Magnolia biondii-Derived Exosome-Like Nanoparticles Ameliorate Ulcerative Colitis by Suppressing Inflammation, Reducing Oxidative Stress, and Modulating Gut Microbiota

Abstract

Background

Ulcerative colitis (UC) is a chronic, relapsing inflammatory bowel disease with limited treatment options and suboptimal remission rates. Plant-derived exosome-like nanoparticles have emerged as promising therapeutic agents due to their natural origin, biocompatibility, and capacity to deliver bioactive cargo. This study aimed to isolate exosome-like nanoparticles from the decoction of Magnolia biondii flower buds (Xinyi) (MFELNs) and evaluate their therapeutic potential in a murine model of UC.

Methods

MFELNs were isolated from Xinyi decoction using differential centrifugation and ultracentrifugation. MFELNs were characterized by transmission electron microscopy, nanoparticle size and potential analysis, and metabolomic profiling. The in vitro antioxidant activity was assessed by DPPH scavenging and intracellular ROS suppression assays. Anti-inflammatory effects were evaluated in LPS-stimulated RAW 264.7 macrophages by measuring cytokine expression and macrophage polarization via flow cytometry. The promotion of cell migration was assessed in NIH3T3 cells by scratch assay. UC was induced in C57BL/6 mice using dextran sodium sulfate (DSS), and MFELNs were orally administered daily for 14 days. Disease activity index, colon length, histopathology, colonic cytokine expression, and gut microbiota composition were evaluated. Bioinformatics analysis, molecular docking, and Western blot were performed to elucidate the underlying molecular mechanisms.

Results

MFELNs exhibited typical exosome-like morphology with a size of 66.7 ± 1.3 nm and a negative surface charge. Metabolomic profiling revealed enrichment in flavonoids, alkaloids, lignans, and phenolic acids. MFELNs demonstrated satisfactory biocompatibility and gastrointestinal stability. In vitro, MFELNs scavenged DPPH radicals, suppressed H₂O₂-induced ROS generation, downregulated LPS-induced pro-inflammatory cytokines (TNF-α, IL-1β, IL-6), restored IL-10 expression, and promoted macrophage polarization from M1 to M2 phenotype. In vivo, MFELNs attenuated DSS-induced body‑weight loss, reduced DAI scores, prevented colon shortening, and alleviated histopathological damage. MFELNs downregulated colonic pro-inflammatory cytokines and restored IL-10 expression. Gut microbiota analysis revealed that MFELNs restored the DSS-elevated Firmicutes/Bacteroidota ratio without significantly affecting α-diversity. Bioinformatics screening identified four potential bioactive components (sesamin, litsoeine/laurotetanine, demethylwedelolactone, and armepavine) that target SRC. Molecular docking demonstrated favorable binding between these components and SRC, while Western blot confirmed that MFELNs suppressed the SRC/NF-κB signaling pathway.

Conclusion

MFELNs represent a novel class of decoction-derived exosome-like nanoparticles that ameliorate UC through integrated anti-inflammatory, antioxidant, and microbiota-modulating mechanisms. Their therapeutic effects are mediated, at least in part, through suppression of the SRC/NF-κB signaling pathway. These findings position MFELNs as a promising natural nanotherapeutic strategy for UC and support the concept that self-assembled nanoparticles formed during herbal decoction constitute an important material basis for the efficacy of traditional Chinese herbals.

Graphical Abstract

Introduction

Ulcerative colitis (UC), a major form of inflammatory bowel disease (IBD), is a chronic, relapsing-remitting inflammatory disorder of the colonic mucosa characterized by mucosal inflammation confined to the colon, typically beginning in the rectum and extending proximally in a continuous manner. The incidence of UC has been steadily increasing over the past decades, making it a growing global health challenge.¹ Although the etiology of UC is not fully understood, it is pathologically considered to result from a complex interplay of intestinal epithelial barrier dysfunction, dysregulated immune responses, and altered gut microbiota.^2–4 In recent years, the management strategy for UC has shifted from merely controlling symptoms to a treat-to-target approach aimed at achieving clinical remission and endoscopic mucosal healing. Therapeutic agents for UC include conventional medications (eg, 5-aminosalicylates, corticosteroids), biologics (eg, anti-TNF monoclonal antibody vedolizumab), and small-molecule drugs (eg, JAK inhibitor tofacitinib).⁵ However, current treatments still face significant challenges. For instance, clinical remission rates during the induction phase generally do not exceed 20–30%, and only about 30–60% of patients maintain remission after one year of maintenance therapy.⁶ Therefore, there is an urgent need to explore novel therapeutic strategies with distinct mechanisms of action, improved efficacy, and enhanced safety profiles in current UC research. Recently, nanosystem strategies such as nanozyme^7–9 or engineered nanovesicle¹⁰ have been reported to remodel colonic immune homeostasis, modulate gut microbiota, repair the intestinal barrier, and alleviate inflammation‑associated pain in UC.

Plant-derived exosomes (PDEs) have emerged as a novel class of naturally occurring nanomedicines with remarkable therapeutic potential for UC.^11,12 PDEs are lipid bilayer-enclosed vesicles (30–500 nm) isolated from fresh plant tissues, such as fruits, vegetables, and herbs, and they carry a diverse cargo of bioactive molecules including proteins, lipids, miRNAs, and secondary metabolites.^13,14 PDEs present distinct advantages for therapeutic applications, including inherently low immunogenicity, high biocompatibility, exceptional stability in the harsh gastrointestinal environment suitable for oral administration, and the unique capacity for cross-kingdom gene regulation.¹⁵ These vesicles were also named as plant-derived exosome-like nanovesicles due to the lack of clear biogenesis pathways and specific markers.^16,17 Building upon the concept of PDEs, an emerging class of exosome-like nanoparticles (ELNs) derived from Chinese herbal medicines has attracted considerable attention due to their unique self-assembly properties and intrinsic pharmacological activities.^18–20 In traditional Chinese medical practice, herbs are typically boiled in water for 30 to 60 minutes to produce decoctions, which serve as the main therapeutic route. Studies have revealed that such decoctions contain hundreds of lipids and millions of small RNAs,²¹ suggesting the presence of exosome-like nanoparticles, also referred to as decoctosomes in some studies.²²

Magnolia biondii flower bud, known as Xinyi in TCM, consist kinds of chemical constituents, such as essential oils and lignans.^23–25 It is traditionally used for treating allergic rhinitis and sinusitis due to its purported anti-inflammatory and immunomodulatory properties.^26–28 It is also demonstrated that lignans isolated from Magnolia fargesii flower buds suppress cigarette smoke-induced airway inflammation in vitro and in vivo by inhibiting the EGFR/ERK/Akt signaling pathway, highlighting their potential as therapeutic agents for chronic obstructive pulmonary disease.²⁹ Given its traditional use in inflammatory conditions and the shared mucosal immune dysregulation features between upper airway and intestinal diseases, it is hypothesized that Magnolia biondii may possess therapeutic efficacy against UC. The anti-inflammatory effect of fargesin, a bioactive lignan from Flos Magnoliae, has been demonstrated to ameliorate chemically induced IBD via NF-κB signaling suppression.³⁰ However, whether its bioactive components are delivered via exosome-like nanoparticles remains entirely unexplored. Therefore, this study aims to isolate exosome-like nanoparticles from Xinyi (MFELNs) and systematically evaluate their protective effects and underlying mechanisms in a mouse model of dextran sodium sulfate (DSS)-induced ulcerative colitis, focusing on their anti-inflammatory, antioxidant, and gut microbiota-modulating activities.

Materials and Methods

Extraction and Characterization of MFELNs

Dried flower buds of Magnolia biondii (Xinyi) were provided by Henglikang Co., Ltd. (Henan, China). For extraction, 25 g of bud kernel was added to 1000 mL of ultrapure water and decocted for 1 hour at 100 °C. The decoction was sequentially filtered through a stainless-steel sieve and then subjected to differential centrifugation (8000 g for 30 min, and 16,000 g for 30 min). The supernatant was subsequently filtered through 0.22 μm membrane. The filtrate was ultracentrifuged at 100,000 × g for 90 min at 4°C (Beckman Type 70 Ti). The final pellet, without washing, was resuspended in phosphate-buffered saline (PBS) to obtain the MFELNs stock solution. The prepared MFELNs were either used immediately or stored at −80°C for future use.

The morphology of MFELNs was characterized by transmission electron microscopy (TEM). The ultraviolet-visible (UV-Vis) absorption spectrum was recorded by a spectrophotometer. The particle size distribution, concentration, and zeta potential were determined using a NanoCoulter counter (Resun Technology, Co., Ltd., Shenzhen) with the resistive pulse sensing (RPS) principle. The protein content was determined by a BCA protein assay kit. Metabolomic profiling of MFELNs was performed by high-performance liquid chromatography-mass spectrometry (HPLC-MS, Metware Biotechnology Co., Ltd). The identified metabolites were categorized based on their chemical classes, and the results were visualized as primary and secondary metabolite classification diagrams.

DPPH Scavenging Assay

The DPPH (1,1-diphenyl-2-trinitrobenzhydrazide) radical scavenging assay was performed to evaluate the antioxidant activity of MFELNs. Briefly, 100 μL DPPH solutions was added to 96-well plate, followed by the addition of 100 μL of MFELNs solution to final concentrations of 200, 500, and 1000 μg/mL. Glutathione (GSH, 1250 μg/mL) was used as a positive control. After gentle mixing, the plate was incubated in the dark at room temperature for 30 min. The absorbance was measured at 517 nm using a microplate reader. The DPPH scavenging rate (%) was calculated according to the following formula:

Gastrointestinal Stability

The gastrointestinal stability of MFELNs was assessed using simulated gastric fluid (SGF, pH=1.2) and simulated intestinal fluid (SIF, pH=6.8), respectively. MFELNs were added to SGF or SIF, and then incubated at 37°C for 30 min, 1 h, 2 h, and 3 h. The particle size was measured to assess their gastrointestinal stability.

Hemolysis Test and CCK-8 Assay

The hemocompatibility of MFELNs was assessed using hemolysis test. Mouse blood was collected and centrifuged at 3000 rpm for 10 min to isolate erythrocytes. The red blood cells were washed repeatedly with PBS until the supernatant became colourless. The samples were divided into three groups: negative control (PBS), positive control (Triton X-100), and MFELNs at 100, 200, and 400 μg/mL. Each group received erythrocyte suspension (V:V = 50:1), followed by incubation at 37 °C for 3 h. After centrifugation at 3000 rpm for 10 min, the supernatant was transferred to a 96-well plate, and the absorbance at 542 nm was recorded. The hemolysis ratio was calculated to assess the hemocompatibility.

The cytotoxicity of MFELNs was assessed using NIH/3T3 cells via CCK-8 assay. Cells were seeded into 96-well plates at a density of 5000 cells/well and allowed to adhere overnight. Subsequently, the culture medium was replaced with fresh complete medium containing 200 μg/mL MFELNs. Following 24 h or 48 h treatment, CCK-8 reagent was added and the absorbance was measured at 450 nm according to the manufacturer’s instructions. Cell viability was calculated to evaluate the potential cytotoxic effects of MFELNs.

Cell Scratch Assay

The cell migration ability was assessed using a scratch assay with NIH/3T3 cells. Briefly, cells were seeded in 6-well plates at a density of 4×10⁵ cells/well and cultured until reaching approximately 80% confluence. A straight scratch wound was created in the cell monolayer. The detached cells were gently removed by washing with PBS. Subsequently, fresh medium containing 2% fetal bovine serum (FBS) and 200 μg/mL MFELNs was added. The initial wound area was photographed under an inverted microscope immediately after scratching (defined as 0 h, W0). Images of the same wound fields were captured after 24 h and 36 h of incubation (defined as Wt). Non-MFELN was served as a control. The migration rate, expressed as the wound‑healing percentage, was calculated using the following formula:

Inhibition of ROS Generation

RAW 264.7 cells (purchased from Sinoresearch (Beijing) Biotechnology Co., Ltd.) were seeded into 24-well plates and cultured overnight to allow adherence. Hydrogen peroxide (H₂O₂) was added into wells to induce intracellular ROS production as the positive control group. 200 μg/mL MFENLs and H₂O₂ were both added to the wells to evaluate the protective effect of MFENLs on oxidative stress. Glutathione was used as a positive control. Following 24 hours of incubation, the DCFH-DA probe was added according to the manufacturer’s instructions. Images were then captured under a fluorescence microscope, and the fluorescence intensity was measured to evaluate the change in intracellular ROS levels.

Anti-Inflammation Assay

Lipopolysaccharide (LPS)-induced macrophage inflammation model was established to evaluate the anti-inflammatory effects of MFELNs in vitro. Macrophages were stimulated with 1 μg/mL LPS for 12 h to induce polarization toward the M1 phenotype. After removal of the LPS-containing medium, the cells were washed and treated with complete medium containing 200 μg/mL MFELNs for 24 h. Subsequently, total RNA was extracted from the harvested cells. Gene-specific primers (listed in Table 1) were used to quantify the mRNA expression levels of TNF-α, IL-1β, IL-6, and IL-10 by quantitative reverse transcription polymerase chain reaction (qRT-PCR) with GAPDH served as a control.

Table 1.

Primers for qRT-PCR

Gene	Primer Sequence (5′ −3′)	Product Size (bp)
β-actin	F: CCACTGTCGAGTCGCGT	17
	R: AAGTCAGTGTACAGGCCAGC	20
TNF-α	F: TAGCCCACGTCGTAGCAAAC	20
	R: GCAGCCTTGTCCCTTGAAGA	20
IL-1β	F: TGCCACCTTTTGACAGTGATG	21
	R: ATGTGCTGCTGCGAGATTTG	20
IL-6	F: CCCCAATTTCCAATGCTCTCC	21
	R: GGATGGTCTTGGTCCTTAGCC	21
IL-10	F: CCTGGGTGAGAAGCTGAAGAC	21
	R: CTTGTAGACACCTTGGTCTTGG	22
GAPDH	F: GACAACTCACTCAAGATTGTCAG	23
	R: ATGGCATGGACTGTGGTCATGAG	23

M1/M2 Phenotype Reversal of Macrophages

RAW264.7 cells were seeded in 6-well plates and divided into the following three groups: negative control group (NC), LPS-induced group (LPS), and LPS-induced plus MFELNs treatment group (MFELNs). After overnight cell adherence, the LPS and MFELNs groups were induced by adding LPS. After 6 h incubation, the medium in the MFELNs group was replaced with complete medium containing 200 μg/mL MFELNs, while the NC and LPS groups received fresh complete medium only. Following 24 hours of culture, the cells were collected and stained with FITC-conjugated CD86 antibody (FITC-CD86) for 40 minutes. Afterward, PE-conjugated CD206 antibody (PE-CD206) was added and incubated for another 40 minutes. The cells were then centrifuged to remove unbound antibodies. The distribution of CD86⁺ M1 and CD206⁺ M2 subtypes was subsequently analyzed by flow cytometry.

Mouse UC Model and Treatment

The ulcerative colitis model was established in eight-week-old male C57BL/6 mice (body weight 23–25 g). This study was approved by the Medical Ethics Committee of the College of Life Sciences and Health, Wuhan University of Science and Technology (Approval No. LL20250018). All animal procedures were performed in accordance with the institutional guidelines for the care and use of laboratory animals. All mice were housed under specific pathogen-free (SPF) conditions with a 12‑h light/dark cycle and allowed free access to food and water. After one week of acclimatization, the mice were randomly assigned to three groups (n = 6 per group): (1) normal control (NC) group, receiving regular drinking water; (2) DSS model group, receiving 2.5% (w/v) DSS (MW 50 kD) in drinking water for 5 days to induce colitis, followed by normal drinking water thereafter; and (3) MFELNs treatment group, receiving 2.5% DSS for 5 days and administered MFELNs (12.5 mg/kg body weight) daily by oral gavage for 14 days, starting from the first day of DSS exposure. Mice in the NC and DSS groups received an equal volume of PBS by oral gavage daily. Body weight and disease activity index (DAI) were recorded daily throughout the experimental period.

Histopathological, Molecular, and Microbiota Analysis

On day 14, all mice were euthanized by cervical dislocation. Subsequently, the entire colon was carefully collected for photographing and length measurement. Part of the colon tissue was fixed in 4% paraformaldehyde, embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E) for histological evaluation. The other part of colon tissue was snap-frozen in liquid nitrogen and stored at −80°C for subsequent molecular analysis. Total RNA was extracted from the tissue, reverse-transcribed into cDNA, and subjected to qRT-PCR to determine the mRNA expression levels of cytokines TNF-α, IL-1β, IL-6, and IL-10 with β-actin served as control. Fresh fecal samples were collected at the endpoint of the experiment and immediately frozen at −80°C. Bacterial genomic DNA was extracted from the feces, and bacterial strains were investigated using 16S rRNA gene sequencing for gut microbiota analysis (Metware Biotechnology Co., Ltd). The data was processed using operational taxonomic unit (OUT) clustering and amplicon sequence variant (ASV) noise reduction; then Chao1 and Simpson indexes were calculated to conduct alpha diversity and beta diversity analysis; and subsequently principal co-ordinate analysis based on unweighted UniFrac distances and analysis of non‑metric multi-dimensional scaling based on Bray-Curtis distances were carried out for comparative analysis of multiple samples. Finally, the community composition of each sample was statistically analyzed at the phylum level and Firmicutes/Bacteroidota (F/B) ratio was calculated.

Bioinformatics Analysis and Molecular Docking

The potential bioactive components of MFELNs were initially screened using the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP, https://www.tcmsp‑e.com/) based on the criteria of oral bioavailability (OB) ≥ 30% and drug‑likeness (DL) ≥ 0.18 (results presented in Table  S1). To further narrow down components with likely biological activity, the above candidates were subjected to a secondary round of screening via the SwissADME online platform (https://www.swissadme.ch/) based on gastrointestinal absorption (HIA) and drug‑likeness (DL) criteria (results shown in Table 2). These screened components were selected for subsequent analysis. Potential target genes were predicted using the Swiss Target Prediction platform (https://www.swisstargetprediction.ch/). Simultaneously, UC-related disease targets were retrieved from the GeneCards (https://www.genecards.org/), OMIM (https://www.omim.org/), and DisGeNET (https://www.disgenet.org/) databases. The intersection between the drug‑predicted targets and the disease targets yielded a set of potential core targets for these components in UC treatment. These overlapping targets were then subjected to Gene Ontology (GO) functional enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis using the Metascape platform (https://metascape.org). A protein‑protein interaction (PPI) network was constructed from the same target set to identify and visualize key hub genes. Finally, molecular docking simulations were performed between these components and the identified core target proteins to predict their binding interactions.

Table 2.

Four Refined Bioactive Components of MFELNs

Molecule Name	MW	OB (%)	DL
Sesamin	354.38	56.55	0.83
Litsoeine/laurotetanine	327.41	55.41	0.51
Demethylwedelolactone	302.25	72.13	0.43
Armepavine	313.43	69.31	0.29

Abbreviations: MW, molecule weight; OB, oral bioavailability; DL, drug‑likeness.

Western Blot Assay

RAW264.7 cells were seeded in 6-well plates and divided into three groups: the control group (NC), LPS-induced group (LPS), and LPS-induced plus MFELNs treatment group (MFELNs). The LPS group was induced by adding LPS into wells. The MFELNs group was constructed by adding LPS and 200 μg/mL MFELNs. After 24 hours co-culture, the cells were collected for protein extraction and SDS-PAGE separation. The separated proteins were incubated with primary antibodies, including p-SRC, SRC, p-p65, p65, p-IκB, and IκB (Cell Signaling Technology) for overnight. β-actin was used as a control. The HRP-labelled secondary antibody was incubated for 1 hour, and then detected through chemiluminescence method. The bands were imaged with a quantitative imaging system (Touch Imager S, eBlot) and the band intensity was analyzed with ImageJ software.

Statistical Analysis

All experiments were performed independently at least three times. Data are presented as mean ± standard deviation (SD). ANOVA with post-hoc test was used for multiple group comparisons and a p‑value < 0.05 was considered statistically significant.

Results

Characterization of MFELNs

MFELNs were isolated from dried kernels of Xinyi using a protocol involving aqueous decoction, filtration, sequentially differential centrifugation, and ultracentrifugation (Figure 1A). TEM images revealed that the isolated MFELNs exhibited a circular and membrane‑encapsulated morphology, which is similar to exosome (Figure 1B). Nanoparticle tracking analysis showed that the exosome-like nanoparticles indicated a size distribution ranging from 50 to 120 nm, with an average hydrodynamic diameter of 66.7 ± 1.3 nm (Figure 1C). The particles displayed a negative surface charge, with a zeta potential of −9.95 ± 0.72 mV (Figure 1D). The negative charge contributes to their colloidal stability in aqueous suspension. The UV-Vis spectrum of MFELNs showed no significant absorption across the UV-Vis region (200–750 nm) (Figure S1). This indicates that MFELNs do not interfere with absorbance measurements at 562 nm during the BCA protein assay. The protein content was about 135 µg/g of dry sample as determined by BCA assay. The yield of MFELNs was approximately 9.6×10¹⁰ particles/g of dry kernels. Compared to mammalian cell-derived extracellular vesicles, MFELNs exhibit a more straightforward isolation protocol and a significantly high production yield. This presents significant advantages to plant-derived nanoparticles, which are superior to cell-derived vesicles.

Figure 1.

Characterization of MFELNs. (A) Protocol for the extraction and purification process of MFELNs; (B) Representative TEM image of MFELNs (scale bar: 100 nm); (C) Size distribution profile of MFELNs measured by NanoCoulter; (D) Zeta potential distribution profile of MFELNs; (E) Metabolomics analysis showing the proportion of metabolites of different classes in MFELNs; (F) Analysis of secondary metabolites in MFELNs.

Metabolomics analysis revealed that MFELNs contain a rich and diverse array of metabolites, highlighting the component complexity of these nanoparticles. Major metabolite categories (Figure 1E) included flavonoids (13.22%), alkaloids (12.99%), lignans and coumarins (11.64%), phenolic acids (10.99%), terpenoids (10.68%), amino acids and derivatives (9.29%). Notably, lipids (8.17%) also accounted for significant proportions. Other important metabolite categories included organic acids (3.54%), nucleotides and derivatives (3.5%), quinones (1.06%), tannins (0.19%), steroids (0.16%), and others (14.57%). This metabolic profile highlights the potential of MFELNs in modulating cellular processes. MFELNs are notably abundant in secondary metabolites, which largely underlie their diverse bioactivities. Thus, profiling these metabolites is essential for understanding their functional potential. Analysis identified a spectrum of key secondary metabolites in MFELNs (Figure 1F), ordered by relative abundance: flavonoids (21.69%), alkaloids (21.32%), lignans and coumarins (19.10%), phenolic acids (18.05%), terpenoids (17.53%), quinones (1.74%), tannins (0.31%), and steroids (0.26%). Among these, several classes, such as alkaloids, lignans and coumarins, and phenolic acids, are known for their anti‑inflammatory and antioxidant properties. These components may act synergistically through multiple pathways, suggesting their potential in preventing or alleviating ulcerative colitis.

Biocompatibility, Gastrointestinal Stability, and Cell Migration Promotion of MFELNs

The blood compatibility was assessed using hemolysis test. The hemolysis ratio grows with the increase of MFELNs concentration (Figure 2A and B). At a concentration of 200μg/mL, the hemolysis ratio of MFELNs was 4.8%, which fell just below the 5% safety limit. Hence, this concentration was selected for subsequent cellular experiments. Cell viability was assessed using the CCK-8 assay following co-culture with MFELNs. As shown in Figure 2C, NIH/3T3 cells maintained good proliferative activity in the concentration range of 100–400 μg/mL MFELNs, even though a slight increase in proliferation was observed at 48 hours with 100 μg/mL MFELNs. Gastrointestinal stability was assessed by monitoring nanoparticle particle size over time. The results (Figure S2) showed that in simulated gastric fluid, particle size gradually increased with time, with a 20% increase observed at 4 hours. However, in simulated intestinal fluid, the particle size first increased and then decreased, with the maximum particle size (an increase of approximately 15%) observed at 1 hour. The nanoparticles were more stable in simulated intestinal fluid than in simulated gastric fluid. These results demonstrate that MFELNs possess satisfactory biocompatibility and gastrointestinal stability, supporting their potential for further evaluation.

Figure 2.

MFELNs possess biocompatibility and promote cell migration. (A) Hemolysis test photographs; (B) Hemolysis ratio for different groups; (C) Proliferation of NIH3T3 cells measured by CCK‑8 assay after 24 h or 48 h treatment with MFELNs; (D) Representative scratch‑wound images showing NIH3T3 cell migration after 24 h and 36 h treatment with MFELNs (red lines indicate the boundary of the wound, scale bars: 500 μm); (E) quantification of cell migration rate from A. (ns means no significant difference, *p < 0.05, **p < 0.01, ****p < 0.0001).

The effect of MFELNs on cell migration was evaluated using a cell scratch assay (Figures 2D). After 24 and 36 hours of incubation with 200 µg/mL MFELNs, NIH/3T3 cells exhibited migration rates of 26.8 ± 2.2% and 38.50 ± 0.37%, respectively, which were significantly higher than those in the untreated control group (Figure 2E). These results indicate that MFELNs promote the migration of fibroblast cells.

Antioxidant and Anti-Inflammatory Effects of MFELNs

The in vitro antioxidant capacity of MFELNs was assessed using the DPPH radical scavenging assay. As shown in Figure 3A, the DPPH radical scavenging ratio increased when MFELNs concentration rose from 200 to 1000 µg/mL. At 200µg/mL, the scavenging rate was about 44.1%, indicating a considerable antioxidant activity. At 1000 µg/mL, the scavenging rate was about 60.1%, a little lower than GSH (70%). This supports the potential of MFELNs in preventing or treating oxidative stress-related diseases. To evaluate the antioxidant effect of MFELNs at the cellular level, intracellular ROS generation was induced by H₂O₂. As illustrated in Figure 3B and C, MFELNs treatment reduced H₂O₂-induced ROS production, even though its inhibition effect is a little weaker than GSH. These results demonstrate that MFELNs can effectively suppress intracellular ROS generation, thereby mitigating oxidative stress-induced cellular damage.

Figure 3.

Antioxidant and anti-inflammatory effects of MFELNs. (A) DPPH radical scavenging of MFELNs at different concentrations and glutathione; (B) Representative fluorescence images showing intracellular reactive oxygen species (ROS) in RAW 264.7 cells following treatment with MFELNs and glutathione (scale bars: 200 μm); (C) Quantitative analysis of relative ROS fluorescence intensity; (D) mRNA expression levels of TNF-α, IL-1β, IL-6 and IL-10 in LPS-induced RAW 264.7 cells treated with MFELNs; (E) Distribution of M1/M2 phenotype of RAW 264.7 cells detected by flow cytometry; (F) Ratio of M1/M2 phenotype. (ns means no significant difference, *p < 0.05, ***p < 0.001, ****p < 0.0001).

The anti-inflammatory effect of MFELNs was evaluated by an LPS-induced macrophage cell inflammation model. The qRT-PCR analysis revealed that MFELNs significantly suppressed the LPS-induced upregulation of pro-inflammatory cytokines TNF-α, IL-1β and IL-6 (Figure 3D). Conversely, MFELNs restored the expression level of IL-10, an anti-inflammatory cytokine downregulated by LPS stimulation (Figure 3D). Furthermore, flow cytometry analysis demonstrated that MFELNs treatment raise the ratio of M2‑phenotype macrophage and decrease the ratio of M1‑phenotype macrophage, illustrating a shift promotion in macrophage polarization from the pro-inflammatory M1 phenotype toward the anti-inflammatory M2 phenotype (Figure 3E and F). Collectively, these findings indicate that MFELNs possess considerable anti-inflammatory property.

MFELNs Ameliorate Mouse Ulcerative Colitis

An UC model was established in mice by administering 2.5% DSS in drinking water for 5 days. The mice were treated with MFELNs via oral gavage daily for a total experimental period of 14 days (Figure 4A). Following DSS induction, mice developed typical symptoms of colitis, including diarrhea and bloody stools. As shown in Figure 4B, body weight in the normal control (NC) group increased steadily throughout the study. In contrast, mice in the DSS group began to lose weight from day 6 onward, and this decline persisted until the end of the experiment. Treatment with MFELNs alleviated this weight loss, with body weight in the MFELNs group beginning to recover from day 12 and showing a significant improvement compared to the DSS group by day 14. Concurrently, the disease activity index (DAI), which integrates body weight change, stool consistency, and fecal bleeding, was evaluated. The DAI score in the MFELNs group was markedly lower than that in the DSS group starting from day 6. Although DAI scores increased in both groups after day 10, the MFELNs group consistently maintained significantly lower scores than the DSS group throughout the later phase of the experiment (Figure 4C).

Figure 4.

MFELNs attenuate DSS-induced ulcerative colitis. (A) Experimental timeline for MFELNs treatment; (B) Weight loss of mice with different treatments; (C) Disease activity index; (D) Colon images; (E) Colon length analysis; (F) H&E staining of colon tissue (scale bars: 500 μm or 200 μm); (G–J) mRNA expression levels of TNF-α, IL-1β, IL-6, and IL-10 analyzed by qRT-PCR. (ns means no significant difference, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

Colon length was measured as a direct morphological indicator of colitis severity. On day 14, colons were collected from each group and their lengths were recorded (Figure 4D). The average colon length was 7.1 ± 0.3 cm in the normal control group, 5.3 ± 0.2 cm in the DSS model group, and 6.5 ± 0.3 cm in the MFELNs-treated group (Figure 4E). These results demonstrate that MFELNs treatment significantly alleviated DSS-induced colon shortening. Histopathological analysis of colon sections via H&E staining (Figure 4F) revealed severe mucosal damage in the DSS group, characterized by irregular crypt architecture, reduced and deformed crypts, and depletion of goblet cells. In contrast, the MFELNs group exhibited markedly preserved tissue integrity. Extensive inflammatory cell infiltration and pronounced thickening of the muscularis mucosa were observed in the DSS group, whereas the MFELNs group showed only minor inflammatory infiltration and muscularis mucosa thickness comparable to that of the normal control group. The mRNA expression levels of inflammatory cytokines in colon tissues were analyzed by qRT-PCR to evaluate the inflammatory status. The results demonstrated that the transcription levels of the pro-inflammatory cytokines TNF-α (Figure 4G), IL-1β (Figure 4H), and IL-6 (Figure 4I) were markedly elevated in the DSS group compared with the normal control. The treatment with MFELNs significantly downregulated the expression of these pro-inflammatory cytokines. In contrast, the expression of the anti-inflammatory cytokine IL-10 (Figure 4J) was substantially reduced in the DSS group, while MFELNs treatment restored its transcription level. Together, these findings indicate that MFELNs effectively attenuate DSS-induced colonic inflammation and ameliorate the severity of ulcerative colitis.

MFELNs Modulate Gut Microbiota

Mice with UC usually exhibit gut microbiota imbalance, characterized by reduced biodiversity, disruption of the balance between beneficial and pathogenic microorganisms, and changes in spatial distribution.³¹ To evaluate the impact of MFELNs treatment on gut microbiota, fecal contents were collected and subjected to 16S rRNA gene sequencing analysis. The Venn diagram based on ASVs (Figure 5A) showed that the DSS group had a numerically lower number of ASVs compared to the NC group, while the MFELNs group exhibited a slightly higher ASV count than the DSS group. However, α-diversity analysis using Chao1 and Simpson indices revealed no statistically significant differences among the three groups (Figure 5B), indicating that MFELNs treatment did not significantly alter the overall richness or evenness of the gut microbiota in UC mice. For β-diversity, although the analysis confirmed no significant differences in microbial community structure among the three groups (p > 0.05), principal co-ordinate analysis based on unweighted UniFrac distances (Figure 5C) and non metric multi-dimensional scaling based on Bray-Curtis distances (Figure 5D) showed a visual tendency for the MFELNs group to cluster closer to the NC group, separating from the DSS group. The bar charts illustrated the microbial community at the phylum level (Figure 5E). The DSS group exhibited an increased ratio of Firmicutes to Bacteroidota (p<0.05), whereas the MFELNs group returned to a ratio comparable to the NC group (ns) (Figure 5F). Collectively, these data imply that MFELNs may alleviate UC by modulating the balance of the gut microbiota.

Figure 5.

MFELNs modulate microbiota in DSS-induced UC. (A) Venn diagram based on amplicon sequence variant (ASV); (B) Chao1 index and Simpson index; (C) Principal co-ordinate analysis (PCoA) based on unweighted UniFrac distances; (D) Analysis of non-metric multi-dimensional scaling (NMDS) based on Bray-Curtis distances (stress<0.2 means significant difference); (E) Intestinal microbiota at the phylum level; (F) Firmicutes/Bacteroidota (F/B) ratio analysis. (ns means no significant difference, *p < 0.05).

Bioinformatics Analysis of MFELNs on Ulcerative Colitis Therapy

To elucidate the molecular mechanisms underlying the protective effect of MFELNs on UC, the potential bioactive components were screened and their bioinformatics analysis and molecular docking were performed. According to the criteria of oral bioavailability (OB) ≥ 30% and drug‑likeness (DL) ≥ 0.18 in TCMSP, 14 bioactive components were screened listed in Table S1. Only four components were retained after SwissADME screening. As listed in Table 2, the four potential effective components were sesamin, litsoeine/laurotetanine, demethylwedelolactone, and armepavine. Their potential target genes were predicted and combined through union process. The Venn diagram identified 88 overlapping target genes between four components and UC (Figure 6A). KEGG pathway enrichment analysis of these targets (Figure 6B) revealed the involvements in NF-κB signaling pathway and EGFR tyrosine kinase inhibitor resistance, respectively, suggesting the anti-inflammatory effect and intestinal mucosal repair of the four bioactive molecules. GO analysis (Figure 6C) further revealed more detailed information, including inflammatory response, positive regulation of cell adhesion, and response to wounding in BP, focal adhesion and receptor complex in CC, integrin binding, scaffold protein binding, and dioxygenase activity in MF. The results of KEGG and GO analysis confirmed that the four bioactive molecules treated UC through multi-dimensional pharmacological activities such as promoting repair and anti-inflammation. The PPI interaction network diagram (Figure 6D) revealed the core gene SRC, in addition to EGFR, NFKB1, AKT1, PIK3CA, ERBB, ESR1. The molecular docking diagrams (Figure 6E) showed the potential ligand-binding domains between SRC and sesamin, litsoeine/laurotetanine, demethylwedelolactone, and armepavine, respectively. The results showed that the four components can enter the active pocket of the SRC protein and form interactions with the amino acid residues through hydrogen bonds. The bonding energy between SRC and sesamin, litsoeine/laurotetanine, demethylwedelolactone, and armepavine were −8.6, −6.7, −7.8, −6.6 kcal/mol. These compounds may directly inhibit the activity of SRC, thereby simultaneously regulating its downstream repair pathways (EGFR/PI3K/AKT) and inflammatory pathways (NF-κB). Western blot results further confirmed this conclusion (Figure 7). The p-SRC/SRC, p-IκB/IκB, and p-p65/p65 ratios were significantly elevated after LPS induction and were significantly reduced following MFELNs treatment, suggesting that MFELNs suppress inflammation through the SRC/NF-κB signaling pathway. Collectively, the results indicate that the inhibition of inflammation driven by SRC/NF-κB may be the mechanism by which MFELNs alleviate UC.

Figure 6.

Potential mechanisms of four bioactive components of MFELNs (litsoeine/laurotetanine, armepavine, demethylwedelolactone, and sesamin) in treating ulcerative colitis. (A) The union of target genes; (B) KEGG analysis; (C) GO analysis in biological process (BP), cellular component (CC), molecular function (MF); (D) Protein–protein interaction (PPI) network; (E) Molecular docking diagrams between four components and SRC.

Figure 7.

Western blot analysis. (A) Western blot images of p-SRC, p-p65, p-IκBα, SRC, p65, IκBα and β-actin in different groups; (B) Quantitative analysis from images A. (ns means no significant difference, *p < 0.05, **p < 0.01, ***p < 0.001).

Discussion

Ulcerative colitis is a chronic and relapsing inflammatory bowel disease whose treatment remains challenging due to low clinical remission rates, high recurrence, and long-term safety concerns associated with existing therapies.^5,6 Therefore, developing novel therapeutic strategies with multi-target effects, favorable safety profiles, and distinct mechanisms is of significant clinical importance. In recent years, an emerging class of nanomedicines derived from Chinese herbal medicines (CHMs) has attracted considerable attention due to their unique self-assembly properties and intrinsic pharmacological activities.^18,19 During the traditional decoction process, various natural compounds in CHMs, such as polyphenols, alkaloids, saponins, and quinones, can spontaneously assemble into nanostructures through non-covalent interactions including hydrogen bonding, π-π stacking, and hydrophobic forces.^18,19 Notably, these self-assembled nanoparticles not only inherit the bioactive components of the herbs but also exhibit enhanced solubility, improved bioavailability, and multi-target therapeutic effects without the need for exogenous carriers.¹⁹ In this study, for the first time, we successfully isolated exosome-like nanoparticles from the flower buds of Magnolia biondii (Xinyi) via a traditional aqueous decoction process and systematically demonstrated their efficacy in ameliorating DSS-induced colitis through an integrated mechanism involving anti-inflammatory, antioxidant, and gut microbiota-modulating activities, with particular emphasis on the SRC/NF-κB signaling axis. Unlike cold-extracted plant-derived exosomes that are isolated from fresh plant tissues via juicing and differential centrifugation, MFELNs are generated through a disruption-reassembly process during high-temperature decoction, a mechanism fundamentally distinct from naturally occurring plant extracellular vesicles.²² This distinction positions MFELNs within the broader category of decoction-derived supramolecular assemblies, which have been increasingly recognized as an important material basis for the pharmacological effects of traditional Chinese medicinal formulas.

MFELNs exhibited typical nanovesicular morphology with a size distribution of 50–120 nm and a negative surface charge (Figure 1B–D), characteristics consistent with the self-assembled nanoparticles previously reported in various classic CHM decoctions. For instance, nanometre aggregates isolated from Baihu-Tang (BHT) decoction displayed spherical morphology with diameters of 25–500 nm and contained bioactive compounds such as mangiferin, neomangiferin, glycyrrhizic acid, and ammonium glycyrrhizinate.³² Similarly, nanoparticles from Maxing-Shigan-Tang (MXSGT) decoction were identified as 50–150 nm spherical structures that carried the majority of ephedrine and pseudoephedrine, thereby modulating the bioactivity of these alkaloids.³³ Beyond carrying small-molecule phytochemicals, these decoction-derived nanoparticles can also encapsulate functional nucleic acids. A notable example is honeysuckle decoction, in which the plant-derived microRNA MIR2911 is packaged within exosome-like nanoparticles and remains intact after boiling due to its unique GC-rich sequence.³⁴ This evidence confirms that the exosome-like character of decoction-derived nanoparticles is not merely a morphological analogy, but reflects their functional capacity to serve as carriers for both small-molecule and nucleic‑acid therapeutics. These findings collectively establish that the formation of bioactive nanoparticles during decoction is a general phenomenon across diverse CHM formulas, and that the presence of such nanoparticles may represent an important material basis for the therapeutic efficacy of traditional herbal decoctions.

Metabolomic profiling revealed that MFELNs are enriched in flavonoids, alkaloids, lignans and coumarins, phenolic acids, and terpenoids (Figure 1E and F). This diverse phytochemical composition is consistent with the known chemical constituents of Magnolia species, which include lignans such as fargesin, magnolin, and sesamin, as well as various alkaloids and essential oil components.^23,24,29 Many of these compounds possess well-documented anti-inflammatory and antioxidant properties,^25,27 providing a material basis for the multifaceted bioactivities of MFELNs. In the context of decoction-derived self-assembled nanoparticles, the presence of multiple component classes, including flavonoids, alkaloids, lignans, and phenolic acids, is thought to contribute to the formation of supramolecular structures through synergistic non-covalent interactions.¹⁹ This compositional complexity may explain why the whole nanoparticle fraction often exhibits superior bioactivity compared to individual monomeric compounds.

Our in vitro experiments demonstrated that MFELNs possess considerable antioxidant capacity, as evidenced by DPPH radical scavenging activity (Figure 3A) and suppression of H₂O₂ induced intracellular ROS generation (Figure 3B and C). These findings align with the established antioxidant properties of lignans and phenolic compounds present in Magnolia species.^28,29 Oxidative stress plays a critical role in UC pathogenesis by perpetuating inflammation, damaging the epithelial barrier, and promoting dysbiosis.³⁵ The ability of MFELNs to scavenge ROS thus represents an important mechanism for mitigating oxidative damage in the colonic mucosa. Furthermore, MFELNs significantly suppressed LPS-induced upregulation of pro-inflammatory cytokines (TNF-α, IL-1β, and IL-6) while restoring IL-10 expression in macrophages (Figure 3D). Flow cytometry analysis confirmed that MFELNs promoted macrophage polarization from the pro-inflammatory M1 phenotype toward the anti-inflammatory M2 phenotype (Figure 3E and F). This immunomodulatory effect is consistent with the anti-inflammatory actions of fargesin, a bioactive lignan from Magnolia species, which has been shown to ameliorate chemically induced IBD via NF-κB signaling suppression.³⁰ Importantly, the anti-inflammatory effects of decoction-derived nanoparticles have been well documented in other CHM formulas. For example, BHT-derived nanoparticles significantly reduced plasma levels of IL-6, IL-1β, and TNF-α in LPS-induced febrile rabbits,³² while MXSGT nanoparticles exhibited antipyretic efficacy equivalent to the full decoction via suppression of PGE2, IL-1β, and TNF-α.³³ These findings suggest that MFELNs modulate macrophage function through pathways involving the bioactive lignans and alkaloids identified in our metabolomic analysis, a mechanism that parallels the activity of nanoparticles from other heat-processed herbal decoctions.

Oral administration of MFELNs significantly attenuated DSS-induced colitis in mice, as evidenced by reduced body weight loss, lower DAI scores, less colon shortening, and diminished histopathological damage. These protective effects were accompanied by downregulation of colonic pro-inflammatory cytokines (TNF-α, IL-1β, and IL-6) and restoration of IL-10 expression (Figure 4), mirroring the in vitro anti-inflammatory profile. The in vivo efficacy of MFELNs is consistent with the therapeutic effects reported for other decoction-derived nanoparticles in various disease models. Analysis of the gut microbiota revealed that MFELNs treatment did not significantly affect α-diversity indices, but did induce notable shifts in microbial community structure, as demonstrated by PCoA and NMDS analyses (Figure 5A–D). At the phylum level, DSS induction increased the Firmicutes/Bacteroidota (F/B) ratio, a key indicator of intestinal microecological homeostasis that is commonly elevated in UC patients.³⁶ MFELNs treatment restored the F/B ratio to a level comparable to that of the normal control group, suggesting that MFELNs help maintain a balanced microbial community structure (Figure 5F). The F/B ratio has been widely recognized as a marker of gut dysbiosis, with an elevated ratio associated with increased inflammation and metabolic disorders.³⁷ The restoration of this ratio by MFELNs correlates with the observed reduction in colonic inflammation, suggesting that the anti-inflammatory effects of MFELNs may be partially mediated through normalization of the gut microbiota. This is consistent with the concept that decoction-derived nanoparticles can influence the gut microbiota. For instance, Gegen-Qinlian-Tang (GQD) was shown to regulate the ecological structure of the gut microbiota by enriching beneficial bacteria in type 2 diabetes patients,³⁸ and nanoparticles from BHT have been reported to modulate the intestinal microenvironment.³² The modulation of the F/B ratio by MFELNs aligns with the broader principle that supramolecular assemblies formed during decoction can exert systemic effects through the gut-axis.

Besides, the ability of MFELNs to promote fibroblast migration, as demonstrated by the wound-healing assay, further supports their role in mucosal repair (Figure 2D and E). Intestinal wound healing requires coordinated migration and proliferation of epithelial cells and fibroblasts, processes that are impaired in the inflamed UC mucosa.² The migration‑promoting effect of MFELNs may be attributed to their bioactive components, as flavonoids^39,40 and terpenoids⁴¹ have been reported to enhance cell migration and tissue repair. This repair‑promoting activity, combined with the anti-inflammatory and antioxidant effects, positions MFELNs as a multi-functional therapeutic agent for UC.

To elucidate the molecular mechanisms underlying the protective effects of MFELNs, we performed bioinformatics screening and molecular docking. Four potential bioactive components (sesamin, litsoeine/laurotetanine, demethylwedelolactone, and armepavine) were identified based on their favorable pharmacokinetic properties (OB ≥ 30%, DL ≥ 0.18) and drug likeness (Table 2). KEGG pathway enrichment analysis of the overlapping targets between these components and UC revealed significant enrichment in the NF-κB signaling pathway. GO analysis further implicated inflammatory response, cell adhesion, and wound healing processes. PPI network analysis identified SRC as a core hub gene within the regulatory network. Molecular docking simulations demonstrated favorable binding potential between the four active components and SRC (Figure 6), suggesting that these compounds may directly interact with SRC to simultaneously modulate downstream inflammatory (NF-κB) and repair (EGFR/PI3K/AKT) pathways. The WB results further confirmed the involvement of SRC/NF-κB signaling pathway in the treatment of MFELNs (Figure 7). SRC is a non-receptor tyrosine kinase that plays a pivotal role in integrating signals from various receptors and serves as a key amplifier of downstream signaling cascades.⁴² Importantly, SRC can directly phosphorylate the IKK complex, leading to activation of the NF-κB signaling pathway.^43,44 Conversely, SRC activation of PI3K/AKT1 can drive cell migration, adhesion, and survival.^45,46 Thus, SRC functions as a molecular switch that coordinately regulates inflammatory responses and tissue repair. By targeting SRC, MFELNs may simultaneously suppress NF-κB‑driven pro-inflammatory cytokine production while promoting EGFR/PI3K/AKT-mediated epithelial cell survival and migration. The identification of SRC as a potential molecular target for MFELNs provides a mechanistic foundation that distinguishes them from other decoction‑derived nanoparticles, which often exert their effects through different signaling pathways. For example, BHT nanoparticles primarily modulate the TLR4/NF-κB pathway,³² while MXSGT nanoparticles influence PGE2 and calcium signaling.³³

However, several limitations of this study should be acknowledged. First, the precise contribution of individual MFELNs components to the overall therapeutic effect remains to be elucidated. The complex phytochemical composition of MFELNs may result in synergistic actions that are not fully captured by single component analysis. Second, while the restoration of F/B ratio was observed, the direct measurements of short‑chain fatty acid levels or other microbial metabolites are still missing, limiting our understanding of the functional consequences of microbiota modulation. Finally, the long‑term safety and efficacy of MFELNs require further investigation in chronic colitis models and, eventually, clinical trials.

Conclusion

In conclusion, this study demonstrates that MFELNs, exosome-like nanoparticles spontaneously assembled during the decoction of Magnolia biondii flower buds (Xinyi), effectively ameliorate DSS-induced ulcerative colitis through an integrated mechanism involving direct anti-inflammatory and antioxidant effects, promotion of macrophage M2 polarization, and modulation of the gut microbiota. Bioinformatics, molecular docking, and Western blot analyses identify SRC as a potential molecular target through which the bioactive components of MFELNs may inhibit NF-κB-mediated inflammation. Our findings contribute to the growing body of evidence supporting Chinese herbal decoction-derived nanoparticles as next‑generation therapeutic platforms and provide a foundation for further development of Xinyi-based nanomedicines for inflammatory diseases.

Funding Statement

This work was financially supported by The 14th Five‑Year Plan Hubei Provincial Advantaged Characteristic Disciplines (Groups) Project of Wuhan University of Science and Technology (2023C0101).

Data Sharing Statement

All data needed to support the conclusions are present in the paper and/or the Supplementary Materials. Additional data is available from the corresponding authors upon request.

Disclosure

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.