A mitochondria-targeted nanoantioxidant restores alveolar bone homeostasis in periodontitis by quenching ROS and suppressing the cGAS-STING pathway

Abstract

Prolonged periodontal inflammation and progressive alveolar bone loss are typical manifestations of periodontitis. Antioxidative therapies targeting the central role of reactive oxygen species (ROS) have been explored, but lack of subcellular specificity limits efficacy. Mitochondria function as an upstream redox hub that drives oxidative stress, inflammatory responses, and alveolar bone resorption, making mitochondrial redox modulation a promising yet underexplored strategy for periodontitis therapy. Herein, we developed a mitochondria-targeted, redox-responsive nanocomposite (TC/pSeSe) that enables programmable redox modulation of the pathological periodontal microenvironment. The antioxidative core consists of a ROS-responsive diselenide-containing copolymer (pSeSe) capable of selenium release, while the triphenylphosphine/chitosan (TC) coating confers mitochondrial-targeted, controlled redox activity, mucosal retention and cationic antibacterial properties. With preferential mitochondrial localization, TC/pSeSe undergoes diselenide bond cleavage and selenium release under oxidative stress, thereby restoring mitochondrial redox homeostasis and attenuating downstream mitochondrial DNA (mtDNA)-cGAS-STING-mediated inflammatory signaling. Through combined ROS scavenging and selenium-mediated support, TC/pSeSe mitigates ferroptosis in a partially glutathione peroxidase 4 (GPX4)-dependent manner and restores osteogenic potential in bone marrow-derived stem cells. In parallel, TC/pSeSe exhibits antibacterial activity against periodontal pathogens through combined selenium and the cationic TC coating functionalities. In vivo, TC/pSeSe restored alveolar bone regeneration and attenuated periodontal inflammation. Collectively, this study proposes a mitochondria-centered redox modulation strategy, providing a comprehensive and promising therapeutic approach for periodontitis treatment.

Graphical abstract

Highlights

• •A mitochondria-targeted redox-responsive nanocomposite was developed to modulate oxidative stress in periodontitis.
• •ROS-responsive diselenide bonds release selenium to attenuate oxidative stress and ferroptosis partially via GPX4.
• •A TC coating facilitates mitochondrial accumulation with controlled redox activity, antibacterial defense, and mucoadhesion.

1. Introduction

Periodontitis is the sixth most prevalent chronic disease worldwide, affecting over one billion people [1]. As a bacteria-triggered inflammatory disease, it is characterized by persistent periodontal inflammation and progressive alveolar bone resorption ultimately leading to tooth loss [2,3]. Periodontitis is fundamentally different from other inflammatory disorders due to its open, bacteria-colonized microenvironment that is continuously exposed to external stimuli [4]. Current standard therapies such as mechanical debridement and antibiotic treatment focus on infection control, but therapeutic outcomes often fall short in addressing the oxidative stress microenvironment and promoting tissue regeneration [5]. These limitations highlight an urgent need for innovative strategies that restore periodontal homeostasis by intervening at upstream pathological drivers.

Reactive oxygen species (ROS) have emerged as a critical factor in the progression of periodontitis with mitochondria serving as a major intracellular source [6]. Importantly, mitochondria themselves are highly susceptible to oxidative stress. Mitochondrial DNA (mtDNA), which lacks histone protection and possesses limited repair capacity, is particularly vulnerable to oxidative damage. Oxidative stress-induced mtDNA damage and cytosolic leakage can subsequently activate the cyclic GMP-AMP synthase (cGAS)-stimulator of interferon genes (STING) pathway, leading to interferon signaling and nuclear factor-κB (NF-κB) activation, thereby exacerbating periodontal inflammation [7]. Moreover, mitochondrial dysfunction further exacerbates ROS production, forming a self-amplifying cycle of ROS formation [8]. In periodontitis, excessive ROS not only intensifies inflammatory signaling, but also drives bone resorption and tissue degradation [9]. Although ROS-scavenging strategies have been widely explored in periodontitis therapy, most existing approaches primarily eliminate bulk ROS and lack subcellular specificity, thereby failing to interrupt ROS amplification at its source. Incorporating mitochondria-targeted antioxidant strategies therefore holds promise for improving therapeutic efficacy through upstream redox modulation and inflammation control.

To enable mitochondria-targeted ROS scavenging, we propose diselenide bond-containing polymers (pSeSe) modified with triphenylphosphine (TPP), a well-established mitochondrial-targeting compound. Diselenide bonds have attracted increasing attention in recent years because of their unique activity to respond to both ROS and glutathione (GSH) [10]. Under oxidative conditions, diselenide bonds are cleaved, and actively participate in ROS neutralization [11]. Beyond direct ROS responsiveness, diselenide-containing polymers exhibit dynamic redox behavior as they enable selenium release that contributes to cellular antioxidant defense. Selenium is an essential component of multiple antioxidant systems, including selenoproteins such as glutathione peroxidase 4 (GPX4) and thioredoxin reductase that play critical roles in maintaining redox homeostasis [12,13]. In addition, selenium has been reported to inhibit osteoclast formation and promote osteogenic differentiation of mesenchymal stem cells through modulation of oxidative stress and phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT) signaling pathway [[14], [15], [16], [17]]. These properties suggest that diselenide-based systems may provide combined antioxidative and pro-osteogenic benefits under pathological oxidative conditions.

TPP, one of the most studied delocalized lipophilic cations (DLCs), can traverse cell membranes and preferentially accumulate within mitochondria driven by the mitochondrial membrane potential gradient [18]. Incorporation of TPP into nanocarriers has been shown to enhance mitochondrial delivery and improve the therapeutic efficiency of antioxidant agents [19]. Accordingly, integrating TPP with ROS-responsive diselenide polymers provides a rational strategy to modulate mitochondrial redox homeostasis and suppress oxidative stress-associated pathological cascades [20].

Despite these advantages, uncontrolled redox reactivity and limited local retention in saliva-rich periodontal environments remain major obstacles for diselenide-based systems, potentially compromising their sustained antioxidative efficacy. Periodontal tissues are continuously exposed to dynamic saliva flushing and bacterial challenge, which imposes physicochemical constraints on redox-active nanomaterials. To address these limitations, we introduced a TPP/chitosan (TC) coating to modulate selenium release behavior, enhance mucoadhesive retention, provide additional antibacterial activity and mitochondria targeting properties. Rather than completely blocking selenium release, chitosan-based coating serves as a stabilizing and regulatory layer that effectively suppresses initial burst release under normal conditions, while still permitting ROS-responsive release in oxidative microenvironments [21]. In addition, the cationic amino groups of chitosan also provide mucoadhesive properties through electrostatic interactions with sialic residues from glycoproteins [22]. Moreover, the cationic nature of chitosan can disrupt bacterial membrane integrity, acting cooperatively with selenium-based antibacterial activity to enhance pathogen elimination [23]. Importantly, beyond its antibacterial and adhesive roles, chitosan also enhances intracellular retention and uptake of the nanosystem, thereby facilitating TPP-mediated mitochondrial accumulation rather than acting as an independent targeting ligand [24]. Based on these considerations, we hypothesize that constructing diselenide-based copolymers with TC modification enables mitochondria-centered controlled redox activity, prolonged mucosal retention and improved antibacterial effects. This design offers a practical strategy for mitochondrial-targeted redox regulation and is particularly suitable for periodontal therapy.

Herein, we aim to integrate mitochondria-targeting redox modulation with infection control for periodontitis treatment (Scheme 1). The proposed TC/pSeSe nanosystem features a dynamic diselenide-bonded polymer core and a functional TC outer coating. Distinct from conventional ROS-scavenging strategies, TC/pSeSe modulates mitochondrial redox homeostasis by intervening at the upstream mitochondrial oxidative source, thereby alleviating oxidative stress and downstream cGAS-STING-mediated inflammatory signaling in the periodontal microenvironment. By restoring redox balance and providing selenium-associated redox support, TC/pSeSe mitigates ROS-driven lipid peroxidation and ferroptosis-associated cellular damage, while rescuing osteogenic potential especially under inflammatory and oxidative conditions. In parallel, the nanosystem exhibits effective antibacterial defense against periodontal pathogens. Collectively, TC/pSeSe represents a mitochondria-centered programmable therapeutic strategy for periodontitis and potentially other oxidative stress-related inflammatory disorders.

Scheme 1.

Schematic illustration of the design and mitochondria-centered redox modulation mechanism of TC/pSeSe nanocomposite. TC/pSeSe preferentially targets mitochondria as the redox hub, restoring mitochondrial redox homeostasis and thereby coordinating antibacterial control, inflammation attenuation, and osteogenic restoration (created with BioRender.com).

2. Materials and methods

Materials: Poly (lactic-co-glycolic acid) (PLGA), TPP, and N, N-dimethylformamide (DMF) were purchased from Sigma-Aldrich (USA). CS was purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. Porphyromonas gingivalis (P. gingivalis)-derived lipopolysaccharide (LPS) was purchased from InvivoGen (USA). Acetic acid was obtained from Sangon Biotech Co., Ltd. MES buffer (0.2 M, pH 5.5, M885670) was purchased from Shanghai Macklin Biochemical Technology Co., Ltd. All reagents were of analytical grade. RAW 264.7 cells were obtained from Baidi Biotech Ltd, China. Escherichia coli (E. coli), Staphylococcus aureus (S. aureus), P. gingivalis, and Fusobacterium nucleatum (F. nucleatum) were obtained from the College of Stomatology, Shanghai Jiao Tong University.

Synthesis of TC/pSeSe nanocomposite: The synthesis of pSeSe segment was based on a method described before [25]. Briefly, Di-(1-hydroxylundecyl) diselenide was synthesized by reacting disodium diselenide with 11-bromoundecanol in THF at 50 °C. The resulting diol was purified by column chromatography. Subsequently, the diselenide diol was reacted with a slight excess of 2,4-toluene diisocyanate (TDI) in anhydrous THF at 50 °C under argon for 12 h to yield the isocyanate-terminated diselenide-containing polyurethane prepolymer (PUSeSe). PLGA was conjugated to the PUSeSe segment following the similar protocol to obtain pSeSe. pSeSe (10 mg) solution in DMF (1 mL) was added dropwise into deionized water (15 mL) under sonication. The resulting mixture was transferred into a dialysis bag and dialyzed against deionized water for 72 h, with the external water replaced every 6 h. After dialysis, the final volume was adjusted to 20 mL using deionized water. CS (10 mg) was dissolved in acetic acid (10 mL, 1 % (v/v)) under stirring until fully dissolved. Separately, TPP was dissolved in MES buffer (10 mL, 0.2 M, pH 5.5) to prepare a 0.5 mg/mL solution, which was then sonicated for 30 min. Then TPP solution was added dropwise to CS solution and mixed, stirred, dialyzed and lyophilized. The TPP-CS solution was mixed with pSeSe solution under vigorous stirring. The mixture was stirred until homogenous and then centrifuged at 12,000 rpm, 4 °C for 20 min. The resulting TC/pSeSe nanoparticles were washed twice with deionized water, collected by discarding the supernatant, and freeze-dried for 48 h for storage and subsequent experiments.

Characterization: The structure of TC/pSeSe nanoparticles was characterized by transmission electron microscopy (TEM, JEOL JEM-F200, Japan). Zeta potential and average particle size were measured by dynamic light scattering (DLS, Malvern Zetasizer Nano ZS90, UK). Elemental composition was further confirmed by X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha, USA) and Fourier-transform infrared spectroscopy (FTIR, Nicolet iS20, Thermo Fisher Scientific, USA). Selenium release was quantified by inductively coupled plasma optical emission spectrometry (ICP-OES, Agilent 7700, USA). Antioxidative ability was assessed using 2,2-diphenyl-1-picrylhydrazyl (DPPH) Free Radical Scavenging Capacity Assay Kit (BC4750, Solarbio, China) and Total Antioxidant Capacity Assay Kit with 2,2'-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) method (S0119, Beyotime, China). Injectability was measured using a periodontal syringe by microcomputer controlled electronic universal testing machine (CMT4503, Jiaxing Siren Machinery, China).

Mitochondrial colocalization: FITC-labeled nanomaterials were prepared by incubation with FITC (3 mg/mL) under gentle stirring overnight, followed by dialysis against distilled water to remove unbound FITC until no detectable fluorescence was observed in the dialysate. RAW 264.7 cells were seeded in glass-bottom confocal dishes at a density of 1 × 10⁶ cells per dish and allowed to adhere overnight. Prior to nanomaterial treatment, the culture medium was replaced with serum-free DMEM for 12 h. FITC-labeled nanomaterials were then added to the cells (50 μg/mL) and incubated for 4 h. After incubation, cells were gently rinsed with PBS, fixed with 4 % (w/v) paraformaldehyde, and stained with MitoTracker Red CMXRos (200 nM, A66443, Thermo Fisher Scientific, USA) at 37 °C for 30 min. Cell nuclei were counterstained with Hoechst 33342 (G1127, Servicebio, China) at room temperature for 5 min. Fluorescence images were acquired using a confocal laser scanning microscope (Leica, Germany) under identical acquisition settings for all groups. Mitochondrial colocalization was quantified by calculating Pearson's correlation coefficient (PCC) using GraphPad Prism software.

Mucoadhesion assay: Fresh rat oral mucosa was harvested, cut into uniform pieces, and mounted onto glass slides. The mucosal samples were incubated with 1 mL of FITC-labeled pSeSe or TC/pSeSe under gentle shaking overnight to allow sufficient adhesion. Subsequently, the samples were rinsed with artificial saliva (R22154, OriLeaf, China) at a constant flow rate of 1 mL/min to simulate salivary flushing. At predetermined time points, fluorescence images were acquired using a fluorescence microscope. The mean fluorescence intensity (MFI) of each sample was quantified using ImageJ software. The retention of nanomaterials on the mucosal surface was calculated according to the following equation: Remaining (%) = (MFI_time/MFI₀) × 100 % where MFI₀ represents the initial fluorescence intensity prior to rinsing.

Antibacterial analysis: The antibacterial activity of TC/pSeSe was assessed using a plate counting assay. Briefly, bacterial suspensions (1 × 10⁶ Colony-forming units (CFUs)/mL) were co-cultured with pSeSe (50 μg/mL) or TC/pSeSe (50 μg/mL) for 24 h in the 96-well culture plate. After serial dilution with saline, samples were spread onto agar plates and incubated for another 24 h. CFUs were counted, and images of the plates were recorded. To evaluate changes in bacterial morphology and membrane integrity, co-cultured samples were fixed with glutaraldehyde fixative (G1102, Servicebio, China) and dehydrated through a graded ethanol series prior to observation by scanning electron microscopy (SEM, HITACHI Regulus 8100, Japan). For live/dead staining, bacterial suspensions (10⁶ CFU/mL) were co-cultured with pSeSe or TC/pSeSe for 24 h in confocal dishes. Samples were fixed and stained with the LIVE/DEAD BacLight Bacterial Viability Kit (L7007, Thermo Fisher, USA) according to the manufacturer's protocol. The antibiofilm effect of TC/pSeSe was investigated using crystal violet staining. Bacterial biofilms were induced in 24-well plates using a sucrose-containing medium for 48 h, followed by treatment with TC/pSeSe for an additional 24 h. Samples were fixed with methanol and stained with 0.1 % crystal violet solution (G1063, Solarbio, China). Excess dye was rinsed with distilled water, and retained stain was eluted with 33 % glacial acetic acid. Quantification was performed by measuring absorbance at 595 nm.

Biocompatibility evaluation: The biocompatibility of TC/pSeSe was assessed using bone marrow-derived mesenchymal stem cells (BMSCs) isolated from Sprague-Dawley (SD) rats. Rats were euthanized following Zoletil overdose, and tibias and femurs were harvested under sterile conditions. Bone marrow was flushed out using syringes pre-filled with DMEM, and cells were collected by centrifugation at 1000 rpm for 5 min, then resuspended in complete DMEM supplemented with 10 % fetal bovine serum (FBS) and 1 % penicillin-streptomycin. Cells were cultured for 3 days prior to the first passage, and BMSCs at passages 2–5 were used for cytocompatibility assays. For cytocompatibility assessment, BMSCs were seeded in 96-well plates (Corning, USA) at a density of 5 × 10³ cells per well. After co-culturing with pSeSe (50 μg/mL) or TC/pSeSe (50 μg/mL) for 1, 4, and 7 days, cell viability was analyzed using the CCK-8 assay (BS350B, Biosharp, China) according to the manufacturer's instructions. Similarly, cells seeded in 24-well plates were co-cultured for predetermined durations and stained with a Calcein-AM/PI Double Stain Kit (40747 ES, Yeason, China). To investigate cytoskeletal morphology, cells were stained with FITC-phalloidin (40735 ES, Yeason, China) and DAPI (G1012, Servicebio, China) and visualized under laser confocal microscope (Zeiss, Germany).

Osteogenesis assay: BMSCs were seeded in 24-well plates at a density of 6 × 10⁴ cells per well and cultured. For direct treatment, BMSCs were incubated with TC/pSeSe for the indicated durations corresponding to subsequent assays. For oxidative conditions, a transwell co-culture system was employed. RAW 264.7 macrophages were seeded in the upper chambers and then stimulated with P. gingivalis-derived lipopolysaccharide (LPS-PG, InvivoGen, USA, 1 μg/mL) to induce inflammation-associated oxidative stress for 24 h. Subsequently, the cells were treated with pSeSe (50 μg/mL) or TC/pSeSe (50 μg/mL), while BMSCs were cultured in the lower chambers. For extract assay, TC/pSeSe were dialyzed against complete DMEM culture media at 37 °C and dialysates were collected at Day 4 (Extract 1) and Day 7 (Extract 2) and used respectively for subsequent BMSC co-culture. Osteogenic differentiation was induced using a commercial osteogenic induction supplement (Oricell, China) in both oxidative stress-related and extract-based assays, following the manufacturer's instructions. The pro-osteogenic potential was assessed using the BCIP/NBT Alkaline Phosphatase (ALP) Color Development Kit (C3206, Beyotime, China), ALP Assay Kit (P0321S, Beyotime, China), and the Alizarin Red S (ARS) Staining Kit for Osteogenesis (C0148S, Beyotime, China), following the manufacturers' protocols.

Antioxidation assay: RAW 264.7 cells were seeded in glass-bottom confocal dishes at a density of 2 × 10⁶ cells/dish and allowed to adhere overnight. Cells were stimulated with P. gingivalis-derived LPS (LPS-PG, InvivoGen, USA, 1 μg/mL) for 24 h. Thereafter, the following treatments were applied for an additional 24 h: PBS, pSeSe (50 μg/mL), TC/pSeSe (50 μg/mL) and RSL3 (200 nM), or 2'3'-cGAMP (2 μg/mL) was added during the last 6 h of treatment. For ROS detection, cells were rinsed with pre-warmed PBS and incubated with 2',7'-dichlorofluorescein diacetate (DCFH-DA) (10 μM, S0033S, Beyotime, China) at 37 °C for 20 min, followed by three PBS washes. For mitochondrial labeling, cells were incubated with MitoTracker Red CMXRos (200 nM, A66443, Thermo Fisher, USA) at 37 °C for 30 min, washed, and then stained with Mitochondrial Superoxide (MitoSOX) Green (5 μM, M36005, Thermo Fisher, USA) at 37 °C for 30 min. For intracellular Fe²⁺, cells were incubated with FerroOrange (1 μM, F374, Dojindo, Japan) at 37 °C for 30 min and washed. For lipid peroxidation, cells were stained with BODIPY 581/591 C11 (10 μM, D3861, Thermo Fisher, USA) at 37 °C for 10 min, followed by PBS washes. Cell nuclei were stained with Hoechst 33342 (G1127, Servicebio, China) at room temperature for 5 min. Fluorescence images were acquired on a laser scanning confocal microscope (Zeiss, Germany) under identical acquisition settings among groups. For biochemical assays, cells were harvested and analyzed using the MDA Assay Kit (M496, Dojindo, Japan) and GSH/GSSG Quantification Kit (G263, Dojindo, Japan) strictly following the manufacturers' instructions.

Immunofluorescence staining: RAW 264.7 cells were seeded in confocal dishes at 2 × 10⁶ cells/dish and fixed at the indicated time points. Cells were rinsed with ice-cold PBS and fixed with 4 % (w/v) paraformaldehyde for 15 min at room temperature. After three PBS washes, cells were permeabilized with 0.5 % Triton X-100 in PBS for 10 min at room temperature, then blocked with 5 % (w/v) BSA in PBS for 30 min at room temperature. Samples were incubated with primary antibodies including p-TBK1 and p-IRF3 (16029, CST, USA), TOM20 (11801-1-AP, Proteintech, China) and TFAM (sc-166965, Santa Cruz, USA) overnight at 4 °C, followed by fluorophore-conjugated secondary antibodies (Abclonal, China) for 1 h at room temperature in the dark. Nuclei were counterstained with DAPI (G1012, Servicebio, China). After final PBS washes, samples were mounted and imaged using a Zeiss laser confocal microscope with identical laser power, gain, and exposure settings across groups. Fluorescence quantification was performed in ImageJ software.

qRT-PCR analysis: To investigate changes in gene expression, cells were seeded in 6-well plates and co-cultured with TC/pSeSe for predetermined durations. Total RNA was extracted using the RNAprep Pure Cell/Bacteria Kit (DP430, Tiangen, China). A total of 1000 ng RNA was reverse transcribed using the PrimeScript RT Reagent Kit (Takara, Japan). Quantitative real-time PCR (qRT-PCR) was performed using Hieff qPCR SYBR Green Master Mix (11201ES03, Yeason, China) on a LightCycler 480 system (Roche, Switzerland). Relative mRNA expression of target genes was calculated using the 2^-ΔΔCt method, with GAPDH or β-actin as the internal reference gene. Primer sequences are listed in Table S1 (Supporting Information).

Western blot analysis: BMSCs were seeded in 6-well plates at a density of 1.5 × 10⁶ cells per well and treated for predetermined durations. Total cellular protein was extracted using RIPA lysis buffer (BL504A, Biosharp, China) and protease inhibitor (P1005, Beyotime, China). After the addition of SDS-PAGE sample loading buffer (AIWB-0025, Affinibody, China), lysates were boiled at 95 °C for 5 min to ensure complete denaturation. Samples were then loaded onto gels prepared with the PAGE Gel Fast Preparation Kit (PG112, EpiZyme, China) and separated by electrophoresis. Proteins were transferred onto PVDF membranes (Millipore, Germany) using NcmBlot Rapid Transfer Buffer (WB4600, NCM, China). Membranes were blocked with Protein-Free Rapid Blocking Buffer (PS108, EpiZyme, China) and incubated overnight at 4 °C with the following primary antibodies: β-actin (AC004, Abclonal, China), RUNX2 (YT5356, Immunoway, China), RANKL (HA500369, HUABIO, China), OPG (A2100, Abclonal, China), OCN (16157-1-AP, Proteintech, China), GPX4 (67763-1-lg, Proteintech, China), Nrf2 (16396-1-AP, Proteintech, China), ACSL4 (22401-1-AP, Proteintech, China), and Mouse-reactive STING Pathway Antibody Sampler Kit (16029, CST, USA). After washing, membranes were incubated with HRP-conjugated secondary antibodies, including HRP Goat Anti-Mouse IgG (H + L) (AS003, Abclonal, China) and HRP Goat Anti-Rabbit IgG (H + L) (AS014, Abclonal, China). Protein bands were visualized using the Immobilon Western Chemiluminescent HRP Substrate (WBKLS0050, Millipore, Germany). Quantification of western blotting results were completed using ImageJ software.

Animal experiments: All animal procedures were approved by the Animal Care and Experiment Committee of the Ninth People's Hospital, affiliated to Shanghai Jiao Tong University School of Medicine (Approval No. SH9H-2024-A1037-1). SD rats were randomly divided into four groups (n = 6 per group) using a random number generator: Sham group: no ligature or treatment, Ligature + PBS group: ligature-induced periodontitis with PBS injection, Ligature + pSeSe group: ligature-induced periodontitis with pSeSe injection, Ligature + TC/pSeSe group: ligature-induced periodontitis with TC/pSeSe injection. All subsequent treatments and analysis were performed in a blinded manner to minimize experimental bias. Rats were anesthetized by intraperitoneal injection of Zoletil (40 mg/kg). To induce experimental periodontitis, 5-0 silk ligatures were placed between the maxillary first and second molars for 14 days. After ligature removal, the PBS, pSeSe and TC/pSeSe groups received intra-gingival injections every other day for an additional 14 days. According to the grouping, 30 μL of PBS or drug solution (0.5 mg/mL) was injected locally at six sites around the maxillary second molar (5 μL per site). At the end of the treatment period, rats were euthanized, and the maxillary jaws were harvested for subsequent histological and biochemical analysis. Pro-inflammatory cytokines including TNF-α, IL-6, IL-1β, and IFN-β in rat serum were quantified using ELISA kits (MultiSciences, China) according to the manufacturer's instructions. DHE staining was conducted on frozen gingival soft-tissue sections due to the transient nature of ROS. All procedures were conducted in accordance with institutional guidelines.

Statistical analysis: All quantitative data were obtained from at least three independent replicates and are presented as mean ± standard deviation (SD). n indicates the number of independent biological replicates. Data were analyzed using Student's t-test or one-way analysis of variance (ANOVA) or two-way ANOVA for multiple comparisons (GraphPad Prism 9.0 software). A p-value <0.05 was considered statistically significant. ns represents no significant difference.

3. Results and discussion

3.1. Characterization of the TC/pSeSe nanocomposite

To fabricate the mitochondria-targeting ROS-responsive TC/pSeSe nanocomposite, a synthetic route was designed (Fig. 1a). The chemical shift region from characteristic ¹H NMR signals (δ ≈ 6–8 ppm for -NH- and δ ≈ 3.8–4.5 ppm for CH₂-O-C(O)NH-) were consistent with previously reported polyurethane systems and confirmed the successful formation of urethane bonds in pSeSe (Fig. S1) [26]. Digital photos of pSeSe and TC/pSeSe are shown in Fig. S2–3. Both pSeSe and TC/pSeSe exist as stable liquid dispersions. Transmission electron microscopy (TEM) revealed that pSeSe nanoparticles exhibit an average diameter of approximately 180.52 nm, while TC/pSeSe nanoparticles display an enlarged diameter of approximately 198.70 nm after TC coating, accompanied by apparent morphological changes (Fig. 1b). Dynamic light scattering (DLS) analysis showed an increase in hydrodynamic diameter from ≈184.49 ± 29.82 nm (pSeSe) to ≈ 192.40 ± 38.44 nm (TC/pSeSe) following TC coating (Fig. 1c). Consistently, zeta potential measurements further confirmed surface modification, as a marked increase from −30.7 ± 0.8 to 23.1 ± 0.1 mV was observed upon TC coating (Fig. 1d), reflecting the cationic nature of this outer layer. X-ray photoelectron spectroscopy (XPS) analysis confirmed the presence of selenium in pSeSe nanoparticles, with characteristic Se 3d peaks observed at 55.6 eV (Fig. 1e–f, S4) [27]. Following surface TC functionalization, additional distinct peaks in the N 1s and P 2p confirmed the successful incorporation of TC components (Fig. 1g–i) while the Se 3d peaks were retained (Fig. S5). Fourier-transform infrared spectroscopy (FTIR) further supported successful coating, with characteristic absorption bands corresponding to both pSeSe and TC components. Broad absorption at 3448 cm⁻¹ was assigned to -OH/-NH stretching from CS and ureido groups (Fig. 1j). A band at 1637 cm⁻¹ corresponds to ureido/amide I of the pSeSe core. Peaks at 1162 and 1076 cm⁻¹ are attributed to C-O-C stretching of PLGA and CS, while 949 and 861 cm⁻¹ indicate glycosidic vibrations of CS. Importantly, signal at ∼545 cm⁻¹ supports the presence of diselenide bonds in the polymer backbone [28]. These results confirm the successful construction of the TC/pSeSe nanocomposite.

Fig. 1.

Structural characterization of TC/pSeSe nanoparticles. a) Schematic illustration of the synthetic route for TC/pSeSe nanoparticles. b) Transmission electron microscopy (TEM) images of pSeSe and TC/pSeSe nanoparticles. Scale bar: 100 nm. c) Hydrodynamic size distribution of pSeSe and TC/pSeSe nanoparticles determined by dynamic light scattering (DLS) (n = 3). d) Zeta potential measurements indicating surface charge (n = 3). e-f) X-ray photoelectron spectroscopy (XPS) spectra of pSeSe. g-i) XPS spectra of TC/pSeSe. j) Fourier-transform infrared spectroscopy (FTIR) spectra. Data are represented as mean ± standard deviation.

Selenium release kinetics were quantified using inductively coupled plasma optical emission spectrometry (ICP-OES) (Fig. 2a). A concentration of 100 μM H₂O₂ was employed as a controlled oxidative stimulus to evaluate the ROS-responsiveness of the nanosystem. Under physiological conditions, pSeSe released 88.56 ± 3.2 % selenium in 48 h, while TC coating markedly reduced selenium release to 31.5 ± 2.0 % in TC/pSeSe over the same period. Under oxidative conditions, TC/pSeSe exhibited sustained 74.7 ± 3.0 % selenium release over a period of at least 48 h. Selenium release rate of TC/pSeSe under simulated oral conditions was similar to PBS environment (Fig. S6). To evaluate the antioxidant capacity, 2,2-diphenyl-1-picrylhydrazyl (DPPH) and 2,2'-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) radical scavenging assays were performed. Both pSeSe and TC/pSeSe demonstrated time-dependent free radical scavenging ability in DPPH and ABTS assays (Fig. 2b-c). Injectability testing showed that the injection force required for TC/pSeSe using a periodontal syringe was comparable to that of PBS, indicating acceptable injectability for local administration (Fig. 2d). Fluorescence imaging revealed enhanced colocalization of TC/pSeSe with mitochondria compared to pSeSe alone (Fig. 2e), which was further supported by quantitative Pearson's correlation coefficient (PCC) analysis. The PCC value for pSeSe was 0.2614, whereas TC/pSeSe exhibited a markedly higher PCC of 0.7789. Additional comparative colocalization analyses of TPP/pSeSe and CS/pSeSe are provided in Fig. S7. Furthermore, TC/pSeSe showed prolonged retention on mucosal surfaces compared to pSeSe, as quantified by fluorescence intensity measurements (Fig. 2f). Collectively, these results characterized the controlled selenium release, redox responsiveness, mitochondrial localization and mucosal retention of TC/pSeSe, providing a physicochemical foundation for its biological investigation.

Fig. 2.

Physicochemical characterization of the TC/pSeSe nanoparticles. a) Cumulative selenium release profiles of pSeSe in PBS and TC/pSeSe in PBS or 100 μM H₂O₂ (n = 3). b) DPPH radical scavenging assay (n = 3). c) ABTS radical scavenging assay (n = 3). d) Injectability test using periodontal syringes. e) Mitochondrial targeting evaluated by fluorescence colocalization analysis. pSeSe or TC/pSeSe were labeled with FITC (green), mitochondria were stained with MitoTracker Red (red), and nuclei were stained with Hoechst 33342 (blue). Pearson's correlation coefficient (PCC) was calculated using GraphPad Prism. f) Mucosal retention evaluated by fluorescence measurement on rat mucosal surfaces (n = 3). Data are represented as mean ± standard deviation. Statistical significance is indicated by ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. ns represents no significant difference.

3.2. Antibacterial and anti-biofilm properties of TC/pSeSe nanoparticles

Given the initiating role of bacterial infection in periodontitis, effective bacterial control was considered a prerequisite for subsequent redox modulation and immune attenuation during periodontal tissue regeneration [29]. The antibacterial activity of TC/pSeSe was first assessed against Escherichia coli (E. coli) and Staphylococcus aureus (S. aureus), representative Gram-negative and Gram-positive pathogens commonly associated with implant-related and systemic infections [30]. In addition, the periodontal pathogens Porphyromonas gingivalis (P. gingivalis) and Fusobacterium nucleatum (F. nucleatum), key members of the red and orange complexes in periodontitis, were also included for evaluation [31]. Quantitative colony-forming unit (CFU) assays revealed a marked reduction in bacterial colonies upon treatment with TC/pSeSe, confirming its potent antibacterial activity (Fig. 3a-b). Similarly, spectrophotometric measurement at 600 nm revealed an up to 64.5 ± 3.0 % reduction in bacterial survival rate, further corroborating the antibacterial efficacy of TC/pSeSe nanoparticles (Fig. 3c). Bacterial morphology following treatment was assessed via scanning electron microscopy (SEM). TC/pSeSe-treated bacteria exhibited notable structural deformation, including cell membrane disruption, shrinkage, and distortion, consistent with previous observations (Fig. 3d). Live/dead staining assays demonstrated a significant increase in red fluorescence (dead bacteria) in the TC/pSeSe group, suggesting effective microbial eradication (Fig. 3e).

Fig. 3.

In vitro antibacterial properties of TC/pSeSe nanoparticles. a-b) Representative images and quantitative analysis of bacterial colonies formed on agar plates after treatment (n = 3). Scale bar: 2 cm. c) Bacterial survival percentage after 24 h co-culture (n = 6). d) Scanning electron microscopy (SEM) images showing morphological changes and membrane integrity of bacteria. Scale bar: 5 μm and 10 μm. e) Live/dead fluorescence staining of bacteria. Scale bar: 50 μm. f) Crystal violet staining and corresponding OD₅₉₅ quantification of biofilm formation (n = 4). Escherichia coli (E. coli), Staphylococcus aureus (S. aureus), Porphyromonas gingivalis (P. gingivalis) and Fusobacterium nucleatum (F. nucleatum) were respectively chosen to evaluate the antibacterial properties of the TC/pSeSe nanoparticles. Data were represented as mean ± standard deviation.

Given the critical role of biofilm formation in periodontitis, the anti-biofilm potential of TC/pSeSe was then investigated. Crystal violet staining further confirmed these findings, highlighting the ability of TC/pSeSe to inhibit biofilm development and maintain antibacterial efficacy (Fig. 3f). Collectively, these results demonstrate that TC/pSeSe exhibits pronounced antibacterial and anti-biofilm activity, potentially associated with the combined selenium-based antibacterial activity and the cationic surface properties of the TC coating.

3.3. Biocompatibility and osteogenic potential of TC/pSeSe nanocomposite

The cytocompatibility of TC/pSeSe nanoparticles was evaluated using bone marrow-derived mesenchymal stem cells (BMSCs). CCK-8 assays demonstrated BMSCs co-culture with TC/pSeSe maintained continuous proliferation over a 7-day period, with viability reaching 94.1 % on Day 7 (Fig. 4a). Correspondingly, live/dead staining revealed that the majority of cells remained viable (green fluorescence), whereas dead cells (red fluorescence) were scarcely observed (Fig. 4b). Cytoskeletal organization, visualized by phalloidin staining, further confirmed that TC/pSeSe treatment did not disrupt typical cellular morphology, and BMSCs maintained well-spread structures with intact actin filaments (Fig. 4c).

Fig. 4.

In vitro biocompatibility and osteogenic properties of TC/pSeSe nanoparticles. a) CCK-8 assay showing cell viability of bone marrow derived stem cells (BMSCs) co-cultured with TC/pSeSe at days 1, 4, and 7 (n = 6). b) Live/dead fluorescence staining of BMSCs. Scale bar: 100 μm. c) Cytoskeletal staining (phalloidin for F-actin and DAPI for nuclei) demonstrating cellular morphology. Scale bar: 50 μm. d) Alkaline phosphatase (ALP) staining indicating early-stage osteogenic differentiation. Scale bar: 100 μm. e) Quantitative analysis of ALP activity (n = 4). f) Alizarin Red S (ARS) staining showing calcified nodule formation. Scale bar: 500 μm. g) qRT-PCR results of osteogenic gene expression (n = 6). h-i) Western blot analysis and quantification of osteogenesis-related proteins and osteoclastogenesis-related protein expression (n = 3). j) ALP and ARS staining of osteogenesis under oxidative conditions. Scale bar: 100 μm and 500 μm. k) Osteogenic gene expression under oxidative conditions (n = 3). l) ALP staining at Day 7 and ARS staining at Day 21 of BMSCs cultured with TC/pSeSe extracts, including Extract 1 (collected at Day 4) and Extract 2 (collected at Day 7). Scale bar: 100 μm and 500 μm. m) Osteogenic gene expression changes after TC/pSeSe extract treatment (n = 3). Data are represented as mean ± standard deviation. ns represents no significant difference.

To assess osteogenic differentiation of BMSCs in response to TC/pSeSe treatment, a series of osteogenic assays were performed. Alkaline phosphatase (ALP), an early biomarker of bone mineralization and extracellular matrix formation, was evaluated after 4 and 7 days of culture. As shown in Fig. 4d, TC/pSeSe-treated cells exhibited markedly increased ALP activity, which was further supported by quantitative ALP enzymatic activity analysis (Fig. 4e). To evaluate osteoblast maturation at later stages, Alizarin Red S (ARS) staining was performed to visualize calcified nodule formation [32]. Compared with the pSeSe-treated group, TC/pSeSe-treated BMSCs displayed markedly higher calcium deposition, as evidenced by more intense ARS staining (Fig. 4f). Osteogenic differentiation was further evaluated at the gene expression level using quantitative real-time PCR. The mRNA levels of osteocalcin (OCN), osteopontin (OPN), ALP and Runt-related transcription factor 2 (RUNX2) were upregulated by 2.5-, 2.6-, 1.7-, and 1.5-fold, respectively, in the TC/pSeSe group compared with control group (Fig. 4g). These findings were further validated by Western blot analysis, which revealed increased expression of key osteogenic proteins including OCN and RUNX2 (Fig. 4h-i). Furthermore, the ratio of receptor activator of nuclear factor-κB ligand (RANKL)/osteoprotegerin (OPG), which is a key axis in periodontal osteoclastogenesis, was decreased in the TC/pSeSe-treated group [33,34].

Given that oxidative and inflammatory stress are known to compromise osteogenic differentiation under pathological conditions, we next sought to distinguish whether the osteogenic effects of TC/pSeSe primarily arise from modulation of the stress-impaired microenvironment or from direct selenium release. To this end, BMSCs were co-cultured with P. gingivalis-derived lipopolysaccharide (LPS)-challenged RAW 264.7 macrophages using a transwell system. Consistent with oxidative challenge-induced osteogenic impairment, LPS led to reduced ALP activity and mineralized nodule formation, whereas TC/pSeSe largely restored these activities (Fig. 4j). These changes were further confirmed by increased expression of osteogenic genes at the transcriptional level (Fig. 4k).

To further evaluate the contribution of direct selenium release to osteogenesis, extracts from TC/pSeSe including Extract 1 (collected at Day 4) and Extract 2 (collected at Day 7) were applied to BMSCs, followed by ALP staining at Day 7 and ARS staining at Day 21 of osteogenic induction. A TC-only group was included to exclude potential effects from the coating components. Both ALP and ARS staining demonstrated that TC/pSeSe extracts supported osteogenic differentiation in a time-dependent manner, with enhanced effects observed at D7 compared with D4 (Fig. 4l). Consistently, the expression of osteogenic-related genes was elevated following extract treatment (Fig. 4m). Collectively, these results indicate that TC/pSeSe-associated osteogenic differentiation of BMSCs is linked to both oxidative stress modulation and selenium support.

3.4. Anti-oxidation and anti-ferroptosis capability of TC/pSeSe nanocomposite

LPS activates macrophages and induces mitochondrial ROS generation, thereby triggering oxidative stress and inflammatory cytokine release [35]. Therefore, LPS-stimulated RAW 264.7 macrophages were selected as a representative in vitro model that recapitulates both inflammatory and oxidative stress conditions characteristic of periodontitis. To evaluate the antioxidative efficacy of the TC/pSeSe nanocomposite, intracellular ROS levels were first assessed using DCFH-DA fluorescent probe. As shown in Fig. 5a-b, LPS stimulation notably increased the green fluorescence intensity in the LPS group, indicating intracellular ROS overload in macrophages. In contrast, the TC/pSeSe group exhibited markedly reduced fluorescence intensity, suggesting TC/pSeSe effectively alleviated ROS accumulation. This result was further validated by flow cytometry analysis (Fig. 5c), which demonstrated a significant decrease in the proportion of ROS-positive cell population from 61.7 % to 4.2 % after TC/pSeSe treatment. We next utilized the Mitochondrial Superoxide (MitoSOX) probe to assess mitochondrial ROS (mtROS) production and co-stained the cells with MitoTracker to confirm the mitochondrial localization of the ROS signals. Compared with the LPS group, TC/pSeSe substantially reduced mtROS signals, whereas pSeSe produced a partial recovery (Fig. 5d and Fig. S8). These results indicate that TC/pSeSe more effectively regulates mitochondrial-associated oxidative stress than non-targeted pSeSe.

Fig. 5.

In vitro anti-oxidative and anti-ferroptotic properties of TC/pSeSe. a-b) Intracellular reactive oxygen species (ROS) levels detected by DCFH-DA fluorescence staining (n = 3). Cell nuclei were counterstained with Hoechst 33342 (blue). Scale bar: 100 μm. c) Flow cytometry analysis of intracellular ROS levels. d) Mitochondrial superoxide assessed by Mitochondrial Superoxide (MitoSOX) Green staining and mitochondria labeled by MitoTracker Red. Cell nuclei were counterstained with Hoechst 33342 (blue). Scale bar: 10 μm. e) Intracellular Fe²⁺ levels measured by FerroOrange probe. Cell nuclei were counterstained with Hoechst 33342 (blue). Scale bar: 25 μm. f) Lipid peroxidation evaluated by BODIPY 581/591 C11 staining, showing reduced (red) and oxidized (green) forms. Cell nuclei were counterstained with Hoechst 33342 (blue). Scale bar: 10 μm. g) Measurement of MDA concentration (n = 6). h) Quantification of the GSH/GSSG ratio (n = 6). i) Western blotting and quantitative analysis of GPX4, ACSL4 and Nrf2 protein expression, normalized to β-actin (n = 3). Data are represented as mean ± standard deviation.

Given that excessive mitochondrial ROS is a key upstream trigger of lipid peroxidation and ferroptotic cell injury, the impact of TC/pSeSe on ferroptosis-related pathways was further investigated in the following studies. Intracellular ferrous iron (Fe²⁺) levels were first assessed using the FerroOrange probe. As shown in Fig. 5e and Fig. S9, LPS triggered iron overload in the LPS group, as indicated by elevated fluorescence intensity. In contrast, TC/pSeSe treatment significantly reduced intracellular Fe²⁺ levels, suggesting alleviation of iron overload. Lipid peroxidation, a hallmark event of ferroptosis, was subsequently evaluated using BODIPY 581/591 C11 dual-staining. Results revealed lowered levels of lipid peroxidation in the TC/pSeSe-treated group compared with the LPS-only group (Fig. 5f and Fig. S10), suggesting suppression of lipid oxidative damage. These results were further supported by malondialdehyde (MDA) assay, an end product of lipid peroxidation marker indicative of ferroptosis [36]. TC/pSeSe treatment significantly reduced MDA levels from 6.8 ± 0.4 to 4.0 ± 0.3 nmol/mg protein compared to the LPS group (Fig. 5g). Given the central role of GSH in maintaining redox balance and supporting ferroptosis defense, intracellular redox status was further examined [37]. The GSH/glutathione disulfide (GSSG) ratio was markedly decreased in the LPS group, indicating disruption of the redox environment and this effect was reversed by TC/pSeSe treatment (Fig. 5h). Western blot analysis further confirmed the regulatory effect of TC/pSeSe on ferroptosis-related proteins (Fig. 5i). GPX4, a selenoprotein and a protective enzyme that detoxifies lipid hydroperoxide and protects cells from ferroptosis, was significantly decreased in response to LPS stimulation and effectively reversed by TC/pSeSe [38]. Consistently, the expression of Acyl-CoA synthetase long-chain family member 4 (ACSL4), a promoter of lipid oxidation, was also markedly reduced. Notably, the expression of Nrf2, a master transcription factor involved in antioxidant responses, was also enhanced following TC/pSeSe treatment.

To unravel whether TC/pSeSe inhibits LPS-induced ferroptosis via GPX4-dependent mechanism, RAS-selective lethal 3 (RSL3) was used as a GPX4 inhibitor and ferroptosis inducer [39]. As shown in Fig. 6a-b and Fig. S11-12, the antioxidative and anti-ferroptotic effects of TC/pSeSe were partially attenuated after RSL3 treatment. Correspondingly, the TC/pSeSe-mediated restoration of GPX4 expression and suppression of ACSL4 were both partially reversed in the presence of RSL3, indicating involvement of the GPX4 axis (Fig. 6c). Notably, TC/pSeSe retained measurable protective effects even under GPX4 inhibition, suggesting that its anti-ferroptotic activity is not exclusively dependent on GPX4. We postulate that, even under GPX4 inhibition, TC/pSeSe retains a certain protective capacity by reducing ROS and lipid peroxidation via its intrinsic redox functionality. Taken together, these findings demonstrate that TC/pSeSe nanoparticles possess effective antioxidative and anti-ferroptotic properties through a partial GPX4-dependent manner, accompanied by GPX4-independent redox regulation (Fig. 6d).

Fig. 6.

RSL3 partially reverses the antioxidative and anti-ferroptotic effects of TC/pSeSe. a) Fluorescence images of intracellular ROS levels detected by DCFH-DA probe after RSL3 treatment. Cell nuclei were counterstained with Hoechst 33342 (blue). Scale bar: 100 μm. b) Lipid peroxidation evaluation by BODIPY 581/591C11 staining after RSL3 treatment. Cell nuclei were counterstained with Hoechst 33342 (blue). Scale bar: 10 μm. c) Western blot analysis and quantification of GPX4 and ACSL4 protein expression after RSL3 addition, normalized to β-actin (n = 3). d) Schematic diagram illustrating the proposed mechanism of TC/pSeSe in restoring mitochondria-centered redox homeostasis and inhibiting ferroptosis (created with BioRender.com). Data are represented as mean ± standard deviation.

3.5. mtDNA-cGAS-STING pathway mitigation and mitochondrial function restoration of TC/pSeSe nanocomposite

Due to oxidative stress, mitochondrial dysfunction not only promotes lipid peroxidation and ferroptotic injury, but also compromises mitochondrial genome integrity [40,41]. Damaged mitochondria and resulting mtDNA cytoplasmic escape can trigger the cGAS-STING pathway, which functions as a critical innate immune surveillance mechanism [42,43]. We hypothesized that the mitochondria-targeting redox modulation of TC/pSeSe may exert anti-inflammatory effects by preserving mitochondrial homeostasis and suppressing aberrant cGAS-STING activation.

To investigate this hypothesis, we first assessed the phosphorylation levels and subcellular location of TANK-binding kinase 1 (TBK1) and interferon regulatory factor 3 (IRF3), two essential markers of cGAS-STING pathway, by immunofluorescence analysis. As expected, LPS challenge upregulated the expression of phosphorylated (p-)TBK1 and nuclear accumulation of p-IRF3 (Fig. 7a), indicating activation of the cGAS-STING axis. pSeSe and TC/pSeSe treatment markedly reduced both p-TBK1 levels and nuclear-localized p-IRF3, suggesting successful inhibition of cGAS-STING transcriptional activity. As a parallel approach to examine cGAS-STING pathway activation, Western blot analysis showed decreased expression of cGAS, STING as well as ratios of p-TBK1/TBK1 and p-IRF3/IRF3 following TC/pSeSe treatment (Fig. 7b and Fig. S13). Furthermore, qRT-PCR revealed a significant downregulation of Ifnb and representative interferon-stimulated genes (ISGs) (Fig. 7c), accompanied by reduced secretion of interferon-β (IFN-β), tumor necrosis factor-α (TNF-α), and interleukin (IL)-6 as determined by enzyme-linked immunosorbent assay (ELISA) (Fig. 7d). To further establish the causal involvement of the cGAS-STING pathway, we introduced 2'3'-cGAMP, a well-established cGAS-STING agonist. cGAMP stimulation re-elevated levels of cGAS and STING, restored the phosphorylation of TBK1 and IRF3 as evidenced by immunofluorescence and western blotting (Fig. 7e–f and Fig. S14), and restored the expression of Ifnb, ISGs, and pro-inflammatory cytokines (Figs. S15–16). Collectively, these findings demonstrate that TC/pSeSe attenuates LPS-induced cGAS-STING activation.

Fig. 7.

The protective effects of TC/pSeSe on cGAS-STING pathway activation and mitochondrial function. a) Immunofluorescence staining of phosphorylated TANK-binding kinase 1 (p-TBK1) (green fluorescence), p-interferon regulatory factor 3 (p-IRF3) (red fluorescence) and nuclei (blue fluorescence). Scale bar: 10 μm. b) Western blot analysis of cGAS, STING, p-TBK1, TBK1, p-IRF3, IRF3, and β-actin. c) Relative mRNA expression of Ifnb and representative interferon-stimulated genes (ISGs) (n = 3). d) Protein expression levels of interferon (IFN)-β, tumor necrosis factor-α (TNF-α), and interleukin (IL-)6 determined by enzyme-linked immunosorbent assay (ELISA) (n = 3). e) Immunofluorescence staining of p-TBK1 (green) and p-IRF3 (red) after treatment with cGAMP. Scale bar: 10 μm. f) Western blot analysis of essential proteins in the cGAS-STING pathway following cGAMP stimulation. g) Co-immunofluorescence of mitochondrial transcription factor A (TFAM, mitochondrial DNA marker) and TOM20 (mitochondrial outer membrane marker). Cell nuclei were counterstained with DAPI (blue). Scale bar: 5 μm. h) Mitochondrial morphology assessment via TOM20 immunofluorescent staining. Cell nuclei were counterstained with DAPI (blue). Scale bar: 5 μm and 2 μm. i) TEM imaging of mitochondrial ultrastructure. Scale bar: 1 μm and 250 nm. j) JC-1 staining to evaluate mitochondrial membrane potential. Scale bar: 50 μm. k) Schematic illustration of the mechanism by which TC/pSeSe protects against mitochondrial oxidative damage and attenuates cGAS-STING pathway activation (created with BioRender.com). Data are represented as mean ± standard deviation.

We next examined whether TC/pSeSe suppresses cGAS-STING activation by preventing cytosolic leakage of mtDNA. Mitochondrial transcription factor A (TFAM), a mtDNA-binding nucleoid protein, was used as a marker for mtDNA, while translocase of outer mitochondrial membrane 20 (TOM20) was used to label the mitochondrial outer membrane [44]. Immunofluorescence co-staining demonstrated clear mtDNA cytoplasmic escape upon LPS challenge (Fig. 7g, white arrows), which was significantly reversed by TC/pSeSe treatment, as indicated by reduced extranuclear TFAM signals. Given that cGAS-STING activation may also be potentiated by nuclear DNA damage, γ-H2AX foci were evaluated as a marker of DNA double-strand breaks. Consistently, TC/pSeSe effectively suppressed γ-H2AX accumulation induced by LPS challenge (Fig. S17). The results above indicated that TC/pSeSe attenuated mtDNA cytoplasmic escape and nuclear DNA damage to modulate cGAS-STING signaling.

Because mitochondrial homeostasis is a key determinant of mtDNA functionality, we further assessed whether TC/pSeSe preserves mitochondrial structure and function. Immunofluorescence staining of TOM20 was first used to assess mitochondrial morphology. Remarkable fragmentation and disorganization of the mitochondrial network in the LPS group were observed (Fig. 7h). The morphological changes were partially restored by pSeSe. By comparison, TC/pSeSe fully restored the well-orchestrated, interconnected mitochondrial architecture. Consistently, TEM analysis revealed that TC/pSeSe reversed the disrupted and fragmented mitochondrial structures induced by LPS and restored mitochondrial ultrastructure, including intact cristae and double-membrane boundaries (Fig. 7i).

Mitochondrial biofunction recovery was further assessed via JC-1 staining and ATP production. JC-1 aggregates (red fluorescence), indicative of intact mitochondrial membrane potential (MMP), were significantly increased in the TC/pSeSe group compared to the LPS group, in which MMP collapse was reflected by JC-1 monomer predominance (green fluorescence) (Fig. 7j and Fig. S18). Additionally, ATP quantification demonstrated that TC/pSeSe restored mitochondrial energy metabolism from 2.0 ± 0.1 mM in LPS group to 5.7 ± 0.4 mM in TC/pSeSe group (Fig. S19). Notably, TC/pSeSe consistently outperformed pSeSe in preserving mitochondrial integrity and bioenergetic function, highlighting the contribution of TC-mediated mitochondrial targeting properties. Collectively, these findings suggest that TC/pSeSe alleviates oxidative stress-induced mitochondrial dysfunction by preserving mitochondrial morphology, stabilizing bioenergetic function, and preventing mtDNA cytosolic release (Fig. 7k). Through these mechanisms, TC/pSeSe effectively suppresses cGAS-STING activation and downstream inflammation, offering a promising strategy for periodontitis treatment.

3.6. Comprehensive therapeutic effect of TC/pSeSe nanocomposite in rat periodontitis model

To further validate the therapeutic efficacy of TC/pSeSe in vivo, a rat ligature-induced periodontitis model was established (Fig. 8a). Silk ligatures were placed around the rat maxillary second molar and retained for 2 weeks. After removal of the ligature, rats with periodontitis received topical injection of PBS, pSeSe or TC/pSeSe every other day for another two weeks. Micro-computed tomography (micro-CT) analysis 3D reconstruction revealed remarkable alveolar bone resorption in the PBS group (Fig. 8b, red area). In comparison, pSeSe partially reduced bone loss area yet the alveolar bone crest remained insufficient to provide periodontal tooth support. TC/pSeSe substantially restored alveolar bone height, as shown by reduced distance from the cementoenamel junction (CEJ) to the alveolar bone crest (ABC), indicating a significant reversal of bone loss (Fig. 8c). Parameters including trabecular number (Tb. N), trabecular thickness (Tb. Th), trabecular separation (Tb. Sp) and bone volume to tissue volume ratio (BV/TV) were also greatly improved by TC/pSeSe treatment. Histological evaluations further corroborated the micro-CT findings. Hematoxylin and eosin (H&E) staining together with Masson staining demonstrated substantial alveolar bone resorption and notable collagen loss in the PBS group (Fig. 8d-e). TC/pSeSe injection promoted new bone formation, restored alveolar bone height and volume, and repaired densely organized collagen fibers (blue), suggesting robust extracellular matrix remodeling and periodontal tissue recovery. Although pSeSe showed modest efficacy, we postulate that the TC coating may contribute to the enhanced in vivo efficacy by improving local retention under salivary flushing (Fig. 2f), thereby exhibiting superior therapeutic effects in this context.

Fig. 8.

In vivo evaluation of bone regeneration in rat ligature-induced periodontitis model. a) Schematic diagram illustrating the establishment of the periodontitis model and subsequent administration of TC/pSeSe. b) Micro-computed tomography (micro-CT) analysis of maxillary bone structure including 3D lingual and 2D sagittal views. Scale bar: 2 mm. c) Quantitative assessment of bone resorption area, cementoenamel junction to alveolar bone crest (CEJ-ABC) distance, trabecular number (Tb.N), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp), and bone volume to tissue volume ratio (BV/TV) (n = 3). d) Representative hematoxylin and eosin (H&E) stained histological sections. Scale bar: 200 μm. e) Representative Masson's trichrome staining images. Scale bar: 500 μm and 100 μm. f) Immunohistochemical (IHC) staining of osteogenic markers osteocalcin (OCN) and osteopontin (OPN) in periodontal tissues. Scale bar: 50 μm and 10 μm. Data are represented as mean ± standard deviation.

To assess osteogenic protein expression, immunohistochemical (IHC) staining for osteoblastogenic markers OCN and OPN was performed. As shown in Fig. 8f, both OCN and OPN expression levels were markedly decreased in PBS group, indicating dampened matrix mineralization. In contrast, the TC/pSeSe group showed significantly increased OCN and OPN expression, confirming the in vivo osteogenic potential of the nanocomposite through matrix maturation. We further adopted tartrate-resistant acid phosphatase (TRAP) staining that marks osteoclast activity. A substantial number of osteoclasts on the alveolar surface was observed in the PBS group, whereas the TC/pSeSe group exhibited a pronounced reduction in osteoclast numbers within the alveolar region (Fig. 9a). These results are consistent with osteogenic properties of TC/pSeSe exhibited in vitro.

Fig. 9.

In vivo evaluation of bone regeneration, STING pathway suppression, antioxidation and systemic biocompatibility in rat ligature-induced periodontitis model. a) Tartrate-resistant acid phosphatase (TRAP) staining images, with black arrows indicating osteoclast presence. Scale bar: 500 μm and 100 μm. b) Immunofluorescence staining of p-STING to assess activation of the cGAS-STING signaling pathway. Cell nuclei were counterstained with DAPI (blue). Scale bar: 50 μm. c) Quantification of pro-inflammatory cytokines IFN-β, TNF-α, IL-6, and IL-1β levels by ELISA (n = 6). d) Gene expression of Ifnb and ISGs after different treatments (n = 3). e) Measurement of MDA levels (n = 3). f) Intracellular superoxide staining by dihydroethidium (DHE, red) conducted on frozen gingival soft-tissue sections. Cell nuclei were counterstained with DAPI (blue). Scale bar: 200 μm. g) Histological evaluation of systemic biocompatibility by H&E staining of major organs, including the heart, liver, spleen, lung, and kidney. Scale bar: 100 μm. Data were represented as mean ± standard deviation.

To quantitatively evaluate the anti-inflammatory effects of TC/pSeSe in vivo, and given its pronounced modulatory action on the cGAS-STING pathway in vitro, we examined the expression of p-STING in periodontal tissues by immunofluorescence staining. The results were consistent with our in vitro findings. The PBS control group exhibited apparent red fluorescence, indicative of elevated p-STING expression, whereas the TC/pSeSe-treated group showed markedly weaker fluorescence signals (Fig. 9b). Effective inflammation control was further confirmed by ELISA analysis of systemic cytokines in rat serum. Pro-inflammatory cytokines, including IFN-β, TNF-α, IL-6, and IL-1β, were significantly upregulated in the PBS group due to the aberrant immune response associated with periodontitis (Fig. 9c). In contrast, these cytokines were notably reduced following pSeSe and TC/pSeSe treatment, demonstrating the potent anti-inflammatory activity of TC/pSeSe in vivo. As IFN-β is a downstream effector of the cGAS-STING pathway, its decline supports cGAS-STING pathway attenuation. Consistently, qPCR analysis of Ifnb and ISGs confirmed significant downregulation, corroborating the attenuation of cGAS-STING pathway by TC/pSeSe (Fig. 9d). The ROS-scavenging capability of TC/pSeSe in vivo was evaluated using MDA levels and the dihydroethidium (DHE) probe as oxidative stress indicators, which were markedly reduced in treated groups (Fig. 9e-f). Antibacterial efficacy was further validated by CFU analysis of oral microbial samples, revealing significantly reduced bacterial burden following TC/pSeSe treatment (Fig. S20). Finally, systematic biocompatibility of TC/pSeSe was evaluated by histopathological examination of major organs including heart, lung, liver, kidney, and spleen using H&E staining (Fig. 9g). No discernible pathological alterations were observed across all treatment groups, indicating an absence of systemic toxicity post-injection.

Collectively, these findings demonstrate that TC/pSeSe effectively mitigates periodontal inflammation, preserves alveolar bone structure and restores redox homeostasis, highlighting its therapeutic potential for periodontitis treatment.

4. Discussion

The open, bacteria-colonized periodontitis microenvironment is continuously exposed to external stimuli and mitochondria serve as a central upstream hub in oxidative stress-driven periodontitis pathogenesis [45]. However, therapeutic strategies that precisely target this subcellular node remain insufficiently explored. In this study, we developed a mitochondria-centered ROS-scavenging nanocomposite system that modulates periodontal redox balance at the upstream source. By harnessing the mitochondria-targeting capability of TC coating, TC/pSeSe effectively targets the organelle central to oxidative stress and restores redox balance through its ROS-scavenging diselenide polymer core, thereby suppressing cGAS-STING-mediated inflammatory reactions. Meanwhile, the TC layer functions as a regulatory interface that modulates selenium release while enhancing local retention and antibacterial performance, collectively enabling sustained therapeutic efficacy within the complex oral microenvironment.

Periodontitis is initiated by bacterial infection, which sustains inflammatory and oxidative stress cascades [29]. Effective infection control is therefore a prerequisite for subsequent redox modulation and bone regeneration. In this study, TC/pSeSe significantly suppressed periodontal bacterial growth and biofilm formation, thereby establishing a foundation for downstream tissue repair. The antibacterial activity of TC/pSeSe is likely attributable to the combined actions of selenium-based bactericidal effects and the cationic nature of the TC coating. Selenium has been reported to disrupt bacterial membrane integrity and induce damage to intracellular proteins and DNA [46]. In parallel, the positively charged TC coating promotes bacterial adhesion and membrane perturbation, acting cooperatively with selenium-mediated effects to enhance antibacterial efficacy [47,48].

Alveolar bone resorption and prolonged inflammation are the two hallmark clinical manifestations of periodontitis [49]. In this study, TC/pSeSe displayed effective osteogenic properties that can be attributed to combined redox modulation and selenium-mediated osteogenic support. Previous studies have demonstrated that selenium enhances bone tissue development, stimulate the osteogenic differentiation of BMSCs, and protect against oxidative stress-induced inhibition of osteogenesis [15,50]. In parallel, selenium nanoparticles suppress osteoclastogenesis and prevent pathological bone destruction through activation of selenoproteins [17]. Consistent with these findings, the sustained release of selenium from TC/pSeSe is likely one of the key contributors to its pronounced osteogenic activity. Compared with the pSeSe group alone, the TC coating guarantees a more controlled and ROS-responsive selenium release profile, thereby enhancing the overall osteogenic performance of TC/pSeSe.

In the context of periodontitis, excessive ROS accumulation and iron metabolism disorder are fundamental causes for ferroptosis [51]. During ferroptosis, Fe²⁺ and ROS form a positive feedback loop in which elevated ROS further enhances iron accumulation through oxidative stress and ferritin degradation [52]. Selenium has been shown to play a protective role toward ferroptosis by driving GPX4 expression and modulating GPX4 activity [53]. In addition, previous studies have demonstrated that selenium supplementation enhances GPX4 expression and activates the Nrf2 antioxidant pathway, thereby suppressing lipid peroxidation and ferroptosis [54,55]. In line with these reports, our nanomaterial significantly upregulated GPX4 and Nrf2 expression and downregulated ACSL4, suggesting that the selenium-containing polymer may modulate ferroptosis through dual mechanisms of selenium ion release and ROS reduction. The modulation of intracellular redox balance and inhibition against ferroptosis observed in TC/pSeSe provides a promising strategy to alleviate oxidative stress-related damage in periodontal disease.

Despite the encouraging outcomes, this study has several limitations. Future work will further dissect by performing RNA-seq analysis to identify pathway-related transcripts and strengthen mechanistic evidence. Moreover, comprehensive long-term safety evaluations and stability assessments will be essential to facilitate clinical translation, particularly given the chronic and recurrent nature of periodontitis.

Numerous studies have employed mitochondria-targeting moieties in anticancer nano-therapy and neurodegenerative diseases [56,57]. However, their application in chronic inflammatory diseases such as periodontitis remains limited. This study presents a mitochondria-centered nanocomposite platform with integrated antibacterial, osteogenic, antioxidative, and anti-inflammatory functions, through a unified mechanistic framework. By targeting mitochondrial dysfunction as an upstream driver of oxidative stress and inflammation, TC/pSeSe is distinguished from conventional antioxidant systems that primarily regulate bulk ROS without subcellular specificity and offers a highly promising strategy for the treatment of periodontitis and other oxidative stress-related disorders.

CRediT authorship contribution statement

Ning Huang: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis, Conceptualization. Lingyan Cao: Writing – review & editing, Writing – original draft, Supervision, Methodology, Funding acquisition, Conceptualization. Yue Xu: Writing – review & editing, Methodology, Investigation, Formal analysis, Conceptualization. Lisha Pan: Writing – review & editing, Investigation, Formal analysis. Ao Zheng: Writing – review & editing, Methodology, Funding acquisition, Conceptualization. Xiao Wang: Writing – review & editing, Supervision, Methodology, Funding acquisition, Conceptualization. Xinquan Jiang: Writing – review & editing, Supervision, Methodology, Funding acquisition, Conceptualization.

Ethics approval and consent to participate

All institutional and national guidelines for the care and use of laboratory animals were followed. All experimental protocols involving animals were approved by the Institutional Animal Care and Use Committee and followed the procedure for Animal Experimental Ethical Inspection of the Ninth People's Hospital, which is affiliated with Shanghai Jiao Tong University School of Medicine. Animal license No. is SCXK (Shanghai) 2023-0041.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Xinquan Jiang reports was provided by the National Natural Science Foundation of China. Lingyan Cao reports financial support was provided by the National Natural Science Foundation of China. Ao Zheng reports financial support was provided by the National Natural Science Foundation of China. Xinquan Jiang reports financial support was provided by the National Key R&D Program of China. Xinquan Jiang reports financial support was provided by the Fundamental Research Funds for the Central Universities. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article: