Colon-targeted microbiota-modulating nanoparticles amplify azathioprine efficacy via gut microbiota axis remodeling in inflammatory bowel disease

Abstract

Background

Inflammatory bowel disease (IBD) is characterized by destruction of the intestinal barrier, dysregulation of mucosal immunity, and imbalance of the gut microbiota homeostasis. Conventional immunosuppressants such as azathioprine (AZA), commonly used in the treatment of IBD, can partially alleviate symptoms but fail to restore normal barrier and immune functions, while exhibiting non-negligible adverse effects. Here, we report an AZA-loaded microbiota-modulating and colon-targeted nanoparticle constructed from pectin, Zein, and Eudragit^®S100 (APZE).

Results

APZE could effectively encapsulate AZA and enhance its cellular and intestinal uptake, thereby improving oral bioavailability in rats compared to AZA suspension. Additionally, its colon targeting ability and mucoadhesive properties prolonged colonic retention, leading to higher colonic accumulation, as indicated by the AUC of colon fluorescence intensity in rats after oral administration of the DiR-labeled formulation. Animal studies demonstrated that APZE significantly reduced inflammation in IBD mice, repaired the intestinal barrier, modulated the gut microbiota, and upregulated short-chain fatty acid levels, exhibiting significant therapeutic efficacy with good safety. The protective effect of APZE against colitis is largely microbiota-dependent. When co-administered with commercially available Bifidobacterium, APZE attenuated colitis effectively, demonstrating excellent therapeutic effects and favorable safety.

Conclusions

In summary, our findings indicate that APZE and Bifidobacterium improved intestinal barrier repair and colitis-related outcomes in a DSS-induced model, supporting their potential as therapeutic agents for IBD.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04136-4.

Background

Inflammatory bowel disease (IBD), which includes Crohn’s disease (CD) and ulcerative colitis (UC), is a chronic inflammatory disorder of the gastrointestinal tract [1–3]. Its course is protracted and prone to relapse, which may cause serious complications or even cancer, seriously affecting the patient’s quality of life [4, 5]. According to statistics, the global age-standardised incidence rate of IBD increased from 4.22 per 100,000 in 1990 to 4.45 per 100,000 in 2021 [6]. The incidence and prevalence of pediatric IBD have also been rising continuously, frequently resulting in growth delays and psychological issues, emerging as an important public health concern [7, 8]. Current treatment strategies for IBD follow a stepwise approach, incorporating aminosalicylates, glucocorticoids, immunosuppressants, biologics, and small-molecule targeted drugs.

Azathioprine (AZA), an immunosuppressant, is recommended by the American College of Gastroenterology guidelines for the maintenance treatment of CD and UC, especially in steroid-dependent patients [9]. AZA is typically administered in oral tablet form. Premature disintegration of the tablets leads to drug absorption in the upper gastrointestinal tract, limiting its accumulation in the colon and reducing therapeutic efficacy [10]. Moreover, systemic exposure to AZA could cause adverse reactions such as hepatotoxicity [11]. To improve efficacy and reduce toxicity, colon-targeted drug delivery systems (CTDDS) have been explored for AZA delivery. Current research mainly focuses on pH-responsive colon-targeted tablets or nanoparticles, which are prepared using polymers such as methacrylate copolymer, polyhydroxyalkanoate terpolymer, and chitosan [12–15]. Eudragit^®S100 (ES100), a methacrylate copolymer, dissolves in the colonic environment at pH values above 7, enabling pH-responsive AZA CTDDS for specific drug release in the colon [12]. However, pH-responsive CTDDS are affected by pH fluctuations in the ileum and the cecum-ascending colon junction. These effects are exacerbated by individual variability and pathological conditions such as inflammation and dysbiosis, potentially causing premature release or insufficient efficacy [16].

Recent studies have increasingly highlighted the association between IBD progression and gut microbiota dysbiosis. The gut microbiota plays a crucial role in maintaining intestinal homeostasis and regulating human health [17, 18]. Dysbiosis of the gut microbiota may trigger the onset of IBD, while the intestinal inflammation caused by IBD may, in turn, affect the composition and function of the gut microbiota, leading to dysbiosis [19–21]. In IBD patients, dysbiosis is primarily characterized by reduced microbial diversity and an imbalance between commensal and pathogenic bacteria. For instance, the abundance of pro-inflammatory bacteria such as Enterobacteriaceae and Clostridiaceae increases, while that of anti-inflammatory bacteria like Firmicutes and protective commensal bacteria, such as Bifidobacterium, Lactobacillus, and Clostridium leptum decreases [22]. This imbalance triggers abnormal immune responses, exacerbating chronic inflammation and compromising the intestinal barrier. Pathogenic bacteria can penetrate deeper tissues, further intensifying local inflammation [23]. Microbiota modulation strategies have demonstrated therapeutic potential [24–26]. In a Phase II clinical trial, fecal microbiota transplantation from healthy donors into mild-to-moderate UC patients achieved clinical remission in approximately 30% of participants [27]. Given these findings, there is a growing need to develop a delivery system with microbiota-modulating and colon-targeting capabilities for AZA, thereby achieving enhanced efficacy while reducing toxicity.

Pectin (PT), a plant-derived prebiotic with mucoadhesive properties, promotes the growth of beneficial bacteria such as Lactobacillus and Bifidobacterium [28, 29]. While resistant to human digestive enzymes, PT is degraded by gut microbiota enzymes like pectinase and eventually metabolized into short-chain fatty acids (SCFAs), which support intestinal immune homeostasis and gut barrier function [30]. Zein, a corn-derived protein, exhibits excellent biocompatibility, biodegradability, and controlled-release properties [31]. Contado, C. et al. developed Eudragit-coated Zein/PT nanoparticles for resveratrol, and in vitro studies demonstrated that this formulation could be used for colon drug delivery [32].

Herein, we designed an AZA-loaded nanoparticle constructed from PT, Zein, and ES100 (APZE) to achieve colon-targeted drug delivery. Given that PT can participate in regulating gut microbiota and the association between IBD and gut microbiota, we further investigated the effectiveness of APZE in the treatment of IBD and in modulating gut microbiota. This oral nano delivery system is capable of targeting the colon, enhancing colonic adhesion and absorption, modulating the gut microbiota, thereby achieving enhanced therapeutic effects while reducing toxicity, effectively alleviating the symptoms of IBD. We investigated the physicochemical properties of APZE, as well as its pharmacokinetic characteristics in vitro and in vivo. Additionally, we evaluated the therapeutic effects of APZE in an acute colitis mouse model induced by dextran sulfate sodium (DSS). Moreover, we assessed its impact on intestinal barrier integrity, mucosal immunity, and gut microbiota homeostasis. Finally, we examined the influence of gut microbiota on the therapeutic efficacy of APZE, as well as the treatment outcomes of APZE in combination with the commercially available probiotic Golden Bifid (Bifidobacterium longum, Lactobacillus bulgaricus, and Streptococcus thermophilus) (Scheme 1).

Scheme 1.

The AZA-loaded microbiota-modulating and colon-targeted nanoparticles are constructed from PT, Zein, and ES100 (APZE). APZE accumulates at inflammation sites through colonic targeting. APZE can improve the bioavailability of the encapsulated drug due to its excellent adhesion and permeability properties. As an oral nanomedicine, APZE exhibits good efficacy for IBD by improving intestinal barrier repair, reducing inflammation, and modulating the gut microbiota. Graphical abstract was created in https://BioRender.com

Methods

Materials and animals

AZA was purchased from Aladdin (Shanghai, China). Zein, ammonia solution, 1,1’-dioctadecyl-3,3,3’,3’-tetramethylindotricarbocyanine iodide (DiR), and 4’,6-diamidino-2-phenylindole (DAPI) were purchased from Macklin (Shanghai, China). PT, ES100, and sodium carboxymethyl cellulose (CMC-Na) were obtained from Sigma-Aldrich (St. Louis, Missouri, USA). Pluronic^®188 (F68) was purchased from yuanye Bio-Technology (Shanghai, China). Methanol, acetonitrile and formic acid were purchased from Thermo Fisher Scientific (Waltham, USA). Methyl-β-cyclodextrin (M-β-CD) was purchased from J&K Scientific (Beijing, China). Genistein (Gen) was purchased from Selleck Chemicals (Texas, USA). Amiloride hydrochloride (AMI-HCl), Brefeldin A (BFA), Monensin (MON), Nocodazole (NOC), bafilomycin A1 (Baf A1), and Chlorpromazine (CPZ) were purchased from MedChemExpress (New Jersey, USA). DSS was obtained from MP Biomedicals (Eschwege, Germany). Metronidazole, and neomycin were purchased from Solarbio (Beijing, China). Vancomycin was purchased from Biorigin (Beijing, China). Ampicillin was obtained from Sangon Biotech (Shanghai, China). Golden Bifid mixed powder (GB) was provided by Wanze Shuangqi Pharmaceutical Co., Ltd. (Inner Mongolia, China). Caco-2 cells were provided by the Institute of Cell Biology, Chinese Academy of Sciences (Shanghai, China). Male C57BL/6J mice (22–25 g) and male SD rats (300–340 g) were obtained from the Department of Laboratory Animal Science, Peking University Health Science Center.

System transition and animal selection

The study followed a stepwise in vitro to in vivo workflow commonly used in oral drug delivery system development. Caco-2 cells and Caco-2 cell monolayers were first used to assess cellular uptake and transport mechanisms under controlled conditions. We then performed distribution, pharmacokinetics, and intestinal transport studies in rats, whose larger body size and blood volume enable ex vivo tissue collection, imaging-based quantification of intestinal distribution, and serial blood sampling. Therapeutic efficacy and microbiota-related outcomes were evaluated in C57BL/6J mice, as the DSS-induced colitis model is well-established in mice and robustly reproduces key IBD features [13, 25]. Mice also allow larger experimental groups and sample sizes, improving statistical power. Notably, there is no intrinsic technical barrier to conducting distribution in mice or DSS-induced colitis in rats. The selection was made to balance feasibility, accuracy, animal welfare, and experimental efficiency for each endpoint.

Preparation and characterization of nanoparticles

The AZA-loaded microbiota-modulating and colon-targeted nanoparticles (APZE) were prepared using a composite solvent-antisolvent co-precipitation method. First, Zein (15 mg, 30 mg/mL in 80% ethanol) was mixed with AZA (3 mg, 75 mg/mL in ammonia solution), followed by the addition of ES100 (15 mg, 30 mg/mL in acetone). The mixture was then slowly injected into a 2.5 mL solution containing 1 mg/mL PT and 0.5% F68 (organic phase/water phase ratio of 2:5, v/v), and was stirred at 600 rpm for 20 min. Ethanol and acetone were removed under reduced pressure at 40 °C using a rotary evaporator. This formed the nanoparticles, which were collected via centrifugation, then washed with deionized water and resuspended in water before being freeze-dried. The blank formulation (PZE) was prepared without AZA. For cellular uptake and in vivo tracking experiments, the DiR-labeled formulation (DAPZE) was prepared by co-encapsulating DiR and AZA, following the same procedure used for APZE [33]. For the control formulation, AZA suspensions (AZAS) were prepared by uniformly dispersing AZA in a 0.5% (w/v) CMC-Na solution. DiR and AZA suspensions (DAZAS) were prepared using the same method as that for AZAS. AZA tablet suspensions (AZAS-Tablet) were prepared by uniformly dispersing AZA tablet powder in a 0.5% (w/v) CMC-Na solution.

The particle size, polydispersity index (PDI), and zeta potential of APZE and PZE were measured by a Zetasizer Nano ZSP (Malvern) at 25 °C. The morphology was observed using a transmission electron microscope (JEM-1400, JEOL). The APZE lyophilized powder was dissolved in 50% aqueous methanol solution via ultrasonication, filtered through a 0.22 μm microporous membrane, and analyzed by HPLC (Shimadzu). Column: Thermo C18 (150 × 4.6 mm, 5 μm); mobile phase: methanol-water-acetic acid (20:80:1, v/v/v); detection wavelength: 280 nm; flow rate: 1.0 mL/min; column temperature: 40 °C; injection volume: 10 µL [34]. The drug loading content (DLC, %) and encapsulation efficiency (EE, %) were calculated using Eqs. (1) and (2):

1

2

The chemical associations in nanoparticles were revealed by a Fourier Transform Infrared spectrometer (Nicolet) and an X-ray diffractometer (Bruker).

In vitro drug release study

The in vitro drug release experiments of AZAS-Tablet, AZAS, and APZE were conducted to simulate drug transport through the gastrointestinal tract, specifically through the stomach (pH 1.2, 0–2 h), small intestine (pH 6.8, 2–6 h), and colon (pH 7.4, 6–24 h). In brief, 300 µL of AZA formulations at a concentration of 1.4 mg/mL were placed in 3000 kDa molecular weight cut-off (MWCO) dialysis tubes. The tubes were then immersed in PBS with gradually changing pH values and incubated on a shaker (100 rpm, 37 °C). To aid the dissolution of AZA, 0.02% (w/v) Tween-80 was added to the PBS. Samples (200 µL) were taken at 0, 0.25, 0.5, 1, 2, 3, 4, 6, 8, 10, 12, and 24 h, with fresh PBS replenished. After centrifugation at 14,500 rpm for 5 min, the supernatants were analyzed for AZA concentration using HPLC.

Cell culture and cell viability analysis

Caco-2 cells were cultured in DMEM with 10% fetal bovine serum, 100 U/mL penicillin, 100 U/mL streptomycin, and 1% non-essential amino acids in a humidified incubator containing 5% CO₂ at 37 °C. The effects of AZAS and APZE on cell viability were determined by CCK-8 assay. Briefly, Caco-2 cells were seeded into a 96-well plate at a density of 1 × 10⁴ cells/mL. After the cells adhered to the plate, the culture medium was replaced with medium containing varying concentrations of AZAS or APZE. The cells were incubated for 4 h, after which the drug-containing medium was replaced with serum-free medium containing 10% CCK-8 solution. After an additional 4-h incubation, the optical density at 450 nm was measured using a microplate reader, and cell viability was calculated according to the instructions.

In vitro cellular uptake assay

Confocal laser scanning microscopy (CLSM) and flow cytometry were used to qualitatively and quantitatively examine the cellular uptake of DAZAS and DAPZE, respectively. Briefly, Caco-2 cells were seeded at 1 × 10⁵ cells/well in 12-well plates or glass-bottomed cell culture dishes and incubated for 24 h. For flow cytometry, the uptake was assessed under varying temperatures (4℃ and 37℃), cell states (adherent and suspended), concentrations (50, 100, 200, and 400 µM), and incubation times (0.5, 1, 2, and 4 h). After incubation, the cells were rinsed with cold PBS, harvested, and analyzed using a flow cytometer (BD Biosciences). For CLSM, cells were incubated under fixed conditions, then rinsed with cold PBS, fixed with 4% paraformaldehyde (PFA), stained with DAPI, and observed with CLSM (Leica Microsystems). DAPI and DiR were excited at 405 and 638 nm, respectively.

Endocytosis and exocytosis mechanism in Caco-2 cells

Before conducting endocytosis and exocytosis experiments, the effects of different concentrations of inhibitors on cell viability were assessed using the CCK-8 assay to ensure the inhibitors would not affect cell activity during co-culture. To investigate endocytic mechanism, Caco-2 cells were pretreated at 37℃ with endocytosis inhibitors, including CPZ (10 µg/mL), Gen (0.1 mM), AMI-HCl (0.1 mM), and M-β-CD (5 mM) for 1 h, followed by incubation with DAPZE for 2 h. After incubation, the cells were rinsed with cold PBS, harvested, and analyzed by flow cytometer. To investigate exocytosis mechanism, Caco-2 cells were incubated at 37℃ with DAPZE for 2 h, and then incubated with intracellular transport inhibitors, including BFA (25 µg/mL), MON (32.5 µg/mL), NOC (10 µg/mL), and Baf A1 (100 nM) for 3 h. Cells were analyzed using a flow cytometer [35].

Transport within Caco-2 cell monolayers

To establish an in vitro Caco-2 cell monolayer model, Caco-2 cells were seeded in 12-well Transwell^® plates. 0.5 mL of cell suspension containing 2 × 10⁵ cells was added to the upper compartment of each Transwell insert, and 1.5 mL of culture medium was added to the lower compartment. The plates were incubated for 21 days to allow cell differentiation and monolayer formation. Permeation experiments were conducted to assess bidirectional transport. For AP-BL permeability (absorptive transport), medium containing 200 µM AZAS or APZE was added to the apical (AP) side, and blank medium was added to the basolateral (BL) side. For BL-AP permeability (secretory transport), the donor and receiver compartments were reversed. During the experiment, samples were withdrawn from the receiving chamber at 30, 60, 90, 120, and 180 min, and each sampling volume was replaced by an equal volume of fresh pre-warmed medium [36]. Samples were analyzed by HPLC, and the apparent permeability coefficient (P_app) was calculated as shown in Eq. (3):

3

where V (cm³) is the volume of the solution received from the receiving chamber; A (cm²) is the membrane area; C₀ is the initial concentration of the test compound; △Q/△t is the rate of permeability. The efflux ratio (ER) was determined by calculating the ratio of P_app (BL→AP) versus P_app (AP→BL).

Transport mechanism across the Caco-2 cell monolayers

For transepithelial transport studies in Caco-2 cell monolayers, cells cultured in Transwell plates were incubated with DAZAS and DAPZE at 37℃ for 2 h. The cells were then rinsed with cold PBS, fixed, and stained with DAPI. Tomographic scanning was performed using CLSM in Z-stack mode for dual-channel detection. Confocal images were analyzed in X-Y-Z mode. To investigate the endocytosis and exocytosis pathways of DAPZE, the cells were treated with inhibitors including M-β-CD and Baf A1.

In vivo imaging and assay in rats

To assess the distribution of DAZAS and DAPZE in various tissues after oral administration, fluorescence imaging studies were performed on rats. SD rats were randomly divided into two groups (n = 24): DAZAS and DAPZE. At 0.5, 1, 2, 3, 5, 9, 12, and 24 h after oral administration, fluorescent images were captured using an In Vivo IVIS Spectral imaging system (PerkinElmer). The gastrointestinal tracts, major organs, and mesenteric lymph nodes were then surgically collected and imaged to evaluate drug distribution. Intestinal segments (duodenum, jejunum, ileum, and colon) were collected, fixed, embedded in paraffin, sectioned, stained with DAPI, and imaged using CLSM.

In vivo pharmacokinetics in SD rats

SD rats were randomly divided into two groups (n = 8) and administered 15 mg/kg AZAS or APZE via oral gavage. Blood samples were collected at 0.5, 1, 2, 3, 4, 5, 7, 9, 12, and 24 h post-administration. The samples were centrifuged at 3,000 rpm for 10 min to separate the plasma. AZA metabolites, including 6-mercaptopurine (6-MP) and 6-methylmercaptopurine (MMP), were then quantified using the LC-MS/MS method previously established by our research group [37].

Intestinal adhesion

To evaluate the mucoadhesive properties of AZAS and APZE in different intestinal segments, ex vivo everted rat gut sac experiments were conducted. In brief, SD rats were fasted overnight with access to water, then sacrificed, and the intestines were exposed by incising along the abdominal midline. Approximately 2 cm segments of the duodenum, jejunum, ileum, and colon were excised and washed with preheated PBS. The intestinal segments were everted using a glass rod, immersed in PBS dispersions of AZAS or APZE, and shaken at 100 rpm at 37℃. The segments were incubated for 10, 30, 60, and 120 min before removal from the dispersions [38]. The remaining AZA concentration in the dispersions was measured by HPLC described in Sect. “Preparation and characterization of nanoparticles”.

Intestinal permeability

Approximately 10 cm segments of the duodenum, jejunum, ileum, and colon were excised from SD rats and washed with preheated PBS. Each segment was tied at one end and everted with a glass rod. The everted sacs were ligated to a hollow glass tube at the other end for sampling. The sacs were filled with PBS and immersed in PBS dispersions containing AZAS or APZE. The temperature was maintained at 37℃, and continuous aeration was provided to the dispersions to ensure the viability of the gut sacs. At 10, 30, 60, 90, 120, and 180 min, samples were withdrawn from the sac interior and replaced with an equal volume of fresh PBS [39]. The AZA content in the samples was measured using HPLC. The flux (F) and P_app for each intestinal segment were calculated using Eqs. (4) and (5):

4

5

where Q (µg) is the cumulative drug amount permeated at each time point. Linear regression of Q versus time provides the slope (dQ/dt). A (cm²) is the intestinal absorption area, and C₀ (µg/mL) is the initial drug concentration outside the intestinal sac.

Intestinal perfusion

SD rats were fasted overnight and anesthetized before surgery. A midline abdominal incision was made, and the desired intestinal segments were carefully exposed to maintain blood supply. The segments were cannulated and secured with tubes at both ends, then rinsed with normal saline. Subsequently, PBS dispersions of AZAS or APZE were perfused through the segment at a flow rate of 0.25 mL/min for 40 min to reach a steady state. Perfusate samples were then collected every 10 min using pre-weighed tubes. This process continued for 90 min. After the experiment, the rats were sacrificed, and the radius and length of the perfused intestinal segments were measured [38]. The AZA content in the collected samples was quantified by HPLC. The absorption percentage (P%), absorption rate constant (K_a) and effective permeability coefficient (P_eff) for each intestinal segment were calculated according to Eqs. (6), (7) and (8):

6

7

8

where C_in and C_out are the drug concentrations in the inlet and outlet perfusate, respectively. V_in and V_out are the perfusion volumes at the inlet and outlet, respectively; r is the internal radius of the intestine (cm), l is the length of the intestine (cm), and Q_in is the perfusion flow rate (mL/min).

In vivo evaluation of the therapeutic study on DSS-induced IBD mice

Acute colitis was induced in 8-week-old male C57/BL6J mice by adding 2.5% (w/v) DSS to their drinking water for 5 days. Prior to the in vivo study, the mice were acclimated for 7 days under standard conditions. Subsequently, the mice were randomly divided into five groups (n = 10): Control, DSS, AZAS, APZE, and PZE. The treatments were as follows: (1) Control: normal water; (2) DSS: DSS from Day 0 to Day 5; (3) AZAS: DSS from Day 0 to Day 5 and daily gavage with AZAS from Day 0 to Day 10; (4) APZE: DSS from Day 0 to Day 5 and daily gavage with APZE from Day 0 to Day 10; (5) PZE: DSS from Day 0 to Day 5 and daily gavage with PZE from Day 0 to Day 10. The AZA dosage was 20 mg/kg. The mice were monitored daily for body weight, diet and water intake, mental state, fecal appearance, and fecal occult blood. Colitis progression was scored using the Disease Activity Index (DAI). DAI = (weight loss + fecal consistency + blood in feces)/3 (Table S1). Feces were collected on Day 0 and Day 10 for microbiome analysis. On Day 10, blood was collected for white blood cell (WBC) count, and then the mice were sacrificed by cervical dislocation. The colon was collected, photographed, and its length was recorded. The distal colon (0.5–1 cm) was fixed for histological analysis, and the remaining tissue was used for myeloperoxidase (MPO) activity and cytokine level measurements. Cecal contents were collected for SCFAs analysis. Major organs were harvested and preserved in 4% PFA for safety analysis.

Histopathological analysis

The fixed colonic tissues and major organs were embedded, sectioned, and stained with hematoxylin and eosin (H&E) according to standard protocols. These sections were analyzed to assess histopathological alterations in colon and evaluate the safety profiles of AZAS, APZE, and PZE. Colonic tissue sections were stained with periodic acid-schiff (PAS) and alcian blue (AB) to evaluate goblet cell distribution and mucin in the colon. Immunofluorescence staining was performed on colonic sections, and the distribution of tight junction proteins ZO-1 and Occludin was observed using CLSM. Fluorescence intensity was quantitatively analyzed to assess protein expression levels. Additionally, the terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining was carried out on colonic and hepatic tissue sections to analyze apoptosis.

MPO assay and inflammatory cytokine analysis

MPO activity in colon tissue was measured using an MPO assay kit (colorimetric method). Cytokine levels (TNF-α, IFN-γ, IL-1β, and IL-6) in colon tissue were quantified using ELISA kits.

Fecal DNA extraction and 16 S rDNA amplicon sequencing

Genomic DNA was extracted from mouse feces using a Fecal DNA extraction kit, and 16 S rDNA amplicon sequencing was performed. Briefly, the V3 + V4 regions were amplified using specific primers, and small fragment libraries were constructed. Paired-end sequencing (PE250) was performed on the Illumina NovaSeq platform. After reads were assembled and filtered, Operational Taxonomic Units (OTUs) clustering was carried out. Based on OTUs and annotation results, species complexity, diversity, and correlation analysis were carried.

Detection of SCFAs

The cecal contents were dehydrated using a freeze dryer, and the powder was extracted with a 50% (v/v) aqueous isopropanol solution. After homogenization and centrifugation, the supernatant was collected. Following derivatization, the concentrations of acetate, propionate, and butyrate in the cecal contents were quantified using an LC-MS/MS method [40].

In vivo therapeutic study of APZE on pseudo-germ-free IBD mice

To investigate the role of the gut microbiota in the protective effects of APZE against DSS-induced colitis, C57BL/6J mice were randomly divided into four groups (n = 10): Control, DSS, AZAS, and APZE. All mice were orally administered an antibiotic cocktail (ABX, 50 mg/kg vancomycin, 100 mg/kg neomycin, 100 mg/kg metronidazole, and 100 mg/kg ampicillin) for 7 days to deplete the resident gut microbiota. During the subsequent treatment period, ABX administration was continued to sustain the microbiota-depleted state. The induction and treatment of colitis followed the same protocol described earlier. The status of mice was monitored daily. At the end of the experiment, blood samples were collected for leukocyte count, the spleen was excised and weighed, and colon tissues were harvested for imaging and length measurement. The distal colon was fixed for histological analysis and immunofluorescence staining.

Combined therapeutic study of APZE and GB

To investigate the role of Bifidobacterium in the therapeutic process, DSS-induced IBD mice were administered commercial GB at different time points. Mice were randomly divided into six groups (n = 8): Control, DSS, GB, pGB, AZAS + GB, and APZE + GB. The treatments were as follows: (1) Control: normal water; (2) DSS: DSS from Day 8 to Day 13; (3) GB (GB treatment): DSS from Day 8 to Day 13 and daily gavage with GB from Day 8 to Day 18; (4) pGB (GB prevention): DSS from Day 8 to Day 13 and daily gavage with GB from Day 0 to Day 18; (5) AZAS + GB: DSS from Day 8 to Day 13, daily gavage with AZAS from Day 8 to Day 18, and daily gavage with GB from Day 14 to Day 18, with at least a 4-h interval between the two oral gavages; (6) APZE + GB: DSS from Day 8 to Day 13, daily gavage with APZE from Day 8 to Day 18, and daily gavage with GB from Day 14 to Day 18. The AZA dosage was 20 mg/kg. The GB powder contained Bifidobacterium longum, Lactobacillus bulgaricus, and Streptococcus thermophilus in a 10:1:1 ratio, with a total viable count of 6 × 10⁸ CFU/g. The daily dose was 1 × 10⁸ CFU per mouse. The status of mice was monitored daily. At the end of the experiment, tissue harvesting and processing were performed as described previously.

Statistical analysis

The experimental data were processed using GraphPad Prism 9.0 software. The quantitative data were expressed as mean±SD. One-way analysis of variance (ANOVA) was used for comparisons between multiple groups, and t-test was used for comparisons between two groups. P < 0.05 was considered statistically significant.

Results

Preparation and characterization of APZE

As shown in Fig. 1A, AZA-loaded microbiota-modulating and colon-targeted nanoparticles (APZE) were successfully prepared using a composite solvent-antisolvent co-precipitation method. The excipients included Zein, microbiota-degradable PT, and pH-responsive ES100. Nanoparticles appeared as light yellow, translucent aqueous dispersions (Fig. S1A and C). Dynamic light scattering (DLS) measurements showed that the hydrodynamic size of APZE and blank PZE were 21.19 ± 3.10 nm and 20.89 ± 9.54 nm, with polydispersity indexes (PDI) of 0.268 ± 0.028 and 0.332 ± 0.010, respectively. Zeta potentials were −32.50 ± 5.06 mV and −34.40 ± 1.35 mV for APZE and PZE, respectively. Both nanoparticles exhibited uniform spherical morphologies under transmission electron microscopy (TEM) (Fig. 1B-E and Fig. S1B and D). These results suggested that the nanoparticles were well-dispersed. The encapsulation efficiency for AZA was 90.79 ± 2.73%, with drug loading capacity of 6.77 ± 0.19% (n = 3). X-ray diffraction and Fourier transform infrared analysis showed the absence of characteristic crystalline peaks of AZA, and no new absorption bands appeared in APZE lyophilized powder (Fig. 1F and G). These findings suggest that AZA exists in an amorphous state within APZE and is loaded through non-covalent interactions.

Fig. 1.

Preparation and characterization of APZE and blank PZE. (A) APZE is formulated by a composite solvent-antisolvent co-precipitation method using Zein, PT, and ES100. (B-C) Representative TEM image of (B) APZE and (C) PZE. Scale bar = 100 nm. (D-E) Particle size and distribution of (D) APZE and (E) PZE. (F) X-ray diffraction patterns of AZA and APZE. (G) Fourier transform infrared spectra of AZA, PT, Zein, ES100, F68, APZE, and PZE

In vitro drug release study

To investigate the release behavior of APZE, we measured the release profiles of AZA from different formulations in simulated media with gradually changing pH values, which represent the pH conditions of the stomach, small intestine, and colon, respectively. As shown in Fig. S2, AZAS-Tablet and AZAS exhibited rapid release, with their release rates unaffected by changes in pH in the simulated release media. In simulated gastric fluid (pH 1.2, 0–2 h) and simulated intestinal fluid (pH 6.8, 2–6 h), more than 80% of AZA was released from both AZAS-Tablet and AZAS. The premature drug release from AZAS-Tablet and AZAS could lead to drug absorption in the stomach and small intestine, potentially causing adverse side effects and preventing the drug from reaching the inflamed colon. In contrast, the release of APZE was below 30% within the first 6 h, and its release rate increased in simulated colonic fluid (pH 7.4), with more than 70% released at 12 h and over 90% at 24 h. These results demonstrated that APZE could effectively retain the encapsulated drug until reaching the colon, improving the drug availability in the colon, which is a promising nano formulation for colon-targeted delivery.

In vitro cellular uptake and transport mechanisms of APZE

To evaluate the absorption advantage of APZE compared to AZAS, we investigated the cellular uptake, endocytosis, and transport pathways using Caco-2 cells and monolayers simulating small intestinal epithelial cells. First, the CCK-8 assay was conducted to determine the toxicity of both formulations on Caco-2 cells at different AZA concentrations. As shown in Fig. 2A, incubation with AZAS or APZE at 400 µM for 4 h showed no apparent toxicity to Caco-2 cells. Next, qualitative and quantitative uptake studies of DAZAS and DAPZE were performed using CLSM and flow cytometry, respectively. DAZAS and DAPZE were used to trace AZAS and APZE. DAPZE, which co-encapsulated AZA and fluorescent probe DiR, had similar physicochemical properties to APZE. Compared with DAZAS, DAPZE obviously enhanced cellular uptake of the encapsulated agent (Fig. 2B-E and Fig. S3A-E). At 37 °C, DAPZE uptake was significantly higher than that at 4 °C, indicating energy-dependent absorption. Additionally, DAPZE uptake increased with incubation time or concentration, demonstrating that internalization is both time- and concentration-dependent. As displayed in Fig. 2F, the absorption of nanoparticles was higher in suspension cells than adherent cells, likely due to the larger contact area between suspension cells and nanoparticles.

Fig. 2.

Uptake, endocytosis, and transport pathways of APZE in Caco-2 cells and cell monolayer models. (A-H) Experiments on Caco-2 cells. (A) The cytotoxicity of AZA suspension (AZAS) and APZE measured by the CCK-8 assay. (B-D) The uptake of DiR-labeled AZAS (DAZAS) and APZE (DAPZE) at different conditions analyzed by flow cytometry. The cellular uptake at different (B) temperatures, (C) times, and (D) concentrations. (E) Confocal laser scanning microscopy (CLSM) imaging of the uptake of DAZAS and DAPZE. From left to right: brightfield, nuclei (blue), DiR (red), and merged images. Scale bars = 200 μm and 100 μm. (F) The uptake of DAPZE by adherent Caco-2 cells (ACU) and suspended Caco-2 cells (SCU). (G) Effect of different endocytosis inhibitors on DAPZE internalization. Endocytosis inhibitors: chlorpromazine (CPZ), genistein (Gen), amiloride hydrochloride (AMI-HCl), and methyl-β-cyclodextrin (M-β-CD). (H) Effect of different intracellular transport inhibitors on DAPZE efflux. Intracellular transport inhibitors: brefeldin A (BFA), monensin (MON), nocodazole (NOC), and bafilomycin A1 (Baf A1). (I-L) Experiments on Caco-2 cell monolayers. (I) Bidirectional cumulative transport of AZAS and APZE over time. (J) CLSM images of the uptake of DAZAS and DAPZE. From top to bottom: brightfield, nuclei (blue), DiR (red), and merged images. (K) Inhibition effect of M-β-CD on DAPZE endocytosis. (L) Inhibition effect of Baf A1 on DAPZE efflux. n = 3 for (A, C, D, F, I), n = 4 for (B, G, H). Data are shown as mean±SD, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

We then explored the cellular endocytosis and efflux mechanisms of DAPZE. First, we confirmed that the inhibitors used did not exhibit noticeable toxicity to Caco-2 cells at applied concentrations (Fig. S3F). As seen in Fig. 2G and Fig. S3G, compared to the control group, Gen, AMI-HCl, and M-β-CD treatment significantly reduced cellular uptake of DAPZE by 28.4%, 30.8%, and 53.8%, respectively. This suggested that both lipid raft/caveolae-dependent and macropinocytosis-dependent endocytosis are involved in DAPZE internalization. As previously reported, nanoparticles generally undergo exocytosis after internalization, being extruded from the apical plasma membrane of Caco-2 cells [35]. To assess the efflux mechanisms, Caco-2 cells were treated with efflux inhibitors along with DAPZE. Figure 2H and Fig. S3H display that the accumulation of DAPZE in cells treated with Baf A1 increased 2.44-fold compared to the control group, indicating that endosomal acidification significantly inhibits DAPZE exocytosis [41].

Subsequently, the bidirectional transport of AZAS and APZE across Caco-2 cell monolayers was studied. As shown in Fig. 2I, the cumulative transport of AZAS and APZE increased with exposure time. As displayed in Table S2, the cumulative permeation of APZE from AP to BL at 180 min (Q_{180 min}, 0.99 µg) and its P_app (2.78 × 10⁻⁴ cm/s) were higher than those of AZAS (Q_{180 min}, 0.70 µg; P_app, 1.91 × 10⁻⁴ cm/s), and the ER for both was less than 0.6. This indicated that both AZAS and APZE were well absorbed and mainly taken up by the transporters on the AP membrane of the small intestine [36]. Compared to AZAS, APZE was more efficiently transported across Caco-2 cell monolayers due to its higher P_app values.

The effectiveness of DAPZE in promoting the transport of the encapsulated drug across Caco-2 cell monolayers was qualitatively observed using CLSM. As shown in Fig. 2J, contrasted with DAZAS group, significantly higher red fluorescence was observed in DAPZE group, which revealed that DAPZE promoted the intracellular accumulation of the encapsulated agent, potentially facilitating transcellular transport. DAPZE is endocytosed through the AP membrane into epithelial cells, and the internalized particles are subsequently extruded through either the AP or BL membrane. As shown in Fig. 2K, Caco-2 cell monolayers treated with M-β-CD exhibited less intracellular red fluorescence than untreated group, confirming that lipid raft/caveolae-mediated endocytosis plays a key role in DAPZE transport across cell monolayers. As seen in Fig. 2L, compared to the control, higher intracellular red fluorescence accumulated in Baf A1 group, indicating that endosomal acidification is involved in the transcellular transport of DAPZE. Therefore, our study demonstrated that DAPZE exhibits superior cellular uptake compared to DAZAS, utilizing lipid rafts/caveolae and endosomal acidification pathways for transport in both cells and cellular monolayers.

In vivo distribution of APZE

The biodistribution and intestinal accumulation of DAZAS and DAPZE after oral administration were investigated using in vivo fluorescence imaging. As displayed in Fig. 3A and Fig. S4A, overall fluorescence signal was consistently stronger in the DAPZE group than that in the DAZAS group at all time points, suggesting that DAPZE may enhance the accumulation and absorption of the encapsulated agent in vivo. As shown in Fig. 3B and C and Fig. S4B and C, fluorescence across different intestinal segments gradually migrated and accumulated from duodenum to jejunum, ileum, and colon over time in the DAPZE group. Within the first 3 h, fluorescence mainly migrated from the duodenum to the ileum. Then, colon fluorescence intensity maintained at a high level from 3 h to the experimental endpoint at 24 h. DAPZE exhibited a 23.38-fold higher area under the curve (AUC) of colon fluorescence intensity than DAZAS, substantially enhancing drug targeting to the colon (Table S3). Fluorescence imaging of major organs showed that the liver was the main aggregation organ in the DAPZE and DAZAS groups (Fig. S4D).

Fig. 3.

In vivo distribution of DAZAS and DAPZE after oral administration in rats. (A-B) Fluorescence images of (A) whole body and (B) gastrointestinal tract at 0.5, 1, 2, 3, 5, 7, 12, and 24 h after drug administration. (C) Quantitative analysis of intestinal fluorescence intensity in the gastrointestinal tract of rats (n = 3). (D-F) CLSM images of sections from different intestinal segments at (D) 2 h, (E) 7 h, and (F) 24 h. From left to right: brightfield, nuclei (blue), DiR (red), and merged images. From top to bottom: duodenum, jejunum, ileum, and colon. Scale bar = 200 μm

Cellular uptake within intestinal tissues was assessed using CLSM. As seen in Fig. 3D-F, distinct red fluorescence was observed in the ileal villi of the DAPZE group at 2 h, indicating that DAPZE was primarily absorbed in ileum at 2 h. By 7 h, the agent had accumulated in the colon, with pronounced fluorescence still detectable in colonic cells until 24 h (Fig. S5A-E). Overall, the biodistribution results demonstrated significant colonic accumulation of DAPZE, supporting further evaluation of its in vivo anti-colitis efficacy.

Pharmacokinetics and intestinal transport studies of APZE in rats

A pharmacokinetic study was conducted to quantitatively assess the ability of APZE to improve bioavailability. As AZA is rapidly metabolized to 6-MP after absorption, we quantified 6-MP concentrations in plasma to reflect the pharmacokinetic profile of AZA [42].Compared with AZAS, APZE showed a 1.94-fold increase in mean AUC and a 2.87-fold increase in mean peak concentration (C_max) (Fig. 4A and Table S4). Given that 6-MP can be further metabolized to MMP, which is considered one of the main contributors to hepatotoxicity, we also quantified MMP concentrations in plasma (Fig. 4B). The results indicated that the AUC of MMP in APZE group was only slightly higher than that in AZAS group, suggesting that while the bioavailability of APZE was remarkably increased, the toxic metabolites did not cause a significant increase in hepatotoxicity.

Fig. 4.

Pharmacokinetics and intestinal transport studies of APZE in rats. (A–B) Blood concentration-time curves of (A) 6-mercaptopurine (6-MP) and (B) methylmercaptopurine (MMP) in rats. (C) Adhesion amounts in duodenum, jejunum, ileum, and colon segments after incubation with AZAS or APZE at different time points. (D) Cumulative permeation amounts in different intestinal segments after incubation with AZAS or APZE at different time points. (E–G) In vivo intestinal absorption parameters of AZAS and APZE in different intestinal segments: (E) absorption percentage (P%), (F) absorption rate constant (K_a), (G) effective permeability coefficient (P_eff). n = 8 for (A, B), n = 3 for (C, D), n = 6 for (E–G). Data are shown as mean ± SD, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Next, we explored the absorption process of APZE through localized intestinal studies. First, we assessed the adhesiveness and permeability of APZE in different intestinal segments using ex vivo everted rat gut sacs [43]. As shown in Fig. 4C, within the first 30 min, APZE exhibited significantly higher mucosal adhesion in all intestinal segments compared to AZAS. This could be attributed to the smaller particle size of APZE and the adhesive properties of its excipients, PT and Zein [44]. After 60 min, the mucosal adhesion amounts of AZAS in intestinal segments became similar to those of APZE, possibly due to gradual permeation and saturation of the agent [39]. In the rat intestinal permeability study (Fig. 4D and Table S5), the cumulative permeation amount exhibited a linear relationship with time. The cumulative permeation amount (Q_180min), F, and P_app of APZE in all intestinal segments were higher than those of AZAS, indicating that APZE is more easily permeated.

An in vivo intestinal perfusion experiment was conducted in rats to compare the absorption of AZAS and APZE in different intestinal segments. As shown in Fig. 4E-G and Table S6, both formulations exhibited the highest absorption in the duodenum. In the jejunum and colon, the absorption parameters of APZE were higher than those of AZAS. Specifically, in the colon (the lesion site of IBD), mean P%, K_a, and P_eff values in the APZE group increased by 32.8%, 31.0%, and 33.6%, respectively, compared to those of AZAS. APZE demonstrates excellent adhesion and permeability properties, and is more easily absorbed by the intestinal mucosa than AZAS.

In vivo evaluation of the therapeutic effects of APZE on DSS-induced IBD mice

Further, the therapeutic efficacy of APZE was evaluated using a DSS-induced IBD mouse model. The modeling and administration process is shown in Fig. 5A. As seen in Fig. 5B-E, mice treated with DSS exhibited significant weight loss, elevated DAI scores, and shortened colon length, confirming successful IBD model establishment. Oral administration of AZAS, APZE, or PZE resulted in varying degrees of DAI score reduction. APZE alleviated colon length shortening, while AZAS and PZE had little effect on colon length recovery. Among the three groups, APZE exhibited the most pronounced reduction in DAI and recovery of colon length. However, no discernible recovery in body weight was observed in any group.

Fig. 5.

Therapeutic effects of APZE on DSS-induced IBD mice. (A) Schematic illustration of the dosing regimen. (B) Body weight changes during the experiment. (C) Changes in disease activity index (DAI) scores. (D) Ex vivo photographs of the colon. (E) Quantification of ex vivo colon length. (F) Myeloperoxidase (MPO) levels in colon tissues. (G-J) Pro-inflammatory cytokine levels in colon tissues: (G) TNF-α, (H) IL-6, (I) IFN-γ, (J) IL-1β. (K) Hematoxylin and eosin (H&E), periodic acid-schiff (PAS), and alcian blue (AB) staining images of distal colon tissues. Scale bar = 100 μm. (L) Goblet cell numbers quantified by PAS staining. (M) Immunofluorescence staining images of colon tissues. From top to bottom: nuclei (blue), tight junction protein Occludin (red), ZO-1 (green), and merged images. Scale bar = 200 μm. (N-O) Quantification of average fluorescence intensity using ImageJ: (N) Occludin, (O) ZO-1. n = 10 for (B, C, E-J), n = 8 for (L, N, O). Data are shown as mean±SD, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

The immune response is dysregulated in DSS-induced mice with colitis, with neutrophils in the gut infiltrating and releasing MPO, as well as pro-inflammatory cytokines production increasing [45, 46]. Additionally, excessive proliferation of immune cells could lead to splenomegaly and leukocytosis. As shown in Fig. 5F, MPO activity in colon tissues were significantly elevated in the DSS group, indicating a high intestinal inflammation. However, treatment with APZE significantly reduced MPO activity in the colon. The inflammatory cytokine response is an important pathophysiological factor that influences the occurrence and progression of intestinal inflammation. As displayed in Fig. 5G-J, APZE treatment restored TNF-α, IL-6, IFN-γ, and IL-1β levels in colon tissues nearly to normal. Additionally, due to the immunosuppressive effects of AZA, mice treated with AZAS and APZE exhibited a significant reduction in spleen index (spleen-to-body weight ratio) and WBC count compared to the DSS group (Fig. S6A and B).

Severe colitis results in disruption of gut epithelial barrier, loss of goblet cells, and cell death. H&E, PAS, and AB staining were used to assess colon morphology, goblet cell numbers, and mucin levels, respectively. As shown in Fig. 5K and L, the colon tissue of DSS group exhibited structural damage, manifested as severe mucosal injury, decreased numbers of goblet cells and crypts, altered crypt morphology, and widespread inflammatory cell infiltration. In contrast, treatment with APZE promoted the regular, tight arrangement of crypt structures and maintained the abundance of goblet cells, significantly alleviating the intestinal morphological damage induced by DSS and restoring it to a near-normal state.

Tight junctions between intestinal epithelial cells are crucial for maintaining the integrity and function of the intestinal mucosal barrier [47]. The expression of tight junction-associated proteins ZO-1 and Occludin in colon was assessed through immunofluorescence staining. As displayed in Fig. 5M-O, the expression of ZO-1 and Occludin in the DSS group was significantly downregulated. In contrast, the expression of both proteins in the APZE group was nearly restored to normal levels, suggesting that APZE has excellent intestinal epithelial barrier recovery capacity. Furthermore, apoptosis in colon tissues was evaluated using TUNEL assay. As shown in Fig. S7A and B, the DSS group exhibited severe apoptosis of colon cells. However, treatment with APZE resulted in minimal apoptosis.

To evaluate the toxicity of AZAS and APZE, H&E staining was performed on sections of major organs. From Fig. S6D, histopathological examination revealed no apparent pathological lesions in the heart, spleen, lungs, or kidneys across all DSS-induced groups. However, the AZAS group exhibited a relatively loose liver tissue structure with signs of cell death, vacuolization, and inflammatory cell infiltration, suggesting signs of hepatotoxicity. In contrast, liver tissue structure in the APZE group appeared more compact with no significant pathological alterations. Meanwhile, TUNEL staining was performed to evaluate hepatocyte apoptosis in each group. The AZAS group exhibited the highest level of liver cell apoptosis, while the APZE group exhibited a substantial reduction (Fig. S7C and D). Overall, our findings demonstrated that APZE could effectively alleviate DSS-induced acute colitis with favorable safety.

Modulation of gut Microbiome by APZE

Previous studies have shown that modulation of the gut microbiota has beneficial effects on IBD treatment. We further explored the microbiota-modulating effect of APZE by examining the gut microbiota and its beneficial metabolites, SCFAs. The composition and abundance of fecal microbiota in mice before and after treatment were analyzed using 16 S rDNA sequencing. As shown in Fig. S8A-C, rarefaction curves plateaued with increasing sequencing depth, indicating sufficient sequencing depth and reliable data. α-Diversity indices (Shannon, Chao1, and Observed_OTUs) revealed that DSS-induced colitis reduced gut microbial diversity, whereas APZE treatment restored microbial diversity and alleviated dysbiosis (Fig. 6A-C). β-Diversity analysis using principal coordinate analysis (PCoA) based on unweighted UniFrac distances (Fig. 6D) showed distinct microbial community structures between DSS-induced mice with colitis and control mice. However, APZE treatment shifted the gut microbiota composition toward a more normal state. As shown in the Venn diagram (Fig. 6E), the APZE group shared the greatest number of common feature sequences with the normal group, indicating that its microbiota composition was most similar to that of healthy mice.

Fig. 6.

16 S rDNA sequencing analysis of gut microbiota. (A-C) α-Diversity analysis of the gut microbiota: (A) Shannon index, (B) Chao1 index, and (C) Observed_OTUs index. (D) β-Diversity analysis through principal coordinate analysis (PCoA) based on unweighted UniFrac distances. (E) Venn diagram of bacteria in different groups. (F-H) Species abundance histograms at (F) phylum level, (G) family level, and (H) genus level (n = 8)

Differences in gut microbiota composition across groups were further analyzed at the phylum, family, and genus levels. At the phylum level, the DSS group exhibited microbial imbalance, characterized by reduced abundances of Firmicutes and Actinobacteria and an increased abundance of Proteobacteria. Administration of APZE reversed this trend (Fig. 6F). Proteobacteria, a major pathogenic phylum and a marker of dysbiosis, increases susceptibility to colitis and impairs both innate and adaptive immunity in DSS-induced mice [48]. In contrast, Firmicutes help limit pathogenic bacteria expansion [49]. Bifidobacterium within Actinobacteria supports gut health by producing SCFAs and modulating immune responses [22]. At the family level, APZE treatment restored the abundance of Muribaculaceae and reduced Enterobacteriaceae (Fig. 6G). Muribaculaceae, known to metabolize mucin glycans and dietary fibers to produce SCFAs, has been shown to increase following PT administration [50, 51]. Enterobacteriaceae normally constitute less than 1% of the gut microbiota but proliferate abnormally in IBD patients [52]. At the genus level, the DSS group showed a marked increase in the harmful Escherichia-Shigella, known to aggravate intestinal inflammation [53], and a decrease in the beneficial Lachnospiraceae_NK4A136_group, which reduces inflammation and promotes gut health via SCFA production [54]. APZE reversed these alterations (Fig. 6H). Collectively, these results suggest that APZE enhances microbial diversity, increases beneficial bacteria, and reduces harmful bacteria, thereby alleviating IBD.

As shown in Fig. 7A-D, pairwise comparisons between groups identified differences in gut microbiota composition at the genus level. Compared to other groups, the abundance of Bifidobacterium was significantly higher in normal group and APZE group. These findings suggested that Bifidobacterium may play an important role in maintaining gut microbiota homeostasis and mediating therapeutic effects. The onset of colitis markedly reduces gut microbial diversity, leading to decreased SCFA production, which exacerbates the disease [55, 56]. As shown in Fig. 7E-G, acetate, propionate, and butyrate levels were significantly decreased in the DSS group, while APZE intervention substantially restored their levels. This may be because APZE could modulate the gut microbiota, thereby influencing SCFA production derived from microbial metabolism. Notably, SCFAs can further regulate gut microbiota composition, forming a bidirectional feedback loop [57].

Fig. 7.

Analysis of gut microbiota and SCFAs levels in cecal contents. (A-D) T-test analysis of gut microbiota between different groups: (A) Control and PBS groups, (B) PBS and APZE groups, (C) APZE and AZAS groups, (D) APZE and PZE groups (n = 8). (E-G) SCFAs levels in cecal contents: (E) acetate, (F) propionate, (G) butyrate (n = 10). Data are shown as mean±SD, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

In summary, APZE developed in our study possesses multiple functions including targeting colon, promoting intestinal drug absorption, and modulating gut microbiota. These effects may be attributed to the colon-targeting property of ES100, the adhesiveness of Zein and PT, and the microbiota-modulating capacity of PT.

In vivo therapeutic effects of APZE on pseudo-germ-free IBD mice

To investigate the impact of gut microbiota on the therapeutic effects of APZE, a pseudo-germ-free mouse model was established by administering an ABX for 7 days prior to DSS treatment [58]. The modeling and administration process is shown in Fig. 8A. Notably, after gut microbiota depletion, there were no significant differences between the AZAS and APZE groups in body weight, DAI scores, colon length, colon index, leukocyte number, spleen index, liver index, histopathological changes, or the expression of tight junction proteins (Fig. 8B-N). Compared with the DSS group, neither treatment showed significant improvements in primary therapeutic outcomes. These results indicated that the superior efficacy of APZE on DSS-induced colitis might be primarily gut microbiota dependent.

Fig. 8.

Therapeutic effects of APZE in pseudo-germ-free IBD mice. (A) Schematic illustration of the dosing regimen. Antibiotic cocktail (ABX): vancomycin, neomycin, metronidazole, and ampicillin. (B) Body weight changes during the experiment. (C) Changes in DAI score. (D) Ex vivo photographs of the colon. (E) Quantification of ex vivo colon length. (F) Ex vivo colon index calculated as the ratio of colon length to mass. (G) Leukocyte counts in whole blood. (H) Spleen index: spleen-to-body weight ratio. (I) Liver index: liver-to-body weight ratio. (J) H&E, PAS, and AB staining images of distal colon tissues. Scale bar = 100 μm. (K) Goblet cell numbers quantified by PAS staining. (L) Immunofluorescence staining images of colon tissues. From top to bottom: nuclei (blue), Occludin (red), ZO-1 (green), and merged images. Scale bar = 200 μm. (M-N) Quantification of average fluorescence intensity using ImageJ: (M) Occludin, (N) ZO-1. n = 10 for (B, C, E-I), n = 8 for (K, M, N). Data are shown as mean±SD, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Combined therapeutic effects of APZE and GB

Previous pharmacodynamic studies have demonstrated that APZE exhibits the best therapeutic effects in IBD mice and shifts the gut microbiota toward a normal profile, with a marked increase in Bifidobacterium abundance. To investigate the role of Bifidobacterium in this process, DSS-induced IBD mice were administered commercial Golden Bifid (GB) consisting of Bifidobacterium longum, Lactobacillus bulgaricus, and Streptococcus thermophilus on different schedules. The modeling and administration process is shown in Fig. 9A. Body weight recovery was most prominent in the predose-GB group (pGB group), followed by the GB therapeutic administration alone (GB group), and then GB combined with AZAS or APZE groups (Fig. 9B). These results indicated that GB exhibited a beneficial effect on weight recovery in mice with colitis. As shown in Fig. 9C-F, upon oral administration of GB alone or in combination with either AZA formulation, the DAI scores and colon length showed varying degrees of improvement.

Fig. 9.

Therapeutic effects of APZE combined with Golden Bifid (GB) in IBD mice. (A) Schematic illustration of the dosing regimen. (B) Body weight changes during the experiment. (C) Changes in DAI score. (D) Ex vivo photographs of the colon. (E) Quantification of ex vivo colon length. (F) Ex vivo colon index calculated as the ratio of colon length to mass. (G) Leukocyte counts in whole blood. (H) Spleen index: spleen-to-body weight ratio. (I) Liver index: liver-to-body weight ratio. (J) H&E, PAS, and AB staining images of distal colon tissues. Scale bar = 100 μm. (K) Goblet cell numbers quantified by PAS staining. (L) Immunofluorescence staining images of colon tissues. From top to bottom: nuclei (blue), Occludin (red), ZO-1 (green), and merged images. Scale bar = 200 μm. (M–N) Quantification of average fluorescence intensity using ImageJ: (M) Occludin, (N) ZO-1. (n = 8). Data are shown as mean±SD, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Additionally, the spleen index and WBC number in both combination therapy groups were substantially reduced, indicating an improvement of the inflammatory status (Fig. 9G and H). The morphology, goblet cell numbers, mucin levels and tight junction proteins expression of colon were shown in Fig. 9J-N, all treatments containing GB displayed certain therapeutic effects on the monitoring indicators. Among these groups, the preventive administration of GB showed better intestinal structure and more expression of tight junction proteins than the monotherapy with GB. The GB administered alone or with AZAS partially alleviated the progression of IBD, while the combined administration of APZE and GB promoted restoration of colon length, intestinal mucosa, crypts, goblet cell numbers, mucin, and tight junction proteins to near normal levels. In addition, we investigated the toxicity of GB alone or in combination with drugs in vivo. As shown in Fig. S9, no obvious lesions were observed in major organs for all groups, indicating the acceptable safety for all regimens.

These results suggested that the APZE combined with GB group exhibited the most significant reduction in DAI. All regimens demonstrated effective treatment for DSS-induced colitis, with preventive administration showing better outcomes than therapeutic administration. The AZAS in combination with GB reduced hepatotoxicity compared with AZAS monotherapy in the aforementioned study. These observations suggested that Bifidobacterium could play a favorable role in enhancing therapeutic efficacy and reducing drug-related hepatotoxicity.

Discussion

IBD is a refractory gastrointestinal disorder characterized by chronic and recurrent inflammation. Current therapies have certain limitations, such as the premature disintegration and drug release of AZA tablets in the upper gastrointestinal tract, which restricts their accumulation in the colon, thereby reducing therapeutic efficacy [10]. Additionally, systemic exposure to AZA may cause adverse effects such as hepatotoxicity [11]. In this study, we developed an AZA-loaded nanoparticle (APZE) constructed from PT, Zein, and ES100 to enhance the therapeutic effects of AZA. The results demonstrated that APZE significantly increased the oral bioavailability and colon accumulation of AZA, offering better efficacy in relieving colitis compared to AZAS, while exhibiting good safety. Notably, the therapeutic efficacy of APZE is linked to its ability to remodel the gut microbiota axis, a mechanism distinct from traditional AZA treatment.

APZE enhances the absorption in Caco-2 cells and cell monolayers, as well as exhibits excellent adhesion, penetration, and absorption properties in the intestine. Compared to AZAS, APZE achieved a 1.94-fold increase in oral bioavailability (as reflected by plasma 6-MP exposure) and a 23.38-fold higher AUC for colon fluorescence intensity. This can be attributed to the colon-targeting capability of ES100 and the mucosal adhesion properties of Zein and PT [12, 28, 31]. These characteristics enhance the retention of nanoparticles in the colon, allowing for sustained drug release at the site of inflammation, thereby exerting immunosuppressive effects. This enhanced colon targeting increases local drug exposure at the disease site while potentially reducing systemic side effects, such as hepatotoxicity.

While APZE significantly increased the bioavailability of AZA compared to AZAS, the formation of MMP, the hepatotoxic metabolite of AZA, did not increase proportionally. We hypothesize that this difference arises from the metabolic diversion effect brought about by the colonic targeting design of APZE. In the case of AZAS, AZA is rapidly absorbed in the small intestine and undergoes extensive first-pass metabolism in the liver, resulting in higher MMP formation. This high first-pass effect contributes to the systemic toxicity associated with AZA. In contrast, APZE delivers the drug to the colon. The drug absorbed from the colon is then transported via lymphatic circulation, bypassing the liver’s first-pass effect [59]. This mechanism allows more 6-MP to reach peripheral immune tissue targets in its active form, while reducing MMP formation in the liver. Therefore, although APZE enhances the bioavailability of AZA, the bypass of hepatic first-pass metabolism leads to less MMP formation, thus minimizing liver toxicity while preserving the therapeutic efficacy at the target sites.

In addition to enhancing the immunosuppressive effect through optimized drug delivery, APZE also modulates the gut microbiota and restores the intestinal barrier to achieve comprehensive therapeutic action. Dysbiosis of the intestinal microbiota is a key contributor to the pathogenesis of IBD, as it exacerbates intestinal inflammation and compromises barrier integrity [21, 23]. The PT component within APZE can modulate the composition and diversity of the gut microbiota. Some in vivo studies have shown that PT can increase the abundance of Clostridiales, Muribaculaceae, Bacteroides, Lactobacillus, Bifidobacterium, and Prevotella [51, 60]. Under APZE intervention, SCFA-producing bacteria, such as Muribaculaceae, Bifidobacterium, and Lachnospiraceae_NK4A136_group were enriched, while pro-inflammatory bacteria like Enterobacteriaceae and Escherichia–Shigella were suppressed, reversing DSS-induced microbial dysbiosis [22, 51, 52, 54]. The restored levels of acetate, propionate, and butyrate further strengthened barrier function and supported immune homeostasis [57]. The therapeutic efficacy of APZE was largely microbiota-dependent, as microbiota depletion abolished the therapeutic advantage of APZE over AZAS.

While APZE alone effectively restored gut barrier integrity and reduced inflammation, the probiotic GB was included with the hypothesis that it could further modulate the microbiota and enhance immune function. Bifidobacterium in GB is known to produce SCFAs, which support intestinal health and modulate the immune system. Additionally, GB could potentially promote intestinal barrier repair by stimulating mucus production and tight junction integrity, thus providing long-term benefits in immune regulation and gut protection [22]. Combined treatment with APZE and GB improved colitis-related outcomes, with restoration of colonic tissue architecture. However, the combination of APZE and GB did not show a significant additive effect on colon length recovery or tight junction protein expression. The lack of synergy in this acute colitis model suggests that the therapeutic effects of APZE alone, through immunosuppression and microbiota modulation, may be sufficient in this setting. The effects of GB may be more prominent in chronic models, where its role in long-term immune modulation and intestinal barrier restoration can be fully realized. Future studies will investigate the dose-response and long-term effects of GB in combination with APZE to explore its full therapeutic potential.

In addition, several limitations of this study should be acknowledged. APZE achieved colonic accumulation and retention via the pH responsiveness of ES100 and the mucoadhesive properties of pectin and Zein. An important question is whether such retention could interfere with physiological gastrointestinal functions, including gut motility, peristalsis and gut secretion. Available evidence suggests that the key materials in APZE are unlikely to cause marked disruption. PT may delay gastric emptying via increased viscosity but does not substantially alter overall transit [61, 62]. Zein and ES100 are widely used as pharmacologically inert oral/colon-targeting carriers with no clear direct effects on gut motility or secretion [31, 63–65]. AZA primarily acts through immunosuppression with minimal direct activity on smooth muscle or the enteric nervous system [11]. Consistently, no apparent new-onset severe dysmotility signs were observed during dosing. Nevertheless, dedicated functional assays will be important in future preclinical safety evaluations to further exclude the effects on motility and secretion.

In our study, only male mice and rats were used for the assessment of APZE’s distribution, pharmacokinetics, therapeutic efficacy, and safety. While male animals were chosen to reduce variability and because of their more consistent response in DSS-induced colitis models, we acknowledge that sex-based differences in microbiota, immune response, and AZA metabolism could potentially influence the outcomes. We plan to extend this study by including both male and female animals in future experiments to assess the role of sex in influencing the treatment outcomes. Additionally, we employed a DSS-induced acute colitis mouse model, which does not fully model the chronic and relapsing nature of human IBD. Future research should evaluate the long-term effects of APZE in chronic IBD models to assess its capacity for inducing and maintaining remission. Furthermore, the precise interactions between microbiota-derived metabolites influenced by APZE and the host immune system require further investigation through integrated metabolomic and immunologic analyses.

In summary, we developed a multifunctional nanoparticle that integrates efficient colon-targeted delivery of AZA with microbiota modulation, thereby addressing the core pathophysiological features of IBD—immune dysregulation, barrier dysfunction, and microbial imbalance—while enhancing both therapeutic efficacy and safety.

Conclusions

In conclusion, we successfully developed a microbiota-modulating and colon-targeted nanoparticle for efficient loading of AZA and alleviation of colitis. APZE enhanced cellular and intestinal uptake, increasing drug bioavailability compared to AZAS. While the level of methylated metabolite MMP, which is considered one of the main contributors to hepatotoxicity, did not increase significantly. Compared with AZA, APZE exhibited favorable therapeutic effects and safety in the treatment of IBD, which may be attributed to its multiple advantages, such as colon targeting ability, mucosal adhesion property, and gut microbiota modulatory capacity. APZE also demonstrated favorable efficacy and safety when used in combination with the commercially available probiotic Bifidobacterium. Overall, this study presents a promising new formulation for IBD treatment, along with a potential microbiota-based combinatorial strategy.

Abbreviations

IBD

Inflammatory bowel disease

AZA

Azathioprine

CD

Crohn’s disease

UC

Ulcerative colitis

CTDDS

Colon-targeted drug delivery systems

ES100

Eudragit^®S100

PT

Pectin

SCFA

Short-chain fatty acid

DSS

Dextran sulfate sodium

DiR

1,1’-dioctadecyl-3,3,3’,3’-tetramethylindotricarbocyanine iodide

DAPI

4’,6-diamidino-2-phenylindole

CMC-Na

Sodium carboxymethyl cellulose

F68

Pluronic^®188

M-β-CD

Methyl-β-cyclodextrin

Gen

Genistein

AMI-HCl

Amiloride hydrochloride

BFA

Brefeldin A

MON

Monensin

NOC

Nocodazole

Baf A1

Bafilomycin A1

CPZ

Chlorpromazine

GB

Golden Bifid

PDI

Polydispersity index

MWCO

Molecular weight cut-off

CLSM

Confocal laser scanning microscopy

AP

Apical

BL

Basolateral

Papp

Apparent permeability coefficient

ER

Efflux ratio

6-MP

6-mercaptopurine

MMP

6-methylmercaptopurine

F

Flux

P%

Absorption percentage

Ka

Absorption rate constant

Peff

Effective permeability coefficient

DAI

Disease activity index

WBC

White blood cell

MPO

Myeloperoxidase

H&E

Hematoxylin and eosin

PAS

Periodic acid-schiff

AB

Alcian blue

TUNEL

The terminal deoxynucleotidyl transferase dUTP nick end labeling ()

OTUs

Operational Taxon omic Units

ABX

Antibiotic cocktail

DLS

Dynamic light scattering

PDI

Polydispersity index

TEM

Transmission electron microscopy

AUC

Area under the curve

PCoA

Principal coordinate analysis

Author contributions

J.Z.: Writing - original draft, Methodology, Investigation, Data curation, Conceptualization. Y.Z.: Methodology, Investigation, Formal analysis, Software. X.L.: Methodology, Formal analysis, Visualization. J.H.: Methodology, Software, Validation. L.W.: Investigation, Data curation. Y.H.: Methodology, Investigation. S.J.: Methodology, Formal analysis. Y.L.: Investigation. X.M.: Validation. H.B.: Methodology. D.M.: Funding acquisition, Methodology, Writing - review & editing, Conceptualization. L.Z.: Funding acquisition, Project administration, Supervision, Writing - review & editing, Conceptualization. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (82173894), the Beijing Natural Science Foundation of China (7232246 and L202042), and the National Key R&D Program of China (2023YFC2706100).

Data availability

Data is provided within the manuscript or supplementary information files.

Declarations

Ethics approval and consent to participate

Animal experiments were approved by the Institutional Animal Care and Use Committee of Peking University (Accreditation number: BCJE0155, BCJH0222).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.