen Medical Research Funds (B2302050), Guangdong Basic and Applied Basic Research Foundation (No. 2022B1515120046, 2023A1515010087, 2023A1515011315), Shenzhen Key Laboratory of Digital Surgical Printing Project (ZDSYS201707311542415), Shenzhen Science and Technology Program (JCYJ20220818103417037, JCYJ20210324115814040, JCYJ 20210324102206016, JSGG20210629144538010, KJZD20230923115 200002), Shenzhen Development and Reform Program (XMHT20220106001), and the Shenzhen Basic Research General Project (JCYJ20220531100408019).

Sidebar

Received 2 August 2024; Received in revised form 1 October 2024; Accepted 18 October 2024 Bioactive Materials 44 (2025) 354–370

Available online 28 October 2024

Footnote

Peer review under responsibility of KeAi Communications Co., Ltd.

* Corresponding author. Centre for Translational Medicine Research & Development, Institute of Biomedical and Health Engineering, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, 518055, China.

** Corresponding author. Department of Orthopedic Surgery, Shenzhen Hospital, Southern Medical University, Shenzhen, 518000, China.

E-mail addresses: [email protected] (H. Sang), [email protected] (Y. Lai).

1 These authors contributed equally to this work.

References

References

[1] M. Liu, Y. Gao, Y. Yuan, K. Yang, S. Shi, J. Zhang, J. Tian, Efficacy and safety of integrated traditional Chinese and western medicine for Corona Virus disease 2019 (COVID-19): a systematic review and meta-analysis, Pharmacol. Res. 158 (2020) 104896, https://doi.org/10.1016/j.phrs.2020.104896.

[2] A. Wang, M. Ren, J. Wang, The pathogenesis of steroid-induced osteonecrosis of the femoral head: a systematic review of the literature, Gene 671 (2018) 103–109, https://doi.org/10.1016/j.gene.2018.05.091.

[3] H. Cao, K. Shi, J. Long, Y. Liu, L. Li, T. Ye, C. Huang, Y. Lai, X. Bai, L. Qin, X. Wang, PDGF-BB prevents destructive repair and promotes reparative osteogenesis of steroid-associated osteonecrosis of the femoral head in rabbits, Bone 167 (2023) 116645, https://doi.org/10.1016/j.bone.2022.116645.

[4] K.C. Hua, X.G. Yang, J.T. Feng, F. Wang, L. Yang, H. Zhang, Y.C. Hu, The efficacy and safety of core decompression for the treatment of femoral head necrosis: a systematic review and meta-analysis, J. Orthop. Surg. Res. 14 (1) (2019) 306,

[5] M. Li, D. Chen, Y. Ma, M. Zheng, Q. Zheng, Stem cell therapy combined with core decompression versus core decompression alone in the treatment of avascular necrosis of the femoral head: a systematic review and meta-analysis, J. Orthop. Surg. Res. 18 (1) (2023) 560, https://doi.org/10.1186/s13018-023-04025-8. [6] Y. Lai, Y. Li, H. Cao, J. Long, X. Wang, L. Li, C. Li, Q. Jia, B. Teng, T. Tang, J. Peng, D. Eglin, M. Alini, D.W. Grijpma, G. Richards, L. Qin, Osteogenic magnesium incorporated into PLGA/TCP porous scaffold by 3D printing for repairing challenging bone defect, Biomaterials 197 (2019) 207–219, https://doi.org/ 10.1016/j.biomaterials.2019.01.013. [7] E.E. Huang, N. Zhang, H. Shen, X. Li, M. Maruyama, T. Utsunomiya, Q. Gao, R. A. Guzman, S.B. Goodman, Novel techniques and future perspective for investigating critical-size bone defects, Bioengineering (Basel) 9 (4) (2022), https://doi.org/10.3390/bioengineering9040171. [8] C. Li, W. Zhang, Y. Nie, D. Jiang, J. Jia, W. Zhang, L. Li, Z. Yao, L. Qin, Y. Lai, Integrated and bifunctional bilayer 3D printing scaffold for osteochondral defect repair, Adv. Funct. Mater. 33 (20) (2023), https://doi.org/10.1002/ adfm.202214158. [9] D.C. Lobb, B.R. DeGeorge Jr., A.B. Chhabra, Bone graft substitutes: current concepts and future expectations, J Hand Surg Am 44 (6) (2019) 497–505 e2, https://doi.org/10.1016/j.jhsa.2018.10.032. [10] G.S. Shi, Y.Y. Li, Y.P. Luo, J.F. Jin, Y.X. Sun, L.Z. Zheng, Y.X. Lai, L. Li, G.H. Fu, L. Qin, S.H. Chen, Bioactive PLGA/tricalcium phosphate scaffolds incorporating phytomolecule icaritin developed for calvarial defect repair in rat model, J Orthop Translat 24 (2020) 112–120, https://doi.org/10.1016/j.jot.2020.05.008. [11] X. Wei, W. Zhou, Z. Tang, H. Wu, Y. Liu, H. Dong, N. Wang, H. Huang, S. Bao, L. Shi, X. Li, Y. Zheng, Z. Guo, Magnesium surface-activated 3D printed porous PEEK scaffolds for in vivo osseointegration by promoting angiogenesis and osteogenesis, Bioact. Mater. 20 (2023) 16–28, https://doi.org/10.1016/j. bioactmat.2022.05.011. [12] C. Li, W. Zhang, Y. Nie, X. Du, C. Huang, L. Li, J. Long, X. Wang, W. Tong, L. Qin, Y. Lai, Time-sequential and multi-functional 3D printed MgO(2)/PLGA scaffold developed as a novel biodegradable and bioactive bone substitute for challenging postsurgical osteosarcoma treatment, Adv. Mater. 36 (34) (2024) e2308875, https://doi.org/10.1002/adma.202308875. [13] S. Li, L. Zhang, C. Liu, J. Kim, K. Su, T. Chen, L. Zhao, X. Lu, H. Zhang, Y. Cui, X. Cui, F. Yuan, H. Pan, Spontaneous immunomodulation and regulation of angiogenesis and osteogenesis by Sr/Cu-borosilicate glass (BSG) bone cement to repair critical bone defects, Bioact. Mater. 23 (2023) 101–117, https://doi.org/ 10.1016/j.bioactmat.2022.10.021. [14] Q. Cui, W.L. Jo, K.H. Koo, E.Y. Cheng, W. Drescher, S.B. Goodman, Y.C. Ha, P. Hernigou, L.C. Jones, S.Y. Kim, K.S. Lee, M.S. Lee, Y.J. Lee, M.A. Mont, N. Sugano, J. Taliaferro, T. Yamamoto, D. Zhao, ARCO consensus on the pathogenesis of non-traumatic osteonecrosis of the femoral head, J. Kor. Med. Sci. 36 (10) (2021) e65, https://doi.org/10.3346/jkms.2021.36.e65.

[15] Y. Peng, S. Wu, Y. Li, J.L. Crane, Type H blood vessels in bone modeling and remodeling, Theranostics 10 (1) (2020) 426–436, https://doi.org/10.7150/ thno.34126.

[16] P. Chen, Y. Liu, W. Liu, Y. Wang, Z. Liu, M. Rong, Impact of high-altitude hypoxia on bone defect repair: a review of molecular mechanisms and therapeutic implications, Front. Med. 9 (2022) 842800, https://doi.org/10.3389/ fmed.2022.842800.

[17] A.J. Sheppard, A.M. Barfield, S. Barton, Y. Dong, Understanding reactive oxygen species in bone regeneration: a glance at potential therapeutics and bioengineering applications, Front. Bioeng. Biotechnol. 10 (2022) 836764, https://doi.org/ 10.3389/fbioe.2022.836764.

[18] Y. Liu, C. Wang, G. Wang, Y. Sun, Z. Deng, L. Chen, K. Chen, J. Tickner, J. Kenny, D. Song, Q. Zhang, H. Wang, Z. Chen, C. Zhou, W. He, J. Xu, Loureirin B suppresses RANKL-induced osteoclastogenesis and ovariectomized osteoporosis via attenuating NFATc1 and ROS activities, Theranostics 9 (16) (2019) 4648–4662, https://doi.org/10.7150/thno.35414.

[19] W. Zhao, X. Zhang, R. Zhang, K. Zhang, Y. Li, F.J. Xu, Self-assembled herbal medicine encapsulated by an oxidation-sensitive supramolecular hydrogel for chronic wound treatment, ACS Appl. Mater. Interfaces 12 (51) (2020) 56898–56907, https://doi.org/10.1021/acsami.0c19492.

[20] S.D. Dutta, K. Ganguly, T.V. Patil, A. Randhawa, K.T. Lim, Unraveling the potential of 3D bioprinted immunomodulatory materials for regulating macrophage polarization: state-of-the-art in bone and associated tissue regeneration, Bioact. Mater. 28 (2023) 284–310, https://doi.org/10.1016/j.bioactmat.2023.05.014.

[21] Y. Zhu, H. Liang, X. Liu, J. Wu, C. Yang, T.M. Wong, K.Y.H. Kwan, K.M.C. Cheung, S. Wu, K.W.K. Yeung, Regulation of macrophage polarization through surface topography design to facilitate implant-to-bone osteointegration, Sci. Adv. 7 (14) (2021), https://doi.org/10.1126/sciadv.abf6654.

[22] M. Locati, G. Curtale, A. Mantovani, Diversity, mechanisms, and significance of macrophage plasticity, Annu. Rev. Pathol. 15 (2020) 123–147, https://doi.org/ 10.1146/annurev-pathmechdis-012418-012718.

[23] Y. Chen, H. Cong, Y. Shen, B. Yu, Biomedical application of manganese dioxide nanomaterials, Nanotechnology 31 (20) (2020) 202001, https://doi.org/10.1088/ 1361-6528/ab6fe1.

[24] M. Wen, N. Yu, S. Wu, M. Huang, P. Qiu, Q. Ren, M. Zhu, Z. Chen, On-demand assembly of polymeric nanoparticles for longer-blood-circulation and disassembly in tumor for boosting sonodynamic therapy, Bioact. Mater. 18 (2022) 242–253, https://doi.org/10.1016/j.bioactmat.2022.03.009.

[25] C. Wang, Y. Guan, M. Lv, R. Zhang, Z. Guo, X. Wei, X. Du, J. Yang, T. Li, Y. Wan, X. Su, X. Huang, Z. Jiang, Manganese increases the sensitivity of the cGAS-STING pathway for double-stranded DNA and is required for the host defense against DNA viruses, Immunity 48 (4) (2018) 675–687 e7, https://doi.org/10.1016/j. immuni.2018.03.017. [26] Y. Li, Q. Pan, J. Xu, X. He, H.A. Li, D.A. Oldridge, G. Li, L. Qin, Overview of methods for enhancing bone regeneration in distraction osteogenesis: potential roles of biometals, J Orthop Translat 27 (2021) 110–118, https://doi.org/10.1016/ j.jot.2020.11.008. [27] J. Li, C. Deng, W. Liang, F. Kang, Y. Bai, B. Ma, C. Wu, S. Dong, Mn-containing bioceramics inhibit osteoclastogenesis and promote osteoporotic bone regeneration via scavenging ROS, Bioact. Mater. 6 (11) (2021) 3839–3850, https://doi.org/10.1016/j.bioactmat.2021.03.039. [28] J. Li, F. Han, J. Ma, H. Wang, J. Pan, G. Yang, H. Zhao, J. Zhao, J. Liu, Z. Liu, B. Li, Targeting endogenous hydrogen peroxide at bone defects promotes bone repair, Adv. Funct. Mater. 32 (10) (2021), https://doi.org/10.1002/adfm.202111208. [29] Q. Chen, J. Li, F. Han, Q. Meng, H. Wang, Q. Wei, Z. Li, F. Li, E. Xie, X. Qin, S. Chen, W. Wang, C. Liu, B. Li, F. Han, A multifunctional composite hydrogel that rescues the ROS microenvironment and guides the immune response for repair of osteoporotic bone defects, Adv. Funct. Mater. 32 (27) (2022), https://doi.org/ 10.1002/adfm.202201067. [30] C. Garot, G. Bettega, C. Picart, Additive manufacturing of material scaffolds for bone regeneration: toward application in the clinics, Adv. Funct. Mater. 31 (5) (2021), https://doi.org/10.1002/adfm.202006967. [31] W. Wang, B. Zhang, M. Li, J. Li, C. Zhang, Y. Han, L. Wang, K. Wang, C. Zhou, L. Liu, Y. Fan, X. Zhang, 3D printing of PLA/n-HA composite scaffolds with customized mechanical properties and biological functions for bone tissue engineering, Compos. B Eng. 224 (2021), https://doi.org/10.1016/j. compositesb.2021.109192. [32] Y. Zhang, C. Li, W. Zhang, J. Deng, Y. Nie, X. Du, L. Qin, Y. Lai, 3D-printed NIRresponsive shape memory polyurethane/magnesium scaffolds with tight-contact for robust bone regeneration, Bioact. Mater. 16 (2022) 218–231, https://doi.org/ 10.1016/j.bioactmat.2021.12.032. [33] O. Andronic, O. Weiss, H. Shoman, P. Kriechling, V. Khanduja, What are the outcomes of core decompression without augmentation in patients with nontraumatic osteonecrosis of the femoral head? Int. Orthop. 45 (3) (2021) 605–613, https://doi.org/10.1007/s00264-020-04790-9. [34] G. Walters, I. Pountos, P.V. Giannoudis, The cytokines and micro-environment of fracture haematoma: current evidence, J Tissue Eng Regen Med 12 (3) (2018) e1662–e1677, https://doi.org/10.1002/term.2593. [35] S. Hassan, T. Wang, K. Shi, Y. Huang, M.E. Urbina Lopez, K. Gan, M. Chen, N. Willemen, H. Kalam, E. Luna-Ceron, B. Cecen, G.D. Elbait, J. Li, L.E. GarciaRivera, M. Gurian, M.M. Banday, K. Yang, M.C. Lee, W. Zhuang, C. Johnbosco, O. Jeon, E. Alsberg, J. Leijten, S.R. Shin, Self-oxygenation of engineered living tissues orchestrates osteogenic commitment of mesenchymal stem cells, Biomaterials 300 (2023) 122179, https://doi.org/10.1016/j. biomaterials.2023.122179.

[36] M.S. Singhvi, S.S. Zinjarde, D.V. Gokhale, Polylactic acid: synthesis and biomedical applications, J. Appl. Microbiol. 127 (6) (2019) 1612–1626, https://doi.org/ 10.1111/jam.14290.

[37] H. Chinnasami, M.K. Dey, R. Devireddy, Three-dimensional scaffolds for bone tissue engineering, Bioengineering (Basel) 10 (7) (2023), https://doi.org/10.3390/ bioengineering10070759.

[38] N.G.A. Willemen, S. Hassan, M. Gurian, J. Li, I.E. Allijn, S.R. Shin, J. Leijten, Oxygen-releasing biomaterials: current challenges and future applications, Trends Biotechnol. 39 (11) (2021) 1144–1159, https://doi.org/10.1016/j. tibtech.2021.01.007.

[39] U.K. Komal, M.K. Lila, I. Singh, PLA/banana fiber based sustainable biocomposites: a manufacturing perspective, Compos. B Eng. 180 (2020), https://doi.org/ 10.1016/j.compositesb.2019.107535.

[40] X. Yao, K. Sun, S. Yu, J. Luo, J. Guo, J. Lin, G. Wang, Z. Guo, Y. Ye, F. Guo, Chondrocyte ferroptosis contribute to the progression of osteoarthritis, J Orthop Translat 27 (2021) 33–43, https://doi.org/10.1016/j.jot.2020.09.006.

[41] A. Robson, Oxidation of miRNAs by ROS leads to cardiac hypertrophy, Nat. Rev. Cardiol. 17 (11) (2020) 678, https://doi.org/10.1038/s41569-020-00436-w.

[42] M. Liu, X. Sun, B. Chen, R. Dai, Z. Xi, H. Xu, Insights into manganese superoxide dismutase and human diseases, Int. J. Mol. Sci. 23 (24) (2022), https://doi.org/ 10.3390/ijms232415893.

[43] W. Sun, Y. Xu, Y. Yao, J. Yue, Z. Wu, H. Li, G. Shen, Y. Liao, H. Wang, W. Zhou, Self-oxygenation mesoporous MnO(2) nanoparticles with ultra-high drug loading capacity for targeted arteriosclerosis therapy, J. Nanobiotechnol. 20 (1) (2022) 88, https://doi.org/10.1186/s12951-022-01296-x.

[44] B. Yu, C.Y. Wang, Osteoporosis: the result of an 'aged' bone microenvironment, Trends Mol. Med. 22 (8) (2016) 641–644, https://doi.org/10.1016/j. molmed.2016.06.002.

[45] T. Wu, H. Shi, Y. Liang, T. Lu, Z. Lin, J. Ye, Improving osteogenesis of calcium phosphate bone cement by incorporating with manganese doped beta-tricalcium phosphate, Mater. Sci. Eng., C 109 (2020) 110481, https://doi.org/10.1016/j. msec.2019.110481.

[46] K. Ye, X. Zhang, L. Shangguan, X. Liu, X. Nie, Y. Qiao, Manganese-implanted titanium modulates the crosstalk between bone marrow mesenchymal stem cells and macrophages to improve osteogenesis, J. Funct. Biomater. 14 (9) (2023), https://doi.org/10.3390/jfb14090456.

[47] M. Wu, S. Wu, W. Chen, Y.P. Li, The roles and regulatory mechanisms of TGF-beta and BMP signaling in bone and cartilage development, homeostasis and disease, Cell Res. 34 (2) (2024) 101–123, https://doi.org/10.1038/s41422-023-00918-9.

[48] H. Li, J. Fan, L. Fan, T. Li, Y. Yang, H. Xu, L. Deng, J. Li, T. Li, X. Weng, S. Wang, R. Chunhua Zhao, MiRNA-10b reciprocally stimulates osteogenesis and inhibits adipogenesis partly through the TGF-beta/SMAD2 signaling pathway, Aging Dis 9 (6) (2018) 1058–1073, https://doi.org/10.14336/AD.2018.0214.

[49] X. Dong, B. Zhao, F.Y. Lin, C. Lu, B.N. Rogers, T.A. Springer, High integrin alpha(V) beta(6) affinity reached by hybrid domain deletion slows ligand-binding on-rate, Proc. Natl. Acad. Sci. U. S. A. 115 (7) (2018) E1429–E1436, https://doi.org/ 10.1073/pnas.1718662115.

[50] A. Armulik, G. Svineng, K. Wennerberg, R. Fassler, S. Johansson, Expression of integrin subunit beta1B in integrin beta1-deficient GD25 cells does not interfere with alphaVbeta3 functions, Exp. Cell Res. 254 (1) (2000) 55–63, https://doi.org/ 10.1006/excr.1999.4722.

[51] B.R. Barrioni, E. Norris, S. Li, P. Naruphontjirakul, J.R. Jones, M.M. Pereira, Osteogenic potential of sol-gel bioactive glasses containing manganese, J. Mater. Sci. Mater. Med. 30 (7) (2019) 86, https://doi.org/10.1007/s10856-019-6288-9.

[52] Q.M. Zhao, Y.Y. Sun, C.S. Wu, J. Yang, G.F. Bao, Z.M. Cui, Enhanced osteogenic activity and antibacterial ability of manganese-titanium dioxide microporous coating on titanium surfaces, Nanotoxicology 14 (3) (2020) 289–309, https://doi. org/10.1080/17435390.2019.1690065.

[53] P.M. Torres, S.I. Vieira, A.R. Cerqueira, S. Pina, O.A. da Cruz Silva, J.C. Abrantes, J. M. Ferreira, Effects of Mn-doping on the structure and biological properties of betatricalcium phosphate, J. Inorg. Biochem. 136 (2014) 57–66, https://doi.org/ 10.1016/j.jinorgbio.2014.03.013.

[54] J. Ye, J. Wang, J. Zhao, M. Xia, H. Wang, L. Sun, W.B. Zhang, RhoA/ROCK-TAZ Axis regulates bone formation within calvarial trans-sutural distraction osteogenesis, Cell. Signal. 121 (2024) 111300, https://doi.org/10.1016/j. cellsig.2024.111300.

[55] K. Xu, W. Fei, W. Gao, C. Fan, Y. Li, Y. Hong, R. Cui, SOD3 regulates FLT1 to affect bone metabolism by promoting osteogenesis and inhibiting adipogenesis through PI3K/AKT and MAPK pathways, Free Radic. Biol. Med. 212 (2024) 65–79, https:// doi.org/10.1016/j.freeradbiomed.2023.12.021.

[56] N. Frangogiannis, Transforming growth factor-beta in tissue fibrosis, J. Exp. Med. 217 (3) (2020) e20190103, https://doi.org/10.1084/jem.20190103.

[57] D.R. Haudenschild, E. Hong, J.H. Yik, B. Chromy, M. Morgelin, K.D. Snow, C. Acharya, Y. Takada, P.E. Di Cesare, Enhanced activity of transforming growth factor beta1 (TGF-beta1) bound to cartilage oligomeric matrix protein, J. Biol. Chem. 286 (50) (2011) 43250–43258, https://doi.org/10.1074/jbc.M111.234716.

[58] H. Newman, Y.V. Shih, S. Varghese, Resolution of inflammation in bone regeneration: from understandings to therapeutic applications, Biomaterials 277 (2021) 121114, https://doi.org/10.1016/j.biomaterials.2021.121114.

[59] Y. Yu, K. Dai, Z. Gao, W. Tang, T. Shen, Y. Yuan, J. Wang, C. Liu, Sulfated polysaccharide directs therapeutic angiogenesis via endogenous VEGF secretion of macrophages, Sci. Adv. 7 (7) (2021), https://doi.org/10.1126/sciadv.abd8217.

[60] L.Z. Zheng, H.J. Cao, S.H. Chen, T. Tang, W.M. Fu, L. Huang, D.H. Chow, Y. X. Wang, J.F. Griffith, W. He, H. Zhou, D.W. Zhao, G. Zhang, X.L. Wang, L. Qin, Blockage of src by specific siRNA as a novel therapeutic strategy to prevent destructive repair in steroid-associated osteonecrosis in rabbits, J. Bone Miner. Res. 30 (11) (2015) 2044–2057, https://doi.org/10.1002/jbmr.2542.

[61] J. Long, Z. Yao, W. Zhang, B. Liu, K. Chen, L. Li, B. Teng, X.F. Du, C. Li, X.F. Yu, L. Qin, Y. Lai, Regulation of osteoimmune microenvironment and osteogenesis by 3D-printed PLAG/black phosphorus scaffolds for bone regeneration, Adv. Sci. 10 (28) (2023) e2302539, https://doi.org/10.1002/advs.202302539.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi. org/10.1016/j.bioactmat.2024.10.019.