Exosome-mediated Piezo1 activation in 3D-printed titanium scaffolds promotes repair of femoral head osteonecrosis

Yang Yu; Zhongyin Ji; Hongjun Xu; Mingyang Ma; Zhaojing Yin; Yiyang Du; Tao Zhang; Guiguan Wang; Feng Tian; Sen Liu; Jiajia Xue; Wenwei Qian

doi:10.1186/s12951-026-04424-z

. 2026 Apr 14;24:472. doi: 10.1186/s12951-026-04424-z

Exosome-mediated Piezo1 activation in 3D-printed titanium scaffolds promotes repair of femoral head osteonecrosis

Yang Yu ^1,^#, Zhongyin Ji ^1,^#, Hongjun Xu ¹, Mingyang Ma ¹, Zhaojing Yin ², Yiyang Du ², Tao Zhang ³, Guiguan Wang ¹, Feng Tian ³, Sen Liu ^1,^✉, Jiajia Xue ^3,^✉, Wenwei Qian ^1,^✉

PMCID: PMC13200330 PMID: 41981591

Abstract

Osteonecrosis of the femoral head (ONFH) is a debilitating bone disorder characterized by ischemic degeneration with limited therapeutic options. Piezo1, a mechanosensitive ion channel, transduces physical stimuli into intracellular signals regulating osteogenesis and angiogenesis. Here, we developed a 3D-printed Ti6Al4V scaffold conformally coated with photocrosslinked hyaluronic acid hydrogel that retains and releases Piezo1-engineered exosomes (P-Exos), forming a hybrid construct (P-Exos@HAMA/Ti) with cancellous-bone-like stiffness and sustained bioactivity. In vitro studies demonstrated that P-Exos enhanced BMSC viability, migration, and osteogenic differentiation (ALP/ARS; Runx2, Osterix, OCN) while promoting endothelial proliferation and network formation (CD31, VEGFA). RNA-seq and mechanistic validation revealed Ca²⁺-responsive YAP1/β-catenin signaling, characterized by elevated intracellular Ca²⁺ levels, upregulated YAP1/β-catenin expression, and nuclear localization, underpinning the scaffold’s osteo-angiogenic synergy. In a rat ONFH model, P-Exos@HAMA/Ti restored trabecular architecture and bone volume, improved micro-CT indices (increased BV/TV, Tb.N, Tb.Th; decreased Tb.Sp), and enhanced osteogenic/angiogenic marker expression, alongside superior locomotor recovery. This multifunctional platform couples mechanical stabilization with Piezo1-mediated mechanotransduction, establishing a pro-regenerative niche integrating biomechanical support with sustained exosomal signaling for effective repair of ischemic bone defects.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04424-z.

Keywords: Piezo1-engineered exosomes, YAP1/β-catenin signaling pathway, 3D printing scaffolds, Osteonecrosis of the femoral head (ONFH)

Background

Osteonecrosis of the femoral head (ONFH), or avascular necrosis (AVN), is a debilitating condition characterized by ischemia-induced bone death in the femoral head due to compromised blood supply [1–3]. This ischemic process leads to osteocyte loss, impaired bone remodeling, and progressive subchondral collapse, ultimately resulting in severe hip pain, joint incongruity, and secondary osteoarthritis [4–7]. The condition primarily affects relatively young and active individuals, typically in their late 30s and 40s, which carries significant socioeconomic implications due to the age group’s high functional demands [8–10]. Clinically, ONFH progresses through distinct stages, from early ischemic changes to irreversible bone collapse, with pathological alterations at each stage requiring different therapeutic approaches [2, 11]. In the early stages, ischemic necrosis leads to the disruption of bone microarchitecture and diminished vascularization, impairing the femoral head’s ability to regenerate [12, 13]. As the disease progresses, subchondral bone collapse occurs, leading to progressive joint incongruity and irreversible articular destruction [5, 14–16]. Therapeutic strategies targeting the underlying pathophysiology, such as regulating the ischemic microenvironment and restoring bone remodeling, are essential for halting disease progression before structural collapse occurs. Currently, treatment options for advanced ONFH are limited, particularly for patients with end-stage disease [17–19]. Despite advances in early detection and conservative management, most ONFH cases progress to end-stage disease, necessitating total hip arthroplasty [20]. In the United States, approximately 20,000 to 30,000 new cases are diagnosed annually, accounting for around 10% of all total hip arthroplasties performed [21–23]. Conservative treatments like bisphosphonates, anticoagulants, and hyperbaric oxygen therapy show some benefit in early stages but fail to address the underlying vascular and mechanical deficiencies that drive the disease’s progression [20, 24–26]. Although interventions such as core decompression or osteotomy may provide short-term relief from intraosseous pressure, they often do not address the pathological microenvironment necessary for long-term repair [27]. Hence, a successful strategy should merge timely mechanical stabilization with precise biological direction, with specific emphasis on synchronized osteogenesis and angiogenesis, to repair the femoral head and halt disease progression.

Additive manufacturing now enables the fabrication of patient-tailored titanium alloy scaffolds with optimized porosity and interconnectivity, which mitigate stress shielding and permit tissue ingrowth. Such scaffolds can provide immediate mechanical support and facilitate bone and vessel infiltration when pore size and architecture are judiciously designed [28–30]. Recent advancements in scaffold materials have led to the development of biologically active composites, incorporating growth factors, drugs, and extracellular vesicles to enhance tissue regeneration [31, 32]. However, the therapeutic efficacy of these scaffolds often depends on precise control of their material properties and the spatial and temporal release of bioactive substances. In the challenging environment of ONFH, characterized by hypoxia and disrupted intercellular communication, titanium’s inherent bioinert properties often result in scaffolds failing to provide the necessary cues for coordinated osteogenesis and angiogenesis [33, 34]. In recent studies, researchers have endeavored to address these limitations by incorporating bioactive molecules, such as exosomes, peptides, and growth factors, into scaffolds to induce localized biological responses [35, 36]. However, the incorporation of mechanobiological signals remains a significant challenge [37]. Overcoming this obstacle necessitates a platform that combines high load-bearing capacity with sustained, localized biological signaling to facilitate osteo-angiogenic interactions and modify the lesion microenvironment to promote regeneration. Specifically, solutions that merge structural reinforcement with mechanotransductive signaling may more effectively mimic how bone and endothelium translate mechanical stimuli into regenerative processes [38].

Exosomes (Exos) serve as promising mediators of cellular communication, functioning as nanoscale extracellular vesicles that convey proteins and nucleic acids with minimal immunogenicity and superior biocompatibility [39–42]. By means of engineering, exosomes can be customized to transport precise mechanobiological cargoes, thereby bypassing the safety and production challenges associated with live cell therapies. Leveraging exosomes enables targeted and regulated delivery of therapeutic agents, offering an efficient strategy to modulate cellular reactions within a specific microenvironment [43–45]. Mechanotransduction plays a central role in skeletal homeostasis, with the mechanosensitive ion channel Piezo1 identified as a crucial mediator of membrane tension and intracellular calcium (Ca²⁺) signaling [46, 47]. Piezo1 channels are expressed on osteoblasts and endothelial cells and regulate osteoblast differentiation, matrix deposition, and endothelial responses to shear stress [48–50]. In ONFH, disruptions in load transmission and cellular dysfunction plausibly attenuate endogenous Piezo1 activity, thereby weakening both bone formation and microvascular remodeling [2, 51, 52]. These disruptions, particularly in the ischemic bone environment, exacerbate the deterioration of the femoral head, impairing tissue repair and vascularization. We reasoned that enhancing local Piezo1-related signaling via engineered exosomes could restore mechano-responsiveness in both osteogenic and endothelial compartments of the necrotic femoral head. Following this rationale, we engineered Piezo1-expressing exosomes (P-Exos), integrated them into a photocrosslinked hyaluronic acid hydrogel (HAMA), and layered the combination onto 3D-printed titanium scaffolds to form a cohesive construct termed P-Exos@HAMA/Ti6Al4V (P-Exos@HAMA/Ti). This strategy merges the immediate load-bearing structural support of Ti with sustained, localised delivery of Piezo1-engineered exosomes within a hydrated polymer network [39, 53]. Upon internalisation by recipient cells, exosome-delivered Piezo1 sensitises BMSCs and endothelial cells to intrinsic biomechanical cues within the femoral head microenvironment, activating Ca²⁺-responsive YAP1/β-catenin mechanotransductive signalling to promote coordinated osteogenesis and angiogenesis. The intent is not merely to provide transient paracrine cues, but to establish a persistent interface capable of orchestrating Piezo1-dependent mechanosensory signaling, enhancing local cell survival, and directing lineage commitment and vascularization in situ. Beyond demonstrating functional efficacy, mechanistic understanding is essential for translational readiness; accordingly, our study incorporated discovery-scale profiling to elucidate how Piezo1-linked cues are transduced at the transcriptomic and metabolic levels.

In summary, we have developed a multifunctional, bioengineered construct—P-Exos@HAMA/Ti—that provides mechanical support and sustained Piezo1-mediated mechanobiological instruction within the necrotic femoral head (Scheme 1). By integrating Piezo1-engineered exosomes within a hydrogel-coated titanium scaffold, this platform addresses the dual challenges of biomechanical instability and impaired biological signaling that typify ONFH. Mechanistically, the system activates a Ca²⁺-dependent YAP1/β-catenin pathway, promoting synchronized osteogenic and angiogenic responses essential for bone regeneration. Preclinical validation in a rat ONFH model confirmed its therapeutic efficacy and systemic biosafety. These findings underscore the therapeutic synergy achieved by coupling structural reinforcement with localized mechanobiological instruction. The modularity of the scaffold and tunable exosome-hydrogel interface offer a clinically adaptable strategy for broader orthopedic applications. Taken together, our work establishes P-Exos@HAMA/Ti as a translationally promising platform that bridges mechanobiology and regenerative bioengineering for effective treatment of ONFH and other load-bearing skeletal defects.

Scheme 1 — Schematic illustration of engineering and mechanism of Piezo1-exosome scaffold therapy for ONFH. (a) Workflow illustrating the electroporation-based preparation of P-Exos and their incorporation into HAMA hydrogel-coated titanium scaffolds (P-Exos@HAMA/Ti) for sustained, controlled exosome release. (b) Sustained release of P-Exos from the scaffold promotes coordinated regeneration, including osteogenic differentiation of BMSCs characterized by upregulated OCN, RUNX2, and Osterix expression, and angiogenic activation of VECs accompanied by enhanced VEGFA and CD31 levels. (c) Mechanistic pathway showing P-Exos-induced Ca²⁺ influx and Piezo1 upregulation, activating YAP1 and β-catenin signaling cascades. (d) In a rat model of ONFH, implantation of P-Exos@HAMA/Ti scaffolds facilitates trabecular bone regeneration, enhances neovascularization, and mitigates necrotic progression, underscoring their therapeutic efficacy

Results and discussion

Design and characterization of the P-Exos@HAMA/Ti platform

Selective laser melting (SLM) was utilized to fabricate 3D-printed titanium alloy scaffolds (Ti, Ti6Al4V) with an open, interconnected lattice architecture. The equivalent pore size determined using the equivalent-circle method was 500 μm, while the filament diameter was 300 μm and the overall porosity of the structure was 67.3% (Fig. 1a). Melt-pool textures and sparsely distributed microspheres generated micro-scale roughness favorable for cell anchorage. Ti6Al4V ELI lattices were printed from 15 to 53 μm powders (270 W, 2100 mm s^–1, 0.10 mm hatch, 30 μm layer). Post-processing comprised heat treatment at 840 ℃ for 2 h, followed by furnace-cooled to 500 ℃, and ultrasonically cleaned sequentially in acetone, ethanol, and deionized water to remove residues and loosely bound particles.

For hydrogel coating and exosome loading, methacrylated hyaluronic acid (HAMA; 0.75–1.5% w/v in PBS with 0.25% w/v LAP) containing Exos or P-Exos was infiltrated into the lattice under mild vacuum and photocrosslinked (405 nm, 10 mW cm^–2, 15–20 s), followed by rinsing with PBS to remove unbound vesicles, yielding Exos@HAMA/Ti or P-Exos@HAMA/Ti. Upon photopolymerization, HAMA formed a uniform porous coating on struts and within the pores (Fig. 1b–c). For in vivo implantation, a 1 mm-diameter Ti variant was additionally fabricated, and representative SEM images at this scale confirmed lattice fidelity (Fig. S1). This trabeculae-scale, hydrogel-coated lattice was designed for localized exosome retention/release and mass transport: pore sizes of several hundred micrometers approximate trabecular length scales to facilitate osteogenic–angiogenic coupling [54]. An interconnected porosity of 60–70% supports convective–diffusive transport and local exosome mobility, while struts of 300 μm provide an appropriate balance between printing fidelity and mechanical stability, as well as sufficient surface area for hydrogel interlocking. This design yields an apparent elastic modulus of approximately 2.8 GPa, closer to cancellous bone than to bulk Ti6Al4V (≈ 110 GPa), thereby mitigating stress shielding [55].

Elemental mapping identified Ti, Al, and V as the primary elements, with uniform distributions of C, N, and O across the coated regions (Fig. 1d). EDS spectrum confirmed the expected elemental composition of the coated construct, with Ti as the dominant element (56.9 wt%) and trace Fe content (0.3 wt%) showing no spatial aggregation (Fig. 1e). Uncoated scaffolds showed the expected Ti–Al–V spectrum without extraneous peaks, confirming the material integrity of the base Ti prior to coating (Fig. S2). Micro-CT reconstructions confirmed the open-pore network and strong interconnectivity (Fig. 1f), consistent with SEM (Fig. S1). At the implantation scale, the 1 mm-diameter Ti retained continuous channels, demonstrating that architectural patency was preserved during scaling down (Fig. S3). To evaluate mechanical performance under physiological loads, geometrically scaled coupons (Φ10 × 20 mm; 0.5 mm min^–1; n = 5) with matched topology and density displayed consistent load–displacement profiles, with an ultimate load of 7929 ± 382 N, stiffness of 10920 ± 248 N mm^–1, and an apparent modulus of 2.78 ± 0.06 GPa, indicative of cancellous-bone-like stiffness (Fig. 1g–h, S4 and Table S1) [56–58]. XRD analysis further showed that the diffraction patterns of bare Ti and P-Exos@HAMA/Ti were essentially consistent, indicating that the coating process preserves the Ti6Al4V scaffold’s crystalline phase composition (Fig. 1i). FTIR analysis confirmed chemical integrity at the hydrogel–metal interface. HAMA exhibited characteristic bands at 3259 cm^–1 (O–H stretching), 2929/2865 cm^–1 (C–H asymmetric stretching), 1614 cm^–1 (C=O stretching), 1403 cm^–1 (CH bending), and 1032 cm^–1 (C–O stretching). The Ti backbone displayed a band near 707 cm^–1, corresponding to Ti6Al4V lattice vibrations. Importantly, the composite retained all spectral features without new peaks, indicating preserved hydrogel chemistry and stable interfacial integration (Fig. 1j).

HAMA coating markedly enhanced surface wettability. Static contact angles decreased relative to bare Ti (Fig. 1k), with sessile-drop profiles approximately 53–54° (Fig. 1l). Macroscopic imaging verified sol-to-gel transition and a self-supporting film suitable for conformal coating (Fig. 1m and Fig. S5). Uniaxial compression curves for 0.75%, 1.0%, and 1.5% w/v HAMA (Fig. 1n) showed a monotonic increase in stiffness and yield stress with concentration, indicating lattice reinforcement without loss of elastic–plastic behavior. Collectively, these findings confirm improved surface wettability and process-compatible gelation at the scaffold interface compared to the uncoated control. To determine the optimal formulation balancing interfacial strength and controlled release, 0.75%, 1.0%, and 1.5% w/v HAMA gels were evaluated using oscillatory rheology and exosome-release assays (Fig. 1o–q). Small-amplitude shear measurements (G’ > G’’) verified solid-like viscoelastic behavior, with the storage modulus increasing proportionally to polymer content (Fig. 1o–p). The release profiles showed an initial burst followed by sustained release; higher solids slowed release, with 1.5% w/v providing the longest duration (Fig. 1q), while lower concentrations offered faster initial release and lower stiffness. This tunability supports early cytoprotection/migration at lower modulus and later differentiation/osteo-angiogenic remodeling at higher modulus. While the post-release bioactivity was not directly quantified here, reports on hyaluronic-acid–based and methacrylated-HA hydrogels suggest that encapsulation with controlled release can help preserve extracellular-vesicle integrity and function (typically 2–8 weeks), supporting osteogenic/angiogenic readouts in vitro and in vivo [53, 59–62]. Based on the balance between sustained release duration and mechanical reinforcement, 1.5% w/v HAMA was selected for all subsequent in vitro and in vivo studies.

Overall, P-Exos@HAMA/Ti establishes a load-bearing yet biologically instructive interface, combining a trabecular-scale Ti framework with a hydrophilic, adherent hydrogel coating. The resulting structure promotes cell infiltration, stable tissue integration, and localized exosome delivery, providing a seamless transition from in vitro characterization to in vivo validation.

Engineering and characterization of P-Exos

To enable controlled delivery of Piezo1, we engineered and characterized Piezo1-loaded exosomes (P-Exos) for subsequent applications (Fig. 2a). For vector construction, a pcDNA3.4-based expression plasmid (Tsingke Biotechnology, Beijing) encoding the full-length Piezo1 open reading frame (ORF) was designed (Fig. 2b). Due to the large size and structural complexity of the gene, a modular two-segment assembly strategy was adopted, in which the ORF was divided into A and B fragments for independent synthesis, PCR verification, and recombination into the vector backbone. A comprehensive primer panel spanning the 5’/3’ vector-insert junctions, internal ORF regions, and the A-B junction enabled full construct coverage. Representative sequencing chromatograms are shown in Fig. 2c, and the complete primer list is provided in Table S2. Gel analysis revealed distinct PCR products for A and B fragments during pre-assembly (Fig. 2d), consistent with the two-segment design. Gel-purified A/B inserts (Fig. S6) were utilized for recombination as part of the initial quality-control step before full-length assembly. Positive colonies were confirmed by Sanger sequencing using primers listed in Table S2 to validate construct identity.

Once the genetic payload had been molecularly qualified, BMSC-derived exosomes were isolated as the loading substrate. Transmission electron microscopy revealed membrane-bound vesicles with the expected morphology (Fig. 2e), and post-loading specimens retained comparable ultrastructural features (Fig. 2f). Nanoparticle tracking analysis revealed unimodal size distributions with a peak diameter of approximately 100 nm for both Exos and P-Exos (Fig. 2g–h), confirming that electroporation did not alter the vesicle size profile.

For genetic cargo incorporation, electroporation was employed to achieve loading. Four preset programs—Program I (CA-137), Program II (DS-120), Program III (CM-137), and Program IV (CM-150)—were evaluated in Solution SF using a 4D-Nucleofector system across distinct program families recommended by the vendor (Table S3). Program II (DS-120) yielded the optimal balance between loading efficiency and vesicle integrity, with a mean loading efficiency of 37.2% (Fig. 2i). To determine the absolute plasmid payload per vesicle—a quantity more directly relevant to dose-controlled delivery than bulk loading efficiency—we performed ddPCR on three independently prepared P-Exos batches, yielding a mean of 144.4 ± 26.3 copies/vesicle (Fig. S7). Unloaded Exos and blank controls produced negligible background signal, confirming that the measured signal reflects genuine intravesicular plasmid rather than surface-adsorbed or carry-over nucleic acid. The ddPCR-derived encapsulation rate (approximately 33% of input plasmid copies) is consistent with the 37.2% efficiency determined by bulk fluorometric quantification, with the two independent methods together confirming the reliability of the loading characterisation. Following electroporation, DNase digestion and subsequent cleanup were performed immediately to remove unencapsulated nucleic acid and minimize background signal, thereby ensuring interfacial stability and compatibility with the controlled-release design of the platform. Subsequently, vesicle preparations were analyzed for surface and population-level characteristics before and after loading. Zeta potential and dispersion indices indicated stable surface charge and colloidal uniformity under all conditions (Fig. 2j–k). At the molecular level, canonical exosomal markers (CD9, CD63, TSG101) were detected in vesicle lysates, while Piezo1 expression was evaluated in BMSC lysates after co-culture with Exos or P-Exos. Relative to unloaded Exos, P-Exos increased cellular Piezo1 protein, confirming successful construction and delivery (Fig. 2l). Time-course imaging of uptake from 1.5% w/v HAMA at 0, 24, and 72 h (Fig. 2m–n), along with quantitative fluorescence analysis (Fig. 2o), showed comparable internalization, indicating that Program II–based electroporation did not measurably impair cellular uptake within the observation window.

To provide direct evidence for membrane localisation of the delivered Piezo1, co-immunofluorescence staining with 594-Phalloidin and an anti-Piezo1 antibody was performed across four groups (Control, Exos, P-Exos, and Yoda1 + Exos). Piezo1 signal was negligible in the Control group and diffusely cytoplasmic in the Exos group. In contrast, P-Exos-treated BMSCs displayed markedly elevated Piezo1 signal co-localising with the cortical F-actin network at the cell periphery, a pattern reproduced in the Yoda1 + Exos group, consistent with membrane-proximal localisation of pharmacologically activated Piezo1 channels (Fig. S8).

Altogether, the findings outline an integrated workflow from vector design to post-loading qualification. The A/B segmented assembly with broad primer coverage secures cassette identity, while the exosome substrate preserves morphology and population homogeneity compatible with loading. Program II offers a workable balance between incorporation and vesicle integrity. After loading, the vesicles maintain stable interfacial attributes and canonical markers with a detectable Piezo1 signal and are efficiently internalized by recipient cells. In sum, the preparation constitutes a defined and reproducible vesicle carrier suitable for controlled delivery in the studies that follow.

In vitro cytocompatibility and baseline cellular performance

To assess cytocompatibility and the preservation of core cellular functions pertinent to later osteogenic and angiogenic analyses, we implemented a seven-group design in bone mesenchymal stem cells (BMSCs), comprising (I) Control, (II) Ti, (III) HAMA, (IV) Exos@HAMA/Ti, (V) P-Exos@HAMA/Ti, (VI) Yoda1 + Exos, and (VII) Normal. In this study, the term “Control” denotes dexamethasone (Dex)-treated BMSCs without scaffold or exosome exposure, representing a pathophysiological baseline. Conversely, “Normal” indicates BMSCs cultured in standard medium without Dex, representing a physiological baseline. Dex was administered to Groups I–VI to emulate ONFH-relevant glucocorticoid stress and provide a uniform injury background; Group VII served solely as an unchallenged reference. Yoda1 was included to benchmark Piezo1 activation independently of exosomal delivery.

In the 24-hour CCK-8 assays, metabolic activity was found to be similar across all seven groups, with live-cell viability remaining within a comparable range under Dex challenge. This collectively confirms the cytocompatibility of the scaffold, hydrogel, and exosome components. Particularly noteworthy is the statistically significant increase in the P-Exos@HAMA/Ti group compared to the Control (p < 0.05), indicating a mild but reproducible cytoprotective effect (Fig. 3a). Over one week, P-Exos@HAMA/Ti (Group V) maintained a continuous and progressive rise in cell proliferation (Fig. 3b), consistent with durable viability for downstream lineage evaluation. The overall stability of CCK-8 readouts across groups establishes the material system as cytocompatible, providing a reliable foundation for subsequent functional analyses.

Fig. 3 — In vitro cytocompatibility and baseline cellular responses of BMSCs to P-Exos@HAMA/Ti scaffolds. (a) CCK-8 viability assay showing cytocompatibility across seven groups. (b) Proliferation curve of BMSCs treated with P-Exos@HAMA/Ti over 7 days. (**c–d**) Live/Dead staining (Calcein-AM green/PI red) after 24 h showing viability ratios and representative fields. (**e–f**) DCFH-DA staining and quantification of mean fluorescence intensity (MFI) showing intracellular ROS modulation among seven groups. (**g–h**) Flow cytometry histograms and quantification of fluorescence intensity showing ROS levels in BMSCs across different treatment groups. (**i–j**) Scratch-wound images and closure analysis at 0 and 24 h demonstrating enhanced cell migration on P-Exos@HAMA/Ti. (**k–l**) Transwell migration assay and quantification showing increased migratory BMSCs under P-Exos@HAMA/Ti treatment. I: Control; II: Ti; III: HAMA; IV: Exos@HAMA/Ti; V: P-Exos@HAMA/Ti; VI: Yoda1 + Exos; VII: Normal. Groups I–VI were treated with Dex, while Group VII served as the untreated normal control. Data are mean ± SD; ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Live/Dead staining further corroborated the excellent cytocompatibility of all constructs. Live-cell ratios exceeded 90% across all seven groups, with fluorescence quantification revealing only a marginal increase in Groups V–VII relative to Groups I–III (Fig. 3c–d). These results confirm that neither the Dex challenge nor any scaffold or exosome component induced appreciable cytotoxicity under the experimental conditions. Concordantly, ROS imaging using DCFH-DA demonstrated attenuated fluorescence in exosome-exposed groups (IV–VI) relative to non-exosome comparators (I–III) (Fig. 3e–f). Flow cytometry analysis confirmed these findings, showing a leftward shift in fluorescence intensity distribution and a reduced median fluorescence intensity (MFI) in Groups IV–VI versus Groups I–III (Fig. 3g–h). Functional preservation was further verified through motility assays. In scratch-wound assays, exosome-coated groups displayed faster gap closure than the Dex control, with P-Exos@HAMA/Ti exhibiting sustained edge migration over 24 h despite inhibitory stressors (Fig. 3j), with quantitative wound closure analysis confirming significantly higher migration ratios in exosome-bearing groups (Fig. 3i). Consistent trends were observed in Transwell migration assays (Fig. 3k–l), indicating that the composite formulation preserves chemotactic responsiveness and adhesion–de-adhesion dynamics. Together, aligned improvements across viability, redox, and motility point to reduced oxidative burden and preserved cell–matrix interaction under Dex.

These results establish P-Exos@HAMA/Ti as cytocompatible, maintaining fundamental cellular functions such as viability, redox balance, and motility consistently across all experimental conditions. The convergence of these readouts with the interface attributes established earlier—enhanced wettability and conformal HAMA coverage—and with maintained vesicle uptake provides a coherent basis for subsequent osteogenic and angiogenic evaluations.

In vitro osteogenic and angiogenic responses

To elucidate the ability of P-Exos@HAMA/Ti to support both osteogenesis and angiogenesis—two tightly linked processes essential for bone regeneration—we examined BMSCs and rat-derived vascular endothelial cells (VECs) across a seven-group experimental framework. Under osteogenic induction, the readouts displayed a consistent hierarchy across groups (Fig. 4a–d). Ti or HAMA alone produced the weakest ARS and ALP staining, whereas incorporation of exosomes (Exos@HAMA/Ti) markedly enhanced signal intensity. Among Dex-challenged groups, P-Exos@HAMA/Ti exhibited the highest levels, with Yoda1 + Exos showing a comparable enhancement. These phenotypic gains were accompanied by upregulated expression of osteogenic markers—Runx2, Osterix, and OCN—at both protein and mRNA levels, along with a distinct Piezo1 band detected in P-Exos@HAMA/Ti lysates (Fig. 4e–j).

Fig. 4 — In vitro osteogenic differentiation of BMSCs treated with P-Exos@HAMA/Ti. (**a–b**) ARS staining and quantification showing enhanced matrix mineralization in BMSCs treated with P-Exos@HAMA/Ti. (**c–d**) ALP staining and quantification indicating elevated early osteogenic activity after P-Exos@HAMA/Ti treatment. (**e–f**) Western blot analysis and relative gray values of Piezo1, Runx2, Osterix, and OCN in BMSCs across different experimental groups. (**g–j**) RT-qPCR analysis of Piezo1, Runx2, Osterix, and OCN mRNA expression in BMSCs after different treatments. (**k–n**) CLSM images showing enhanced Piezo1 expression with pericellular distribution, stronger nuclear localisation of Osterix and Runx2, and greater OCN deposition in BMSCs after P-Exos@HAMA/Ti stimulation. I: Control; II: Ti; III: HAMA; IV: Exos@HAMA/Ti; V: P-Exos@HAMA/Ti; VI: Yoda1 + Exos; VII: Normal. Groups I–VI were treated with Dex, while Group VII served as the untreated normal control. Results are presented as mean ± SD. Statistical significance was calculated with two-tailed t-test; ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Immunofluorescence analysis likewise revealed enhanced Piezo1 expression, characterized by a more continuous pericellular and cytoplasmic distribution, stronger nuclear localization of Runx2 and Osterix, and greater extracellular OCN deposition under exosome-bearing conditions. These features were supported by a higher nuclear-to-cytoplasmic intensity ratio for Runx2/Osterix, more continuous OCN-positive matrices along cell–cell interfaces, and intensified Piezo1 signals. Among the groups, P-Exos@HAMA/Ti exhibited the most uniform and pronounced staining patterns, followed by Yoda1 + Exos and Exos@HAMA/Ti, whereas Ti or HAMA alone showed minimal signals. The Normal group provided the upper reference bound (Fig. 4k–n and Fig. S9). This hierarchy mirrored the ALP and ARS outcomes observed under osteogenic induction. Mechanistically, these findings are consistent with Piezo1-mediated mechanotransduction in mesenchymal lineages, where channel activation—including pharmacologic stimulation by Yoda1—promotes osteogenic differentiation and biases MSC fate toward an osteoblastic phenotype [63–65]. In parallel, the previously observed reduction in intracellular ROS (Fig. 3e–h) provides a permissive context for osteogenesis, as excessive oxidative stress impairs MSC differentiation, and redox normalization would be expected to favor lineage progression under osteoinductive cues [64, 66–69]. Together, the concordant gains in ALP/ARS, osteogenic transcripts/proteins, and nuclear marker distribution support a bona fide differentiation effect rather than a nonspecific metabolic shift.

Because osteogenesis and angiogenesis are interdependent in bone repair, we next evaluated endothelial cell behavior through proliferation, gene/protein expression, and tube-formation assays. P-Exos@HAMA/Ti exhibited progressively higher CCK-8 values than the Control from day 3 onwards, reaching statistical significance at days 3, 5, and 7 (Fig. 5a). EdU staining confirmed greater DNA synthesis in P-Exos@HAMA/Ti-treated VECs, with a significantly higher percentage of EdU-positive cells compared with all other Dex-challenged groups (Fig. 5b–c). VEGFA and CD31 expression levels were elevated at both the mRNA and protein levels (Fig. 5d–g). On Matrigel, endothelial morphogenesis was significantly enhanced under exosome-coated conditions, with significantly greater total tube length and junction number in P-Exos@HAMA/Ti-treated cells compared with all other Dex-challenged groups (Fig. 5h–j). Immunofluorescence staining of VEGFA and CD31 (Fig. 5m–n) and corresponding MFI quantification (Fig. 5k–l) were consistent with these findings. These results align with the established role of Piezo1 as a mechanosensor in endothelial cells, mediating angiogenic remodeling in response to shear stress and extracellular matrix interactions [70]. Yoda1 was again used as an internal comparator for Piezo1 activation, serving as a pathway-oriented reference under identical conditions [71, 72].

Fig. 5 — In vitro angiogenesis and cellular responses to P-Exos@HAMA/Ti. (a) Time-course CCK-8 assay of VECs viability with or without P-Exos@HAMA/Ti. (**b–c**) Images and percentage of EdU-positive cells in VECs after different treatment. (**d–e**) Western blot analysis and relative gray values of CD31 and VEGFA in VECs across different experimental groups. (**f–g**) RT-qPCR analysis of VEGFA and CD31 mRNA levels in VECs, showing significant upregulation in the P-Exos@HAMA/Ti group. (h) Matrigel tube formation assay (Calcein-AM fluorescence and bright field) showing enhanced tubular network formation after P-Exos@HAMA/Ti exposure. (**i–j**) Quantification of the total tube length and number of junctions through different treatments in VECs. (**k–n**) Representative immunofluorescence (DAPI counterstain) and MFI quantification showing intensified VEGFA and CD31 expression in VECs. Results are presented as mean ± SD. Statistical significance was calculated with a two-tailed t-test; ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Across the osteogenic and angiogenic datasets, the remarkable synergistic performance of P-Exos@HAMA/Ti can be attributed to several interrelated design features. The hydrophilic and conformal HA-based hydrogel coating on the porous lattice ensures tight cell–material contact and localized bioactivity, enabling effective paracrine communication between endothelial and osteogenic cells that strengthens osteo-angiogenic coupling [73]. This architecture also facilitates nutrient transport and mechanical compatibility, creating a microenvironment favorable for concurrent vascular and bone tissue growth. Moreover, the HA matrix provides sustained retention and controlled release of exosomes, extending their local bioavailability and maintaining biological activity over time—a profile consistent with our multi-day observations [53, 74, 75]. Such sustained signaling likely contributes to the coordinated activation of angiogenic and osteogenic pathways. In addition, by maintaining efficient exosome internalization under Dex-induced stress and modulating redox balance at the material interface, the composite supports cellular reprogramming toward lineage-specific functions—matrix synthesis and mineralization in BMSCs and tubulogenesis in endothelial cells—without requiring external growth factors that could disrupt the intrinsic osteo–angiogenic equilibrium [76].

Collectively, these attributes position P-Exos@HAMA/Ti as a multifunctional, bioactive interface that integrates and harmonizes osteogenic and angiogenic cues within a single platform. The findings support a mechanistic model in which HA-based hydrogels act as localized reservoirs for spatiotemporally controlled exosome delivery, providing a strong rationale for subsequent in vivo validation.

RNA-seq and metabolomics reveal Ca²⁺-YAP1/β-catenin activation

To investigate the osteogenic and angiogenic effects of P-Exos@HAMA/Ti and understand the underlying mechanisms, we conducted bulk RNA-seq and untargeted metabolomics analyses on the Control and P-Exos groups exposed to Dex challenge. Our omics profiling provided an unbiased molecular portrait of the Piezo1-mediated networks. Mechanistic validation was then carried out across five groups—(I) Control, (II) VP + P-Exos, (III) P-Exos, (IV) Yoda1 + Exos, and (V) Normal. Groups I–IV were Dex-stimulated, while Group V served as an untreated baseline. VP was used as a YAP inhibitor to confirm pathway specificity.

Bulk RNA-seq revealed substantial transcriptomic remodeling between the P-Exos and Control groups, with 1,143 upregulated and 846 downregulated differentially expressed genes identified (|log₂FC| ≥ 1, adjusted p < 0.05), as shown by the volcano plot and hierarchical clustering heat map (Fig. 6a–b). Gene-level heat maps highlighted the upregulation of osteogenic, angiogenic, and mechanotransductive genes, including Tek (Tie2), Ibsp, Alpl, Col1a1/Col1a2, Ctnnb1, Tcf7, Tead4, Wnt5a, Axin2, Ccnd1, Myc, Peg3, Ccdc80, Rgs2, Stc1, Ccl20, Arg1, Il11/Il6, Ptx3, and Ednrb, indicating ECM remodeling with Wnt/YAP crosstalk (Fig. 6b). GSEA confirmed significant enrichment of the calcium signaling pathway (p < 0.0001, q = 0.2324), supporting Ca²⁺ influx–dependent signaling in the P-Exos group (Fig. 6c). GO enrichment analysis identified focal adhesions, collagen-containing ECM, adherens junctions, and the actin cytoskeleton as enriched cellular components, while molecular functions emphasized integrin binding, ECM structural constituents, and Wnt receptor activity; biological processes prominently featured focal adhesion, integrin binding, actin remodeling, lamellipodia, stress fibers, angiogenesis, and osteoblast differentiation, highlighting the integration of mechanical cues with lineage specification (Fig. 6d–f). KEGG pathway mapping implicated focal adhesion, ECM–receptor interaction, actin cytoskeleton regulation, Wnt, Hippo, and TGF-β signaling, together with PI3K–Akt/MAPK, JAK–STAT, and HIF-1, as well as immune-modulatory and metabolic circuits such as TNF/IL-17 signaling, glycolysis/gluconeogenesis, and complement cascades (Fig. 6g). These findings collectively delineate a Ca²⁺-responsive adhesion–cytoskeleton–transcription relay consistent with Piezo1-mediated mechanotransduction.

Fig. 6 — RNA-seq reveals Ca²⁺-YAP1/β-catenin activation in P-Exos@HAMA/Ti-treated BMSCs. (a) Volcano plot showing differentially expressed genes between P-Exos and control. (b) Heatmap clustering of differentially expressed genes, clustering BMSCs exposed to P-Exos and control, with specific upregulation in osteogenic and angiogenic pathways. (c) GSEA highlighting enrichment of the calcium-signaling pathway in the P-Exos group. (**d–f**) GO enrichment for cellular components (focal adhesion/ECM), molecular functions (integrin binding/Wnt receptor activity), and biological processes (angiogenesis/osteoblast differentiation/cytoskeleton regulation). (g) KEGG enrichment analysis showing pathways related to focal adhesion, actin cytoskeleton regulation, ECM-receptor interaction, and calcium signaling. (h) Immunofluorescence images showing higher calcium influx in P-Exos-treated BMSCs. (i) Western blot analysis showing increased YAP1/β-catenin expression in P-Exos-treated BMSCs. (j) mRNA quantification showing upregulation of YAP1 and β-catenin in P-Exos groups. (**k–l**) Immunofluorescence showing enhanced nuclear translocation of YAP1 and β-catenin in P-Exos-treated BMSCs. I: Control; II: VP + P-Exos; III: P-Exos; IV: Yoda1 + Exos; V: Normal. Groups I–IV were exposed to Dex, while Group V was given standard medium. Results are presented as mean ± SD. Statistical significance was calculated with two-tailed t-test; ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Network-level analysis provided further mechanistic resolution. A protein–protein interaction (PPI) network analysis revealed a mechanotransduction–adhesion module centered on CAV3 and CAVIN3, coupled with Wnt/Hippo regulators WIF1 and MAP4K2. Upregulation of CAV3 and CAVIN3 together with downregulation of WIF1 and MAP4K2 indicates activation of the YAP1/β-catenin axis. Vascular-associated receptors P2RY1 and EDNRB were also incorporated, reinforcing pro-angiogenic signaling (Fig. S10). The resulting PPI topology functionally connects the transcriptomic signatures to the phenotypic osteo-angiogenic effects.

Untargeted metabolomics revealed mechanically activated anabolic signatures (Fig. S11). Pathway analysis indicated enrichment in the TCA cycle, glycolysis/gluconeogenesis, and amino acid biosynthesis (notably alanine, aspartate, glutamate, glycine, serine, and threonine), reflecting enhanced bioenergetic and biosynthetic flux. Upregulated nicotinate/nicotinamide metabolism and purine/pyrimidine turnover suggested metabolic activation, while increased glycerophospholipid metabolism implied active membrane remodeling linked to cytoskeletal reorganization and mechanotransduction.

To further confirm that these transcriptomic and metabolomic signatures reflect Piezo1-specific effects rather than generic exosome cargo, additional pairwise comparisons between Control and Exos groups, and between Exos and P-Exos groups, were performed for both transcriptomic and metabolomic datasets (Fig. S12–S14). At the transcriptomic level, the Control vs. Exos comparison revealed moderate activation of osteogenic and angiogenic pathways, while the Exos vs. P-Exos comparison demonstrated a substantially amplified response, including specific enrichment of mechanosensory processes (response to mechanical stimulus) and Piezo1 downstream targets (Ctnnb1, Tcf7, Tead4). At the metabolomic level, Piezo1-specific metabolic reprogramming became progressively more pronounced from the Exos to the P-Exos group. Collectively, these analyses confirm that the transcriptomic and metabolomic signatures reported are predominantly attributable to Piezo1 overexpression rather than generic exosome cargo effects.

Following omics guidance, we examined the Ca²⁺-responsive YAP1/β-catenin signaling axis. Live-cell Ca²⁺ imaging revealed markedly elevated intracellular calcium levels in the P-Exos group relative to the Control. VP treatment substantially attenuated this elevation, while Yoda1 + Exos produced a comparable Ca²⁺ response to P-Exos, and the Normal group served as an untreated physiological reference (Fig. 6h). YAP1 and β-catenin expression increased with P-Exos and decreased with VP at the mRNA and protein levels (Fig. 6i–j). The reciprocal pattern observed with Yoda1 + Exos and VP + P-Exos supports a Piezo1-dependent but partially YAP-mediated activation cascade. Confocal imaging demonstrated pronounced nuclear accumulation of YAP1 and β-catenin in P-Exos-treated cells, which was largely reversed to cytoplasmic localisation upon VP treatment (Fig. 6k–l). These patterns emphasize the transcriptional roles of YAP1/β-catenin in mechanotransduction, in agreement with reports linking YAP nuclear translocation to osteogenic differentiation in response to mechanical signals [77, 78].

Mechanistically, these convergent data support a model in which P-Exos elevate intracellular Ca²⁺, activate YAP1, and promote nuclear translocation, where it cooperates with β-catenin to drive osteogenic and angiogenic gene expression under glucocorticoid challenge (Fig. 6). Partial VP inhibition confirms YAP-dependent contribution. This model aligns with RNA-seq evidence of calcium/adhesion–cytoskeleton enrichment, PPI-derived caveolae/Wnt–Hippo modules, and metabolomic indicators of enhanced energy remodeling.

In vivo biosafety and therapeutic evaluation in ONFH model

To comprehensively evaluate the in vivo biosafety and therapeutic performance of P-Exos@HAMA/Ti scaffold, a sequence of animal studies was performed in Sprague–Dawley (SD) rats, including hematological and histopathological analyses, fluorescence-based biodistribution, micro-CT and histological evaluation of bone regeneration, and behavioral assessments for functional recovery.

For the acute-phase biosafety assessment, six SD rats were divided into two groups: P-Exos@HAMA/Ti implantation (n = 3) and sham-operated control (n = 3). Hematological and biochemical parameters measured seven days post-surgery remained within physiological ranges, with no significant differences between the two groups. Routine blood indices (RBC, HGB, HCT, MCV, MCH, MCHC, PLT, LYMPH, MON, GRAN) and organ function markers (ALT, AST, ALP, BUN, CREA, UA) all indicated normal liver and kidney function, confirming the absence of systemic toxicity. These findings are consistent with the favorable biosafety profiles of hydrogel-based delivery platforms (Fig. S15). Histopathological evaluation further showed excellent in vivo compatibility. Hematoxylin and eosin staining of major organs (heart, liver, spleen, lung, kidney) revealed intact tissue architecture without inflammation, necrosis, or fibrosis (Fig. S16). No granulomatous or immune activation was observed. To assess in vivo exosome release kinetics, rats were injected with fluorescence-labeled P-Exos-HAMA or Exos-HAMA into the proximal femoral cavity (n = 3 per group). In vivo fluorescence imaging at days 3, 6, and 9 demonstrated a gradual decline in fluorescence, indicating sustained exosome release (Fig. S17). Comparable profiles between the two groups suggest that Piezo1 modification did not alter exosomal retention, consistent with hydrogel encapsulation prolonging exosome persistence.

For therapeutic efficacy validation, an ONFH rat model was induced using a sequential lipopolysaccharide and methylprednisolone protocol (Fig. 7a), with animals divided into five groups: (I) Control, (II) Ti only, (III) Exos@HAMA/Ti, (IV) P-Exos@HAMA/Ti, and (V) Normal group (n = 3). Micro-CT imaging at 8 weeks post-surgery showed that P-Exos@HAMA/Ti induced robust trabecular regeneration and preserved femoral head morphology, whereas Groups I–III exhibited osteolysis, trabecular loss, and partial collapse (Fig. 7b). BV/TV and BS/TV were significantly higher in Group IV, while BS/BV was reduced (Fig. 7c-e). Tb.N and Tb.Th were likewise elevated, and Tb.Sp was markedly decreased (Fig. 7f-h), collectively approaching normal values and indicating near-physiological trabecular restoration. At 4 weeks, micro-CT already showed early trabecular reconstruction in Group IV (Fig. S18 and S19), with early BV/TV and Tb.N increases confirming rapid repair initiation.

Fig. 7 — In vivo evaluation of P-Exos@HAMA/Ti in an ONFH rat model. (a) The experimental design for evaluating P-Exos@HAMA/Ti in a rat ONFH model involves scaffold implantation and assessment at 4 and 8 weeks after surgery. (b) Micro-CT reconstructions at 8 weeks showing that P-Exos@HAMA/Ti markedly restored trabecular continuity and prevented femoral head collapse, indicating effective bone regeneration. (**c–h**) Quantitative micro-CT indices (increased BV/TV, BS/TV, Tb.N, Tb.Th and reduced BS/BV, Tb.Sp) confirm superior bone microarchitecture recovery in the P-Exos@HAMA/Ti group. (n = 3; I: Control; II: Ti; III: Exos@HAMA/Ti; IV: P-Exos@HAMA/Ti; V: Normal). Results are presented as mean ± SD. Statistical significance was calculated with a two-tailed t-test; ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Histological findings corroborated the radiological results (Fig. 8a–b). Masson’s trichrome staining revealed dense collagen deposition and organized bone lamellae in Group IV, contrasting with the disordered and necrotic trabeculae in Groups I–III. H&E staining confirmed abundant new bone tissue, intact marrow cavities, and osteoblast-lined trabecular surfaces, indicative of active remodeling. Immunohistochemistry further demonstrated elevated expression of osteogenic markers (Runx2, Osterix) and the mechanosensitive protein Piezo1, reflecting mechanotransductive activation (Fig. 8a). Angiogenic markers (CD31, VEGFA) were likewise upregulated, consistent with coordinated osteogenesis and angiogenesis (Fig. 8b). At 4 weeks, incipient osteoid formation and early endothelial sprouting matured into dense collagen networks and established vasculature by 8 weeks, demonstrating progressive bone–vessel co-regeneration (Fig. S20 and S21). Furthermore, no pathological changes were observed in distant organs at either 4 or 8 weeks post-surgery (Fig. S22 and S23), confirming no systemic adverse effects. Together, these results demonstrate that P-Exos@HAMA/Ti promotes spatially coordinated osteogenesis and angiogenesis while maintaining systemic biosafety.

Functional recovery was assessed through gait tracking and hindlimb grip testing (Fig. 9a–b). Gait parameters including stride length, stance width, and forelimb weight support ratio were quantified using computer-assisted analysis. At 8 weeks, footprint mapping revealed more consistent and symmetric hindlimb stepping patterns in P-Exos@HAMA/Ti-treated rats, closely resembling those of healthy controls (Fig. 9c). Quantitative locomotor parameters—stride length, stance width, and forelimb weight support—were significantly improved in the P-Exos@HAMA/Ti group (Fig. 9d–f), with no significant differences from normal controls, indicating near-complete functional recovery. At 4 weeks, Group IV rats exhibited improved stride symmetry and paw-print regularity (Fig. S24 and S25), though stride length and contact force remained below normal values, reflecting a gradual recovery trajectory. These results underscore the value of longitudinal behavioral tracking as a sensitive indicator of scaffold-mediated healing. Hindlimb grip strength was enhanced in Group IV, approaching normal values and exceeding Groups I–III (Fig. 9g). Dynamic paw-pressure mapping revealed synchronized bilateral paw forces and prolonged ground contact, consistent with enhanced balance and coordination (Fig. 9h). Partial recovery of paw-pressure dynamics was already evident at week 4 (Fig. S26), which further matured by week 8, confirming progressive restoration of locomotor function. Improvements in Group IV were more pronounced than those in Group III, highlighting the contribution of Piezo1-mediated mechanotransduction to functional recovery. Collectively, these behavioral results demonstrate that P-Exos@HAMA/Ti not only facilitates anatomical regeneration but also accelerates functional restitution—an essential criterion for clinical translation.

Taken together, P-Exos@HAMA/Ti exhibits excellent biocompatibility, robust osteogenic efficacy, and substantial locomotor improvement in the ONFH model. These outcomes likely arise from synergistic mechanisms: prolonged exosome release, Piezo1-mediated mechanotransduction, and scaffold-provided osteoconduction, recapitulating the physiological healing microenvironment. Behavioral endpoints provide a translational bridge between structural restoration and clinically relevant functional recovery. The 4- and 8-week datasets delineate temporal healing dynamics, illustrating early therapeutic engagement and long-term durability, positioning P-Exos@HAMA/Ti as a promising platform for ONFH and other orthopedic disorders requiring coupled regeneration and mechanosensory reprogramming.

Several limitations of the present study merit consideration. The in vivo assessment was carried out in a SD rat model of ONFH induced by a sequential LPS-steroid protocol, which is well established for replicating key histopathological features of glucocorticoid-induced ONFH, including trabecular disruption and microvascular compromise [79, 80]. Although this model facilitates robust group comparisons, the small size of the femoral head limits the volume of implantable scaffolds, and the quadrupedal loading pattern differs from human bipedal weight-bearing. Subsequent investigations in larger animal models, such as New Zealand white rabbits or beagle dogs, are crucial to validate the scalability and translational potential of the P-Exos@HAMA/Ti platform. Additionally, the mechanotransductive effects described herein reflect Piezo1-mediated intracellular signaling rather than externally applied mechanical loading; future studies incorporating dynamic bioreactor conditions would further validate the mechano-responsive release and regenerative potential of the P-Exos@HAMA/Ti platform.

Conclusions

In summary, we have developed a bioinspired, mechanosensitive exosome-hydrogel scaffold (P-Exos@HAMA/Ti) with rigorously characterised per-vesicle Piezo1 payload and confirmed membrane-proximal localisation, capable of promoting osteogenesis and angiogenesis through sustained delivery of Piezo1-engineered exosomes in a mechanically supportive microenvironment. In vitro results indicate that P-Exos significantly enhance the osteogenic differentiation and vascularization potential of BMSCs and VECs through Piezo1-mediated mechanosensory activation, recapitulating key aspects of the mechanobiological niche of bone remodeling. In vivo studies in a glucocorticoid-induced ONFH rat model confirm the biocompatibility and therapeutic effectiveness of P-Exos@HAMA/Ti, leading to notable enhancements in bone microarchitecture, neovascularization, and functional limb recovery, with no observable systemic toxicity. Mechanistically, the therapeutic effects were attributed to Piezo1-mediated Ca²⁺ influx, which activated the YAP1/β-catenin signalling axis to drive osteogenic and angiogenic gene expression, as further evidenced by transcriptomic, metabolomic, and pathway-specific inhibitor studies, and corroborated by upregulation of Runx2, Osterix, VEGFA, and CD31 in histological analyses. The integration of Piezo1-engineered exosomes with a structurally stable and bioactive scaffold not only facilitates localized delivery and mechanical transduction but also provides a dual-regulatory strategy for the repair of necrotic bone. Importantly, longitudinal assessment revealed a time-dependent trajectory of tissue regeneration and functional restoration, highlighting the translational potential of this platform in orthopedic regenerative applications. Collectively, our findings establish P-Exos@HAMA/Ti as a promising therapeutic strategy for treating ONFH and potentially other mechanically compromised bone disorders. Future work should focus on large-animal validation, dynamic bioreactor-based mechanostimulation studies, and clinical-scale scaffold fabrication to advance P-Exos@HAMA/Ti toward translational application.

Materials and methods

Chemicals and biological reagents

Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F-12), fetal bovine serum (FBS), penicillin-streptomycin, phosphate buffered saline (PBS), trypsin-EDTA, and growth factor-reduced Matrigel were purchased from Gibco (Thermo Fisher Scientific, USA). Competent cells, antibiotic-coated plates, plasmid extraction kits, and control plasmids were obtained from Tsingke Biotechnology Co., Ltd. (Beijing, China). Transfection reagent kits were obtained from Lonza (Basel, Switzerland). Ultracentrifuge tubes were purchased from Beckman Coulter (Brea, CA, USA). PKH67 and DiO fluorescent dyes, Calcein-AM/PI Live/Dead Cell Staining Kit, DCFH-DA probe, RIPA lysis buffer, protease and phosphatase inhibitor cocktail, SDS loading buffer, BCA protein assay kits and anti-fade mounting medium were obtained from Beyotime Institute of Biotechnology (Shanghai, China). Primary antibodies against CD9, TSG101, Runx2, OCN, CD31, VEGFA, YAP1 and β-catenin were purchased from ABclonal Biotechnology Co., Ltd. (Wuhan, China). Anti-Osterix was obtained from Abcam (Cambridge, UK). Anti-Piezo1 antibody was purchased from Invitrogen (Carlsbad, USA). Anti-CD63 and HRP-conjugated Goat Anti-Rabbit IgG(H + L) were purchased from Proteintech Group (Rosemont, USA). Methacrylated hyaluronic acid (HAMA, EFL-HAMA-400K) and photoinitiator (LAP, 0.25% w/v) were purchased from Engineering For Life (Soochow, China). Cell Counting Kit-8 (CCK-8) was purchased from Biosharp (Anhui, China). Alkaline phosphatase (ALP) and Alizarin Red S (ARS) staining kits as well as paraformaldehyde solution (4%) and Calcium Ion Detection Kit were obtained from Servicebio (Wuhan, China). EdU Cell Proliferation Kit and qPCR reagents (cDNA synthesis kit and SYBR Green Master Mix) were obtained from Vazyme Biotech Co., Ltd. (Nanjing, China). PVDF membranes, SDS-PAGE gels, and ECL chemiluminescence reagents were purchased from LABLEAD Trading Co., Ltd. (Beijing, China). Deionized (DI) water with a resistivity of 18.2 MΩ·cm was used throughout this project.

Fabrication and key parameters of titanium lattice scaffold (Ti)

Ti6Al4V ELI lattices were generated via selective laser melting (SLM) utilizing powder size between 15 and 53 μm. The process parameters consisted of a laser power set at 270 W, scan speed at 2100 mm/s, hatch spacing of 0.1 mm, and layer thickness of 30 μm. Two variations of scaffolds were produced: flat cylindrical porous scaffolds with dimensions of 3 mm in diameter and 2 mm in height and elongated cylindrical porous scaffolds with dimensions of 1 mm in diameter and 18 mm in height. Following the printing process, the scaffolds were subjected to heat treatment at 840 ℃ for 2 h, then furnace-cooled to 500 ℃, and subsequently cleaned using acetone, ethanol, and water with the aid of ultrasonic technology.

Synthesis of exosomes and P-Exos

Exosomes were isolated from bone marrow mesenchymal stem cells (BMSCs) cultured in DMEM/F-12. When the cells reached 80–90% confluence, they were washed with PBS and incubated in serum-free medium for 48 h. The conditioned medium was collected on ice and processed through sequential centrifugation steps at 300 × g for 10 min, 2,000 × g for 20 min, and 10,000 × g for 30 min at 4 ℃ to eliminate cells, cell debris, and large vesicles/organelles. The resulting supernatant was filtered through a 0.22 μm PES membrane and then subjected to ultracentrifugation at 110,000 × g for 70 min at 4 ℃. The pellet obtained was washed with PBS, re-centrifuged, and resuspended in sterile PBS (50–200 µL per 10‑cm dish equivalent). Aliquots were stored at 4 ℃ for 24 h or at − 80 ℃ for longer periods, avoiding freeze-thaw cycles. The purified vesicles were subsequently designated as ‘Exos’ for use in the following experiments.

Subsequently, the Piezo1 coding sequence was assembled into a pcDNA3.4-derived backbone (Tsingke Biotechnology, Beijing) using a two-segment strategy (A/B). PCR products for each segment were confirmed by agarose gel and gel-extracted prior to recombination (Fig. 2d). Following transformation, colonies were screened by colony-PCR using a diagnostic primer pair from the Piezo1 panel (20240828-ZB102-E02/20240828-ZB102-F02; 5’-CACTGTGCATCGACTACCCG-3’/5’-TGTCTGGGCGTTGGTTACCC-3’), and positives were advanced to plasmid preparation (Fig. S6) and Sanger confirmation (sequencing primers in Table S2). Endotoxin-free plasmid stocks were used for loading. For electroporation, exosome-plasmid mixtures were prepared in Nucleofector™ Solution SF and transferred to 100-µL cuvettes of a 4D-Nucleofector (square-wave). Four preset programs were prospectively screened, Program I (CA-137), Program II (DS-120), Program III (CM-137), and Program IV (CM-150), as an initial grid spanning distinct program families recommended by the manufacturer for vesicle/primary-type systems (Table S3). Based on pilot comparison, Program II (DS-120) was adopted for production runs (Fig. 2i). Immediately after pulsing, samples were treated with DNase I and purified (110,000 × g ultracentrifugation or size-exclusion chromatography) to remove unencapsulated DNA, yielding P-Exos for downstream assays. To quantify the absolute plasmid payload per vesicle, P-Exos preparations produced using Program II (DS-120) were subjected to ddPCR. Each P-Exos stock was first confirmed at 1.0 × 10⁹ particles/mL by NTA. Stocks were then diluted 1:100,000 in DNase/RNase-free water and analysed using a Piezo1-specific primer–probe set on a ddPCR platform (Bio-Rad). Piezo1 copy numbers per µL were determined from positive droplet fractions by Poisson statistics, and per-vesicle copy number was calculated by dividing ddPCR-derived copies/µL by the NTA-determined particle concentration. Unloaded Exos and a water blank were included as background and negative controls, respectively. Three independent P-Exos batches were quantified on separate days.

Synthesis of HAMA hydrogels and assembly of P‑Exos@HAMA/Ti

HAMA was dissolved in PBS with LAP and sterile‑filtered through a 0.22 μm membrane. Three concentrations (0.75%, 1.0%, and 1.5% w/v) were prepared to create materials with adjustable mechanical and release properties. The hydrogels were stored at 4 ℃ before use. Ti scaffold was combined with HAMA solutions containing Exos or P-Exos and gently mixed to a final loading of 1 × 10¹¹ particles·mL^− 1, to ensure even distribution without disrupting vesicle integrity. The Ti6Al4V scaffold was fully immersed in the exosome-laden HAMA precursor solution and subjected to gentle agitation for approximately 1 h at 4 °C to allow complete infiltration via capillary action prior to photocrosslinking. Photocrosslinking was then done with 405 nm UV light (10 mW/cm², 15–20 s) to create stable hydrogel coatings. The resulting constructs were named Exos@HAMA/Ti6Al4V (Exos@HAMA/Ti) or Piezo1‑Exos@HAMA/Ti6Al4V (P‑Exos@HAMA/Ti). All composites were rinsed with sterile PBS to remove unbound vesicles and stored at 4 ℃ until further use.

Material characterization

Transmission electron microscopy (TEM, JEM‑1400Plus, Japan) was employed to visualize the morphology of Exos and Piezo1‑Exos. Ultra‑high‑resolution field‑emission scanning electron microscopy (FE‑SEM, HITACHI, Japan) was used to observe the pore architecture of both scaffold types as well as the surface of HAMA coatings and P‑Exos@HAMA/Ti constructs. In addition, field‑emission SEM (Thermo Fisher Scientific, USA) equipped with an EDS system (EDAX ELECT SUPER) was used for elemental mapping and compositional analysis. X‑ray diffraction (XRD, Bruker, Germany) was conducted to determine crystalline phases, and Fourier transform infrared spectroscopy (FTIR, Thermo Fisher Scientific, USA) was performed to validate HAMA crosslinking. Static water contact angles were recorded using a Drop Shape Analyzer (DSA100, Germany), and rheological properties of hydrogels (frequency-modulus curves) were characterized on a rotational rheometer (Anton Paar, Austria) using 1–100 rad/s frequency sweeps. Nanoparticle tracking analysis (NTA) was performed to determine particle size distribution and concentration, and zeta potential was measured on a Nano-ZS90 Zetasizer (Malvern Instruments, UK), and exosome markers (CD9, CD63, TSG101, Piezo1) were verified by Western blotting. Micro-computed tomography (Micro-CT, Quantum GX2, PerkinElmer, USA) was performed to confirm porosity and pore interconnectivity. Mechanical properties were evaluated under uniaxial compression on an MTS E44.304 universal testing system (AK-SB-520, AK Medical, China) at a crosshead speed of 0.5 mm/min (24.6 ℃, 42% RH). Frequency-dependent modulus of hydrogel composites was additionally profiled on an Instron 5967 (Instron, USA) at 25 ℃.

In vitro biocompatibility and cell function assays

BMSCs were cultured in DMEM/F-12 with 10% FBS and 1% penicillin-streptomycin, while VECs were cultured in endothelial cell medium. Both cell types were maintained at 37℃ in a humidified atmosphere with 5% CO₂. BMSCs were divided into seven experimental groups: (I) Control, (II) Ti, (III) HAMA, (IV) Exos@HAMA/Ti, (V) P-Exos@HAMA/Ti, (VI) Yoda1 + Exos, and (VII) Normal. Groups I − VI were treated with dexamethasone (Dex) to simulate glucocorticoid-associated stress characteristic of ONFH, while Group VII served as an untreated control. For short-term viability, BMSCs were seeded in 96-well plates at 5 × 10³ cells/well and cultured for 24 h in the different groups. Cell viability was assessed using CCK-8 (Biosharp, Anhui, China), with absorbance at 450 nm measured on a microplate reader (SYNERGY Neo2, USA). For long-term viability, Group V cells were cultured for 7 days with daily proliferation monitoring. Cell viability was further evaluated using Calcein-AM/PI staining. Following 24 h treatment, cells were stained with 2 µM Calcein-AM and 4.5 µM PI for 15 min at 37 ℃, rinsed with PBS, and visualized by confocal laser scanning microscopy (CLSM, Andor, UK). Intracellular ROS levels were measured using a ROS detection kit (Beyotime Institute of Biotechnology, Shanghai, China). After 48 h of treatment, BMSCs were incubated with DCFH-DA (10 µM) at 37 ℃ for 30 min and rinsed twice with PBS. For confocal imaging, stained cells were directly imaged by CLSM, and mean fluorescence intensity (MFI) was quantified using ImageJ. For flow cytometry, cells were detached with trypsin, neutralized, centrifuged, resuspended in cold PBS, and filtered through a 40 μm strainer. Single-cell suspensions were analyzed using a flow cytometer at 488 nm excitation and 525–530 nm emission (FITC channel), and MFI was calculated for inter-group comparison. For wound-healing assays, confluent BMSCs were scratched with a 200 µL tip, washed with PBS, and cultured in serum-free DMEM/F-12 for 24 h. Wound closure % at 0/24 h was computed. For Transwell migration, 1 × 10⁴ cells (seeded in upper chambers with 8 μm pore inserts) migrated toward 10% FBS for 24 h, then were fixed, stained with crystal violet, and counted in three random fields per membrane. All experiments were performed in triplicate; data represent mean ± SD. For Piezo1 membrane localisation analysis, BMSCs were seeded on confocal dishes, treated according to the four-group design (Control, Exos, P-Exos, and Yoda1 + Exos), fixed with 4% paraformaldehyde for 15 min, permeabilised with 0.1% Triton X-100 for 10 min, and blocked with 5% BSA for 1 h. Cells were incubated overnight at 4°C with an anti-Piezo1 primary antibody, followed by an Alexa Fluor 488-conjugated secondary antibody for 1 h at room temperature. F-actin was labelled with 594-Phalloidin for 30 min, and nuclei were counterstained with DAPI. Images were acquired by CLSM and mean fluorescence intensity was quantified from at least three randomly selected fields per group using ImageJ.

In vitro osteogenic differentiation and angiogenesis assays

Following 14 days of osteogenic induction, cells were fixed with 4% paraformaldehyde for 15 min and rinsed with PBS. ALP activity was evaluated using an ALP staining kit and positively stained areas were examined under a light microscope. For mineralization assessment, cells were incubated with 0.1% Alizarin Red S solution for 30 min at room temperature, washed with PBS, and mineralized nodules were observed under a microscope. For western blot analysis, cells were lysed in RIPA buffer supplemented with protease and phosphatase inhibitors after 7 days of induction. Protein (20–30 µg) were separated by SDS-PAGE and transferred to PVDF membranes. After blocking with 5% BSA for 1 h, membranes were incubated overnight at 4 ℃ with primary antibodies against Piezo1, Runx2, Osterix, and Osteocalcin, using β-actin as the loading control. HRP-conjugated secondary antibodies were applied for 1 h, and signals were detected using ECL. Band intensities were quantified using ImageJ. Total RNA was extracted using TRIzol, reverse-transcribed to cDNA, and analyzed by qPCR with SYBR Green Master Mix. Expression of Piezo1, Runx2, Osterix, and OCN was normalized to GAPDH using the 2^−ΔΔCt method. For immunofluorescence, cells were fixed after 7 days, permeabilized with 0.1% Triton X-100, blocked in 5% BSA, and incubated overnight with primary antibodies against Piezo1, Runx2, Osterix, and OCN. Alexa Fluor-conjugated secondary antibodies and DAPI were sequentially applied. Images were acquired by CLSM and quantified using ImageJ. For angiogenesis assays, VECs were used for CCK-8 proliferation assessment at days 1, 3, 5, and 7, and EdU staining to quantify DNA synthesis. For tube formation, VECs were seeded onto polymerised Matrigel and incubated for 12 h; total tube length and junction number were quantified using ImageJ. CD31 and VEGFA expression were assessed by Western blot, RT-qPCR, and immunofluorescence using the same protocols as described above for osteogenic markers.

Transcriptomic profiling, untargeted metabolomics, and Ca²⁺-YAP1/β-catenin pathway validation

Total RNA was extracted from samples using TRIzol reagent (Invitrogen, USA), and RNA purity and integrity were confirmed by NanoDrop and Agilent 2100 Bioanalyzer. Qualified RNA was used for library construction and sequenced on the Illumina HiSeq platform, yielding more than 50 Gb of clean reads with Q30 scores above 94%. Clean reads were mapped to the rat reference genome (mRatBN7.2, NCBI) using HISAT2, assembled with StringTie, and quantified as FPKM/TPM values. Differentially expressed genes (DEGs) were identified using DESeq2 (|log₂FC| ≥ 1, adjusted p < 0.05), and subjected to GO, KEGG, and GSEA enrichment analyses. To isolate the transcriptomic and metabolomic effects of exosome treatment from potential scaffold material interactions, all omics profiling experiments were conducted in standard cell culture conditions in the absence of HAMA hydrogel or Ti6Al4V scaffold components. For metabolomics, metabolites were extracted using a methanol-based method and profiled by UHPLC-MS/MS (Thermo Fisher Scientific) under both positive and negative ion modes. BMSC-derived samples were rapidly quenched on ice and extracted with pre-chilled methanol: acetonitrile: water (2:2:1, v/v/v) containing internal standards. A pooled QC sample was injected periodically to monitor system stability. Chromatographic separation was performed on a hydrophilic-interaction (HILIC) amide column with a water/acetonitrile gradient (autosampler at 4℃). MS data were acquired on a high-resolution Orbitrap in ESI-positive and ESI-negative modes with data-dependent MS/MS. Data processing—including peak alignment, normalization, and metabolite annotation—was conducted using Compound Discoverer software with the KEGG, HMDB, and mzCloud databases. Peak areas were normalized to internal standards, log-transformed, and Pareto-scaled. Multivariate statistical analyses (PCA, OPLS-DA) were conducted, and differential metabolites were defined by VIP > 1 and FDR-adjusted p < 0.05, with KEGG pathway enrichment used to explore relevant metabolic alterations. To validate the Ca²⁺-responsive YAP1/β-catenin axis identified by omics analysis, intracellular Ca²⁺ was quantified using a cell-permeant fluorescent indicator (488-nm excitation), and mean fluorescence intensity was measured from confocal images. YAP1 and β-catenin were assessed by western blot (RIPA extraction, SDS-PAGE, PVDF transfer, 5% BSA blocking; primary antibodies overnight at 4 ℃; HRP secondary; GAPDH as loading control), RT-qPCR (SYBR Green; 2^−ΔΔCt vs. GAPDH), and immunofluorescence to assess subcellular localization (4% PFA fixation, 0.1% Triton X-100 permeabilization, 5% BSA blocking, Alexa Fluor secondaries, DAPI counterstain). Nuclear-to-cytoplasmic fluorescence ratios were quantified using ImageJ.

In vivo biocompatibility evaluation

Six 10-week-old Sprague-Dawley (SD) rats (Vital River Laboratory Animal Technology, Beijing, China) were used. The animals were randomized (n = 3 per group) to either no treatment (Control group) or P-Exos@HAMA/Ti implantation (Treatment group) and were monitored for 7 days. On day 7, the rats were euthanized; blood was collected for routine hematology and serum biochemistry (RBC, HGB, HCT, MCV, MCH, MCHC, PLT, MON, LYMPH, GRAN, ALT, AST, ALP, BUN, CREA, UA) using automated analyzers (BC-2800vet, Chemray 240, Rayto). The major organs (heart, liver, spleen, lung, and kidney) were harvested, fixed in 4% paraformaldehyde, paraffin-embedded, sectioned, and H&E-stained for histopathological assessment.

In vivo ONFH model on rats

Ten-week-old Sprague-Dawley rats were housed under standard conditions (22 − 26 ℃, 37 − 42% humidity, 12-h light/dark cycle; ad libitum food and water; two per cage) and randomly allocated to five groups: Control, Ti, Exos@HAMA/Ti, P-Exos@HAMA/Ti and Normal. ONFH was induced in all groups except Normal using a sequential LPS-steroid protocol: two doses of lipopolysaccharide (20 µg kg^− 1, i.p.; Sigma, USA) 24 h apart followed by three injections of methylprednisolone (40 mg kg^− 1, i.m.; Pfizer, USA) into alternating gluteal muscles at 24 h intervals, beginning 24 h after the second LPS dose. Two weeks after the final steroid injection, rats underwent surgery under isoflurane anesthesia (1 − 2% in oxygen) with aseptic technique. Femoral heads were exposed during surgery and either implanted with designated materials (Ti, Exos@HAMA/Ti, or P-Exos@HAMA/Ti) or underwent sham treatment (Control); the Normal group did not undergo induction or surgery. Locomotor function was evaluated at 4- and 8-weeks post-surgery through computer-assisted gait analysis and hindlimb grip-strength testing (mean of three consecutive trials per rat). At both time points, the animals were euthanized, and femora were collected for micro-computed tomography (micro-CT) and subsequent 3D reconstruction. Trabecular morphometrics—bone volume/total volume (BV/TV), bone surface/volume ratio (BS/BV), bone surface/total volume (BS/TV), trabecular number (Tb. N), trabecular thickness (Tb. Th), and trabecular separation (Tb. Sp) —were quantified within a standardized femoral-head region of interest using CTAn (Bruker, Germany). For histological and immunohistochemical analyses, the femoral heads were fixed in 4% paraformaldehyde, decalcified in 10% EDTA (pH 7.4) at 4 ℃ with regular solution changes, paraffin-embedded, and sectioned for H&E and Masson’s trichrome staining. Sections were subjected to antigen retrieval and immunostained for Osterix, Runx2, Piezo1, CD31, and VEGFA, with visualization achieved through HRP-DAB staining and hematoxylin counterstaining. All imaging was performed under consistent settings.

Statistical analysis

The quantitative evaluation of the data in the experiments was repeated at least three times and expressed as mean ± SD. Two-tailed t-test and two-way analysis of variance test were performed to assess statistical significance of values through GraphPad Prism 9.0 (GraphPad Software Inc.). Statistical significance was set at: ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1.^{(21.1MB, docx)}

Acknowledgements

None.

Author contributions

Y. Yu and Z. Ji contributed equally to this work. S. Liu, J. Xue, and W. Qian designed the project. Y. Yu, Z. Ji, H. Xu, M. Ma, Z. Yin, Y. Du, T. Zhang and G. Wang carried out the experiments and characterization. Y. Yu, F. Tian, J. Xue, S. Liu and W. Qian wrote and revised the paper. All authors participated in the discussion of the results.

Funding

This work was financially supported by the National Natural Science Foundation of China (Grant No. 82372410, 81972046, 52473121, and 52221006) and the National High Level Hospital Clinical Research Funding (Grant number: 2022-PUMCH-B-001).

Data availability

All data supporting the findings of this study are available within the paper and its Supplementary Information.

Declarations

Ethics approval and consent to participate

All procedures were approved by the Ethics Committee on Animal Experiments of Peking Union Medical College Hospital (Approval No. XHDW-2024-128) and conducted in accordance with China’s GB/T 35892 − 2018 (Guidelines for Ethical Review of Laboratory Animal Welfare) and the ARRIVE guidelines.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

The original online version of this article was revised: Fig. 9 has been updated.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Yang Yu and Zhongyin Ji contributed equally to this work.

Change history

6/30/2026

A Correction to this paper has been published: 10.1186/s12951-026-04712-8

Contributor Information

Sen Liu, Email: liusen@pumch.cn.

Jiajia Xue, Email: jiajiaxue@mail.buct.edu.cn.

Wenwei Qian, Email: qianww@pumch.cn.

References

1.Peyravian N, Milan PB, Kebria MM, Mashayekhan S, Ghasemian M, Amiri S, Hamidi M, Shavandi A, Moghtadaei M. Designing and synthesis of injectable hydrogel based on carboxymethyl cellulose/carboxymethyl chitosan containing QK peptide for femoral head osteonecrosis healing. Int J Biol Macromol. 2024;270:132127. [DOI] [PubMed] [Google Scholar]
2.Chen Y, Miao Y, Liu K, Xue F, Zhu B, Zhang C, Li G. Evolutionary course of the femoral head osteonecrosis: histopathological - radiologic characteristics and clinical staging systems. J Orthop Translat. 2022;32:28–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Li B, Yu L, Huang Z, Liang Y, Li G, Zhao Y. A novel device for treatment of osteonecrosis of femoral head: feasibility and preliminary efficacy of animal study. J Orthop Translat. 2021;31:20–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Fan Y, Chen Z, Wang H, Jiang M, Lu H, Wei Y, Hu Y, Mo L, Liu Y, Zhou C, et al. Isovitexin targets SIRT3 to prevent steroid-induced osteonecrosis of the femoral head by modulating mitophagy-mediated ferroptosis. Bone Res. 2025;13:18. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Lamb JN, Holton C, O’Connor P, Giannoudis PV. Avascular necrosis of the hip. BMJ. 2019;365:l2178. [DOI] [PubMed] [Google Scholar]
6.Rajeev PB, Jyothiprasanth M, Savio A, Valsan S, Jithin CR. A minimally invasive technique of core decompression and cancellous bone grafting of femoral head osteonecrosis. J Orthop Rep. 2024;3:100318. [Google Scholar]
7.Gangji V, Toungouz M, Lechanteur C, Beguin Y, Baudoux E, De maertelaer V, Pather S, Katz R, Ino J, Egrise D, et al. THU0540 autologous osteoblastic cells versus concentrated bone marrow implantation in osteonecrosis of the femoral head: a randomized controlled single blind study. Ann Rheum Dis. 2016;75:387. [Google Scholar]
8.Lavernia CJ, Sierra RJ, Grieco FR. Osteonecrosis of the femoral head. J Am Acad Orthop Surg. 1999;7:250–61. [DOI] [PubMed] [Google Scholar]
9.Campbell P, Beaulé PE, Ebramzadeh E, Le Duff MJ, De Smet K, Lu Z, Amstutz HC. The John Charnley Award: a study of implant failure in metal-on-metal surface arthroplasties. Clin Orthop Relat Res. 2006;453:35–46. [DOI] [PubMed] [Google Scholar]
10.Yang P, He W, Yang W, Jiang L, Lin T, Sun W, Zhang Q, Bai X, Sun W, Guo D. Panoramic heat map for spatial distribution of necrotic lesions. Bone Joint Res. 2024;13:294–305. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Tang H, Ling T, Zhao E, You M, Chen X, Chen G, Zhou K, Zhou Z. The efficacy of core decompression combined with regenerative therapy in early femoral head necrosis: a systematic review and meta-analysis involving 954 subjects. Front Pharmacol. 2024;15:1501590. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Zhao D, Huang S, Lu F, Wang B, Yang L, Qin L, Yang K, Li Y, Li W, Wang W, et al. Vascularized bone grafting fixed by biodegradable magnesium screw for treating osteonecrosis of the femoral head. Biomaterials 2016;81:84–92. [DOI] [PubMed] [Google Scholar]
13.Yang W, Pan Q, Peng Y, Hu Y, Cui Y, Shao Z, Meng C, Wang H, Huang W. Dual-target nanotherapy for vascular endothelium and bone mesenchymal stem cells halt steroid-induced osteonecrosis of the femoral head progression. J Controlled Release. 2025;380:219–39. [DOI] [PubMed] [Google Scholar]
14.Yamamoto T, Bullough PG. Subchondral insufficiency fracture of the femoral head: a differential diagnosis in acute onset of coxarthrosis in the elderly. Arthritis Rheum. 1999;42:2719–23. [DOI] [PubMed] [Google Scholar]
15.Berkowitz JL, Potter HG. Advanced MRI techniques for the hip joint: focus on the postoperative hip. AJR Am J Roentgenol. 2017;209:534–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Zu H, Zheng L, Huo M, Liu K, Dreyer C, Zhang Y, He X, Li Y, Zou L, Huang L, et al. Tree-inspired magnesium hybrid column for preventing hip collapse in steroid-associated osteonecrosis in bipedal emus. Mater Today. 2024;80:113–38. [Google Scholar]
17.Moya-Angeler J, Gianakos AL, Villa JC, Ni A, Lane JM. Current concepts on osteonecrosis of the femoral head. World J Orthop. 2015;6:590–601. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Guo M, Qi B, Pei Z, Ni H, Ren J, Luo H, Shi H, Meng C, Yu Y, Tang Z, et al. Therapeutic effect of low-dose BMSCs-Loaded 3D microscaffold on early osteonecrosis of the femoral head. Mater Today Bio. 2025;30:101426. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Aoyama T, Goto K, Kakinoki R, Ikeguchi R, Ueda M, Kasai Y, Maekawa T, Tada H, Teramukai S, Nakamura T, Toguchida J. An exploratory clinical trial for idiopathic osteonecrosis of femoral head by cultured autologous multipotent mesenchymal stromal cells augmented with vascularized bone grafts. Tissue Eng Part B Rev. 2014;20:233–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Cao H, Guan H, Lai Y, Qin L, Wang X. Review of various treatment options and potential therapies for osteonecrosis of the femoral head. J Orthop Transl. 2016;4:57–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Torres A, Lorenzo V, Salido E. Calcium metabolism and skeletal problems after transplantation. J Am Soc Nephrol. 2002;13:551–8. [DOI] [PubMed] [Google Scholar]
22.Torres A, Fernández A, Castillo P, Muñoz M, Páez F, Macías M, Zambrano A, Guarnizo J, Serrano J, García B. Osteonecrosis of femoral head scoping review. EPRA Int J Multidiscip Res. 2025;11(2):187–91. 10.36713/epra20136.
23.Klumpp R, Trevisan C. Aseptic osteonecrosis of the hip in the adult: current evidence on conservative treatment. Clin Cases Min Bone Metab. 2015;12:39–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Lai KA, Shen WJ, Yang CY, Shao CJ, Hsu JT, Lin RM. The use of alendronate to prevent early collapse of the femoral head in patients with nontraumatic osteonecrosis. A randomized clinical study. J Bone Joint Surg Am. 2005;87:2155–9. [DOI] [PubMed] [Google Scholar]
25.Camporesi EM, Vezzani G, Bosco G, Mangar D, Bernasek TL. Hyperbaric oxygen therapy in femoral head necrosis. J Arthroplast. 2010;25:118–23. [DOI] [PubMed] [Google Scholar]
26.Sen RK, Tripathy SK, Aggarwal S, Marwaha N, Sharma RR, Khandelwal N. Early results of core decompression and autologous bone marrow mononuclear cells instillation in femoral head osteonecrosis: a randomized control study. J Arthroplast. 2012;27:679–86. [DOI] [PubMed] [Google Scholar]
27.Sadile F, Bernasconi A, Russo S, Maffulli N. Core decompression versus other joint preserving treatments for osteonecrosis of the femoral head: a meta-analysis. Br Med Bull. 2016;118:33–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Sun X, Zhong R, Wu C, Ye S, Yuan H, Fang Z, Chen J, Cheng D, Hao L, Chu L, Wang L. 3D printed titanium scaffolds with bi-directional gradient QK-functionalized surface. Adv Mater. (Deerfield Beach Fla). 2025;37:e2406421. [DOI] [PubMed] [Google Scholar]
29.Villegas M, Bayat F, Kramer T, Schwarz E, Wilson D, Hosseinidoust Z, Didar TF. Emerging strategies to prevent bacterial infections on titanium-based implants. Small 2024;20:e2404351. [DOI] [PubMed] [Google Scholar]
30.Zhang Y, Cheng Z, Liu Z, Shen X, Cai C, Li M, Luo Z. Functionally tailored metal-organic framework coatings for mediating Ti implant osseointegration. Adv Sci. (Weinheim Baden-Wurttemberg Germany). 2023;10:e2303958. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Ju Y, Hu Y, Yang P, Xie X, Fang B. Extracellular vesicle-loaded hydrogels for tissue repair and regeneration. Mater Today Bio. 2023;18:100522. [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Oliveira ÉR, Nie L, Podstawczyk D, Allahbakhsh A, Ratnayake J, Brasil DL, Shavandi A. Advances in growth factor delivery for bone tissue engineering. Int J Mol Sci. 2021;22(2):903. [DOI] [PMC free article] [PubMed]
33.Mirkhalaf M, Men Y, Wang R, No Y, Zreiqat H. Personalized 3D printed bone scaffolds: areview. Acta Biomater. 2023;156:110–24. [DOI] [PubMed] [Google Scholar]
34.Wang Z, Sun Y, Li C. Advances in 3D printing technology for preparing bone tissue engineering scaffolds from biodegradable materials. Front Bioeng Biotechnol. 2024;12:1483547. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Janmohammadi M, Nazemi Z, Salehi AOM, Seyfoori A, John JV, Nourbakhsh MS, Akbari M. Cellulose-based composite scaffolds for bone tissue engineering and localized drug delivery. Bioactive Mater. 2023;20:137–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Farjaminejad R, Farjaminejad S, Garcia-Godoy F. Regenerative endodontic therapies: harnessing stem cells, scaffolds, and growth factors. Polymers 2025;17(11):1475. [DOI] [PMC free article] [PubMed]
37.Xu W, Wang F, Stein J, Wang S, Jiang P, Yang L, Cheng L, Lee K. Engineering topographical cues to enhance neural regeneration in spinal cord injury: overcoming challenges and advancing therapies. Adv Funct Mater. 2025;35:2508435.
38.Islam M, Ordoño J, Tang W, Monclús M, Echeverry-Rendón M, Wang D. Pyrolytic carbon microlattices from 3d‐printed polyethylene glycol diacrylate and their in vitro assessment for bone regeneration. Small Structures; 2025.
39.Deng L, Liu Y, Wu Q, Lai S, Yang Q, Mu Y, Dong M. Exosomes to exosome-functionalized scaffolds: a novel approach to stimulate bone regeneration. Stem Cell Res Ther. 2024;15:407. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Tenchov R, Sasso JM, Wang X, Liaw WS, Chen CA, Zhou QA. ExosomesNature’s lipid nanoparticles, a rising star in drug delivery and diagnostics. ACS Nano. 2022;16:17802–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Zhu L, Sun HT, Wang S, Huang SL, Zheng Y, Wang CQ, Hu BY, Qin W, Zou TT, Fu Y, et al. Isolation and characterization of exosomes for cancer research. J Hematol Oncol. 2020;13:152. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Hade MD, Suire CN, Suo Z. Mesenchymal stem cell-derived exosomes: applications in regenerative medicine. Cells 2021;10(8):1959. [DOI] [PMC free article] [PubMed]
43.Zhang M, Hu S, Liu L, Dang P, Liu Y, Sun Z, Qiao B, Wang C. Engineered exosomes from different sources for cancer-targeted therapy. Signal Transduct Target Ther. 2023;8:124. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Liang Y, Duan L, Lu J, Xia J. Engineering exosomes for targeted drug delivery. Theranostics 2021;11:3183–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Yao T, Dong X, Wang X, Liu X, Fu L, Li L. Engineering exosomes for mRNA delivery: a review. Int J Biol Macromol. 2025;316:144662. [DOI] [PubMed] [Google Scholar]
46.Qin L, Liu W, Cao H, Xiao G. Molecular mechanosensors in osteocytes. Bone Res. 2020;8:23. [DOI] [PMC free article] [PubMed]
47.Chen N, Danalache M, Liang C, Alexander D, Umrath F. Mechanosignaling in osteoporosis: when cells feel the force. Int J Mol Sci. 2025;26(9):4007. [DOI] [PMC free article] [PubMed]
48.Zhou T, Gao B, Fan Y, Liu Y, Feng S, Cong Q, Zhang X, Zhou Y, Yadav PS, Lin J et al. Piezo1/2 mediate mechanotransduction essential for bone formation through concerted activation of NFAT-YAP1-ß-catenin. Elife 2020;9:e52779. [DOI] [PMC free article] [PubMed]
49.Ranade SS, Qiu Z, Woo S-H, Hur SS, Murthy SE, Cahalan SM, Xu J, Mathur J, Bandell M, Coste B et al. Piezo1, a mechanically activated ion channel, is required for vascular development in mice. Proceedings of the National Academy of Sciences 2014;111:10347–10352. [DOI] [PMC free article] [PubMed]
50.Chen X, Ji X, Lao Z, Pan B, Qian Y, Yang W. Role of YAP/TAZ in bone diseases: a transductor from mechanics to biology. J Orthop Transl. 2025;51:13–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Li X, Han L, Nookaew I, Mannen E, Silva MJ, Almeida M, Xiong J. Stimulation of Piezo1 by mechanical signals promotes bone anabolism. eLife 2019;8:e49631. [DOI] [PMC free article] [PubMed]
52.Sun W, Chi S, Li Y, Ling S, Tan Y, Xu Y, Jiang F, Li J, Liu C, Zhong G et al. The mechanosensitive Piezo1 channel is required for bone formation. eLife 2019;8:e47454. [DOI] [PMC free article] [PubMed]
53.Deng H, Wang J, An R. Hyaluronic acid-based hydrogels: as an exosome delivery system in bone regeneration. Front Pharmacol. 2023;14:1131001. [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Yang X, Gao J, Yang S, Wu Y, Liu H, Su D, Li D. Pore size-mediated macrophage M1 to M2 transition affects osseointegration of 3D-printed PEEK scaffolds. Int J Bioprint. 2023;9:755. [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Yang X, Wu L, Li C, Li S, Hou W, Hao Y, Lu Y, Li L. Synergistic amelioration of osseointegration and osteoimmunomodulation with a microarc oxidation-treated three-dimensionally printed Ti-24Nb-4Zr-8Sn scaffold via surface activity and low elastic modulus. ACS Appl Mater Interfaces 2024;16:3171–86. [DOI] [PubMed] [Google Scholar]
56.Yang F, Li Y, Wang L, Che H, Zhang X, Jahr H, Wang L, Jiang D, Huang H, Wang J. Full-thickness osteochondral defect repair using a biodegradable bilayered scaffold of porous zinc and chondroitin sulfate hydrogel. Bioactive Mater. 2024;32:400–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Chmielewska A, Dean D. The role of stiffness-matching in avoiding stress shielding-induced bone loss and stress concentration-induced skeletal reconstruction device failure. Acta Biomater. 2024;173:51–65. [DOI] [PubMed] [Google Scholar]
58.Mushtaq RT, Askari GH, Bao C, Wang Y, Ahmed K, Khan AM, Sharma S, Alkahtani M. Optimization of 3D-printed bio-based super tough PLA (ST-PLA) scaffolds for cancellous bone regeneration: mechanical properties, lattice architecture, and osseointegration potential. Int J Biol Macromol. 2025;316:144466. [DOI] [PubMed] [Google Scholar]
59.Lv Z, Xu J, Cai X, Wang Y, Yang X, Wang H, Zhu Y, Feng B, Weng X. Exosome-based therapeutics for musculoskeletal disorders: advances in engineering, targeting, and biomaterial integration. ACS Nano. 2025;19:33681–716. [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Hashemi A, Ezati M, Nasr MP, Zumberg I, Provaznik V. Extracellular vesicles and hydrogels: an innovative approach to tissue regeneration. ACS Omega. 2024;9:6184–218. [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Zhao X, Chen X, Deng Y, Wu C, Ruan Z, Li C, Chen Y, Bei C, Zhu L, Yu H, Zhang X. A novel adhesive dual-sensitive hydrogel for sustained release of exosomes derived from M2 macrophages promotes repair of bone defects. Mater Today Bio. 2023;23:100840. [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Lu W, Zeng M, Liu W, Ma T, Fan X, Li H, Wang Y, Wang H, Hu Y, Xie J. Human urine-derived stem cell exosomes delivered via injectable GelMA templated hydrogel accelerate bone regeneration. Mater Today Bio. 2023;19:100569. [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Huang Z, Huang Y, Ning X, Li H, Li Q, Wu J. The functional effects of Piezo channels in mesenchymal stem cells. Stem Cell Res Ther. 2023;14:222. [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Wang L, You X, Lotinun S, Zhang L, Wu N, Zou W. Mechanical sensing protein PIEZO1 regulates bone homeostasis via osteoblast-osteoclast crosstalk. Nat Commun. 2020;11:282. [DOI] [PMC free article] [PubMed] [Google Scholar]
65.Lin X, Su S, Duan X, Li W, Hu D, Lin Y, Jiang S, Wang L, Li J, Xu D, Qingchun Z. Abstract 4136901: mechanosensitive channel protein Piezo1 synergizes with TLR4 to enhance osteogenic responses in human aortic valve interstitial cells. Circulation 2024;150(Suppl_1):A4136901.
66.Atashi F, Modarressi A, Pepper MS. The role of reactive oxygen species in mesenchymal stem cell adipogenic and osteogenic differentiation: a review. Stem Cells Dev. 2015;24:1150–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Lin C, Yang Q, Guo D, Xie J, Yang Y-S, Chaugule S, DeSouza N, Oh W-T, Li R, Chen Z, et al. Impaired mitochondrial oxidative metabolism in skeletal progenitor cells leads to musculoskeletal disintegration. Nat Commun. 2022;13:6869. [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Dai X, Yan X, Wintergerst KA, Cai L, Keller BB, Tan Y. Nrf2: redox and metabolic regulator of stem cell state and function. Trends Mol Med. 2020;26:185–200. [DOI] [PubMed] [Google Scholar]
69.Li S, Gao L, Zhang W, Yu Y, Zheng J, Liang X, Xin S, Ren W, Zhi K. MiR-152-5p suppresses osteogenic differentiation of mandible mesenchymal stem cells by regulating ATG14-mediated autophagy. Stem Cell Res Ther. 2022;13:359. [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Santana Nunez D, Malik AB, Lee Q, Ahn SJ, Coctecon-Murillo A, Lazarko D, Levitan I, Mehta D, Komarova YA. Piezo1 induces endothelial responses to shear stress via soluble adenylyl Cyclase-IP(3)R2 circuit. iScience 2023;26:106661. [DOI] [PMC free article] [PubMed] [Google Scholar]
71.Wang S, Chennupati R, Kaur H, Iring A, Wettschureck N, Offermanns S. Endothelial cation channel PIEZO1 controls blood pressure by mediating flow-induced ATP release. J Clin Investig. 2016;126:4527–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
72.Syeda R, Xu J, Dubin AE, Coste B, Mathur J, Huynh T, Matzen J, Lao J, Tully DC, Engels IH et al. Chemical activation of the mechanotransduction channel Piezo1. eLife 2015;4. [DOI] [PMC free article] [PubMed]
73.Grosso A, Burger MG, Lunger A, Schaefer DJ, Banfi A, Di Maggio N. It takes two to tango: coupling of angiogenesis and osteogenesis for bone regeneration. Front Bioeng Biotechnol. 2017;5:68. [DOI] [PMC free article] [PubMed] [Google Scholar]
74.Ma S, Wu J, Hu H, Mu Y, Zhang L, Zhao Y, Bian X, Jing W, Wei P, Zhao B, et al. Novel fusion peptides deliver exosomes to modify injectable thermo-sensitive hydrogels for bone regeneration. Mater Today Bio. 2022;13:100195. [DOI] [PMC free article] [PubMed] [Google Scholar]
75.Ma Z, Song W, He D, Zhang X, He Y, Li H. Smart µ-fiber hydrogels with macro‐porous structure for sequentially promoting multiple phases of articular cartilage regeneration. Adv Funct Mater. 2022;32:2113380.
76.Grosso A, Lunger A, Burger MG, Briquez PS, Mai F, Hubbell JA, Schaefer DJ, Banfi A. Di Maggio N: VEGF dose controls the coupling of angiogenesis and osteogenesis in engineered bone. npj Regen Med. 2023;8:15. [DOI] [PMC free article] [PubMed] [Google Scholar]
77.Dupont S, Morsut L, Aragona M, Enzo E, Giulitti S, Cordenonsi M, Zanconato F, Le Digabel J, Forcato M, Bicciato S, et al. Role of YAP/TAZ in mechanotransduction. Nature 2011;474:179–83. [DOI] [PubMed] [Google Scholar]
78.Aragona M, Panciera T, Manfrin A, Giulitti S, Michielin F, Elvassore N, Dupont S, Piccolo S. A mechanical checkpoint controls multicellular growth through YAP/TAZ Regulation by actin-processing factors. Cell 2013;154:1047–59. [DOI] [PubMed] [Google Scholar]
79.Zheng LZ, Wang JL, Kong L, Huang L, Tian L, Pang QQ, Wang XL, Qin L. Steroid-associated osteonecrosis animal model in rats. J Orthop Translat. 2018;13:13–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
80.Dong Y, Li Y, Huang C, Gao K, Weng X. Systemic application of teriparatide for steroid induced osteonecrosis in a rat model. BMC Musculoskelet Disord. 2015;16:163. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1.^{(21.1MB, docx)}

Data Availability Statement

All data supporting the findings of this study are available within the paper and its Supplementary Information.

[CR1] 1.Peyravian N, Milan PB, Kebria MM, Mashayekhan S, Ghasemian M, Amiri S, Hamidi M, Shavandi A, Moghtadaei M. Designing and synthesis of injectable hydrogel based on carboxymethyl cellulose/carboxymethyl chitosan containing QK peptide for femoral head osteonecrosis healing. Int J Biol Macromol. 2024;270:132127. [DOI] [PubMed] [Google Scholar]

[CR2] 2.Chen Y, Miao Y, Liu K, Xue F, Zhu B, Zhang C, Li G. Evolutionary course of the femoral head osteonecrosis: histopathological - radiologic characteristics and clinical staging systems. J Orthop Translat. 2022;32:28–40. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR3] 3.Li B, Yu L, Huang Z, Liang Y, Li G, Zhao Y. A novel device for treatment of osteonecrosis of femoral head: feasibility and preliminary efficacy of animal study. J Orthop Translat. 2021;31:20–5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR4] 4.Fan Y, Chen Z, Wang H, Jiang M, Lu H, Wei Y, Hu Y, Mo L, Liu Y, Zhou C, et al. Isovitexin targets SIRT3 to prevent steroid-induced osteonecrosis of the femoral head by modulating mitophagy-mediated ferroptosis. Bone Res. 2025;13:18. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR5] 5.Lamb JN, Holton C, O’Connor P, Giannoudis PV. Avascular necrosis of the hip. BMJ. 2019;365:l2178. [DOI] [PubMed] [Google Scholar]

[CR6] 6.Rajeev PB, Jyothiprasanth M, Savio A, Valsan S, Jithin CR. A minimally invasive technique of core decompression and cancellous bone grafting of femoral head osteonecrosis. J Orthop Rep. 2024;3:100318. [Google Scholar]

[CR7] 7.Gangji V, Toungouz M, Lechanteur C, Beguin Y, Baudoux E, De maertelaer V, Pather S, Katz R, Ino J, Egrise D, et al. THU0540 autologous osteoblastic cells versus concentrated bone marrow implantation in osteonecrosis of the femoral head: a randomized controlled single blind study. Ann Rheum Dis. 2016;75:387. [Google Scholar]

[CR8] 8.Lavernia CJ, Sierra RJ, Grieco FR. Osteonecrosis of the femoral head. J Am Acad Orthop Surg. 1999;7:250–61. [DOI] [PubMed] [Google Scholar]

[CR9] 9.Campbell P, Beaulé PE, Ebramzadeh E, Le Duff MJ, De Smet K, Lu Z, Amstutz HC. The John Charnley Award: a study of implant failure in metal-on-metal surface arthroplasties. Clin Orthop Relat Res. 2006;453:35–46. [DOI] [PubMed] [Google Scholar]

[CR10] 10.Yang P, He W, Yang W, Jiang L, Lin T, Sun W, Zhang Q, Bai X, Sun W, Guo D. Panoramic heat map for spatial distribution of necrotic lesions. Bone Joint Res. 2024;13:294–305. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR11] 11.Tang H, Ling T, Zhao E, You M, Chen X, Chen G, Zhou K, Zhou Z. The efficacy of core decompression combined with regenerative therapy in early femoral head necrosis: a systematic review and meta-analysis involving 954 subjects. Front Pharmacol. 2024;15:1501590. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Zhao D, Huang S, Lu F, Wang B, Yang L, Qin L, Yang K, Li Y, Li W, Wang W, et al. Vascularized bone grafting fixed by biodegradable magnesium screw for treating osteonecrosis of the femoral head. Biomaterials 2016;81:84–92. [DOI] [PubMed] [Google Scholar]

[CR13] 13.Yang W, Pan Q, Peng Y, Hu Y, Cui Y, Shao Z, Meng C, Wang H, Huang W. Dual-target nanotherapy for vascular endothelium and bone mesenchymal stem cells halt steroid-induced osteonecrosis of the femoral head progression. J Controlled Release. 2025;380:219–39. [DOI] [PubMed] [Google Scholar]

[CR14] 14.Yamamoto T, Bullough PG. Subchondral insufficiency fracture of the femoral head: a differential diagnosis in acute onset of coxarthrosis in the elderly. Arthritis Rheum. 1999;42:2719–23. [DOI] [PubMed] [Google Scholar]

[CR15] 15.Berkowitz JL, Potter HG. Advanced MRI techniques for the hip joint: focus on the postoperative hip. AJR Am J Roentgenol. 2017;209:534–43. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR16] 16.Zu H, Zheng L, Huo M, Liu K, Dreyer C, Zhang Y, He X, Li Y, Zou L, Huang L, et al. Tree-inspired magnesium hybrid column for preventing hip collapse in steroid-associated osteonecrosis in bipedal emus. Mater Today. 2024;80:113–38. [Google Scholar]

[CR17] 17.Moya-Angeler J, Gianakos AL, Villa JC, Ni A, Lane JM. Current concepts on osteonecrosis of the femoral head. World J Orthop. 2015;6:590–601. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR18] 18.Guo M, Qi B, Pei Z, Ni H, Ren J, Luo H, Shi H, Meng C, Yu Y, Tang Z, et al. Therapeutic effect of low-dose BMSCs-Loaded 3D microscaffold on early osteonecrosis of the femoral head. Mater Today Bio. 2025;30:101426. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR19] 19.Aoyama T, Goto K, Kakinoki R, Ikeguchi R, Ueda M, Kasai Y, Maekawa T, Tada H, Teramukai S, Nakamura T, Toguchida J. An exploratory clinical trial for idiopathic osteonecrosis of femoral head by cultured autologous multipotent mesenchymal stromal cells augmented with vascularized bone grafts. Tissue Eng Part B Rev. 2014;20:233–42. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Cao H, Guan H, Lai Y, Qin L, Wang X. Review of various treatment options and potential therapies for osteonecrosis of the femoral head. J Orthop Transl. 2016;4:57–70. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Torres A, Lorenzo V, Salido E. Calcium metabolism and skeletal problems after transplantation. J Am Soc Nephrol. 2002;13:551–8. [DOI] [PubMed] [Google Scholar]

[CR22] 22.Torres A, Fernández A, Castillo P, Muñoz M, Páez F, Macías M, Zambrano A, Guarnizo J, Serrano J, García B. Osteonecrosis of femoral head scoping review. EPRA Int J Multidiscip Res. 2025;11(2):187–91. 10.36713/epra20136.

[CR23] 23.Klumpp R, Trevisan C. Aseptic osteonecrosis of the hip in the adult: current evidence on conservative treatment. Clin Cases Min Bone Metab. 2015;12:39–42. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR24] 24.Lai KA, Shen WJ, Yang CY, Shao CJ, Hsu JT, Lin RM. The use of alendronate to prevent early collapse of the femoral head in patients with nontraumatic osteonecrosis. A randomized clinical study. J Bone Joint Surg Am. 2005;87:2155–9. [DOI] [PubMed] [Google Scholar]

[CR25] 25.Camporesi EM, Vezzani G, Bosco G, Mangar D, Bernasek TL. Hyperbaric oxygen therapy in femoral head necrosis. J Arthroplast. 2010;25:118–23. [DOI] [PubMed] [Google Scholar]

[CR26] 26.Sen RK, Tripathy SK, Aggarwal S, Marwaha N, Sharma RR, Khandelwal N. Early results of core decompression and autologous bone marrow mononuclear cells instillation in femoral head osteonecrosis: a randomized control study. J Arthroplast. 2012;27:679–86. [DOI] [PubMed] [Google Scholar]

[CR27] 27.Sadile F, Bernasconi A, Russo S, Maffulli N. Core decompression versus other joint preserving treatments for osteonecrosis of the femoral head: a meta-analysis. Br Med Bull. 2016;118:33–49. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR28] 28.Sun X, Zhong R, Wu C, Ye S, Yuan H, Fang Z, Chen J, Cheng D, Hao L, Chu L, Wang L. 3D printed titanium scaffolds with bi-directional gradient QK-functionalized surface. Adv Mater. (Deerfield Beach Fla). 2025;37:e2406421. [DOI] [PubMed] [Google Scholar]

[CR29] 29.Villegas M, Bayat F, Kramer T, Schwarz E, Wilson D, Hosseinidoust Z, Didar TF. Emerging strategies to prevent bacterial infections on titanium-based implants. Small 2024;20:e2404351. [DOI] [PubMed] [Google Scholar]

[CR30] 30.Zhang Y, Cheng Z, Liu Z, Shen X, Cai C, Li M, Luo Z. Functionally tailored metal-organic framework coatings for mediating Ti implant osseointegration. Adv Sci. (Weinheim Baden-Wurttemberg Germany). 2023;10:e2303958. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR31] 31.Ju Y, Hu Y, Yang P, Xie X, Fang B. Extracellular vesicle-loaded hydrogels for tissue repair and regeneration. Mater Today Bio. 2023;18:100522. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Oliveira ÉR, Nie L, Podstawczyk D, Allahbakhsh A, Ratnayake J, Brasil DL, Shavandi A. Advances in growth factor delivery for bone tissue engineering. Int J Mol Sci. 2021;22(2):903. [DOI] [PMC free article] [PubMed]

[CR33] 33.Mirkhalaf M, Men Y, Wang R, No Y, Zreiqat H. Personalized 3D printed bone scaffolds: areview. Acta Biomater. 2023;156:110–24. [DOI] [PubMed] [Google Scholar]

[CR34] 34.Wang Z, Sun Y, Li C. Advances in 3D printing technology for preparing bone tissue engineering scaffolds from biodegradable materials. Front Bioeng Biotechnol. 2024;12:1483547. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Janmohammadi M, Nazemi Z, Salehi AOM, Seyfoori A, John JV, Nourbakhsh MS, Akbari M. Cellulose-based composite scaffolds for bone tissue engineering and localized drug delivery. Bioactive Mater. 2023;20:137–63. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR36] 36.Farjaminejad R, Farjaminejad S, Garcia-Godoy F. Regenerative endodontic therapies: harnessing stem cells, scaffolds, and growth factors. Polymers 2025;17(11):1475. [DOI] [PMC free article] [PubMed]

[CR37] 37.Xu W, Wang F, Stein J, Wang S, Jiang P, Yang L, Cheng L, Lee K. Engineering topographical cues to enhance neural regeneration in spinal cord injury: overcoming challenges and advancing therapies. Adv Funct Mater. 2025;35:2508435.

[CR38] 38.Islam M, Ordoño J, Tang W, Monclús M, Echeverry-Rendón M, Wang D. Pyrolytic carbon microlattices from 3d‐printed polyethylene glycol diacrylate and their in vitro assessment for bone regeneration. Small Structures; 2025.

[CR39] 39.Deng L, Liu Y, Wu Q, Lai S, Yang Q, Mu Y, Dong M. Exosomes to exosome-functionalized scaffolds: a novel approach to stimulate bone regeneration. Stem Cell Res Ther. 2024;15:407. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR40] 40.Tenchov R, Sasso JM, Wang X, Liaw WS, Chen CA, Zhou QA. ExosomesNature’s lipid nanoparticles, a rising star in drug delivery and diagnostics. ACS Nano. 2022;16:17802–46. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR41] 41.Zhu L, Sun HT, Wang S, Huang SL, Zheng Y, Wang CQ, Hu BY, Qin W, Zou TT, Fu Y, et al. Isolation and characterization of exosomes for cancer research. J Hematol Oncol. 2020;13:152. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Hade MD, Suire CN, Suo Z. Mesenchymal stem cell-derived exosomes: applications in regenerative medicine. Cells 2021;10(8):1959. [DOI] [PMC free article] [PubMed]

[CR43] 43.Zhang M, Hu S, Liu L, Dang P, Liu Y, Sun Z, Qiao B, Wang C. Engineered exosomes from different sources for cancer-targeted therapy. Signal Transduct Target Ther. 2023;8:124. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Liang Y, Duan L, Lu J, Xia J. Engineering exosomes for targeted drug delivery. Theranostics 2021;11:3183–95. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR45] 45.Yao T, Dong X, Wang X, Liu X, Fu L, Li L. Engineering exosomes for mRNA delivery: a review. Int J Biol Macromol. 2025;316:144662. [DOI] [PubMed] [Google Scholar]

[CR46] 46.Qin L, Liu W, Cao H, Xiao G. Molecular mechanosensors in osteocytes. Bone Res. 2020;8:23. [DOI] [PMC free article] [PubMed]

[CR47] 47.Chen N, Danalache M, Liang C, Alexander D, Umrath F. Mechanosignaling in osteoporosis: when cells feel the force. Int J Mol Sci. 2025;26(9):4007. [DOI] [PMC free article] [PubMed]

[CR48] 48.Zhou T, Gao B, Fan Y, Liu Y, Feng S, Cong Q, Zhang X, Zhou Y, Yadav PS, Lin J et al. Piezo1/2 mediate mechanotransduction essential for bone formation through concerted activation of NFAT-YAP1-ß-catenin. Elife 2020;9:e52779. [DOI] [PMC free article] [PubMed]

[CR49] 49.Ranade SS, Qiu Z, Woo S-H, Hur SS, Murthy SE, Cahalan SM, Xu J, Mathur J, Bandell M, Coste B et al. Piezo1, a mechanically activated ion channel, is required for vascular development in mice. Proceedings of the National Academy of Sciences 2014;111:10347–10352. [DOI] [PMC free article] [PubMed]

[CR50] 50.Chen X, Ji X, Lao Z, Pan B, Qian Y, Yang W. Role of YAP/TAZ in bone diseases: a transductor from mechanics to biology. J Orthop Transl. 2025;51:13–23. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR51] 51.Li X, Han L, Nookaew I, Mannen E, Silva MJ, Almeida M, Xiong J. Stimulation of Piezo1 by mechanical signals promotes bone anabolism. eLife 2019;8:e49631. [DOI] [PMC free article] [PubMed]

[CR52] 52.Sun W, Chi S, Li Y, Ling S, Tan Y, Xu Y, Jiang F, Li J, Liu C, Zhong G et al. The mechanosensitive Piezo1 channel is required for bone formation. eLife 2019;8:e47454. [DOI] [PMC free article] [PubMed]

[CR53] 53.Deng H, Wang J, An R. Hyaluronic acid-based hydrogels: as an exosome delivery system in bone regeneration. Front Pharmacol. 2023;14:1131001. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR54] 54.Yang X, Gao J, Yang S, Wu Y, Liu H, Su D, Li D. Pore size-mediated macrophage M1 to M2 transition affects osseointegration of 3D-printed PEEK scaffolds. Int J Bioprint. 2023;9:755. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR55] 55.Yang X, Wu L, Li C, Li S, Hou W, Hao Y, Lu Y, Li L. Synergistic amelioration of osseointegration and osteoimmunomodulation with a microarc oxidation-treated three-dimensionally printed Ti-24Nb-4Zr-8Sn scaffold via surface activity and low elastic modulus. ACS Appl Mater Interfaces 2024;16:3171–86. [DOI] [PubMed] [Google Scholar]

[CR56] 56.Yang F, Li Y, Wang L, Che H, Zhang X, Jahr H, Wang L, Jiang D, Huang H, Wang J. Full-thickness osteochondral defect repair using a biodegradable bilayered scaffold of porous zinc and chondroitin sulfate hydrogel. Bioactive Mater. 2024;32:400–14. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR57] 57.Chmielewska A, Dean D. The role of stiffness-matching in avoiding stress shielding-induced bone loss and stress concentration-induced skeletal reconstruction device failure. Acta Biomater. 2024;173:51–65. [DOI] [PubMed] [Google Scholar]

[CR58] 58.Mushtaq RT, Askari GH, Bao C, Wang Y, Ahmed K, Khan AM, Sharma S, Alkahtani M. Optimization of 3D-printed bio-based super tough PLA (ST-PLA) scaffolds for cancellous bone regeneration: mechanical properties, lattice architecture, and osseointegration potential. Int J Biol Macromol. 2025;316:144466. [DOI] [PubMed] [Google Scholar]

[CR59] 59.Lv Z, Xu J, Cai X, Wang Y, Yang X, Wang H, Zhu Y, Feng B, Weng X. Exosome-based therapeutics for musculoskeletal disorders: advances in engineering, targeting, and biomaterial integration. ACS Nano. 2025;19:33681–716. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR60] 60.Hashemi A, Ezati M, Nasr MP, Zumberg I, Provaznik V. Extracellular vesicles and hydrogels: an innovative approach to tissue regeneration. ACS Omega. 2024;9:6184–218. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR61] 61.Zhao X, Chen X, Deng Y, Wu C, Ruan Z, Li C, Chen Y, Bei C, Zhu L, Yu H, Zhang X. A novel adhesive dual-sensitive hydrogel for sustained release of exosomes derived from M2 macrophages promotes repair of bone defects. Mater Today Bio. 2023;23:100840. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR62] 62.Lu W, Zeng M, Liu W, Ma T, Fan X, Li H, Wang Y, Wang H, Hu Y, Xie J. Human urine-derived stem cell exosomes delivered via injectable GelMA templated hydrogel accelerate bone regeneration. Mater Today Bio. 2023;19:100569. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR63] 63.Huang Z, Huang Y, Ning X, Li H, Li Q, Wu J. The functional effects of Piezo channels in mesenchymal stem cells. Stem Cell Res Ther. 2023;14:222. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR64] 64.Wang L, You X, Lotinun S, Zhang L, Wu N, Zou W. Mechanical sensing protein PIEZO1 regulates bone homeostasis via osteoblast-osteoclast crosstalk. Nat Commun. 2020;11:282. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR65] 65.Lin X, Su S, Duan X, Li W, Hu D, Lin Y, Jiang S, Wang L, Li J, Xu D, Qingchun Z. Abstract 4136901: mechanosensitive channel protein Piezo1 synergizes with TLR4 to enhance osteogenic responses in human aortic valve interstitial cells. Circulation 2024;150(Suppl_1):A4136901.

[CR66] 66.Atashi F, Modarressi A, Pepper MS. The role of reactive oxygen species in mesenchymal stem cell adipogenic and osteogenic differentiation: a review. Stem Cells Dev. 2015;24:1150–63. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR67] 67.Lin C, Yang Q, Guo D, Xie J, Yang Y-S, Chaugule S, DeSouza N, Oh W-T, Li R, Chen Z, et al. Impaired mitochondrial oxidative metabolism in skeletal progenitor cells leads to musculoskeletal disintegration. Nat Commun. 2022;13:6869. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR68] 68.Dai X, Yan X, Wintergerst KA, Cai L, Keller BB, Tan Y. Nrf2: redox and metabolic regulator of stem cell state and function. Trends Mol Med. 2020;26:185–200. [DOI] [PubMed] [Google Scholar]

[CR69] 69.Li S, Gao L, Zhang W, Yu Y, Zheng J, Liang X, Xin S, Ren W, Zhi K. MiR-152-5p suppresses osteogenic differentiation of mandible mesenchymal stem cells by regulating ATG14-mediated autophagy. Stem Cell Res Ther. 2022;13:359. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR70] 70.Santana Nunez D, Malik AB, Lee Q, Ahn SJ, Coctecon-Murillo A, Lazarko D, Levitan I, Mehta D, Komarova YA. Piezo1 induces endothelial responses to shear stress via soluble adenylyl Cyclase-IP(3)R2 circuit. iScience 2023;26:106661. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR71] 71.Wang S, Chennupati R, Kaur H, Iring A, Wettschureck N, Offermanns S. Endothelial cation channel PIEZO1 controls blood pressure by mediating flow-induced ATP release. J Clin Investig. 2016;126:4527–36. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR72] 72.Syeda R, Xu J, Dubin AE, Coste B, Mathur J, Huynh T, Matzen J, Lao J, Tully DC, Engels IH et al. Chemical activation of the mechanotransduction channel Piezo1. eLife 2015;4. [DOI] [PMC free article] [PubMed]

[CR73] 73.Grosso A, Burger MG, Lunger A, Schaefer DJ, Banfi A, Di Maggio N. It takes two to tango: coupling of angiogenesis and osteogenesis for bone regeneration. Front Bioeng Biotechnol. 2017;5:68. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR74] 74.Ma S, Wu J, Hu H, Mu Y, Zhang L, Zhao Y, Bian X, Jing W, Wei P, Zhao B, et al. Novel fusion peptides deliver exosomes to modify injectable thermo-sensitive hydrogels for bone regeneration. Mater Today Bio. 2022;13:100195. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR75] 75.Ma Z, Song W, He D, Zhang X, He Y, Li H. Smart µ-fiber hydrogels with macro‐porous structure for sequentially promoting multiple phases of articular cartilage regeneration. Adv Funct Mater. 2022;32:2113380.

[CR76] 76.Grosso A, Lunger A, Burger MG, Briquez PS, Mai F, Hubbell JA, Schaefer DJ, Banfi A. Di Maggio N: VEGF dose controls the coupling of angiogenesis and osteogenesis in engineered bone. npj Regen Med. 2023;8:15. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR77] 77.Dupont S, Morsut L, Aragona M, Enzo E, Giulitti S, Cordenonsi M, Zanconato F, Le Digabel J, Forcato M, Bicciato S, et al. Role of YAP/TAZ in mechanotransduction. Nature 2011;474:179–83. [DOI] [PubMed] [Google Scholar]

[CR78] 78.Aragona M, Panciera T, Manfrin A, Giulitti S, Michielin F, Elvassore N, Dupont S, Piccolo S. A mechanical checkpoint controls multicellular growth through YAP/TAZ Regulation by actin-processing factors. Cell 2013;154:1047–59. [DOI] [PubMed] [Google Scholar]

[CR79] 79.Zheng LZ, Wang JL, Kong L, Huang L, Tian L, Pang QQ, Wang XL, Qin L. Steroid-associated osteonecrosis animal model in rats. J Orthop Translat. 2018;13:13–24. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR80] 80.Dong Y, Li Y, Huang C, Gao K, Weng X. Systemic application of teriparatide for steroid induced osteonecrosis in a rat model. BMC Musculoskelet Disord. 2015;16:163. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Exosome-mediated Piezo1 activation in 3D-printed titanium scaffolds promotes repair of femoral head osteonecrosis

Yang Yu

Zhongyin Ji

Hongjun Xu

Mingyang Ma

Zhaojing Yin

Yiyang Du

Tao Zhang

Guiguan Wang

Feng Tian

Sen Liu

Jiajia Xue

Wenwei Qian

Abstract

Graphical abstract

Supplementary Information

Background

Scheme 1.

Results and discussion

Design and characterization of the P-Exos@HAMA/Ti platform

Fig. 1.

Engineering and characterization of P-Exos

Fig. 2.

In vitro cytocompatibility and baseline cellular performance

Fig. 3.

In vitro osteogenic and angiogenic responses

Fig. 4.

Fig. 5.

RNA-seq and metabolomics reveal Ca2+-YAP1/β-catenin activation

Fig. 6.

In vivo biosafety and therapeutic evaluation in ONFH model

Fig. 7.

Fig. 8.

Fig. 9.

Conclusions

Materials and methods

Chemicals and biological reagents

Fabrication and key parameters of titanium lattice scaffold (Ti)

Synthesis of exosomes and P-Exos

Synthesis of HAMA hydrogels and assembly of P‑Exos@HAMA/Ti

Material characterization

In vitro biocompatibility and cell function assays

In vitro osteogenic differentiation and angiogenesis assays

Transcriptomic profiling, untargeted metabolomics, and Ca2+-YAP1/β-catenin pathway validation

In vivo biocompatibility evaluation

In vivo ONFH model on rats

Statistical analysis

Supplementary Information

Acknowledgements

Author contributions

Funding

Data availability

Declarations

Ethics approval and consent to participate

Consent for publication

Competing interests

Footnotes

Contributor Information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

RNA-seq and metabolomics reveal Ca²⁺-YAP1/β-catenin activation

Transcriptomic profiling, untargeted metabolomics, and Ca²⁺-YAP1/β-catenin pathway validation