Ternary mTOR-targeted conductive nanofibrous scaffolds with bioactive peptides orchestrate immune-metabolic-fibrotic balance for diabetic bone regeneration

Abstract

Diabetic bone defects (DBD) are difficult to heal because persistent inflammation, metabolic dysfunction and fibrotic repair disrupt the transition from injury to regeneration. Here, we report PP@IL^ES, a peptide functionalized electroconductive nanofibrous scaffold that combines local electrical cues with immobilized and gradually released bioactive peptide signals to remodel the diabetic defect niche. Single cell RNA sequencing and mechanistic analyses showed that PP@IL^ES reshaped macrophage and fibroblast states at the early repair stage. In macrophages, PP@IL^ES was associated with reduced mTORC1 activity, enhanced mitochondrial oxidative metabolism and fatty acid oxidation, increased CD206 expression, and decreased iNOS, TNF alpha and glycolytic output. In fibroblasts, PP@IL^ES attenuated activation through KLF4 associated downregulation of alpha SMA, limiting scar like fibrotic remodeling. In diabetic rats with 5 mm cranial defects, PP@IL^ES promoted bone regeneration and mature lamellar bone formation with reduced fibrotic tissue compared with conductive scaffold alone or peptide scaffold alone. These findings suggest that coordinated regulation of immune metabolism, fibrotic remodeling and osteogenesis through an electrobiochemical scaffold may provide a promising strategy for diabetic bone repair.

Graphical abstract

Highlights

• •A conductive peptide-functionalized scaffold was developed for diabetic bone defect repair.
• •PP@IL^ES enabled ternary regulation of mTORC1, macrophages and fibroblasts.
• •PP@IL^ES drove macrophage metabolism from glycolysis to OXPHOS/FAO.
• •Fibroblast activation and α-SMA-mediated fibrosis were markedly reduced.
• •PP@IL^ES promoted mature bone regeneration with minimal scar formation in diabetic defects.

1. Introduction

Diabetic bone defect (DBD) refers to disruption of bone continuity or bone loss in diabetic individuals caused by severe trauma, surgery, infection, or other factors, in which the diabetic pathological microenvironment leads to persistent impairment of bone healing and long-term existence of the bone defect gap [[1], [2], [3]]. Although bone tissue possesses substantial regenerative capabilities in healthy individuals, approximately 5–10% of fractures still experience delayed healing or non-union, especially when combined with bone defects [4]. This clinical challenge is significantly exacerbated in patients with diabetes. Diabetes is a common metabolic disorder worldwide, and its prevalence continues to rise [5]. And people with diabetes often exhibit fragile bones, increased fracture risk, and impaired bone repair [6,7]. Diabetes mellitus increases fracture healing time by 87% and the risk of complications such as bone defects [8], resulting in a significant socio-economic and medical burden [3,9].

Georg N. Duda points out that current regenerative therapeutic strategies have had some success, but clinical trials have often failed in the face of a complex metabolic and immune environment. To successfully develop regenerative fracture therapies, researchers must move beyond a “one-size-fits-all” strategy and gain a deeper understanding of the mechanisms that influence healing. Focusing on the early healing phase and the dynamic immune microenvironment in an effort to promote scarless healing of fractures [10]. For refractory bone defects, a multifaceted approach is necessary [11]. Bone fracture healing involves distinct yet dynamically interwoven phases [12], including inflammation, repair, callus formation, and remodeling, coordinated by multiple cell types [13,14]. Early inflammatory responses rapidly recruit and activate immune cells to clear necrotic debris, releasing cytokines and chemokines that attract mesenchymal stem cells (MSCs) and osteoblasts to initiate regeneration [15]. However, persistently activated inflammation inhibits bone repair [16]. Effective healing critically depends on a timely transition from pro-inflammatory M1 macrophages to anti-inflammatory M2 macrophages [17]. Macrophage polarization is closely regulated by cellular metabolism, and in diabetes, sustained hyperglycemia and excessive reactive oxygen species (ROS) accumulation cause macrophages to rely on a glycolytic mode for rapid ATP production, resulting in a sustained polarization of macrophages toward the pro-inflammatory M1 type [18].

This prolonged M1 dominance exacerbates the release of inflammatory mediators, intensifying local inflammation [19,20]. This inflammatory state further induces excessive fibrosis, leading to impaired local healing [21]. Previous studies suggest that macrophages and fibroblasts have close cellular communication [[22], [23], [24]], and macrophage dysregulation triggers a fibroinflammatory response [25]. Inflammatory macrophages signal fibroblasts to transform into myofibroblasts and deposit collagen in an abnormal manner [24], leading to imperfect wound healing [26]. In response to tissue injury, transient activation of myofibroblasts contributes to tissue repair, but sustained activation triggers pathological fibrosis [27,28]. Under normal conditions, modest myofibroblasts can generate contractile tension via α-SMA driven cytoplasmic stress fibers, which can fine-tune fracture-end alignment, compress the fibro-cartilaginous scab, and promote mineralization, with the majority of myofibroblasts cleared by apoptosis in the pre-mineralization phase [[29], [30], [31]]. However, myofibroblasts can be persistently activated in the inflammatory environment of diabetes [32,33]. Consequently, fibrotic scarring ensues, hindering angiogenesis and osteogenesis, and ultimately leading to delayed or impaired bone healing.

Previous studies have mostly emphasized the enhancement of osteogenic differentiation and mineralization [34], as well as the promotion of macrophage M2-type polarization at the functional level [35]. Less attention has been paid to the way macrophages are metabolized in the diabetic setting, and the impact of immunofibrosis, leading to little clinical efficacy [36,37]. How to efficiently repair diabetic bone defects remains a clinical challenge. To address these multifaceted challenges, current strategies emphasize conductive biomaterials and electrical stimulation (ES), capable of safely and flexibly modulating cellular responses [38]. Recent evidence indicates that ES effectively promotes osteogenesis, angiogenesis and macrophage polarization toward the M2 phenotype [39,40]. In addition, electrical stimulation can restore physiological electrical adaptation in tissues and regulate mitochondrial activity [41]. However, its immune-regulatory mechanisms and metabolic regulation remain poorly understood.

Current single target therapies, such as mTOR inhibitors alone or conductive scaffolds alone, fail to address the interconnected immune, metabolic, and fibrotic disorders in DBD, and their clinical translation is limited by insufficient tissue integration. Inspired by Trombley's tri regulatory approach in nerve regeneration [42], we proposed a ternary regulatory strategy centered on mechanistic target of rapamycin complex 1 (mTORC1)-mediated macrophage-fibroblast regulation. In this study, this ternary regulatory strategy refers to the coordinated modulation of mTORC1 related signaling, macrophage immune metabolic remodeling, and fibroblast associated fibrotic remodeling, in which mTORC1 serves as a signaling node linking inflammatory macrophage activation and fibroblast mediated scar like repair in the diabetic defect microenvironment. Specifically, we engineered a multifunctional scaffold by homogeneously blending polyvinyl butyral (PVB) with the conductive poly(3,4-ethylenedioxythiophene): polystyrene sulfonate (PEDOT:PSS) [43] and subsequently fabricating the mixture into conductive nanofibrous scaffolds, abbreviated as PP, via electrospinning. A bioactive peptide that simultaneously promotes osteogenesis and mitigates fibrotic scarring [44,45], (DOPA)₂ -IEELEEELEER- (PEG4)- LYENRL- (DOPA)₂, was synthesized by solid phase peptide synthesis followed by an amidation reaction and is denoted IL. Leveraging the mussel inspired catechol adhesion of its terminal DOPA residues, IL was immobilized onto the PP surface, yielding a biofunctional conductive nanofibrous scaffold referred to as PP@IL. We hypothesized that the combination of exogenous electrical stimulation and endogenous bioactive peptide cues would remodel the pathological microenvironment of diabetic bone defects by coordinating macrophage immune metabolism, fibroblast activation, and osteogenic repair, thereby promoting improved regenerative outcomes (Fig. 1).

Fig. 1.

Schematic diagram of PP@IL^ES promoting diabetic bone regeneration. a) Schematic diagram of PP@IL material composition. b) Diagram of the mechanism by which PP@IL^ES promotes diabetic bone defect repair by regulating mTORC1, macrophage metabolism, and fibroblast activity. c) Schematic diagram of the action of PP@IL^ES to promote diabetic bone defect repair in vivo by improving the immune microenvironment and inhibiting scarring. TNF-α, tumor necrosis factor-alpha; iNOS, inducible nitric oxide synthase; IL-1β, interleukin-1 beta; TGF-β, transforming growth factor-beta; BMP-2, bone morphogenetic protein-2; ARG-1, arginase-1; IL-10, interleukin-10; M1 macrophage, pro-inflammatory (classically activated) macrophage; M2 macrophage, anti-inflammatory (alternatively activated) macrophage; BMSC, bone marrow mesenchymal stem cell; ROS, reactive oxygen species.

2. Results

2.1. Characterization of conductive nanofibrous scaffolds

The preparation process of conductive nanofibrous scaffolds is shown in Fig. 2a. Material groupings and abbreviations are listed in Supplementary Tables 1–2. Conductive nanofibrous scaffolds (PP@IL) were successfully fabricated via electrospinning. Scanning electron microscopy (SEM) revealed smooth fiber surfaces for the PVB scaffold (P group), with an average diameter of approximately 2.2 μm. After incorporating PEDOT: PSS (PP group), fibers exhibited rougher surfaces with an increased diameter averaging approximately 2.7 μm. Energy-dispersive X-ray spectroscopy (EDS) analysis detected a sulfur signal in the PP fibers (absent in the P group), consistent with the presence of PEDOT:PSS. These data support successful incorporation of PEDOT:PSS into the nanofibrous scaffold. In the PP@IL group, catechol-mediated immobilisation of the IL peptide on the PP conductive nanofibers produced a smoother surface and increased the average fiber diameter to approximately 3.1 μm (Fig. 2b–c, Fig. S1, Supporting Information). Fiber orientation analysis indicated that fibers were well-aligned (Fig. S2a–c, Supporting Information). This alignment is expected to facilitate microcurrent transmission. Surface wettability analysis demonstrated that the P scaffold exhibited a water contact angle around 111.9°. Incorporating PEDOT: PSS increased surface hydrophobicity, raising the contact angle to approximately 114.2°. Although PEDOT:PSS exhibits some hydrophilicity, the water contact angle may have increased due to its low concentration and the resulting increase in the roughness of the fiber surface. However, in the PP@IL group, the contact angle decreased to approximately 90°. This may be due to the increased hydrophilicity conferred by the peptide, which may favor cell adhesion and migration (Fig. 2d). Atomic force microscopy (AFM) showed that the PP group was rougher, while the surface became smooth after IL peptide coating. AFM analysis also revealed variations in surface roughness. Specifically, surface roughness values for P, PP, and PP@IL scaffolds were 66.12 ± 11.25 nm, 118.1 ± 22.31 nm, and 97.11 ± 15.32 nm, respectively (Fig. 2e, Fig. S2d, Supporting Information). This may be attributed to the incorporation of PEDOT, which increased the surface roughness, and the uniform coating of polypeptides which reduces the surface roughness. This is consistent with the images observed by SEM.

Fig. 2.

Physical and chemical characterization of conductive nanofibrous scaffolds. a) Schematic diagram of conductive nanofibrous scaffolds preparation. b) Surface morphology SEM photographs and EDS elemental mapping of P、PP and PP@IL nanofibrous scaffolds. c) Diameter distributions for the P、PP and PP@IL nanofibrous scaffolds. d)Water contact angles of nanofibrous scaffolds. e) Atomic Force Microscope(AFM)spectra acquired for the P、PP and PP@IL nanofibrous scaffolds. f) Fourier transform infrared (FTIR) spectra acquired for the P、PP and PP@IL nanofibrous scaffolds. g) X-ray photoelectron spectroscopy (XPS) spectra acquired for the P、PP and PP@IL nanofibrous scaffolds. (n = 3 per group, one-way ANOVA with Dunnett's post hoc test, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns: not significant (p > 0.05)).

Fourier-transform infrared spectroscopy (FTIR) analysis further confirmed the chemical composition. In the P group, characteristic absorption peaks at 1102.63 cm⁻¹ and 1129.86 cm⁻¹ corresponded to ester-group (C–O) stretching vibration. The PP group exhibited an additional peak at 962.91 cm⁻¹, which may be attributed to PEDOT:PSS. The PP@IL group showed peaks at 1652.47 cm⁻¹ (amide I band). The intensity of the peak at 3413.79 cm⁻¹was enhanced compared with the P and PP groups, which should be the broad O–H/N–H Overlap peak (Fig. 2f). This indicates that IL peptides were successfully coated on the conductive nanofibrous scaffolds. X-ray photoelectron spectroscopy (XPS) provided chemical characterization of scaffolds. The C 1s spectrum of the P group showed peaks at 284.8 eV (C–C), 285.7 eV (C–O), and 287.0 eV (O=C–O). The PP group displayed additional characteristic peaks for S 2p at 168.2 eV and 169.6 eV (-SO₃^-), confirming PEDOT: PSS incorporation. The PP@IL group exhibited peaks at 284.2 eV (C–S), 286.0 eV (C–N), as well as clear N 1s peaks at 399.4 eV (C–N bonds), confirming successful coating with multifunctional peptides (Fig. 2g, Fig. S3, Supporting Information). The stress strain curves showed that the P scaffold exhibited the highest tensile stress among the tested groups (Figure S4, Supporting Information). After incorporation of PEDOT, the PP scaffold showed a little decrease in tensile performance, suggesting that the conductive component may partially affect the continuity or compactness of the PVB nanofibrous network. Notably, after IL peptide coating, PP@IL showed improved tensile behavior compared with PP. This improvement may be associated with DOPA mediated interfacial adhesion and enhanced interactions among nanofibers after peptide functionalization. The degradation-associated pH and Zeta potential of PP@IL were monitored to evaluate its physicochemical stability (Figure S5a-b, Supporting Information). During degradation, the pH showed only mild fluctuations and remained within a near-physiological range. The Zeta potential exhibited a transient negative shift during peptide release and then gradually returned toward neutrality, possibly related to reduced peptide release and DOPA-associated interfacial stabilization. These results suggest that PP@IL maintained relatively stable physicochemical behavior during degradation. The apparent surface binding efficiency of IL peptide on PP scaffolds was quantified by an indirect depletion method. The calculated amount of a binding efficiency of 69.1%. These results indicate effective DOPA-mediated peptide retention on the conductive nanofibrous scaffold. IL peptide release from PP@IL scaffolds was further evaluated at 37 °C. In PBS, the peptide showed a gradual release profile and approached a plateau at approximately 14 days. Electrical stimulation accelerated the early release, with approximately 76% cumulative release by day 7, followed by a slower release phase (Fig. S6, Supporting Information). The degradation behavior of PP@IL scaffolds was further evaluated under simulated in vivo-like conditions. Both PP@IL and PP@IL^ES exhibited gradual mass loss without rapid degradation (Fig. S7, Supporting Information). PP@IL^ES showed a slightly faster degradation trend than PP@IL, possibly due to electrical stimulation-enhanced hydration, ion migration, and interfacial remodeling. These results suggest that PP@IL scaffolds possess mild and controllable degradation behavior suitable for local implantation.

2.2. Electrochemical properties of scaffold

A key design feature of the PP@IL conductive nanofibrous scaffolds is their capacity to facilitate microcurrent transmission. The current transfer of the conductive nanofiber scaffold is shown in Fig. 3a. We conducted detailed assessments of electrical conductivity and electrochemical properties. As depicted in Fig. 3b, pure PVB electrospun fibers (P group) exhibited negligible electrical conductivity (0 mS/cm). When PEDOT:PSS was incorporated at concentrations of 0.1% (PP¹), the conductivity increased to approximately 2.04 × 10⁻⁵ mS/cm. Significantly greater increases were observed at 0.2% (PP²) and 0.3% (PP³), reaching approximately 1.23 × 10⁻³ mS/cm and 11.99 × 10⁻² mS/cm, respectively. This conductivity value suggests improved microcurrent transmission capability. When multifunctional peptides (IL) were coated onto PP³ scaffolds, the conductivity reached approximately 6.01 × 10⁻² mS/cm. Wet-state conductivity was measured to evaluate electrical stability under hydrated conditions (Fig. S8, Supporting Information). After PBS wetting, PP³ and PP³@IL showed conductivities of 0.176 ± 0.012 and 0.118 ± 0.012 mS cm⁻¹, respectively. After 14 days of PBS immersion, the values decreased to 0.144 ± 0.010 and 0.087 ± 0.004 mS cm⁻¹, corresponding to retention rates of 81.5% and 73.7%. These results suggest that although IL peptide coating partially reduced conductivity, PP³@IL maintained stable wet-state conductivity after prolonged PBS exposure. Attempts to use 0.4% PEDOT:PSS concentration were unsuccessful, as stable conductive nanofibrous scaffolds could not be formed.

Fig. 3.

Electrical characterization of conductive nanofibrous scaffolds. a) Schematic illustration of conductive nanofibrous scaffolds for local electrical cue transmission. b) Conductivity of PP¹, PP², PP³ and PP³@IL nanofibrous scaffolds. c) Cyclic voltammetry curves of PP², PP³ and PP³@IL. d) Overpotentials required to reach a current density of 10 mA cm⁻² for PP¹, PP², PP³ and PP³@IL. e) Tafel plots of PP¹, PP², PP³ and PP³@IL. f) Nyquist plots of PP¹, PP², PP³ and PP³@IL. (n = 3 per group, one-way ANOVA with Dunnett's post hoc test, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns: not significant (p > 0.05)).

CV curves further showed enlarged capacitive areas with increasing PEDOT content, while PP³@IL retained a clear capacitive response, although its current density was lower than that of PP³ (Fig. 3c). PP³ and PP³@IL required lower potentials to reach 10 mA cm⁻² than PP¹ and PP² (Fig. 3d), and showed lower Tafel slopes (Fig. 3e), suggesting more favorable apparent interfacial charge transfer characteristics. EIS further revealed reduced impedance for PP³ and PP³@IL, as reflected by smaller semicircle diameters in the Nyquist plots (Fig. 3f). Electrochemical double-layer capacitance analysis showed that PP³ had the highest Cdl value (36.68 mF cm⁻²), followed by PP³@IL (28.34 mF cm⁻²), both of which were higher than PP¹and PP² (Fig. S9, Supporting Information), indicating improved interfacial charge storage after PEDOT incorporation. After 1000 LSV cycles, the scaffolds maintained stable electrochemical polarization behavior (Fig. S10, Supporting Information). Collectively, these results indicate that PP³ and PP³@IL possess favorable conductivity, interfacial charge transfer capability, low impedance, and electrochemical stability, supporting their use for local electrical cue transmission in subsequent experiments.

2.3. Characterization and function of bioactive peptides

The bioactive peptide (DOPA)₂-IEELEEELEER-(PEG4)-LYENRL-(DOPA)₂ was synthesized by solid-phase peptide synthesis followed by amidation. The peptide purity was 96.85% and the molecular weight was 3170.35 Da, confirmed by HPLC and MS (Fig. S11–12, Supporting Information). The structural formula and schematic illustration of the IL peptide are shown in Fig. S13. Initial molecular docking suggested favorable binding poses between the peptide and selected targets (Fig. S14, Supporting Information). Network pharmacology analysis identified a PPI network containing 37 nodes and 192 edges, from which ten hub genes were screened using the maximal clique centrality algorithm; the top five targets were STAT3, IL1B, NFKB1, SRC, and TLR4 (Fig. S15, Supporting Information). GO and KEGG enrichment analyses further suggested that these targets may be associated with peptide related cellular responses, inflammatory regulation, and tissue repair pathways (Fig. S16, Supporting Information).

To further explore potential interactions between the peptide and the five hub proteins, semi flexible molecular docking was performed. As shown in Supplementary Table 3, all targets except IL1B showed binding energies below −9 kcal/mol, and the corresponding binding modes are presented in Fig. S17, Supporting Information. Functionally, STAT3 and SRC are closely related to osteogenic differentiation, bone homeostasis, and extracellular matrix remodeling [46,47], whereas NFKB1 and TLR4 are involved in inflammatory signaling that may contribute to impaired healing and scar like fibrotic remodeling [48,49]. These in silico results suggest that the IL peptide may exert pro-osteogenic and anti-fibrotic bioactivity through multiple inflammation and repair related targets. However, these analyses were used as hypothesis supporting evidence during peptide design and do not by themselves prove direct target engagement in the biological context.

2.4. Regulation of the macrophage polarization phenotype

Persistent M1 macrophage activation exacerbates inflammatory responses within the diabetic bone defect microenvironment. Additionally, foreign biomaterial implantation could potentially trigger localized inflammation [50,51]. Therefore, the capability of conductive nanofibrous scaffolds to regulate macrophage polarization is critical. Initially, live/dead assays were conducted to assess macrophage viability. In this study, BMDMs from 6-week-old C57BL/6 mice were treated with different scaffold groups under high glucose and LPS stimulation, with 40 mV electrical stimulation applied for 15 min where indicated. As depicted in Fig. 4a, macrophages in all groups displayed high viability (>95%), indicated by dominant green fluorescence (live cells), with minimal red fluorescence (dead cells), demonstrating that neither scaffold materials nor electrical stimulation exhibited significant cytotoxicity. Quantitative analysis showed greater cell viability with electrical stimulation interventions (Fig. S18, Supporting Information). To further investigate the impact of scaffolds on macrophage polarization, immunofluorescent staining was performed for M1 marker CD86 and M2 marker CD206 (Fig. 4b–c). In the P, PP, and PP@IL groups, stronger fluorescence signals for CD86 (red) and weaker signals for CD206 (green) were observed, indicating a tendency towards M1 polarization. Quantitative fluorescence analysis showed no statistically significant differences among these three groups, suggesting that the IL peptide alone does not influence macrophage polarization direction. In contrast, the electrically stimulated groups (PP^ES and PP@IL^ES) demonstrated significant M2 polarization features, with notably reduced CD86-positive cells and significantly enhanced CD206 fluorescence intensity (Fig. S19, Supporting Information).

Fig. 4.

Regulation of macrophage polarization by different scaffold treatments. BMDMs from 6-week-old C57BL/6 mice were treated under high-glucose/LPS-induced inflammatory conditions, with electrical stimulation applied at 40 mV for 15 min where indicated. a) Living cells (green) and dead cells (red) of macrophages under different treatment conditions. b)Fluorescence microscope images of immunofluorescent staining of CD86: red (CD86) and blue (DAPI). Scale bars: 100 μm. c) Fluorescence microscope images of immunofluorescent staining of CD206: green (CD206) and blue (DAPI). Scale bars: 100 μm. d) Flow cytometry analysis of CD86 and CD206 expression with macrophages under different treatment conditions. e-h)iNOS, TNF-α, IL-10 and TGF-β gene expression of macrophages evaluated by qRT-PCR. i-j) Percentages of F4/80+CD86⁺ and F4/80+CD206+ BMDM cells cultured on the different material groups, as determined by flow cytometry. (n = 5 per group, one-way ANOVA with Dunnett's post hoc test, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns: not significant (p > 0.05)).

Flow cytometry results (Fig. 4d) were consistent with immunofluorescence findings, confirming the highest proportion of CD86-positive cells and lowest CD206-positive cells in the P, PP, and PP@IL groups. Conversely, CD86-positive cells were markedly decreased, and CD206-positive cells significantly increased in the PP^ES and PP@IL^ES groups. These findings highlight the pivotal role of conductive nanofibrous scaffolds under microcurrent stimulation in promoting M2 polarization. To elucidate the polarization mechanism further, qRT-PCR analyses were conducted for M1-related genes (iNOS, TNF-α) and M2-related genes (IL-10, TGF-β). Expression levels of iNOS and TNF-α were highest in the P, PP, and PP@IL groups, whereas IL-10 and TGF-β levels were relatively low. In contrast, the PP^ES and PP@IL^ES groups exhibited significantly downregulated iNOS and TNF-α expression and markedly upregulated IL-10 and TGF-β (Fig. 4e–h). Quantitative flow cytometry analysis of CD86 showed that there were no statistically significant differences between PP^ES and PP@IL^ES groups. Quantitative analysis of CD206 also showed no significant difference between the PP^ES and PP@IL^ES groups. (Fig. 4i–j). The decrease in CD86, iNOS and TNF-α together with the increase in CD206, IL-10 and TGF-β reflects not only a phenotypic switch towards an M2-like state, but also a profound reshaping of the cytokine milieu from a pro-inflammatory, tissue-damaging environment to a pro-resolving and pro-regenerative one. This anti-inflammatory cytokine profile is expected to reduce osteoblast inhibition and oxidative stress, support BMSC survival and osteogenic differentiation [52].

2.5. Biocompatibility, oxidative stress and mitochondrial function in BMSCs

Initially, live/dead staining was employed to evaluate the biocompatibility of conductive nanofibrous scaffolds. Bone marrow mesenchymal stem cells (BMSCs) co-cultured with scaffolds showed excellent viability across all groups, with over 95% live cells (green fluorescence) and very few dead cells (red fluorescence) (Fig. S20a–b, Supporting Information). To further examine the proliferation of BMSCs, cells were co-cultured with the scaffolds for 5 days, and cell proliferation was assessed on days 1, 3, and 5 using the Cell Counting Kit-8 (CCK-8) assay. Results indicated that all groups maintained good proliferation rates at each time point. Notably, the PP@IL, PP^ES, and PP@IL^ES groups demonstrated significantly higher proliferation rates compared to the other groups, with the PP@IL^ES group showing the highest proliferation (Fig. S20c, Supporting Information). These findings suggest that both IL peptides and electrical stimulation (ES) can enhance the proliferation of BMSCs. Collectively, the live/dead staining and CCK-8 assays confirm that PP@IL^ES scaffolds possess excellent biocompatibility suitable for subsequent biological applications.

SOD2 immunofluorescence revealed weak signals in the P and PP groups, a modest increase in the PP@IL group, and significantly enhanced SOD2 expression in the PP^ES and PP@IL^ES groups, indicating an improved antioxidant response and partially restored mitochondrial function in these two groups (Fig. S21, Supporting Information)To directly assess mitochondrial bioenergetics, we measured ATP content and mitochondrial membrane potential (ΔΨm) in BMSCs. In the P and PP groups, ATP levels and the JC-1 red/green fluorescence ratio were markedly reduced, indicating impaired mitochondrial function under diabetic conditions. PP@IL partially restored ATP production and ΔΨm, whereas conditioned media from the PP^ES and PP@IL^ES groups significantly increased both ATP levels and JC-1 red/green ratios, with PP@IL^ES yielding the highest values among all groups (Fig. S22, Supporting Information).The alleviation of oxidative stress may be attributed to the shift of macrophages from pro-inflammatory to anti-inflammatory states, the secretion of anti-inflammatory factors, and metabolic regulation, thereby breaking the vicious cycle of high Oxidative stress levels and M1 macrophage polarization. Additionally, ES can also enhance mitochondrial function [53,54]. Under diabetic conditions, AGEs accumulation and persistent inflammatory signaling will disrupts mitochondrial homeostasis, inhibits cell proliferation, and impairs osteogenic differentiation capacity. Hyperglycemia and AGE-induced oxidative stress activate the NF-κB/MAPK pathway, promoting DNA damage and cellular senescence, thereby diverting BMSC differentiation away from the osteogenic pathway. This results in reduced ALP activity, decreased matrix mineralization, and increased apoptosis. Concurrently, mitochondrial dysfunction is recognized as a core mechanism suppressing BMSC osteogenic function, while restoring mitochondrial redox balance can salvage osteogenic gene expression and bone-forming capacity [55,56].

2.6. The effect of scaffold on modulating osteogenesis in vitro

To further evaluate the osteogenic differentiation potential of the conductive nanofibrous scaffolds, in vitro osteogenic experiments were conducted. To investigate the coupling of osteoimmunomodulation with osteogenesis, BMDMs pretreated with P, PP, PP@IL, PP^ES and PP@IL^ES were stimulated with LPS and collected in a conditioned medium (CM) (Fig. 5a). The osteoimmunomodulatory function of PP@IL^ES was then evaluated by collecting the macrophage CM for culturing BMSCs. As illustrated in Fig. 5b, alkaline phosphatase (ALP) staining after 7 days demonstrated relatively weak staining intensity and low ALP activity levels in both the P and PP groups, with no significant differences observed between them (Fig. 5c). In contrast, the PP@IL scaffold group exhibited significantly enhanced ALP staining intensity and activity. Notably, the scaffolds subjected to electrical stimulation, PP^ES and PP@IL^ES, showed even stronger ALP activity, indicating enhanced early-stage osteogenic differentiation. Among these, the PP@IL^ES group displayed the highest ALP activity, demonstrating significant differences compared to other groups. After 21 days of osteogenic induction, Alizarin Red S (ARS) staining was conducted to assess mineralized nodule formation. Consistent with ALP results, mineralization was minimal in both P and PP groups, without significant differences. The PP@IL group showed markedly improved mineral deposition. Furthermore, the electrically stimulated groups, particularly PP@IL^ES, exhibited the highest level of mineralization, significantly surpassing other groups. Quantitative analysis of ARS staining confirmed these results, highlighting the pronounced effect of electrical stimulation in enhancing calcium deposition (Fig. 5d–e). Early extracellular matrix formation was evaluated by collagen type I (COL1) immunofluorescent staining on day 3. Results demonstrated faint staining in P and PP groups, with slightly increased staining intensity in the PP@IL group. However, the PP^ES and PP@IL^ES groups exhibited remarkably stronger COL1 signals and clearer fiber arrangements (Fig. 5f). Quantitative fluorescence intensity analysis further supported these findings, confirming the PP@IL^ES group had the highest COL1 expression level, significantly exceeding all other groups (Fig. 5g). RUNX2 immunofluorescence staining was further performed to evaluate early osteogenic activation in BMSCs (Fig. S23, Supporting Information). Compared with the control and other scaffold groups, the PP@IL^ES group showed stronger RUNX2 fluorescence intensity and more evident nuclear RUNX2 signal after osteogenic induction. This result was consistent with the ALP, ARS, COL1A1 staining, and osteogenic gene expression results, further supporting that PP@IL^ES promotes early osteogenic commitment and osteogenic differentiation of BMSCs.

Fig. 5.

The effect of different scaffold treatments on modulating osteogenesis in vitro. a) Description of the experimental setup. b) Alkaline phosphatase (ALP) staining of bone marrow mesenchymal stem cells (BMSCs) in cultures for 7 days. Scale bars: 200 μm. c) Quantitative evaluation of ALP activity for 7 days. d) Alkaline phosphatase (ARS) staining of BMSCs in cultures for 21 days. Scale bars: 200 μm. e) Quantification of ARS at optical density value (420 nm). f) Immunofluorescent staining of BMSCs cells cultured for 3 days: green (COL1), red (Phalloidin) and blue (DAPI). Scale bars: 100 μm. g) Fluorescence intensity of COL1.h-l) Real-time quantitative reverse transcription polymerase chain reaction (RT‒qPCR) results for osteogenesis-related genes expression (Col1a1, Alpl, Runx2, Bglap, and Spp1) in BMSCs. (n = 5 per group, one-way ANOVA with Dunnett's post hoc test, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns: not significant (p > 0.05)).

To validate these findings at the molecular level, the expression of osteogenesis-related genes, including Col1a1, Alpl, Runx2, Bglap, and Spp1, was assessed via RT-qPCR. Gene expression results demonstrated minimal differences between P and PP groups. In contrast, PP@IL scaffolds significantly upregulated Col1a1, Alpl, Runx2, Bglap, and Spp1 expression. The electrically stimulated PP^ES group showed further significant increases in the expression of key osteogenic regulators. The PP@IL^ES group consistently exhibited the highest expression of all examined osteogenic markers, suggesting enhanced osteogenic differentiation. (Fig. 5h–l). Collectively, these results indicate that IL peptide coating alone moderately promotes osteogenic differentiation, whereas electrically conductive nanofibrous scaffolds under electrical stimulation significantly enhance ALP activity, accelerate mineral deposition, and elevate key osteogenic gene and collagen protein expressions. The combined strategy in PP@IL^ES scaffolds consistently demonstrated more favorable osteogenic performance across multiple indicators.

2.7. In vivo bone regeneration evaluation of scaffold in a rat calvarial defect model

Based on the aforementioned in vitro results, we further investigated the in vivo therapeutic efficacy of conductive nanofibrous scaffolds in promoting bone regeneration. Fig. 6a illustrates the schematic design of the in vivo calvarial defect experiment. The schematic diagram of electrical stimulation of diabetic rats is shown in Fig. S24, Supporting Information. Micro-CT imaging and 3D reconstruction of defect areas were conducted at 4 and 8 weeks post-implantation (Fig. 6b). Results demonstrated minimal new bone formation in the P and PP groups, with most defects remaining filled with fibrous tissues or empty spaces, exhibiting low bone mineral density (BMD) and bone volume fraction (BV/TV), as well as greater trabecular spacing (Tb.Sp). The PP@IL group exhibited notable new bone growth, significantly improved BMD and BV/TV, and correspondingly reduced Tb.Sp. The electrically stimulated PP^ES group demonstrated superior new bone filling compared to the PP@IL group, showing further increases in BMD and BV/TV values. Notably, the PP@IL^ES group exhibited the most extensive and continuous bone regeneration, displaying the highest BMD and BV/TV and the lowest Tb.Sp values, with statistically significant differences compared to all other groups (Fig. 6c, Fig. S25, Supporting Information). These findings suggest that combining conductive nanofibrous scaffolds with electrical stimulation and IL peptide coating provides greater enhancement of bone regeneration than either component alone in this model. H&E staining of heart, liver, spleen, lungs and kidneys after 8 weeks of scaffolds implantation in vivo showed that the scaffolds material and electrical stimulation had good in vivo tissue biosafety (Fig. S26, Supporting Information). To evaluate the long-term biosafety of scaffold implantation, we performed peripheral blood cell counts and serum biochemical analyses 8 weeks after surgery. All haematological parameters, including WBC, RBC, NE, HCT, PLT and HGB (Fig. S27, Supporting Information), remained within normal reference ranges, indicating no evident haematological abnormalities. Likewise, liver and kidney function markers and cardiac enzymes, including ALT, AST, TBIL, CR, CK and LDH (Fig. S28, Supporting Information), were all within physiological limits, suggesting the absence of systemic organ toxicity. Together with the histological evaluation of major organs, these data further support the favorable long-term in vivo biocompatibility of the PP@IL^ES scaffold under the stimulation regimen used in this study.

Fig. 6.

In vivo bone regeneration after different scaffold treatments in a diabetic rat calvarial defect model. a) Schematic diagram of the modeling and animal experiment procedures. b) Micro-CT scanning and reconstructed 3D images of cranial bone defects. c) Quantitative analysis of bone regeneration within the original 5 mm defect ROI, including BMD and BV/TV. d-e) H&E and Masson's trichrome staining of new bone at 4 and 8 weeks. (n = 5 animals per group, one-way ANOVA with Dunnett's post hoc test, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns: not significant (p > 0.05)).

To validate the Micro-CT findings, histological analyses including H&E and Masson's trichrome staining were performed at 4 and 8 weeks post-implantation (Fig. 6d–e). The defect areas in the P and PP groups primarily contained fibrous tissues with sparse and scattered trabeculae, indicative of slower repair processes. The PP@IL group exhibited increased bone-like tissue and trabecular structures, alongside reduced fibrous tissue infiltration, suggesting that the IL peptides contributed to enhanced bone regeneration and reduced fibrosis. Compared to PP@IL, the PP^ES group displayed markedly increased trabecular bone formation at 4 weeks and more mature bone structures at 8 weeks. The PP@IL^ES group consistently showed more favorable bone regeneration, featuring dense trabecular structures at 4 weeks and mature, continuous lamellar bone by 8 weeks, with significantly reduced fibrous tissue presence. Masson's staining further confirmed these differences, highlighting the PP@IL^ES group's denser collagen deposition and more mature bone matrix compared to other groups. To further assess fibrotic responses within the defect area, we performed Sirius Red staining and α-smooth muscle actin (α-SMA) immunohistochemistry on the sections (Fig. S29a–b, Supporting Information). In the P and PP groups, extensive collagen fiber deposition and a high proportion of α-SMA-positive cells were observed, indicative of pronounced fibrotic scar formation. In the PP@IL group, both collagen accumulation and α-SMA expression were reduced, consistent with the antifibrotic activity of the IL peptide. The PP^ES group likewise showed decreased fibrotic components, which may be attributable to an improved immune microenvironment and attenuated inflammatory scarring under electrical stimulation. The PP@IL^ES group exhibited the lowest levels of collagen deposition and α-SMA positivity, indicating that the combined intervention reduced α-SMA positive myofibroblast activation and collagen accumulation, accompanied by enhanced bone regeneration.

To verify whether the beneficial effects of PP@IL^ES extend to load-bearing sites, we next evaluated the therapeutic efficacy of the conductive nanofibrous scaffold in a load-bearing femoral defect model. Micro-CT imaging and three-dimensional reconstruction of the defect region at 4 and 8 weeks post-implantation (Fig. 7a) revealed a pattern consistent with that observed in the calvarial model. In the P and PP groups, only minimal new bone formation was detected, and the defects remained largely occupied by fibrous tissue or voids. In contrast, the PP@IL group showed appreciable regeneration of bone-like trabeculae with significantly increased BMD and BV/TV values compared with P and PP (P < 0.001). However, these parameters were still lower than those in the electrically stimulated PP^ES group. This may reflect partial attenuation of IL peptide activity in the inflammatory diabetic microenvironment, limiting its osteoregenerative efficacy when used alone. Notably, the PP@IL^ES group exhibited the most extensive and continuous bone regeneration, with the highest BMD and BV/TV among all groups and statistically significant differences compared with each of the other conditions (Fig. 7b–c), indicating a synergistic effect between exogenous electrical stimulation and the endogenous bioactive peptide.

Fig. 7.

In vivo bone regeneration after different scaffold treatments in a diabetic rat femoral defect model. a) Micro-CT scanning and reconstructed 3D images of femoral bone defects. b-c) Quantitative evaluation of bone generation in the defect region according to the BV/TV and BMD. d) H&E's trichrome staining of new bone at 4 and 8 weeks. e) Masson's trichrome staining of new bone at 4 and 8 weeks. (n = 5 per group, one-way ANOVA with Dunnett's post hoc test, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns: not significant (p > 0.05)).

Histological analyses further corroborated these findings. H&E and Masson's trichrome staining at 4 and 8 weeks (Fig. 7d–e) showed that defects in the P and PP groups contained sparse, discontinuous trabeculae interspersed with abundant fibrous tissue. In line with the calvarial data, the PP@IL group displayed increased bone-like tissue and more organised trabecular structures, accompanied by reduced fibrous infiltration. The PP^ES group exhibited markedly greater trabecular bone formation at 4 weeks and more mature, continuous bone architecture by 8 weeks. Among all groups, PP@IL^ES showed the most pronounced regeneration, characterized by a dense trabecular network at 4 weeks and the presence of mature, continuous lamellar bone at 8 weeks, with only minimal residual fibrous tissue. Masson's staining confirmed these differences, highlighting denser collagen deposition and a more mature bone matrix in the PP@IL^ES group compared with the other groups. Collectively, these results demonstrate that PP@IL^ES not only promotes regeneration in calvarial defects but also significantly enhances repair in femoral defects, supporting its potential applicability to load-bearing skeletal sites.

These findings collectively demonstrate that combining conductive nanofibrous scaffolds, external microcurrent stimulation, and bioactive peptide coatings can create a highly favorable in vivo environment for bone regeneration, enhance bone mineralization, and effectively inhibit excessive fibrosis and pathological scar formation. This integrated strategy holds significant promise for clinical bone repair and tissue engineering applications.

2.8. Assessment of the inflammatory microenvironment and osteogenic activity in vivo

To evaluate the inflammatory microenvironment in the defect areas, immunohistochemical staining for TNF-α and TGF-β was performed (Fig. S30a, Supporting Information), and quantitative analysis of the positively stained areas was conducted (Fig. S30b, Supporting Information). Results showed extensive TNF-α positive areas and relatively weak TGF-β signals in the P、PP and PP@IL groups, indicating a pronounced inflammatory environment with limited tissue repair signaling. The groups subjected to electrical stimulation, PP^ES and PP@IL^ES, displayed significantly reduced TNF-α-positive staining areas and significantly enhanced TGF-β expression, indicating that electrical stimulation effectively modulated the immune microenvironment, suppressed excessive inflammation, and promoted activation of tissue repair pathways. A schematic diagram illustrating the regulation of the immune microenvironment, as shown in Fig. S30c, Supporting Information.

Immunohistochemical staining for COL1 was performed to observe early extracellular matrix deposition in the defect areas (Fig. S31a, Supporting Information). Quantitative analysis demonstrated limited COL1-positive areas in the P and PP groups, whereas a significant increase was noted in the PP@IL group, suggesting the IL peptide coating effectively accelerated collagen deposition. Importantly, the electrically stimulated PP^ES group exhibited more extensive and intense COL1 staining compared to PP@IL, indicating enhanced collagen synthesis. Notably, the PP@IL^ES group showed the highest COL1 expression, significantly surpassing all other groups, supporting a combined contribution of electrical stimulation and IL peptide coating to extracellular matrix deposition and bone tissue maturation (Fig. S31b, Supporting Information). RUNX2 immunohistochemistry was conducted to assess osteogenic activity at the cellular level (Fig. S32a, Supporting Information). The number of RUNX2-positive cells was limited in the P and PP groups, whereas an increase in RUNX2 expression was observed in the PP@IL group, suggesting partial promotion of osteogenic differentiation by IL peptides. Further enhancement was observed in the electrically stimulated PP^ES group, with significantly increased RUNX2-positive cell populations. The PP@IL^ES group exhibited the highest number of RUNX2-positive cells, demonstrating significant enhancement in osteogenic differentiation and proliferation compared to other groups (Fig. S32b, Supporting Information). This indicates that the dual intervention of electrical stimulation and IL peptide coating synergistically activated the osteogenic gene network, effectively enhancing osteoblast proliferation and differentiation in the defect regions.

2.9. Mechanistic exploration of ternary regulation of immune microenvironmental processes in bone regeneration

Single-cell RNA sequencing (scRNA-seq) was performed on the tissue surrounding the cranial defect to elucidate the molecular mechanisms underlying the scaffold's therapeutic effects. Dimensionality reduction and clustering analyses were conducted, and the resulting cells were grouped into different clusters based on corresponding marker genes (Fig. 8a–b). Cell-type annotation and composition analysis showed that PP@IL^ES treatment altered the cellular landscape of the early defect microenvironment (Fig. 8c). Quantification of the original scRNA-seq annotation data showed that macrophages decreased from 44.51% in the control dataset to 39.70% in the PP@IL^ES treated dataset, whereas fibroblasts increased from 0.66% to 12.28%. These results suggest that PP@IL^ES is associated with macrophage and fibroblast-related remodeling in the diabetic defect microenvironment. Further pathway enrichment analyses using MsigDB (KEGG and GO) highlighted central biological processes related to inflammation and metabolism, including the concurrent inhibition of the mTORC1 pathway in both fibroblasts and macrophages (Fig. 8d). This represents the combined effect of bioactive peptides interacting with exogenous electrical stimulation in a DBD environment. It reflects a multi-level intervention rather than a simple peptide-protein interaction.

Fig. 8.

Single-cell sequencing to study the ternary regulatory mechanisms of PP@IL^ES. a) Cell type intergroup scatterplot. b) Marker gene. c) Percentage of cell types. d) Molecular Signature Database (MsigDB) enrichment analysis. e-f) GSEA enrichment analysis of macrophages and fibroblasts. g-h) Cellular communication analysis. i-j) Representative immunoblots of PI3K/AKT/mTOR signaling pathway regulation in differently treated BMDMs. k) Representative immunoblots of P-S6, S6, α-SMA and KLF4 in fibroblasts from different treatments. l) Schematic representation of the molecular mechanism of macrophage metabolic reprogramming.

Gene set enrichment analysis (GSEA) of macrophages and fibroblasts under PP@IL^ES intervention showed that macrophages upregulated complement activation pathways while downregulating glycolytic metabolic pathways (Fig. 8e). In macrophages, the activation of the complement and coagulation cascades is critical for immune defense in the early stages of inflammation. The complement system marks pathogens, clears necrotic tissues, and recruits immune cells to the site of injury, while the coagulation cascade helps form a barrier in damaged tissue to limit the spread of infection and create a stable environment for tissue repair. These pathways have been implicated in anti-inflammatory responses and tissue repair in previous studies. In addition, key metabolic pathways were significantly downregulated in macrophages. Metabolic status plays a major role in macrophage function: highly active glycolysis is generally associated with pro-inflammatory M1-type macrophages, whereas reduced metabolic activity (relying more on fatty acid oxidation and oxidative phosphorylation) is commonly linked to anti-inflammatory M2-type macrophages. Studies have confirmed that regulation of the PI3K-Akt-mTOR axis can drive M2 polarization in macrophages along with associated metabolic reprogramming [57]. Overall, the activation of complement and coagulation signaling, coupled with the suppression of metabolic pathways, suggests that PP@IL^ES intervention is associated with a shift in macrophages toward an M2-like, anti-inflammatory, and tissue-repair phenotype. In fibroblasts, PP@IL^ES intervention activated the Wnt and MAPK signaling pathways, consistent with enhanced bone repair (Fig. 8f).

To investigate dynamic cell-state transitions, we performed pseudotime trajectory analysis via Monocle, revealing a shift from pro-inflammatory M1-like macrophages to anti-inflammatory M2-like macrophages (Figure S33, Supporting Information). Network analysis further indicated that the PP@IL^ES group exhibited intensified pro-angiogenic signaling from macrophages to endothelial cells, whereas the control group was dominated by pro-fibrotic signaling from macrophages to fibroblasts—emphasizing a transition in the local microenvironment from inflammatory to reparative (Fig. 8g). A cell–cell interaction heatmap showed crosstalk among macrophages, neutrophils, and fibroblasts in the diabetic defect region. Moreover, macrophages were identified as the primary source of CXCL, which may aggravate local inflammation; endothelial cells and fibroblasts, regulated by CD45, could hinder angiogenesis and encourage inflammatory scar formation (Figure S34a-b, Supporting Information). In diabetic bone defects, the MK signaling pathway drives imbalances in inflammation and fibrosis via macrophage–fibroblast interactions. Additionally, fibroblasts and macrophages increase FN1 secretion, with neutrophils and T cells intensifying inflammatory signals and ultimately leading to excessive fibrosis (Fig. 8h). Macrophages receiving FN1 and MAPK signals may, in turn, regulate the classic PI3K-AKT axis involved in tissue repair [58].

Consistent with single-cell analysis results, the treatment group exhibited activation of the PI3K/AKT/mTORC2 signaling pathway in murine bone marrow-derived macrophages (BMDMs). Specifically, the expression of PTEN, an upstream inhibitory molecule in the AKT signaling pathway, was downregulated, while key proteins in the pathway, including PI3K, phosphorylated AKT at Thr308, and phosphorylated AKT at Ser473 (a critical mTORC2 protein), were progressively elevated. This signaling pattern was accompanied by increased expression of M2 associated markers, such as MRC1 (CD206) [59,60], ARG1 [61], and ACOD1 [62], and a concurrent downregulation of the inflammatory marker NLRP3 [63]. Functional assays using PI3K and AKT inhibitors confirmed that the PI3K/AKT [64]/mTORC2 pathway is essential for regulating M2 macrophage polarization markers. Simultaneously, the mTORC1 signaling pathway was inhibited in macrophages, as evidenced by reduced phosphorylation of S6, a key downstream effector molecule of mTORC1. Interestingly, this inhibition also resulted in the upregulation of M2 macrophage polarization markers such as MRC1 (CD206) and ARG1. Treatment with glutamine, an mTORC1-activating metabolite, suggested that mTORC1 might be involved in the feedback regulation of M2 macrophage polarization (Fig. 8i–j). ES may play a major role. Recent studies have shown that ES can reshape cellular metabolism and autophagy via axes such as AMPK–mTOR [65] and Akt2 –miR-29b–mTOR [66], thereby altering cell phenotype and function. Specific reports indicate that particular electrical stimulation paradigms can suppress mTORC1 signaling [67], reduce protein synthesis, and attenuate aberrant glycolysis in certain cell types, including tumor cells. Taken together, these findings support the concept that ES can influence mTORC1 and glycolytic pathways, enabling fine-tuned regulation of macrophage metabolism and phenotype. The aforementioned research further supports our research findings. In summary, the treatment may modulates the PI3K/AKT/mTOR signaling pathway homeostasis, particularly through phosphorylation and dephosphorylation of downstream effector molecules, to regulate the protein levels of M2 macrophage polarization markers and induce M2 macrophage polarization.

Additionally, single-cell analysis data of fibroblasts in the bone microenvironment were validated through Western blot analysis (Fig. 8k). In murine bone marrow-derived fibroblasts, the treatment group demonstrated inhibition of the mTORC1 signaling pathway, resulting in decreased phosphorylation of downstream S6K. This inhibition led to increased expression of KLF4, a transcription factor known to bind the α-SMA promoter and repress its transcription. Consequently, α-SMA levels were significantly reduced, suggesting a decreased potential for scar formation.These findings are consistent with modulation of the mTORC1/KLF4 signaling axis and a concomitant reduction in fibrotic markers within the bone microenvironment [68]. A Schematic diagram of the mechanism regulation is shown in Fig. 8l.This is a transitional schematic summary of the signaling events associated with PP@IL^ES mediated macrophage reprogramming. The upstream PI3K/AKT/mTOR-related changes were supported by the data in Fig. 8, while downstream metabolic regulators such as HO-1 and NRF1 were further validated in Fig. 9.

Fig. 9.

Macrophage metabolomics assay and immunoblotting experiments. BMDMs were induced under high-glucose and LPS conditions, and the PP@IL^ES group received electrical stimulation at 40 mV for 15 min. a) Macrophage subtyping. b-c) Analysis of macrophage subtype transcription factors. d-e)Representative immunoblots P-AKT-473/308, HO-1, NRF1, ATF4 and TFAM signaling pathway regulation and quantified results in differently treated BMDMs. f) The scheme of metabolic pattern in M1 and M2 type macrophages. g-h) Detection of ATP and lactate levels in BMDM cells under different treatments. i-l) Citrate Acid, Fatty Acid Oxidation, Itaconic Acid and Pyruvate metabolism levels were assayed in BMDM cells. m) Representative ECAR profiles and corresponding parameter analysis of BMDMs with different treatments. n) Representative OCR profiles and corresponding parameter analysis of BMDMs with different treatments. (n = 3 per group, one-way ANOVA with Dunnett's post hoc test, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns: not significant (p > 0.05)).

Dimensionality reduction and clustering analyses identified five distinct macrophage subpopulations (Identities 0–4) within the samples (Fig. 9a and Fig. S35, Supporting Information). Identity 0 and Identity 1 both featured gene expressions typically associated with tissue repair and anti-inflammatory (M2-type) macrophages, such as Arg1 and Spp1.In contrast, Identity 2 showed high expression of immune-related genes (C1q family), indicating a pro-inflammatory or classically activated M1-like phenotype potentially involved in innate immune responses. Identity 3 and 4, which comprised fewer cells, displayed distinct transcriptional signatures pointing to specialized or transitional populations; for instance, Identity 3 prominently expressed Ccl6 and Pltp, whereas Identity 4 was characterized by Gzma and Xcl1, suggestive of cytotoxic or other unique immune functions. Overall, Identities 0 and 1 likely correspond to anti-inflammatory, tissue-repair macrophages (M2 or M2-like), and Identity 2 corresponds to pro-inflammatory or classically activated M1-type macrophages.

Transcription factor analysis of the macrophage subtypes (Fig. 9b–c) revealed significant upregulation of NFE2L1 in the M2 groups, suggesting its pivotal role in these cells. As the precursor of NRF1, NFE2L1 participates in regulating mitochondrial metabolic homeostasis and macrophage polarization and plays a critical role in inflammation, immune response, and disease progression [[69], [70], [71]]. For the in vitro validation experiments, BMDMs were induced under high-glucose and LPS conditions. The PP@IL^ES group received electrical stimulation at 40 mV for 15 min. Consistent with these single-cell sequencing results, Western blotting and quantitative analyses (Fig. 9d–e) confirmed a marked increase in NRF1 expression in the treatment group. Furthermore, previous signal-pathway validation demonstrated activation of the PI3K and AKT pathways in macrophages, accompanied by elevated levels of HO-1—an AKT downstream mediator. Based on prior studies [72], it appears the treatment may modulate mitochondrial function in macrophages through the PI3K/AKT/HO-1/NRF1 axis. Additionally, we observed increased levels of TFAM, an essential regulator of mitochondrial metabolism, in the treatment group, aligning with previous findings on the NRF1–TFAM axis [70,71]. Notably, there was no significant increase in the protein level of ATF4, a negative regulatory molecule downstream of mTORC1 [73]. This absence of upregulation may be attributed to negative feedback, wherein elevated NRF1 suppresses ATF4 to mitigate oxidative stress [69,71].

Based on the Western blot results, a significant increase in TFAM often coincides with enhanced mitochondrial metabolism and marked activation of oxidative phosphorylation [74,75]. To further substantiate the mechanistic findings, qPCR analysis was performed to assess the expression of genes related to inflammation regulation, mitochondrial metabolic remodeling, and oxidative stress (Fig. S36a–f, Supporting Information). Compared with the control group, the PP@IL^ES group exhibited downregulation of Casp1 and upregulation of Il1rn, indicating suppression of inflammatory amplification and enhancement of anti-inflammatory regulation. Meanwhile, the expression levels of Nfe2l1 and Ppargc1b were significantly increased, supporting activation of mitochondrial transcriptional regulation and metabolic reprogramming. In addition, Slc25a1 and Sod2 were markedly upregulated, suggesting improved metabolic adaptation and enhanced antioxidant capacity. These qPCR results further support that PP@IL^ES promotes macrophage transition toward a reparative phenotype through coordinated regulation of inflammatory signaling, mitochondrial homeostasis, and immunometabolic remodeling. Macrophage metabolic reprogramming is shown in Fig. 9f. The metabolic profile of BMDMs were further evaluated using metabolic inhibitors and biochemical assays. In the control group, treatment with oligomycin (an inhibitor of oxidative phosphorylation, OXPHOS) significantly enhanced glycolytic activity, as evidenced by increased lactate and ATP production. Conversely, treatment with 2-deoxyglucose (2-DG, a glycolysis inhibitor) markedly reduced lactate and ATP production, indicating a predominant reliance on glycolysis for energy production in untreated BMDMs. In sharp contrast, BMDMs subjected to PP@IL^ES intervention exhibited reduced lactate and ATP production upon oligomycin treatment, while 2-DG treatment still markedly decreased ATP and lactate levels (Fig. 9g–h).

These findings suggest a metabolic shift in PP@IL^ES-treated BMDMs, demonstrating increased reliance on OXPHOS rather than glycolysis. Consistent with this observation, PP@IL^ES-treated BMDMs, suggesting enhanced TCA cycle activity and further validated the enhanced mitochondrial metabolism (Fig. 9i). Fatty acid oxidation (FAO) assays revealed that FAO levels were significantly elevated in the PP@IL^ES group, aligning with increased acetyl-CoA flux into the TCA cycle and enhanced mitochondrial metabolism (Fig. 9j). Moreover, itaconate (an anti-inflammatory metabolite) levels have increased (Fig. 9k). In addition, BMDMs showed decreased pyruvate levels, further indicating suppressed glycolysis and enhanced TCA cycle activity (Fig. 9l).

On this basis, we further measured the extracellular acidification rate (ECAR) and oxygen consumption rate (OCR). Compared with the control group, the PP@IL^ES group exhibited significantly reduced Glycolysis, Glycolytic capacity, and Glycolytic reserve (Fig. 9m). Glycolysis, indicated by ECAR after adding saturating glucose, reflects the cell's basal glycolytic capability. Glycolytic capacity, measured as the maximum ECAR following oligomycin treatment, signifies the cell's maximal energy production via glycolysis once oxidative phosphorylation is inhibited. The difference between these two metrics, referred to as Glycolytic reserve, represents the cell's ability to meet additional energy needs, while non-glycolytic acidification denotes acidification sources outside of glycolysis. In contrast, Basal respiration, Maximal respiration, and ATP production rose markedly (Fig. 9n). Basal respiration reflects oxygen consumption used to fulfill ATP demands and compensate for proton leak under basal conditions, whereas Maximal respiration (induced by FCCP) indicates the highest achievable respiratory rate. Proton leak, calculated by subtracting ATP-linked respiration from basal respiration, can signal mitochondrial damage or function as a regulatory mechanism of ATP synthesis. ATP production gauges the mitochondria's capacity to meet cellular energy requirements, and Spare respiratory capacity—defined as the difference between Maximal and Basal respiration—serves as an indicator of the cell's metabolic flexibility. These findings are consistent with metabolic features typically associated with an M2-like phenotype, including decreased glycolytic activity and increased oxidative phosphorylation [76].

To further examine whether mTORC1 suppression was functionally associated with PP@IL^ES induced macrophage metabolic reprogramming, we performed a rapamycin-based mTORC1 inhibition assay (Fig. S37, Supporting Information). Compared with the Ctrl group, PP@IL^ES markedly reduced lactate production while maintaining a relatively favorable ATP level, indicating decreased glycolytic output and improved cellular energy status. Rapamycin treatment showed a partially similar metabolic trend, especially in reducing lactate production, suggesting that mTORC1 suppression can partly mimic the glycolysis-reducing effect observed after PP@IL^ES treatment. Although rapamycin reduced lactate production, its effect on ATP maintenance was less favorable than that of PP@IL^ES. This result suggests that pharmacological mTORC1 inhibition alone may suppress glycolytic activity but does not fully reproduce the coordinated metabolic remodeling induced by PP@IL^ES.

Oxidative phosphorylation plays a key role in macrophage metabolic homeostasis and M2-type regulation [77,78]. ELISA assays revealed that PP@IL^ES-treated BMDMs exhibited a pronounced M2 anti-inflammatory polarization phenotype, characterized by decreased secretion of pro-inflammatory cytokines (TNF-α, IL-6 and IL-12) and increased levels of anti-inflammatory cytokine IL-10 (Fig. S38, Supporting Information). These cytokine changes are consistent with previously reported metabolic characteristics of M2 macrophages, including higher OXPHOS dependency and increased accumulation of the anti-inflammatory metabolite itaconate. These results were further corroborated by the previously observed increase in ACOD1 protein levels, a critical enzyme associated with itaconate production [[79], [80], [81]]. Additionally, the enhanced FAO pathway aligns with established mechanisms underlying M2 macrophage polarization. In summary, metabolic reprogramming of macrophages critically influences their polarization states, wherein distinct polarization phenotypes exhibit specific metabolic dependencies. M1 macrophages primarily rely on glycolysis, exhibiting increased glucose consumption and lactate production, with reduced oxygen utilization. In contrast, M2 macrophages preferentially utilize oxidative phosphorylation, characterized by enhanced TCA cycle activity and increased FAO, ultimately facilitating a more anti-inflammatory and tissue-reparative macrophage phenotype [82].

3. Discussion

In diseases such as diabetes and related metabolic disorders, the regenerative capacity of bone tissue is significantly impaired, leading to delayed fracture healing and even bone nonunion [83]. Within the diabetic milieu, advanced glycation end-products (AGEs) and oxidative stress sustaining macrophage M1 polarization [84]. Hyperglycemia-induced mitochondrial dysfunction further impedes the transition of macrophages toward the anti-inflammatory M2 phenotype. Immune cells play a key role in bone repair, and persistent M1-type polarization of macrophages often leads to bone repair failure [85]. In addition, chronic hyperglycemia and inflammation also stimulates abnormal activation of fibroblasts [33,86]. Thus, multi-target regulatory strategies that can target both macrophages as well as fibroblasts may be promising approaches to address the challenges of diabetic bone repair.

Electroactive biomaterials have been shown to restore local electrophysiological cues and enhance osteogenic signaling during bone repair [[87], [88], [89]]. In diabetic bone regeneration, piezoelectric or immunomodulatory hydrogels, engineered exosome-loaded hydrogels and metabolic small molecule-functionalized hydrogels have recently been developed to regulate immune imbalance, oxidative stress, metabolic dysfunction, angiogenesis, and osteogenesis [52,[90], [91], [92]]. In addition, peptide-conjugated scaffolds have been used to improve the bioactivity, cell adhesion, and osteogenic performance of synthetic bone-repair materials [93]. Although recent bone regenerative systems have achieved important advances in electroactive stimulation, immune regulation, metabolic remodeling, or peptide mediated biofunctionalization, many approaches still primarily emphasize one dominant regulatory dimension. In contrast, PP@IL^ES was designed as an integrated electro biochemical scaffold that combines a PEDOT containing conductive nanofibrous network, DOPA mediated IL peptide functionalization, and transcutaneous electrical stimulation within one therapeutic mode. After implantation at the bone defect site, the conductive PP@IL scaffold is positioned within the transcutaneous stimulation region and may facilitate local electrical cue transmission and distribution in the defect microenvironment. Meanwhile, the surface immobilized IL peptide provides osteogenic and antifibrotic bioactivity. Through this design, PP@IL^ES is not merely a passive membrane, but a conductive bioactive scaffold capable of coordinating macrophage metabolic reprogramming, mTOR related immune regulation, fibroblast deactivation, antifibrotic remodeling, and osteogenesis in the diabetic defect microenvironment.

Based on this electro biochemical scaffold design, we further explored whether PP@IL^ES could regulate common pathological nodes shared by macrophage mediated inflammation and fibroblast associated fibrotic remodeling. Single cell transcriptomic analysis suggested that PP@IL^ES treatment was associated with reduced mTORC1 related signaling in both macrophages and fibroblasts. Given that the mTOR pathway is a central metabolic axis linking nutrient sensing, inflammatory activation, and cellular remodeling [94], these findings indicate that mTORC1 may serve as a shared signaling node connecting macrophage immune metabolism and fibroblast activation in the diabetic defect microenvironment. Therefore, the ternary regulatory strategy proposed in this study refers to the coordinated modulation of mTORC1 related signaling, macrophage metabolic reprogramming, and fibroblast associated fibrotic remodeling. Through this mechanism, PP@IL^ES may redirect the diabetic microenvironment from persistent inflammation and scar like repair toward a more regenerative state, thereby supporting effective bone defect repair.

Macrophage polarization is closely regulated by cellular metabolism, with mTOR playing a pivotal role. In diabetes, elevated glucose levels cause macrophages to rely heavily on glycolysis for rapid ATP and ROS production, processes positively correlated with increased mTORC1 activity [95,96]. The mechanistic target of mTORC1 pathway has emerged as a crucial regulator of macrophage metabolic reprogramming [97,98], shifting metabolism from glycolysis toward oxidative phosphorylation (OXPHOS) and fatty acid oxidation (FAO), thereby facilitating M2 polarization [99]. In this study, PP@IL^ES reduced glycolytic dependence and promoted mitochondrial metabolic recovery, thereby helping to disrupt the glycolysis inflammation loop. The increased citrate level suggested enhanced TCA cycle activity, while the elevated itaconate level, an anti-inflammatory metabolite associated with M2 polarization [100], further supported a reparative metabolic shift.

PP@IL^ES treatment increased FAO activity in macrophages, which was consistent with a metabolic shift toward mitochondrial oxidative metabolism and was accompanied by reduced mTORC1 activity. Since excessive mTORC1 activation is generally associated with enhanced glycolysis and anabolic metabolism, its suppression may contribute to the increased prominence of FAO under PP@IL^ES treatment [101]. This metabolic remodeling may help reduce glycolysis associated ROS accumulation and inflammatory amplification, while enhanced TCA cycle activity and FAO provide a more stable energy supply for reparative macrophage responses. Electrical stimulation may further support mitochondrial metabolic activity through the PI3K/AKT/mTORC2 axis, thereby favoring M2 like polarization. By reducing glycolytic dependence and enhancing mitochondrial oxidative metabolism, PP@IL^ES may help disrupt the ROS glycolysis inflammation cascade [96,102]. The rapamycin-based inhibition assay further suggested that mTORC1 suppression is functionally involved in PP@IL^ES induced macrophage metabolic remodeling; however, the more favorable ATP/lactate profile induced by PP@IL^ES indicates that this process is likely coordinated with broader mitochondrial regulatory pathways rather than being solely attributable to pharmacological mTORC1 inhibition.

Transcription factor analysis further suggested that NFE2L1, which encodes NRF1, was enriched in reparative macrophage clusters. Together with the PI3K, AKT and HO1 related changes observed in PP@IL^ES treated macrophages, NRF1 may participate in mitochondrial homeostasis and support OXPHOS associated macrophage remodeling through TFAM. The unchanged ATF4 level suggests that mTORC1 related stress signaling was not broadly activated under PP@IL^ES treatment. These findings are consistent with emerging evidence that NRF1 mediated mitochondrial regulation contributes to macrophage polarization and redox balance [[103], [104], [105]].

The mTOR pathway is also closely linked to fibroblast activation and fibrotic repair. Previous studies have shown that mTORC1 inhibition can attenuate excessive fibroblast activation and fibrosis, thereby favoring regenerative healing [106,107]. In this study, PP@IL^ES was associated with reduced fibroblast mTORC1 activity, increased KLF4 expression, and decreased α-SMA expression, suggesting attenuated myofibroblast activation [27]. Under diabetic conditions, persistent inflammatory and metabolic stress may simultaneously sustain macrophage glycolytic inflammation and promote fibroblast mediated scar like remodeling, ultimately impairing bone repair [108]. Therefore, PP@IL^ES may improve diabetic bone regeneration through a ternary regulatory framework involving mTORC1 related signaling, macrophage immune metabolic remodeling, and fibroblast associated fibrotic regulation.This study mainly focused on tissue-level bone repair outcomes, α-SMA and Sirius Red results were interpreted together with Micro-CT, histology, COL1, and RUNX2 staining. Future studies should further assess collagen I/III organization and MMP/TIMP-mediated matrix turnover to better define regenerated matrix quality.

In current orthopedic practice, adjunctive electrical stimulation has been explored to promote fracture healing and postoperative recovery, particularly in patients at risk of delayed repair or impaired regeneration [109]. By modulating the local bioelectrical microenvironment, electrical stimulation may support tissue repair and osteogenic activity, which is relevant to diabetic defects characterized by persistent inflammation, metabolic dysfunction, microvascular impairment, and compromised healing capacity. From a practical perspective, PP@IL may serve as a local conductive bioactive scaffold for diabetic bone defects caused by trauma, tumor or infection related debridement, revision surgery, delayed union, or postoperative bone loss. After implantation at the defect site, the scaffold can be placed within the transcutaneous stimulation region and combined with external electrical stimulation devices to form the PP@IL^ES therapeutic mode, offering potential compatibility with individualized postoperative rehabilitation and Enhanced Recovery After Surgery oriented management [110].

The bioactivity and degradation behavior of PP@IL^ES were also considered in relation to diabetic bone repair requirements. PP@IL^ES is not intended to function as a load bearing substitute, but rather as a local electro biochemical adjunct that supports immune metabolic remodeling, fibrotic microenvironment regulation, and osteogenic repair. The surface immobilized IL peptide provides bioactive cues, and its gradual release over approximately 14 days may match the early inflammatory to reparative transition phase of bone healing. In addition, degradation associated pH and Zeta potential monitoring suggested relatively stable physicochemical behavior without marked pH disturbance or abrupt charge reversal. Although PP@IL was positioned within the transcutaneous stimulation region in vivo, the local electric field strength and current density at the defect site were not directly measured. Future modeling, in situ mapping, large animal studies, and long term biosafety evaluation will be needed to optimize stimulation parameters and assess clinical translational feasibility. For future translation, validated terminal sterilization procedures should also be established to preserve scaffold morphology, peptide bioactivity, conductivity, and cytocompatibility while ensuring sterility.

In summary, PP@IL^ES integrates a conductive nanofibrous scaffold, bioactive IL peptide cues, and transcutaneous electrical stimulation to remodel the diabetic bone defect microenvironment. Rather than acting as a passive membrane, PP@IL^ES may facilitate local electrical cue transmission while providing sustained peptide mediated bioactivity. Mechanistically, PP@IL^ES was associated with coordinated regulation of mTORC1 related signaling, PI3K AKT mTORC2 NRF1 mediated mitochondrial remodeling, macrophage immune metabolic reprogramming, and fibroblast associated fibrotic regulation, thereby supporting osteogenic repair under diabetic conditions. This electro biochemical scaffold can be fabricated by electrospinning and may serve as a local adjunctive strategy for diabetic bone defect repair.

4. Conclusion

This study established a ternary regulatory strategy for diabetic bone defect repair by coordinating macrophage immune metabolism, fibroblast-associated fibrotic remodeling, and mTORC1 related signaling. The PP@IL^ES scaffold, which integrates conductive nanofibers, bioactive IL peptide cues, and electrical stimulation, was associated with reduced mTORC1 activity, macrophage metabolic remodeling from glycolysis toward oxidative phosphorylation and fatty acid oxidation, and enhanced M2-like polarization. Meanwhile, PP@IL^ES attenuated abnormal fibroblast activation and reduced α-SMA expression, thereby limiting scar-like fibrotic remodeling and creating a more favorable microenvironment for bone regeneration. These findings suggest that PP@IL^ES may provide a multifunctional local adjunctive strategy for diabetic bone defect repair.

5. Materials and methods

5.1. Fabrication of PP@IL

PVB (Huizhi Electrospinning, 120S) powder was dissolved in absolute ethanol at room temperature under continuous magnetic stirring to form a homogeneous solution with a concentration of 10% (w/v). Subsequently, add PEDOT:PSS (Sigma-Aldrich, Cat. No. 483095) to the PVB solution as needed to achieve a final concentration of 0.1/0.2/0.3%. Then, vigorously shake and stir the mixture for 24 h to ensure thorough mixing and uniform dispersion. The resulting solution was loaded into a 10 mL syringe equipped with a stainless-steel needle (inner diameter: 0.8 mm). Electrospinning was conducted at an applied voltage of 15 kV, a flow rate of 1.0 mL/h, and a distance of 15 cm between the needle tip and a rotating drum collector (speed: 500 rpm) to produce aligned conductive nanofibrous scaffolds. After electrospinning, the PP scaffolds were sterilized by immersion in 75% ethanol followed by ultraviolet irradiation before peptide coating. The IL peptide was dissolved in sterile PBS and filtered through a 0.22 μm sterile filter. Peptide coating and PBS washing were performed under aseptic conditions in a laminar flow hood. The electrospun fibers (PP scaffold) were immersed in the bioactive IL peptide solution for 24 h to obtain PP@IL scaffolds. The IL peptide used in this study had a purity of 96.85% and a molecular weight of 3170.35 Da. After peptide coating, the scaffolds were dried under aseptic conditions, and stored in sterile containers before further use.

5.2. Characterization of PP@IL

Scanning electron microscopy (SEM; FEI Scios 2 HiVac, USA) was employed to evaluate the surface morphology of the P, PP, and PP@IL scaffolds. Fiber diameters and orientation were quantitatively analyzed from SEM images using ImageJ software (NIH, USA). The chemical compositions of the scaffolds were characterized using X-ray photoelectron spectroscopy (XPS; AXIS Ultra DLD, Japan) and Fourier-transform infrared spectroscopy (FTIR; Thermo Scientific Nicolet iS20, USA). The wettability of scaffold surfaces was evaluated through water contact angle measurements using a contact angle goniometer (SDC 350 KS, China). Additionally, surface roughness and topography were examined via atomic force microscopy (AFM; Bruker Dimension Icon, Germany) in tapping mode. Mechanical properties were assessed using a universal testing machine (INSTRON 3343, USA). To evaluate physicochemical changes during degradation, PP@IL scaffolds were immersed in PBS at pH 7.4 and incubated at 37 °C under gentle shaking. At predetermined time points, the degradation media were collected for pH and Zeta potential measurements using a calibrated pH meter and a Zeta potential analyzer, respectively. To assess scaffold degradation under simulated physiological conditions, PP@IL scaffolds were immersed in PBS containing 10% fetal bovine serum and incubated at 37 °C under gentle shaking. PP@IL scaffolds exposed to the same degradation environment with electrical stimulation were defined as the PP@IL^ES group. At predetermined time points, the scaffolds were collected, rinsed with deionized water, lyophilized, and weighed. The residual weight was calculated as follows: residual weight percentage = Wt divided by W0 × 100%, where W0 represents the initial dry weight and Wt represents the dry weight after degradation. To evaluate IL peptide release, PP@IL scaffolds were immersed in 2 mL PBS and incubated at 37 °C with or without electrical stimulation. At predetermined time points, the supernatants were collected and replaced with equal volumes of fresh PBS. The absorbance of the collected supernatants at 280 nm was measured using a spectrophotometer, and the released IL peptide amount was calculated according to a standard curve. The cumulative release percentage was calculated to compare peptide release profiles under non-stimulated and electrically stimulated conditions. Electrical conductivity of the scaffolds was measured using a four-point probe method (Keithley 2400 SourceMeter, USA). Electrochemical characterization, including microcurrent detection and electrochemical impedance spectroscopy (EIS) were performed with a three-electrode system (CHI660E, China).

5.3. Cell biocompatibility

Bone marrow mesenchymal stem cells (BMSCs) were isolated from the femurs of 6-week-old Sprague–Dawley rats for in vitro experiments. All animal procedures were conducted following protocols approved by the Institutional Animal Care and Use Committee of Soochow University. BMSCs were cultured in α-Minimum Essential Medium (α-MEM; Gibco) supplemented with 10% fetal bovine serum (FBS; Gibco) and 1% penicillin-streptomycin (P/S; Gibco). To simulate hyperglycemic conditions, the glucose concentration in the medium was adjusted to 35 mmol/L. To evaluate biocompatibility, BMSCs were seeded at a density of 2 × 10⁴ cells onto well plates and cultured under standard cell culture conditions (37 °C, 5% CO₂) with different scaffolds as interventions. Cell proliferation was evaluated at days 1, 3, and 5 using a Cell Counting Kit-8 (CCK-8; Beyotime), following the manufacturer's protocol. Additionally, cell viability on different scaffolds was assessed using a Live/Dead Cell Staining Kit (Beyotime), and fluorescence images were captured using a fluorescence microscope (Olympus IX73, Japan). BMDMs were isolated and cultured following established protocols. BMDMs were seeded onto well plates and cultured under different scaffolds. Viability and proliferation of BMDMs on scaffolds were evaluated using the same CCK-8 and Live/Dead assays as described above.

5.4. Macrophage polarization analysis

BMDMs were isolated from the femurs of 6-week-old C57BL/6 mice. Macrophage polarization was evaluated through immunostaining and flow cytometry. BMDMs were seeded in 12-well plates at a density of 4×104 cells per well for 12 h. The cells were induced using 100 ng/mL of LPS (MCE, Cat. No. HY-D1056) and 25 mM of glucose (high glucose), and were co-cultured with various scaffold groups under standard conditions (37 °C, 5% CO₂). Using a custom-built high-throughput ES device [111,112], electrical stimulation was applied to BMDMs at 40 mV, 15min. The scaffold thickness (L) was approximately 0.2 mm. Electric field strength (E) = V/L. Therefore, E = 40 mV/0.2 mm. E = 200 mV/mm. Previous studies have shown that low-intensity electrical stimulation can effectively modulate macrophage polarization and metabolic programming, thereby restoring the balance between M1 and M2 phenotypes [[113], [114], [115]]. In addition, an electric field strength of approximately 200 mV/mm has been demonstrated to be a safe and effective range that promotes mesenchymal stem cell proliferation and improves mitochondrial function [53]. Macrophage polarization was evaluated through immunofluorescence staining and flow cytometry. For immunofluorescence staining, macrophages were fixed, permeabilized, and incubated with primary antibodies against inducible nitric oxide synthase (CD86; 1:500, ABclonal) and CD206 (1:500, ABclonal) overnight at 4 °C, followed by incubation with corresponding fluorescently-labeled secondary antibodies for 1 h at room temperature. Images were acquired using an inverted fluorescence microscope. Flow cytometry was employed to quantify the expression levels of CD86 and CD206. A combination of various conditioned media was collected to induce osteogenesis and alleviate oxidative stress levels. For this purpose, BMDMs were seeded into 6-well plates and cultured in an incubator containing 5% CO₂. After incubation, each conditioned medium was treated with LPS (100 ng/mL) for 12 h. Subsequently, the supernatants from different groups were collected under sterile conditions and filtered to remove cellular debris [116]. Flow cytometry data were analyzed using a sequential gating approach. Initially, cell debris was excluded by plotting forward scatter area (FSC-A) against side scatter area (SSC-A), and the main cell population was gated while avoiding events in the lower-left region corresponding to debris. Next, doublets and cell aggregates were removed by plotting FSC-A versus forward scatter height (FSC-H) and applying a rectangular gate along the diagonal (approximately 45°) to retain single cells. Finally, macrophage subsets were identified by plotting F4/80 on the x-axis and either CD86 or CD206 on the y-axis. Double-positive cells in the upper-right quadrant were considered F4/80+CD86⁺ or F4/80+CD206+. This sequential gating strategy was applied consistently across all samples to ensure reliable identification of target cell populations.

5.5. Oxidative stress and mitochondrial function

For SOD2 (MnSOD) immunofluorescence, BMSCs on coverslips were fixed with 4% paraformaldehyde for 15 min, permeabilised with 0.1% Triton X-100 for 20 min, and blocked with 1% BSA for 40 min at room temperature. Cells were incubated with anti-SOD2 primary antibody (1:200, ABclonal, China) overnight at 4 °C, followed by fluorescent secondary antibody (1:500, ABclonal) for 1 h at room temperature, and counterstained with DAPI. Images were acquired by fluorescence microscopy, and SOD2 fluorescence intensity was analyzed using ImageJ.Intracellular ATP levels in BMSCs were measured using an ATP assay kit (Beyotime, China) according to the manufacturer's instructions. BMSCs were washed with cold PBS, lysed in ATP lysis buffer on ice, and centrifuged at 12,000 g for 5 min at 4 °C. Supernatants were collected, mixed with ATP detection working solution, and luminescence was measured with a microplate reader. ATP concentrations were calculated from a standard curve and normalized to total protein content determined by a BCA assay. Mitochondrial membrane potential (ΔΨm) was evaluated using a JC-1 Mitochondrial Membrane Potential Assay Kit (Beyotime, China). BMSCs were incubated with JC-1 working solution at 37 °C for 20 min in the dark, washed twice with JC-1 buffer, and immediately imaged. The ratio of red (JC-1 aggregates, high ΔΨm) to green (JC-1 monomers, low ΔΨm) fluorescence was calculated from at least five random fields per sample using ImageJ and expressed relative to the P group.

5.6. Evaluation of osteogenic differentiation

BMSCs were cultured in osteogenic induction medium under high glucose conditions (35 mmol/L glucose). They were then subjected to interventions with different scaffolds and the aforementioned collected supernatant according to experimental groupings. For osteogenic induction, BMSCs were cultured in α-MEM supplemented with 10% FBS, 1% penicillin-streptomycin, 100 nM dexamethasone (Sigma-Aldrich, Cat. No. D4902; purity ≥97%; molecular weight, 392.46 Da), 50 μg/mL L-ascorbic acid (Sigma-Aldrich, Cat. No. A5960; BioXtra, purity ≥99.0%; molecular weight, 176.12 Da), and 10 mM β-glycerophosphate disodium salt hydrate (Sigma-Aldrich, Cat. No. G9422; molecular weight, 216.04 Da on an anhydrous basis). ALP staining was performed using a BCIP/NBT Alkaline Phosphatase Color Development Kit (Beyotime, Cat. No. C3206), and mineralized matrix deposition was evaluated using an Alizarin Red S Staining Kit for Osteogenesis (Beyotime, Cat. No. C0148S). ALP activity quantification was performed using an ALP assay kit (Beyotime), following the manufacturer's protocol. Semi-quantitative analysis of ARS staining was carried out by dissolving stained mineralized nodules in 10% cetylpyridinium chloride solution and measuring absorbance at 562 nm.

BMSCs were treated with different scaffold-conditioned media under osteogenic induction conditions. After induction, cells were fixed, permeabilized, blocked, and incubated with anti-RUNX2 primary antibody overnight at 4 °C, followed by fluorescent secondary antibody incubation and DAPI counterstaining. Images were captured by fluorescence microscopy, and RUNX2 fluorescence intensity was quantified using ImageJ. For immunofluorescence analysis of COL1, cells were fixed with 4% PFA for 15 min, permeabilized with 0.1% Triton X-100 for 20 min, and blocked with 1% bovine serum albumin (BSA) for 40 min. Cells were then incubated overnight at 4 °C with a primary antibody against COL1 (1:500, ABclonal), followed by incubation with a fluorescently labeled secondary antibody (1:500, ABclonal) for 1 h at room temperature. Cell nuclei were stained using DAPI (Thermo Fisher Scientific). Fluorescent images were captured using an inverted fluorescence microscope (Zeiss), and fluorescence intensity was quantified using ImageJ softwar.

5.7. Evaluation of diabetic bone regeneration in vivo

All animal experiments were approved by the Ethics Committee of Soochow University (Approval No. SUDA20230625A01). The protocol was approved by the Animal Care and Experiment Committee of Soochow University. All methods were performed in accordance with relevant guidelines and regulations. Male Sprague-Dawley (SD) rats, aged 6 weeks, were fed a high-sugar, high-fat diet for one month. Diabetes was induced by intraperitoneal injection of streptozotocin (STZ; Sigma-Aldrich; C8H15N3O7; molecular weight, 265.22) dissolved in citrate buffer (pH 4.5) at a dose of 30 mg/kg. Rats with sustained blood glucose levels above 16.7 mmol/L were considered diabetic and randomly allocated into five groups (n = 10 for each group): P, PP, PP@IL, PP^ES, and PP@IL^ES. Under anesthesia with 1.5% sodium pentobarbital and strict aseptic conditions, two critical-sized calvarial defects (5 mm in diameter) were surgically created using a dental trephine drill in each rat. Respective scaffolds were implanted into the defects. For groups receiving electrical stimulation (PP^ES and PP@IL^ES), stimulation was delivered every other day starting 24 h post-surgery. Rats were immobilized in a custom-made restraint device to minimize movement and stress during stimulation. For in vivo electrical stimulation, electrical cues were delivered transcutaneously using surface electrodes placed on the skin overlying opposite sides of the cranial defect region. The PP@IL scaffold was implanted at the cranial defect site and located within the stimulation region between the two electrodes. Electrical stimulation was applied at 10 mA and 10 Hz for 15 min using an external stimulator. The electrical stimulation parameters used in this study were selected based on a synthesis of previous clinical and experimental evidence. Clinical studies have reported that transcutaneous electrical stimulation within the range of 5–20 mA, 5–20 Hz, for 10–20 min per session is generally well tolerated and can improve local microcirculation and promote healing in conditions such as non-union fractures [117,118]. In vitro, investigations on human osteoblasts have shown that electrical stimuli of comparable amplitude can accelerate bone tissue regeneration by modulating cell adhesion and Ca²⁺-dependent signalling pathways [119]. In addition, several studies have indicated that ES within this intensity range can attenuate local inflammatory responses and thereby support tissue repair [120,121].

At predetermined time intervals (4 and 8 weeks post-implantation), rats were euthanized by CO₂ asphyxiation, and skull tissues were harvested for further analysis. In addition, major organs, including the heart, liver, lung, spleen, and kidney, were collected from the rats for H&E staining at 8 weeks to evaluate the biocompatibility of the scaffolds. No signs of pain or discomfort were observed after surgery or throughout the study. Harvested skulls were scanned using a Micro-CT system (SkyScan 1176). Raw projection images were reconstructed using NRecon software version 1.7.4.2 with alignment correction. Quantitative analysis was performed using CTAn software. For each calvarial defect, the original 5 mm circular defect was used as the fixed ROI, with an analysis thickness of 0.8 mm. Adjacent native calvarial bone outside the defect margin was excluded. Based on the overall grayscale distribution of the calvarial samples, the bone segmentation threshold was set to 90–255 and applied consistently to all samples. BMD was calibrated using a hydroxyapatite phantom, and BMD, BV/TV, and Tb.Sp were calculated within the defined ROI. The animal was used as the statistical unit, with 5 animals per group.

Histological Analysis: Skull specimens were decalcified in 10% ethylenediaminetetraacetic acid (EDTA; Sigma-Aldrich) for approximately one month, embedded in paraffin, and sectioned at 6 μm thickness. Sections were stained with hematoxylin and eosin (H&E) and Masson's trichrome for general morphological evaluation and collagen deposition analysis. Immunohistochemical analyses for tumor necrosis factor-alpha (TNF-α; 1:500, ABclonal) and transforming growth factor-beta (TGF-β; 1:500, ABclonal) were conducted at 4 weeks post-implantation to evaluate inflammation and tissue remodeling. Immunohistochemical staining results were quantitatively analyzed using Image software.

5.8. Quantitative PCR assay

Total RNA extraction was performed with TRIzol reagent (Invitrogen), and cDNA synthesis was conducted using the PrimeScript RT reagent kit (Takara), and qRT-PCR was conducted using Thermo's Maxima SYBR Green/ROX qPCR Master Mix, with β-actin serving as the internal control. The mRNA levels of iNOS, TNF-α, IL-10, TGF-β, ALPL, COL1A1, RUNX2, Bglap, Spp1, β-actin were determine using the 2 –ΔΔCT method. Primer sequences used for gene analysis are provided in Supplementary Tables 4–5.

5.9. Single-cell sequencing analysis

Following euthanasia at 1week post-implantation, tissues surrounding cranial defects were harvested. The scRNA-seq analysis was performed at week 1 to capture early immune and metabolic changes after scaffold implantation. This timing is relevant because diabetic conditions often impair the transition from inflammatory M1 like macrophages to reparative M2 like macrophages. Fresh tissues surrounding the defect region were collected and preserved in MACS Tissue Storage Solution until processing. Samples were enzymatically dissociated using a Tumor Dissociation Kit and neutral protease at 37 °C with gentle agitation. After digestion, the cell suspension was filtered through a 70 μm cell strainer and centrifuged. Red blood cells were removed using red blood cell lysis buffer. The remaining cells were washed with PBS containing 0.04% BSA, resuspended in PBS containing 0.04% BSA, and filtered through a 35 μm cell strainer. Cell viability was assessed by AO/PI staining using a Countstar Fluorescence Cell Analyzer.

Single-cell transcriptomic libraries were prepared using the BD Rhapsody system. Single-cell suspensions were loaded into microwells, and barcoded beads were used to capture mRNA from individual cells. After cell lysis, reverse transcription, ExoI digestion, and cDNA amplification, whole-transcriptome libraries were generated using the BD Rhapsody single-cell WTA workflow. Library quality and concentration were assessed using a Bioanalyzer and Qubit assay. Libraries were sequenced on a DNBSEQ-T7 platform with 150 bp paired-end reads.

Raw sequencing data were processed using fastp to remove adapters and low-quality reads. UMI-tools was used to identify cell barcodes and process UMI-based count data. Clean reads were aligned to the rat genome using STAR, and gene expression matrices were generated for downstream analysis. Cells expressing more than 200 genes and with mitochondrial UMI proportions below 20% were retained for analysis. Mitochondrial genes were removed from the expression matrix.

Downstream analysis was performed using Seurat. Data were normalized and scaled, and principal component analysis was conducted based on the top 2000 highly variable genes. The top principal components were used for dimensionality reduction and clustering. Cell clusters were identified using a graph-based clustering method, and marker genes were calculated using the FindAllMarkers function with the Wilcoxon rank-sum test. Cell types were annotated according to canonical marker genes, and selected cell populations were further reclustered when necessary.

Pseudotime trajectory analysis was performed using Monocle2 with DDRTree. Branch-dependent genes were identified using BEAM analysis. Cell-cell communication analysis was conducted using CellPhoneDB based on normalized expression matrices. Transcription factor regulatory activity was analyzed using pySCENIC. Gene set activation and pathway enrichment were evaluated using QuSAGE. Differentially expressed genes between groups were identified using the FindMarkers function in Seurat with the Wilcoxon rank-sum test. Gene co-regulatory modules were analyzed using the find_gene_modules function in Monocle3.

5.10. Immunoblotting and metabolite assays

For the in vitro macrophage mechanism experiments, BMDMs were induced under high-glucose and LPS conditions and then treated with different scaffold groups. In the PP@IL^ES group, electrical stimulation was applied at 40 mV for 15 min. Cells were washed twice with ice-cold PBS and lysed in ice-cold RIPA lysis buffer (Beyotime, China) supplemented with 1 mM PMSF (Beyotime, China) and a 1× protease and phosphatase inhibitor cocktail (Selleck, China). Following lysis, samples were briefly sonicated on ice for 10–15 cycles of 2 s each to ensure complete disruption while minimizing protein denaturation. The lysates were then clarified by centrifugation at 10,000 × g for 10 min at 4 °C, and protein concentrations were determined using a BCA protein assay kit. Equal amounts of protein were mixed with 5× SDS sample loading buffer, denatured, separated by SDS-PAGE, and transferred onto 0.45-μm PVDF membranes at 100 V for 1 h. After blocking with 5% non-fat dry milk in TBST for 1 h, the membranes were incubated overnight at 4 °C with the following primary antibodies: p-PI3K (CST, Cat. No. 17366), PI3K (CST, Cat. No. 4292), p-AKT Ser473 (CST, Cat. No. 4060), p-AKT Thr308 (CST, Cat. No. 13038), AKT (CST, Cat. No. 9272), PTEN (CST, Cat. No. 9188), ARG1 (CST, Cat. No. 93668), ACOD1 (CST, Cat. No. 19857), p-S6 (CST, Cat. No. 4858), S6 (CST, Cat. No. 2217), MRC1 (CST, Cat. No. 24595), α-SMA (Proteintech, Cat. No. 14395-1-AP), KLF4 (Abcam, Cat. No. ab215036), HO-1 (Abcam, Cat. No. ab305290), NRF1 (Abcam, Cat. No. ab175932), ATF4 (Abcam, Cat. No. ab216839), TFAM (Abcam, Cat. No. ab307302), and β-actin (CST, Cat. No. 4970). The following day, the membranes were washed three times with TBST for 10 min each and incubated with HRP-conjugated secondary antibodies, including goat anti-mouse or goat anti-rabbit IgG (Proteintech), for 90 min at room temperature. After four additional washes with TBST, protein bands were visualized using enhanced chemiluminescence detection reagent (Vilber, France). Densitometric quantification was performed using ImageJ software. Phosphorylated protein levels were normalized to their corresponding total protein levels, while other target proteins were normalized to β-actin.

Metabolite measurements were performed using BMDMs after different treatments. Cell culture supernatants or cell lysates were collected according to the requirements of each assay. Lactate and ATP levels were measured using commercial assay kits according to the manufacturers’ instructions. For ATP source analysis, oligomycin or 2-deoxyglucose was used to distinguish mitochondrial and glycolytic ATP production. Pyruvate, itaconate, and fatty acid oxidation levels were detected using corresponding commercial assay kits, and the results were normalized to protein content or control values when appropriate. To assess the functional involvement of mTORC1 suppression in macrophage metabolic remodeling, BMDMs were cultured under high-glucose/LPS inflammatory conditions and assigned to Ctrl, PP@IL^ES, and Rapamycin groups. The PP@IL^ES group was treated with PP@IL scaffolds under electrical stimulation, whereas the Rapamycin group was treated with rapamycin for 24 h to inhibit mTORC1 signaling. ATP and lactate levels were measured to evaluate cellular energy status and glycolytic output, respectively.

The extracellular acidification rate and oxygen consumption rate of BMDMs were measured using a Seahorse XFe96 Extracellular Flux Analyzer. For OCR analysis, cells were incubated in Seahorse assay medium supplemented with glucose, pyruvate, and glutamine, followed by sequential injection of oligomycin, FCCP, and rotenone plus antimycin A. For ECAR analysis, cells were incubated in Seahorse assay medium supplemented with pyruvate and glutamine, followed by sequential injection of glucose, oligomycin, and 2-deoxyglucose. Before measurement, cells were washed and equilibrated in assay medium in a non-CO₂ incubator. OCR and ECAR parameters were calculated according to the standard Seahorse protocols.

5.11. ELISA analysis

Cytokine concentrations were determined by ELISA using commercially available kits: TNF-α (ABclonal), IL-10 (Cusabio, CSB-E04594m), IL-6 (CSB-E04639m) and IL-12 (YOBIBIO). Cell culture supernatants and tissue extracts were diluted according to the kit instructions, added to pre-coated ELISA plates alongside standards, and incubated at 37 °C for 1 h. After washing, detection reagents were applied, and absorbance was measured at 450 nm. Cytokine concentrations were calculated using standard curves and normalized to relative levels.

5.12. Statistical analysis

The data were expressed as mean ± SD; for comparisons among three or more groups, one-way ANOVA followed by Dunnett's post hoc test was used; for datasets involving two independent variables, two-way ANOVA with Šídák's multiple-comparisons test was applied. All statistical analyses were performed using SPSS 20.0, and graphs were generated with Microsoft Excel and GraphPad Prism 9.0. Statistical significance was defined as ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001 and ∗∗∗∗p < 0.0001.

Ethics approval and consent to participate

All animal experimental procedures in this study were reviewed and approved by the Ethics Committee of Soochow University, with the approval number SUDA20230625A01. All experiments were conducted in accordance with relevant institutional and national guidelines for the care and use of laboratory animals. Consent to participate was not applicable, as this study did not involve human participants.

CRediT authorship contribution statement

Haifu Sun: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Validation, Visualization, Writing – original draft, Writing – review & editing. Wei Geng: Conceptualization, Data curation, Formal analysis, Investigation, Methodology. Peng Yang: Conceptualization, Data curation, Formal analysis, Methodology, Software. Xuequan Zhao: Conceptualization, Data curation, Formal analysis. Yu Chen: Conceptualization, Data curation, Software, Validation. Yu Wang: Conceptualization, Data curation, Methodology. Yuchen Qian: Conceptualization, Data curation, Formal analysis. Yonggang Li: Data curation, Formal analysis, Methodology. Zhonglai Qian: Project administration, Supervision. Longpo Zheng: Funding acquisition, Investigation, Project administration, Resources, Supervision. Nanning Lv: Conceptualization, Data curation, Formal analysis, Funding acquisition, Project administration, Software, Supervision, Validation. Yusen Qiao: Conceptualization, Data curation, Funding acquisition, Investigation, Project administration, Resources, Supervision, Writing – review & editing. Dechun Geng: Conceptualization, Data curation, Investigation, Project administration, Resources, Supervision, Validation, Writing – review & editing.

Declarations of competing interest

The authors declare no competing interests. All authors approved the manuscript and its publication.

Appendix A. Supplementary data

The following is the supplementary data to this article: