Bioelectric signals promote diabetic bone regeneration through Piezo1-mediated activation of the efferocytic immune microenvironment Authors: Dilixiati Duolikun, Lei Zhang, Tianlong Wang, Xinhui Wu, Xuran Guo, Waheed Muhammad Zeeshan, Zhiqing Liu, Haijun Zhang, Jinhui Zhao, Longpo Zheng Published: 2026 ABSTRACT Defective macrophage efferocytosis sustains inflammation and compromises bone repair in diabetic defects, yet strategies that restore efferocytic immune homeostasis remain limited. Here, we engineered an ultrasound-responsive piezoelectric hydrogel (Gel BC@ZnO) by incorporating zinc oxide (ZnO) nanoparticles into a bacterial cellulose (BC)-reinforced gelatin methacryloyl (GelMA) network to provide localized bioelectric cues. Gel BC@ZnO exhibited good cytocompatibility, structural stability, and robust ultrasound-triggered piezoelectric output. Under diabetic-mimicking conditions, ultrasound activation generated bioelectric signals that engaged Piezo1, induced Piezo1-dependent Ca2+ influx, and rescued macrophage efferocytosis. This efferocytic reactivation attenuated ROS accumulation, shifted macrophages toward a pro-regenerative phenotype, enhanced osteogenic differentiation, and suppressed osteoclast genesis, thereby restoring bone-remodeling homeostasis. In a diabetic femoral condyle defect model, Gel BC@ZnO combined with ultrasound accelerated bone repair, increased local Piezo1 expression, promoted collagen deposition and osteogenic marker expression, and rebuilt a favorable immune microenvironment. These results define a bioelectric signal-Piezo1-Ca2+-efferocytosis axis for diabetic bone regeneration and establish a non-invasive, ultrasound-controllable immunomodulatory strategy for compromised bone repair. Introduction Diabetes mellitus (DM) imposes a major health burden and frequently delays bone healing [[1], [2], [3]]. The diabetic bone microenvironment combines hyperglycemia, chronic inflammation, vascular dysfunction, and immune dysregulation, leading to excessive apoptosis and accumulation of uncleared dying cells [4]. Failure to remove these cells sustains inflammatory signaling, disrupts local immune homeostasis, and impairs regeneration [5,6]. Efferocytosis, the macrophage-mediated clearance of apoptotic cells, is a critical cellular debridement process that resolves inflammation, prevents secondary necrosis, and initiates repair [7]. Thus, defective efferocytosis represents a key mechanistic link between diabetes and impaired bone regeneration [8]. Efferocytosis is not simply a clearance event; it actively reprograms tissue immunity through anti-inflammatory, pro-resolving, and pro-regenerative signaling networks [[9], [10], [11]]. Macrophages are the principal efferocytic cells in bone repair and coordinate inflammatory resolution, immune homeostasis, and regenerative signaling [8]. Under physiological conditions, macrophages rapidly engulf apoptotic cells, limit inflammatory amplification, and establish a pro-resolving microenvironment that supports tissue regeneration [12]. This process also promotes M2-like, pro-regenerative macrophage remodeling [13]. In diabetes, excessive apoptosis overwhelms macrophage efferocytic capacity, reinforces M1-biased inflammation, and sustains immune dysfunction [14,15]. The resulting loop of apoptotic burden, impaired clearance, and inflammatory polarization fundamentally restricts diabetic bone repair [16]. Restoring macrophage efferocytosis and reprogramming the diabetic immune microenvironment are therefore rational strategies for improving bone regeneration. Piezo1 has recently emerged as a mechanosensitive regulator of macrophage efferocytosis and immune reprogramming, making it a relevant signaling node for diabetic bone repair [[17], [18], [19]]. Piezo1 also functions as a stretch-activated cation channel in bone cells, where it transduces mechanical cues to regulate osteoblast and osteoclast activities and maintain bone homeostasis [20,21]. Therefore, Piezo1 offers a target that may link immune regulation with bone remodeling [22]. Bioelectric cues generated by mechanically responsive materials provide a non-invasive and spatiotemporally controllable approach to modulate Piezo1 activity [23]. Piezoelectric biomaterials can potentiate Piezo1-related osteogenic signaling while regulating macrophage function [24]. However, the mechanistic sequence by which localized bioelectric signals activate Piezo1, trigger Ca2+ entry, restore macrophage efferocytosis, and ultimately reshape diabetic bone repair remains insufficiently defined. Piezoelectric materials convert mechanical stimulation into electrical signals and are therefore attractive platforms for programmable bioelectric regulation of tissue repair [25,26]. Zinc oxide (ZnO) is particularly suitable because of its facile synthesis, favorable semiconductor properties, and strong piezoelectric performance compared with many conventional ceramics and organic materials [[27], [28], [29]]. For bone-defect repair, gelatin methacryloyl (GelMA) hydrogels are widely used as cell-compatible scaffolds [30,31], but their limited mechanical strength restricts bone applications [32]. Bacterial cellulose (BC), with its nanofibrous architecture and high mechanical robustness, can reinforce GelMA hydrogels [[33], [34], [35]]. In this work, BC strengthened the GelMA network, whereas optimized ZnO incorporation endowed the scaffold with ultrasound-triggered piezoelectric activity, yielding a biocompatible hydrogel capable of delivering localized bioelectric stimulation. We therefore hypothesized that an ultrasound-responsive piezoelectric scaffold could promote diabetic bone regeneration through a defined bioelectric signal-Piezo1-Ca2+-efferocytosis pathway. In this mechanism, ultrasound activates ZnO-mediated piezoelectric output within Gel BC@ZnO, the resulting bioelectric cues engage Piezo1 and induce Ca2+ influx, and Piezo1-dependent Ca2+ signaling restores macrophage efferocytosis, reduces inflammatory oxidative stress, and shifts macrophages toward a pro-regenerative state. The restored efferocytic immune microenvironment is expected to couple inflammatory resolution with osteogenic enhancement and osteoclast suppression (Scheme 1). By validating this pathway in vitro and in a diabetic bone-defect model, this study aims to clarify how bioelectric cues coordinate immune regulation and bone remodeling, providing a mechanistic basis for bioelectric immunomodulatory therapy in diabetic bone repair.Scheme 1Schematic illustration of Gel BC@ZnO synthesis and the proposed bioelectric signal-Piezo1-Ca2+-efferocytosis mechanism for diabetic bone repair. Schematic illustration of Gel BC@ZnO synthesis and the proposed bioelectric signal-Piezo1-Ca2+-efferocytosis mechanism for diabetic bone repair. Results and discussion Preparation and characterization of Gel BC@ZnO composite hydrogels Scaffold-derived three-dimensional cues, particularly ECM-like architecture and electrical microenvironments, strongly influence early cell attachment, migration, and osteogenic differentiation of bone marrow-derived mesenchymal stem cells [36]. Gelatin methacryloyl (GelMA; abbreviated as Gel) resembles extracellular matrix components and exhibits favorable cytocompatibility, making it a widely used hydrogel scaffold for bone regeneration [37,38]. Here, we designed a biodegradable Gel BC@ZnO hydrogel with a hierarchical porous structure and an ultrasound-responsive electrical microenvironment inspired by ECM fibers and endogenous bone bioelectricity (Fig. 1A). SEM showed that the ZnO nanoparticles were spherical with an average diameter of approximately 50 nm (Fig. S1). Macroscopic images confirmed changes in hydrogel appearance after BC and ZnO incorporation (Fig. 1B). FTIR and XRD analyses verified the chemical composition and crystalline phase of the composites (Fig. 1C and D). The FTIR spectrum of Gel BC@ZnO showed characteristic GelMA amide bands and a broad -OH/-NH band at 3200-3600 cm−1, together with a Zn-O vibration band at 525-430 cm−1, confirming ZnO incorporation [39]. XRD peaks at 31.5° (100), 34.3° (002), 36.2° (101), 47.6° (102), 56.5° (110), 62.8° (103), and 67.8° (112) matched the hexagonal wurtzite phase of ZnO [40]. SEM images revealed interconnected porous structures that support cell infiltration and migration, and EDS mapping showed homogeneous distribution of C, N, O, and Zn throughout Gel BC@ZnO (Fig. 1E). Degradation testing showed that BC and ZnO slightly slowed scaffold degradation, whereas ultrasound accelerated degradation to a limited extent (Fig. S2). The surrounding solution remained at pH 7.2-7.5 (Fig. S3), indicating that scaffold degradation did not markedly disturb local pH.Fig. 1Structural and physicochemical characterization of Gel BC@ZnO hydrogels. A) Schematic of GelMA synthesis and incorporation of BC and ZnO. B) Photographs of GelMA-based hydrogels with different BC and ZnO contents. C) FTIR spectra showing characteristic functional groups of Gel BC@ZnO composites. D) XRD patterns confirming ZnO crystalline phases in the composites. E) SEM images of GelMA, GelMA@ZnO, Gel BC, and Gel BC@ZnO hydrogels, with elemental mapping of C, N, O, and Zn. Structural and physicochemical characterization of Gel BC@ZnO hydrogels. A) Schematic of GelMA synthesis and incorporation of BC and ZnO. B) Photographs of GelMA-based hydrogels with different BC and ZnO contents. C) FTIR spectra showing characteristic functional groups of Gel BC@ZnO composites. D) XRD patterns confirming ZnO crystalline phases in the composites. E) SEM images of GelMA, GelMA@ZnO, Gel BC, and Gel BC@ZnO hydrogels, with elemental mapping of C, N, O, and Zn. Properties and biosafety of the Gel BC@ZnO composite hydrogel Mechanical properties are critical for osteogenic regulation [41]. Increasing BC content significantly enhanced the compressive modulus, demonstrating effective reinforcement of GelMA (Fig. 2A and B). Piezoelectric output was then measured under ultrasound (US; 1.0 W cm−2). The peak-to-peak voltage (Vpp) of Gel BC@ZnO increased with ZnO content (0-5 wt%), confirming composition-dependent piezoelectric responsiveness (Fig. 2C). Because excessive ZnO may induce cytotoxicity, CCK-8 assays were used to select the ZnO dose. Gel BC and Gel BC@1%ZnO maintained high cell viability, whereas Gel BC@3%ZnO and Gel BC@5%ZnO significantly reduced viability (Fig. S4). Therefore, 1 wt% ZnO was used for subsequent experiments. RAW264.7 macrophages cultured with different scaffolds showed no significant viability differences during the first 3 days and no obvious cytotoxicity compared with the blank control. By day 5, Gel BC@ZnO + US promoted higher cell viability than the other groups (Fig. 2D). Live/dead staining confirmed minimal cell death and increasing cell density over time (Fig. 2E), and flow cytometry showed no detectable early or late apoptosis under any condition (Fig. 2F). These results demonstrate that Gel BC@ZnO is cytocompatible and that the applied ultrasound does not compromise macrophage viability.Fig. 2Mechanical, piezoelectric, and biosafety properties of Gel BC@ZnO hydrogels. A) Photographs of GelMA-based hydrogels with different BC and ZnO contents. B) Stress-strain curves of Gel BC@ZnO composites. C) Voltage-time response under ultrasound stimulation. D) CCK-8 analysis of cell viability on days 1, 3, and 5. E) Live/dead staining images after 1 and 3 days of culture. F) Flow-cytometry analysis of apoptosis using 7-AAD and Annexin V-PE staining. Data are presented as mean ± SD; error bars indicate SD. Statistical analysis for Fig. 2D was performed using two-way ANOVA followed by Tukey's multiple-comparison test. ∗P < 0.05, n = 3. Mechanical, piezoelectric, and biosafety properties of Gel BC@ZnO hydrogels. A) Photographs of GelMA-based hydrogels with different BC and ZnO contents. B) Stress-strain curves of Gel BC@ZnO composites. C) Voltage-time response under ultrasound stimulation. D) CCK-8 analysis of cell viability on days 1, 3, and 5. E) Live/dead staining images after 1 and 3 days of culture. F) Flow-cytometry analysis of apoptosis using 7-AAD and Annexin V-PE staining. Data are presented as mean ± SD; error bars indicate SD. Statistical analysis for Fig. 2D was performed using two-way ANOVA followed by Tukey's multiple-comparison test. ∗P < 0.05, n = 3. Ultrasound-triggered piezoelectric activation restores macrophage efferocytosis To define the diabetic inflammatory baseline, macrophages were first compared under normal-glucose conditions (Control) and high-glucose conditions (35 mM; diabetic group). Subsequent scaffold-based experiments were conducted under high-glucose conditions, with the untreated high-glucose group defined as Control. Compared with the normal-glucose Control, the diabetic group showed a markedly impaired bone-repair microenvironment. High glucose increased apoptosis-related gene expression (Fig. S5) and strengthened pro-inflammatory macrophage signals (Fig. S6). More importantly, macrophage efferocytosis was sharply reduced: under the same gating strategy, the proportion of macrophages engulfing apoptotic debris decreased from 48.36% in the Control group to 6.59% in the diabetic group (Fig. S7). High glucose also changed macrophage morphology from a rounded shape to a protrusion-rich, activated phenotype (Fig. S8), and flow cytometry confirmed increased M1 marker expression (Fig. S9). These data show that hyperglycemia simultaneously increases apoptotic burden, weakens efferocytic clearance, and drives pro-inflammatory macrophage polarization. Piezo1 is a mechanically gated cation channel and a key mechanotransducer in skeletal homeostasis [23,42]. We next examined whether ultrasound-activated Gel BC@ZnO could engage Piezo1 in macrophages. Piezo1 immunofluorescence was comparable between the Control and Gel BC groups, and Gel BC@ZnO produced only a modest, statistically insignificant increase. In contrast, Gel BC@ZnO + US markedly enhanced Piezo1 fluorescence, indicating activation of Piezo1-associated signaling by ultrasound-triggered bioelectric cues (Fig. 3A and D) [24,43]. Flow cytometry showed similar CFSE-positive macrophage fractions in the Control and Gel BC groups, a moderate increase in the Gel BC@ZnO group, and a pronounced increase in the Gel BC@ZnO + US group, where CFSE-positive macrophages exceeded 50% (Fig. 3B).Fig. 3Ultrasound-triggered Gel BC@ZnO promotes Piezo1 activation and macrophage efferocytosis. A) Piezo1 immunofluorescence staining. B) Representative flow-cytometry plots showing CFSE signal in gated macrophages after co-culture with CFSE-labeled apoptotic Jurkat cells. Before co-culture, camptothecin-induced Jurkat cells were quality-checked by Annexin V/7-AAD staining to ensure an apoptosis-enriched target-cell population and minimize necrotic/debris contamination. C) Fluorescence images showing intracellular CFSE signals in macrophages. D) Quantification of Piezo1 immunofluorescence. E) Quantification of CFSE-positive macrophages by flow cytometry. F) Percentage of efferocytic macrophages, defined as F4/80+CFSE + events among total F4/80+ macrophages after exclusion of debris, doublets, and F4/80-negative free apoptotic cells/debris. G) Quantification of intracellular CFSE fluorescence intensity. Data are presented as mean ± SD; error bars indicate SD. Statistical comparisons were performed using one-way ANOVA followed by Tukey's multiple-comparison test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n = 3. Ultrasound-triggered Gel BC@ZnO promotes Piezo1 activation and macrophage efferocytosis. A) Piezo1 immunofluorescence staining. B) Representative flow-cytometry plots showing CFSE signal in gated macrophages after co-culture with CFSE-labeled apoptotic Jurkat cells. Before co-culture, camptothecin-induced Jurkat cells were quality-checked by Annexin V/7-AAD staining to ensure an apoptosis-enriched target-cell population and minimize necrotic/debris contamination. C) Fluorescence images showing intracellular CFSE signals in macrophages. D) Quantification of Piezo1 immunofluorescence. E) Quantification of CFSE-positive macrophages by flow cytometry. F) Percentage of efferocytic macrophages, defined as F4/80+CFSE + events among total F4/80+ macrophages after exclusion of debris, doublets, and F4/80-negative free apoptotic cells/debris. G) Quantification of intracellular CFSE fluorescence intensity. Data are presented as mean ± SD; error bars indicate SD. Statistical comparisons were performed using one-way ANOVA followed by Tukey's multiple-comparison test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n = 3. To directly assess efferocytosis, apoptotic Jurkat cells were labeled with CFSE and co-cultured with F4/80-identified macrophages. Gel BC@ZnO + US significantly increased intracellular CFSE signals within macrophages and raised the proportion of F4/80+CFSE + cells, demonstrating more efficient engulfment of apoptotic Jurkat cells (Fig. 3C, F and 3G). Because Piezo1 is the predominant mechanosensitive ion channel in macrophages, it provides a plausible conduit for converting ZnO-derived bioelectric stimulation into immune function [44,45]. These results support the first step of the proposed mechanism: ultrasound activates ZnO-mediated piezoelectric output, which engages Piezo1-associated macrophage mechanotransduction and restores efferocytic clearance under diabetic-mimicking conditions [28,46]. Piezo1-dependent Ca2+ influx couples bioelectric activation to macrophage immunomodulation in vitro To determine whether Gel BC@ZnO + US functionally activated Piezo1, rather than only increasing Piezo1 expression, we measured intracellular Ca2+ influx in macrophages and osteoblasts under high-glucose conditions and used GsMTx4 to inhibit Piezo1. In macrophages, FITC-Ca2+-positive cells were comparable in the Control and Gel BC groups (22.49% and 22.59%) and increased to 33.78% in the Gel BC@ZnO group. Gel BC@ZnO + US further elevated this population to 53.74%, whereas GsMTx4 reduced it to 38.19% (Fig. 4A and B). Fluorescence imaging showed the same pattern, with the strongest Ca2+ signal in Gel BC@ZnO + US and a clear reduction after Piezo1 blockade (Fig. 4C and D). In osteoblasts, Gel BC@ZnO + US increased FITC-Ca2+-positive cells from 10.66% to 12.59% in the Control/Gel BC/Gel BC@ZnO groups to 21.82%, and GsMTx4 decreased this value to 15.30% (Fig. 4E-H). These results verify that ultrasound-responsive Gel BC@ZnO induces Piezo1-dependent Ca2+ entry in both immune and osteogenic cells, establishing a functional bioelectric-Piezo1-Ca2+ mechanotransduction pathway [44,45,[47], [48], [49]].Fig. 4Gel BC@ZnO + US induces Piezo1-dependent Ca2+ influx and macrophage immunomodulation in vitro. A, B) Representative flow-cytometry plots and quantification of FITC-Ca2+-positive macrophages. C, D) Representative fluorescence images and MFI quantification of intracellular Ca2+ in macrophages. E, F) Representative flow-cytometry plots and quantification of FITC-Ca2+-positive osteoblastic cells. G, H) Representative fluorescence images and MFI quantification of intracellular Ca2+ in osteoblastic cells. I) Flow-cytometry analysis of macrophage polarization based on CD206 and CD86 expression. J, K) Representative flow-cytometry histograms and quantification of intracellular ROS. GsMTx4 was used as a Piezo1 inhibitor. Data are presented as mean ± SD; error bars indicate SD. Statistical significance was determined by one-way ANOVA with Tukey's multiple-comparison test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n = 3. Scale bars: 100 μm in C and 150 μm in G. Gel BC@ZnO + US induces Piezo1-dependent Ca2+ influx and macrophage immunomodulation in vitro. A, B) Representative flow-cytometry plots and quantification of FITC-Ca2+-positive macrophages. C, D) Representative fluorescence images and MFI quantification of intracellular Ca2+ in macrophages. E, F) Representative flow-cytometry plots and quantification of FITC-Ca2+-positive osteoblastic cells. G, H) Representative fluorescence images and MFI quantification of intracellular Ca2+ in osteoblastic cells. I) Flow-cytometry analysis of macrophage polarization based on CD206 and CD86 expression. J, K) Representative flow-cytometry histograms and quantification of intracellular ROS. GsMTx4 was used as a Piezo1 inhibitor. Data are presented as mean ± SD; error bars indicate SD. Statistical significance was determined by one-way ANOVA with Tukey's multiple-comparison test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n = 3. Scale bars: 100 μm in C and 150 μm in G. We then tested whether this Piezo1-Ca2+ response translated into immune remodeling. CD206/CD86 flow cytometry showed that Gel BC@ZnO + US increased the CD206-associated pro-regenerative macrophage population and reduced the CD86-single-positive pro-inflammatory fraction, whereas GsMTx4 partially weakened this shift (Fig. 4I). ROS analysis further showed that ROS-positive macrophages decreased from 26.30% in the Control group and 24.42% in the Gel BC group to 13.31% in the Gel BC@ZnO group and 9.13% in the Gel BC@ZnO + US group; GsMTx4 partially reversed this decrease to 15.47% (Fig. 4J and K). Together with the efferocytosis data, these results clarify the mechanistic sequence: ultrasound-triggered ZnO piezoelectricity activates Piezo1, promotes Ca2+ influx, restores macrophage efferocytosis, attenuates oxidative inflammatory stress, and drives pro-regenerative immune remodeling in the diabetic-like microenvironment. Ultrasound-induced piezoelectric stimulation restores bone-remodeling homeostasis ALP and ARS staining showed no significant differences between the Control and Gel BC groups, indicating that BC alone exerted limited osteogenic activity under hyperglycemic conditions (Fig. 5A and B). Gel BC@ZnO significantly increased ALP activity and calcium nodule formation, suggesting that ZnO promotes osteoblastic differentiation and matrix mineralization. This effect may involve ZnO-induced AMPK-ULK1 activation and zinc-ion release, both of which have been associated with bone regeneration and osteogenic differentiation [50,51]. After ultrasound stimulation, Gel BC@ZnO further enhanced osteogenesis. Mechanistically, Gel BC@ZnO + US activates Piezo1-dependent Ca2+ influx, which can stimulate ERK1/2 phosphorylation and F-actin assembly; F-actin remodeling further promotes YAP nuclear translocation and Cyclin D1 expression to support osteoblast cell-cycle progression and differentiation [[47], [48], [49]]. Quantitative ALP and ARS analyses confirmed that ultrasound-induced piezoelectric stimulation enhances osteogenic differentiation through Piezo1-associated signaling (Fig. 5E and F).Fig. 5Ultrasound-induced piezoelectric stimulation regulates bone-remodeling homeostasis. A) ALP staining. B) ARS staining. C) TRAP staining. D) F-actin staining. E) Quantitative analysis of ALP activity. F) Quantitative analysis of calcium nodules. G) Quantitative analysis of osteoclast area. H) Quantitative analysis of osteoclast number. Data are presented as mean ± SD; error bars indicate SD. Statistical comparisons were performed using one-way ANOVA followed by Tukey's multiple-comparison test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n = 3. Ultrasound-induced piezoelectric stimulation regulates bone-remodeling homeostasis. A) ALP staining. B) ARS staining. C) TRAP staining. D) F-actin staining. E) Quantitative analysis of ALP activity. F) Quantitative analysis of calcium nodules. G) Quantitative analysis of osteoclast area. H) Quantitative analysis of osteoclast number. Data are presented as mean ± SD; error bars indicate SD. Statistical comparisons were performed using one-way ANOVA followed by Tukey's multiple-comparison test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n = 3. TRAP staining and F-actin immunofluorescence showed abundant multinucleated osteoclast-like cells with clear actin rings in the Control and Gel BC groups, whereas Gel BC@ZnO reduced osteoclast size, number, and F-actin organization (Fig. 5C and D). This suppression was further strengthened by ultrasound in the Gel BC@ZnO + US group. Quantification confirmed that Gel BC alone had little effect on osteoclast area or number, whereas ZnO incorporation significantly decreased both parameters and ultrasound amplified the inhibition (Fig. 5G and H). Because actin-ring formation is required for sealing-zone formation and bone resorption [52], these data indicate that Gel BC@ZnO + US inhibits osteoclast differentiation and functional maturation under hyperglycemic conditions. Zinc ions may contribute by modulating NF-κB activation and the Ca2+-calcineurin-NFATc1 pathway [53,54], while ultrasound-responsive piezoelectric stimulation may further restrain osteoclastogenesis [55]. Although these downstream pathways require direct validation, the present data support a dual bone-homeostatic effect: enhanced osteogenesis and suppressed osteoclast formation. Ultrasound-activated Gel BC@ZnO accelerates diabetic bone-defect repair in vivo To assess whether ultrasound-induced piezoelectric stimulation promotes bone repair in vivo, we established a critical-sized femoral condyle defect (φ 3 mm × 4 mm) in diabetic rats (Fig. S10–S12). After scaffold implantation, ultrasound was applied to activate the piezoelectric response. Micro-CT reconstruction showed minimal spontaneous repair in the Control group, confirming the impaired healing capacity of diabetic bone defects (Fig. 6A). Gel BC partially reduced the defect, and Gel BC@ZnO further improved defect closure. The Gel BC@ZnO + US group achieved the most extensive repair, with near-complete reconstruction of the femoral condyle architecture.Fig. 6Gel BC@ZnO + US accelerates diabetic bone-defect repair in vivo. A) Micro-CT images showing new bone formation in diabetic rat femoral condyle defects 8 weeks after surgery. B) Quantitative micro-CT analysis of regenerated bone. C) H&E and Masson's trichrome staining after 8 weeks. D) Piezo1 immunofluorescence staining in the defect area. E) Semi-quantitative analysis of Piezo1-positive area. Data are presented as mean ± SD; error bars indicate SD. Statistical comparisons were performed using one-way ANOVA followed by Tukey's multiple-comparison test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n = 3. Gel BC@ZnO + US accelerates diabetic bone-defect repair in vivo. A) Micro-CT images showing new bone formation in diabetic rat femoral condyle defects 8 weeks after surgery. B) Quantitative micro-CT analysis of regenerated bone. C) H&E and Masson's trichrome staining after 8 weeks. D) Piezo1 immunofluorescence staining in the defect area. E) Semi-quantitative analysis of Piezo1-positive area. Data are presented as mean ± SD; error bars indicate SD. Statistical comparisons were performed using one-way ANOVA followed by Tukey's multiple-comparison test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n = 3. Quantitative micro-CT analysis supported these observations (Fig. 6B). Bone volume fraction (BV/TV), trabecular thickness (Tb.Th), and bone mineral density (BMD) progressively increased across the treatment groups, whereas trabecular separation (Tb.Sp) decreased. The Gel BC@ZnO + US group showed the most favorable microstructural profile, indicating that ultrasound-triggered piezoelectric stimulation substantially improved diabetic bone regeneration. H&E and Masson's trichrome staining further confirmed enhanced tissue repair (Fig. 6C). Gel BC@ZnO and Gel BC@ZnO + US promoted new bone formation and collagen deposition, with Gel BC@ZnO + US showing a more continuous and mature bone matrix. Piezo1 immunofluorescence provided in vivo support for engagement of the proposed pathway (Fig. 6D and E). Piezo1 expression was higher in Gel BC than in Control, further increased in Gel BC@ZnO, and reached the highest level in Gel BC@ZnO + US. This graded response is consistent with scaffold-mediated mechanotransduction and ultrasound-amplified ZnO piezoelectric activation. Together with the in vitro GsMTx4 and Ca2+ data, these findings support a mechanistic link among localized bioelectric stimulation, Piezo1 activation, efferocytic immune remodeling, and enhanced diabetic bone repair. Osteogenic maturation and late-stage immune remodeling in the femoral condyle defect area COL I and OCN were used to evaluate late-stage osteogenic maturation. COL I was mainly localized in the extracellular matrix (Fig. 7A). The Control group showed minimal COL I expression, whereas Gel BC induced scattered collagen deposition within fibrous tissue. Gel BC@ZnO increased COL I deposition within newly formed bone matrix, and Gel BC@ZnO + US produced the most pronounced matrix-associated COL I signal, indicating more advanced bone formation. Quantification showed a higher COL I-positive area in Gel BC than in Control; although Gel BC and Gel BC@ZnO showed no significant difference in COL I-positive area, Gel BC@ZnO displayed more organized bone tissue. Gel BC@ZnO + US showed the largest COL I-positive area (Fig. 7E). OCN staining showed the same trend: minimal expression in Control, increased expression in Gel BC and Gel BC@ZnO, and the highest OCN positivity in Gel BC@ZnO + US (Fig. 7B and F). These data confirm that ultrasound-activated Gel BC@ZnO promotes osteogenic maturation in diabetic bone defects.Fig. 7Osteogenic maturation and late-stage immune remodeling in femoral condyle defects at 8 weeks after surgery. A) COL I immunohistochemical staining. B) OCN immunohistochemical staining. C) iNOS immunofluorescence staining. D) Arg1 immunofluorescence staining. E) Quantification of COL I-positive area. F) Quantification of OCN-positive area. G) Quantification of iNOS immunofluorescence. H) Quantification of Arg1 immunofluorescence. The iNOS/Arg1 data reflect late healing/remodeling immune status rather than the initial inflammatory transition. Data are presented as mean ± SD; error bars indicate SD. Statistical comparisons were performed using one-way ANOVA followed by Tukey's multiple-comparison test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n = 3. Osteogenic maturation and late-stage immune remodeling in femoral condyle defects at 8 weeks after surgery. A) COL I immunohistochemical staining. B) OCN immunohistochemical staining. C) iNOS immunofluorescence staining. D) Arg1 immunofluorescence staining. E) Quantification of COL I-positive area. F) Quantification of OCN-positive area. G) Quantification of iNOS immunofluorescence. H) Quantification of Arg1 immunofluorescence. The iNOS/Arg1 data reflect late healing/remodeling immune status rather than the initial inflammatory transition. Data are presented as mean ± SD; error bars indicate SD. Statistical comparisons were performed using one-way ANOVA followed by Tukey's multiple-comparison test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n = 3. To evaluate whether the efferocytosis-centered mechanism was accompanied by immune remodeling in vivo, we analyzed iNOS and Arg1 expression at 8 weeks after surgery. These data represent the late healing/remodeling phase rather than the early inflammatory-to-reparative transition. iNOS and Arg1 were used as representative markers of pro-inflammatory M1 and pro-regenerative M2 macrophage phenotypes, respectively [56]. iNOS fluorescence remained high in the Control and Gel BC groups, indicating persistent inflammation during late repair (Fig. 7C and G). Gel BC@ZnO partially reduced iNOS expression, while Gel BC@ZnO + US produced the strongest suppression, consistent with improved inflammatory resolution. Conversely, Arg1 fluorescence was lowest in the Control group and only slightly increased in the Gel BC group. Gel BC@ZnO significantly enhanced Arg1 expression, potentially reflecting zinc-ion-mediated M2-like remodeling [57], whereas Gel BC@ZnO + US induced the highest Arg1 signal (Fig. 7D and H). These results are consistent with ultrasound-enhanced efferocytosis promoting a pro-regenerative immune state [[58], [59], [60]]. Importantly, because samples were collected at 8 weeks, the iNOS/Arg1 data should be interpreted as late-stage immune remodeling; future time-course studies are needed to resolve early macrophage dynamics. Subcutaneous implantation further showed detectable scaffold residues without severe inflammation at 3 weeks, accompanied by tissue ingrowth and collagen deposition around Gel BC@ZnO, particularly under ultrasound (Fig. S14). Histological examination of the heart, liver, spleen, lung, kidney, and brain revealed no obvious abnormalities, supporting the in vivo biocompatibility and low systemic toxicity of Gel BC@ZnO (Fig. S15). To position Gel BC@ZnO among biomaterial therapies for inflammatory bone defects, we compared it with representative stimulus-responsive systems. Mild photothermal hydrogels, soft-hard composite scaffolds, injectable multifunctional hydrogels, and polyphenol-functionalized photothermal platforms have shown that regulating inflammation, ROS stress, angiogenesis, macrophage polarization, and osteogenic differentiation can improve bone repair under diabetic or inflammatory conditions [[61], [62], [63], [64], [65]]. Gel BC@ZnO differs by converting external ultrasound into localized ZnO-mediated bioelectric cues. Rather than acting only through anti-inflammatory or osteogenic effects, the platform activates Piezo1-dependent Ca2+ signaling to restore macrophage efferocytosis, thereby linking immune resolution, osteoclast suppression, osteogenic enhancement, and diabetic bone regeneration. Existing Piezo1-based bone-regeneration strategies mainly rely on mechanical loading/exercise or pharmacological Piezo1 activation. Mechanical stimulation transmits deformation-dependent cues to osteoblasts and osteocytes and regulates osteoblast-osteoclast crosstalk, thereby promoting bone anabolism [22,23,47]. Pharmacological Piezo1 activation can enhance mesenchymal stem-cell proliferation and osteogenic differentiation, but dose, localization, and systemic exposure require careful control [18]. By contrast, Gel BC@ZnO + US converts external ultrasound into defect-localized bioelectric signals through ZnO piezoelectricity. This design enables non-invasive and temporally controllable Piezo1 activation while coupling Piezo1 signaling to macrophage efferocytosis, inflammatory resolution, osteoclast suppression, and osteogenic differentiation. Thus, the mechanism differs from conventional mechanical or pharmacological activation by integrating bioelectric stimulation with immune-microenvironment remodeling in diabetic bone defects. Piezo1-regulated outcomes are highly context dependent. In osteoporotic or unloading-related bone loss, Piezo1 activation mainly restores bone homeostasis by stimulating osteogenic mechanotransduction and balancing osteoblast-osteoclast activity [49,55]. In infected bone defects, ultrasound-responsive piezoelectric or Zn-containing scaffolds often emphasize antibacterial and immunoregulatory functions in addition to osteogenesis [28,46]. Diabetic bone defects, however, are characterized by hyperglycemia, persistent inflammation, excess apoptosis, and impaired efferocytosis. Therefore, the central feature of the present study is not simply increasing Piezo1 expression, but using localized bioelectric Piezo1 activation to restore efferocytic immune clearance and convert a chronically inflamed diabetic microenvironment into a pro-resolving, pro-regenerative niche. This microenvironment-specific mechanism supports the advantage of Gel BC@ZnO + US for regeneration-impaired diabetic bone repair, although direct comparisons with mechanical stimulation and Piezo1 agonists remain needed. Compared with clinical approaches such as bone grafting and growth factor delivery, which mainly provide structural or osteoinductive support and may be limited by donor-site morbidity, graft availability, variable efficacy in diabetic microenvironments, and dose-related complications, Gel BC@ZnO + US offers a localized, ultrasound-tunable bioelectric strategy that actively remodels the immune microenvironment through Piezo1-mediated efferocytosis [66,67]. Several limitations remain. Although scaffold degradation was evaluated in a subcutaneous implantation model, the observation period was shorter than the 8-week bone-repair endpoint. Thus, the long-term degradation profile and duration of bioelectric output require further investigation. Clinical translation will also require optimization of ultrasound delivery. Insufficient acoustic activation may weaken the piezoelectric response, whereas off-target or excessive exposure may compromise reproducibility and safety. Image-guided, defect-specific ultrasound protocols should therefore integrate preoperative CT/ultrasound localization, acoustic-field calibration, and standardized treatment windows for intensity, duty cycle, duration, frequency, and coupling conditions. Monitoring local temperature and tissue responses would help balance Piezo1 activation with biosafety. Finally, Gel BC@ZnO implantation must be compatible with irregular defects and routine surgical workflows. The injectable precursor and in situ photocrosslinking design support minimally invasive defect filling, but translation will require sterile closed delivery devices, imaging-assisted patient-specific molds or 3D-printed templates, and clinically compatible crosslinking strategies for deep or poorly accessible defects. Because diabetic patients differ in glycemic control, vascular perfusion, infection burden, immune status, and bone quality, individualized treatment should combine perioperative glycemic and infection control, vascular assessment, stratification by diabetic severity/comorbidities, and adaptive ultrasound regimens. Long-term systemic biosafety, zinc-ion release, scaffold degradation, and large-animal efficacy must be systematically evaluated before clinical application [66]. Conclusion In summary, we developed an ultrasound-activated piezoelectric hydrogel, Gel BC@ZnO, that delivers localized bioelectric cues to remodel the diabetic bone immune microenvironment. The data support a mechanistic sequence in which ultrasound-triggered ZnO piezoelectricity activates Piezo1, induces Piezo1-dependent Ca2+ influx, restores macrophage efferocytosis, attenuates inflammatory oxidative stress, and promotes pro-regenerative immune remodeling. This immune reprogramming cooperates with enhanced osteogenesis and suppressed osteoclastogenesis to restore bone-remodeling homeostasis. In a critical-sized diabetic bone defect model, Gel BC@ZnO + US markedly accelerated bone repair and showed favorable biocompatibility. These findings identify a bioelectric signal-Piezo1-Ca2+-efferocytosis axis as a therapeutic route for overcoming diabetes-associated bone-repair impairment. With injectable, defect-adaptable, and ultrasound-responsive features, Gel BC@ZnO represents a promising bioelectric scaffold for diabetic fracture healing, delayed union/nonunion, and other regeneration-impaired bone defects. Future studies should optimize ultrasound targeting, clinical delivery, crosslinking workflows, scaffold degradation, long-term biosafety, and large-animal validation. Experimental section Preparation of Gel BC@ZnO composite hydrogel Lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP; 0.25 wt%; EFL, China) was dissolved in 20 mL PBS as the photoinitiator solution. GelMA (EFL, China) was added to prepare a 5% (w/v) GelMA solution. BC was prepared as previously reported [68]. Briefly, Komagataeibacter xylinus (BNCC339798) was statically fermented in medium containing 100 g L−1 D-fructose, 5 g L−1 peptone, and 3 g L−1 yeast extract, with the pH adjusted to 5.0 ± 0.2 using 0.1 M sodium citrate buffer. GelMA was mixed with different BC contents (1, 3, and 5 wt%) and stirred until homogeneous. The mixtures were cast into cylindrical molds and UV-crosslinked to obtain Gel BC scaffolds; the BC content was optimized according to mechanical properties and bioactivity. ZnO nanoparticles (Aladdin Biochemical Technology, 97%) were then incorporated at predetermined contents, and the optimal ZnO concentration was selected based on cytotoxicity and bioactivity assays. The optimized Gel BC/ZnO precursor was cast into cylindrical molds and UV-crosslinked to obtain Gel BC@ZnO hydrogel scaffolds. Characterization, mechanical testing, and piezoelectric assessment GelMA, Gel BC, and Gel BC@ZnO hydrogels were freeze-dried to obtain lyophilized scaffolds. Scaffold morphology and elemental distribution were examined by scanning electron microscopy (SEM) and energy-dispersive spectroscopy (EDS). Chemical composition and crystalline structure were analyzed by Fourier-transform infrared spectroscopy (FTIR) and X-ray diffraction (XRD), respectively. Degradation behavior The initial dry mass of each scaffold was recorded as WI. Scaffolds were immersed in PBS and incubated at 37 °C and 100 rpm for 4 weeks. At each weekly time point, samples were washed three times with ddH2O, freeze-dried, and weighed to obtain the residual dry mass (WR).Massremainingratio(%)=WR/WI×100% where WI and WR represent the initial and residual dry scaffold masses, respectively. The pH of the incubation solution was monitored at each time point. Mechanical properties were measured using a uniaxial mechanical testing system (HY-940FS; Heng Yu Instrument Co., Ltd., Shanghai, China). Hydrogels were prepared in molds with a uniform inner diameter and depth of 1.0 cm, and the diameter and height of each sample were recorded before testing. Piezoelectric performance of Gel BC@ZnO containing different ZnO mass ratios (1, 3, and 5 wt%) was measured under ultrasound stimulation (1.0 W cm−2), and voltage outputs were analyzed using Origin software (OriginLab, USA). Cell culture Mouse RAW264.7 macrophages were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Jurkat cells were purchased from Servicebio (Wuhan, China), and MC3T3-E1 Subclone 14 cells (MC-14) were obtained from Pricella (Wuhan, China). Bone marrow-derived mesenchymal stem cells (BMSCs) were isolated from the tibiae and femurs of 4–6-week-old Sprague-Dawley (SD) rats (80–100 g) supplied by the Animal Center of Shanghai Tenth People's Hospital (Shanghai, China). Bone marrow-derived macrophages (BMMs) were isolated from the tibiae and femurs of 4-6-week-old C57BL/6 mice (16-18 g) supplied by the Animal Center of Shanghai Tenth People's Hospital (Shanghai, China). Cells were cultured in α-minimum essential medium (α-MEM; Gibco, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, USA) and 1% penicillin-streptomycin (P/S; Gibco, USA) at 37 °C in a humidified 5% CO2 atmosphere. Except for the in vitro biocompatibility assay, all subsequent experiments were performed under high-glucose conditions (35 mM) to mimic the diabetic microenvironment in vitro. In vitro biocompatibility assessment RAW 264.7 macrophages and rBMSCs were separately cultured in complete medium alone (Control), Gel BC, Gel BC@ZnO, or Gel BC@ZnO + US. Cell viability was assessed on days 1, 3, and 5 using CCK-8 (Dojindo, Japan). Culture medium containing 10% (v/v) CCK-8 reagent was added and incubated for 2 h at 37 °C. Then, 100 μL supernatant from each well was transferred to a 96-well plate, and absorbance at 450 nm was measured. Cytotoxicity was further assessed by live/dead staining using calcein-AM for viable cells and propidium iodide (PI; Beyotime Biotechnology, China) for dead cells. Apoptosis was analyzed with an apoptosis detection kit (Servicebio, Wuhan, China). Culture supernatants were collected, adherent cells were detached with EDTA-free trypsin, and combined suspensions were centrifuged at 500 × g for 5 min at 4 °C. Cell pellets were washed twice with pre-chilled PBS under the same conditions. Cells were resuspended in pre-chilled 1× binding buffer at 1-5 × 106 cells mL−1. Aliquots (100 μL) were stained with 5 μL Annexin V-PE and 5 μL 7-AAD for 8-10 min in the dark at room temperature. Pre-chilled 1× binding buffer (400 μL) was then added, and samples were analyzed by flow cytometry or fluorescence microscopy within 1 h. Quantitative real-time PCR (qRT-PCR) Total RNA was extracted using TRIzol according to the manufacturer's instructions. RNA concentration and purity were determined by absorbance at 260/280 nm. Equal amounts of RNA were reverse-transcribed into cDNA, and qRT-PCR was performed using SYBR Green PCR Master Mix on a real-time PCR system. Relative mRNA expression was normalized to GAPDH and calculated by the 2^−ΔΔCt method. Primer sequences are listed in Table S1. Data were obtained from at least three independent biological replicates. Piezo1 immunofluorescence staining Cells were fixed with 4% paraformaldehyde for 30 min at room temperature and washed three times with PBS. Cells were permeabilized with 0.5% Triton X-100, blocked with 5% BSA, and incubated overnight at 4 °C with anti-Piezo1 primary antibody (Abclonal; 1:100) in immunostaining dilution buffer (Beyotime Biotechnology, China). After washing, cells were incubated with Alexa Fluor 488-conjugated goat anti-rabbit secondary antibody (Abcam; 1:200) for 1 h at room temperature. F-actin was counterstained with FITC- or TRITC-labeled phalloidin (Yeasen, Shanghai, China), and nuclei were stained with DAPI (Beyotime Biotechnology, China). Images were acquired using a fluorescence microscope (DMI8, Leica, Germany). Efferocytosis-related experiments Jurkat cells were treated with 10 μM camptothecin for 8-10 h to induce apoptosis. Before CFSE labeling, an aliquot of the induced cell suspension was examined by Annexin V-PE/7-AAD staining and flow cytometry to confirm apoptotic enrichment and exclude preparations with excessive necrotic cells, secondary necrosis, or debris. Only intact suspensions with predominantly Annexin V-positive apoptotic profiles and minimal nonspecific damage were used as apoptotic targets. Cells were centrifuged, resuspended in 10 μM carboxyfluorescein succinimidyl ester (CFSE; MCE, USA) staining solution for 20 min in the dark, washed to remove excess CFSE, and resuspended in macrophage culture medium. RAW264.7 macrophages were pre-seeded and allowed to adhere. CFSE-labeled apoptotic Jurkat cells (2 mL suspension) were gently added to each well and co-cultured with macrophages for 6 h at 37 °C in the dark. Wells were then washed three times with PBS to remove free apoptotic cells, apoptotic debris, and loosely attached CFSE-positive material before macrophage collection. During flow cytometry, debris and small apoptotic fragments were excluded by forward/side scatter, doublets were excluded by singlet gating, and macrophages were identified as F4/80-positive events. Residual free CFSE-positive Jurkat cells or debris were excluded because they were F4/80-negative and/or outside the macrophage FSC/SSC gate. Efferocytic macrophages were defined as F4/80+CFSE + cells, and efferocytosis was calculated as F4/80+CFSE + events divided by total F4/80+ macrophage events × 100%. For immunofluorescence analysis, macrophages were stained with anti-F4/80 primary antibody (Proteintech, USA) followed by Alexa Fluor 594-conjugated goat anti-rabbit secondary antibody (Abcam; 1:200). CFSE-labeled apoptotic cells were visualized simultaneously. After co-culture and washing, intracellular CFSE signals within F4/80-positive macrophages were imaged by fluorescence microscopy to corroborate flow-cytometry quantification and reduce misinterpretation of free or non-internalized CFSE-positive material as true efferocytosis. Intracellular Ca2+ influx, macrophage polarization, and ROS detection To evaluate Piezo1 functional activation and downstream immunomodulation, RAW264.7 macrophages and MC-14 osteoblasts were cultured under high-glucose conditions and assigned to Control, Gel BC, Gel BC@ZnO, Gel BC@ZnO + US, or Gel BC@ZnO + US + GsMTx4 groups. For Piezo1 inhibition, cells were pretreated with GsMTx4 (MedChemExpress, USA) before ultrasound stimulation. After treatment, cells were incubated with a FITC-labeled Ca2+ detection probe according to the manufacturer's protocol, washed with PBS, and analyzed by flow cytometry to quantify FITC-Ca2+-positive cells. Intracellular Ca2+ fluorescence was also imaged by fluorescence microscopy, and relative mean fluorescence intensity (MFI) was quantified using ImageJ under identical exposure and threshold settings. For macrophage polarization, RAW264.7 cells were stained with PE-conjugated anti-CD206 and APC-conjugated anti-CD86 antibodies and analyzed by flow cytometry. For ROS analysis, macrophages were stained with a FITC-ROS probe, and ROS-positive cells were quantified by flow cytometry. Debris and doublets were excluded using standard FSC/SSC and singlet gating, the same gating strategy was applied to all groups, and all experiments included three independent biological replicates. In vitro osteogenesis and osteoclast genesis MC-14 cells were seeded in the lower chamber of 12-well Transwell inserts (0.4 μm pore size; Corning, NY, USA), and corresponding scaffolds were placed in the upper chamber. Osteogenic medium consisted of α-MEM supplemented with 10% FBS, 100 U mL−1 penicillin-streptomycin, 50 mg L−1 ascorbic acid, 10 mM β-glycerophosphate, and 10−8 M dexamethasone (Gibco, Carlsbad, CA, USA). Inserts without scaffolds served as controls, and medium was refreshed every 3 days. ALP and ARS staining were performed on days 7 and 14, respectively. For mineralization quantification, cells were incubated with 10% (w/v) hexadecylpyridinium chloride solution (Macklin) at room temperature on a shaker for 30 min, and absorbance was measured at 570 nm. For osteoclast genesis assays, BMMs were seeded in the lower chamber of Transwell inserts, with scaffolds placed in the upper chamber under high-glucose (35 mM) conditions. Osteoclastogenic medium consisted of high-glucose DMEM supplemented with 30 ng mL−1 M-CSF and 100 ng mL−1 RANKL (Novoprotein, China) and was replaced every other day. After 7 days of induction, TRAP staining (Sigma) and cytoskeletal staining were performed. In vivo femoral condyle bone defect model Male Sprague-Dawley rats received a single intraperitoneal injection of streptozotocin (STZ; 35 mg kg−1) in freshly prepared citrate buffer after overnight fasting. Random blood glucose was monitored for three consecutive days, and rats with sustained diabetic-range glucose levels were used for surgery. Twelve diabetic rats were randomly assigned to four groups: Control, Gel BC, Gel BC@ZnO, and Gel BC@ZnO + US. All animal procedures complied with the Guidelines for the Care and Use of Laboratory Animals and were approved by the Animal Care and Use Committee of Shanghai Tenth People's Hospital (Approval No. SHDSYY-2025-2530-N1). After general anesthesia, the medial femoral condyle region was shaved, sterilized, and surgically prepared. A longitudinal incision was made, and the skin, muscle, and surrounding soft tissues were dissected to expose the medial femoral condyle. A critical-sized cylindrical defect (3 mm in diameter) was created using a 3 mm trephine drill. Control animals received no material implantation. Rats in the Gel BC, Gel BC@ZnO, and Gel BC@ZnO + US groups received in situ injection of the corresponding precursor solutions into the defect, followed by photocrosslinking for 30 s to form hydrogels. Postoperatively, the Gel BC@ZnO + US group received low-frequency ultrasound stimulation (1.0 W cm−2). At 8 weeks, rats were euthanized, and femoral condyle specimens were harvested for Micro-CT. Samples were then processed for H&E, Masson's trichrome, Piezo1 immunofluorescence, iNOS/Arg1 immunofluorescence, and COL I/OCN immunohistochemical analyses. Subcutaneous implantation Gel BC@ZnO scaffolds were implanted into dorsal subcutaneous pockets in Sprague-Dawley rats to evaluate in vivo degradation and tissue compatibility (Fig. S13). At predetermined time points, scaffolds with surrounding tissues were harvested for histological and degradation analyses. Statistical analysis Data are presented as mean ± standard deviation (SD), and figure legends specify that error bars represent SD. Two-group comparisons were performed using an independent-samples Student's t-test. Comparisons among more than two groups were performed using one-way ANOVA followed by Tukey's multiple-comparison test. Time-course cell-viability assays involving treatment and culture time were analyzed by two-way ANOVA followed by Tukey's test. All tests were two-sided, and P < 0.05 was considered statistically significant. Significance levels are denoted as ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001. CRediT authorship contribution statement Dilixiati Duolikun: Writing – review & editing, Writing – original draft, Data curation. Lei Zhang: Writing – review & editing, Data curation. Tianlong Wang: Writing – review & editing, Software, Data curation. Xinhui Wu: Investigation, Formal analysis. Xuran Guo: Investigation, Data curation. Waheed Muhammad Zeeshan: Formal analysis, Data curation. Zhiqing Liu: Writing – review & editing, Visualization, Methodology. Haijun Zhang: Visualization, Resources, Conceptualization. Jinhui Zhao: Writing – review & editing, Writing – original draft, Software, Methodology, Conceptualization. Longpo Zheng: Supervision, Project administration, Funding acquisition. Ethics approval and consent to participate All surgical procedures strictly adhered to the guidelines of the National Institutes of Health (NIH) for the care and use of laboratory animals. The study was approved by the Animal Ethics Committee of Shanghai Tenth People's Hospital affiliated with Tongji University School of Medicine (approval number: SHDSYY-2025-2530-N1). Declaration of competing interest The authors declare no conflict of interest.