Triptolide loaded cerium oxide nanozyme based oral nanoplatform for targeted colitis therapy Authors: Jing Lu, Yiming Chen, Wenxing Ding, Shuhui Zheng, Liping Yang, Yiting Wang, Qingdong Ji, Chendong Qi, Feng Gao Published: 2026 ABSTRACT Ulcerative colitis (UC) is a type of colonic inflammation characterized by immune dysregulation and intestinal inflammation. Cerium oxide (CeO₂) nanozyme can effectively downregulate reactive oxygen species (ROS), but it lacks anti-inflammatory and other pharmacological activities. Triptolide (TP) has anti-inflammatory, antioxidant, antifibrotic, and autophagy-regulating pharmacological properties. However, severe gastrointestinal side effects hinder its clinical application. Therefore, this study aimed to develop TP-loaded CeO₂ nanozymes (T-C). T-C effectively reduces ROS at sites of colitis. In addition, chitosan, chondroitin sulfate, and pectin were employed as carriers to construct T-C-loaded nanoparticles (T-C@pNP), in which chitosan promoted nanoparticle assembly, chondroitin sulphate enhances uptake of nanoparticle, and pectin enabled pectinase-responsive release, resulting in a novel oral nanoplatform for colitis treatment. In a mouse colitis model, T-C@pNP significantly alleviated inflammation-induced weight loss, colon shortening, and colon damage, demonstrating notable therapeutic effects. T-C@pNP not only downregulated ROS, but also reduced the levels of proinflammatory factors (TNF-α, IL-6, and IFN-γ), possibly through inhibition of the nuclear factor kappa-B (NF-κB) pathway, and diminished neutrophil infiltration. Furthermore, it regulated macrophage polarization, potentially alleviated inflammation-associated fibrosis, and disrupted the vicious cycle of inflammation. Moreover, T-C@pNP may activate epithelial cell autophagy, enhance the intestinal mucus barrier, promote the expression of ZO-1 and occludin, and partially modulate the gut microbiota composition, thereby improving intestinal barrier repair. In conclusion, complemented by satisfactory anti-inflammatory and biosafety results, this nanoplatform represents a promising treatment option for colitis in rats. Introduction Inflammatory bowel disease (IBD), including Crohn's disease (CD) and ulcerative colitis (UC), is a group of intestinal inflammatory disorders of unknown etiology (Cui et al., 2025; Wang et al., 2022). These conditions are characterized by persistent inflammation of the intestinal mucosa (Fu et al., 2024; Lopetuso et al., 2023), which leads to a substantial global disease burden(Kaplan, 2015; Loftus, 2021). IBD seems to be caused by a combination of multiple interacting factors, and its pathogenesis is not yet fully understood (Roda et al., 2020). Accordingly, IBD exhibits a heterogeneous disease phenotype and can lead to various complications, including strictures, fistulas, ulcers, and colon cancer-related lesions, which pose significant challenges in diagnosis and treatment (Verstockt et al., 2022). These diseases have a profound impact on patients' quality of life, with symptoms including abdominal pain, diarrhoea, rectal bleeding, weight loss, and life-threatening complications such as bowel obstruction, malnutrition, and malignancies (Bedke et al., 2024; Song et al., 2023). Despite advances in the use of monoclonal antibodies and small molecules to interrupt the malignant cycle of IBD, treatment continues to face significant limitations (Honap et al., 2024; Yao et al., 2023). Research has indicated that IBD is closely associated with the overproduction of reactive oxygen species (ROS) in its microenvironment (Bao et al., 2023; Ge et al., 2024). The oxidative metabolites produced by epithelial cells in the IBD microenvironment lead to oxidative stress, excessive ROS production, and damage to epithelial cell structural proteins and tight junctions, impairing the intestinal mucosal barrier (Kan et al., 2024; Zhang et al., 2023). Moreover, excessive ROS in the microenvironment is a key factor in maintaining the proinflammatory phenotype of macrophages (Cao et al., 2024; Chen et al., 2025). Hollow cerium oxide (CeO2), with its superoxide dismutase (SOD)-like and catalase (CAT)-like activities, effectively counteracts diseases associated with excessive ROS production (Cao et al., 2023; Su et al., 2025). However, owing to its lack of specificity, the use of CeO₂ alone may result in systemic side effects, and the limited functionality of CeO₂ restricts its therapeutic application (He et al., 2025). Triptolide (TP) has significant anti-inflammatory effects in various models of inflammatory and autoimmune diseases. TP downregulates inflammatory factors through inhibition of the nuclear factor kappa-B (NF-κB) pathway, restores tight junction permeability, and suppresses the abnormal localisation of the zonula occludens 1 (ZO-1) protein under TNF-α stimulation (Yuan et al., 2019; Zhang et al., 2021). However, the gastrointestinal toxicity of TP significantly limits its clinical oral application (Xi et al., 2017). Nanodelivery platforms can help overcome these issues (Liu et al., 2022; Nie et al., 2020; Shubhra, 2023; Shubhra et al., 2017). Therefore, there is an urgent need for an oral nanoparticle drug delivery system that targets sites of inflammation in the colon to enhance local treatment and reduce off-target toxicity (Qiu et al., 2023; Tu et al., 2022). To address this, we loaded TP into the pores of CeO₂ to prepare TP-CeO₂ nanozymes (T-C), utilizing the mesoporous structure of CeO₂ nanozymes to achieve the dual effects of ROS downregulation and drug loading. However, the effectiveness of CeO₂ nanozymes is hindered when orally administered, due to the complexity of the gastrointestinal environment, which impedes their efficient delivery to inflamed intestinal sites (Huang et al., 2025). To improve the oral delivery and targeting of T-C, chitosan (CS) has been identified as an effective carrier for drug encapsulation (Lemos et al., 2021). CS is a positively charged polysaccharide that not only acts as a carrier to facilitate the formation of subsequent coating layers through electrostatic interactions during nanoparticle assembly, but also increases the residence time of nanoparticles at the inflammation site. Chondroitin sulphate (CH), a naturally occurring acidic glycosaminoglycan formed by the polymerisation of N-acetylgalactosamine and D-glucuronic acid, is negatively charged and has a strong affinity for overexpressed CD44 receptors under inflammatory conditions (Qiao et al., 2025; Song et al., 2023). Pectin (PC) is a natural polymer that can respond to pectinase (Pase) secreted by colonic bacteria (Bora and Mishra, 2016; Günter and Popeyko, 2016). After considering these facts, we synthesized T-C to scavenge excessive ROS. CS and CH were used to encapsulate T-C to achieve inflammation-targeting properties. Then, the particle was coated with PC to produce T-C@pNP, an innovative oral nanoplatform with excellent stability in the stomach. This nanoplatform accumulates in the site of colitis, thus mitigating the gastrointestinal side effects of TP. Furthermore, its multiple weakly acidic functional groups endow it with a negatively charged surface. This characteristic further promotes its binding to positively charged sites within colonic regions affected by inflammation (Zou et al., 2025). Previous studies have developed nanozyme-based or triptolide delivery systems for inflammatory disease therapy (Cao et al., 2023; Rao et al., 2023). However, these systems mainly focus on either nanozyme-mediated ROS regulation or triptolide delivery alone. In contrast, the present study constructed a CeO₂ nanozyme loaded with triptolide (T-C) using CS, CH, and PC as carriers to develop a multifunctional drug delivery system. This system is a novel oral nanoformulation for the treatment of IBD, with Pase responsiveness and colonic inflammation-targeting capabilities. This drug delivery system facilitates the targeted delivery of TP to the inflamed colon, which reduces proinflammatory cytokines by downregulating ROS levels, possibly inhibiting the NF-κB pathway, and regulating macrophage polarization. The administration of T-C@pNP might play a role in promoting a more abundant and varied gut microbial community in models of inflammatory bowel disease (Scheme 1). Conclusively, the developed nanoplatform introduces a promising therapeutic avenue for IBD.Scheme 1Preliminary study on the mechanism of oral treatment of T-C@pNP in colonic inflamed mice.Scheme 1 Preliminary study on the mechanism of oral treatment of T-C@pNP in colonic inflamed mice. Materials and methods Synthesis and characterization of T-C CeO₂ nanozymes were synthesized and characterized based on previously reported methods (Kempaiah Devaraju et al., 2012). First, 0.5 g of Ce(NO₃)₃·6H₂O and 0.2 g of PVP K29/31 were dissolved in 15 mL of ethylene glycol, and 1 mL of HCl solution was added slowly under vigorous stirring. After stirring for 2 h and sonicating for 3 min, the clear solution was transferred to a 100 mL PTFE-lined autoclave and heated at 160 °C for 180 min. After cooling to room temperature, the gray product was collected, centrifuged at 10,000 rpm for 20 min using deionized water, washed twice, and the precipitate was dispersed in deionized water. The dispersion was then sonicated for 10 min, and the product was vacuum-dried at 45 °C for 12 h to obtain CeO₂ nanozymes. For the preparation of TP-CeO₂ nanoparticles (T-C), the hollow mesoporous CeO₂ nanozyme surface was used to adsorb the free drug via co-incubation. TP was dissolved in a methanol solution (0.5 mg/mL, 0.6 mL) and co-incubated with the synthesized cerium oxide (2 mg/mL, 1 mL) stock solution for 12 h. The mixture was then washed with deionized water at 10,000 rpm for 20 min, and the gray-white product was collected. The product was dispersed in deionized water, sonicated for 10 min to achieve uniform dispersion, and T-C were obtained. The particle size, zeta potential, and polydispersity index (PDI) of T-C were measured at 25 °C using a Malvern Zetasizer (Nano-ZS90, Malvern, UK). To prepare samples for transmission electron microscope (TEM) imaging, the T-C (1 mg/mL) solution was placed on a nickel grid for 20 min, and excess solution was absorbed using filter paper. The morphology of the nanozyme was then evaluated using TEM (JEM 2100, JEOL Ltd., Japan), with point and line resolution of 0.38 and 0.20 nm, respectively. After vacuum drying, an equal amount of T-C, CeO₂, and TP was analyzed by X-ray diffraction (XRD, PXRD, Ultimate IV, Rigaku, Japan) with Cu Kα radiation, operating at 40 kV and 30 mA, scanning speed of 5°/min, and recording diffraction peaks within the 2θ range of 5° to 45°. Additionally, elemental qualitative and semi-quantitative analysis of T-C was performed using Energy Dispersive Spectroscopy (EDS, Tecnai G2 F30 S-Twin, USA) and X-ray Photoelectron Spectroscopy (XPS, Thermo Fisher Scientific ESCALAB 250Xi, USA), which also analyzed the valence state composition of cerium ions on the surface of T-C. Additionally, the existing form of drug in nanozymes was determined by differential scanning calorimetry (DSC) (DSC2500, TA Instruments, USA). The ability of CeO₂, TP, and T-C at different concentrations to modulate four types of free radicals: •OH, H₂O₂, •ABTS+, and •O₂− was evaluated (Lin et al., 2025). First, the TMB method was used to assess T-C's modulation of •OH. H₂O₂ modulation capacity was tested using a hydrogen peroxide detection kit. Additionally, the ABTS method was employed to evaluate T-C's effect on •ABTS+. The •O₂− modulation activity was also assessed using an •O₂− detection kit. Preparation and characterization of the nanoplatform CS was acidified and adjusted to pH 4.5, and then 1 mg/mL of TPP solution was slowly added to 4 mL of 1 mg/mL CS acetic acid solution, followed by stirring for 30 min to obtain CS nanoparticles (CS NP). Next, 1.0 mg/mL CH solution was slowly added to the CS NP, and electrostatic interactions were utilized to form CS nanoparticles encapsulating CH (CS/CH NP). Subsequently, a low-ester PC solution was mixed with CS/CH NP, and 2.5% calcium chloride solution was added for crosslinking to obtain PC-coated CS/CH NP nanoparticles (pNP) (Zou et al., 2024). The CS acetic acid stock solution was diluted and then slowly added dropwise to 4 mL of T-C solution, followed by stirring and incubation for 3 h to obtain CS-encapsulated T-C nanoparticles (T-C@CS). Then, varying amounts of CH solution (1 mg/mL) were slowly added to the T-C@CS (4 mL, 1 mg/mL) solution. The pH was maintained at 4.5–5.0, throughout the process, followed by stirring for 4 h to obtain CH-encapsulated T-C@CS nanoparticles (T-C@CS/CH). The T-C@CS/CH solution was slowly added to a 0.025 mg/mL low-ester PC solution, stirred for 12 h, and calcium chloride solution was added for solidification. After stirring for 4 h, the mixture was centrifuged to obtain PC-coated T-C@CS/CH (T-C@pNP). The formulation was optimized based on particle size and PDI by varying the amounts of T-C, CS, CH, and PC. T-C@pNP was dispersed in deionized water and homogenized by probe sonication to obtain a uniform suspension prior to use. The particle size of T-C@pNP was determined using a Malvern Zetasizer (Nano-ZS90, Malvern, UK), and the morphology was examined using Transmission Electron Microscopy (TEM, JEM 1400, JEOL Ltd., Japan). Additionally, the drug loading rate (DL) was determined using a UV–Vis spectrophotometer after dissolving the T-C@pNP and treating them with sonication. To prepare FITC-loaded CS nanoparticles (FITC@CS), 0.1 mg/mL FITC stock solution was mixed with the acidified CS solution under stirring, and then 1.0 mg/mL TPP solution was slowly added until the solution became clear, stirred for 30 min, followed by an ice bath and sonication to obtain FITC@CS. Next, 0.06 mL of CH (1 mg/mL) solution was slowly added to the FITC@CS solution, and the mixture was stirred at room temperature for 4 h to obtain FITC-loaded CS/CH nanoparticles (FITC@CS/CH). ICG was used in place of FITC, and ICG@CS/CH was prepared according to the above method, followed by PC coating and calcium chloride solidification to obtain ICG@pNP. Storage stability of T-C@pNP The prepared T-C@pNP nanoparticles were stored at 4 °C in a refrigerator for 8 days. Samples were taken at 0, 2, 4, 6, and 8 days to measure the particle size distribution and zeta potential of the nanoparticles, in order to assess the short-term storage stability of T-C@pNP. In vitro release of T-C@pNP The in vitro release of T-C@pNP was investigated using the dialysis bag method. The prepared T-C and T-C@pNP nanoparticles were resuspended in deionized water. A 5 mL solution of T-C and T-C@pNP was placed into MW 8000–14,000 Da dialysis bags and tightly sealed with dialysis clips. The bags were immersed in release media at different stages and incubated at 37 °C with slow shaking at 110 rpm. At predefined time points (0, 0.5, 1, 2, 4, 6, 8, 10, 12, and 24 h), release samples were collected, and an equivalent volume of release medium was replenished. The concentration of TP in the samples was measured using UV–vis spectroscopy, and the in vitro release profile was plotted. The first 0–2 h simulated the stomach environment after oral administration, using a pH 1.2 buffer. The next 24 h simulated the intestinal environment, with a pH 6.8 buffer containing 2.26 U/mL Pase. Degradation study of T-C@pNP To study the acid resistance and responsive degradation of the nanoparticle shell, the prepared pNP was incubated in pH 1.2 and pH 6.8 buffer solutions. The effect of 2.26 U/mL Pase on nanoparticle degradation was also investigated. Changes in particle size, zeta potential, and PDI were monitored(Lv et al., 2023). In vitro cytotoxicity of T-C@pNP The cytotoxicity of T-C@pNP in RAW 264.7 and IEC-6 cells was evaluated using the CCK-8 assay kit. RAW 264.7 cells were seeded at a density of 1 × 104 cells/well in a 96-well plate. After 24 h of attachment, a series of concentrations of TP, CeO₂, T-C, T-C@CS/CH, and T-C@pNP were added, and the cells were incubated. IEC-6 cells were treated in the same manner. Cell viability was determined by measuring the optical density (OD) at 450 nm using a multimode plate reader (Varioskan LUX, Thermo Fisher Scientific, USA) after a 24-h incubation, and the cell viability for each group was calculated. In vitro cellular uptake Cellular uptake of nanoparticles by RAW 264.7 and IEC-6 cells was evaluated using confocal laser scanning microscopy (CLSM, Nikon, Japan). RAW 264.7 and IEC-6 cells (5 × 105 cells/well) were cultured overnight in glass-bottomed confocal dishes. RAW 264.7 and IEC-6 cells were induced to an inflammatory state with LPS (500 ng/mL) and IFN-γ (2000 U/mL). The inflammatory cells were then exposed to FITC, FITC@CS/CH, and CH + FITC@CS/CH (FITC: 1.0 μg/mL) for 4 h. To study the effect of CH targeting on uptake, cells were pre-incubated with 1 mg/mL CH for 4 h to block surface binding sites, followed by drug administration. The cell nuclei were stained with Hoechst 33342. Fluorescence intensity was quantified using ImageJ software. In vitro anti-inflammatory capacity The levels of inflammatory cytokines (including IL-6, TNF-α, and IFN-γ) in the inflamed RAW 264.7 cells were quantified using an enzyme-linked immunosorbent assay (ELISA) kit. RAW 264.7 cells were seeded at an initial density of 2 × 105 cells/well in a 24-well plate and cultured for 24 h to allow attachment. The control group (Control) was maintained in fresh serum-free medium, while the other groups were incubated with LPS + IFN-γ for 4 h to induce inflammation. First, serum-free medium containing 0.1, 0.5, 1, and 1.5 μg/mL T-C was used for intervention for 24 h. The culture supernatants were collected by centrifugation at 3000 rpm for 20 min, and TNF-α was measured by enzyme-linked immunosorbent assay (ELISA), using a standard curve according to the manufacturer's instructions and measuring at 450 nm with a microplate reader. Afterwards, serum-free medium containing TP + CeO₂ (TP: 0.5 μg/mL, CeO₂: 4.5 μg/mL), T-C (5 μg/mL), and T-C@CS/CH (5 μg/mL) was used for intervention for 24 h. The culture supernatants were collected by centrifugation at 3000 rpm for 20 min, and cytokine concentrations (IL-6, TNF-α, and IFN-γ) were measured by enzyme-linked immunosorbent assay (ELISA), using a standard curve according to the manufacturer's instructions and measuring at 450 nm with a microplate reader. In vitro cell scratch assay The scratch assay was used to investigate the ability of the nanoparticles to promote cell migration and repair intestinal tight junctions. IEC-6 cells were seeded at a density of 2 × 106 cells/well in a 6-well plate, with a higher initial cell density to shorten the time for cell confluence. After the cells had adhered to the plate, a 10 μL sterile pipette tip was used to create parallel scratches across the well, establishing a wound model. PBS buffer was used to wash the wells multiple times to remove cell debris. The control group (Control) was cultured with serum-free medium, while the other groups were treated with TP (0.5 μg/mL), CeO₂ (4.5 μg/mL), T-C (5 μg/mL), and T-C@CS/CH (5 μg/mL) in serum-free medium. Wound healing was observed under a microscope at 0, 6, and 12 h, and all experiments were independently repeated three times (n = 3). ImageJ software (NIH, USA) was used to analyze the wound area. Relative wound closure was calculated as:(1)Relative scratch area%=AreatArea0×100%where Area0 is the initial scratch area, and Areat is the wound area at the indicated time point. In vivo biodistribution measurement The colitis model in Balb/c mice was induced by continuously administering 2.5% (w/v) DSS in drinking water for 7 days. ICG was used to assess the distribution of nanoparticles in the mice. Mice with acute colitis were orally administered ICG, ICG@CS/CH, and ICG@pNP (ICG: 1 mg/kg). At 1, 12, and 24 h, the mice were anesthetized, and biodistribution was recorded using an IVIS Imaging System (IVIS Spectrum CT, PerkinElmer, USA) (Chen et al., 2024). At 24 h, the mice were euthanized, and major organs, including the heart, liver, spleen, lungs, kidneys, small intestine, and colon, were collected. Fluorescence distribution in the isolated organs was observed and recorded using the IVIS system to evaluate the in vivo distribution of the nanoparticles. In vivo anti-colitis efficiency Mice were randomly divided into five groups: healthy group (Healthy), DSS-induced model group (DSS), free TP treatment group (TP), free CeO₂ nanoenzymes treatment group (CeO₂), and T-C@pNP treatment group (T-C@pNP) (n = 5). The healthy group (Healthy) was allowed to drink water freely, and saline was administered via gavage on days 9, 12, 15, 18, and 21. The DSS, TP, CeO₂, and T-C@pNP treatment groups were administered 2.5% DSS solution from days 1–5, 9–13, and 17–21. On days 9, 12, 15, 18, and 21, these groups were gavaged with saline, TP solution (0.15 mg/kg), CeO₂ solution (1.32 mg/kg), and T-C@pNP solution (TP: 0.15 mg/kg; CeO₂: 1.32 mg/kg). All mice were euthanized on day 24. Mouse weight, stool, and behaviour were recorded daily to obtain the disease activity index (DAI). Colon length was measured, and the colon was processed by a cryogenic grinder. The fixed colon was sectioned for H&E and AB/PAS staining (Fu et al., 2024). The spleen index was calculated as the ratio of the spleen weight to the body weight of the mouse (Wang et al., 2023). The heart, liver, spleen, lungs, and kidney tissues of the healthy group, DSS group, TP group, CeO₂ group, and T-C@pNP group were collected for H&E staining. Blood samples from each group were collected, and biochemical parameters, including alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), and creatinine (CRE), were measured according to the kit instructions. Additionally, tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6), interferon-γ (IFN-γ), and lipid carrier protein-2 (LCN-2) levels were determined using ELISA kits. Histochemical analysis The changes in ROS levels in the tissues of each group were detected using cryosectioning. Tissue sections were incubated with the DHE fluorescent probe under light-protected conditions. Cell nuclei were counterstained with DAPI staining solution, and sections were mounted with an anti-fade mounting medium. Fluorescence signals were observed and recorded using Confocal Laser Scanning Microscopy (CLSM). The ROS levels in colon tissues were quantitatively assessed using ImageJ software. The level of MPO in the tissues of each group was detected using paraffin embedding and sectioning techniques. Colon tissues were fixed with universal tissue fixative and subjected to standard dehydration, clearing, and paraffin embedding. Following sectioning, deparaffinization, rehydration, and washing, antigen retrieval was performed. Heat-induced epitope retrieval was carried out using either citrate antigen retrieval buffer or pH 9.0 Tris-EDTA retrieval buffer. Sections were then blocked with 3% bovine serum albumin (BSA) at room temperature, followed by sequential incubation with primary antibody against MPO and a corresponding secondary antibody. The secondary antibody was CY3-conjugated goat anti-rabbit IgG (Servicebio, GB21303, diluted 1:300). Nuclei were counterstained with DAPI. The images obtained under the microscope were analyzed semi-quantitatively using ImageJ software by measuring the signal intensity in the labelled area, which was used to evaluate the relative activity of MPO in the colon tissue of each treatment group. The expression of macrophage M1 polarization marker CD86 and M2 polarization marker CD206 in colon tissues was analyzed using immunofluorescence staining to further investigate the regulatory effect of T-C@pNP on macrophage polarization. Colon tissues from the distal colon of mice in each group were fixed, embedded, and sectioned. After antigen retrieval, immunostaining was performed with primary antibodies for CD86 and CD206, followed by incubation with FITC- and Cy3-labelled secondary antibodies for double staining. The nuclei were counterstained with DAPI, and images were captured using a fluorescence microscope. The secondary antibodies were FITC-conjugated Donkey anti-rat IgG (Servicebio, GB22302, diluted 1:200) and Cy3-conjugated Donkey anti-rabbit IgG (Servicebio, GB21403, diluted 1:300). Collagen deposition in colon tissues was evaluated using Masson's trichrome staining. Colon tissues were fixed, paraffin-embedded, and sectioned. Sections were deparaffinized and rehydrated following standard protocols. Staining was performed using a Masson's Trichrome Staining Kit according to the manufacturer's instructions. After differentiation with 1% glacial acetic acid and dehydration with anhydrous ethanol, sections were cleared in xylene and mounted with neutral balsam. Images were observed and captured under a light microscope, with collagen fibers appearing blue and muscle fibers, fibrin, etc., appearing red. The collagen volume fraction (CVF) in colon tissues were quantitatively assessed using ImageJ software. The mucous layer of the colon was marked using AB/PAS staining. The expression levels of tight junction proteins (ZO-1 and occludin) in colon tissue sections were assessed using immunofluorescence analysis to evaluate the integrity of the intestinal barrier and the repair effect after treatment. The colon tissue samples of mice were fixed, embedded, and sectioned. After blocking nonspecific binding sites with PBS containing 1% BSA, the sections were incubated with the primary antibody against occludin, followed by incubation with HRP-conjugated secondary antibodies. After antigen retrieval, the sections were incubated with the primary antibody against ZO-1, followed by Cy3-conjugated secondary antibody treatment. Finally, the nuclei were stained with DAPI. The relative activity of ZO-1 and occludin in the colon tissues of each treatment group was evaluated by measuring the signal intensity. The secondary antibody for occludin was Alexa Fluor 488-conjugated goat anti-rabbit IgG (Servicebio, GB25303, diluted 1:400), and for ZO-1 was Cy3-conjugated Donkey anti-rabbit IgG (Servicebio, GB21403, diluted 1:300). Western blot Western blotting was used to investigate the effects of healthy, DSS, TP, CeO₂, and T-C@pNP treatments on the expression of NF-κB p-p65 and Beclin-1 in colon tissues. Protein samples were harvested from fresh colon tissue using RIPA lysis buffer mixed with phosphatase inhibitor cocktail and protease inhibitor cocktail. Subsequently, the samples underwent SDS-PAGE electrophoresis, membrane transfer, blocking, and incubation with primary antibody and secondary antibody. The expression of proteins in different colon tissues was detected using enhanced chemiluminescence (ECL). Primary antibodies used were NF-κB p65 (diluted 1:1000) and Beclin-1 (diluted 1:1000). The second antibody, HRP-goat anti-rabbit, was diluted by a factor of 5000 times before use. The chemiluminescence signals obtained were analyzed using ImageJ software. Gut microbiota analysis The fecal samples from healthy, DSS, TP, CeO₂, and T-C@pNP groups (n = 4) were promptly harvested and flash-frozen after chronic treatment. Subsequently, the FastPure stool DNA isolation kit (Shanghai Majorbio Biopharm Technology Co., Ltd., Shanghai, China) was used to isolate the 16S rRNA fragments from the above fecal samples. Then, the microbiota V3-V4 hypervariable regions of the 16S rRNA were amplified by PCR (primers information (5′-3′): 338F: ACTCCTACGGGAGGCAGCAG, and 806R: GGACTACHVGGGTWTCTAAT), analyzing for 16S rRNA sequencing following a standard protocol on the Illumina NextSeq platform (2000 PE300, Illumina, USA). The extraction and analysis operations were conducted by Major Biopharm Technology Co., Ltd. (China) (Zou et al., 2024). Bioinformatic analysis of the gut microbiota was carried out using the Majorbio Cloud platform (https://cloud.majorbio.com). The similarity among the microbial communities in different samples was determined by principal coordinate analysis (PCoA) based on the genus-level relative abundance table using Bray-Curtis dissimilarity. Statistical analysis All experimental data in this study are presented as the mean ± standard deviation (mean ± SD). The statistical data processing and analysis were performed using GraphPad Prism software. For multi-group experimental data, statistical analysis was carried out using one-way ANOVA, and corresponding statistical graphs were plotted. The following notation was used to indicate significant differences: *p < 0.05, **p < 0.01, ***p < 0.001. Results and discussion Characterization of T-C The optimal UV absorption wavelength of TP is 231 nm (Fig. S1). The particle size of the prepared CeO₂ nanozyme was 152.6 ± 3.1 nm, with a zeta potential of 5.0 ± 0.8 mV and a PDI of 0.11 ± 0.02 (Table S1). The CeO₂ nanozyme appeared gray–white in colour. TP was loaded onto the CeO₂ nanozymes using a coincubation method to obtain T-C (Fig. 1A). After coincubation, the particle size of T-C increased to 161.8 ± 2.9 nm, and the zeta potential shifted to −9.8 ± 0.7 mV (Table S1), with a uniform particle size, indicating successful drug encapsulation. The change in zeta potential was consistent with previous reports in the literature (Liu et al., 2023). The drug loading rate of TP was 10.2 ± 0.8% (Table S1). The TEM images exhibited a quasispherical morphology of CeO₂ nanozymes, characterized by alternating light and dark spots resulting from differences in light transmittance (Fig. S2). The TEM images of T-C displayed a darker appearance with the disappearance of bright spots compared to blank CeO₂ nanozymes (Fig. 1B).Fig. 1Synthesis and characterization of T-C and T-C@pNP. (A) Synthesis diagram of T-C and T-C@pNP. (B) TEM image of T-C (scale bar = 100 nm). (C) XRD pattern of T-C. (D) EDS pattern of T-C. (E) XPS spectra of the Ce ion. (F) ·OH, H2O2,•ABTS+, and ·O2− regulating activity of T-C (n = 3). (G) TEM image of T-C@pNP (scale bar = 100 nm). (H) Storage stability at 4 °C (RH 40% ± 15%) for 8 days of T-C@pNP (n = 3). (I) TP release from T-C@pNP at different release media (n = 3). Data are presented as mean ± SD. **p < 0.01 and ***p < 0.001.Fig. 1 Synthesis and characterization of T-C and T-C@pNP. (A) Synthesis diagram of T-C and T-C@pNP. (B) TEM image of T-C (scale bar = 100 nm). (C) XRD pattern of T-C. (D) EDS pattern of T-C. (E) XPS spectra of the Ce ion. (F) ·OH, H2O2,•ABTS+, and ·O2− regulating activity of T-C (n = 3). (G) TEM image of T-C@pNP (scale bar = 100 nm). (H) Storage stability at 4 °C (RH 40% ± 15%) for 8 days of T-C@pNP (n = 3). (I) TP release from T-C@pNP at different release media (n = 3). Data are presented as mean ± SD. **p < 0.01 and ***p < 0.001. The XRD and DSC patterns of TP, CeO₂ and T-C revealed that TP had obvious crystal diffraction peaks, and melting peaks while its characteristic peaks disappeared in T-C, which proved that TP mainly existed in CeO₂ in amorphous form (Fig. 1C, S3). The EDS results revealed the presence of C, Ce and O in T-C, with the percentage of Ce being approximately 28.63%, further corroborating the successful preparation of T-C (Fig. 1D). Additionally, Ce 3D XPS spectra were obtained from the surface of the T-C sample to determine the ratio of Ce3+ to Ce4+. Both Ce3+ and Ce4+ were present on the surface of T-C. Characteristic peaks for Ce3+ were observed at 903.7, 897.1, and 884.9 eV, whereas those for Ce4+ were observed at 916.4, 907.3, 898.9, 888.2, and 882.1 eV. Quantitative analysis revealed that Ce3+ and Ce4+ accounted for 36.6% and 63.5% of the total, respectively. The presence and interconversion of these two oxidation states form the basis of the antioxidant activity of T-C (Fig. 1E). The loading of TP onto CeO₂ nanozymes did not alter the surface Ce3+/Ce4+ redox pair, indicating that T-C retained the enzyme-like activity of CeO₂ and has broad potential for ROS reduction. As shown in Fig. 1F, T-C exhibited significant reductive effects on all four representative free radicals. Moreover, as the concentration of T-C increased, its ability to reduce •OH, H₂O₂, •ABTS+, and •O₂− increased, indicating a concentration-dependent trend. The 10 μg/mL T-C group exhibited the strongest free radical reduction efficiency. Overall, T-C demonstrated a markedly high reductive capacity for these four representative types of ROS, indicating its potential to reduce excessive ROS at the site of inflammation. Notably, 5 μg/mL CeO₂ and TP also resulted in some free radical reduction ability; however, TP exhibited relatively poor free radical reduction capacity, suggesting that T-C compensates for the weak point of TP in free radical reduction while carrying the drug. Characterization of T-C@pNP T-C@pNP was constructed through electrostatic interactions to possess pectinase (Pase) responsiveness and colonic inflammation-targeting properties. Chitosan (CS) was electrostatically adsorbed onto the surface of the T-C nanoparticles, resulting in an increase in the particle size of T-C@CS from 161.8 ± 2.9 nm to 181.9 ± 2.1 nm and causing a zeta potential inversion, which confirmed the successful adsorption of CS onto T-C. The addition of negatively charged CH caused the particle size of T-C@CS/CH to increase to 191.3 ± 2.6 nm, and the zeta potential decreased to 11.7 ± 1.3 mV. Finally, after pectin (PC), which is strongly negatively charged, was encapsulated, the resulting T-C@pNP nanoparticles were obtained. When the mass ratio of T-C/CS/CH/PC was 80:10:4:1, the particle size was 230.5 ± 0.9 nm, the zeta potential was −9.6 ± 0.9 mV, and the PDI was 0.22 (Table S3). Furthermore, the drug loading efficiency of TP was 8.8 ± 1.2%, indicating efficient encapsulation of TP (Table S1). The formulation of the blank nanoparticle carrier pNP was optimized, resulting in a particle size of 213.0 ± 6.8 nm, a zeta potential of −9.3 ± 1.1 mV, and a PDI of 0.22 (Table S4). TEM revealed that the T-C@pNP nanoparticles exhibited a relatively uniform spherical shape, with a consistent particle size (Fig. 1G). T-C@pNP demonstrated excellent storage stability, with almost no change in particle size or zeta potential over 8 days, as shown in Fig. 1H. In the pH 1.2 and pH 6.8 buffer solutions, both the CS NPs and the CS/CH NPs showed significant swelling behaviour, indicating that the outer shell of the nanoparticles ruptured, resulting in the release of the encapsulated drug under these conditions. In contrast, PC-coated nanoparticles (pNPs) remained relatively stable in a pH 1.2 buffer. When 2.26 U/mL Pase was present, the outer layer of pNP (coated with negatively charged PC) was degraded by the enzyme, causing the system to gradually become turbid, with a zeta potential inversion, suggesting that the degradation of the negatively charged PC exposed the positively charged CS/CH beneath it (Table S5, S6). These results further validate the acid resistance and Pase responsiveness of T-C@pNP. In vitro release of T-C@pNP In the pH 1.2 environment, over the first 2 h, T-C showed significant drug release behaviour due to its instability in acidic conditions, while it released the drug gradually in the pH 6.8 environment. T-C@pNP showed strong acid resistance because of the protective effect of the PC outer layer. After the pH increased, the PC shell did not degrade in response to the environment and maintained a relatively low drug release level. In the presence of 2.26 U/mL Pase in a pH 6.8 environment, the Pase-responsive degradation of the outer PC shell exposed the inner nanoparticles, and TP was released gradually (Fig. 1I). These results indicate that T-C@pNP substantially reduces the premature release of TP, exhibits excellent acid resistance and Pase responsiveness, and facilitates colon-specific, Pase-responsive drug release, highlighting its potential for oral colon targeting. It should be noted that this in vitro model is simplified and does not fully replicate the complex gastrointestinal environment, including the presence of bile salts, digestive enzymes, and variable transit times, which may affect nanoparticle degradation and drug release in vivo. In vitro cytological assessment of T-C@pNP We studied the therapeutic effect of the nanoparticles at the cellular level using FITC@CS and FITC@CS/CH. The green fluorescence represents FITC-labelled nanoparticles, and the blue fluorescence represents Hoechst-labelled cell nuclei. After the nanoparticles were coincubated with inflammatory RAW 264.7 and IEC-6 cells for 4 h, the FITC@CS/CH group exhibited the strongest fluorescence intensity, indicating efficient cellular uptake of the nanoparticles. To further explore the uptake mechanism, we employed a CH preincubation blocking group. In this group, the green fluorescence intensity significantly decreased (Fig. 2A, B). This is possibly due to the high affinity between CH and the CD44 receptor. Preincubation with CH impair the uptake of nanoparticles by potentially blocking the CD44 receptors on the surface of inflammatory RAW 264.7 cells and IEC-6 cells. Additionally, owing to the positive charge of FITC@CS/CH, electrostatic adsorption occurred with the negatively charged inflammatory cells, further enhancing the uptake efficiency of the nanoparticles.Fig. 2In vitro assessment of uptake and therapeutic effects of T-C@pNP. (A) In vitro fluorescence images of cellular uptake in RAW264.7 and IEC-6 cells (Scale bar = 50 μm). (B) MFI of free FITC, FITC@CS/CH, and CH + FITC@CS/CH after 4 h of incubation in RAW 264.7 and IEC-6 cells (n = 3). (C) Fluorescence images (Scale bar = 25 μm) and (D) MFI of DCFH-DA in inflamed RAW264.7 cells (n = 3). (E) Pro-inflammatory cytokines of TNF-α, IFN-γ, and IL-6 in RAW264.7 cells following different treatments (n = 3). (F) The quantification of scratch area in the different groups at 6 and 12 h (n = 3). (G) The result of scratching in IEC-6 cells (Scale bar = 100 μm). Data are presented as mean ± SD. *p < 0.05, **p < 0.01 and ***p < 0.001.Fig. 2 In vitro assessment of uptake and therapeutic effects of T-C@pNP. (A) In vitro fluorescence images of cellular uptake in RAW264.7 and IEC-6 cells (Scale bar = 50 μm). (B) MFI of free FITC, FITC@CS/CH, and CH + FITC@CS/CH after 4 h of incubation in RAW 264.7 and IEC-6 cells (n = 3). (C) Fluorescence images (Scale bar = 25 μm) and (D) MFI of DCFH-DA in inflamed RAW264.7 cells (n = 3). (E) Pro-inflammatory cytokines of TNF-α, IFN-γ, and IL-6 in RAW264.7 cells following different treatments (n = 3). (F) The quantification of scratch area in the different groups at 6 and 12 h (n = 3). (G) The result of scratching in IEC-6 cells (Scale bar = 100 μm). Data are presented as mean ± SD. *p < 0.05, **p < 0.01 and ***p < 0.001. In vitro anti-ROS and anti-inflammatory capacity The ROS levels in the LPS + IFN-γ group were significantly elevated, as indicated by abundant green fluorescence, while the fluorescence intensity in the treatment groups markedly decreased. This is attributed to the presence of Ce3+ and Ce4+ on the surface of CeO₂, where Ce3+ provides superoxide dismutase (SOD)-like activity and Ce4+ provides catalase (CAT)-like activity(Dhall and Self, 2018). This dual enzyme-like activity effectively reduces ROS levels at the site of inflammation. Additionally, TP itself contributes to the antioxidative mechanism. Compared with the T-C group, the T-C@CS/CH group exhibited stronger anti-ROS capacity, mainly because of the enhanced cellular uptake ability of T-C@CS/CH (Fig. 2C, D). T-C reduced TNF-α levels in RAW 264.7 cells in a dose-dependent manner within the 0.1–1.5 μg/mL range, with the inhibitory effect approaching saturation at higher concentrations (Fig. S4). Compared with the control group, the T-C@CS/CH group showed the most significant inhibition of proinflammatory factor expression (Fig. 2E). This could be attributed to TP and CeO₂, which possibly block the NF-κB pathway, effectively downregulate ROS, and reduce the expression of proinflammatory factors. Furthermore, the T-C@CS/CH group, with its high uptake capacity, showed the most notable anti-inflammatory effects. Wound healing capacity As time progressed, the wound area in the control group did not significantly change, whereas the treatment groups exhibited varying degrees of cell migration and wound healing. At 12 h, the wound area in the T-C@CS/CH group was only 0.46 times the initial size, indicating significant therapeutic effects. This is attributed mainly to the high uptake efficiency resulting from electrostatic adsorption and the potential role of the CH moiety. The T-C, TP and CeO₂ treatment groups also showed some reduction in the wound area, likely due to the proliferative effect of CeO₂ on cells and the anti-inflammatory effects of TP (Fig. 2F, G). These results suggest that T-C@CS/CH promotes IEC-6 cell migration and may have the potential to promote the repair of intestinal tight junctions. In vivo biodistribution The targeting ability and distribution characteristics of T-C@pNP in IBD mice were studied using ICG. At 1 h, the fluorescence intensity in all groups was concentrated in the stomach. After 12 h, the fluorescence intensity in the free ICG group decreased rapidly and accumulated primarily in the liver, indicating that free ICG was metabolized primarily by the liver. Possibly owing to the high affinity of the CH targeting moiety for the CD44 receptors, highly expressed on inflammatory cells, ICG@CS/CH showed some gastrointestinal targeting ability, but the fluorescence intensity also weakened significantly because of degradation by stomach acid. In contrast, ICG@pNP showed stronger fluorescence in the intestinal region. Compared with the ICG@CS/CH group, the ICG@pNP group maintained stronger fluorescence signals after 24 h (Fig. 3A), effectively demonstrating that the PC coating significantly enhanced the acid resistance of the nanoparticles and facilitated their retention in the colon, indicating enhanced colonic accumulation and retention ability. Compared with the healthy group, the inflammation group that was orally administered ICG@pNP showed a more significant retention effect, which may be related to the interaction between the highly expressed CD44 receptors at the inflammation site and the CH-targeting moiety.Fig. 3In vivo distribution and therapeutic effects. (A) Free ICG, ICG@CS/CH, and ICG@pNP-mediated fluorescence distribution in acute colitis mice. (B) Ex vivo fluorescence images of different organs from different groups at 24 h. (C) Schematic of the pharmacodynamic experimental protocol. (D) Body weight change after being treated with T-C@pNP (n = 5). (E) DAI score of mice (n = 5). (F) Colon length changes after being treated with T-C@pNP (n = 5). Quantitative analysis of (G) colon lengths and (H) spleen index (n = 5). (I) histopathological score of colon sections and (J) H&E staining from different groups (n = 5). Data are presented as mean ± SD. ***p < 0.001.Fig. 3 In vivo distribution and therapeutic effects. (A) Free ICG, ICG@CS/CH, and ICG@pNP-mediated fluorescence distribution in acute colitis mice. (B) Ex vivo fluorescence images of different organs from different groups at 24 h. (C) Schematic of the pharmacodynamic experimental protocol. (D) Body weight change after being treated with T-C@pNP (n = 5). (E) DAI score of mice (n = 5). (F) Colon length changes after being treated with T-C@pNP (n = 5). Quantitative analysis of (G) colon lengths and (H) spleen index (n = 5). (I) histopathological score of colon sections and (J) H&E staining from different groups (n = 5). Data are presented as mean ± SD. ***p < 0.001. The fluorescence in the free ICG group mainly accumulated in the liver, whereas the fluorescence in the ICG@CS/CH group was detected in the liver, small intestine, and colon. However, no fluorescence was detected in the liver in the ICG@pNP group, and the greatest amount of fluorescence accumulated in the colon (Fig. 3B). These results are consistent with the in vivo imaging results, indicating that ICG@pNP has good acid resistance and potential colonic targeting ability, achieving accumulation at the site of colonic inflammation. The ex vivo organ imaging was performed only at the last time point. It serves to support the in vivo trends rather than provide precise anatomical localization. In vivo therapeutic effects On the basis of the in vitro anti-ROS and anti-inflammatory studies and the in vivo biodistribution results, the therapeutic efficacy of T-C@pNP was evaluated in a DSS-induced colitis mouse model (Fig. 3C). The length of the colon is a visual indicator of the severity of IBD, as it typically results in colon shortening (Zou et al., 2024). During the experiment, the body weight of the mice in the healthy group gradually increased by approximately 17%. In contrast, all DSS-induced colitis mice experienced significant weight loss on day 5, accompanied by symptoms such as ruffled fur and reduced activity. During the 6–8-day water phase, their weight either plateaued or increased by less than 5%. The trend in body weight changes correlated with the modelling process, with weight decreasing during DSS administration and increasing slowly during the water phase, indicating a zigzag pattern. The weight loss of the mice in the DSS group significantly decreased after three cycles, and oral administration of TP and CeO₂ did not alleviate this trend. However, in the T-C@pNP treatment group, the body weight of the mice tended to stabilise with long-term treatment, with significant weight gain and gradual recovery (Fig. 3D). In addition, the disease activity index (DAI) was monitored daily to assess the disease status of the mice (Wang et al., 2023). During the modelling phase, the DAI index of the DSS group significantly increased, accompanied by symptoms such as haematochezia and diarrhoea. The DAI of the T-C@pNP treatment group was significantly lower than that of the DSS group, TP treatment group, and CeO₂ treatment group (Fig. 3E). Colitis triggered by DSS resulted in a shortened colon length and vascular congestion. Following treatment with T-C@pNP, however, both the colon length and tissue morphology were similar to those observed in healthy animals (Fig. 3F, G). The occurrence of colitis is often accompanied by splenomegaly, so the spleen index is also an important indicator for assessing therapeutic effects. The spleen index of the DSS group was elevated, while the T-C@pNP treatment group showed a significant decrease in the spleen index (Fig. 3H). Additionally, microscopic tissue analysis demonstrated the treatment efficacy of T-C@pNP. In the DSS group, typical inflammatory symptoms, including epithelial erosion, goblet cell depletion, crypt abscesses, and significant interstitial oedema, were observed. Treatment with TP and CeO₂ alleviated some of the DSS-induced colonic inflammation, but some degree of inflammatory cell infiltration and erosion still occurred. In contrast, the colon tissue in the T-C@pNP group was similar to that in the healthy group. Compared with those in the DSS, TP, and CeO₂ groups, the histopathological scores in the T-C@pNP group were significantly lower, indicating the in vivo anti-inflammatory and protective effects of T-C@pNP (Fig. 3I, J). The experimental data thus confirm the substantial protective and inflammation-reducing properties of T-C@pNP nanozymes in the DSS-induced colitis system. In vivo therapeutic mechanism of T-C@pNP To further investigate the ability of T-C@pNP to reduce ROS levels, we quantified the ROS levels in colon tissue sections from the different groups of mice. The cell nuclei were labelled with blue fluorescence, while the ROS were labelled with red fluorescence (Zou et al., 2026). The strongest increase in red fluorescence was detected in the DSS group, indicating that DSS treatment induced significant oxidative stress. However, compared with the control group, the T-C@pNP treatment group showed a significant reduction in red fluorescence intensity, with a fluorescence intensity similar to that of the control group. Notably, the ROS reduction ability of the TP and CeO₂ treatment groups was limited, possibly because of the low oral dosage and bioavailability of TP and CeO₂, resulting in a limited therapeutic effect (Fig. 4A, B). The ROS fluorescence in colon tissue sections revealed that oral administration of T-C@pNP, after effectively avoiding degradation by stomach acid, achieved colonic inflammation site targeting through the Pase-responsive outer shell, thereby exhibiting a strong, targeted in vivo ability to reduce the level of ROS.Fig. 4In vivo therapeutic mechanism of T-C@pNP. (A) Representative images of ROS, MPO, CD86/CD206, AB/PAS, zonula occludens 1 (ZO-1)/occludin, and Masson of colon tissue after being treated with T-C@pNP (Scale bar = 100 μm, red: ROS, ZO-1 or CD86, green: occludin or CD206, blue: DAPI). (B) MFI of ROS (n = 5). (C) The level of LCN-2 after treatment with T-C@pNP (n = 5). MFI of (D) CD86 and (E) CD206 expression levels (n = 5). MFI of (F) occludin and (G) ZO-1 (n = 5). (H) The protein expression and (I) quantitative analysis of NF-κB p-p65 and Beclin 1 (n = 4). The concentration of (J) IFN-γ, (K) TNF-α, and (L) IL-6 in colonic mucosa (n = 5). Data are presented as mean ± SD. *p < 0.05, **p < 0.01 and ***p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)Fig. 4 In vivo therapeutic mechanism of T-C@pNP. (A) Representative images of ROS, MPO, CD86/CD206, AB/PAS, zonula occludens 1 (ZO-1)/occludin, and Masson of colon tissue after being treated with T-C@pNP (Scale bar = 100 μm, red: ROS, ZO-1 or CD86, green: occludin or CD206, blue: DAPI). (B) MFI of ROS (n = 5). (C) The level of LCN-2 after treatment with T-C@pNP (n = 5). MFI of (D) CD86 and (E) CD206 expression levels (n = 5). MFI of (F) occludin and (G) ZO-1 (n = 5). (H) The protein expression and (I) quantitative analysis of NF-κB p-p65 and Beclin 1 (n = 4). The concentration of (J) IFN-γ, (K) TNF-α, and (L) IL-6 in colonic mucosa (n = 5). Data are presented as mean ± SD. *p < 0.05, **p < 0.01 and ***p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) MPO (myeloperoxidase) is an enzyme predominantly found in neutrophils, and its activity reflects the recruitment of neutrophils to the intestine, where it serves as an important marker of inflammation (Zhang et al., 2022). In DSS-induced colitis mice, significant MPO-positive staining was observed, and MPO activity was higher in the DSS-induced colitis group than in the healthy group, indicating that neutrophil infiltration into the intestinal mucosa occurred and confirming the occurrence of inflammation. After T-C@pNP treatment, MPO-positive staining was significantly reduced and returned to levels similar to those in the healthy group (Fig. 4A). Additionally, LCN-2 (lipocalin-2), which is produced primarily by neutrophils in response to inflammatory stimuli, is widely used as a biomarker for assessing intestinal inflammation. LCN-2 levels in fecal samples, as quantified by ELISA, were notably higher in the DSS group than in the healthy group. In contrast, the group receiving T-C@pNP exhibited a considerable decrease in LCN-2 expression (Fig. 4C). These results indicate that T-C@pNP can effectively alleviate neutrophil infiltration and control the progression of intestinal inflammation. Additionally, owing to the established link between colonic inflammation and macrophage phenotypes, we examined the regulation of CD86 (stained red) and CD206 (stained green), which are markers of M1 and M2 macrophages, in colon tissues from mice treated with T-C@pNP (Zou et al., 2024). The DSS-induced changes in macrophage phenotype were significantly reversed in the T-C@pNP treatment group. Moreover, the increase in the proportion of M2 macrophages may contribute to the alleviation of colitis (Fig. 4A, D and E). Furthermore, occludin (stained green) and zonula occludens-1 (ZO-1, stained red) are critical for maintaining intestinal homeostasis. Immunofluorescence staining showed that T-C@pNP intervention enhanced the expression of the tight junction-related proteins ZO-1 and occludin in DSS-induced mice, suggesting partial restoration of colonic barrier-associated structural markers (Fig. 4A, F and G). A frequent complication associated with IBD involves intestinal fibrosis, which typically presents as a heightened accumulation of collagen and additional extracellular matrix (ECM) elements(Rieder and Fiocchi, 2008). Masson's trichrome staining was employed to evaluate intestinal collagen deposition. In this method, collagen fibers are stained blue, whereas the muscular tissue and cytoplasm are stained red, with erythrocytes appearing light red. In DSS-induced colitis mice, the colonic microstructure was severely damaged, with significant collagen deposition between the submucosa and the muscular layer. After TP or CeO₂ treatment, collagen deposition remained, but mice treated with T-C@pNP exhibited alleviated intestinal inflammation and reduced collagen deposition (Fig. 4A and S5). These results suggest that T-C@pNP effectively reduce collagen deposition and potentially alleviates intestinal fibrosis. To elucidate the potential mechanism underlying the anti-inflammatory effects of T-C@pNP, the expression levels of NF-κB p-p65 and unphosphorylated p65 in the colon tissue of mice were analyzed by Western blotting (Zou et al., 2025). The p65 levels remained similar among the groups, but the expression of NF-κB p-p65 increased in the DSS group, indicating that inflammatory stimulation promoted the phosphorylation of p65, potentially activating the NF-κB pathway. Both free TP and CeO₂ treatment had some inhibitory effects, but after T-C@pNP treatment, the phosphorylation induced by inflammatory stimulation was inhibited, and the expression of NF-κB p-p65 was reduced (Fig. 4H and I). The ability of T-C@pNP to inhibit NF-κB p-p65 expression may stem from the inhibitory effects of TP and CeO₂ on the NF-κB pathway, as well as the regulation of ROS downregulation and macrophage polarization by T-C@pNP. By inhibiting NF-κB p-p65 and downregulating colonic ROS levels, T-C@pNP alleviated neutrophil infiltration, regulated macrophage polarization, and reduced the expression of inflammatory factors locally in the colon, thereby interrupting the inflammatory cycle at the site of colonic inflammation. Autophagy plays a crucial regulatory role in maintaining intestinal homeostasis, participating in the regulation of the gut microbiota and intestinal immunity and providing protective effects during microbial invasion of the intestine (Mizushima, 2018). Therefore, the level of the autophagy marker Beclin-1 in colon tissue was determined by Western blotting (Wang et al., 2025). Treatment with T-C@pNP increased the expression level of Beclin-1 to some extent, which is consistent with previous reports and is likely due to its strong ability to reduce ROS levels (Fig. 4H and I). When intracellular ROS levels are reduced, CeO₂ can induce the formation of autophagy initiation complexes by regulating the mTOR signalling pathway or activating the AMPK pathway (Annangi et al., 2021). T-C@pNP treatment upregulated Beclin-1 levels. This increase suggests that T-C@pNP may promote autophagy activation in intestinal epithelial cells. In vivo regulation of inflammatory cytokines Additionally, we measured the levels of proinflammatory cytokines, including IL-6, TNF-α, and IFN-γ. The DSS group displayed typical features of an inflammatory response, with significantly elevated levels of the proinflammatory cytokines TNF-α, IFN-γ, and IL-6. Treatment with TP, CeO₂, or T-C@pNP had some effects on their levels, especially in the T-C@pNP treatment group, which restored TNF-α, IFN-γ, and IL-6 levels to near normal ranges (Fig. 4J, K and L). These results are associated with the regulation of intestinal oxidative stress by T-C@pNP, the effective reduction of excess ROS, and the anti-inflammatory effects of T-C@pNP through multiple signalling pathways. In summary, T-C@pNP strongly regulated inflammatory cytokines in a colitis mouse model. While the above results clarify the regulatory effects of T-C@pNP on the intestinal microenvironment, further research is needed to determine whether these changes influence the gut microbiota. Gut microbiota regulation In addition to maintaining the intestinal barrier, the gut microbiota serves as a key regulator of inflammatory conditions within the colonic microenvironment. We evaluated the regulatory effect of T-C@pNP on the intestinal microbiota by measuring the diversity and quantity of fecal bacteria. Venn diagram analysis revealed that T-C@pNP administration enhanced gut microbial diversity in mice with DSS-induced colitis (Fig. 5A). A PCoA plot was constructed to display the differences and similarities among the healthy group, DSS group, and T-C@pNP treatment group. The DSS-induced group exhibited a clear difference from the healthy group, whereas the T-C@pNP treatment group presented a distinct microbiota that resembled that of the healthy group (Fig. 5B).Fig. 5Bioinformatic analysis of gut microbiome. (A) Venn diagram of gut microbiota, and values between different circles represent the number of shared families at the family level. (B) PCoA plot on genus level for the gut microbiota from different groups (n = 4). (C) Relative abundance of Family-level taxa from gut microbiota in different groups. (D) Relative abundance of Lactobacillaceae and Desulfovibrionaceae (n = 4). Data are presented as mean ± SD.Fig. 5 Bioinformatic analysis of gut microbiome. (A) Venn diagram of gut microbiota, and values between different circles represent the number of shared families at the family level. (B) PCoA plot on genus level for the gut microbiota from different groups (n = 4). (C) Relative abundance of Family-level taxa from gut microbiota in different groups. (D) Relative abundance of Lactobacillaceae and Desulfovibrionaceae (n = 4). Data are presented as mean ± SD. Lactic acid bacteria (Lactobacillaceae) are important members of the intestinal commensal microbiota, and their subgroups are thought to affect the progression of IBD, showing protective effects in experimental colitis mouse models. Desulfovibrionaceae, on the other hand, are commonly enriched in the intestines of IBD patients and release hydrogen sulphide (H₂S), a product of sulphate reduction, as a signalling molecule that damages the gut. The community composition was visualised through a community bar plot (Fig. 5C), and the relative abundance of Lactobacillaceae and Desulfovibrionaceae was assessed. In the T-C@pNP treatment group, the abundance of Lactobacillaceae increased, and the abundance of Desulfovibrionaceae decreased, indicating that T-C@pNP may partially modulate the gut microbiota composition (Fig. 5D). Biosafety evaluation TP exhibited significant toxicity to both cell lines, whereas CeO₂ demonstrated a high cell survival rate in the 0–2.5 μg/mL range, with a certain proliferative effect observed at lower concentrations. This mechanism may be related to the ability of CeO₂ to affect the cell cycle, cell adhesion, and oxidative stress, thereby promoting cell proliferation. T-C, T-C@CS/CH, and T-C@pNP all showed good biocompatibility within the 0–2.5 μg/mL range (Fig. 6A).Fig. 6Biocompatibility evaluation of T-C@pNP. (A) The cell viability of different concentrations of TP, CeO2, T-C, T-C@CS/CH, and T-C@pNP exposure in RAW 264.7 and IEC-6 cells after a 24-h incubation (n = 4). (B) H&E staining of major organ sections from different groups (Scale bar = 100 μm). Data are presented as mean ± SD. (C) AST, ALT, BUN, and CRE in mice serum after being treated with T-C@pNP (n = 5).Fig. 6 Biocompatibility evaluation of T-C@pNP. (A) The cell viability of different concentrations of TP, CeO2, T-C, T-C@CS/CH, and T-C@pNP exposure in RAW 264.7 and IEC-6 cells after a 24-h incubation (n = 4). (B) H&E staining of major organ sections from different groups (Scale bar = 100 μm). Data are presented as mean ± SD. (C) AST, ALT, BUN, and CRE in mice serum after being treated with T-C@pNP (n = 5). To evaluate the in vivo safety profile of T-C@pNP, haematoxylin and eosin (H&E) staining of principal organs was performed, and serum biochemical parameters were analyzed. The H&E staining results revealed no significant pathological changes in the major organs of any of the experimental groups (Fig. 6B). Moreover, there were no significant differences in AST, ALT, BUN, or CRE levels among the experimental groups (Fig. 6C). These results indicate that T-C@pNP has good in vivo safety. Conclusions Nanotechnology holds tremendous potential for application (Li et al., 2025; Oyane et al., 2016; Xie et al., 2024). In this study, we developed an inflammation-targeting, colonic pectinase-responsive oral nanoparticle drug delivery system, T-C@pNP, which encapsulates TP-loaded cerium oxide (CeO₂) nanozymes for targeted drug delivery to the site of colitis. T-C@pNP is a mesoporous CeO₂ nanozyme that not only reduces ROS in the colonic tissue but also delivers TP, achieving anti-inflammatory effects, regulating oxidative stress, and potentially repairing the colonic barrier. The outer layers, electrostatically adsorbed with CS and CH, enable inflammation targeting, whereas the PC‑calcium coating aims to prevent premature drug release in the gastrointestinal tract after oral administration and to achieve pectinase-responsive release in the colon. In vitro and in vivo studies have clearly confirmed its excellent ability to downregulate colonic inflammation markers, highlighting its potential for the safe treatment of colitis. Additionally, T-C@pNP significantly decreased the levels of proinflammatory cytokines (IL-6, TNF-α, and IFN-γ) by reducing the levels of ROS, MPO, and LCN2 and possibly inhibiting the NF-κB pathway. Moreover, T-C@pNP also decreased the levels of ROS, which play a key role in modulating the colonic inflammatory microenvironment and inflammatory cycles. T-C@pNP exhibited favourable preliminary biosafety in the present study. However, considering the potential systemic toxicity of triptolide, its clinical translation still requires cautious evaluation (Xi et al., 2017). This nanoparticle-based drug delivery system demonstrated promising therapeutic efficacy in a murine colitis model. 3D bioprinted colon organoids provide a novel platform for further investigation (Xu et al., 2024). In conclusion, our study highlights the potential of T-C@pNP in treating IBD, representing a promising direction for future investigations. CRediT authorship contribution statement Jing Lu: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation. Yiming Chen: Formal analysis, Data curation. Wenxing Ding: Investigation, Formal analysis, Data curation. Shuhui Zheng: Writing – review & editing, Formal analysis. Liping Yang: Formal analysis, Data curation. Yiting Wang: Formal analysis, Data curation. Qingdong Ji: Writing – review & editing, Formal analysis. Chendong Qi: Writing – review & editing, Supervision, Methodology. Feng Gao: Writing – review & editing, Supervision, Methodology, Funding acquisition. Declaration of competing interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.