Assembly of bioinspired multifunctional microspheres for enhanced alveolar bone regeneration

Abstract

Regeneration of the alveolar bone remains a major clinical challenge due to the complex oral microenvironment and the need for coordinated restoration of multiple tissue types. To overcome these hurdles, biomaterials designed for periodontal regeneration must meet a rigorous set of criteria, including excellent injectability, mechanical stability, selective cell repopulation, and strong osteoinductive capacity. In this study, we developed a bioinspired, multifunctional microsphere system that fulfills these requirements. The system is injectable, mechanically robust, selectively binds bone marrow-derived stem cells (BMSCs), and exhibits potent osteoinductivity. These multifunctional properties were achieved by UV-assembling nanofibrous hollow microspheres (NFH-MS), conjugating the BMSC-specific E7 peptide to the nanofibrous shell, and encapsulating a bone-forming peptide (BFP) within the hollow core. UV-assembly enhanced the scaffold's mechanical integrity, generated interconnected macropores to support cell infiltration, and promoted intercellular communication. Notably, it significantly upregulated Connexin 43 and N-cadherin-mediated junctions, further facilitating cellular interactions. In synergy with E7 and BFP, the UV-assembled NFH-MS scaffold markedly improved BMSC adhesion, osteogenic differentiation, and biomineralization. This bioinspired multifunctional NFH-MS platform demonstrated superior alveolar bone regeneration in a rat fenestration defect model, offering a promising and minimally invasive strategy for periodontal tissue engineering.

Graphical abstract

Highlights

• •Injectable nanofibrous microsphere-based 3D scaffolds are UV-assembled in situ in defect sites.
• •Interconnected macropores and high scaffold modulus accelerate 3D cell infiltration and multicellular communication.
• •The assembled 3D scaffold promotes CX43/N-cadherin coupling and p-STAT3/RUNX2 signaling.
• •A synergistic microenvironment within the scaffold promotes periodontal tissue regeneration.

1. Introduction

Periodontitis remains one of the most prevalent chronic diseases, leading to the progressive destruction of the tooth's supporting structures, including cementum, periodontal ligament (PDL), and alveolar bone [1]. This destruction compromises not only the stability of the tooth but also the overall oral health and quality of life of affected individuals. Conventional therapeutic approaches, such as scaling, root planning, and surgical interventions, primarily aim to arrest disease progression, but often fall short in restoring the lost periodontal structures and functions. As a result, the development of effective strategies for periodontal regeneration has become a critical focus in dental research and clinical practice [[2], [3], [4], [5]].

Despite notable advancements in dental tissue engineering, current approaches to periodontal regeneration exhibit significant limitations [6,7]. Techniques such as guided tissue regeneration (GTR), bone grafting, and the application of growth factors have demonstrated partial success but often yield inconsistent and unpredictable clinical outcomes [8]. Given the complex and irregular geometry of periodontal defects, injectable biomaterials are particularly appealing due to their minimally invasive delivery and ability to conform to intricate defect shapes [9]. Among them, hydrogel-based materials have been extensively investigated for periodontal tissue regeneration [10]. However, many hydrogels exhibit limited biological activity and fail to effectively promote the differentiation of stem cells into alveolar bone in periodontium [11]. Additionally, conventional hydrogels often restrict cell migration within their matrix until the outer layers have been sufficiently degraded, thereby impeding their overall effectiveness in supporting tissue regeneration [12].

Besides injectability, a biomaterial designed for periodontal regeneration should also fulfill several other critical criteria [13]. First, it must facilitate selective cell repopulation within the defect site, meaning to promote the adhesion and proliferation of osteogenic stem cells and inhibit the infiltration of fibroblasts and epithelial cells. This selectivity in cell recruitment is critical for guiding the regenerative process toward the formation of functional periodontal tissues [14]. Second, it must exhibit intrinsic biological activity, including the capacity to deliver bioactive molecules that stimulate stem cell differentiation into specialized periodontal tissues [14]. Third, the biomaterial should be stable within the defect area under physiological conditions, providing a supportive scaffold for tissue in-growth and spatial organization [15,16]. Despite extensive research efforts, no existing biomaterial fully satisfies all these requirements, particularly in the context of alveolar bone loss, highlighting the need for innovative solutions [6,13].

In our previous work, we sought to overcome some challenges by developing a series of innovative biomaterial strategies centered on nanofibrous hollow microspheres (NFH-MS) [[17], [18], [19], [20]]. These microspheres were injectable biomaterials, and offered a more favorable microenvironment for cell growth compared to conventional hydrogels [17]. Their nanofibrous architecture closely mimics the extracellular matrix (ECM), enhancing cell attachment, proliferation, and differentiation [18]. To refine cell selectivity, we identified and incorporated the E7 peptide comprising the amino acid sequence of EPLQLKM with high specific affinity to bone marrow-derived stem cells (BMSCs), onto the surface of the microspheres [19]. E7 acts as a biological barrier that selectively promotes the adhesion and proliferation of BMSCs while inhibiting the infiltration of fibroblasts and gingival epithelial cells [19]. This targeted recruitment is essential for directing the regenerative process toward the formation of functional periodontal tissues. Additionally, we developed a double emulsion process to encapsulate a bone-forming peptide (BFP) within the hollow core of the microspheres [20]. This design enables sustained release of BFP, providing continuous osteogenic stimulation and significantly enhancing bone regeneration.

Another challenge is ensuring the stability of the injectable biomaterial. Microspheres often show limited stability, with some becoming dislodged after administration. This unintended dislodgement from the targeted defect site diminishes therapeutic efficacy and introduces potential risks, including localized inflammation and interference with surrounding tissues. Such spatial instability compromised the precision and reliability of the regenerative process [7]. Moreover, when microspheres function as isolated units, they lack the ability to support multicellular communication—an essential component of tissue development and function. In particular, they fail to facilitate the formation of adequate gap junctions, which are specialized intercellular connections that enable direct cell-to-cell signaling. Gap junction–mediated communication has been shown to significantly contribute to osteogenic efficiency, underscoring its critical role in bone regeneration [21]. The absence of these signaling pathways in our earlier microsphere system compromises its capacity to replicate the complex cellular interactions required for robust and functional tissue formation.

To prevent microsphere dislodgement and enhance multicellular communication, microspheres need to be assembled into a stable three-dimensional (3D) macroporous scaffold after injection into the defect site. Traditionally, such microsphere-based scaffolds are formed using sintering techniques, such as heat sintering, solvent-based sintering, or selective laser sintering [22]. However, these methods form pre-formed 3D scaffolds and lose the injectability, which limits their suitability for periodontal regeneration. Moreover, no studies to date have successfully demonstrated the assembly of nanofibrous microspheres into a stable 3D scaffold.

In this work, we present an approach that assembles NFH-MS into 3D macroporous scaffolds. This strategy transforms the microspheres from discrete units into an integrated scaffold structure after injecting the NFH-MS into the defect site and offers enhanced mechanical stability and supports multicellular interactions. Specifically, we incorporated double bonds onto the surfaces of NFH-MS, which enables crosslinking of the microspheres after injection and allows them to form a stable 3D macroporous scaffold within the defect site. This crosslinked structure prevents dislocation and maintains the spatial organization necessary for tissue regeneration. Moreover, the macroporous architecture of the scaffold facilitates the formation of gap junctions, promoting robust intercellular communication and significantly enhancing osteogenic efficiency.

In addition to structural improvements, we retained the dual-functionalization strategy from our previous work. E7 was incorporated onto the nanofibers of the microspheres to maintain selective cell adhesion, ensuring that BMSCs preferentially populate the scaffold. Simultaneously, BFP was encapsulated within the hollow core of the NFH-MS to provide sustained osteogenic stimulation. This combination of selective recruitment and continuous bioactivity creates a synergistic environment conducive to periodontal tissue regeneration (Scheme 1). We evaluated the performance of the assembled microspheres in a periodontal defect model, focusing on their ability to promote alveolar bone regeneration.

Scheme 1.

Schematic illustration of synthesis and UV-assembly of bioinspired multifactional microspheres for enhanced alveolar bone regeneration.

2. Materials and methods

2.1. Materials

Gelatin (∼225g Bloom, Type B), methacrylic anhydride (MA), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), 2-(N-morpholino) ethanesulfonic acid (MES), N-hydroxy succinimide (NHS), glycine, calcium dichloride, 2-Hydroxy-4′-(2-hydroxyethoxy)-2-methylpropiophenone, and disodium hydrogen phosphite were purchased from Sigma (St Louis, MO). Mineral oil, isopropanol, ethanol, and hexane were ordered from VWR Scientific (Seattle, WA). Bone forming peptide-1 (BFP) and E7 peptide were synthesized by Biomatik Co. (Wilmington, DE). PrimeScript™ RT Reagent Kit and TB Green Premix Ex TaqII were ordered from Takara Bio (Mountain View, CA). Rat bone marrow derived stem cells (BMSCs) were purchased from Lonza (PT-2501, Basel, Switzerland). LIVE/DEAD™ Viability/Cytotoxicity Assay Kit (Fisher, L32250).

2.2. Preparation of NFH-MS

GelMA was synthesized by the reaction of gelatin with methacrylic anhydride (MA). Briefly, 10 g of gelatin was dissolved in 100 ml of PBS (pH 7.5) at 60 °C. After stirring for 3 h, the solution was diluted to 2 % (w/v) with PBS, followed by dialysis against distilled water (MWCO = 14 kDa) at 60 °C for 7 d. The water was replaced every 12 h. The resulting GelMA was lyophilized for 1 week and stored at −20 °C until further use.

NFH-MS were fabricated using a combination of a oil-in-water-in-oil (O/W/O) double emulsion technique to create the hollow structure and a thermally induced phase separation method to generate the nanofibrous surface. In detail, 1.2 g of GelMA was dissolved in 10 ml of a 50 % (v/v) ethanol–water solution at 50 °C for 20 min. Under mechanical stirring, 7 ml of mineral oil was added, followed by 50 ml of additional mineral oil to form the O/W/O double emulsion. The mixture was stirred for 30 s and immediately poured into 800 ml of a precooled (−80 °C) isopropanol/hexane/ethanol solution (4:3:1, v/v/v) under mild magnetic stirring (200 rpm) to induce phase separation and solidify the microspheres. The resulting microspheres were suspended in acetone/water (9:1, v/v) and crosslinked using EDC and NHS in MES buffer (0.05 M, pH 5.3) at 4 °C for 24 h. After crosslinking, microspheres were washed with 100 % and 50 % ethanol (10 min each), incubated twice in 0.1 M glycine solution for 2 h to quench residual crosslinkers, and sequentially washed with 50 %, 70 %, and 100 % ethanol. Microspheres were size fractionated using stacked mesh sieves, and particles in the range of 63–105 μm were collected, lyophilized, and stored for future use. Dried NFH-MS were packed into a cylindrical mold and exposed to UV light to form interconnected microsphere networks. Morphology, hollow structure, and microsphere connectivity were characterized using scanning electron microscopy (SEM) and confocal microscopy (Leica STP6000).

Incorporation of E7 and BFP into NFH-MS scaffolds.

BFP was encapsulated within NFH-MS during the thermally induced phase separation process [20]. To modify the microsphere surface with E7, sulfo-SMCC was used to activate NFH-MS, followed by dropwise addition of E7 peptide as previously described. [20]. Briefly, NFH-MS (60 mg) were dispersed in deionized water, followed by the addition of sulfo-SMCC (5.2 mg, 12 μmol). The reaction mixture was gently stirred at room temperature for 30 min to allow amine-to-NHS ester coupling. Next, E7 peptide (14 mg, 12 μmol) was added dropwise and the reaction was continued for 8 h under gentle stirring to enable thiol–maleimide conjugation. The resulting E7-modified microspheres were thoroughly washed to remove unreacted peptide and lyophilized for subsequent experiments.

To encapsulate BFP into CaP nanoparticles, 6 mg of BFP was dissolved in 1 mL of aqueous CaCl₂ solution (400 mM) and subsequently dispersed in cyclohexane containing surfactant to form microemulsion A. Separately, 1 mL of Na₂HPO₄ solution (400 mM) was dispersed in cyclohexane to generate microemulsion B. Next, microemulsion B was slowly added to microemulsion A at a rate of 10 mL/h under continuous stirring at room temperature. The reaction was allowed to proceed for 1 h to facilitate CaP nucleation and growth. The resulting CaP/BFP nanoparticles were collected by centrifugation at 10,000 rpm for 15 min, washed four times with ethanol to remove residual surfactants, redispersed in deionized water, and lyophilized for further use.

To encapsulate CaP/BFP NPs into NFH-MS, the lyophilized CaP/BFP nanoparticles were first uniformly dispersed in the GelMA solution at the initial stage of the oil-in-water emulsion. Subsequently, a double emulsification combined with thermally induced phase separation was performed to generate hollow microspheres with a nanofibrous shell structure. During this process, CaP/BFP nanoparticles were preferentially entrapped within the hollow interior and partially embedded within the nanofibrous matrix of the microspheres. The obtained NFH-MS were subsequently crosslinked, washed, and lyophilized as described above.

To evaluate the release profile of BFP, both UV-assembled and non-assembled microspheres were immersed in 1 ml of PBS and incubated at 37 °C on a shaking bed. At designated time intervals, the supernatant was collected and replaced with an equal volume of fresh PBS. At the final time point, each sample was sonicated to extract any residual BFP retained within the microspheres, representing the unreleased fraction. All collected supernatants were analyzed using UV–vis spectrophotometry to quantify cumulative BFP release.

To evaluate scaffold degradation, UV-assembled and non-UV-assembled NFH-MS were first freeze-dried and weighed to obtain the initial dry mass (W₀). The samples were then immersed in artificial saliva (Sigma-Aldrich, USA) and incubated at 37 °C on a shaker. At predetermined time points, the samples were collected, thoroughly rinsed with deionized water, freeze-dried, and weighed again (W_t). The remaining mass was calculated as the percentage of W_t relative to W₀.

2.3. Cell migration assay

To visualize cell migration within the scaffold, FITC-labeled microspheres were prepared. Briefly, the microspheres were sterilized in 75 % ethanol for 2 h and rinsed three times with PBS. The sterilized microspheres were packed into cylindrical molds (6 mm in diameter × 3 mm in height), immersed in a photo-initiator solution, and crosslinked under UV light to form stable scaffolds.

A total of 1 × 10⁶ cells were seeded onto the top surface of each scaffold and allowed to attach for 40 min before gently adding cell culture medium. At 2, 6, and 12 h post-seeding, the samples were fixed with 4 % paraformaldehyde, permeabilized, and stained with phalloidin-633 for 45 min followed by Hoechst staining to label cell nuclei. Cell morphology and migration depth were examined using a Leica SP8 digital light-sheet microscope. Briefly, z-stack images were acquired using a Leica SP8 light-sheet microscope and converted to maximum-intensity projections. The scaffold surface was delineated based on the phalloidin (F-actin) signal, and the migration depth was measured as the perpendicular distance from this surface to the deepest phalloidin-positive cell layer in ImageJ. For each scaffold, measurements were taken at three randomly selected positions and averaged to obtain one migration depth value per sample. For the quantification of cell penetration, rectangular regions of interest (200 μm in width × 50 μm in height) were placed at 0–50 μm and 50–100 μm below the surface, and the number of Hoechst-positive nuclei within each ROI was counted with the Cell Counter plugin in ImageJ.

2.4. Cell proliferation and live/dead staining

The viability and proliferation of BMSCs (Lonza, PT-2501, Basel, Switzerland) on microspheres were assessed using the MTT assay on days 1, 3, and 5 after cell seeding, following the manufacturer's protocol. Cells at passages 3–5 were used for all in vitro experiments. BMSCs were cultured in α-MEM (Gibco, USA) supplemented with 10 % fetal bovine serum (FBS, Gibco, USA) and 1 % penicillin–streptomycin (Gibco, USA) and maintained at 37 °C in a humidified atmosphere containing 5 % CO₂. To further evaluate cell viability, live/dead staining was conducted on the microspheres. After appropriate incubation, live cells were stained with Calcein-AM (green) and dead cells with propidium iodide (red). The samples were imaged under a fluorescence microscope.

2.5. Intercellular communication and immunofluorescence staining

To evaluate intercellular communication within microsphere scaffolds, Lucifer Yellow dye diffusion was performed. BMSCs were seeded onto microsphere scaffolds and cultured for 24 h. Lucifer Yellow (0.5 mg/ml) was added to the culture medium and incubated at 37 °C for 10 min. Following incubation, the scaffolds were gently rinsed with PBS, fixed with 4 % paraformaldehyde, and counterstained with DAPI. Dye diffusion patterns were visualized using confocal microscopy (Leica SP8), and fluorescence intensity was quantified to assess the extent of intercellular dye transfer.

To investigate cell–cell junctions and signaling activation, immunofluorescence staining was conducted. For adhesion-related markers, samples were fixed after 24 h of culture and incubated with primary antibodies against Connexin 43 (CX43) and N-cadherin (NCAD), followed by fluorescently labeled secondary antibodies. Nuclei were stained with DAPI, and FITC-labeled microspheres were used to visualize the scaffold architecture.

For osteogenic signaling analysis, phosphorylated STAT3 (p-STAT3) expression was evaluated under the same conditions. After fixation, samples were stained with anti–p-STAT3 antibody and imaged using confocal microscopy. The fluorescence intensities of CX43, NCAD, and p-STAT3 were quantified using ImageJ software. Data were collected from three independent biological replicates.

2.6. Western blot analysis

To evaluate intercellular adhesion and downstream signaling activation, Western blot analysis was performed on BMSCs cultured on microsphere scaffolds. After 24 h of culture, cells were lysed using RIPA buffer containing protease and phosphatase inhibitors. Total protein concentration was quantified using the BCA assay (Thermo Fisher Scientific).

Equal amounts of protein (25 μg) were separated by SDS-PAGE and transferred onto PVDF membranes (Millipore). Membranes were blocked with 5 % non-fat milk in TBST (Tris-buffered saline with 0.1 % Tween-20) for 1 h at room temperature, followed by incubation with primary antibodies overnight at 4 °C. The following primary antibodies were used: anti–NCAD, anti–CX43, anti–phospho-CX43 (Ser368), anti–STAT3, anti–phospho-STAT3 (Tyr705), and anti–β-actin (loading control). After washing, the membranes were incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. Protein bands were visualized using an enhanced chemiluminescence detection system and quantified with ImageJ software. Band intensities were normalized to β-actin, and phosphorylation levels were normalized to corresponding total protein levels. All experiments were performed in triplicate (n = 3).

2.7. Cell adhesion, cytoskeletal staining, and early osteogenic signaling

To evaluate the effects of UV crosslinking and E7/BFP functionalization on BMSC adhesion and cytoskeletal organization, cells were seeded onto various microsphere scaffolds and analyzed after 24 h of incubation. The experimental groups included UV-assembled and non-assembled NFH-MS scaffolds, with or without E7 and/or BFP functionalization.

Following 24h of culture, BMSCs were fixed with 4 % paraformaldehyde, permeabilized with 0.1 % Triton X-100, and stained with phalloidin-633 to visualize F-actin. Nuclei were counterstained with DAPI. Samples were imaged using confocal microscopy (Leica SP8), and cell spreading, morphology, and cytoskeletal organization were assessed across all groups. Quantitative analysis of cell spreading area was performed using ImageJ software.

To investigate early osteogenic signaling, immunofluorescence staining was conducted after 24 h of incubation. Samples were incubated with primary antibodies against RUNX2 and phosphorylated STAT3 (p-STAT3, Tyr705), followed by fluorescent secondary antibodies. Signal localization was visualized using confocal microscopy, and fluorescence intensity within the nuclear region was quantified using ImageJ. Co-localization analysis was performed to assess potential synergistic effects of E7 and BFP functionalization.

2.8. Assessment of osteogenic differentiation via ALP and ARS staining

Osteogenic activity of BMSCs on microsphere scaffolds was evaluated by alkaline phosphatase (ALP) and Alizarin Red S (ARS) staining. For ALP staining, cells were cultured in osteoinductive medium for 7 d. Samples were then fixed in 4 % paraformaldehyde and treated with a chromogenic substrate solution composed of nitro blue tetrazolium (NBT) and 5-bromo-4-chloro-3-indolyl phosphate (BCIP) in ALP buffer. After color development for 20 min at ambient temperature, the reaction was stopped by rinsing with deionized water. For ARS staining, after 14 d of osteogenic induction, constructs were fixed and incubated in ARS solution (adjusted to pH 4.2) for 30 min to visualize calcium deposition. Excess dye was removed by thorough washing. Stained constructs were photographed under a stereomicroscope. To quantify mineralization, ARS was eluted using 10 % cetylpyridinium chloride (CPC), and the absorbance at 560 nm was measured using a microplate reader (Cytation 5, BioTek).

2.9. Quantitative real-time PCR (qRT-PCR) analysis

To investigate the expression of osteogenic genes, total RNA was extracted from BMSCs cultured on scaffolds (with or without UV crosslinking) after 14 d of osteogenic induction. RNA isolation was performed using the RNeasy Mini Kit (Qiagen) following the manufacturer's instructions. First-strand cDNA was synthesized using the PrimeScript™ RT Reagent Kit (Takara Bio, Japan).

Quantitative PCR was carried out using TB Green® Premix Ex Taq™ II (Takara) on a CFX96™ Real-Time PCR Detection System (Bio-Rad, USA). GAPDH served as the housekeeping gene for normalization. Relative gene expression levels were calculated using the 2^−ΔΔCT method. The sequences of primers used for qRT-PCR are provided in Table S1 (Supporting Information).

2.10. In vivo study

All animal procedures were approved by the Institutional Animal Care and Use Committee (Protocol# IACUC 2020-0284-CD). Twenty Sprague-Dawley (SD) rats (180–200 g, Charles River) were equally divided into four groups: Blank (no scaffold), MS (UV-assembled), MS-BFP-E7 (No UV), and MS-BFP-E7 (UV).

Rats were anesthetized with isoflurane inhalation followed by an intraperitoneal injection of ketamine/xylazine. After shaving the mandibular area, the skin was disinfected sequentially with iodine and 70 % ethanol. A 15 mm incision was made to expose the masseter muscle. A secondary incision was then made through the muscle below the inferior ligament line to access the mandibular buccal plate. A standardized fenestration defect (3 mm × 2 mm × 1 mm) was created distal to the mandibular first molar using a #4 round bur followed by a #1/2 bur to remove the cementum bilaterally. All tissues on the buccal side of the distal root, including the buccal cortical bone, PDL, and associated soft tissues, were completely removed, following our previously published protocol [19]. Sterile NFH-MS microspheres were injected into the defect cavity to fully fill the defect volume in a conformal manner, without intentional compression, and allowed to assemble in situ in groups receiving UV crosslinking. For groups receiving in situ UV crosslinking, microspheres were photo-crosslinked (0.06 W/cm², 365 nm). Following scaffold placement, the masseter muscle was repositioned and sutured with absorbable sutures, and the skin was closed using interrupted sutures. Animals were maintained on a soft diet postoperatively and monitored daily. All animals were euthanized at 8 weeks post-surgery for further analysis. Five animals were initially assigned to each group. Due to perioperative loss of one animal in a group, four animals per group were included in the final micro-CT and immunohistochemical analyses.

2.11. Micro-CT, histology, and immunohistochemistry

Mandibles were scanned using a μCT35 system (Scanco Medical, Switzerland) to evaluate bone regeneration within the defect area. Scans were acquired at an X-ray energy of 70 kV and a current of 114 μA, with an integration time of 600 ms and an isotropic voxel size of 10 μm. Bone mineral density (BMD) measurements were calibrated using a hydroxyapatite phantom (maximum density 1200 mg HA/cm³) scanned under identical conditions. A standardized 3D volume of interest (VOI), corresponding to the surgically created fenestration defect (3 mm × 2 mm × 1 mm), was defined based on reproducible anatomical landmarks relative to the distal root of the mandibular first molar. Tooth and root structures were excluded from the VOI by manual contouring. Parameters including bone volume to total volume ratio (BV/TV), bone mineral density (BMD), and new bone height were quantified using Scanco software. After imaging, samples were fixed in 10 % neutral-buffered formalin, decalcified, dehydrated, and embedded in paraffin. Serial sections (5 μm thickness) were prepared for histological and immunohistochemical analyses. Hematoxylin and eosin (H&E) staining was performed to assess overall tissue morphology and defect bridging. Masson's trichrome staining was used to evaluate collagen deposition and new bone matrix formation.

Immunohistochemistry (IHC) was conducted to detect osteogenic markers including specificity protein 7 (SP7), runt-related transcription factor 2 (RUNX2), and phosphorylated STAT3 (p-STAT3). After deparaffinization and antigen retrieval, tissue sections were incubated with primary antibodies followed by detection using the Vectastain Elite ABC-HRP kit (Vector Laboratories, Burlingame, CA) according to the manufacturer's protocol. Images were captured and quantified using ImageJ software.

2.12. Statistical analysis

All quantitative data were expressed as mean ± standard deviation (SD). One-way analysis of variance (ANOVA) with post hoc Tukey's test was used for multiple group comparisons. For pairwise comparisons, unpaired Student's t-test was applied. A p-value <0.05 was considered statistically significant.

3. Results

3.1. Characterization of UV-assembled NFH-MS scaffolds

To enable the assembly of NFH-MS into 3D macroporous scaffolds, we first modified gelatin with methacrylate groups to produce gelatin methacrylate (GelMA), introducing double bonds necessary for subsequent UV crosslinking. NFH-MS were fabricated from GelMA using an approach that combines oil-in-water-in-oil (O/W/O) double emulsification with thermally induced phase separation, as described in our recent work [20]. As shown in Fig. 1A–C, the resulting NFH-MS featured a nanofibrous shell with a thickness of approximately 8–15 μm, and the nanofibrous structure closely mimics the architecture of natural collagen fibers found in bone extracellular matrix (ECM). Confocal imaging revealed a hollow core within the shell of the NFH-MS (Fig. 1D–F), which served as a reservoir for bioactive molecules to enable controlled release in subsequent experiments. Following fabrication, the NFH-MS were injected into a designated site and assembled into a 3D scaffold via UV crosslinking (Fig. 1G and H). The assembled structure maintained inter-spherical connectivity while preserving the integrity of individual microspheres. For instance, when injected around a rat tooth, the NFH-MS formed a porous scaffold and readily filled the space around the tooth (Fig. 1I). This high shape fidelity suggests strong potential for adapting to the irregular contours of periodontal defects. The interconnected microsphere network created a continuous yet porous architecture, as confirmed by 3D confocal reconstruction (Fig. 1J and K), which is expected to facilitate cell infiltration and nutrient diffusion. Cross-sectional imaging further verified that the hollow internal cavities remained intact post-assembly, along with abundant inter-microsphere spacing (Fig. 1L). These findings demonstrate that NFH-MS assembly preserves structural integrity at both intra- and inter-microsphere levels, establishing a robust physical framework for subsequent biological functionalization.

Fig. 1.

Morphological characterization of NFH-MS and the assembled NFH-MS scaffold. (A–C) SEM images of NFH-MS, revealing the uniform spherical morphology with a distinct nanofibrous surface architecture. (D–F) Confocal images of FITC-labeled NFH-MS, displaying consistent hollow spherical structures. (G–H) SEM images of the assembled NFH-MS, showing interconnected microspheres with macropores between them. (I) Photograph of a rat molar embedded in a cylindrical mold filled with the assembled NFH-MS, demonstrating the assembled NFH-MS can readily fill the space around the tooth and form a continuous porous scaffold. (J–K) 3D confocal reconstructions of the assembled NFH-MS scaffold, highlighting the interconnected microsphere network. (L) 2D cross-sectional confocal image, indicating that internal cavities remained intact following assembly.

Next, the E7 peptide was conjugated onto the surface of NFH-MS using a sulfo-SMCC-assisted coupling reaction. To visualize the conjugation, NFH-MS were labeled with fluorescein isothiocyanate (FITC, green), and E7 was labeled with rhodamine (red). As shown in Fig. 2A–C, a bright red ring encircles each green microsphere, indicating that E7 was uniformly distributed across the surface of the NFH-MS.

Fig. 2.

Incorporation of functional E7 and BFP into NFH-MS and characterizations of the multifunctional NFH-MS scaffold. (A–C) Confocal images of rhodamine-labeled E7 (red) conjugated onto the NFH-MS (green). (D–F) Confocal images of rhodamine-labeled BFP (red) encapsulated into the NFH-MS (green). (G) Cumulative release profiles of BFP from the non-assembled NFH-MS (No UV) and UV-assembled NFH-MS. (H–I) Mechanical strengths of assembled NFH-MS scaffolds under the conditions of different UV crosslinking time and methacrylate content in the GelMA precursor. (J) In vitro degradation (percent mass remaining) of non-assembled NFH-MS (No UV) and UV-assembled NFH-MS scaffolds. n = 3, ∗∗p < 0.01.

In addition, BFP was encapsulated within NFH-MS during the thermally induced phase separation process. As shown in Fig. 2D–F, the majority of the rhodamine-labeled BFP (red) was localized within the hollow core of the NFH-MS (green), while a smaller fraction was embedded in the surrounding nanofibrous shell. Thanks to the hollow architecture of the NFH-MS, an encapsulation efficiency of up to 69 % was achieved, as reported in our recent publication [20]. Release profile analysis revealed a sustained release of BFP from both the NFH-MS and the assembled NFH-MS scaffold over a period of 14 days. Notably, the assembled NFH-MS exhibited a slower release rate compared to individual NFH-MS, particularly during the first three days, followed by a slow release to plateau at later time points (Fig. 2G). This suggests that the assembled scaffold offers enhanced control over peptide release, potentially improving temporal regulation of osteogenic stimulation.

The mechanical strength of the assembled NFH-MS scaffold could be effectively tuned by adjusting the UV exposure time and methacrylate concentration during GelMA synthesis. As UV exposure time increased from 30 s to 5 min, the compressive modulus of the scaffold rose from 6.28 ± 0.92 kPa to 12.43 ± 2.62 kPa (Fig. 2H), demonstrating a time-dependent enhancement in mechanical strength. Additionally, increasing the content of methacrylate in the GelMA precursor from 5 ml to 10 ml significantly boosted the compressive modulus from 4.93 ± 0.23 kPa to 13.52 ± 3.01 kPa (Fig. 2I). Accordingly, we selected 1 min of UV exposure and l0 ml of methacrylate for GelMA synthesis in all subsequent experiments. The assembled NFH-MS scaffold exhibited a markedly slower degradation rate compared to NFH-MS alone. As shown in Fig. 2J and 67.3 % of the assembled scaffold's weight remained after 25 days, whereas only 42.6 % of the standalone NFH-MS were retained over the same period.

3.2. Assembled NFH-MS scaffolds promote BMSC migration

To assess whether the UV-assembled NFH-MS scaffold supports 3D cell infiltration, we monitored the migration of BMSCs over time and compared their behavior to that in conventional gelatin hydrogels. Confocal imaging at 2, 6, and 12 h revealed a progressive, time-dependent increase in BMSC penetration within the microsphere-assembled scaffolds (Fig. 3A–I). By 12 h, cells had migrated 272.0 ± 9.1 μm into the NFH-MS scaffold—significantly deeper than in gelatin hydrogels (5.0 ± 1.3 μm, p < 0.05), where most cells remained confined to the surface (Fig. 3J). Remarkably, BMSCs on NFH-MS exhibited early actin protrusions as early as 2 h and demonstrated robust inter-microsphere bridging by 6 h, indicative of microstructural guidance and enhanced migratory capacity.

Fig. 3.

Time-dependent migration of BMSCs within the UV-assembled NFH-MS scaffolds compared to gelatin hydrogels. Images are organized by time points: 2 h, 6 h, and 12 h (top to bottom). (A, D, G): Overlay images showing FITC-labeled NFH-MS (green) with phalloidin-stained BMSCs (red). (B, E, H): Corresponding red-channel-only images from (A, D, G), highlighting the migration front marked by a white dashed line (yellow arrow indicates depth). (C, F, I): Red-channel-only images of gelatin hydrogel controls, showing the migration front and depth markings. (J) Quantitative analysis of cell migration depth over time. (K) Cell number quantification at two depths beneath the scaffold surface: 0–50 μm and 50–100 μm. n = 3, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Quantitative dynamic analysis revealed a progressive increase in cell migration depth within the assembled NFH-MS scaffold, rising from 30.8 ± 3.3 μm at 2 h to 272.0 ± 9.1 μm at 12 h. This was accompanied by a significant increase in cell density across both shallow (0–50 μm) and deeper (50–100 μm) regions over time (Fig. 3K), with the 50–100 μm zone exhibiting a 2.3-fold increase between 6 and 12 h. In contrast, cells displayed minimal infiltration in the gelatin hydrogel. Collectively, these findings demonstrate that the assembled NFH-MS scaffold offers interconnected and macroporous channels that facilitate directional cell migration and rapid 3D infiltration that is an essential prerequisite for accelerated tissue regeneration.

In addition, the biocompatibility of the UV-assembled NFH-MS scaffold was evaluated. Live/dead cell staining revealed minimal cell death in both the UV-assembled and non-UV-assembled groups after up to five days of culture (Supporting Information, Fig. S1). Furthermore, the number of BMSCs increased steadily over time in both groups, with no significant difference observed between them. These findings indicate that the UV-assembly process did not impair BMSC adhesion or proliferation.

3.3. Assembled NFH-MS scaffolds enhance intercellular coupling and junction-mediated communication and activate downstream signaling

The dye diffusion method was initially employed to assess intercellular dye transfer as functional readout associate with cell–cell communication. As shown in Fig. 4A, Lucifer Yellow dye exhibited extensive intercellular transfer among BMSCs cultured on UV-assembled NFH-MS scaffolds, with an average spread of 218.25 ± 6.68 μm at 24 h. In contrast, minimal dye transfer was observed in cells on the non-UV control group. Connexin 43 (CX43) and N-cadherin (NCAD), markers of gap junctions and adherens junctions, respectively, were used to further characterize intercellular communication. Immunofluorescence staining revealed markedly elevated expression of both CX43 and NCAD in the UV-assembled group compared to the control (No UV) (Fig. 4B). Quantitative analysis confirmed that fluorescence intensities of both markers were several-fold higher in the UV-assembled group, indicating significantly enhanced cell–cell communication.

Fig. 4.

UV-assembled NFH-MS scaffolds enhanced cell–cell communication at the junctional level and associated downstream signaling pathways. (A1–A2) Fluorescence images showing Lucifer Yellow (yellowish green) diffusion in BMSCs cultured on UV-assembled and non-UV-assembled NFH-MS after 24 h (A3) Quantitative comparison of dye diffusion distance between the two groups. (B1–B6) Immunofluorescence images displaying elevated expression of Connexin 43 (CX43, yellow) and N-cadherin (NCAD, red) in BMSCs on UV-assembled versus non-UV-assembled NFH-MS after 24 h (B7–B8) Quantitative analysis of fluorescence intensity for CX43 and NCAD between the two groups. (C1–C6) Immunofluorescence images showing phosphorylated STAT3 (p-STAT3, red) expression in BMSCs cultured on UV-assembled and non-UV-assembled NFH-MS after 24 h (C7) Quantitative analysis of p-STAT3 fluorescence intensity between the two groups. n = 3, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

In parallel, phosphorylated STAT3 (p-STAT3), a downstream effector of CX43-mediated signaling, was substantially upregulated in BMSCs on the UV-assembled NFH-MS scaffolds, whereas only weak signals were detected in the control group (Fig. 4C). Quantitative data further demonstrated that p-STAT3 expression in the UV-assembled group was approximately 10-fold higher than in the control, underscoring the activation of intracellular signaling cascades in response to improved intercellular connectivity.

The temporal dynamics of intercellular connectivity among BMSCs cultured on UV-assembled and non-UV-assembled NFH-MS were further investigated. As shown in Fig. 5, CX43 expression in the UV-assembled group increased markedly from 6 h to 24 h, indicating progressive enhancement of gap junction formation over time. In contrast, CX43 expression remained minimal throughout the same period in the non-UV-assembled group. These findings collectively demonstrate that UV-assembled NFH-MS scaffolds promote dynamic and time-dependent intercellular communication.

Fig. 5.

Time-dependent expression of CX43 in BMSCs cultured on UV-assembled and non-UV-assembled NFH-MS scaffolds. (A–C, G–I) Immunofluorescence images of BMSCs cultured on UV-assembled NFH-MS for 6 h (A–C) and 24 h (G–I), respectively, showing a marked increase in CX43 expression over time. (D–F, J–L) Immunofluorescence images of BMSCs cultured on non-UV-assembled NFH-MS for 6 h (D–F) and 24 h (J–L), exhibiting minimal CX43 signal and limited intercellular contact.

Western blot analysis was conducted to further investigate the impact of UV assembly on intercellular and intracellular communication within the NFH-MS scaffolds. As shown in Fig. 6A–C, UV-assembled NFH-MS significantly upregulated the expression of NCAD and CX43 by 2.7-fold and 1.6-fold, respectively, compared to the non-UV-assembled group. Additionally, the expression of the inhibitory phosphorylated form of CX43 (p-CX43) was reduced by 2.8-fold in the UV-assembled group (Fig. 6D). Notably, the p-CX43/CX43 ratio in the UV-assembled group was 6.8-fold lower than that in the non-UV-assembled group (Fig. 6E), indicating a substantially higher proportion of functional CX43 available for gap junction formation.

Fig. 6.

Western blot analysis of intercellular adhesion and STAT3 signaling in BMSCs cultured on UV-assembled and non-UV-assembled NFH-MS. (A) Western blot images showing the expression levels of NCAD, phosphorylated CX43 at Ser368 (p-CX43), total CX43, phosphorylated STAT3 at Tyr705 (p-STAT3), and total STAT3. (B–G) Quantitative analysis of protein expression levels: (B) NCAD, (C) CX43, (D) p-CX43, (E) p-CX43/CX43 ratio, (F) p-STAT3, and (G) total STAT3. β-actin was used as an internal reference for normalization. n = 3, ∗p < 0.05, ∗∗p < 0.01.

Analysis of downstream intracellular signaling revealed a 1.9-fold increase in phosphorylated STAT3 (Tyr705) in the UV-assembled group (Fig. 6F), accompanied by a 2.9-fold elevation in total STAT3 expression (Fig. 6G). These findings suggest activation of the intracellular CX43– STAT3 signaling pathway, which is closely linked to intercellular communication.

3.4. Assembled NFH-MS scaffolds combined with E7 and BFP synergistically enhance BMSC adhesion and osteogenic differentiation

We first investigated the synergistic effects of assembled NFH-MS scaffolds combined with E7 peptide on BMSC adhesion and cytoskeletal organization. As shown in Fig. 7A–J, conjugation of E7 to the microspheres significantly enhanced BMSC adhesion on both UV-assembled and non-UV-assembled NFH-MS after 24 h. Notably, cell attachment was greater on the UV-assembled NFH-MS compared to the non-UV-assembled group. Quantitative analysis revealed a 3.5-fold increase in cell number on the UV-assembled NFH-MS and a 6.3-fold increase on the E7-conjugated UV-assembled NFH-MS, highlighting the synergistic effect of UV assembly and E7 functionalization in promoting BMSC adhesion (Fig. 7P). Additionally, cells cultured on E7-conjugated NFH-MS displayed well-spread morphologies, whereas those on the NFH-MS without E7 remained isolated or clustered, exhibiting poor cytoskeletal organization. Quantitative analysis showed that the average cell spreading areas were 101.8 ± 18.0 μm² for the non-UV-assembled group, 172.9 ± 13.4 μm² for the UV-assembled group, and 477.7 ± 16.6 μm² for the E7-conjugated UV-assembled NFH-MS group (Fig. 7Q).

Fig. 7.

Effects of UV assembly, E7 conjugation, and BFP loading on BMSC adhesion, cytoskeletal organization, and early osteogenic signaling. (A–E) Confocal images of BMSCs cultured for 24 h on E7-conjugated NFH-MS: UV-assembled (A–C) and non-UV-assembled (D–E). (F–J) Confocal images of BMSCs cultured for 24 h on NFH-MS without E7 conjugation: UV-assembled (F–H) and non-UV-assembled (I–J). (K–O) Immunofluorescence staining of RUNX2 (yellow) and phosphorylated STAT3 (p-STAT3, red) in BMSCs cultured on BFP-loaded NFH-MS scaffolds with E7: UV-assembled (K–M) and non-UV-assembled (N–O). (P) Quantification of BMSC adhesion across four groups (No UV, UV, E7+No UV, and E7+UV) after 24 h. (Q) Quantification of cell spreading area in the same four groups. (R) Pearson correlation analysis of nuclear fluorescence intensity between RUNX2 and p-STAT3 in UV-assembled and non-UV-assembled groups. (S–T) Quantitative analysis of nuclear fluorescence intensity for p-STAT3 (S) and RUNX2 (T) in UV-assembled versus non-UV-assembled groups that were loaded with E7 and BFP. n = 3, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Next, we investigated the synergistic effects of UV-assembled NFH-MS scaffolds combined with E7 and BFP on BMSC osteogenic signaling. As shown in Fig. 7K–O, intense nuclear co-localization of p-STAT3 and RUNX2 fluorescence signals was observed in BMSCs cultured on UV-assembled, BFP-E7-loaded scaffolds. In contrast, only weak signals of p-STAT3 and RUNX2 were detected in the control group. Quantitative analysis revealed that the nuclear fluorescence intensity of p-STAT3 and RUNX2 in the UV-assembled, BFP-E7-loaded group was 21.9-fold and 2.6-fold higher, respectively, compared to the non-UV-assembled, BFP-E7-loaded group (Fig. 7S and T). Moreover, a strong positive correlation between p-STAT3 and RUNX2 expression was identified, indicating activation of the STAT3–RUNX2 signaling axis by UV assembly (Fig. 7R).

Alkaline phosphatase (ALP) activity and calcium deposition were assessed to evaluate the synergistic effects of UV-assembled NFH-MS scaffolds combined with E7 and BFP on BMSC osteogenic differentiation and mineralization at days 7 and 14, respectively. As shown in Fig. 8A–B, BMSCs cultured in the BFP + E7+UV group exhibited the most intense ALP and Alizarin Red S (ARS) staining among all experimental groups. Specifically, ALP levels were 19.1 ± 1.4 ng/ml in the BFP + E7+UV group, compared to 10.8 ± 1.2 ng/ml in the No UV group and 8.2 ± 1.5 ng/ml in the UV-only group (Fig. 8C). Similarly, ARS absorbance at 560 nm (indicative of calcium deposition) was highest in the BFP + E7+UV group (0.57 ± 0.05), compared to 0.15 ± 0.03 in the No UV group and 0.21 ± 0.01 in the UV-only group (Fig. 8D), confirming enhanced mineralization.

Fig. 8.

UV-assembly combined with BFP and E7 synergistically enhanced osteogenic differentiation and mineralization of BMSCs in vitro. (A, B) Representative images of ALP staining on day 7 (A) and Alizarin Red S (ARS) staining on day 14 (B) for BMSCs cultured under different conditions. (C, D) Quantitative analysis of ALP activity (C) and ARS-based calcium deposition (D) derived from the staining images in (A, B). (E–G) Real-time PCR quantification of osteogenic gene expression on day 14, including ALP (E), RUNX2 (F), and OCN (G). n = 3, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Real-time PCR analysis further demonstrated that the BFP + E7+UV group showed the most robust upregulation of key osteogenic markers, including an 11.6-fold increase in ALP, 9.7-fold in RUNX2, and 15.6-fold in osteocalcin (OCN), relative to the MS-only control (Fig. 8E–G). Notably, all UV-assembled groups exhibited higher gene expression levels than their non-UV-assembled counterparts, underscoring the role of UV assembly in synergistically enhancing BMSC differentiation and mineralization.

3.5. Multifunctional assembled NFH-MS significantly enhanced alveolar bone regeneration in a rat fenestration defect model

To assess the in vivo therapeutic efficacy of the multifunctional NFH-MS scaffold, a rat fenestration defect model was established, and four treatment groups were compared: Blank (no NFH-MS), UV-assembled NFH-MS (UV group), NFH-MS combined with E7 and BFP without UV assembly (E7+BFP + No UV), and UV-assembled NFH-MS combined with E7 and BFP (E7+BFP + UV). Micro-CT imaging conducted eight weeks post-surgery revealed marked differences among the groups (Fig. 9A–D). The E7+BFP + UV group demonstrated the most extensive new bone formation, characterized by continuous trabecular bridging across the defect site. In contrast, the Blank and UV groups exhibited minimal bone regeneration with clearly defined defect margins, while the E7+BFP + No UV group showed moderate improvement. Quantitative μ-CT analysis confirmed these observations: the E7+BFP + UV group achieved the highest bone volume fraction (BV/TV: 86 ± 6.1 %), bone mineral density (BMD: 700.4 ± 33.4 mg HA/cm³), and new bone height (3.43 ± 0.05 mm), significantly outperforming the E7+BFP + No UV group (BV/TV: 51.75 ± 8.1 %, BMD: 499.5 ± 63.2 mg HA/cm³, height: 2.30 ± 0.07 mm) and both control groups (Fig. 9M–O).

Fig. 9.

In vivo evaluation of alveolar bone regeneration using multifunctional NFH-MS in a rat fenestration defect model. (A–D) Representative 3D reconstructions and sagittal μ-CT images of alveolar bone defects at 8 weeks post-surgery across four groups: Blank (no NFH-MS), UV (UV-assembled NFH-MS), E7+BFP + No UV (NFH-MS combined with E7 and BFP without UV assembly), and E7+BFP + UV (UV-assembled NFH-MS combined with E7 and BFP). (E–H) Hematoxylin and eosin (H&E)-stained histological sections showing tissue morphology in each group, where residual microspheres were marked with white arrows. (I–L) Masson's trichrome-stained sections illustrating collagen deposition and bone matrix formation. In these images, blue-stained regions indicate collagen-rich matrix and osteoid, while native bone is identified based on its compact morphology and anatomical continuity. (M − O) Quantitative analysis of bone volume fraction (BV/TV, M), bone mineral density (BMD, N), and new bone height (O) derived from μ-CT data. White dashed lines in panels (A–L) indicate defect boundaries. n = 4, ∗∗p < 0.01, ∗∗∗p < 0.001.

Histological analysis further supported these findings. H&E-stained sections revealed densely packed, well-organized mineralized tissue in the E7+BFP + UV group, whereas the Blank and UV groups showed fibrous tissue infiltration and incomplete defect closure (Fig. 9E–H). Masson's trichrome staining highlighted robust collagen deposition and abundant newly formed bone matrix (blue-stained) in the E7+BFP + No UV group, indicating enhanced osteogenesis and seamless integration with surrounding native tissue (Fig. 9I–L).

To further investigate osteogenic activation within the regenerated alveolar bone, immunohistochemical (IHC) staining was performed at 8 weeks post-implantation to assess the expression of key transcription factors—SP7 and RUNX2—as well as phosphorylated STAT3 (p-STAT3), a mechanosensitive signaling effector. Representative IHC images and corresponding quantitative analyses revealed significant differences among the treatment groups (Fig. 10). In the E7+BFP + UV group, SP7 expression was prominently localized along the bone–scaffold interface, particularly near the leading edge of newly formed bone (Fig. 10D–D1). In contrast, SP7 signals were sparse and confined in the UV and Blank groups (Fig. 10A–B1). Quantitative analysis showed that the SP7-positive area reached 2.6 ± 0.1 % in the E7+BFP + UV group, significantly higher than in the E7+BFP + No UV group (1.8 ± 0.1 %) and both controls (UV: 0.6 ± 0.02 %, Blank: 0.4 ± 0.1 %) (Fig. 10M).

Fig. 10.

Immunohistochemical evaluation of osteogenic transcription factors SP7 and RUNX2 and signaling marker p-STAT3 in regenerated alveolar bone tissue. (A–D) Representative low- and high-magnification IHC images showing SP7 expression at 8 weeks post-implantation across the four treatment groups. (E–H) Representative low- and high-magnification IHC images illustrating RUNX2 expression at 8 weeks post-surgery in the same groups. (I–L) Representative low- and high-magnification IHC images depicting p-STAT3 expression at 8 weeks post-implantation. (M − O) Quantitative analysis of SP7 positive area (M), RUNX2 positive area (N), and p-STAT3 positive area (O) across the four groups. n = 4, ∗∗∗p < 0.001.

A similar trend was observed for RUNX2 expression. The E7+BFP + UV group exhibited robust RUNX2 staining (Fig. 10H–H1), with positive cells distributed along the scaffold–bone interface, indicating active osteoblast differentiation. The RUNX2-positive area was quantified at 3.2 ± 0.1 %, significantly exceeding values in the E7+BFP + No UV group (2.2 ± 0.2 %) and controls (UV: 0.9 ± 0.1 %, Blank: 0.8 ± 0.1 %) (Fig. 10N).

Additionally, p-STAT3 expression was strongly detected in the nuclei of cells within the E7+BFP + UV group (Fig. 10L–L1), while minimal signal was observed in the Blank and UV groups. The p-STAT3 positive area reached 5.7 ± 0.3 % in the E7+BFP + UV group, significantly higher than in the E7+BFP + No UV group (1.4 ± 0.03 %) and controls (UV: 1.4 ± 0.1 %, Blank: 0.4 ± 0.1 %) (Fig. 10O). Collectively, these findings demonstrate that the multifunctional, UV-assembled NFH-MS scaffold markedly enhances new bone formation, likely through synergistic activation of SP7/RUNX2-mediated transcription and STAT3 signaling at the bone–material interface.

4. Discussion

Due to the irregular geometry of periodontal defects and the complex architecture of the periodontium, injectable biomaterials are desirable for periodontal tissue regeneration [23]. These materials can be delivered in a minimally invasive manner and readily conform to irregular defect shapes. However, currently available injectable biomaterials, including hydrogels and microspheres, exhibit limitations that hinder their clinical translation [7]. Hydrogels often suffer from low mechanical strength, slow cell migration, and poor tissue integration, whereas microspheres may disperse beyond the target area after administration [24]. Although some microsphere systems can be sintered to form a 3D porous scaffold, these constructs lose injectability and must be implanted surgically [25]. Moreover, most current microspheres fail to provide an ECM-like nanofibrous surface structure to facilitate cell growth and tissue regeneration [26]. In this work, we developed a unique approach to address the above challenges of traditional injectable systems. We introduced a UV-assembled nanofibrous hollow microsphere system that integrates structural architecture, biochemical signaling, and functional performance through a hierarchical design (Scheme 1). Upon UV-induced assembly, discrete NFH-MS undergo covalent crosslinking in situ to form a mechanically stable, macroporous scaffold. This NFH-MS scaffold works synergistically with bioactive peptides—E7, which promotes BMSC binding, and BFP, which enhances osteogenesis—to support BMSC adhesion, migration, differentiation, and biomineralization. Collectively, this multifunctional microsphere-based platform significantly accelerated alveolar bone regeneration in vivo.

To address the issue of NFH-MS dislocation following administration, we implemented a UV-assembly strategy that enables the microspheres to form a cohesive 3D macroporous scaffold in situ after injection. Traditional solid-walled microspheres, due to their high density and rigid structure, pose significant challenges for chemical crosslinking into a stable scaffold configuration [7]. In contrast, NFH-MS are composed entirely of nanofibers and possess a density less than one-thirtieth that of their solid counterparts [[27], [28], [29]], making them highly amenable to UV-induced crosslinking. Upon injection into the defect site, NFH-MS rapidly assemble into a structurally stable scaffold under UV exposure. Additionally, the mechanical properties of the scaffold can be finely tuned by adjusting the polymer concentration and UV-crosslinking duration (Fig. 2). Notably, the NFH-MS scaffold achieved a relatively high compressive modulus within 60 s of UV exposure, which is much faster than the most hydrogel gelling systems that usually take multiple minutes [30]. The fast fixation of the injectable NFH-MS scaffold meets clinical requirements by quickly forming a mechanically stable structure following administration. The enhanced stability of NFH-MS ensures that bioactive cues within the microspheres remain localized, allowing for sustained interaction with resident and recruited cells. The stabilized NFH-MS also facilitate the formation of a continuous tissue interface, promoting integration with surrounding native structures.

Besides stabilizing the NFH-MS after injection, UV-assembly has two additional functionalities. First, the UV-assembly process generated interconnected macropores within the NFH-MS scaffold (Fig. 1). Such macroporosity is essential for supporting cell infiltration, nutrient exchange, and waste removal. The resulting 3D macroporous architecture enables deep and uniform cell infiltration (Fig. 3). In parallel, UV-assembly duration regulates the stiffness of the NFH-MS scaffold, which in turn influences cell–material interactions and cytoskeletal organization. Higher matrix stiffness is known to enhance focal adhesion formation and promote more organized cytoskeletal structures [31]. Together, the macroporosity and tunable stiffness of the UV-assembled NFH-MS synergistically guide cell migration and adhesion and facilitate the formation of CX43-and N-cadherin–mediated intercellular junctions, thereby enhancing STAT3 phosphorylation and activating downstream osteogenic signaling pathways (Fig. 4, Fig. 5, Fig. 6).

Second, the UV-assembly process significantly enhanced intercellular communication that is a critical factor to regulate cell migration, proliferation, differentiation, and ECM production. The assembled NFH-MS scaffold facilitated robust cell–cell interactions by promoting the formation of gap junctions and adherens junctions. These junctions enable the exchange of signaling molecules and ions, thereby coordinating cellular responses and maintaining tissue homeostasis [21]. Enhanced intercellular communication within the scaffolding microenvironment activates key intracellular signaling pathways involved in osteogenesis and tissue regeneration. Notably, the UV-assembled NFH-MS scaffold has been shown to enhance the nuclear localization of STAT3, a mechanosensitive transcription factor that regulates cell proliferation and differentiation. Activation of STAT3 signaling contributes to the scaffold's ability to support early osteogenic commitment of BMSCs, as corroborated by previous studies [32].

In addition, the degradation behavior of the UV-assembled NFH-MS scaffold further supports its suitability as an early-stage regenerative matrix. Under artificial saliva conditions, approximately 67 % of the scaffold mass remained after 25 days, indicating sufficient temporal stability to sustain early cell adhesion, functional cell-cell coupling, and osteogenic commitment, while maintaining biodegradability. Further investigation under enzymatic or cell-mediated degradation conditions will be valuable to elucidate long-term scaffold remodeling and clearance behavior.

The multifunctionality of the UV-assembled NFH-MS scaffold was further enhanced through its integration with bioactive peptides E7 and BFP. The E7 peptide was conjugated to the microsphere surface to selectively promote BMSC adhesion via integrin-mediated interactions, thereby improving cell attachment and spreading on the UV-assembled scaffold (Fig. 7A–C). Meanwhile, BFP was encapsulated within the hollow core of the NFH-MS to provide sustained osteoinductive stimulation. When combined with the structural and signaling benefits conferred by UV-assembly, these peptides synergistically activated key transcriptional programs essential for bone formation. As illustrated in Fig. 7K–O, the nuclear co-localization of phosphorylated STAT3 (p-STAT3) and RUNX2 was most pronounced in the UV-assembled scaffold integrated with both E7 and BFP. RUNX2, a master regulator of osteoblast differentiation and bone matrix production, was significantly upregulated in this environment. The UV-assembled scaffold, enriched with targeted bioactive cues, created optimal conditions for RUNX2 expression, thereby driving BMSC commitment to the osteogenic lineage. This coordinated activation of STAT3 and RUNX2 pathways led to enhanced biomineralization and accelerated bone regeneration (Fig. 8).

In vivo studies using a rat fenestration defect model further validated the superior regenerative performance of the UV-assembled NFH-MS scaffold functionalized with E7 and BFP (Fig. 9). Compared to unassembled or peptide-free controls, the multifunctional NFH-MS scaffold achieved significantly higher bone volume, mineral density, and tissue integration. These results underscore the therapeutic potential of this bioinspired injectable platform for periodontal tissue engineering.

Despite the many advantages demonstrated in this study, two questions remain to be addressed in future study. First, although our findings support a scaffold-induced CX43–STAT3–RUNX2 association in osteogenesis, we interpret this as an associated mechanism rather than a confirmed causal pathway. In bone, CX43 is a central mediator of osteolineage gap-junction communication and mechanotransduction [33], and STAT3 is required for skeletal development and osteogenesis [32]. Studies in non-bone systems have shown that CX43 perturbation affects p-STAT3 expression in a complicated, tissue- and channel-state-dependent manner [34,35]. Further validation using pathway-specific inhibitors or genetic knockdown approaches will be essential to establish causality [36]. Second, the NFH-MS in this study exhibited limited antibacterial capacity, which is a critical consideration for periodontal applications where infection risk is high. The present work focuses on validating the core regenerative functions of NFH-MS, including in situ mechanical stabilization, enhanced intercellular communication, and osteogenic induction, which are mediated by UV-assisted assembly, E7-guided selective BMSC adhesion, and sustained BFP delivery. Incorporation of antibacterial modules into the NFH-MS in the current study would introduce additional design variables, such as dosing, release kinetics, and potential interference with cytocompatibility and osteogenesis, which makes this system too complex. Once the NFH-MS system with UV-assisted assembly, E7-guided selective BMSC adhesion, and sustained BFP release have been validated and optimized, the next step is the incorporation of antibacterial components into the NFH-MS. One straightforward strategy would be to encapsulate the antibacterial agent into the hollow core and/or nanofibers of NFH-MS for controlled release. Additionally, evaluating the scaffold's performance in infected or chronic animal models will be crucial for assessing its translational potential and clinical applicability in real-world scenarios.

In addition, future studies incorporating in vivo visualization of endogenous BMSC recruitment using lineage tracing, dynamic mineralization labeling, vascular analysis, and longitudinal micro-CT assessment will be important to capture the temporal and multi-tissue aspects of periodontal regeneration.

5. Conclusion

In this study, we developed an injectable nanofibrous hollow microsphere-based scaffold for alveolar bone regeneration. The multifunctional NFH-MS scaffold was engineered through a combination of UV-induced assembly, E7 peptide conjugation to the nanofibrous shell, and BFP encapsulation within the hollow core. The UV-assembly process imparted three key functionalities: enhanced mechanical stability, formation of interconnected macropores to facilitate cell infiltration, and promotion of intercellular communication. This assembly significantly improved CX43/NCAD-mediated junctions and facilitated nuclear localization of STAT3 and RUNX2, working synergistically with E7 and BFP to enhance BMSC adhesion, osteogenic differentiation, and biomineralization. As a result, the UV-assembled NFH-MS scaffold integrated with E7 and BFP demonstrated superior alveolar bone regeneration in a rat fenestration defect model. Collectively, our findings present a promising approach for alveolar bone regeneration.

CRediT authorship contribution statement

Yingzi Li: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation. Qian Li: Methodology. Zhaoming Deng: Formal analysis. Xiaohua Liu: Writing – review & editing, Writing – original draft, Supervision, Resources, Project administration, Investigation, Funding acquisition, Formal analysis, Conceptualization.

Ethics approval and consent to participate

All animal protocols were approved by the University Committee on Use and Care of Animals of University of Missouri (Protocol # 44035).

Declaration of competing interest

The authors declare no conflict of interests.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.