Turmeric-derived nanovesicles reprogram pathological NETosis to mitigate inflammatory bone loss Authors: Ming Wang, Kamoran Tuerhong, Ying Xie, Qi Huang, Qianyu Zhang, Kehao Liu, Huiya Wang, Xiaoke Su, Xiaoyang Luo, Xinzhi Jiang, Haixia Dang, Danfeng Shen, Yuzhou Li, Sheng Yang Published: 2026 ABSTRACT BackgroundInflammatory bone destruction is a progressive pathological process driven by persistent immune-mediated inflammation and osteoclast-dependent bone resorption, exemplified by periodontitis. Increasing evidence indicates that dysregulated inflammatory microenvironments and aberrant osteoclast activation form a self-amplifying loop that accelerates bone loss. Neutrophils are central to this process through the formation of neutrophil extracellular traps (NETs), while excessive NETosis further amplifies inflammation and promotes bone resorption. Although NETosis has emerged as a promising therapeutic target, current strategies largely rely on single-molecule inhibitors, which often suffer from limited targeting efficiency, poor bioavailability, and uncertain long-term safety. Therefore, developing biomimetic delivery-based, multi-component and multi-target therapeutic strategies to precisely regulate NET-associated signalling pathways remains a major challenge.ResultsSingle-cell transcriptomic analysis of Gene Expression Omnibus (GEO) datasets identified aberrant NET formation as a key pathogenic feature and a potential therapeutic target in periodontitis. To address this challenge, we developed a natural biomimetic nanodelivery system based on turmeric-derived nanovesicles (TNVs). This platform may attenuate inflammatory responses and NET formation through vesicle-mediated delivery of intrinsic bioactive cargo. TNVs possess a plant-derived lipid bilayer structure enriched with bioactive compounds, exhibiting excellent biocompatibility and efficient uptake by neutrophils. Functional studies demonstrated that TNVs markedly suppressed NET formation, accompanied by reduced intracellular reactive oxygen species (ROS) levels, decreased expression of citrullinated histone H3 (CitH3), myeloperoxidase (MPO), neutrophil elastase (ELANE), and reduced extracellular DNA release. Transcriptomic profiling further confirmed that TNVs downregulated NET-associated pathways while enhancing antioxidant and anti-inflammatory responses. In a murine periodontitis model, TNVs preferentially accumulated at inflamed sites, inhibited NET formation and osteoclastogenesis, alleviated inflammation, reduced bone loss, and preserved bone microarchitecture.ConclusionTNVs represent a promising plant-derived biomimetic nanoplatform that preferentially accumulates in inflamed tissues and exerts immunomodulatory effects. By integrating intrinsic bioactive cargo with efficient delivery capacity, TNVs offer a compelling therapeutic strategy for inflammation-induced bone loss, including periodontitis.Graphical abstract Supplementary InformationThe online version contains supplementary material available at 10.1186/s12951-026-04682-x. Background Inflammatory bone destruction is a pathological process characterised by persistent immune-mediated inflammation and progressive bone resorption [1]. It occurs in a variety of immune-mediated disorders, including periodontitis [2], rheumatoid arthritis [3], and peri-implantitis [4], ultimately leading to structural bone damage and functional tissue loss. Increasing evidence suggests that dysregulated inflammatory microenvironments and aberrant osteoclast activation form a self-amplifying feedback loop that drives pathological bone resorption [5]. Therefore, precisely modulating inflammatory immune responses to interrupt inflammation-driven bone loss remains a major challenge in the treatment of bone-related diseases [1, 6]. Periodontitis represents a typical model of inflammatory bone destruction and is a chronic inflammatory disease initiated by plaque biofilms [7]. Its pathological hallmarks include sustained immune activation and progressive alveolar bone resorption, which may ultimately result in tooth mobility and tooth loss [8]. Although conventional therapies such as mechanical debridement and anti-inflammatory treatments can partially alleviate clinical symptoms, their ability to restore inflammatory microenvironmental homeostasis and prevent progressive bone loss remains limited [9]. Moreover, current therapeutic strategies often face challenges including antibiotic resistance, recurrent inflammation, and insufficient bone regeneration [8]. These limitations highlight the need for innovative strategies targeting the key inflammatory mechanisms underlying bone destruction. Recent studies have highlighted the central role of neutrophils in inflammation-associated bone loss [10]. As one of the earliest recruited innate immune cells, neutrophils regulate inflammatory responses not only through pathogen clearance but also through the formation of NETs [11, 12]. However, excessive or dysregulated NET formation can amplify inflammatory signalling, exacerbate tissue injury, and promote osteoclast-mediated bone resorption [11–13]. NET formation is regulated by several key enzymes, including myeloperoxidase (MPO), neutrophil elastase (ELANE) and peptidylarginine deiminase 4 (PAD4) [11, 13, 14]. PAD4-mediated histone citrullination drives chromatin decondensation, facilitating NET release [11, 13, 14]. Importantly, reactive oxygen species (ROS) signalling plays a crucial role in regulating NETosis [11]. Suppression of ROS generation has been shown to effectively inhibit NET release by preventing ELANE- and MPO-mediated chromatin decondensation [11]. Dysregulation of the ROS–NETosis axis has therefore emerged as a key pathogenic mechanism linking innate immune imbalance with inflammation-driven bone destruction [11, 13]. Although NETosis represents a promising therapeutic target for inflammatory diseases, current strategies mainly rely on single-molecule inhibitors [15]. Such approaches frequently exhibit limited targeting efficiency, poor bioavailability, and uncertain long-term safety [16]. Consequently, developing therapeutic strategies capable of modulating NET-associated signalling through multi-component and multi-target mechanisms remains an important challenge for the treatment of inflammatory bone destruction. Extracellular vesicles (EVs) have attracted considerable attention as natural nanocarriers due to their roles in intercellular communication and immune regulation [9, 17–20]. However, mammalian-derived EVs face practical limitations, including limited availability, high production costs, and potential immunogenicity [21–23]. Recently, plant-derived extracellular vesicles (PDEVs) have emerged as a promising natural nanocarrier system [21–24]. With favorable biocompatibility, low immunogenicity, and the intrinsic ability to encapsulate diverse bioactive molecules, PDEVs show substantial therapeutic potential for inflammatory and bone-related diseases [25, 26]. Beyond serving as delivery carriers, PDEVs can act as biologically active nanoregulatory platforms that modulate complex immune responses via multi-component, multi-target mechanisms [25, 27–31]. These properties make PDEVs particularly suitable for achieving multi-component and multi-target regulation of complex inflammatory signalling pathways. Turmeric is a medicinal and edible plant whose bioactive constituents exhibit well-established anti-inflammatory and antioxidant properties [32]. Our previous study demonstrated that curcumin effectively suppresses NET formation and attenuates periodontal inflammation [16]. Building on these findings, we propose that turmeric-derived nanovesicles (TNVs), owing to their intrinsic bioactive cargo and natural nanovesicular architecture, may serve as a unique biomimetic platform that integrates efficient delivery with immunomodulatory functions. Accordingly, we hypothesised that TNVs may function as a natural biomimetic nanosystem capable of alleviating inflammatory bone destruction through the regulation of NET formation and inflammatory responses. In this study, we employed periodontitis as a representative model of inflammatory bone destruction. We first performed single-cell transcriptomic analysis of gingival tissues from patients with periodontitis to characterise the immune microenvironment and identify NET-associated signalling pathways during disease progression. Turmeric-derived nanovesicles (TNVs) were subsequently isolated and characterised, and their effects on ROS levels, NET formation and related inflammatory responses were systematically investigated. Through integrated in vitro experiments, in vivo disease models, and transcriptomic analyses, this study elucidated the immunomodulatory mechanisms of TNVs in inflammatory bone destruction and demonstrated their potential as a natural nanotherapeutic strategy for inflammation-associated bone diseases (Scheme 1). Scheme 1Turmeric-Derived Nanovesicles (TNVs) Suppress NET Formation to Alleviate Inflammatory Bone Loss. This scheme outlines the experimental workflow and the osteoprotective mechanism of TNVs in inflammatory bone-destructive diseases, with periodontitis serving as a representative model. Single-cell RNA sequencing (scRNA-seq) identified neutrophil extracellular trap (NET) formation as a key driver of disease progression. Liquid chromatography–mass spectrometry (LC–MS) profiling, combined with molecular docking, was employed to screen turmeric-derived bioactive compounds targeting NET-associated pathways. TNVs are internalised via the endosomal–lysosomal pathway, carrying intrinsic bioactive cargo including curcumin, (R)-ar-turmerone, and curcumenol, and may suppress NET formation through modulation of intracellular ROS levels. TNVs also preferentially accumulate at inflamed sites through vascular extravasation. In vivo, TNVs suppress pathological NET formation by downregulating PAD4, ELANE, and MPO, thereby reducing inflammatory cytokine production, inhibiting osteoclastogenesis, and attenuating inflammation-induced bone resorption Turmeric-Derived Nanovesicles (TNVs) Suppress NET Formation to Alleviate Inflammatory Bone Loss. This scheme outlines the experimental workflow and the osteoprotective mechanism of TNVs in inflammatory bone-destructive diseases, with periodontitis serving as a representative model. Single-cell RNA sequencing (scRNA-seq) identified neutrophil extracellular trap (NET) formation as a key driver of disease progression. Liquid chromatography–mass spectrometry (LC–MS) profiling, combined with molecular docking, was employed to screen turmeric-derived bioactive compounds targeting NET-associated pathways. TNVs are internalised via the endosomal–lysosomal pathway, carrying intrinsic bioactive cargo including curcumin, (R)-ar-turmerone, and curcumenol, and may suppress NET formation through modulation of intracellular ROS levels. TNVs also preferentially accumulate at inflamed sites through vascular extravasation. In vivo, TNVs suppress pathological NET formation by downregulating PAD4, ELANE, and MPO, thereby reducing inflammatory cytokine production, inhibiting osteoclastogenesis, and attenuating inflammation-induced bone resorption Materials and methods Single-cell RNA-Seq data processing and analysis Single-cell RNA sequencing data were obtained from the Gene Expression Omnibus (GEO) database (GSE171213, platform GPL24676), comprising gingival tissues from five patients with periodontitis and four healthy controls. Expression matrices were processed using Seurat (v5.1.0). Cells with fewer than 200 detected genes, as well as genes expressed in fewer than three cells, were excluded from the analysis. Quality control was conducted based on nFeature_RNA, nCount_RNA, and mitochondrial gene percentage (percent.mt), retaining cells with nFeature_RNA between 200 and 5,500, nCount_RNA < 40,000, and percent.mt < 15%. After filtering, 15,269 high-quality cells were retained. Data were normalised using the LogNormalize method, followed by identification of highly variable genes and scaling with regression of nCount_RNA and percent.mt. Batch effects were minimised using Harmony-based integration. Principal component analysis (PCA) was performed for dimensionality reduction, and the optimal number of principal components was determined using ElbowPlot and JackStraw analyses. Cells were clustered using the FindNeighbors and FindClusters algorithms (resolution = 0.2), followed by visualisation with t-distributed stochastic neighbour embedding (t-SNE). Cell-type annotation was carried out based on established marker genes. The ggplot2 package was utilised to display cell-type proportions across samples. Differences in cellular composition between groups were assessed using Fisher’s exact test, with individual samples as statistical units. Differentially expressed genes in neutrophils were determined using FindMarkers (|avg_log2FC| > 0.5, adjusted P < 0.05). Subsequently, Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analyses were performed to explore the associated biological pathways. Isolation and extraction of turmeric-derived nanovesicles Fresh turmeric (sourced from Guangxi, China) was washed, homogenised, and subjected to sequential differential centrifugation at 1,000, 2,000, 3,000, and 10,000 g for 10, 20, 30, and 60 min, respectively [33]. The supernatant was then ultracentrifuged at 150,000 g for 1.5 h using an Optima XPN-100 ultracentrifuge (Beckman Coulter). The resulting pellet was resuspended in PBS by ultrasonic dispersion. The suspension was then layered onto a discontinuous sucrose gradient (8%, 30%, 45%, and 60%) and centrifuged again at 150,000 g for 1.5 h. TNVs located at the 30–45% interface were collected, diluted with PBS, and subjected to another round of ultracentrifugation to remove residual sucrose. The resulting pellet was finally resuspended in PBS. Protein concentration was measured using a BCA assay, and the samples were stored at −80 °C. Characterisation of turmeric-derived plant nanovesicles TNVs were prepared in PBS at a concentration of 1 mg/mL, resuspended by ultrasonication (40 kHz), and filtered through a 0.22 μm membrane. Particle size distribution and zeta potential were determined using a nanoparticle tracking analyser (ZetaView PMX120, Particle Metrix, Germany). For transmission electron microscopy (TEM), 5 µL of the sample was placed onto carbon-coated copper grids, negatively stained with phosphotungstic acid for 30 s, air-dried, and subsequently observed using a transmission electron microscope (JEM-1400plus, JEOL Ltd., Japan) operated at 100 kV. The morphology of TNVs was analysed, and their protein composition was evaluated by Coomassie Brilliant Blue staining (Solarbio, China). LC-MS analysis The metabolite composition of TNVs was analysed using liquid chromatography–mass spectrometry (LC–MS) (Shanghai Majorbio Bio-pharm Technology Co., Ltd., China; instrumental parameters are provided in Supplementary Information. Raw data were processed using Progenesis QI (Waters, USA) for baseline correction, peak extraction, alignment, and integration, generating a data matrix containing retention time, mass-to-charge ratio (m/z), and peak intensity. MS and MS/MS spectra were matched against the MJBIOTCM database with a mass error of < 10 ppm. Metabolite identification was performed using accurate mass data together with spectral similarity scores. Molecular docking analysis Three-dimensional structures of curcumin, curcumenol, and (R)-ar-turmerone were obtained from PubChem in SDF format. Crystal structures of human PAD4, ELANE, and MPO were retrieved from the RCSB Protein Data Bank and cross-checked using UniProt. Molecular docking was performed using AutoDock Vina (v1.1.2), as previously described [34]. Protein structures were pre-processed by removing water molecules and co-crystallised ligands and adding hydrogen atoms, while ligands were structurally optimised. Docking was conducted within defined active-site regions. Binding energy was used to evaluate affinity, and interaction modes were subsequently analysed. Cell culture MC3T3-E1 cells obtained from the Cell Bank of the Chinese Academy of Sciences were cultured in α-MEM (Gibco, USA) supplemented with 10% FBS (Gibco, USA) and 1% penicillin–streptomycin (Solarbio, China) at 37 °C in a humidified atmosphere containing 5% CO₂. The medium was changed every two days, and cells were subcultured at 80–90% confluence. Primary neutrophils were isolated from 6–8-week-old male C57BL/6 mice. Femurs and tibias were harvested under sterile conditions, and bone marrow cells were flushed with RPMI-1640 (Gibco, USA) containing 10% FBS (Gibco, USA). Neutrophils were purified using a commercial isolation kit (P8550, Solarbio, China) by density gradient centrifugation. HL-60 cells were grown in IMDM (Gibco, USA) supplemented with 20% FBS (Gibco, USA), followed by differentiation into neutrophil-like cells using 1.3% DMSO for 3–5 days. Live/Dead cell staining MC3T3-E1 cells were seeded in 24-well plates at a density of 1 × 10⁴ cells per well and cultured for 24 h before treatment with TNVs (10, 20, 30 and 50 µg/mL) for 4, 8–24 h. Cell viability was evaluated using a Live/Dead staining kit (C2015S, Beyotime, China). Live cells were labelled with Calcein AM (green fluorescence), while dead cells were stained with propidium iodide (red fluorescence). CCK-8 assay MC3T3-E1 cells were seeded in 96-well plates at a density of 5 × 10³ cells per well and cultured for 24 h prior to treatment with TNVs (10, 20, 30 and 50 µg/mL) for 8, 24–48 h. Cell viability was determined using a CCK-8 assay (HY-K0301, MedChemExpress, USA), and the absorbance was recorded at 450 nm. All experiments were conducted in quadruplicate. Wound healing assay MC3T3-E1 cells were seeded in 12-well plates at a density of 5 × 10⁴ cells per well and cultured to confluence. Two perpendicular scratches were created using a sterile 200 µL pipette tip. After washing with PBS to remove detached cells, the cultures were treated with TNVs at concentrations of 10, 20, and 30 µg/mL. Images were captured at 0, 6, 12, 24, and 48 h using an optical microscope (DMi1, Leica Microsystems, Germany). Wound closure was quantified using ImageJ software to assess cell migratory capacity. All experiments were performed in triplicate. In addition, a Transwell migration assay was conducted to further evaluate the effect of TNVs on MC3T3-E1 cell migration, with detailed procedures provided in Supplementary Information. Cellular uptake assay MC3T3-E1 and dHL-60 cells were seeded onto glass coverslips in 24-well plates at densities of 2 × 10⁴ and 2 × 10⁵ cells per well, respectively, and stimulated with 100 ng/mL LPS (Beyotime, China). TNVs were labelled with PKH26 (Sigma-Aldrich, USA) and purified using dye removal columns (E0015-5T, Hubei Milian Biotechnology, China). Labelled TNVs (30 and 50 µg/mL) were co-incubated with cells for 0–24 h. After PBS washing and fixation, cells were stained with phalloidin (40736ES75, Yeasen, China) and DAPI (Beyotime, C1006, China), mounted, and imaged using confocal microscopy (Leica TCS SP8) to assess intracellular distribution. For uptake quantification, cells were harvested and analysed by flow cytometry, and PKH26 fluorescence was quantified using FlowJo (v10.8.1). Lysosomal co-localisation analysis MC3T3-E1 cells and primary neutrophils were stimulated with 100 ng/mL LPS for lysosomal co-localisation analysis. PKH26-labelled TNVs (50 µg/mL) were co-incubated with the cells for 2–4 h. After washing with PBS, nuclei were stained with Hoechst 33258 (C1011, Beyotime, China) and lysosomes with LysoTracker (C1047S, Beyotime, China). Fluorescence images were acquired using a confocal microscope (TCS SP8, Leica Microsystems, Germany). Co-localisation between TNVs and lysosomes was quantitatively analysed using ImageJ software to evaluate intracellular trafficking pathways. Inhibitory effects of different concentrations of TNVs on NET formation Primary neutrophils were seeded onto poly-L-lysine-coated coverslips in 24-well plates (1 × 10⁶ cells per well) and subsequently stimulated with 5 µM ionomycin (I3909, Sigma-Aldrich, USA) or 70 nM phorbol 12-myristate 13-acetate (PMA; HY-18739, MCE) to induce the formation of NETs. Cells were simultaneously treated with PKH26-labelled TNVs (10, 20, and 30 µg/mL) and incubated for 4 h. Following PBS washing, cells were fixed with 4% paraformaldehyde, permeabilised using 0.5% Triton X-100, and blocked with 5% donkey serum. Samples were then incubated overnight at 4 °C with an anti-CitH3 antibody (ab5103, Abcam, UK), followed by incubation with a fluorescent secondary antibody (A-11034, Invitrogen, USA) for 1 h. Nuclei were subsequently counterstained with DAPI (C1006, Beyotime, China), and images were captured using confocal microscopy. For quantitative analysis, neutrophils were seeded at 3 × 10⁶ cells per well and co-treated with ionomycin and TNVs for 4 h. Cells were collected, resuspended in PBS, and stained with SYTOX Green (R37168, Invitrogen, USA) for 10 min. Fluorescence intensity was analysed by flow cytometry (CytoFLEX, Beckman Coulter, USA) using the 488 nm channel to quantify extracellular DNA release. Effects of different TNV concentrations on ROS production Primary neutrophils were seeded in 24-well plates (1 × 10⁶ cells per well) and treated with 200 µM H₂O₂ in the presence or absence of TNVs (10, 20, and 30 µg/mL) for 2 h. Cells were incubated with DCFH-DA (S0033S, Beyotime, China) for 20 min at 37 °C, followed by nuclear staining with Hoechst 33258 (C1011, Beyotime, China). Intracellular ROS levels were visualised using confocal microscopy (TCS SP8, Leica Microsystems, Germany) and quantified by flow cytometry (CytoFLEX, Beckman Coulter, USA). Investigation of ROS involvement in TNV-mediated NETosis inhibition To investigate the role of ROS in TNV-mediated regulation of NETosis, primary neutrophils were divided into seven groups: (1) control (CTR), (2) ION, (3) ION + H₂O₂, (4) ION + TNVs, (5) ION + NAC, (6) ION + TNVs + NAC, and (7) ION + TNVs + H₂O₂. Cells were seeded at 1 × 10⁶ cells/well in 24-well plates containing poly-L-lysine-coated coverslips. TNVs (30 µg/mL), H₂O₂ (200 µM), or N-acetyl-L-cysteine (NAC, 5 mM; HY-B0215, MedChemExpress, USA) were added as indicated. Groups 2–7 were stimulated with ionomycin (ION; 5 µM, I3909, Sigma-Aldrich, USA) for 4 h at 37 °C to induce NETosis [35]. Extracellular DNA release was detected using SYTOX Green staining (S7020, Invitrogen) [35, 36]. NET formation was visualized by confocal microscopy (TCS SP8, Leica Microsystems, Germany), and the SYTOX Green-positive area was quantified using ImageJ software. All experiments were independently performed in triplicate. Inhibitory effects of TNVs and free bioactive compounds on NET formation To assess the effects of TNVs and their major bioactive components on NET formation, primary neutrophils were co-treated with 30 µg/mL TNVs or individual compounds at concentrations equivalent to those in TNVs, including curcumin (HY-N0005, MedChemExpress, USA), (R)-ar-turmerone (HY-N6703, MedChemExpress, USA), and curcumenol (HY-N2259, MedChemExpress, USA), in the presence of 5 µM ionomycin (I3909, Sigma-Aldrich, USA) for 4 h. TNV samples were diluted with ethanol, subjected to ultrasonic treatment for 30 min (200 W, 30 °C), and centrifuged to collect the supernatant for absorbance measurement. All procedures were conducted under light-protected conditions. The concentration of curcumin in TNVs was determined by UV–Vis spectrophotometry (UV2500, Techcomp, China) at 425 nm using an external calibration curve constructed with a curcumin analytical standard (HY-N0005R, MedChemExpress, USA) (Fig. S1). All measurements were performed in six replicates, and the results were expressed as mean values. Due to the lack of certified reference standards, the levels of curcumenol and (R)-ar-turmerone were semi-quantitatively estimated based on their relative LC–MS peak area ratios to curcumin. Individual compounds were subsequently tested at concentrations corresponding to their estimated levels in TNVs. NETosis was assessed using a double immunofluorescence staining method with tyramide signal amplification (TSA). Cells were fixed, permeabilised, blocked, and incubated overnight with an anti-CitH3 antibody (ab5103, Abcam, UK), followed by incubation with an HRP-conjugated secondary antibody and signal development using TYR-480 Plus. After antibody stripping and re-blocking, samples were incubated with an anti-ELANE antibody (ab314916, Abcam, UK) and visualised using the TYR-620 Plus substrate. Nuclei were counterstained with DAPI (C1006, Beyotime, China), and images were captured using a confocal microscope (TCS SP8, Leica Microsystems, Germany). Fluorescence intensity was subsequently quantified using ImageJ software. For quantitative analysis, neutrophils were co-incubated with TNVs or individual compounds in the presence of ionomycin for 2 h, followed by staining with SYTOX Green (R37168, Invitrogen, USA). Extracellular DNA release was measured by flow cytometry. All experiments were independently repeated at least three times. To assess the effects of TNVs, individual bioactive components, and their mixture on NETosis, primary neutrophils were stimulated with ionomycin (ION; I3909, Sigma-Aldrich, USA) at 5 µM to induce NET formation. Cells were subsequently stained with SYTOX Green (S7020, Invitrogen), and extracellular DNA release was visualised by confocal microscopy (TCS SP8, Leica Microsystems, Germany). The SYTOX Green-positive area was quantified using ImageJ software. RNA sequencing assay Primary neutrophils were seeded in 24-well plates at 5 × 10⁶ cells per well. Cells were treated with 2 µM ionomycin alone (ION group) or with 2 µM ionomycin plus 30 µg/mL TNVs (TNVs group), in triplicate. After 2 h, total RNA was extracted and subjected to high-throughput RNA sequencing by Novogene (Beijing, China). Ligature-induced periodontitis model and TNV treatment in C57BL/6 mice A ligature-induced periodontitis (LIP) model was established on bilateral second molars of male C57BL/6 mice (6–8 weeks). Mice were anaesthetised by intraperitoneal injection of tribromoethanol (100 µL/10 g body weight), after which a 5 − 0 silk ligature was positioned around the cervical region of the second molars. After modelling, mice received intravenous TNVs (10 mg/kg, once daily) for 7 days (LIP+TNVs) or PBS (LIP) under isoflurane anaesthesia. Control mice (CTR) received no ligature or treatment. All animals were euthanised on Day 7. The TNV dose was based on previous studies showing that intravenous administration of 10 mg/kg plant-derived nanovesicles was well tolerated without tissue toxicity, thereby supporting the safety of the dosing approach used in this study [37–39]. In vivo biodistribution analysis A LIP model was established in C57BL/6 mice, followed by 24 h of fasting. TNVs were labelled with the near-infrared probe DiR (HY-D1048, MedChemExpress, USA) and administered intravenously (10 mg/kg), with free DiR in PBS used as the control. In vivo fluorescence imaging was performed at 0, 6, 12, 24, 48, and 72 h post-injection using an in vivo imaging system (IVIS Spectrum2, Revvity, USA). Mice were euthanised at 6, 24 and 48 h, and major organs, including the heart, liver, spleen, lungs, kidneys, brain and maxillary alveolar bone, were harvested for ex vivo fluorescence imaging. The fluorescence intensity of each organ was subsequently quantified to evaluate the biodistribution of TNVs. Experiments were performed with four biological replicates. To further evaluate TNV accumulation at inflammatory sites, equal doses of DiR-labelled TNVs (10 mg/kg) were intravenously injected into healthy (CTR) and LIP mice. At 6, 24, and 48 h post-injection, the maxillae and major organs were harvested for ex vivo fluorescence imaging using the IVIS Spectrum2 system (Revvity, USA). Fluorescence intensity was quantified with Living Image software to compare TNV accumulation in inflamed maxillary tissues. Experiments were performed in quadruplicate. Micro-computed tomography (Micro-CT) analysis Maxillary specimens were fixed in 4% paraformaldehyde for 24 h and scanned using a high-resolution micro-CT system (Skyscan, Belgium) at 17.5 μm resolution, 70 kV, and 800 µA. Three-dimensional reconstruction was carried out using 3D Slicer software. Alveolar bone loss was evaluated by measuring the distance between the cemento–enamel junction and the alveolar bone crest (CEJ–ABC). Bone microstructural parameters, including BV/TV, Tb.N, Tb.Th and Tb.Sp, were quantitatively assessed. Histological staining analysis The maxillary specimens were decalcified in 0.5 M EDTA solution for 10 days, followed by paraffin embedding and sectioning into 5 μm-thick slices. The sections were processed for haematoxylin and eosin (H&E) and tartrate-resistant acid phosphatase (TRAP) staining. Immunofluorescence staining of tissue sections Maxillary specimens were paraffin-embedded and sectioned. Sections were deparaffinised, rehydrated and subjected to antigen retrieval, then blocked with 3% BSA and incubated overnight at 4 °C with a rabbit anti-CitH3 antibody (ab5103, Abcam, UK). Following incubation with an HRP-conjugated secondary antibody (GB23303, Servicebio, China), signals were developed using iF555-Tyramide (G1233, Servicebio, China). After antibody stripping and re-blocking, sections were incubated with anti-ELANE (ab314916, Abcam, UK) or anti-MPO antibodies (ab208670, Abcam, UK) and visualised using iF488-Tyramide (G1231, Servicebio, China). Nuclei were counterstained with DAPI. Images were acquired using a high-resolution digital microscope (Nikon, Japan) and analysed using ImageJ software. Real-time quantitative PCR (RT-qPCR) Total RNA was isolated from cells and gingival tissues using the RNAeasy Plus Animal RNA Isolation Kit (R0032, Beyotime, China) following the manufacturer’s instructions. RNA was reverse-transcribed into cDNA at 37 °C for 15 min, followed by 85 °C for 5 s. RT-qPCR was carried out using a TaKaRa system with a reaction mixture consisting of 1 µL cDNA, 0.4 µL of each primer, 3.2 µL ddH₂O and 5 µL TB Green. The amplification programme included an initial denaturation at 95 °C for 30 s, followed by 39 cycles of 95 °C for 5 s and 60 °C for 30 s, with melt curve analysis performed afterwards. Primers were synthesised by TaKaRa and are listed in Table S1. All experiments were conducted with six biological replicates. Western blot analysis of tissue samples Gingival tissues were homogenised, and total protein was extracted using RIPA buffer and centrifuged at 15,000 g for 5 min at 4 °C. Protein concentrations were measured using an Enhanced BCA assay. Equal amounts of protein were separated by SDS–PAGE and subsequently transferred onto PVDF membranes, which were then blocked with 5% skimmed milk for 1.5 h. Membranes were incubated overnight at 4 °C with primary antibodies against PAD4, ELANE, CitH3, MPO, and β-actin, followed by secondary antibodies for 2 h. Target protein bands were visualised using chemiluminescent reagents (Merck Millipore, USA) and captured with an imaging system (LAS4000M, USA), followed by densitometric analysis using ImageJ software. Experiments were performed with five biological replicates. Histological assessment of major organs (Heart, liver, spleen, lung and kidney) Histological evaluation was conducted on the major organs of mice. The heart, liver, spleen, lung and kidney were collected, rinsed, fixed, embedded in paraffin and sectioned. The resulting sections were subsequently stained with haematoxylin and eosin (H&E). Histological images were acquired using a light microscope (Olympus, Japan) and subjected to morphological analysis. Haemolysis assay Fresh mouse blood (5 mL) was collected from the orbital venous plexus to prepare a red blood cell (RBC) suspension. RBCs were incubated with TNVs (10, 20 and 30 µg/mL) at 37 °C for 1 h. RBCs treated with distilled water or PBS were used as positive and negative controls, respectively. The samples were centrifuged at 5,000 rpm for 5 min, and the absorbance of the supernatant was measured at 540 nm. The haemolysis rate was calculated using the following equation: Haemolysis (%) = (Abs_sample − Abs_negative control)/(Abs_positive control − Abs_negative control) × 100%. Haematological and serum biochemical analysis in mice Healthy mice received intravenous TNVs (10 mg/kg) once daily for 7 days, with body weight monitored throughout. Blood samples were collected for biosafety evaluation. Complete blood count parameters (WBC, RBC, NE, LY, MO, HGB, and PLT) were measured. Serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), urea (UREA) and creatinine (CREA) were measured to evaluate liver and kidney function. Statistical analysis Data are expressed as mean ± standard deviation (SD). Differences between two groups were assessed using unpaired two-tailed Student’s t-tests, whereas comparisons among multiple groups were conducted by one-way analysis of variance (ANOVA). Time-dependent effects were analysed using two-way repeated-measures ANOVA followed by Šídák’s multiple comparisons test. Statistical analyses and graphical presentation were performed using GraphPad Prism 10.0 or Origin. A p value < 0.05 was considered statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001). All experiments were repeated independently at least three times. Results Single-cell transcriptomics reveals dysregulated NET activation in human periodontitis To characterise immune cell heterogeneity and regulatory mechanisms associated with inflammatory bone destruction, this study selected periodontitis as a representative disease model. Publicly available periodontitis-related single-cell RNA sequencing data (GSE171213, platform GPL24676) were retrieved from the GEO database, comprising gingival tissue samples from healthy individuals and patients with periodontitis [40]. Quality control analysis demonstrated consistent sample distributions without evident batch effects, with most cells showing mitochondrial gene proportions below 15%, indicating reliable data quality (Fig. S2A, B). Highly variable genes associated with cellular heterogeneity were identified (Fig. S2C). Additional details are shown in Fig. S2D–H. t-SNE analysis revealed clear segregation of multiple cell populations in both groups, including fibroblasts, epithelial cells, T cells, monocytes, macrophages, and neutrophils (Fig. 1A–C). Marker gene expression patterns are presented in Fig. 1D. Although overall cellular compositions were comparable between groups, immune cells—particularly neutrophils—were markedly expanded in periodontitis tissues. Quantitative analysis showed that neutrophil proportions increased significantly from 12.65% in healthy samples to 22.59% in periodontitis samples (Fig. 1E), indicating extensive immune cell recruitment and suggesting neutrophil-driven immune dysregulation as a central pathological feature. Functional enrichment analysis of differentially expressed genes in neutrophils revealed significant involvement in immune regulation and inflammatory responses. Gene Ontology (GO) analysis showed enrichment in myeloid leukocyte activation, regulation of innate immune response, and phagocytosis (Fig. 1F), reflecting enhanced innate immune activity. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis further revealed significant enrichment of the NOD-like receptor, TNF, and Toll-like receptor signalling pathways, as well as pathways involved in neutrophil extracellular trap formation (Fig. 1G). These enriched pathways were associated with enhanced phagocytosis, degranulation, NETs formation, and activation of the TLR/NOD–NF-κB/MAPK signalling axis. Notably, the NET formation pathway was significantly enriched, suggesting that aberrant neutrophil activation and excessive NETs formation may be key drivers of periodontitis initiation and progression, consistent with previous studies on inflammatory bone destruction [1–3]. Fig. 1Single-cell RNA sequencing analysis reveals neutrophil activation and NETs-associated pathways in periodontitis. (A-C) t-SNE visualization of major cell populations in healthy samples, periodontitis samples, and combined datasets, respectively. (D) Dot plot illustrating the expression of canonical marker genes across the identified cell clusters. (E) Proportional analysis of major cell types between healthy and periodontitis groups. (F) GO enrichment analysis of differentially expressed genes, highlighting myeloid leukocyte activation, innate immune responses, and granule-related biological processes. (G) KEGG pathway enrichment analysis showing significant enrichment of pathways associated with neutrophil extracellular trap formation, NOD-like receptor signalling, NF-κB signalling, TNF signalling and MAPK signalling Single-cell RNA sequencing analysis reveals neutrophil activation and NETs-associated pathways in periodontitis. (A-C) t-SNE visualization of major cell populations in healthy samples, periodontitis samples, and combined datasets, respectively. (D) Dot plot illustrating the expression of canonical marker genes across the identified cell clusters. (E) Proportional analysis of major cell types between healthy and periodontitis groups. (F) GO enrichment analysis of differentially expressed genes, highlighting myeloid leukocyte activation, innate immune responses, and granule-related biological processes. (G) KEGG pathway enrichment analysis showing significant enrichment of pathways associated with neutrophil extracellular trap formation, NOD-like receptor signalling, NF-κB signalling, TNF signalling and MAPK signalling Isolation, purification, and physicochemical characterisation of turmeric-derived nanovesicles To systematically characterise the physicochemical properties of TNVs, they were successfully isolated using differential centrifugation combined with ultracentrifugation and density gradient centrifugation (Fig. 2A). Density gradient analysis demonstrated that TNVs were predominantly enriched in the 30%–45% fractions, forming clear and stable layers (Fig. 2B), indicating effective removal of impurities and high sample purity. Transmission electron microscopy revealed typical vesicular morphologies, mainly spherical or cup-shaped structures with intact membranes and well-defined boundaries (Fig. 2C, D). Nanoparticle tracking analysis showed that TNVs had hydrodynamic diameters primarily within 0–300 nm, with an average size of approximately 150 nm (Fig. 2F), consistent with plant-derived nanovesicles and indicative of good uniformity [41, 42]. Zeta potential analysis revealed a pronounced negative surface charge (−36.50 ± 0.37 mV) (Fig. 2G), suggesting favourable colloidal stability. SDS–PAGE analysis confirmed the presence of multiple low-molecular-weight proteins (Fig. 2E). LC–MS-based metabolite profiling indicated that TNVs were mainly enriched in lipids and lipid-like molecules (36.4%), together with organic oxides, organic acids and their derivatives, and nitrogen- and oxygen-containing compounds (Fig. 2H). The high lipid content provides structural support for the bilayer membrane and may enhance biocompatibility and delivery performance [43]. Further LC–MS analysis identified several characteristic turmeric-derived bioactive compounds, including curcumenol, (R)-ar-turmerone, and curcumin (Fig. 2I), which possess well-documented anti-inflammatory and antioxidant activities [44–46]. The MS/MS spectrum are shown in Fig. S3. Their co-existence suggests efficient encapsulation and protection of hydrophobic molecules within TNVs, potentially improving stability and bioavailability. Molecular docking analysis further showed that these bioactive compounds could stably bind to the active sites of key NETs-related molecules (MPO, ELANE, and PAD4) via hydrogen bonding and hydrophobic interactions (Fig. 2J). Binding energy analysis revealed strong affinities, with curcumin displaying the highest binding stability, followed by (R)-ar-turmerone and curcumenol (Fig. 2K; Table S2). These findings suggest that TNV-derived bioactive components may regulate key mediators of NETs formation through a multi-component mode of action. Biosafety is a critical determinant for the clinical translation of nanomedicines [47]. To evaluate the biocompatibility of TNVs for potential clinical translation, their in vitro and in vivo toxicological profiles were systematically investigated. TNVs exhibited negligible cytotoxicity towards MC3T3-E1 cells, as demonstrated by live/dead staining and CCK-8 assays (Fig. S4). Moreover, wound-healing and transwell assays further revealed that TNVs significantly enhanced the migratory capacity of MC3T3-E1 cells (Fig. S5). The in vivo biosafety of TNVs was subsequently assessed. No apparent histopathological abnormalities were observed in the major organs based on haematoxylin and eosin (H&E) staining (Fig. S6A). In addition, no significant differences were observed in body weight, haemolysis rate, serum biochemical parameters, or haematological indices among the groups, indicating the favourable biocompatibility of TNVs (Fig. S6B–E). Collectively, these results demonstrate that TNVs exhibit excellent in vitro and in vivo biocompatibility and biosafety, supporting their potential for future translational development. Fig. 2Preparation, characterization, and compositional analysis of TNVs. (A) Schematic representation of the TNVs isolation process, including juicing, differential centrifugation, ultracentrifugation, and density gradient centrifugation. (B) Macroscopic appearance of the TNVs-enriched fraction after density gradient separation. (C–D) Transmission electron microscopy (TEM) images showing the morphology and ultrastructure of TNVs. (E) SDS–PAGE analysis of TNVs protein profiles. (F) Nanoparticle Tracking Analysis (NTA) of TNVs size distribution. (G) Zeta potential measurement of TNVs. (H) Classification of bioactive compounds in TNVs identified by LC–MS analysis. (I) Representative LC–MS chromatogram identifying Curcumenol, (R)-ar-Turmerone, and Curcumin. (J) Molecular docking models predicting the interactions between TNV-derived compounds and PAD4, MPO, and ELANE. (K) Assessment of the binding energies of compounds interacting with target proteins Preparation, characterization, and compositional analysis of TNVs. (A) Schematic representation of the TNVs isolation process, including juicing, differential centrifugation, ultracentrifugation, and density gradient centrifugation. (B) Macroscopic appearance of the TNVs-enriched fraction after density gradient separation. (C–D) Transmission electron microscopy (TEM) images showing the morphology and ultrastructure of TNVs. (E) SDS–PAGE analysis of TNVs protein profiles. (F) Nanoparticle Tracking Analysis (NTA) of TNVs size distribution. (G) Zeta potential measurement of TNVs. (H) Classification of bioactive compounds in TNVs identified by LC–MS analysis. (I) Representative LC–MS chromatogram identifying Curcumenol, (R)-ar-Turmerone, and Curcumin. (J) Molecular docking models predicting the interactions between TNV-derived compounds and PAD4, MPO, and ELANE. (K) Assessment of the binding energies of compounds interacting with target proteins Cellular uptake and lysosomal co-localisation of TNVs To evaluate the uptake of TNVs under inflammatory conditions, MC3T3-E1 cells and differentiated HL-60 (dHL-60) neutrophil-like cells were stimulated with 100 ng/mL lipopolysaccharide. Time-dependent uptake of PKH26-labelled TNVs (50 µg/mL) by both cell types was observed using confocal microscopy (Fig. 3A, B). In MC3T3-E1 cells, minimal fluorescence was observed within 0–1 h, whereas weak signals appeared at 4 h and increased progressively thereafter. After 12–24 h, pronounced red fluorescence was observed uniformly within the cytoplasm, suggesting continuous cellular uptake and intracellular accumulation. In contrast, dHL-60 neutrophil-like cells exhibited more rapid internalisation, with detectable fluorescence at 1 h, marked enhancement at 4–6 h, and strong signals in nearly all cells at 12–24 h, reflecting higher endocytic efficiency. Flow cytometric analysis confirmed these observations (Fig. 3C–E). In MC3T3-E1 cells, TNVs uptake increased from 15.9% at 1 h to 30.2% at 4 h and 63.3% at 6 h, reaching near saturation at 12–24 h (79.1% and 98.1%). By comparison, dHL-60 cells showed significantly higher early uptake, reaching 32.2% at 1 h and 88.1% at 4 h, and maintaining > 99% internalisation at later time points. Significant differences between the two cell types were observed at multiple time points. In both cell types, internalised TNVs were predominantly localised to the perinuclear cytoplasmic region, indicating stable intracellular accumulation. To further evaluate cellular uptake of TNVs at a lower concentration, qualitative and quantitative analyses were performed using confocal microscopy and flow cytometry. The results demonstrated that TNVs (30 µg/mL) were efficiently internalised by both cell types, with dHL-60 cells exhibiting higher and time-dependent uptake. After 24 h of co-incubation, the uptake rates reached approximately 96.9% in dHL-60 cells and 83.6% in MC3T3-E1 cells (Fig. S7). To investigate intracellular trafficking, lysosomal co-localisation analysis was performed using LysoTracker staining (Fig. 3F–K). Partial overlap between TNVs and lysosomes was observed at 2 h and became more pronounced at 4 h in both cell types. Quantitative analysis using ImageJ confirmed a high degree of co-localisation, indicating that TNVs were predominantly trafficked to and accumulated within the endosome–lysosome pathway. These findings suggest that TNVs are internalised via endocytosis and subsequently transported through the endosome–lysosome system. As illustrated in the schematic diagram (Fig. 3L), TNVs undergo endosomal–lysosomal trafficking, during which their cargos may be released into the cytoplasm via lysosomal membrane fusion or permeabilisation, thereby mediating their biological functions, consistent with previous studies [48, 49]. Fig. 3Cellular uptake and intracellular trafficking of TNVs. (A–B) Confocal fluorescence microscopy images showing time-dependent uptake of PKH26-labeled TNVs by MC3T3-E1 osteoblasts and dHL-60 neutrophil-like cells. (C–D) Flow cytometric analysis quantifying cellular uptake efficiency of TNVs in MC3T3-E1 and dHL-60 cells. (E) Quantitative analysis of TNVs uptake kinetics over time. Data are expressed as mean ± SD (n = 3). The time–effect curves with repeated measurements at multiple time points were analyzed using two-way repeated-measures ANOVA, with treatment and time as factors, followed by Šídák’s multiple comparisons test to compare treatments at individual time points. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. (F, I) LysoTracker staining showing the co-localisation of TNVs with lysosomes. (G, H, J, K) Line-scan analysis demonstrating spatial overlap between TNVs and lysosomal compartments. (L) Schematic illustration of TNVs internalization via endocytosis, followed by endosomal–lysosomal trafficking and subsequent intracellular cargo release Cellular uptake and intracellular trafficking of TNVs. (A–B) Confocal fluorescence microscopy images showing time-dependent uptake of PKH26-labeled TNVs by MC3T3-E1 osteoblasts and dHL-60 neutrophil-like cells. (C–D) Flow cytometric analysis quantifying cellular uptake efficiency of TNVs in MC3T3-E1 and dHL-60 cells. (E) Quantitative analysis of TNVs uptake kinetics over time. Data are expressed as mean ± SD (n = 3). The time–effect curves with repeated measurements at multiple time points were analyzed using two-way repeated-measures ANOVA, with treatment and time as factors, followed by Šídák’s multiple comparisons test to compare treatments at individual time points. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. (F, I) LysoTracker staining showing the co-localisation of TNVs with lysosomes. (G, H, J, K) Line-scan analysis demonstrating spatial overlap between TNVs and lysosomal compartments. (L) Schematic illustration of TNVs internalization via endocytosis, followed by endosomal–lysosomal trafficking and subsequent intracellular cargo release Effects of different concentrations of TNVs on NET and ROS formation To evaluate the regulatory effects of TNVs on NET formation under controlled and reproducible conditions, ionomycin (ION) and phorbol 12-myristate 13-acetate (PMA) were used to induce NETosis, respectively, thereby enabling precise assessment of TNVs-mediated modulation of pathological NETosis while avoiding confounding non-specific immune activation associated with bacterial or lipopolysaccharide (LPS) stimulation. Immunofluorescence confocal microscopy showed that control neutrophils exhibited intact nuclear morphology with minimal NETs-associated structures, whereas ionomycin stimulation induced pronounced chromatin decondensation, extensive extracellular DNA networks, and markedly enhanced CitH3 signals, indicating robust NETs release (Fig. 4A). TNVs significantly attenuated ionomycin-induced NET formation in a dose-dependent manner. Increasing TNVs concentrations from 10 to 30 µg/mL progressively reduced CitH3-positive networks and restored nuclear morphology, indicating effective inhibition of NETosis execution. PKH26 labelling confirmed efficient interaction and internalisation of TNVs by neutrophils. Similar inhibitory effects were observed under PMA stimulation (Fig. S8A–C). Extracellular DNA release was quantitatively assessed using SYTOX Green staining and flow cytometry. Ionomycin significantly increased the proportion of SYTOX Green-positive cells to approximately 25.2%, whereas TNVs treatment reduced this proportion in a concentration-dependent manner (Fig. 4B, C). Medium and high doses of TNVs markedly reduced extracellular DNA release. Comparable results were obtained following PMA stimulation (Fig. S8D, E), confirming the potent inhibitory effects of TNVs on NETs formation. Given the central role of reactive oxygen species (ROS) in NETosis [11], we next examined whether TNVs regulate intracellular ROS production. H₂O₂-induced oxidative stress markedly increased ROS levels, whereas TNVs reduced ROS generation in a dose-dependent manner (Fig. 4D). Flow cytometry confirmed decreases in both ROS-positive cells and fluorescence intensity (Fig. 4E, F), and similar antioxidant effects were observed in MC3T3-E1 cells (Fig. S9). The reduction in ROS levels coincided with decreased CitH3 formation and extracellular DNA release, indicating that TNVs may suppress NETosis via ROS-mediated signalling. To further elucidate the molecular mechanisms underlying TNV-mediated regulation of NETosis and inflammation, RT-qPCR was performed to quantify the expression of NETs-related enzymes and pro-inflammatory cytokines. Ionomycin markedly upregulated IL-1β, TNF-α, PAD4, ELANE, MPO, and CTSG, whereas TNVs treatment significantly reduced their expression (Fig. 4G–M). IL-6 displayed a decreasing trend without statistical significance. These findings suggest that TNVs may suppress NET formation and attenuate inflammation-associated responses, accompanied by reduced ROS levels and altered expression of NET-related molecules. To investigate the role of ROS in TNV-mediated regulation of NETosis, functional interventions were performed using the ROS scavenger NAC and the ROS enhancer H₂O₂. NET formation was assessed by SYTOX Green staining. Both TNVs and NAC reduced ION-induced NETosis, whereas ROS enhancement with H₂O₂ promoted NET formation. Notably, the inhibitory effect of TNVs was partially reversed by H₂O₂, suggesting that modulation of ROS levels may contribute to the inhibitory effects of TNVs on NETosis. Representative images and quantitative data are shown in Fig. S10. Fig. 4TNVs suppress neutrophil extracellular trap formation and ROS production in vitro. (A) Immunofluorescence images showing ionomycin-induced NETs formation, with staining for citrullinated histone H3 (CitH3) and PKH26-labeled TNVs. ION represents ionomycin. L, M, and H represent TNVs at 10, 20, and 30 µg/mL, respectively. (B–C) SYTOX Green staining and quantitative analysis of extracellular DNA release (n = 3). (D) Fluorescence images of intracellular ROS generation in neutrophils detected by DCFH-DA following H₂O₂ stimulation with or without TNVs treatment. (E–F) Flow cytometric analysis and quantification of ROS production in neutrophils. Data are expressed as mean ± SD (n = 3). Comparisons among multiple groups were made using one-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. (G–M) Quantitative real-time PCR analysis of inflammatory cytokines (IL-1β, IL-6, TNF-α) and NETs-associated genes (PAD4, ELANE, MPO, CTSG). Data are expressed as mean ± SD (n = 3). Comparisons among multiple groups were made using one-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 TNVs suppress neutrophil extracellular trap formation and ROS production in vitro. (A) Immunofluorescence images showing ionomycin-induced NETs formation, with staining for citrullinated histone H3 (CitH3) and PKH26-labeled TNVs. ION represents ionomycin. L, M, and H represent TNVs at 10, 20, and 30 µg/mL, respectively. (B–C) SYTOX Green staining and quantitative analysis of extracellular DNA release (n = 3). (D) Fluorescence images of intracellular ROS generation in neutrophils detected by DCFH-DA following H₂O₂ stimulation with or without TNVs treatment. (E–F) Flow cytometric analysis and quantification of ROS production in neutrophils. Data are expressed as mean ± SD (n = 3). Comparisons among multiple groups were made using one-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. (G–M) Quantitative real-time PCR analysis of inflammatory cytokines (IL-1β, IL-6, TNF-α) and NETs-associated genes (PAD4, ELANE, MPO, CTSG). Data are expressed as mean ± SD (n = 3). Comparisons among multiple groups were made using one-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 Effects of TNVs and their bioactive components on NET formation To investigate the effects of TNVs and their major bioactive components on NET formation, neutrophils were stimulated with ionomycin (ION) to induce NETosis and subsequently treated with TNVs (30 µg/mL) or equivalent concentrations of curcumin (Cur), curcumenol (CUM), and (R)-ar-turmerone (Tur) (Fig. 5A). Immunofluorescence staining revealed that control neutrophils exhibited intact nuclear morphology with weak ELANE and CitH3 signals. In contrast, ION stimulation induced pronounced chromatin decondensation, enhanced ELANE and CitH3 release, and typical NETs-like structures. Notably, TNVs treatment markedly inhibited NET formation, as evidenced by reduced marker expression and preservation of nuclear integrity. However, Cur, CUM, and Tur alone only partially attenuated these changes. Consistently, quantitative analysis showed that CitH3 and ELANE fluorescence intensities were significantly increased in the ION group but were more effectively suppressed by TNVs than by individual compounds (Fig. 5B, C). Furthermore, SYTOX Green flow cytometric analysis demonstrated that ION stimulation increased extracellular DNA release to 30.3%, whereas TNVs reduced this proportion to 8.23%, showing stronger inhibitory effects than Cur (12.5%), CUM (19.2%), and Tur (16.7%) (Fig. 5D, E). To compare the effects of TNVs, individual bioactive components, and their simple mixture on NETosis, NET formation was assessed following ION stimulation (Fig. S11). ION markedly induced NETosis, whereas TNVs significantly inhibited this process. Individual components (Cur, CUM, or Tur) and their simple mixture exhibited varying degrees of inhibitory effects, but were less effective than TNVs. Quantitative analysis further supported these observations, suggesting that the enhanced inhibitory effect of TNVs on NETosis may be associated with vesicle-mediated delivery of intrinsic bioactive cargo. As illustrated in the schematic diagram (Fig. 5F), TNVs mediate the delivery of bioactive constituents, including curcumenol, (R)-ar-turmerone, and curcumin, thereby inhibiting NET formation. Fig. 5Effects of TNVs and Their Bioactive Components on NET Formation. (A) Representative immunofluorescence images of neutrophils treated with ionomycin (ION), ionomycin + TNVs (ION + TNVs), ionomycin + curcumin (ION + Cur), ionomycin + curcumenol (ION + CUM), or ionomycin + (R)-ar-turmerone (ION + Tur) tested at estimated equivalent concentrations. Nuclei were stained with DAPI (blue), CitH3 (green), and ELANE (red). (B–C) Quantification of ELANE and CitH3 fluorescence intensities (n = 3). (D–E) Flow cytometric analysis of NETs formation based on SYTOX Green staining, showing representative histograms and quantitative percentages. (F) Schematic illustration showing that TNVs deliver bioactive compounds (curcumenol, (R)-ar-turmerone, and curcumin) to inhibit NET formation. Data are presented as mean ± SD (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 Effects of TNVs and Their Bioactive Components on NET Formation. (A) Representative immunofluorescence images of neutrophils treated with ionomycin (ION), ionomycin + TNVs (ION + TNVs), ionomycin + curcumin (ION + Cur), ionomycin + curcumenol (ION + CUM), or ionomycin + (R)-ar-turmerone (ION + Tur) tested at estimated equivalent concentrations. Nuclei were stained with DAPI (blue), CitH3 (green), and ELANE (red). (B–C) Quantification of ELANE and CitH3 fluorescence intensities (n = 3). (D–E) Flow cytometric analysis of NETs formation based on SYTOX Green staining, showing representative histograms and quantitative percentages. (F) Schematic illustration showing that TNVs deliver bioactive compounds (curcumenol, (R)-ar-turmerone, and curcumin) to inhibit NET formation. Data are presented as mean ± SD (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 TNVs reprogram neutrophil transcriptomes to suppress NETosis and oxidative burst To investigate transcriptomic alterations induced by TNVs, neutrophils were treated with 2 µM ionomycin alone (ION group) or 2 µM ionomycin plus 30 µg/mL TNVs (TNVs group), followed by RNA sequencing of samples from both groups. Each group contained three biological replicates. Principal component analysis revealed clear separation between the two groups along the PC1 and PC2 axes, accounting for 65.22% and 12.02% of the total variance, respectively (Fig. 6A), reflecting significant differences in global gene expression patterns. Pearson correlation analysis demonstrated high consistency among biological replicates, confirming data reliability (Fig. 6B). Venn diagram analysis revealed 558 and 425 uniquely expressed genes in the ION and TNVs groups, respectively, with 11,421 shared genes (Fig. 6C). Using thresholds of |log₂ fold change| ≥ 0.322 (1.25-fold change) and adjusted P < 0.05, differential expression analysis identified 6,087 DEGs, comprising 3,083 upregulated and 3,004 downregulated genes (Fig. 6D). The volcano plot revealed extensive transcriptomic remodelling following TNVs treatment. GO enrichment analysis of upregulated DEGs showed significant enrichment in biological processes related to antigen presentation, innate immune signalling, bacterial defence, and inflammatory regulation. Molecular function analysis highlighted cytokine activity, immunoglobulin receptor binding, and antioxidant activity, while cellular component analysis indicated localisation to MHC complexes and phagosome membranes (Fig. 6E). KEGG analysis further demonstrated enrichment in immune-related pathways, including antigen processing, cytokine–receptor interaction, apoptosis, autophagy, and phagosome formation (Fig. 6G). In contrast, downregulated DEGs were significantly enriched in processes associated with respiratory burst, ROS generation, neutrophil activation, chemotaxis, degranulation, migration, and osteoclast differentiation (Fig. 6F). KEGG analysis revealed suppression of chemokine signalling, actin cytoskeleton regulation, leukocyte transendothelial migration, FcγR-mediated phagocytosis, and endocytosis pathways. Notably, neutrophil extracellular trap formation and osteoclast differentiation pathways were markedly downregulated (Fig. 6H). Heatmap analysis further confirmed distinct expression patterns of NETosis-related, antioxidant, and anti-inflammatory genes between groups (Fig. 6I). NETosis-associated genes were downregulated in the TNVs group, whereas antioxidant and anti-inflammatory genes were upregulated. Collectively, these findings indicate that TNVs profoundly remodel the transcriptomic landscape by suppressing NET formation and osteoclast-related pathways while enhancing immune regulation and antioxidant responses. Fig. 6Transcriptomic analysis of ION and TNVs groups by RNA sequencing. (A) Principal component analysis (PCA) showing the separation of ION and TNVs samples based on global gene expression profiles (n = 3). (B) Pearson correlation heatmap displaying the correlation coefficients among all samples, indicating the reproducibility and reliability of the RNA-seq data. (C) Venn diagram illustrating the numbers of unique and shared genes expressed in the ION and TNVs groups. (D) Volcano plot of differentially expressed genes (DEGs). Significantly upregulated genes are shown in red, significantly downregulated genes in blue, and genes without significant changes in grey. DEGs were identified based on the criteria of |log₂ fold change| ≥ 0.322 (corresponding to a 1.25-fold change) and adjusted P value (P.adjust) < 0.05. (E) Gene Ontology (GO) enrichment analysis of upregulated DEGs, including biological process (BP), molecular function (MF), and cellular component (CC) categories. (F) GO enrichment analysis of downregulated DEGs. (G) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of upregulated DEGs. (H) KEGG pathway enrichment analysis of downregulated DEGs. (I) Heatmap showing the expression patterns of NETosis-related, antioxidant-related, and anti-inflammatory–related genes in the ION and TNVs groups. The colour scale from blue to red indicates increasing gene expression levels Transcriptomic analysis of ION and TNVs groups by RNA sequencing. (A) Principal component analysis (PCA) showing the separation of ION and TNVs samples based on global gene expression profiles (n = 3). (B) Pearson correlation heatmap displaying the correlation coefficients among all samples, indicating the reproducibility and reliability of the RNA-seq data. (C) Venn diagram illustrating the numbers of unique and shared genes expressed in the ION and TNVs groups. (D) Volcano plot of differentially expressed genes (DEGs). Significantly upregulated genes are shown in red, significantly downregulated genes in blue, and genes without significant changes in grey. DEGs were identified based on the criteria of |log₂ fold change| ≥ 0.322 (corresponding to a 1.25-fold change) and adjusted P value (P.adjust) < 0.05. (E) Gene Ontology (GO) enrichment analysis of upregulated DEGs, including biological process (BP), molecular function (MF), and cellular component (CC) categories. (F) GO enrichment analysis of downregulated DEGs. (G) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of upregulated DEGs. (H) KEGG pathway enrichment analysis of downregulated DEGs. (I) Heatmap showing the expression patterns of NETosis-related, antioxidant-related, and anti-inflammatory–related genes in the ION and TNVs groups. The colour scale from blue to red indicates increasing gene expression levels In vivo biodistribution of TNVs To evaluate the in vivo biodistribution of TNVs under pathological conditions, a LIP mouse model was established and intravenously injected with DiR-labelled TNVs (DiR-TNVs). Whole-body fluorescence imaging was performed at 0, 6, 12, 24, 48, and 72 h using an IVIS system (Fig. 7A). Both free DiR and DiR-TNVs were rapidly distributed throughout the body within 6 h after injection; however, distinct differences in tissue accumulation and retention were observed (Fig. 7B). A similar rise–decline pattern was observed in both groups; however, the DiR-TNVs group showed significantly higher average radiant efficiency than the free DiR group during 6–24 h (Fig. 7C), indicating enhanced in vivo accumulation mediated by TNVs. After 48 h, fluorescence signals gradually converged and returned to near-background levels by 72 h, suggesting favourable and controllable metabolic behaviour. Ex vivo organ imaging demonstrated that DiR signals were predominantly localised in the liver, alveolar bone, brain, and kidneys, with minimal fluorescence detected in other organs (Fig. 7D, G, J). At 6 h, both groups exhibited detectable fluorescence in the maxillary region, with slightly higher signals in the free DiR group, accompanied by stronger hepatic fluorescence (Fig. 7D–F), indicating more diffuse distribution and faster clearance of the free dye. At 24 h, DiR-TNVs maintained stable and significantly enhanced fluorescence in the maxillary region, while hepatic signals were comparable between groups (Fig. 7G–I), suggesting preferential accumulation at periodontal inflammatory sites during the therapeutic window. At 48 h, hepatic fluorescence in the DiR-TNVs group was markedly increased, whereas substantial maxillary signals were still detectable (Fig. 7J–L), indicating gradual hepatic metabolism following local action. Additional data are provided in Fig. S12 and Fig. S13. To examine the accumulation of TNVs in inflamed periodontal sites, DiR-labelled TNVs were administered to healthy (CTR) and ligature-induced periodontitis (LIP) mice, and their distribution was assessed by ex vivo fluorescence imaging (Fig. S14). TNVs were mainly detected in the liver at all time points, consistent with hepatic clearance. Fluorescence in the maxilla was weak in healthy mice but markedly higher in the inflamed maxilla of LIP mice at 24 h and 48 h. These findings indicate preferential accumulation of TNVs in inflammatory tissues, which may be associated with passive enrichment within the inflammatory microenvironment. Collectively, TNVs exhibited a time-dependent biodistribution profile, establishing an effective therapeutic window between 6 and 24 h, preferential accumulation in inflamed tissues, and controlled hepatic clearance, thereby reducing the risk of long-term systemic accumulation. Fig. 7In vivo biodistribution of TNVs in ligature-induced periodontitis mice. (A) Experimental workflow illustrating intravenous injection of DiR-labelled TNVs followed by in vivo imaging system (IVIS) analysis. (B) Representative whole-body fluorescence images acquired at indicated time points (0–72 h) after injection of DiR or DiR-TNVs. (C) Quantitative analysis of average radiant efficiency over time (n = 4). (D, G, J) Ex vivo fluorescence imaging of major organs at 6, 24, 48 h, including maxilla, brain, heart, liver, spleen, lung, and kidney (n = 4). (E, H, K) Fluorescence intensity in the maxilla quantified at 6, 24 and 48 h. (F, I, L) Quantification of fluorescence intensity in the liver at 6, 24, 48 h. Data are expressed as mean ± SD (n = 4). The time–effect curves with repeated measurements at multiple time points were analyzed using two-way repeated-measures ANOVA, with treatment and time as factors, followed by Šídák’s multiple comparisons test for pairwise comparisons at individual time points. Statistical significance between two groups was assessed using an unpaired t-test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 In vivo biodistribution of TNVs in ligature-induced periodontitis mice. (A) Experimental workflow illustrating intravenous injection of DiR-labelled TNVs followed by in vivo imaging system (IVIS) analysis. (B) Representative whole-body fluorescence images acquired at indicated time points (0–72 h) after injection of DiR or DiR-TNVs. (C) Quantitative analysis of average radiant efficiency over time (n = 4). (D, G, J) Ex vivo fluorescence imaging of major organs at 6, 24, 48 h, including maxilla, brain, heart, liver, spleen, lung, and kidney (n = 4). (E, H, K) Fluorescence intensity in the maxilla quantified at 6, 24 and 48 h. (F, I, L) Quantification of fluorescence intensity in the liver at 6, 24, 48 h. Data are expressed as mean ± SD (n = 4). The time–effect curves with repeated measurements at multiple time points were analyzed using two-way repeated-measures ANOVA, with treatment and time as factors, followed by Šídák’s multiple comparisons test for pairwise comparisons at individual time points. Statistical significance between two groups was assessed using an unpaired t-test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 TNVs attenuate ligature-induced alveolar bone resorption and reduce inflammatory cytokine expression To evaluate the therapeutic efficacy of TNVs in vivo, a ligature-induced periodontitis mouse model was established and treated with intravenous TNVs (10 mg/kg, q.d.) (Fig. 8A). Three-dimensional micro-computed tomography (micro-CT) reconstruction revealed marked alveolar bone resorption in the ligature-induced periodontitis (LIP) group, characterised by reduced alveolar crest height and increased cementoenamel junction–alveolar bone crest (CEJ–ABC) distance compared with controls. In contrast, TNVs treatment substantially attenuated alveolar bone loss, with evident improvements on buccal, palatal, and sagittal sections (Fig. 8B). Quantitative analysis showed significantly increased mesial and distal CEJ–ABC distances in the LIP group, which were markedly reduced following TNVs administration (Fig. 8C, D). Further microstructural evaluation revealed that TNV treatment markedly increased bone volume fraction (BV/TV) and trabecular thickness (Tb.Th), decreased trabecular separation (Tb.Sp), and partially restored trabecular number (Tb.N) compared with the LIP group (Fig. 8E–H), suggesting an improvement in alveolar bone microarchitecture. Histological analysis using haematoxylin and eosin staining revealed extensive inflammatory infiltration and periodontal ligament disruption in the LIP group, whereas TNVs treatment markedly alleviated inflammation and restored tissue integrity (Fig. 8I). Tartrate-resistant acid phosphatase staining showed significantly elevated osteoclast numbers in the LIP group, which were substantially reduced following TNVs administration, as reflected by decreased TRAP-positive areas (Fig. 8J, K). Furthermore, RT-qPCR analysis further confirmed that TNV treatment significantly downregulated the expression of the pro-inflammatory cytokines IL-1β, IL-6, and TNF-α (Fig. 8L–N). Collectively, these findings indicate that TNVs effectively suppress inflammation, inhibit osteoclast-mediated bone resorption, and promote alveolar bone preservation in periodontitis. Fig. 8TNVs attenuate alveolar bone loss in ligature-induced periodontitis mice. (A) Schematic illustration of the experimental timeline. (B) Representative three-dimensional micro-computed tomography (micro-CT) reconstructions of alveolar bone. (C–D) Quantification of alveolar bone loss expressed as the distance from the cementoenamel junction to the alveolar bone crest (CEJ–ABC) (n = 6). (E–H) Quantitative analysis of bone morphometric parameters, including BV/TV, Tb.Th, Tb.Sp, and Tb.N (n = 6). (I–J) Haematoxylin and eosin (H&E) staining and tartrate-resistant acid phosphatase (TRAP) staining of periodontal tissues (n = 4). (K) Quantification of TRAP-positive area. (L–N) Quantitative real-time PCR analysis of inflammatory cytokines (n = 6). Data are expressed as mean ± SD. Comparisons among multiple groups were made using one-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 TNVs attenuate alveolar bone loss in ligature-induced periodontitis mice. (A) Schematic illustration of the experimental timeline. (B) Representative three-dimensional micro-computed tomography (micro-CT) reconstructions of alveolar bone. (C–D) Quantification of alveolar bone loss expressed as the distance from the cementoenamel junction to the alveolar bone crest (CEJ–ABC) (n = 6). (E–H) Quantitative analysis of bone morphometric parameters, including BV/TV, Tb.Th, Tb.Sp, and Tb.N (n = 6). (I–J) Haematoxylin and eosin (H&E) staining and tartrate-resistant acid phosphatase (TRAP) staining of periodontal tissues (n = 4). (K) Quantification of TRAP-positive area. (L–N) Quantitative real-time PCR analysis of inflammatory cytokines (n = 6). Data are expressed as mean ± SD. Comparisons among multiple groups were made using one-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 TNVs attenuate ligature-induced periodontitis by suppressing NET formation To determine whether TNVs suppress periodontitis-associated NET formation in vivo, immunofluorescence staining was performed on periodontal tissues from ligature-induced periodontitis (LIP) mice. In the control group (CTR), periodontal tissues exhibited intact architecture with minimal expression of CitH3, ELANE, and MPO and negligible NETs-related signals (Fig. 9A, B). In contrast, the LIP group displayed abundant CitH3-positive signals that extensively co-localised with ELANE or MPO, indicating pronounced NETosis. Notably, TNVs treatment (LIP + TNVs) markedly reduced the fluorescence intensities of CitH3, ELANE, and MPO, accompanied by a substantial decrease in NETs-associated structures. Quantitative analysis confirmed that TNVs significantly decreased the expression levels of CitH3, ELANE, and MPO compared with the LIP group (Fig. 9C–E), demonstrating effective inhibition of NET accumulation in vivo. To further validate these regulatory effects at the molecular level, NETs-related gene expression in periodontal tissues was analysed by quantitative PCR. PAD4, MPO, ELANE, and Ctsg were significantly upregulated in the LIP group, whereas TNVs treatment markedly reduced PAD4, MPO, and CtSG gene expression (Fig. 9F, H, I). ELANE expression also exhibited a downward trend, although the change did not reach statistical significance (Fig. 9G). Western blot analysis further confirmed these findings. Protein levels of MPO, PAD4, ELANE, and CitH3 were markedly elevated in the LIP group compared with controls. In contrast, TNV treatment significantly reduced the protein levels of these markers, particularly CitH3, ELANE, and MPO (Fig. 9J–M), whereas PAD4 showed a decreasing trend without statistical significance (Fig. 9N). The detailed Western blot bands are presented in Supplementary Fig. S15. Collectively, these results demonstrate that TNVs effectively suppress NETosis at both transcriptional and protein levels in vivo, thereby contributing to their therapeutic efficacy in ligature-induced periodontitis. Fig. 9TNVs suppress NETs formation in vivo in periodontal tissues. (A–B) Immunofluorescence staining of periodontal tissues for CitH3, ELANE, and MPO, with enlarged views of representative regions. (C–E) Quantification of fluorescence intensities (n = 6). (F–I) Quantitative real-time PCR analysis of NET-associated genes, including PAD4, ELANE, MPO, and CTSG (n = 6). (J) Western blot analysis of PAD4, ELANE, MPO, and CitH3 expression in periodontal tissues (n = 5). (K–N) Densitometric measurement of protein expression was normalised to β-actin. Results are expressed as mean ± SD, and comparisons among groups were conducted using one-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 TNVs suppress NETs formation in vivo in periodontal tissues. (A–B) Immunofluorescence staining of periodontal tissues for CitH3, ELANE, and MPO, with enlarged views of representative regions. (C–E) Quantification of fluorescence intensities (n = 6). (F–I) Quantitative real-time PCR analysis of NET-associated genes, including PAD4, ELANE, MPO, and CTSG (n = 6). (J) Western blot analysis of PAD4, ELANE, MPO, and CitH3 expression in periodontal tissues (n = 5). (K–N) Densitometric measurement of protein expression was normalised to β-actin. Results are expressed as mean ± SD, and comparisons among groups were conducted using one-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 Discussion Inflammatory bone destruction represents a common pathological outcome of various chronic inflammatory diseases, including periodontitis, rheumatoid arthritis and peri-implantitis [1–5]. In these conditions, dysregulated host immune responses drive excessive osteoclast activation and progressive bone loss [10–12]. In the present study, we developed turmeric-derived nanovesicles (TNVs) as a natural nanoplatform for the regulation of neutrophil extracellular trap (NET) formation and the attenuation of inflammatory cascades associated with bone destruction. Through single-cell transcriptomic analysis, mechanistic investigations and both in vitro and in vivo validation, we demonstrated that TNVs effectively suppress NET formation, accompanied by reduced intracellular ROS levels and decreased expression of NET-associated markers. This regulation subsequently alleviates periodontal inflammation, inhibits osteoclastogenesis and preserves the microstructural integrity of alveolar bone. Collectively, these findings suggest that plant-derived nanovesicles may represent a multifunctional immunomodulatory strategy for controlling inflammation-driven bone resorption. Increasing evidence suggests that excessive NET formation plays an important role in the pathogenesis of various inflammatory diseases [50]. In periodontitis and other inflammatory bone-destructive disorders, neutrophils represent a dominant immune cell population within the inflammatory microenvironment and play a crucial role in inflammatory regulation [51]. Consistent with previous reports, our single-cell transcriptomic analysis revealed pronounced neutrophil hyperactivation and persistent upregulation of NET-associated signalling pathways in periodontal tissues, providing direct clinical evidence supporting the involvement of NETosis in inflammatory amplification and alveolar bone destruction [13, 15]. NET formation results in the release of extracellular DNA, histones and proteolytic enzymes, which function as damage-associated molecular patterns capable of sustaining inflammatory responses and promoting osteoclast differentiation [13]. Mechanistically, NETosis is primarily regulated by ROS-dependent signalling cascades involving key molecules such as neutrophil elastase (ELANE), myeloperoxidase (MPO) and peptidylarginine deiminase 4 (PAD4) [13]. These molecules have therefore been widely regarded as potential therapeutic targets for NETosis-associated diseases. However, most currently available pharmacological strategies focus on single molecular targets and frequently suffer from limited bioavailability and insufficient tissue targeting [16, 52]. Our findings suggest that TNVs may provide an alternative approach for NETosis modulation through multi-target regulatory mechanisms. Physicochemical characterisation demonstrated that TNVs exhibit typical EV-like morphology, nanoscale size, and favourable colloidal stability comparable to that of mammalian extracellular vesicles. Furthermore, metabolomic profiling combined with molecular docking analysis revealed that TNVs encapsulate multiple turmeric-derived bioactive molecules capable of interacting with MPO, ELANE and PAD4. A notable finding of this study is that TNVs exhibited markedly stronger inhibitory effects on NET formation than free turmeric-derived compounds. Although numerous turmeric constituents possess recognised anti-inflammatory properties, their therapeutic efficacy is often limited by poor solubility, low bioavailability and rapid metabolic clearance. In contrast, TNVs significantly suppressed NET formation, reduced intracellular ROS levels and decreased the expression of NETosis-associated markers, including citrullinated histone H3 (CitH3), MPO and ELANE. This enhanced inhibitory effect is likely attributable to the vesicle-mediated delivery architecture of TNVs. Cellular uptake studies demonstrated that TNVs can be efficiently internalised by neutrophil-like cells and predominantly trafficked through endosomal–lysosomal pathways, where their cargos may be released into the cytoplasm via lysosomal membrane fusion or membrane permeabilisation, thereby enabling the delivery of bioactive molecules to intracellular inflammatory compartments, consistent with previous reports [48, 49]. In addition, in vivo imaging revealed preferential accumulation of TNVs within inflamed periodontal tissues, likely mediated by passive targeting through blood circulation and vascular extravasation [53], accompanied by delayed systemic clearance and controlled hepatic metabolism [54]. Intravenous administration was adopted in this study based on several considerations. First, this work was designed as a proof-of-concept study to evaluate the potential of TNVs as a systemic immunomodulatory platform, with periodontitis serving as a representative model for validation within a local inflammatory microenvironment. Second, periodontitis is characterised by pronounced systemic inflammatory features, as inflammatory mediators and bacterial products can enter the circulation, suggesting that systemic immunomodulation may confer broader therapeutic benefits than local intervention alone. Notably, following intravenous administration, TNVs exhibited preferential accumulation in inflamed tissues, including periodontal lesions, suggesting their potential utility for inflammation-associated delivery applications. This phenomenon may be associated with passive enrichment within the inflammatory microenvironment. In addition, systemic administration enabled comprehensive evaluation of in vivo biodistribution and potential safety profiles, thereby providing important evidence to support the clinical translational potential of TNVs. From a broader perspective, the present study provides additional insight into the mechanisms underlying inflammatory bone destruction. Increasing evidence indicates that excessive NET formation establishes a self-amplifying pathological loop linking oxidative stress, immune activation and osteoclast-mediated bone resorption [10–12]. NET-derived molecular components can continuously stimulate inflammatory signalling, further recruit immune cells and promote osteoclast differentiation, thereby accelerating bone destruction in chronic inflammatory diseases [13, 15]. By suppressing ROS generation and inhibiting NETosis, TNVs effectively interrupt this pathological cascade. Transcriptomic analysis further demonstrated that TNVs significantly downregulated NET-associated signalling pathways, while antioxidant and anti-inflammatory gene expression exhibited an increasing trend, suggesting that TNVs may modulate inflammatory signalling networks and thereby improve the immune microenvironment associated with inflammatory bone destruction. Importantly, plant-derived nanovesicles represent a natural multi-component delivery system capable of regulating multiple molecular targets simultaneously. Compared with conventional single-component therapeutics, this natural nanosystem may represent a promising therapeutic strategy through vesicle-mediated delivery of intrinsic bioactive cargo. In multifactorial diseases such as periodontitis, this multi-target regulatory strategy may offer distinct advantages. ROS-related mechanisms and NET formation may play important roles in linking innate immune activation to inflammatory bone resorption, and modulation of these processes through multifunctional nanovesicles may provide new therapeutic opportunities for inflammatory bone-destructive disorders. Despite the promising therapeutic potential demonstrated in this study, several limitations should be considered. Since Cur, CUM, and Tur are highly hydrophobic compounds with poor aqueous dispersibility and limited in vivo stability, their in vivo administration generally requires auxiliary solubilizing or carrier systems, which may introduce formulation-dependent confounding factors and thereby compromise the reliability of direct comparisons [55–57]. Consequently, direct in vivo comparisons between TNVs and free multi-component formulations may not provide reliable results; therefore, the in vivo effects of free Cur, CUM and Tur were not further investigated in this study. This limitation may also reflect the potential advantage of TNVs in facilitating the delivery and bioavailability of hydrophobic bioactive compounds [23, 58]. Moreover, large-scale production, batch consistency, standardisation and long-term storage stability of plant-derived nanovesicles still require further optimisation for future clinical translation. Further improvements in isolation and purification procedures, together with the development of stable and broadly applicable marker systems for plant-derived EVs, will also be essential for enhancing reproducibility and standardisation in this field. Finally, the ligature-induced periodontitis model used in this study mainly reflects an acute inflammatory condition and may not fully reproduce the pathological characteristics of chronic human periodontitis. Therefore, future studies using chronic periodontal disease models and long-term observations are still required to further evaluate the therapeutic potential of TNVs. Conclusion In summary, this study demonstrates that turmeric-derived nanovesicles act as a natural nanocarrier capable of modulating NET formation and attenuating inflammation-driven bone destruction.These findings highlight plant-derived nanovesicles as a promising biomimetic nanoplatform for immunomodulatory therapy in inflammatory bone-destructive diseases. Supplementary Information Supplementary Material 1 Supplementary Material 1