Anion-Dependent Layered Double Hydroxide-Reinforced Gelatin Methacryloyl Hydrogels for Osteoimmunomodulation and Bone Regeneration Authors: Zhong Wang, Mengle Shi, Hao Xu, Huanbin Huang, Jin Xing, Qinwei Gu, Siyu Wang, Jiajie Zheng Published: 2026 ABSTRACT PurposePersistent inflammation and insufficient osteogenesis jointly limit the repair of critical-sized bone defects. This study aimed to develop an anion-programmable osteoimmunomodulatory hydrogel based on magnesium-aluminum layered double hydroxide (LDH) nanoparticles and gelatin methacryloyl (GelMA), determine whether LDH interlayer anions regulate immune-skeletal crosstalk, and evaluate whether this strategy enhances repair of critical-sized calvarial bone defects.MethodsChloride- and nitrate-intercalated MgAl-LDH nanoparticles (LDH-Cl− and LDH-NO3−) were synthesized and characterized. Their cytocompatibility, cellular uptake, macrophage-polarizing effects, and osteogenic activity toward bone marrow mesenchymal stem cells (BMSCs) were evaluated in vitro. LDH nanoparticles were incorporated into photocrosslinked GelMA hydrogels and assessed in a murine critical-sized calvarial defect model. Transcriptomic sequencing was performed with three biological replicates per group, and cell-type-specific qRT-PCR was used to validate representative pathway-associated genes in vitro.ResultsThoroughly washed low- and high-magnification confocal images demonstrated FITC-LDH association/internalization in macrophages and BMSCs without apparent cytotoxicity. In lipopolysaccharide-activated macrophages, LDH nanoparticles reduced TNF-α and IL-1β expression while increasing TGF-β and IL-10 expression, with LDH-NO3− producing the strongest macrophage repolarization. LDH-NO3− also enhanced alkaline phosphatase activity, extracellular matrix mineralization, and osteogenic gene expression in BMSCs. After incorporation into GelMA hydrogels, LDH-NO3− promoted greater defect bridging, bone mineral density, collagen deposition, and osteocalcin expression in vivo. Transcriptomic analysis was consistent with increased PI3K/Akt-associated regenerative signaling and attenuation of NF-κB-associated inflammatory programs, and in vitro qRT-PCR showed concordant Akt1 and Rela transcriptional changes.ConclusionInterlayer anion engineering of LDH nanoparticles, particularly nitrate intercalation, provides a strategy for constructing immunoactive GelMA-based hydrogels that coordinate inflammation resolution and osteogenic repair in critical-sized bone defects. Introduction Critical-sized bone defects resulting from trauma, tumor resection, infection, or congenital malformation remain a major clinical challenge because spontaneous healing is insufficient and current grafting strategies (autografts/allografts or inert substitutes) often fail to restore both structural continuity and long-term function. In neurosurgical practice, calvarial defects arising after decompressive craniectomy, cranio-cerebral trauma, tumor extirpation, or infection represent a particularly demanding subset, because successful reconstruction must simultaneously re-establish mechanical protection of the brain, restore cranial contour, and support recovery of the local intracranial milieu, rather than merely cover the defect. Although cranioplasty using autologous bone or patient-specific alloplastic implants has become standard for skull reconstruction, these approaches are still frequently complicated by bone flap resorption, infection, implant exposure, poor tissue integration, or long-term foreign-body burden, highlighting the limitations of structurally restorative yet biologically passive materials. Moreover, cranial defect healing is shaped by a distinct regenerative niche involving the dura mater, pericranium, and overlying soft tissues, making calvarial repair biologically different from conventional load-bearing skeletal regeneration. Increasing evidence therefore indicates that successful regeneration in large defects is not limited by “filling space”, but by the inability to reconstruct a permissive regenerative microenvironment that synchronizes early inflammation resolution and osteoprogenitor recruitment/differentiation. Accordingly, bone biomaterials are increasingly expected to actively coordinate osteo–immuno coupling, rather than delivering a single osteoinductive signal.1–3 Bone healing is tightly governed by the immune system, especially macrophages that dynamically switch phenotypes across the repair timeline. While an initial pro-inflammatory phase is essential for debris clearance and host defense, persistent activation of inflammatory programs compromises osteogenesis, ultimately contributing to fibrotic encapsulation or non-union. Immunomodulatory design principles—tuning material chemistry, topography, mechanics, and degradation to program macrophage fate—have therefore become central in regenerative biomaterials research.4–6 Notably, osteoimmunomodulatory surfaces and scaffolds have been shown to bias macrophages toward reparative phenotypes, thereby enhancing mesenchymal stem cell (MSC) osteogenesis and improving osseointegration outcomes.2,5 Hydrogels are particularly attractive for engineering such dynamic niches because they are highly hydrated, extracellular matrix (ECM)-mimetic, and readily adaptable for minimally invasive delivery. Recent studies highlight how hydrogels can be engineered beyond passive carriers—integrating immunoregulation, controlled release, and microenvironment responsiveness to enhance therapeutic efficacy in regeneration.7,8 Among them, gelatin methacryloyl (GelMA) is widely used due to its cytocompatibility, photocrosslinkable injectability, and biochemical motifs that support cell adhesion and remodeling. However, GelMA alone typically lacks robust osteoinductive and immunoregulatory potency, motivating composite strategies that incorporate bioactive nanomaterials or ionic cues.8,9 Importantly, not only biochemical composition but also hydrogel crosslinking and mechanics can reshape macrophage behavior and downstream stromal responses, underscoring the need for materials that can actively guide immune–stromal communication rather than merely provide a 3D matrix.10 Bioactive 2D nanomaterials embedded in GelMA-based printed hydrogels can enhance osteogenesis through sustained ion release and cell-instructive surface chemistry.9 Layered double hydroxides (LDHs) are particularly intriguing because they combine tunable metal composition, a positively charged lamellar structure favorable for biomolecule adsorption and cellular interaction, and exchangeable interlayer anions that can modulate dissolution, interfacial chemistry, and biological responses. LDH-based biomaterials have already demonstrated osteogenic benefits in orthopedics, including LDH-modified bone cement that promotes osseointegration via multi-pathway osteogenic signaling.11 More recently, LDH derivatives have been advanced as bioinspired regulators of inflammatory/redox microenvironments to overcome hostile conditions that otherwise limit stem-cell-driven bone repair.12,13 Despite these advances, how interlayer anion identity in LDH nanoplates can be used to orchestrate osteo–immuno coupling within a hydrogel scaffold remains insufficiently defined—especially in critical-sized defects where immune dysregulation often compromises osteogenic repair. Here, we address this gap by preparing anion-dependent MgAl-LDH-GelMA composite hydrogels, in which LDH nanoplates bearing either chloride (LDH-Cl−) or nitrate (LDH-NO3−) interlayer anions are embedded within a photocrosslinked GelMA network. We hypothesized that changing the interlayer anion while maintaining the same MgAl-LDH metal framework and broadly comparable particle size and surface charge would differentially regulate macrophage polarization and downstream osteogenesis. This design is distinct from recent LDH-based hydrogel studies that primarily optimized nanoplatelet size, introduced additional bioactive cargo such as calcein, or used different polymer matrices.14–16 By isolating interlayer anion identity as the principal material variable, the present study directly tests whether anion chemistry itself can serve as a programmable regulator of osteoimmune coupling. We further used transcriptomics to identify signaling programs associated with the regenerative response. Recent oral and maxillofacial biomaterial studies using nano-hydroxyapatite-containing constructs also underscore the broader value of nanoscale inorganic cues for apatite formation and osteoblast compatibility.17,18 Experimental Section Materials Magnesium chloride hexahydrate (MgCl2·6H2O), aluminum chloride hexahydrate (AlCl3·6H2O), magnesium nitrate hexahydrate (Mg(NO3)2·6H2O), sodium hydroxide (NaOH), lipopolysaccharide (LPS), Triton X-100, β-glycerophosphate, dexamethasone, cetylpyridinium chloride, and Alizarin Red S were purchased from Sigma-Aldrich, Merck KGaA, Germany. Aluminum nitrate nonahydrate (Al(NO3)3·9H2O) and L-ascorbic acid were obtained from Shanghai Macklin Biochemical Co., Ltd., China. Gelatin methacryloyl (GelMA) was obtained from Suzhou Intelligent Manufacturing Research Institute/EFL, China. Irgacure 2959 was purchased from Tokyo Chemical Industry Co., Ltd., Japan. Phosphate-buffered saline (PBS), α-minimum essential medium (α-MEM), RPMI-1640 medium, fetal bovine serum (FBS), penicillin–streptomycin, fluorescein isothiocyanate (FITC), DAPI, Alexa Fluor 594-conjugated phalloidin, and fluorophore-conjugated secondary antibodies were supplied by Thermo Fisher Scientific, USA. Paraformaldehyde, EDTA decalcification solution, hematoxylin and eosin staining reagents, and Masson’s trichrome staining reagents were purchased from Servicebio Technology Co., Ltd., China. Cell Counting Kit-8 (CCK-8) was purchased from MedChemExpress, USA. Calcein-AM/propidium iodide live/dead staining reagents were obtained from Yeasen Biotechnology, China. Annexin V–FITC/PI apoptosis assay kit and flow cytometry antibodies against CD86, CD206, CD29, CD44, CD90, CD34, and CD45 were obtained from BioLegend, USA. Mouse ELISA kits for TNF-α, IL-1β, TGF-β, and IL-10 were purchased from R&D Systems, Bio-Techne, USA. Primary antibodies against osteocalcin (OCN), TNF-α, and IL-10 were obtained from Proteintech Group, USA. Synthesis of Anion-Defined MgAl-LDH Nanoplatelets Chloride- and nitrate-intercalated magnesium–aluminum layered double hydroxide nanoplatelets, denoted LDH–Cl− and LDH–NO3−, respectively, were prepared by alkaline coprecipitation followed by hydrothermal crystallization. For LDH–Cl−, MgCl2·6H2O (0.6099 g) and AlCl3·6H2O (0.2414 g) were dissolved in 10 mL of deionized water to generate a mixed metal precursor solution. For LDH–NO3−, Mg(NO3)2·6H2O (0.7692 g) and Al(NO3)3·9H2O (0.3751 g) were dissolved in the same volume of deionized water. In each synthesis, the metal precursor was added dropwise into 40 mL of freshly prepared NaOH solution containing 0.272 g NaOH under vigorous stirring. The alkaline medium was continuously protected under nitrogen flow to minimize carbonate contamination from atmospheric CO2. After complete addition, the milky suspension was maintained under stirring for an additional 30 min to allow nucleation and initial growth of LDH platelets. The precipitates were collected by centrifugation at 8500 rpm for 10 min, washed repeatedly with deionized water until the supernatant approached neutral pH, and then re-dispersed in deionized water. The dispersions were transferred into Teflon-lined stainless-steel autoclaves and hydrothermally aged at 100 °C for 16 h to improve crystallinity. The resulting LDH products were harvested by centrifugation at12000rpm for 15 min, washed thoroughly with deionized water, and finally dispersed in sterile deionized water to obtain stock suspensions for subsequent physicochemical and biological experiments. Physicochemical Characterization of LDH Nanoparticles The morphology and crystalline features of LDH–Cl− and LDH–NO3− were examined by transmission electron microscopy using a Talos F200X G2 transmission electron microscope (Thermo Fisher Scientific, USA). Diluted LDH dispersions were dropped onto carbon-coated copper grids and dried at room temperature before observation. Surface morphology was further characterized by field-emission scanning electron microscopy using a Regulus 8100 FE-SEM system (Hitachi High-Tech Corporation, Japan). For SEM imaging, dried nanoparticle powders were mounted on conductive adhesive tape and sputter-coated with a thin conductive layer when necessary. Hydrodynamic diameter, polydispersity, and ζ-potential were measured at 25 °C using a Litesizer 500 particle analyzer (Anton Paar, Austria). Prior to measurement, LDH stock dispersions were diluted in deionized water or PBS and equilibrated at room temperature. The surface chemical composition and successful incorporation of chloride or nitrate species were evaluated by X-ray photoelectron spectroscopy using an ESCALAB QXi XPS microprobe (Thermo Fisher Scientific, USA). Survey spectra were collected to identify Mg, Al, O, Cl, and N signals, and high-resolution scans were acquired for selected elements to verify the anion-dependent chemical signatures. Preparation and Structural Characterization of GelMA–LDH Composite Hydrogels GelMA was dissolved in sterile PBS at 37 °C to obtain a 5% (w/v) precursor solution. Irgacure 2959 was added at a final concentration of 0.5% (w/v) and fully dissolved under light-protected conditions. To prepare LDH-reinforced hydrogels, concentrated LDH–Cl− or LDH–NO3− stock suspensions were first sterilized and homogeneously dispersed by brief sonication. The LDH stock suspension was then mixed with the GelMA precursor to obtain a final LDH concentration of 1 mg mL−1 in the hydrogel precursor. The final formulations were defined as GelMA, GelMA–LDH–Cl−, and GelMA–LDH–NO3−, all containing 5% (w/v) GelMA and 0.5% (w/v) Irgacure 2959, with the two composite hydrogel groups containing 1 mg mL−1 LDH. For each hydrogel scaffold, 100 μL of precursor solution was transferred into a sterile cylindrical mold and photocrosslinked under 365 nm UV irradiation for 1 min. Thus, each 100 μL GelMA–LDH hydrogel contained 100 μg of LDH nanoparticles. For in vivo implantation, hydrogel constructs were prepared with a diameter matching the 5-mm calvarial defect and an appropriate thickness to fill the defect space. The same volume and geometry were used for GelMA, GelMA–LDH–Cl−, and GelMA–LDH–NO3− groups to ensure comparable scaffold loading among groups. Fluorescent Labeling and Cellular Internalization of LDH Nanoparticles To visualize cellular uptake, LDH–Cl− and LDH–NO3− were fluorescently labeled through adsorption-assisted incubation with FITC. Briefly, LDH suspensions were mixed with FITC solution in PBS and incubated overnight at room temperature in the dark. Unbound FITC was removed by repeated centrifugation and washing until negligible fluorescence was detected in the supernatant. The resulting FITC-labeled LDH nanoparticles were re-dispersed in sterile PBS and used immediately for cell uptake experiments. RAW264.7 murine macrophages (Procell Life Science and Technology Co., Ltd., Wuhan, China; Cat. No. CL-0190; passages 5–15) and mouse bone marrow-derived mesenchymal stem cells (BMSCs; Procell Life Science and Technology Co., Ltd., Wuhan, China; Cat. No. CP-M131; passages 2–5) were seeded in glass-bottom dishes and allowed to adhere overnight. Cells were incubated with FITC-LDH-Cl− or FITC-LDH-NO3− at 10 μg mL−1 for 12 h. Before imaging, cells were thoroughly washed three times with PBS to remove extracellular nanoparticles, fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and stained with Alexa Fluor 594-phalloidin for F-actin visualization. Nuclei were counterstained with DAPI. Low- and high-magnification fluorescence images were acquired using an FV3000 confocal laser scanning microscope (Olympus/Evident, Japan). The proportion of FITC-LDH-positive cells was quantified from randomly selected fields by dividing the number of cells with intracellular/cytoplasmic FITC-LDH signal by the total number of DAPI-positive cells. Three independent biological experiments were analyzed (n = 3). Cell Culture and Cytocompatibility Evaluation RAW264.7 macrophages (Procell, Cat. No. CL-0190; passages 5–15) were maintained in RPMI-1640 medium supplemented with 10% FBS and 1% penicillin-streptomycin. BMSCs (Procell, Cat. No. CP-M131; passages 2–5) were cultured in α-MEM containing 10% FBS and 1% penicillin-streptomycin. All cells were incubated at 37 °C in a humidified atmosphere containing 5% CO2, and culture media were replaced every 2 days. All quantitative in vitro experiments were performed using three independent biological replicates (n = 3). For cytocompatibility testing, RAW264.7 cells and BMSCs were seeded in 96-well plates at 5 × 103 cells per well and cultured overnight. Cells were then treated with LDH–Cl− or LDH–NO3− at 10 μg mL−1 for 72 h. Cell metabolic activity was quantified using CCK-8 reagent. After treatment, 10 μL of CCK-8 working solution was added to each well, followed by incubation at 37 °C for 2 h. Absorbance at 450 nm was measured using a SpectraMax iD5 multimode microplate reader (Molecular Devices, USA), and cell viability was calculated relative to untreated controls. Apoptosis and necrosis were analyzed using an Annexin V–FITC/PI apoptosis assay kit. After 72 h of LDH exposure, cells were detached, washed with cold PBS, resuspended in binding buffer, and stained according to the manufacturer’s instructions. Samples were analyzed using a BD FACSCelesta flow cytometer (BD Biosciences, USA). Live/dead staining was performed using Calcein-AM and propidium iodide. Following LDH treatment, cells were incubated with the staining solution for 20 min at 37 °C, washed gently with PBS, and imaged with a DMi8 inverted fluorescence microscope (Leica Microsystems, Germany). In vitro Macrophage Polarization Assay To establish an inflammatory macrophage model, RAW264.7 cells were seeded in 6-well plates and cultured until approximately 70% confluence. Cells were stimulated with LPS at 100 ng mL−1 for 24 h to induce an M1-like inflammatory phenotype. The LPS-containing medium was then removed, and cells were washed gently with PBS before being cultured in fresh medium containing LDH–Cl− or LDH–NO3− at 10 μg mL−1 for another 48 h. Macrophage phenotype was evaluated by transcriptional, secretory, and flow cytometric analyses. Total RNA was collected for quantitative real-time PCR analysis of M1-associated genes, including Tnf-α and Il-1β, and M2-associated genes, including Tgf-β and Il-10. Culture supernatants were collected, clarified by centrifugation, and stored at −80 °C before ELISA measurement of TNF-α, IL-1β, TGF-β, and IL-10. For flow cytometry, cells were harvested, blocked with staining buffer, and incubated with fluorophore-conjugated antibodies against CD86 and CD206. The relative proportions of CD86-positive and CD206-positive populations were quantified to assess the shift from a pro-inflammatory toward a pro-regenerative macrophage phenotype. Quantitative Real-Time PCR Total RNA from cultured cells and regenerated calvarial tissues was extracted using a FastPure Cell/Tissue Total RNA Isolation Kit (Vazyme Biotech, China), following the manufacturer’s protocol. RNA concentration and purity were determined using a NanoDrop One spectrophotometer (Thermo Fisher Scientific, USA). For reverse transcription, 1000 ng of total RNA was converted into cDNA using HiScript III RT SuperMix for qPCR (Vazyme Biotech, China). Quantitative real-time PCR was performed on a CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories, USA) using ChamQ Universal SYBR qPCR Master Mix (Vazyme Biotech, China). Relative gene expression was calculated using the 2−ΔΔCt method, with Gapdh serving as the internal reference gene. Primer sequences are provided in Table S1. For in vitro validation of the transcriptomic signatures, RAW264.7 cells were analyzed in Ctrl, LPS, LPS + LDH-Cl−, and LPS + LDH-NO3− groups for Rela expression, while BMSCs cultured in osteogenic induction medium (OM), OM + LDH-Cl−, or OM + LDH-NO3− were analyzed for Akt1 expression. These validation experiments were performed with n = 3 independent biological replicates per group. Immunofluorescence Staining For cellular immunofluorescence, cells grown on coverslips were fixed with 4% paraformaldehyde, permeabilized using 0.1% Triton X-100, and blocked with 5% normal donkey serum for 1 h at room temperature. Samples were then incubated with primary antibodies against OCN, TNF-α, or IL-10 overnight at 4 °C. After three washes with PBS, cells were incubated with appropriate fluorophore-conjugated secondary antibodies for 1 h at room temperature in the dark. Nuclei were counterstained with DAPI before imaging. For tissue immunofluorescence, paraffin-embedded calvarial sections were deparaffinized, rehydrated, subjected to antigen retrieval, and processed using the same blocking and antibody incubation procedures. Fluorescence images were acquired using an FV3000 confocal laser scanning microscope. Mean fluorescence intensity within the defect region was quantified using ImageJ software under identical thresholding conditions. ELISA Analysis Cell culture supernatants were collected after the indicated treatments and centrifuged at 1000 × g for 20 min to remove cell debris. The concentrations of TNF-α, IL-1β, TGF-β, and IL-10 were quantified using mouse-specific ELISA kits according to the manufacturer’s instructions. Absorbance was measured using a SpectraMax iD5 multimode microplate reader. Cytokine concentrations were calculated from standard curves generated with recombinant cytokine standards and normalized when appropriate. Characterization and Osteogenic Induction of BMSCs Before osteogenic experiments, BMSCs were immunophenotypically characterized by flow cytometry. Cells were stained with antibodies against CD29, CD44, CD90, CD34, and CD45. A typical mesenchymal phenotype was confirmed by high expression of CD29, CD44, and CD90, together with minimal expression of CD34 and CD45. For osteogenic induction, BMSCs were seeded in 24-well plates and cultured until they reached suitable confluence. The medium was then replaced with osteogenic induction medium consisting of α-MEM supplemented with 10% FBS, 50 μM L-ascorbic acid, 10 mM β-glycerophosphate, and 100 nM dexamethasone. LDH-Cl− or LDH-NO3− was added to the induction medium at 10 μg mL−1 where indicated. A non-osteogenic blank group was maintained in basal α-MEM containing 10% FBS and 1% penicillin-streptomycin without L-ascorbic acid, β-glycerophosphate, or dexamethasone. The blank group was included in ALP, Alizarin Red S, qRT-PCR, and OCN immunofluorescence analyses. The medium was refreshed every 2 days throughout the induction period. All quantitative assays were performed with n = 3 independent biological replicates. Early osteogenic differentiation was assessed after 7 days by alkaline phosphatase staining and quantitative ALP activity assay. For qualitative staining, cells were fixed and incubated with ALP staining solution according to the manufacturer’s protocol. For quantitative analysis, cell lysates were prepared, and ALP activity was measured colorimetrically and normalized to total protein content. Late-stage matrix mineralization was examined after 14 days using Alizarin Red S staining. Briefly, cells were fixed, stained with 2% Alizarin Red S solution, and washed thoroughly to remove nonspecific dye. Bound dye was extracted using 10% cetylpyridinium chloride, and absorbance was measured at 562 nm. In parallel, osteogenic gene expression, including Runx2, Opn, Bmp2, Alpl, and Ocn, was examined by qRT-PCR, and OCN protein expression was visualized by immunofluorescence staining. Animal Experiments and Critical-Sized Calvarial Defect Model All animal experiments were performed in accordance with institutional and national animal welfare guidelines and were approved by the Institutional Animal Care and Use Committee of Shanghai Pudong Hospital (approval No. KCZX20251128001). The animal anesthesia, postoperative monitoring, humane endpoint assessment, and euthanasia procedures were conducted under veterinary oversight and in accordance with the principles of the American Veterinary Medical Association (AVMA) Guidelines for the Euthanasia of Animals. Male BALB/c mice, 8 weeks of age, were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd., China, and acclimatized under specific-pathogen-free conditions before surgery. Animals were housed in a temperature- and humidity-controlled facility under a 12 h light/dark cycle with free access to standard chow and water. For calvarial defect surgery, mice were anesthetized with inhaled isoflurane delivered in oxygen. Anesthesia was induced in an induction chamber with 3–4% isoflurane and maintained at 1.5–2.5% isoflurane via a nose cone during the procedure. Adequate anesthetic depth was confirmed by loss of the pedal withdrawal reflex and continuously monitored by respiratory pattern and response to toe pinch. Ophthalmic ointment was applied to prevent corneal drying, and body temperature was maintained using a warming pad. Pre-emptive analgesia was provided with buprenorphine (0.05 mg kg−1, subcutaneous injection) before incision and every 12 h for 48 h after surgery; additional analgesia or supportive care was provided when signs of pain or distress were observed. After shaving and disinfection of the cranial region, a midline sagittal incision was made to expose the parietal bone. A circular full-thickness calvarial defect of 5 mm in diameter was generated using a sterile trephine bur under continuous sterile saline irrigation, with careful preservation of the underlying dura mater. The animals were randomly assigned to three treatment groups (n = 6 per group): GelMA, GelMA–LDH–Cl−, and GelMA–LDH–NO3−. The corresponding hydrogel constructs were implanted into the defect sites, and the periosteum and skin were closed in layers. Postoperatively, animals were allowed to recover in a warmed cage and were monitored until they regained normal ambulation. Mice were then checked at least once daily for general condition, wound healing, body weight, food and water intake, and signs of pain, infection, or distress. Humane endpoints included severe wound infection or dehiscence, persistent bleeding, inability to access food or water, marked lethargy, or body weight loss exceeding 20%, in which case animals were euthanized immediately according to the approved protocol. At 12 weeks after implantation, mice were euthanized before tissue harvest by gradual-fill carbon dioxide (CO2) inhalation using compressed CO2 delivered into a non-prefilled chamber at a displacement rate of 30–70% of the chamber volume per minute. CO2 exposure was continued for at least 1 min after respiratory arrest. Death was confirmed by the absence of respiration, heartbeat, corneal reflex, and response to firm toe pinch, followed by cervical dislocation as a secondary physical method to ensure death. Calvarial specimens containing the defect region and surrounding native bone were then harvested for micro-computed tomography, histological staining, immunofluorescence analysis, and molecular assays. Micro-Computed Tomography Analysis Harvested calvarial specimens were fixed in 4% paraformaldehyde and scanned using a Quantum GX3 microCT system (Revvity, USA). Three-dimensional reconstructions were generated from the acquired projection data using the manufacturer’s reconstruction software. A standardized cylindrical region of interest corresponding to the original 5-mm defect was selected for all specimens. Bone volume fraction (BV/TV) and bone mineral density (BMD) were quantified using identical thresholding parameters across all groups to enable unbiased comparison of newly formed mineralized tissue. Histological and Tissue Immunofluorescence Evaluation After microCT scanning, calvarial samples were decalcified in 10% EDTA solution at 4 °C with regular solution changes until complete decalcification was achieved. Specimens were dehydrated through graded ethanol, cleared, embedded in paraffin, and sectioned through the central region of the defect at approximately 6 μm thickness. Hematoxylin and eosin staining was performed to evaluate general tissue morphology, defect bridging, and the integration of newly formed tissue with host bone. Masson’s trichrome staining was used to visualize collagen deposition, osteoid formation, and maturation of regenerated bone matrix. For immunofluorescence evaluation, sections were deparaffinized, rehydrated, subjected to antigen retrieval, blocked with normal donkey serum, and incubated with primary antibodies against TNF-α, IL-10, or OCN overnight at 4 °C. After incubation with fluorophore-conjugated secondary antibodies, nuclei were counterstained with DAPI. Images were collected by confocal microscopy, and fluorescence intensity within the defect area was quantified using ImageJ. Transcriptomic Sequencing and Bioinformatic Analysis Transcriptomic sequencing was performed on regenerated calvarial defect tissues from the GelMA, GelMA-LDH-Cl−, and GelMA-LDH-NO3− groups (n = 3 independent biological replicates per group; nine samples in total). Total RNA concentration and purity were initially assessed by NanoDrop, and RNA integrity was evaluated using an Agilent 2100 Bioanalyzer; samples with RNA integrity number (RIN) ≥ 7.0 were used for library construction. Poly(A)-enriched RNA libraries were prepared according to the standard BGI eukaryotic transcriptome workflow and sequenced by BGI Genomics (Shenzhen, China) on a DNBSEQ-T7 platform with 150-bp paired-end reads (PE150). At least 30 million clean reads were obtained per sample, with ≥80% of bases reaching Q30. Adapter sequences and low-quality reads were removed using SOAPnuke (v2.1.9). Clean reads were aligned to the Mus musculus GRCm39 reference genome using HISAT2 (v2.2.1), and gene-level expression was quantified using RSEM (v1.3.3). Raw count data were analyzed with DESeq2 (v1.42.0), and differentially expressed genes were defined using an adjusted p value < 0.05 and |log2 fold change| ≥ 1. Principal component analysis (PCA) was performed on variance-stabilized expression values from all nine samples to evaluate sample-level clustering. The GelMA-LDH-NO3− versus GelMA comparison was used as the principal contrast for pathway-focused interpretation, while the GelMA-LDH-Cl− group was retained for comparative clustering and expression analyses. Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis, gene set enrichment analysis (GSEA), and protein-protein interaction (PPI) network analysis were subsequently performed to identify signaling programs associated with LDH-mediated osteoimmunomodulation. Statistical Analysis All quantitative data are presented as mean ± standard deviation. Unless otherwise stated, in vitro experiments were performed with n = 3 independent biological replicates, RNA-seq was performed with n = 3 biological replicates per group, and animal experiments used n = 6 mice per group. Statistical analysis was conducted using GraphPad Prism software (GraphPad Software, USA). Data normality was assessed using the Shapiro–Wilk test and homogeneity of variance was assessed using the Brown-Forsythe test before parametric analysis. Comparisons among multiple groups were performed using one-way analysis of variance followed by Tukey’s post hoc multiple-comparison test when the assumptions for parametric analysis were satisfied. A value of p < 0.05 was considered statistically significant. Statistical significance is denoted as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Results and Discussion Structural and Chemical Verification of Anion-Defined MgAl-LDH Nanoplatelets The successful construction of anion-regulated MgAl-LDH nanoplatforms was first examined by electron microscopy, colloidal characterization, and surface chemical analysis. As shown in Figure 1A and B, both LDH–Cl− and LDH–NO3− displayed a well-defined two-dimensional platelet-like morphology with a typical hexagonal outline. High-resolution TEM further revealed clear lattice fringes in both formulations, indicating that the hydrothermal aging process effectively promoted crystalline ordering within the LDH nanosheets (Figure 1C). The measured lattice spacings were approximately 0.11 nm for LDH–Cl− and 0.12 nm for LDH–NO3−, respectively, which supports the formation of crystallized inorganic layers rather than amorphous hydroxide precipitates.Figure 1Characterization of LDH-Cl− and LDH-NO3−. (A) TEM and (B) SEM images showing the platelet-like morphology of LDH nanoparticles. (C) High-resolution TEM images revealing crystalline lattice fringes. The red parallel lines in the images indicate the lattice fringes used for interplanar spacing measurements. (D) Hydrodynamic size distribution profiles determined by DLS. (E) ζ-Potential values of LDH dispersions. (F) XPS survey spectra confirming the elemental composition and anion-specific chemical signatures of LDH-Cl− and LDH-NO3−. DLS and ζ-potential measurements were repeated three times (n = 3 independent measurements).TEM, SEM, DLS and XPS analysis of LDH-Cl and LDH-NO subscript 3 nanoparticles.The image A shows TEM images of LDH-Cl and LDH-NO subscript 3, revealing platelet-like morphology. The image B shows SEM images of the same, highlighting their hexagonal outline. The image C shows high-resolution TEM images with lattice fringes, measuring 0.11 nm for LDH-Cl and 0.12 nm for LDH-NO subscript 3. The image D shows hydrodynamic size distribution profiles from DLS, with LDH-Cl at 131.6 nm and LDH-NO subscript 3 at 120.1 nm. The image E shows zeta-potential measurements, with LDH-Cl at 16.1 mV and LDH-NO subscript 3 at 16.6 mV. The image F shows XPS survey spectra, indicating elemental composition: LDH-Cl with Cl 2p, Al 2p and Mg 1s; LDH-NO subscript 3 with N 1s, O 1s, Al 2p and Mg 1s. Characterization of LDH-Cl− and LDH-NO3−. (A) TEM and (B) SEM images showing the platelet-like morphology of LDH nanoparticles. (C) High-resolution TEM images revealing crystalline lattice fringes. The red parallel lines in the images indicate the lattice fringes used for interplanar spacing measurements. (D) Hydrodynamic size distribution profiles determined by DLS. (E) ζ-Potential values of LDH dispersions. (F) XPS survey spectra confirming the elemental composition and anion-specific chemical signatures of LDH-Cl− and LDH-NO3−. DLS and ζ-potential measurements were repeated three times (n = 3 independent measurements). Dynamic light scattering analysis showed that LDH–Cl− and LDH–NO3− exhibited narrow and monomodal hydrodynamic size distributions centered at approximately 130 and 120 nm, respectively (Figure 1D). This nanoscale size range is advantageous for biological interactions because particles within this dimension can efficiently contact the cell membrane, undergo endocytic internalization, and distribute within hydrated polymeric matrices without severely compromising scaffold homogeneity. Importantly, the two LDH formulations exhibited similar hydrodynamic dimensions, minimizing the possibility that subsequent differences in cellular responses were simply caused by particle size effects. ζ-Potential analysis showed that both nanoplatelets were positively charged under aqueous conditions, with surface potentials of approximately +16.1 mV for LDH–Cl− and +16.6 mV for LDH–NO3− (Figure 1E). Such moderately cationic surfaces are characteristic of LDH materials and may facilitate electrostatic interactions with negatively charged cell membranes, extracellular matrix components, and serum proteins. X-ray photoelectron spectroscopy further confirmed the anion-specific chemical composition of the two nanomaterials (Figure 1F). The LDH–Cl− spectrum contained characteristic Mg, Al, O, and Cl signals, whereas LDH–NO3− displayed Mg, Al, O, and N signals without detectable Cl contribution. The selective appearance of Cl and N signals in the corresponding samples verifies that chloride and nitrate were successfully incorporated into the LDH interlayer region. Since the two LDH groups shared comparable morphology, colloidal size, and surface charge, the present system provides a rational platform for evaluating how interlayer anion chemistry regulates biological behavior. This point is important because LDHs are increasingly recognized not only as ion-releasing inorganic carriers, but also as chemically programmable two-dimensional nanomaterials in which metal composition, layer charge density, and interlayer species jointly determine cellular interactions and therapeutic performance.13 Previous work has shown that LDH-modified orthopedic materials can activate multiple osteogenic pathways and improve osseointegration, highlighting the relevance of this material family in bone repair.11 In the present study, the controlled comparison between LDH–Cl− and LDH–NO3− therefore enables a more focused interrogation of anion-dependent osteoimmunomodulation. LDH Nanoparticles Reprogram LPS-Activated Macrophages Toward a Pro-Regenerative Phenotype The interaction between LDH nanoparticles and macrophages was next investigated because macrophages constitute an early and dominant immune cell population after biomaterial implantation and critically influence whether the local microenvironment evolves toward chronic inflammation or constructive regeneration. After FITC-LDH exposure, cells were thoroughly washed three times with PBS before imaging. Figure 2A demonstrates intracellular/cytoplasmic-perinuclear FITC-LDH signals in RAW264.7 macrophages. Quantification of FITC-LDH-positive cells confirmed reproducible cellular uptake of both LDH formulations. F-actin staining showed preserved cytoskeletal organization, and no evident cell shrinkage, rounding, or membrane disruption was observed.Figure 2LDH nanoparticles attenuate inflammatory macrophage activation and promote a reparative macrophage phenotype. (A) Low- and high-magnification confocal images acquired after thorough PBS washing show intracellular/cytoplasmic FITC-labeled LDH signals (green) in RAW264.7 macrophages; F-actin and nuclei are labeled with phalloidin (red) and DAPI (blue). White dashed boxes in the low-magnification images indicate the regions shown at higher magnification in the adjacent panels. The proportion of FITC-LDH-positive cells was quantified from three independent experiments. (B–D) CCK-8, live/dead staining, and apoptosis analysis confirming cytocompatibility of LDH-Cl− and LDH-NO3−. (E) qRT-PCR analysis of M1-associated genes Tnf-α and Il-1β and M2-associated genes Tgf-β and Il-10 in LPS-activated macrophages after LDH treatment. (F) ELISA quantification of TNF-α, IL-1β, TGF-β, and IL-10 in culture supernatants. (G) Flow cytometric analysis of CD86 and CD206 expression. Quantitative in vitro data are presented as mean ± SD (n = 3 independent biological replicates). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.LDH-Cl/LDH-NO subscript 3 impact on macrophages: uptake, viability, cytokines, polarization.The figure (A to G) displays experimental results. Panel A shows confocal images of RAW264.7 macrophages with FITC-labeled LDH-Cl and LDH-NO subscript 3, indicating intracellular uptake, with LDH-NO subscript 3 slightly higher. Panel B presents OD450 values for cell viability, similar to control. Panel C shows live/dead staining, maintaining viability with LDH treatments. Panel D displays flow cytometry plots of FITC-Annexin V, indicating low apoptosis in LDH-treated cells. Panel E shows qRT-PCR results for Tnf-alpha, Il-1beta, Tgf-beta and Il-10, with LDH treatments reducing pro-inflammatory and increasing anti-inflammatory gene expression. Panel F presents ELISA data for cytokines, showing similar trends. Panel G shows flow cytometry plots of CD206 and CD86, indicating macrophage polarization, with LDH-NO subscript 3 increasing CD206 expression. Statistical significance is marked with asterisks: p < 0.05, p < 0.01, p < 0.001, p < 0.0001. LDH nanoparticles attenuate inflammatory macrophage activation and promote a reparative macrophage phenotype. (A) Low- and high-magnification confocal images acquired after thorough PBS washing show intracellular/cytoplasmic FITC-labeled LDH signals (green) in RAW264.7 macrophages; F-actin and nuclei are labeled with phalloidin (red) and DAPI (blue). White dashed boxes in the low-magnification images indicate the regions shown at higher magnification in the adjacent panels. The proportion of FITC-LDH-positive cells was quantified from three independent experiments. (B–D) CCK-8, live/dead staining, and apoptosis analysis confirming cytocompatibility of LDH-Cl− and LDH-NO3−. (E) qRT-PCR analysis of M1-associated genes Tnf-α and Il-1β and M2-associated genes Tgf-β and Il-10 in LPS-activated macrophages after LDH treatment. (F) ELISA quantification of TNF-α, IL-1β, TGF-β, and IL-10 in culture supernatants. (G) Flow cytometric analysis of CD86 and CD206 expression. Quantitative in vitro data are presented as mean ± SD (n = 3 independent biological replicates). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Cytocompatibility was further assessed using complementary viability and apoptosis assays. CCK-8 analysis indicated that both LDH–Cl− and LDH–NO3− maintained macrophage metabolic activity after 72 h of exposure at 10 μg mL−1 (Figure 2B). Consistently, live/dead staining revealed abundant Calcein-AM-positive cells and only sporadic PI-positive cells across LDH-treated groups (Figure 2C). Flow cytometric apoptosis analysis further confirmed that neither LDH formulation substantially increased early apoptotic, late apoptotic, or necrotic populations compared with the untreated control (Figure 2D). These results establish that the tested LDH concentration is well tolerated by macrophages, thereby allowing the following immunomodulatory changes to be interpreted as bioactive regulation rather than a secondary consequence of cytotoxic stress. The favorable cytocompatibility of MgAl-LDH is consistent with recent reports describing LDHs as versatile biomedical nanomaterials capable of interacting with immune and stromal cells while maintaining acceptable biosafety windows.13,19 This feature is essential for osteoimmunomodulatory biomaterials because excessive immune cell death or membrane damage could amplify damage-associated inflammatory signaling, whereas a cytocompatible material-cell interface enables regulated phenotypic remodeling. Together, the cellular internalization of LDH nanoparticles analysis and viability data support efficient cellular accessibility of the LDH nanoparticles without overt cytotoxicity at the tested concentration. To determine whether LDH nanoparticles could reverse an established inflammatory state, RAW264.7 macrophages were first activated with LPS to mimic a pro-inflammatory bone-defect microenvironment and then treated with LDH–Cl− or LDH–NO3−. Compared with the LPS group, both LDH formulations markedly reduced the transcription of Tnf-α and Il-1β while increasing the expression of Tgf-β and Il-10 (Figure 2E). This reciprocal regulation of inflammatory and reparative markers indicates that LDH nanoparticles do not simply suppress macrophage activity but rather redirect macrophage polarization toward a resolution-associated phenotype. The cytokine secretion profile measured by ELISA corroborated the transcriptional results. LDH-treated macrophages secreted lower levels of TNF-α and IL-1β, two cytokines that are strongly associated with sustained inflammatory amplification and impaired osteogenic differentiation. In contrast, TGF-β and IL-10 production was significantly increased following LDH treatment (Figure 2F). These soluble mediators are highly relevant to bone repair because IL-10 contributes to inflammatory resolution and macrophage phenotypic stabilization, whereas TGF-β participates in matrix remodeling, mesenchymal cell recruitment, and osteogenic lineage progression. Flow cytometry further confirmed that LDH exposure increased the CD206⁺CD86− macrophage fraction while decreasing the CD86⁺CD206− population (Figure 2G), providing cellular-level evidence for a shift from an M1-like state toward an M2-like phenotype. Notably, LDH–NO3− consistently exerted a stronger immunoregulatory effect than LDH–Cl−, as evidenced by more pronounced suppression of M1-associated markers, higher induction of M2-associated genes and cytokines, and the largest increase in the CD206⁺CD86−/CD86⁺CD206− ratio. Because LDH–Cl− and LDH–NO3− exhibited similar particle size and surface charge, this biological divergence is likely linked to interlayer anion-dependent differences in dissolution behavior, ion exchange, surface hydration, or protein adsorption. Recent work on nitrate-intercalated LDH systems has similarly shown that NO3−-containing LDHs can display superior pro-osteogenic and anti-inflammatory activities compared with chloride-intercalated analogues, supporting the concept that interlayer anions function as biologically instructive components rather than passive charge-balancing species.20 From an osteoimmunological perspective, these findings are particularly important. Macrophage polarization is not a binary process but a dynamic continuum, and successful bone repair requires timely transition from early inflammatory activation to a reparative immune state. Persistent M1-like activation can maintain high levels of TNF-α and IL-1β, thereby inhibiting osteoblast differentiation and promoting fibrotic tissue formation. Conversely, M2-skewed macrophages secrete pro-regenerative cytokines and growth factors that support extracellular matrix deposition, vascular ingrowth, and osteoprogenitor differentiation. Recent high-impact studies on immunomodulatory bone biomaterials emphasize that materials capable of reshaping macrophage phenotype can generate a permissive osteoimmune niche and thereby enhance bone regeneration.2,21,22 The present data demonstrate that LDH nanoparticles, especially LDH–NO3−, possess this macrophage-instructive capability and may therefore serve as nanoscale immune regulators for defect repair. Anion-Dependent Promotion of BMSC Osteogenic Differentiation Given that bone regeneration requires the coordinated recruitment, survival, and osteogenic differentiation of mesenchymal progenitors, the direct influence of LDH nanoparticles on BMSCs was systematically evaluated. Flow cytometric characterization confirmed that the cells used in this study exhibited a typical mesenchymal stem cell immunophenotype, with high expression of CD29, CD44, and CD90 and low expression of hematopoietic markers CD34 and CD45 (Figure S1). This phenotype supports the suitability of the cell model for osteogenic differentiation experiments. FITC-labeled LDH-Cl− and LDH-NO3− produced clear cytoplasmic signals in BMSCs after thorough PBS washing, without obvious morphological impairment (Figure 3A). CCK-8 and live/dead assays demonstrated that BMSC viability was preserved after LDH exposure (Figure 3B and C). Together with the macrophage observations, these data show that LDH nanoparticles can interact with both immune and stromal cell compartments while maintaining cytocompatibility.Figure 3LDH nanoparticles enhance BMSC osteogenic differentiation in an anion-dependent manner. (A) Low- and high-magnification confocal images acquired after thorough PBS washing show intracellular/cytoplasmic FITC-LDH signals (green) in BMSCs; F-actin and nuclei are labeled with phalloidin (red) and DAPI (blue), with quantitative analysis of FITC-LDH-positive cells. (B and C) CCK-8 and Calcein-AM/PI staining showing BMSC cytocompatibility after LDH exposure. (D and E) ALP staining and quantitative ALP activity after 7 days. (F and G) Alizarin Red S staining and quantitative mineralization analysis after 14 days. (H) qRT-PCR analysis of Runx2, Opn, Alpl, Ocn, and Bmp2. (I and J) OCN immunofluorescence and fluorescence quantification; scale bars are shown in (I). The BMSCs group represents the non-osteogenic blank control cultured in basal medium without osteogenic supplements; Ctrl represents osteogenic induction medium without LDH. Quantitative data are presented as mean ± SD (n = 3 independent biological replicates). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.LDH-Cl and LDH-NO3 boost BMSC osteogenic differentiation, enhancing markers and mineralization.A multi-panel infographic summarizes the effects of LDH-Cl and LDH-NO3 on BMSC osteogenic differentiation. Panels are organized in rows, showing uptake, viability, ALP activity, mineralization, gene expression and OCN staining. Confocal images (Panel A) show FITC-LDH signals in BMSCs, with LDH-NO3 showing higher uptake. Panel B presents a bar chart of cell viability, with no significant differences. Panel C shows Calcein-AM/PI staining, indicating cell viability. Panel D displays ALP staining, with LDH-NO3 showing stronger signals. Panel E′s bar chart shows relative ALP activity, with LDH-NO3 highest. Panel F shows Alizarin Red S staining, with LDH-NO3 indicating more mineralization. Panel G′s bar chart confirms this trend. Panel H shows gene expression, with LDH-NO3 increasing Alpl, Runx2, Ocn, Bmp2 and Opn. Panel I shows OCN immunofluorescence, with LDH-NO3 having higher intensity. Panel J′s bar chart supports this. Significant differences are noted (p<0.01/p<0.0001). LDH nanoparticles enhance BMSC osteogenic differentiation in an anion-dependent manner. (A) Low- and high-magnification confocal images acquired after thorough PBS washing show intracellular/cytoplasmic FITC-LDH signals (green) in BMSCs; F-actin and nuclei are labeled with phalloidin (red) and DAPI (blue), with quantitative analysis of FITC-LDH-positive cells. (B and C) CCK-8 and Calcein-AM/PI staining showing BMSC cytocompatibility after LDH exposure. (D and E) ALP staining and quantitative ALP activity after 7 days. (F and G) Alizarin Red S staining and quantitative mineralization analysis after 14 days. (H) qRT-PCR analysis of Runx2, Opn, Alpl, Ocn, and Bmp2. (I and J) OCN immunofluorescence and fluorescence quantification; scale bars are shown in (I). The BMSCs group represents the non-osteogenic blank control cultured in basal medium without osteogenic supplements; Ctrl represents osteogenic induction medium without LDH. Quantitative data are presented as mean ± SD (n = 3 independent biological replicates). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Under osteogenic induction, LDH-treated BMSCs displayed enhanced early osteoblastic differentiation. The non-osteogenic blank group showed minimal ALP staining and substantially lower ALP activity than the osteogenic-medium control, confirming the induction-dependent nature of the assay. Among osteogenically induced groups, ALP staining at day 7 was more extensive and intense after LDH treatment, and the LDH-NO3− group showed the strongest staining intensity (Figure 3D). Quantitative ALP activity measurements confirmed this trend (Figure 3E), indicating that LDH nanoparticles enhance early osteogenic commitment, with the nitrate-intercalated formulation producing the strongest response. Matrix mineralization was assessed by Alizarin Red S staining after 14 days. The non-osteogenic blank group showed little mineral deposition, whereas osteogenic induction increased mineralization and LDH treatment further enhanced this response. Among the induced groups, the LDH-NO3− group exhibited the greatest mineralized nodule formation and Alizarin Red S signal (Figure 3F and G). Consistent with these phenotypic findings, the non-osteogenic blank group showed lower expression of Runx2, Opn, Bmp2, Alpl, and Ocn and lower OCN immunofluorescence than the osteogenic control, while LDH treatment increased these osteogenic markers, with the strongest response observed for LDH-NO3− (Figure 3H–J). Scale bars have been added to the OCN immunofluorescence images in Figure 3I. These data demonstrate that LDH exposure is associated with enhanced BMSC osteogenic differentiation in addition to its immunomodulatory effects on macrophages. However, the present experiments do not directly establish the molecular mechanism responsible for this effect. Plausible contributors include magnesium-related biochemical cues, electrostatic cell-material interactions, and interlayer-anion-dependent changes in ion exchange, hydration, and protein adsorption, but these possibilities require direct mechanistic testing. Accordingly, the stronger response observed with LDH-NO3− should be interpreted as evidence of an anion-dependent association rather than proof of a specific causal signaling mechanism. Future studies using ion-release profiling, protein-corona analysis, and pathway-specific perturbation will be needed to distinguish these mechanisms. LDH-Reinforced GelMA Hydrogels Promote Critical-Sized Calvarial Bone Regeneration Encouraged by the in vitro immunomodulatory and osteogenic results, LDH nanoparticles were incorporated into photocrosslinked GelMA hydrogels to generate implantable composite scaffolds for bone repair. The LDH loading concentration was fixed at 1 mg mL−1 in the GelMA precursor to provide sufficient inorganic bioactivity while avoiding excessive nanoparticle aggregation or disruption of the photocrosslinked hydrogel network. This formulation allowed homogeneous dispersion of LDH nanoplates within the GelMA matrix and preserved the interconnected porous architecture after lyophilization. Because all hydrogel groups were prepared with the same GelMA concentration, photoinitiator concentration, crosslinking condition, and scaffold volume, the improved osteoimmunomodulatory and osteogenic outcomes could be primarily attributed to the incorporated LDH nanoplates and their interlayer anion chemistry. The LDH concentration of 1 mg mL−1 was guided by previous studies on LDH/GelMA and bioactive inorganic nanofiller-reinforced photocurable hydrogels,14–16,23–25 in which appropriately controlled nanofiller incorporation enhanced osteogenic bioactivity while preserving hydrogel handling, crosslinking, and structural integrity. SEM imaging of lyophilized hydrogels revealed a porous and interconnected architecture in GelMA, GelMA-LDH-Cl−, and GelMA-LDH-NO3− groups (Figure 4A). The introduction of LDH nanoparticles did not collapse the hydrogel network or eliminate pore interconnectivity, indicating that LDH incorporation was compatible with GelMA gelation. Cytocompatibility assays further confirmed that GelMA and LDH-reinforced GelMA hydrogels supported the viability of RAW264.7 macrophages and BMSCs (Figure S2), supporting their suitability as cell-interactive regenerative matrices.Figure 4LDH-reinforced GelMA hydrogels enhance calvarial defect repair. (A) SEM images showing porous and interconnected structures of GelMA and GelMA-LDH hydrogels. (B) Representative 3D micro-CT reconstructions of calvarial defects at 12 weeks after implantation. (C) Quantitative analysis of BV/TV and BMD within the defect region. (D) qRT-PCR analysis of inflammatory and reparative macrophage-related genes in regenerated tissues. (E) qRT-PCR analysis of osteogenic genes in regenerated tissues. The x-axis group labels are shown in all quantitative panels. Animal study n = 6 mice per group. Data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.Composite image showing SEM, micro-CT and gene analysis of GelMA and GelMA-LDH hydrogels.The image A shows SEM images of GelMA, GelMA-LDH-Cl superscript minus and GelMA-LDH-NO subscript 3 hydrogels, revealing porous structures. The image B shows 3D micro-CT reconstructions of calvarial defects at 12 weeks post-implantation for the same groups. The image C shows bar graphs of relative BV/TV and BMD, comparing GelMA, GelMA-LDH-Cl superscript minus and GelMA-LDH-NO subscript 3, with statistical significance indicated by asterisks. The image D shows qRT-PCR analysis of inflammatory and reparative macrophage-related genes, including Tnf-alpha/Gapdh, Il-1beta/Gapdh, Tgf-beta/Gapdh and Il-10/Gapdh. The image E shows qRT-PCR analysis of osteogenic genes, including Alpl/Gapdh, Runx2/Gapdh, Ocn/Gapdh and Bmp2/Gapdh, with statistical significance indicated by asterisks. LDH-reinforced GelMA hydrogels enhance calvarial defect repair. (A) SEM images showing porous and interconnected structures of GelMA and GelMA-LDH hydrogels. (B) Representative 3D micro-CT reconstructions of calvarial defects at 12 weeks after implantation. (C) Quantitative analysis of BV/TV and BMD within the defect region. (D) qRT-PCR analysis of inflammatory and reparative macrophage-related genes in regenerated tissues. (E) qRT-PCR analysis of osteogenic genes in regenerated tissues. The x-axis group labels are shown in all quantitative panels. Animal study n = 6 mice per group. Data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. GelMA is widely used in regenerative biomaterials because it combines photocrosslinkability, ECM-mimetic biochemical motifs, and adjustable physical properties. However, GelMA alone is generally insufficient to drive robust bone regeneration in critical-sized defects, particularly in inflammatory environments. Recent studies emphasize that hydrogel systems should be designed not merely as passive space fillers but as active regulators of immune–stromal communication, cell infiltration, and local tissue remodeling.7,10 In this context, LDH incorporation endows GelMA with inorganic bioactivity and anion-dependent immunomodulatory capacity, thereby converting a cytocompatible hydrogel matrix into an osteoimmune-instructive scaffold. The regenerative performance of the hydrogels was evaluated in a murine 5-mm critical-sized calvarial defect model. The in vivo design was intended to compare the effect of LDH incorporation and interlayer anion identity within a common GelMA scaffold; therefore, GelMA served as the scaffold control. At 12 weeks after implantation, micro-CT reconstruction showed limited bone regeneration in the GelMA-only group, with a large unbridged defect area remaining (Figure 4B). In contrast, both GelMA-LDH groups displayed more extensive radiopaque tissue formation within the defect region. Among all groups, GelMA-LDH-NO3− exhibited the most pronounced defect bridging and mineralized tissue coverage. Quantitative micro-CT analysis demonstrated higher BV/TV and BMD in LDH-reinforced groups than in GelMA alone, with the highest values in the GelMA-LDH-NO3− group (Figure 4C). Because an empty-defect model group and a sham-operated group were not included, the present data primarily support comparative efficacy among the tested scaffold formulations rather than a complete separation of spontaneous healing, surgery-only effects, and GelMA-specific effects; this limitation is acknowledged below. To determine whether enhanced bone formation was associated with modulation of the local immune microenvironment, inflammatory and osteogenic gene expression was examined in regenerated tissues. Compared with GelMA alone, GelMA-LDH scaffolds reduced Tnf-α and Il-1β expression while increasing Tgf-β and Il-10 (Figure 4D). Osteogenic markers, including Runx2, Bmp2, Ocn, and Alpl, were also upregulated in LDH-containing groups (Figure 4E), indicating that inflammatory resolution occurred alongside active osteogenic matrix formation. Histological staining was evaluated before immunofluorescence to emphasize tissue-level regeneration. H&E and Masson’s trichrome staining are presented first in Figure 5A, followed by immunofluorescence analyses in Figure 5B and C. H&E staining showed that GelMA-only defects were mainly occupied by fibrous connective tissue with limited osteoid deposition, whereas GelMA-LDH hydrogels supported more substantial new bone-like tissue formation, improved defect bridging, and closer integration with adjacent native bone. Masson’s staining showed denser collagen-rich matrix and more organized mineralizing tissue in LDH-containing groups, particularly in the GelMA-LDH-NO3− group. Immunofluorescence analysis then showed lower TNF-α signals and higher IL-10 and OCN signals in GelMA-LDH-treated defects, with the strongest IL-10 and OCN signals in the GelMA-LDH-NO3− group. Histological examination of the brain, heart, liver, spleen, lung, and kidney did not reveal apparent pathological injury (Figure 5D), supporting preliminary in vivo biocompatibility.Figure 5Histological and immunofluorescence evaluation of regenerated calvarial tissues. (A) H&E and Masson’s trichrome staining showing tissue bridging, collagen deposition, and new bone matrix formation. (B) Immunofluorescence staining of TNF-α, IL-10, and OCN in the defect region. (C) Quantitative fluorescence analysis. (D) Histological evaluation of the brain, heart, liver, spleen, lung, and kidney for preliminary tissue biocompatibility. Animal study n = 6 mice per group. Data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.Microscopy figure: H&E, Masson stains, immunofluorescence, bar graphs, organ histology on varied backgrounds. Histological and immunofluorescence evaluation of regenerated calvarial tissues. (A) H&E and Masson’s trichrome staining showing tissue bridging, collagen deposition, and new bone matrix formation. (B) Immunofluorescence staining of TNF-α, IL-10, and OCN in the defect region. (C) Quantitative fluorescence analysis. (D) Histological evaluation of the brain, heart, liver, spleen, lung, and kidney for preliminary tissue biocompatibility. Animal study n = 6 mice per group. Data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Together, these in vivo data demonstrate that LDH–GelMA hydrogels promote bone regeneration through coordinated immunomodulation and osteoinduction. The superiority of GelMA–LDH–NO3− is consistent with the in vitro macrophage and BMSC results, indicating that the biological advantage of nitrate intercalation is maintained after incorporation into a hydrogel matrix. This is highly relevant for advanced bone-regenerative material design, where the challenge is not only to provide an osteoconductive scaffold but also to actively reconstruct a microenvironment that suppresses chronic inflammation while stimulating osteogenic tissue formation. Recent immunomodulatory bone biomaterials, including collagen-based membranes and mineralized scaffolds, have similarly shown that controlling the osteoimmune niche is critical for achieving advanced bone regeneration.2,22 The present LDH–GelMA system extends this principle by demonstrating that LDH interlayer anion chemistry can be used as a subtle yet effective design parameter to amplify osteoimmune coupling. Size-optimized LDH nanoplatelet-reinforced GelMA hydrogels have shown that nanoparticle geometry can influence osteoimmunomodulation and critical-sized defect repair, while calcein-loaded LDH/GelMA scaffolds and photocurable LDH-containing hydrogels have demonstrated osteogenic and regenerative benefits through cargo delivery or matrix engineering.14–16 In contrast, the present study focuses specifically on interlayer anion identity within a comparable MgAl-LDH framework. This comparison supports the novelty of using anion chemistry itself as a material-design variable, while also emphasizing that the downstream mechanisms identified by transcriptomics require further causal validation. Transcriptomic Analysis Associates PI3K/Akt Signaling with Regeneration and Reduced NF-κB-Associated Inflammation To gain mechanistic insight into how LDH-reinforced hydrogels remodel the defect microenvironment, RNA sequencing was performed on calvarial tissues from the GelMA, GelMA-LDH-Cl−, and GelMA-LDH-NO3− groups (n = 3 biological replicates per group). PCA of all nine samples showed group-wise clustering and separation among the three hydrogel conditions (Figure S3), supporting reproducibility within groups and treatment-associated transcriptional differences. Hierarchical clustering and heatmap analysis likewise showed distinct expression patterns across the three groups (Figure 6A). For pathway-focused interpretation, the GelMA-LDH-NO3− versus GelMA comparison was used as the principal contrast because the nitrate-intercalated formulation produced the strongest regenerative phenotype.Figure 6Transcriptomic analysis of anion-dependent GelMA-LDH-mediated bone regeneration. (A) Heatmap and hierarchical clustering of gene-expression patterns in GelMA, GelMA-LDH-Cl−, and GelMA-LDH-NO3− groups. (B) KEGG enrichment analysis showing major pathways associated with the regenerative response. (C) GSEA indicating enrichment of PI3K/Akt-associated signaling and NF-κB-associated inflammatory signaling in the GelMA-LDH-NO3− group. (D) PPI network analysis highlighting representative hub genes, including Akt1 and Rela. (E) qRT-PCR validation of selected pathway-related genes in regenerated tissues. RNA-seq n = 3 biological replicates per group; tissue qRT-PCR data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.Analysis of gene expression and pathways in GelMA, GelMA-LDH-Cl and GelMA-LDH-NO subscript 3 groups.The image consists of multiple panels analyzing gene expression and pathways. A shows a heatmap and hierarchical clustering of gene-expression patterns in GelMA, GelMA-LDH-Cl and GelMA-LDH-NO subscript 3 groups, with a log(TPM+1) scale. B lists KEGG pathways, including PI3K-Akt, NF-kappaB, cytokine-cytokine receptor interaction and others, with -log subscript 10 (Q value) on the x-axis. C presents enrichment score (ES) graphs for PI3K-Akt and NF-kappaB signaling pathways. D depicts a PPI network highlighting hub genes Akt1 and Rela, with node links and ppi edges. E shows bar graphs of Akt1/Gapdh and Rela/Gapdh expression levels in the three groups, with significance indicated by asterisks: p < 0.05, *p < 0.01. Transcriptomic analysis of anion-dependent GelMA-LDH-mediated bone regeneration. (A) Heatmap and hierarchical clustering of gene-expression patterns in GelMA, GelMA-LDH-Cl−, and GelMA-LDH-NO3− groups. (B) KEGG enrichment analysis showing major pathways associated with the regenerative response. (C) GSEA indicating enrichment of PI3K/Akt-associated signaling and NF-κB-associated inflammatory signaling in the GelMA-LDH-NO3− group. (D) PPI network analysis highlighting representative hub genes, including Akt1 and Rela. (E) qRT-PCR validation of selected pathway-related genes in regenerated tissues. RNA-seq n = 3 biological replicates per group; tissue qRT-PCR data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. KEGG pathway enrichment analysis revealed that the differentially expressed genes were associated with PI3K–Akt signaling, NF-κB signaling, cytokine–cytokine receptor interaction, TNF signaling, JAK–STAT signaling, MAPK signaling, Hedgehog signaling, ECM–receptor interaction, and HIF-1 signaling pathways (Figure 6B). These enriched pathways can be functionally interpreted as three interconnected modules. The first module involves inflammatory regulation, including NF-κB, TNF, JAK–STAT, and cytokine–cytokine receptor interaction pathways. The second module relates to osteogenic differentiation and cell survival, including PI3K–Akt, MAPK, Hedgehog, and ECM–receptor interaction pathways. The third module reflects tissue adaptation and matrix remodeling, including HIF-1 signaling and ECM-related pathways. Such pathway convergence is biologically plausible because bone regeneration is governed by continuous communication among immune cells, osteoprogenitors, endothelial cells, and extracellular matrix components. Recent work has emphasized that osteogenesis is tightly coupled to broader cell–cell communication networks within the regenerative microenvironment, while macrophage-centered studies have highlighted the importance of inflammatory signaling balance in tissue repair.21,26 Among the enriched pathways, PI3K/Akt signaling emerged as a central axis associated with GelMA–LDH–NO3−-mediated repair. GSEA demonstrated significant enrichment of PI3K/Akt signaling in the GelMA–LDH–NO3− group relative to GelMA alone (Figure 6C). PPI network analysis further identified Akt1 as a representative hub gene within the differentially expressed gene network (Figure 6D), and qRT-PCR validation confirmed increased Akt1 expression in GelMA–LDH–NO3−-treated tissues (Figure 6E). PI3K/Akt signaling is a multifunctional pathway that integrates extracellular matrix cues, growth factor stimulation, ion-mediated signaling, and mechanical inputs. In bone regeneration, activation of this pathway can support cell survival, osteogenic differentiation, matrix production, and reparative tissue remodeling. Accordingly, the PI3K/Akt enrichment observed here is biologically consistent with the increased ALP activity, mineralized matrix deposition, OCN expression, and micro-CT-based bone formation in the LDH-NO3− groups; however, these parallel changes do not establish pathway causality. In parallel, the transcriptomic data revealed suppression of inflammatory signaling. GSEA showed reduced enrichment of NF-κB-associated gene sets in GelMA–LDH–NO3−-treated defects (Figure 6C), and qRT-PCR confirmed the downregulation of Rela (Figure 6E). NF-κB is a canonical inflammatory transcriptional regulator that promotes the expression of TNF-α, IL-1β, and other pro-inflammatory mediators. Sustained NF-κB activation after biomaterial implantation may prolong inflammatory macrophage activation, inhibit osteogenic differentiation, and interfere with matrix maturation. Thus, the attenuation of NF-κB signaling is consistent with the reduced TNF-α/IL-1β expression and elevated IL-10/TGF-β levels observed both in vitro and in vivo. This anti-inflammatory transcriptional profile is consistent with a shift from an inflammatory defect environment toward a more pro-regenerative osteoimmune state, but direct pathway perturbation is required to establish causality. The enrichment of ECM-receptor interaction and HIF-1-related pathways further suggests that the regenerative response may involve matrix-associated tissue organization in addition to immune and osteogenic signaling. ECM remodeling is an active regulatory process during bone repair, governing cell adhesion, mechanotransduction, osteoprogenitor differentiation, and vascular-associated tissue maturation. HIF-1 signaling may reflect adaptive responses to the hypoxic and metabolically dynamic microenvironment of healing bone defects. Recent studies have shown that ECM organization and metabolic adaptation can be tightly interconnected in musculoskeletal repair, and dysregulation of these processes may lead to aberrant matrix remodeling.27 In the present study, the coordinated changes in PI3K/Akt-, NF-κB-, ECM-receptor-, and HIF-1-associated gene sets are consistent with a network-level response involving immune resolution, osteogenic activation, and matrix remodeling; they do not by themselves establish that GelMA-LDH-NO3− directly controls each pathway. Overall, the RNA-seq results provide pathway-level associations that connect the cellular and animal observations, but they should not be interpreted as proof of pathway causality. In macrophages, LDH-NO3− reduced inflammatory mediators and favored a reparative phenotype; in BMSCs, it enhanced osteogenic differentiation and mineralization; and in vivo, it increased bone volume and mineral density while reducing inflammatory signals. Transcriptomic profiling was consistent with increased PI3K/Akt-associated regenerative signaling and reduced NF-κB-associated inflammatory programs. Direct causal validation using pathway inhibitors, genetic perturbation, or dedicated in vitro pathway assays remains necessary. Representative pathway-associated genes were examined in the relevant cell types (Figure S4). In RAW264.7 macrophages, LPS increased Rela expression, whereas subsequent treatment with LDH-Cl− or LDH-NO3− reduced Rela, with the greatest reduction in the LDH-NO3− group. In osteogenically induced BMSCs, Akt1 expression increased after LDH treatment, with the highest expression in the LDH-NO3− group. These cell-type-specific transcriptional changes are concordant with the NF-κB- and PI3K/Akt-associated signatures identified in vivo, while remaining insufficient by themselves to establish pathway causality. Several limitations should be acknowledged. First, the in vivo study focused on comparisons among GelMA, GelMA-LDH-Cl−, and GelMA-LDH-NO3− and did not include an empty-defect model group or a sham-operated group; therefore, spontaneous healing and surgery-only effects cannot be fully separated from scaffold-specific effects. Second, although RNA-seq, tissue qRT-PCR, and the additional in vitro Akt1/Rela qRT-PCR validation support PI3K/Akt- and NF-κB-associated transcriptional changes, these data do not directly demonstrate pathway activation or causality at the protein/phosphorylation level; pharmacological inhibition or genetic perturbation will be required in future studies. Third, the anion comparison was limited to chloride and nitrate, and the study did not systematically measure anion-release kinetics, Mg2⁺ release, protein-corona composition, or long-term hydrogel degradation. Finally, longer observation periods, larger-animal models, and additional mechanical and translational assessments will be required before clinical application. Conclusion This study demonstrates that interlayer anion engineering can regulate the osteoimmunomodulatory performance of MgAl-LDH-reinforced GelMA hydrogels. Although LDH-Cl− and LDH-NO3− exhibited broadly comparable physicochemical characteristics and cytocompatibility, LDH-NO3− produced stronger suppression of inflammatory macrophage activity and greater BMSC osteogenic differentiation. In the critical-sized calvarial defect model, GelMA-LDH-NO3− enhanced mineralized defect bridging, increased BV/TV and BMD, promoted collagen-rich matrix maturation, and increased OCN expression. Transcriptomic analysis was consistent with increased PI3K/Akt-associated regenerative signaling and reduced NF-κB-associated inflammatory signaling, although direct causal validation remains necessary. These findings identify LDH interlayer anions as tunable bioactive regulators of immune-bone communication and support further development of anion-engineered GelMA-LDH nanocomposite hydrogels for bone regeneration.