Mannan-oligosaccharides-based self-assembly nanoparticles enable gut-specific quercetin release, providing multi-pathway protection against AD via gut-brain axis

Abstract

Mannan-oligosaccharides (MOS), as indigestible functional prebiotic, support intestinal health. Here, we engineered MOS to undergo linoleic acid (LA)-mediated self-assembly with quercetin, forming lipid bilayer nanoparticles (MOS-LA/LA-Q NPs). These nanoparticles (NPs) achieved a high quercetin encapsulation efficiency (92.4%) and demonstrated superior stability in aqueous solution at 4 °C. The MOS outer layer effectively protected the quercetin core from gastrointestinal stress. Triggered by microbial enzymes, the system enabled a sustained quercetin release, up to 60% over 48 h, representing a novel microbe-specific delivery mechanism. In an Alzheimer's disease (AD) mice models induced by AlCl₃/D-galactose, MOS-LA/LA-Q NPs significantly enhanced gut microbiota diversity, restored gut microbiota balance, suppressed pathogenic bacteria, and regulated metabolites short-chain fatty acid (SCFA) and lactic acid levels. Compared with free quercetin, these NPs alleviated cognitive deficits, reduced inflammation and oxidative stress in both the colon and brain, improved gut-brain barrier integrity, and restored gut-brain function more significantly. Mechanistically, these effects were associated with modulation of the Kelch-like ECH-associated protein 1 (Keap1)/nuclear factor erythroid 2-related factor 2 (Nrf2)/heme oxygenase-1 (HO-1) and toll-like receptor 4 (TLR4)/nuclear factor kappa-B (NF-κB) pathways, reducing tau phosphorylation and Aβ amyloid levels, thus demonstrating superior neuroprotection. This study proposes a novel strategy for addressing the complex pathology of AD through multi-target intervention, highlighting the potential of MOS-based NPs to enhance quercetin's therapeutic efficacy.

Graphical abstract

Highlights

• •The nano-formulation enables gut-specific release, overcoming the limitations of poor oral bioavailability of quercetin.
• •Multi-pathway action: modulates gut microbiota, reduces inflammation, and protects the brain.
• •Offers a novel nano-strategy for Alzheimer's disease prevention/therapy via the gut-brain axis.

1. Introduction

Mannan-oligosaccharides (MOS), natural prebiotics, have attracted growing interest in the nutraceutical and functional food industries due to their health benefits [1]. MOS selectively promote the growth of beneficial bacteria (e.g., Lactobacillus, Bifidobacterium), enhancing short-chain fatty acid (SCFA) production, particularly butyric acid, and maintaining intestinal barrier integrity, thus reducing lipopolysaccharide (LPS) leakage [2]. Unlike synthetic materials, natural polysaccharides such as MOS and chitosan are biodegradable, low in toxicity, and suitable for long-term therapeutic use [[3], [4], [5]]. They are also promising as nanodrug delivery carriers. For instance, ferulic acid-grafted oligofructose nanoparticles (NPs) alleviate gastrointestinal oxidative stress while improving ferulic acid colonic absorption [6].

Alzheimer's disease (AD) is a neurodegenerative disorder characterized by progressive cognitive decline. Core pathological features include β-amyloid (Aβ) plaque deposition, tau hyperphosphorylation leading to neurofibrillary tangles, and chronic neuroinflammation [7]. Despite progress in Aβ and tau targeted therapies, clinical trials often fail due to limited efficacy and side effects, highlighting the need for multi-target strategies. Recent evidence underscores oxidative stress-mediated neuroinflammation as a key AD progression driver. Specifically, Aβ deposition triggers excessive reactive oxygen species (ROS) production, leading to neuronal damage [8]. Moreover, Aβ accumulation activates microglia, triggering the release of inflammatory cytokines that further exacerbate neuroinflammation and accelerate AD progression [9]. Consequently, targeting oxidative stress and neuroinflammation is a promising therapeutic approach for AD.

Quercetin, a natural flavonoid, possesses anti-inflammatory, free radical scavenging, and antioxidant properties that can mitigate neuroinflammation and reduce AD severity [10]. However, its clinical potential is limited by poor oral bioavailability, rapid intestinal metabolism into inactive derivatives, and poor blood-brain barrier penetration. To address these challenges, various nano-delivery systems, including liposomes and polymer nanoparticles, have been developed to enhance the stability and target delivery of polyphenolic compounds such as quercetin [11,12]. Nevertheless, these systems alone remain insufficient to address the complex, multifactorial nature of AD.

The "gut-brain axis" concept has recently introduced novel possibilities for AD intervention. The gut microbiota influences central nervous system oxidative stress, inflammation, and synaptic plasticity through mechanisms such as SCFA production, immune modulation, and vagal nerve signaling. In patients with AD, the abundance of Firmicutes and SCFA-producing bacteria (e.g., Bifidobacterium, Ruminococcus spp.) is significantly reduced, weakening anti-inflammatory and antioxidant defenses [13]. Additionally, the overgrowth of pro-inflammatory bacteria such as Proteobacteria increases systemic inflammation and compromises blood-brain barrier integrity [14]. These observations support the use of microbiota-targeted interventions, including prebiotics and probiotics, as a promising strategy for AD prevention and treatment.

In this study, we developed a novel nanosystem in which linoleic acid (LA) bridges MOS and quercetin (Q) to form amphiphilic lipid bilayer nanoparticles (MOS-LA/LA-Q NPs) via molecular self-assembly. These NPs offer high quercetin encapsulation efficiency, enhanced gastrointestinal stability, and sustained release. Beyond supporting intestinal health, this system modulates the gut-brain axis and exerts significant neuroprotective effects in AD model mice, providing a multi-targeted approach for addressing AD.

2. Experimental

2.1. Chemical reagents

Quercetin (97% purity), dicyclohexylcarbodiimide (DCC), 4-dimethylaminopyridine (DMAP), LA, and dimethyl sulfoxide (DMSO) were purchased from Macklin Inc. (Shanghai, China). MOS (910.83 Da) were supplied by Yingxin Laboratory Equipment Co., Ltd. (Shanghai, China) (Fig. S1 and Table S1). Superoxide dismutase (SOD), catalase (CAT), and malondialdehyde (MDA) assay kits were obtained from Jiancheng Bioengineering Institute (Nanjing, Jiangsu, China). Enzyme-linked immunosorbent assay (ELISA) kits for interleukin-6 (IL-6), interleukin-1β (IL-1β), and tumor necrosis factor-alpha (TNF-α) were purchased from Animaluni Biotechnology Co., Ltd. (Shanghai, China).

2.2. Animals

Male ICR mice (6-week-old, specific pathogen free (SPF) grade) were obtained from Changsheng Biological Co., Ltd. (Benxi, Liaoning, China) and housed under standard conditions (12 h light/dark cycle, 25 °C ± 1 °C, 50% ± 10% relative humidity). Mice were acclimatized for one week before the experiment, with body weights monitored weekly. All animal experiments and procedures were strictly conducted in accordance with relevant laws, regulations, and institutional guidelines, and were approved by the Ethics Review Committee of Xinhua Hospital Affiliated to Dalian University (Approval number: K2024-002).

2.3. Synthesis of MOS-LA/LA-Q NPs

For mannan-oligosaccharides-linoleic acid (MOS-LA) synthesis, MOS, DCC, and DMAP were dissolved in 80 mL of DMSO, while LA was dissolved separately in 20 mL of DMSO. The two solutions were then combined and stirred at 25 °C for 24 h to ensure complete reaction between MOS and LA, thereby forming MOS-LA. The resulting solution was slowly added dropwise into 95% ethanol. Since MOS-LA is ethanol-insoluble, it was purified and precipitated via centrifugation (6000 g, 5 min, 4 °C). The obtained precipitate was resuspended in 30 mL of anhydrous ethanol and vacuum-dried at 45 °C to yield MOS-LA as a white solid (Fig. S2)

For the synthesis of quercetin and linoleic acid conjugates (LA-Q), Q, LA, DCC, and DMAP were dissolved in 30 mL of ethyl acetate and stirred at 25 °C for 24 h. The molar ratio of LA to Q was systematically optimized using different ratios (Q:LA = 1:3, 1:4, 1:5, and 1:6), The production of LA-Q and the remaining amount of Q at different feeding ratios were analyzed by high performance liquid chromatography (HPLC). Complete reaction between Q and LA was achieved when the ratio reached 1:6, leading to the successful formation of the LA-Q conjugate. Subsequently, the reaction mixture was filtered through filter paper to remove by-products, which remained on the surface of the filter medium. The residue was further washed thoroughly with ethyl acetate to ensure complete recovery of LA-Q. The resulting filtrate was concentrated under vacuum to afford LA-Q as a yellow oil (Fig. S2).

For the self-assembly of MOS-LA and LA-Q, after proportioning optimization, the MOS-LA and LA-Q prepared above were respectively taken in a ratio of 4:1 (3.2 g and 0.8 g) was placed in 40 mL of methanol solution, and rotary distillation was carried out in a rotary distillation flask at 40 °C and 80 rpm. Through film dispersion, MOS-LA and LA-Q were fully mixed and a film was formed on the flask wall. Finally, through dehydration and ultrasonic treatment, MOS-LA/LA-Q NPs solution was obtained (Fig. S2). The LA-Q and MOS-LA structures were confirmed using nuclear magnetic resonance hydrogen spectrum (¹H NMR).

2.4. Encapsulation efficiency (EE%) and drug loading (DL%) of MOS-LA/LA-Q NPs using HPLC

DL% and EE% of quercetin in MOS-LA/LA-Q NPs were determined using HPLC (SPD-M20A, Shimadzu, Kyoto, Japan). An Elite C₁₈ column (Supersil ODS2 4.6 mm × 250 mm × 5 μm) was used (Elite Analytical Instruments Co., Ltd., Dalian, Liaoning, China). The mobile phase A was ultrapure water containing 0.2% phosphoric acid, and the mobile phase B was methanol. Isocratic elution of B/A (55:45 = v/v) was used. The detection wavelength was 254 nm; the flow rate was 1 mL/min, the column temperature was 30 °C, and the injection volume was 20 μL [15]. In total, 1 mL of MOS-LA/LA-Q NPs aqueous solution was mixed with 9 mL of methanol, vortex-sonicated for 5 min, and centrifuged (3000 g, 10 min). The supernatant was analyzed by HPLC to quantify LA-Q concentration. Specific chromatograms and standard curves can be found in the Supplementary material (Fig. S3). DL% and EE% were calculated using the following two equations:

$$\text{EE}\%=\frac{\mathrm{W}\text{eight}\text{of}\text{the}\text{drug}\text{in}\text{the}\text{nanoparticle}}{\mathrm{W}\text{eight}\text{of}\text{the}\text{feeding}\text{drug}}\times 100\%$$

$$\text{DL}\%=\frac{\mathrm{W}\text{eight}\text{of}\text{the}\text{drug}\text{in}\text{the}\text{nanoparticle}}{\mathrm{W}\text{eight}\text{of}\text{the}\text{whole}\text{nanoparticle}}\times 100\%$$

2.5. Characterization of MOS-LA/LA-Q NPs

LA-Q and MOS-LA were dissolved in 40 mL of methanol, followed by rotary evaporation to remove the solvent. The residue was redispersed in 25 mL of mono-distilled water and sonicated to form MOS-LA/LA-Q NPs. MOS-LA/LA-Q NPs morphology and zeta potential were characterized using transmission electron microscopy (TEM) (JEOL JEM-2100, Nippon Electronica Co., Ltd., Tokyo, Japan) and a laser particle size analyzer (Nanobrook 90Plus PALS, Brookhaven Instruments Co., Ltd., New York, NY, USA). Structural features were further analyzed using ultraviolet-visible (UV-Vis) spectroscopy (V-560, JASCO Co., Ltd., Tokyo, Japan), fluorescence spectroscopy (FP6500, JASCO Co., Ltd., Tokyo, Japan), and Fourier transform infrared spectroscopy (FTIR) (Shimadzu IR Prestige-21, Shimadzu, Kyoto, Japan).

2.6. Verification of the core-shell structure of MOS-LA/LA-Q NPs

MOS-LA@Nil and MOS-LA/LA-Q@Nil NPs were prepared by the same method as that used to prepare MOS-LA/LA-Q NPs. The fluorescence intensity was observed by IVIS imaging (IVIS Lumina Series III, PerkinElmer, Waltham, MA, USA), and then the specific fluorescence intensity values were measured by microplate reader.

2.7. In vitro drug release

Drug release from MOS-LA/LA-Q NPs was evaluated using the dialysis bag diffusion method. One milliliter of NP suspension was placed in a dialysis bag and immersed in 5 mL of phosphate-buffered saline (PBS) at 37 °C under shaking conditions (100 rpm). At predetermined intervals, the release medium was replaced, and the quercetin content was quantified using HPLC. For microbe-mediated release studies, mouse cecal contents (MCC) were prepared by dissolving 2 g of MCC in 100 mL of PBS, followed by centrifugation (500 g, 5 min). The supernatant was used as the release medium. Samples were centrifuged (500 g, 5 min), filtered using a 0.45 μm membrane, and analyzed using HPLC.

2.8. In vitro stability of MOS-LA/LA-Q NPs in simulated gastrointestinal fluid

The stability of MOS-LA/LA-Q NPs was assessed in simulated gastric fluid (SGF; pH 1.2) and intestinal fluid (SIF; pH 6.8). One milliliter of NP solution was diluted to 6 mL with distilled water, and the pH was adjusted to 2.0. For SGF, 3 mL of pepsin solution was added, and the mixture was incubated at 37 °C. For SIF, the pH was adjusted to 7.5, trypsin solution was added, and the reaction was similarly conducted at 37 °C. Particle size measurements were taken at 0.5, 1, 2, 4, 6, and 8 h in both conditions.

2.9. In vitro antioxidant activity assay

The antioxidant activity of MOS-LA/LA-Q NPs was assessed by measuring their scavenging capacity against 2,2-diphenyl-1-picrylhydrazyl (DPPH) radicals using the BC4750 DPPH Radical Scavenging Capacity Assay Kit (Solarbio, Beijing, China). Total antioxidant capacity was assessed using the ABTS Total Antioxidant Capacity Assay Kit (S0119, Beyotime Biotechnology, Shanghai, China), following the manufacturer's protocols.

2.10. Animal design

Following one week of acclimatization, 32 male ICR mice were randomly assigned to four groups: control, model, Q, and MOS-LA/LA-Q NPs. Specifically, random sequences were generated by the sample function of the R software, and 32 mice were divided into four groups. Before grouping, we determined that all mice were male and weighed between 20 and 22 g. Except for the control group, all mice received daily intraperitoneal injections of AlCl₃ (40 mg/kg) and D-galactose (100 mg/kg). The Q and MOS-LA/LA-Q NPs groups were also treated by gavage with either free quercetin (50 mg/kg) or an equivalent dose of encapsulated quercetin based on the EE%. The treatment period lasted 8 weeks. Body weights were measured weekly. Behavioral tests were performed during the final week, followed by euthanasia and collection of serum, cecal contents, and tissues for further analyses.

2.11. Morris water maze (MWM) test

Spatial learning and memory were assessed using the MWM consisting of a circular pool (150 cm diameter, 50 cm height) with a hidden platform (6 cm diameter) submerged 1 cm below the water surface in the fourth quadrant. The water temperature was maintained at 22 °C. During training, mice were placed facing the pool wall, and escape latency (time to find the platform) was recorded. If unsuccessful within 120 s, mice were guided to the platform and allowed to rest for 30 s. Training occurred daily across all four quadrants over 5 days. On the final day, the platform was removed, and mice were allowed to swim freely from a fixed position in the second quadrant. A computerized tracking system (Smart 3.0, Panlab Co., Ltd., Barcelona, Spain) recorded model swimming trajectories, platform crossings, and time spent in the target quadrant.

2.12. Y-maze test

Short-term memory was assessed using the Y-maze apparatus (Model 63,006, Shenzhen Reward Life Technology Co., Ltd., Shenzhen, China). Mice were placed at the end of one arm and allowed to explore freely for 5 min without food stimuli. The number and sequence of arm entries were recorded, and the percentage of spontaneous alternations was calculated.

2.13. Nesting test

Nesting behavior was assessed as a cognitive function indicator. Mice were individually housed and acclimatized for two days. A cotton square (5 × 5 cm²) was placed in each cage, and nesting behavior was assessed at 12, 24, and 48 h using the following scale: 1, no tearing; 2, slight tearing and no nest; 3, obvious tearing and partial nest; 4, obvious tearing and clear nest.

2.14. The distribution of MOS-LA/LA-Q NPs in vivo

To explore the distribution of MOS-LA/LA-Q NPs in vivo, MOS-LA@DiR NPs was prepared by the same way as MOS-LA/LA-Q NPs. The mice were randomly divided into two groups (n = 3 per group) and orally administered with free DiR (Aladdin Co., Ltd., Shanghai, China) or MOS-LA@DiR (DiR concentration: 4 mg/kg). After anesthesia, all mice were photographed by IVIS. (IVIS Lumina Series III, PerkinElmer, Waltham, MA, USA) at certain time points (1, 4, 8, and 24 h). The fluorescence of DiR and the bioluminescence were both analyzed.

For quantitative analysis, the mice were randomly divided into two groups (n = 12 per group) and orally administered with free LA-Q or MOS-LA/LA-Q NPs (80 mg/kg LA-Q). Then the major organs (heart, liver, spleen, lung, brain and kidney) of 3 mice in each group were taken out at certain time points (1, 4, 8, and 24 h), homogenized, and centrifuged. The LA-Q concentrations in the supernatant were tested by HPLC.

2.15. Pharmacokinetics

Mice were randomly divided into two groups (n = 24 per group) and administered orally with free LA-Q or MOS-LA/LA-Q NPs (80 mg/kg LA-Q). The blood of mice (500 μL) was taken into eppendorf (EP) tubes equipped with 50 μL of ethylenediaminetetraacetic acid (EDTA) at 0 min, 30 min, 1 h, 2 h, 4 h, 8 h, 12 h, and 24 h, and then centrifuged (2000 g, 20 min). 200 μL of the supernatant was mixed with methanol (500 μL) and vortexed well. The mixture was centrifuged (2000 g, 20 min) again, and finally the LA-Q concentration in the supernatant was tested by HPLC.

2.16. Intestinal permeability was measured by FITC-dextran-4000 (FD4)

All mice were divided into four groups (control, model, Q, MOS-LA/LA-Q NPs, n = 3 per group). After 4 h of fasting, FD4 (Meilunbio Co., Ltd., Dalian, Liaoning, China) (500 mg/kg) was administered to mice by gavage, and the intestinal tract of mice was imaged by IVIS (IVIS Lumina Series III, PerkinElmer, Waltham, MA, USA) 4 h after gavage. Following collection, blood was centrifuged at 3000 rpm for 10 min. Serum was collected and fluorescence intensity was measured using a microplate reader with excitation wavelength of 485 nm and emission wavelength of 535 nm.

2.17. Hematoxylin-eosin (H&E) staining

Brain tissues were fixed in 4% paraformaldehyde, paraffin-embedded, and sectioned into 3-μm slices. Sections were deparaffinized, rehydrated using a graded alcohol series, and stained with H&E staining. Histopathological changes were observed using a light microscope, and images were captured for analysis.

2.18. Nissl staining

Nissl staining was performed on fixed brain tissue using Nissl staining solution (C0117, Beyotime, Shanghai, China) according to the manufacturer's instructions. Nissl-positive neurons were visualized using a light microscope (Nikon E100, Nikon Co., Ltd., Tokyo, Japan).

2.19. Immunofluorescence

Brain sections were incubated overnight at 4 °C with primary antibodies against neuronal nuclei (NeuN) (1:300, Abcam, Shanghai, China), Kelch-like ECH-associated protein 1 (Keap1) (1:5000, Servicebio, Wuhan, Hubei, China), nuclear factor erythroid 2-related factor 2 (Nrf2) (1:8000, Servicebio, Wuhan, Hubei, China), glial fibrillary acidic protein (GFAP) (1:300, Abcam, Shanghai, China), and ionized calcium-binding adapter molecule 1 (IBA-1) (1:300, Abcam, Shanghai, China). After washing, sections were incubated with horseradish peroxidase (HRP)-labeled secondary antibodies (1:500, Servicebio, Wuhan, Hubei, China) for 50 min at room temperature, followed by tyramide signal amplification dye (G1231, Servicebio, Wuhan, Hubei, China) in the dark for 10 min. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI), and fluorescence images were acquired using a Nikon Eclipse C1 orthogonal fluorescence microscope (Nikon Eclipse C1, Nikon Co., Ltd., Tokyo, Japan).

2.20. Immunohistochemistry (IHC) staining

Brain tissue sections were incubated overnight at 4 °C with primary antibodies against mucin 2 (MUC2) (1:500, Servicebio, Wuhan, Hubei, China), occludin (1:500, Servicebio, Wuhan, Hubei, China), and claudin-2 (1:200, Bioss, Beijing, China). After washing, HRP-labeled secondary antibodies were applied for 50 min at room temperature. Following 3,3′-diaminobenzidine (DAB) chromogenic development and hematoxylin counterstaining, sections were dehydrated, mounted, and imaged using a Nikon E100 white light microscope (Nikon E100, Nikon Co., Ltd., Tokyo, Japan).

2.21. Enzyme-linked immunosorbent (ELISA) assay

Brain homogenates were prepared and centrifuged to collect supernatants. IL-1β, IL-6 and TNF-α levels were measured using ELISA kits (Animaluni Biotechnology Co., Ltd., Shanghai, China) according to the manufacturer's protocol.

2.22. Western blotting

Brain tissues were weighed, lysed with radioimmunoprecipitation assay buffer (RIPA) lysis buffer (1:10, w/v), and sonicated. After centrifugation (12,000 rpm, 10 min, 4 °C), total protein concentrations were determined using a bicinchoninic acid (BCA) protein assay kit (PC0020, Solarbio, Beijing, China). Proteins were denatured by boiling, separated using 12.5% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), transferred onto polyvinylidene difluoride (PVDF) membrane, and blocked with 5% bovine serum albumin (BSA) for 1 h at room temperature. Membranes were incubated overnight at 4 °C with primary antibodies against p-tau (Abcam, Shanghai, China), Aβ1-42 (Bioss, Beijing, China), neurofilament light chain (NF-L) (Bioss, Beijing, China), GFAP (Bioss, Beijing, China), low-density lipoprotein receptor-related protein 1 (LRP-1) (Bioss, Beijing, China), receptor for advanced glycation endproducts (RAGE) (Bioss, Beijing, China), Nrf2 (Bioss, Beijing, China), Keap1 (Bioss, Beijing, China), toll-like receptor 4 (TLR4) (Bioss, Beijing, China), nuclear factor kappa-B (NF-κB) (Bioss, Beijing, China), monocarboxylate transporter 1 (MCT-1) (Bioss, Beijing, China), occludin (Servicebio, Wuhan, Hubei, China), claudin-1 (CLDN1) (Bioss, Beijing, China), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Bioss, Beijing, China). After washing three times with tris-buffered saline with tween 20 (TBST), membranes were incubated with HRP-conjugated secondary antibodies for 2 h at room temperature. Protein bands were visualized using an ECL chemiluminescence kit (Biosharp, Hefei, Anhui, China) and analyzed via grayscale density.

2.23. 16S rRNA gene sequencing analysis

Genomic DNA was extracted from 300 mg of mouse feces using hexadecyltrimethylammonium bromide (CTAB) or sodium dodecyl sulfate (SDS), and DNA purity and concentration were assessed using agarose gel electrophoresis. DNA samples were diluted to 1 ng/μL with sterile water and used as templates for polymerase chain reaction (PCR) amplification using specific primers (Phusion® High-Fidelity PCR Master Mix with GC Buffer, New England Biolabs, Ipswich, MA, USA) to ensure amplification efficiency and accuracy. PCR products were detected using 2% agarose gel electrophoresis, purified using magnetic beads, and quantified using an enzyme labeler. Samples were pooled in equal concentrations, mixed thoroughly, and re-detected using 2% agarose gel electrophoresis. Target bands were recovered using a QIAGEN's Gel Recovery Kit (QIAGEN Co., Ltd., Shenzhen, Guangdong, China) Sequencing libraries were prepared using the TruSeq® DNA PCR-Free Sample Preparation Kit (Illumin Co., Ltd., San Diego, CA, USA), and quantified using Qubit (Thermo Fisher Scientific Co., Ltd., Waltham, MA, USA) and qPCR (Bio-Rad Laboratories, Hercules, CA, USA). Libraries meeting quality criteria were sequenced using the NovaSeq6000 platform (Illumin Co., Ltd., San Diego, CA, USA).

2.24. Effect of MOS-LA/LA-Q NPs on the in vitro proliferation of Lactobacillus plantarum and Bifidobacterium spp.

To evaluate the effect of MOS-LA/LA-Q NPs on Lactobacillus plantarum and Bifidobacterium proliferation, De Man, Rogosa and Sharpe (MRS) liquid medium was prepared, and single colonies were inoculated and incubated anaerobically at 37 °C. MRS agar plates were prepared for both control and MOS-LA/LA-Q NPs groups, with 1 mL of MOS-LA/LA-Q NPs added per 10 mL of agar medium in the treatment group. Diluted bacterial suspensions were spread onto the plates and incubated at 37 °C. After incubation, colonies were counted and photographed for analysis.

2.25. Quantification of SCFA

For SCFAs analysis, 360 mg of feces was placed into a 5-mL sterile EP tube, and an internal standard (deutero-n-pentanoic acid) was added. The sample volume was adjusted to 1 mL with anhydrous ethanol, followed by ultrasonic lysis and centrifugation (13,000 rpm, 4 °C). The supernatant was filtered using a 0.22-μm membrane, concentrated under vacuum, and re-dissolved in 100 μL of chromatographically pure solvent for analysis. Blood and brain tissues were processed similarly. SCFA levels in plasma, brain, and feces were analyzed using liquid chromatography-mass spectrometry (LC-MS) (9030, Shimadzu, Kyoto, Japan).

2.26. Biocompatibility

Healthy ICR mice were randomly assigned into 3 groups (Control, Q, MOS-LA/LA-Q NPs). Mice were orally treated with various formulations (50 mg/kg for Q, 50 mg/kg for MOS-LA/LA-Q NPs). The mice were sacrificed, then liver and kidney were sectioned and stained with H&E.

2.27. Statistical analysis

All in vitro and in vivo experiments were carried out with at least 3 biological replicates for each experimental group. Biological replicates are indicated within figure legends. Data are expressed as the mean ± standard deviation (SD). Mice were assigned randomly to experimental groups. No statistical method was used to predetermine sample size. No samples were excluded from the analyses. Investigators were not blinded to allocation during experiments and outcome assessment, and the analyses were based on objectively measurable data. Unpaired two-sided Student's test were used when two groups were compared, with Welch's correction when variances are not equal. One-way analysis of variance (ANOVA) followed by Tukey's multiple comparisons post hoc test was used for testing differences between groups. Data were analyzed using GraphPad Prism 9.5 and IBM SPSS Statistics 27.

3. Results and discussion

3.1. Preparation and characterization of MOS-LA/LA-Q NPs

MOS-LA/LA-Q NPs were successfully prepared using the thin-film dispersion method (Fig. 1A). Compared with free Q, the aqueous solution of MOS-LA/LA-Q NPs remained stable after standing for 1 h (Fig. S2). TEM revealed that the NPs were spherical and uniformly distributed. According to statistics, the particle size distribution histogram shows that the particle size is approximately 54.3 nm (Figs. 1B and C). Dynamic light scattering analysis showed an average hydrated particle size of 58.79 ± 1.27 nm with a zeta potential of −37.54 ± 0.46 mV (Fig. 1D). Stability testing at 4 °C showed that the NP size remained stable in the 40–60 nm range over 14 days (Fig. 1E).

Fig. 1.

Preparation and characterization of mannan-oligosaccharides-linoleic acid/linoleic acid-quercetin nanoparticles (MOS-LA/LA-Q NPs). (A) Schematic diagram of the MOS-LA/LA-Q NPs preparation process. (B) Representative transmission electron microscope (TEM) images of MOS-LA/LA-Q NPs. (C) Particle size distribution histogram of MOS-LA/LA-Q NPs. (D) Hydrated particle size vs. zeta potential analysis of MOS-LA/LA-Q NPs. (E) Long-term stability of MOS-LA/LA-Q NPs stored at 4 °C for 14 days. (F) Ultraviolet-visible (UV-Vis) spectra of quercetin (Q), linoleic acid-quercetin (LA-Q), mannan-oligosaccharides-linoleic acid (MOS-LA), and MOS-LA/LA-Q NPs. (G) Fluorescence spectra of Q, LA-Q, MOS-LA, and MOS-LA/LA-Q NPs. (H) Fourier transform infrared spectra (FTIR) of Q, LA-Q, MOS-LA, and MOS-LA/LA-Q NPs. (I) In vitro stability of MOS-LA/LA-Q NPs in a simulated gastrointestinal environment. (J) Cumulative release kinetics of MOS-LA/LA-Q NPs in mouse cecal contents (MCC)-containing medium, in inactivated MCC-containing medium and PBS. (K) In vitro antioxidant activities of MOS-LA/LA-Q NPs evaluated using 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS) and 1,1-diphenyl-2-picrylhydrazyl radical (DPPH) assays. Data are presented as mean ± standard deviation (SD). In Fig. 1K, n = 3 for each group. ns: not siginificant. PDI: polydispersity index; T-AOC: total antioxidant capacity; IC₅₀: half maximal inhibitory concentration; DCC: N,N′-dicyclohexylcarbodiimide; DMAP: 4-dimethylaminopyridine; SGF: simulated gastric fluid; SIF: simulated intestinal fluid.

The MOS-LA/LA-Q NP chemical structure was further characterized using UV-Vis spectrum, fluorescence, and FTIR. In the UV-Vis spectrum, free Q displayed characteristic absorption peaks at 373 nm and 256 nm. After esterification with LA, the peak at 373 nm disappeared, shifting to 260 nm. Following self-assembly, MOS-LA and LA-Q formed MOS-LA/LA-Q NPs, with the characteristic UV absorption peak shifting to 277 nm (Fig. 1F), confirming successful esterification between LA and quercetin and MOS-LA/LA-Q NPs formation.

Fluorescence spectroscopy revealed that free Q exhibited an emission peak at 484 nm, whereas fluorescence intensity progressively decreased in LA-Q and MOS-LA/LA-Q NPs after esterification and NP assembly, indicating effective molecular bridging (Fig. 1G). To verify the core-shell structure of MOS-LA/LA-Q NPs, Nil environment responsive fluorescent dye was used. When Nil is in a hydrophobic environment, the fluorescence intensity will increase greatly. The fluorescence intensity of MOS-LA/LA-Q@Nil NPs and MOS-LA@Nil was significantly higher than that of Nil in PBS, indicating that the MOS-LA core has a hydrophobic inner cavity, which greatly increases the fluorescence intensity of Nil, indirectly confirming that the MOS-LA/LA-Q NPs have a core-shell structure with hydrophilic outer shell and hydrophobic inner core (Fig. S4). FTIR analysis further confirmed these structural changes (Fig. 1H). Specifically, the methylene (−CH₂−) stretching vibration peak at 2927 cm⁻¹ and the C=C stretching vibration peak at 1655 cm⁻¹ corresponded to the LA chain. A broad peak at 3200−3600 cm⁻¹, attributed to hydroxyl (−OH) stretching, was significantly reduced after esterification, reflecting the acylation of quercetin's phenolic hydroxyl groups (e.g., 3-OH, 7-OH, 4′-OH). The characteristic peaks at 1738 and 1095 cm⁻¹, corresponding to the ester carbonyl (C=O) and ester groups (C−O−C), respectively, provided evidence of successful ester bond formation.

In MOS-LA, the peak at 1028 cm⁻¹ corresponded to the glycan ring backbone (C−O−C vibrations), while pyran ring vibrations were observed at 809 cm⁻¹, 843 cm⁻¹, and 899 cm⁻¹, confirming that the glycosidic bonds and ring structure of the MOS remained intact. In MOS-LA/LA-Q NPs, ester bond peaks (1740 cm⁻¹ and 1089 cm⁻¹) matched those of LA-Q, while sugar ring peaks (1021 cm⁻¹ and 811 cm⁻¹) were consistent with MOS-LA, confirming the coexistence of both components. These results indicate that both LA-Q and MOS-LA underwent esterification, acquiring amphiphilic properties. The structures of MOS-LA and LA-Q were further confirmed by ¹H NMR spectroscopy (Figs. S5 and S6). The resulting MOS-LA/LA-Q NPs self-assembled into a structure featuring a hydrophilic mannan oligosaccharide shell and a hydrophobic quercetin core.

HPLC was used to qualitatively and quantitatively assess quercetin and LA-Q, with LA-Q serving as the standard for EE% calculations. Following esterification with LA, the quercetin retention time was shortened, and the quercetin EE in MOS-LA/LA-Q NPs reached 92.4%, significantly higher than the 76%−82.4% reported for chitosan-encapsulated quercetin nanoparticles [16].

MOS-LA/LA-Q NPs stability in simulated gastrointestinal conditions was evaluated by monitoring particle size in SGF and SIF. No significant size changes were observed over 8 h (Fig. 1I), likely due to the resistance of β-1,4 glycosidic bonds in the MOS shell to hydrolysis by digestive enzymes such as α-glucosidase and amylase, which primarily target α-glycosidic bonds. This structural stability protects the encapsulated LA-Q and preserves its biological activity.

The release behavior of LA-Q from MOS-LA/LA-Q NPs was further investigated under simulated gastrointestinal conditions. In PBS and inactivated MCC-containing PBS, cumulative LA-Q release from MOS-LA/LA-Q NPs was below 17% after 48 h, whereas in MCC-containing PBS, release reached 60% within the same period (Fig. 1J). This enhanced release in MCC-containing medium suggests that intestinal bacteria, such as Bifidobacterium and Lactobacillus, enzymatically degrade the MOS shell, triggering the gradual release of LA-Q. Moreover, the LA pro-ester chain may facilitate Q transport across physiological barriers, enhancing bioavailability. These results highlight the dual advantage of MOS-LA/LA-Q NPs as both a solubility enhancer and a gut microbe-specific, slow-release delivery system.

The in vitro antioxidant activities of MOS, MOS-LA, Q, LA-Q and MOS-LA/LA-Q NPs were investigated using 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS) and DPPH assays [17]. Esterification with LA slightly reduced the in vitro antioxidant activity of quercetin (LA-Q) (Fig. 1K), likely due to the decreased number of hydroxyl groups and steric hindrance caused by the introduction of long-chain LA, which affects the ability of quercetin molecules to bind free radicals. Additionally, the MOS-LA/LA-Q NPs antioxidant activity was further reduced, presumably because the MOS-LA outer layer encapsulating LA-Q provided additional shielding, limiting Q's interaction with free radicals. However, after oral administration, MOS-LA/LA-Q NPs still exhibited strong antioxidant activity following digestion of the MOS outer layer by gut microbiota. In fact, existing research has confirmed that lipophilic antioxidant phenols have a significant ability to scavenge ROS in ARPE-19 cells treated with H₂O₂ [18]. These findings indicate that MOS-LA/LA-Q NPs or LA-Q maintain a protective antioxidant function upon release. Therefore, this NP system can be considered a prodrug design.

3.2. Pharmacokinetic study and drug tissue distribution of MOS-LA/LA-Q NPs

To validate the enhanced oral bioavailability and sustained-release characteristics of MOS-LA/LA-Q NPs, we comprehensively evaluated their performance as a nano-delivery system through multi-dimensional pharmacokinetic and tissue distribution studies. In vivo live imaging revealed a significant advantage of the nanosystem in terms of tissue distribution (Fig. 2A). While free DiR was rapidly cleared after oral administration and became undetectable by 8 h, MOS-LA@DiR NPs maintained strong fluorescence signals in mice up to 24 h, demonstrating markedly prolonged systemic retention. Ex vivo organ imaging (Fig. 2B) further confirmed that MOS-LA@DiR NPs exhibited stronger and more persistent fluorescence distribution in the heart, liver, brain, lungs, spleen, and kidneys compared to the free drug at all time points (1, 4, 8, and 24 h). Particularly notable accumulation was observed in the liver and brain, indicating the MOS-LA/LA-Q NPs potential for sustained release and enhanced tissue accumulation. These findings align with the in vitro cumulative release profile of the nanoparticles (Fig. 1J). To further verify whether the nano-delivery system provides sustained release in the intestinal tract, we performed fluorescence imaging of mouse intestinal distribution (Fig. 2C). MOS-LA@DiR NPs showed significantly stronger fluorescence signals in the intestines than free Dir at all time points, confirming that the nanoparticles improve drug release efficiency in the gut through a sustained-release mechanism. We speculate that this enhanced intestinal retention and release, particularly the strong fluorescence signal maintained at 24 h, explains the persistent fluorescence observed in the brain and liver in ex vivo organ imaging at the same time point.

Fig. 2.

Pharmacokinetic study and drug tissue distribution of mannan-oligosaccharides-linoleic acid/linoleic acid-quercetin nanoparticles (MOS-LA/LA-Q NPs). (A) In vivo fluorescence imaging of mice after oral administration of free 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindotricarbocyanine iodide (DiR) or mannan-oligosaccharides-linoleic acid @DiR (MOS-LA@DiR). (B) Fluorescence imaging of each major organ in mice after oral administration of free DiR or MOS-LA@DiR. (C) Fluorescence imaging of the intestine of mice after oral administration of free DiR or MOS-LA@DiR. (D) Linoleic acid-quercetin (LA-Q) concentration-time curves in plasma of mice after oral administration of free LA-Q and MOS-LA/LA-Q NPs. (E) LA-Q concentrations in various tissues and organs of mice after oral administration of free LA-Q and MOS-LA/LA-Q NPs. Data are presented as mean ± standard deviation (SD). In Figs. 2A–E, n = 3 for each group.

To further investigate the improvement in oral bioavailability, we conducted pharmacokinetic studies of LA-Q. The analysis (Fig. 2D) demonstrated that the nanoformulation significantly enhanced the oral bioavailability of LA-Q, as evidenced by a substantially increased plasma AUC _0−t, a 122% prolongation of elimination half-life (t_1/2), a 123% extension of mean residence time (MRT), and a 66.5% reduction in plasma clearance (Table S2). These parameters collectively indicate altered drug disposition kinetics with prolonged systemic exposure. Correspondingly, brain distribution studies (Fig. 2E) showed that MOS-LA/LA-Q NPs maintained high drug concentrations during the critical 1−8 h window, with levels progressively increasing and peaking at 4 and 8 h, highlighting sustained brain retention. Similar trends of increasing drug concentrations were observed in other major organs, consistent with the extended circulation time.

The successful delivery and accumulation of MOS-LA/LA-Q NPs in brain tissue provide a crucial foundation for treating central nervous system diseases. The peak brain drug concentrations at 4−8 h, synchronized with prolonged plasma exposure, indicate effective drug delivery from systemic circulation to brain targets through improved blood circulation. This "gut reservoir-prolonged circulation-brain accumulation" cascade effect explains the superior neuroprotective efficacy observed in our AD model.

3.3. Effects of MOS-LA/LA-Q NPs on cognitive dysfunction in AD model mice

AlCl₃ combined with D-gal induces cognitive decline in mice, serving as a widely accepted AD mouse model [19]. D-gal promotes the accumulation of ROS and accelerates the formation of advanced glycation end products, ultimately leading to oxidative stress and better mimicking the aging process [20]. Aluminum exposure, an environmental risk factor for AD, enhances abnormal cleavage of amyloid precursor protein, increases the production and deposition of Aβ-40 and Aβ-42, and induces tau phosphorylation, resulting in neurofibrillary tangle formation [21]. Here, we evaluated the effects of MOS-LA/LA-Q NPs on memory and cognitive dysfunction in AD model mice using behavioral experiments, including the MWM, Y-maze, and nesting behavior tests, following AlCl₃/D-gal-induced modeling in ICR mice. The experimental timeline is shown in Fig. 3A.

Fig. 3.

Effects of mannan-oligosaccharides-linoleic acid/linoleic acid-quercetin nanoparticles (MOS-LA/LA-Q NPs) on cognitive dysfunction in Alzheimer's disease (AD) model mice. (A) Schematic diagram of the animal experimental design. (B) Representative swimming trajectories and trajectory heat maps of mice in the water maze with or without a platform. (C) Time spent in the target quadrant during the spatial probe trial. (D) Number of crossings over the former platform location during the probe trial. (E) The distance taken to first reach the target platform. (F) The time taken to first reach the target platform. (G) Proportion of time spent on the target and starting quadrants. (H) Average swimming speed during the morris water maze (MWM) test. (I) The total number of arm entry in Y-maze test. (J) Spontaneous alternation rate in the Y-maze test. (K) Escape latency in the MWM test. (L) Representative images of mouse nesting behavior at 48 h. (M) Quantitative analysis of nesting scores over 48 h. Data are presented as mean ± standard deviation (SD). In Figs. 3C–M, n = 6 for each group. ns: not siginificant. Q: quercetin.

In the MWM localization navigation test, representative swimming trajectories of the mice during the water maze test are shown in Fig. 3B. In the spatial exploration experiments, the MOS-LA/LA-Q NPs group displayed a significant increase in the number of platform crossings and longer exploration time in the target quadrant than the model group. Furthermore, these mice exhibited enhanced spatial learning, reaching the platform more quickly (shorter latency and path length) and displaying an improved search strategy, characterized by a greater preference for the target and starting quadrants. This improvement was also more pronounced than that observed in the unmodified Q group (Figs. 3C−G). In the Y-maze spontaneous alternation test, the MOS-LA/LA-Q NPs group showed significantly enhanced memory performance than the model and Q groups. No significant differences in average swimming speed and number of arm entry in Y-maze were observed among the groups (Figs. 3H−J), indicating that locomotor abilities remained comparable. The model group exhibited significantly longer escape latency than the other groups, whereas the MOS-LA/LA-Q NPs intervention led to significantly shorter escape latency than that in the model group (Fig. 3K).

Mouse nesting behavior, a natural and sensitive indicator of cognitive decline, was also evaluated. In our nesting experiment, model mice showed only partial shredding of the nesting material within 48 h, without forming well-structured spherical nests. The Q group exhibited more active shredding but limited nest formation, whereas the MOS-LA/LA-Q NPs group displayed significantly greater shredding and stacking behavior, with nesting scores approaching those of the control group (Figs. 3L and M).

In conclusion, across multiple behavioral experiments, MOS-LA/LA-Q NPs demonstrated effective improvement of spatial memory deficits and cognitive impairments in AD model mice, with superior outcomes compared to those obtained from the Q treatment alone.

3.4. Neuroprotective effects of MOS-LA/LA-Q NPs

Neurons are the primary components of the nervous system and serve as the foundation for neuroelectric signal transmission, as well as for the mechanisms underlying memory information construction and storage [22]. In AD model mice, memory loss is often associated with neuronal damage [23].

We hypothesized that the design of MOS-LA/LA-Q NPs enhances Q bioavailability and its ability to cross physiological barriers, thereby improving its neuroprotective efficacy. To investigate the MOS-LA/LA-Q NPs neuroprotective effects in AD model mice, we employed immunofluorescence experiments to explore the expression of NeuN protein. Results showed that positive NeuN expression in the Q group was only slightly higher than that in the AD model group, without significant differences, indicating that Q had limited effects on regulating NeuN protein expression. This finding was consistent with its inability to effectively inhibit neuronal sequestration under pathological conditions. Conversely, the MOS-LA/LA-Q NPs intervention group had significantly increased NeuN fluorescence intensity than the AD model group (P < 0.05), with dense fluorescence signals observed in the neuronal cytoplasm (Figs. 4A and B).

Fig. 4.

Neuroprotective effects of mannan-oligosaccharides-linoleic acid/linoleic acid-quercetin nanoparticles (MOS-LA/LA-Q NPs). (A) Immunofluorescence images of hippocampal neurons showing neuronal nuclei (NeuN) expression. (B) Quantitative analysis of NeuN fluorescence intensity. (C) Representative histopathological images of the cerebral cortex and hippocampus. Arrows indicate areas of neuronal sequestration and edema. (D) Western blot images for phosphorylated-tau at Ser396 (p-tau), amyloid beta 1-42 (Aβ1-42), and neurofilament light chain (NF-L). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as the internal control. (E) Aβ1-42/GAPDH ratio. (F) p-tau/GAPDH ratio. (G) NF-L/GAPDH ratio. Data are presented as mean ± standard deviation (SD). In Figs. 4A, B and E–G, n = 6 for each group. DAPI: 4′,6-diamidino-2-phenylindole; H&E: hematoxylin-eosin staining; Q: quercetin.

Brain histopathology was examined using H&E staining. Results showed that the model group exhibited significant neuronal sequestration and edema compared to the control group. After Q intervention, edema was reduced to a small cortex area, suggesting that while Q partially alleviates edema, its effect may be limited due to single-target action or insufficient dosage. Conversely, the MOS-LA/LA-Q NPs intervention group displayed the mildest pathological changes, with only occasional neuronal sequestration observed in the cornu ammonis 3 (CA3) and dentate gyrus (DG) regions, and complete disappearance of edema. These findings demonstrate the superior neuroprotective potential of MOS-LA/LA-Q NPs compared to Q treatment alone. Nissl staining further confirmed the potential of the MOS-LA/LA-Q NPs group in alleviating neuronal injury (Fig. 4C).

These results suggest that MOS-LA/LA-Q NPs exhibit stronger effects on nerve regeneration and neuronal protection. This neuroprotective effect may be achieved by inhibiting pro-apoptotic proteins via Akt phosphorylation, activating anti-apoptotic proteins to maintain neuronal survival, or suppressing oxidative stress-related pathways [24].

To further investigate the regulatory effects of MOS-LA/LA-Q NPs on AD-related pathological indicators, we focused on evaluating phospho-tau (Ser396) and Aβ1-42 as markers of AD progression, along with NF-L proteins as potential AD biomarkers. The results demonstrated that both Aβ1-42 and phospho-tau (p-tau) at Ser396 protein levels were significantly more upregulated in the brains of the AD model group than in those of the control group (P < 0.05 and P < 0.01, respectively). MOS-LA/LA-Q NPs significantly reduced the Aβ1-42 and p-tau expression levels, although p-tau did not fully recover to the level observed in the control group. However, the marked improvement in neuronal morphology (e.g., reduction of sequestration and edema) suggests that the MOS-LA/LA-Q NPs neuroprotective effects may be primarily mediated through anti-inflammatory, antioxidant, or synaptic stabilization mechanisms, rather than direct inhibition of tau phosphorylation (Figs. 4D−F). NF-L, a potential AD biomarker, exhibited experimental trends consistent with previous findings, showing significant reductions in both the Q and MOS-LA/LA-Q NPs groups (Figs. 4D and G). Our results indicate that MOS-LA/LA-Q NPs exert significant neuroprotective effects and can effectively alleviate AD pathology via multiple mechanisms.

3.5. MOS-LA/LA-Q NPs remodel gut microbiota disorders in AD model mice

The gut microbiota plays a critical role in AD progression [25]. To evaluate the effects of MOS-LA/LA-Q NPs on the gut microbiota of AD model mice, we performed 16S rRNA gene sequencing of fecal samples. Operational taxonomic units (OTUs) were clustered for all sample species compositions, and species annotation was conducted. Venn diagram analysis revealed that the control, model, Q, and MOS-LA/LA-Q NPs groups contained 11, 11, 6, and 13 unique OTUs, respectively (Fig. 5A). Principal coordinates analysis (PCoA) showed that the microbial composition of the model group was significantly different from that of the other groups, whereas the MOS-LA/LA-Q NPs samples clustered closely with those of the control group after intervention. This suggests that MOS-LA/LA-Q NPs effectively remodeled the gut microbiota in AD model mice (Fig. 5B). Compared with those of the model group, the Ace index and Chao1 index were significantly higher in the MOS-LA/LA-Q NPs group (P < 0.01 for both), indicating increased species richness. The Simpson index showed a slight increase without statistical significance (P > 0.05), while the Shannon index significantly increased (P < 0.05) (Fig. 5C). Through the ternary phase diagram, the distribution of flora related to genus-level classifications within each group can be elucidated (Fig. 5D). These results indicate that MOS-LA/LA-Q NPs intervention enhanced both the richness and diversity of the gut microbiota in AD model mice. Beta diversity differences between groups were further analyzed using a distance matrix heatmap (Fig. 5E) and UPGMA (Unweighted pair group method with arithmetic mean) clustering tree (Fig. 5F). The results confirmed that MOS-LA/LA-Q NPs intervention altered the abundance of intestinal microbial communities and improved species diversity compared to those in the model group. At the phylum level, the UPGMA clustering tree analysis showed that the clustering pattern of the MOS-LA/LA-Q NPs group was more similar to that of the control group than to that of the model group.

Fig. 5.

Mannan-oligosaccharides-linoleic acid/linoleic acid-quercetin nanoparticles (MOS-LA/LA-Q NPs) remodel gut microbiota disorders in Alzheimer's disease (AD) model mice. (A) Venn diagram showing the shared and unique operational taxonomic unit (OTUs) among groups. (B) Principal coordinates analysis (PCoA) illustrating beta diversity differences between groups. (C) α-Diversity analysis (based on Ace, Chao1, Shannon, and Simpson indices). (D) Ternary phase diagram at the genus level. (E) Distance matrix heatmap representing dissimilarity of microbial communities between groups. (F) Unweighted pair group method with arithmetic mean (UPGMA) clustering tree showing hierarchical clustering of gut microbiota. (G) Linear discriminant analysis effect size (LEfSe) analysis identifying differentially abundant taxa between groups. (H) Linear discriminant analysis (LDA) scores indicating the effect size of identified microbial biomarkers. Data are presented as mean ± standard deviation (SD). n = 6 for each group. Q: quercetin.

We also examined specific bacterial communities within the gut flora using linear discriminant analysis effect size (LEfSe) to identify key bacterial taxa following MOS-LA/LA-Q intervention (Figs. 5G and H). The genera Klebsiella and Lachnoclostridium were identified as characteristic intestinal microbiome biomarkers in the AD model group, with significantly higher abundances compared to that in the control group (P < 0.05). Klebsiella and Lachnoclostridium contribute to intestinal barrier disruption, leading to the “leaky gut” phenomenon [26]. After MOS-LA/LA-Q NPs intervention, the dominant intestinal microorganisms shifted to Dubosiella and Bifidobacterium pseudolongum, which alleviate oxidative stress and support gut flora remodeling [27], consistent with our findings. Overall, MOS-LA/LA-Q NPs intervention effectively regulated the structural composition and diversity of the gut microbiota, increased the abundance of beneficial dominant species, and reduced the presence of harmful species, thereby remodeling gut dysbiosis in AD model mice.

3.6. Effects of MOS-LA/LA-Q NPs on intestinal function, gut microbiota composition, and proliferation in AD model mice

To further elucidate the detailed differences in gut microbiota composition and abundance among groups, we conducted genus-level analyses. The relative abundances of Frisingicoccus, Turicibacter, and Proteus were significantly higher in the model group than in the control group (P < 0.01, P < 0.05, P < 0.05). In contrast, the abundances of these three bacteria were significantly lower in both the Q and MOS-LA/LA-Q NPs intervention groups than in the model group (P < 0.01, P < 0.05, P < 0.05) (Fig. 6A). Frisingicoccus is strongly associated with neuroinflammation, as well as gut barrier integrity; the reduced abundance of this bacterium following intervention correlates with delayed cognitive decline and reduced gut barrier inflammation [28]. In an autism mouse model, Turicibacter was negatively correlated with SCFAs, particularly butyric acid, which plays a critical role in maintaining blood-brain barrier function and modulating neuroinflammation [29]. Additionally, significantly elevated Proteus levels have been observed in ulcerative colitis, depression, and cognitive impairment models, consistent with our results [30]. Conversely, the relative abundances of Bifidobacterium, Muribaculum, Lactobacillaceae, Odoribacter, and Mucispirillum were significantly lower in the model group than in the control group. After intervention with either Q or MOS-LA/LA-Q NPs, the abundances of these five beneficial bacteria were significantly increased (P < 0.05). Notably, the MOS-LA/LA-Q NPs group showed a more substantial increase than the Q group, with relative abundance increases of 1667.03% (P < 0.05), 385.17% (P < 0.05), 120.05% (P > 0.05), 221.20% (P < 0.05), and 160.36% (P < 0.05), respectively. Although the relative abundances of Faecalibaculum, Allobaculum, Dubosiella, and Parvibacter did not differ significantly between the model and control groups, the relative abundance of these bacteria was significantly higher in the MOS-LA/LA-Q NPs intervention group than in the model group (Fig. 6A). Moreover, their relative abundances were also significantly higher in the MOS-LA/LA-Q NPs group than in the Q group, showing a more pronounced upward trend (P < 0.05 or P > 0.05). Specifically, the increases were of 201.68% (P > 0.05), 3.72% (P > 0.05), 937.61% (P < 0.05), and 206.15% (P < 0.05), respectively. Further analysis revealed that Muribaculum is negatively associated with cognitive impairment and contributes to SCFA production through the degradation of endogenous mucinoglycans and exogenous dietary fiber. This strain also exhibits a cross-feeding relationship with probiotics such as Bifidobacterium and Lactobacillus, supporting gut microbiota remodeling [31]. Bifidobacterium has been used in AD mouse models to mediate neuroinflammation primarily by modulating gut microbiota composition and enhancing SCFA production [32]. Moreover, Faecalibaculum, Allobaculum, and Dubosiella are key SCFAs producers, particularly butyric acid, which regulates the intestinal microenvironment. Butyric acid significantly upregulates intestinal occludin expression, reduces serum LPS concentrations, and protects the integrity of both the intestinal and blood-brain barriers [33]. Additionally, Dubosiella is positively correlated with IL-10 and negatively correlated with IL-6, suggesting its involvement in modulating neuroinflammation and alleviating cognitive impairment. Collectively, these findings indicate that MOS-LA/LA-Q NPs, as a nano-delivery system, offer significant advantages in restoring gut microbiota balance. Importantly, MOS functions not only as a structural component in nanoparticle assembly but also as a prebiotic to regulate and remodel the gut microbiota.

Fig. 6.

Effects of mannan-oligosaccharides-linoleic acid/linoleic acid-quercetin nanoparticles (MOS-LA/LA-Q NPs) on intestinal function, gut microbiota composition, and proliferation in Alzheimer's disease (AD) model mice. (A) Relative abundances of differential bacterial genera in the gut microbiota across experimental groups. (B) Analysis of colon length with representative images of colons from each group. (C) Representative hematoxylin and eosin (H&E) staining of colon tissues showing histological differences among groups. (D) In vitro validation of Lactobacillus and Bifidobacterium proliferation in media supplemented with MOS-LA/LA-Q NPs vs. the control group. Data are presented as mean ± standard deviation (SD). In Fig. 6A, for each group n = 6, in Fig. 6B, n = 3 for each group. ns: not siginificant. Q: quercetin.

Owing to the regulatory effect of MOS-LA/LA-Q NPs on the gut microbiota, we further investigated their impact on the colonic health of mice mediated by modulating the gut microbiota. Results demonstrated that the colon length in the AD model group was significantly shorter compared to the control group, whereas the colon length in the MOS-LA/LA-Q NPs treatment group was markedly restored to a level comparable to the control group. H&E staining analysis corroborated these findings, showing that the colonic crypt structure in the MOS-LA/LA-Q NPs group was more intact and the number of goblet cells was significantly higher than in the AD model group (Figs. 6B and C).

To preliminarily verify the effects of MOS-LA/LA-Q NPs on gut microbiota proliferation, we conducted in vitro proliferation experiments using representative Lactobacillus and Bifidobacterium strains, The results showed that colony numbers were significantly higher in the medium supplemented with MOS-LA/LA-Q NPs than in the control group, confirming the proliferation-promoting effect of MOS-LA/LA-Q NPs on Lactobacillus and Bifidobacterium. This finding aligns with our previous results showing increased abundance of Lactobacillaceae following MOS-LA/LA-Q NPs treatment (Fig. 6D). Exogenous supplementation with Lactobacillus plantarum, as well as modulation of Lactobacillus abundance through natural compounds, can effectively delay AD progression, significantly reduce amyloid expression in the brain, and remodel the gut microbiota [34].

In summary, our experimental results further support existing evidence that gut microbiota disorganization, characterized by increased abundance of pathogenic bacteria and decreased abundance of beneficial bacteria, contributes to leaky gut development, thereby exacerbating neurological damage through pathogenic factors. The MOS-LA/LA-Q NPs intervention successfully restored gut microbiota balance in AD model mice by reducing the abundance of pathogenic bacteria and promoting the proliferation of beneficial bacteria. The superior therapeutic outcomes observed with MOS-LA/LA-Q NPs, compared to free quercetin, are likely attributable to a synergistic effect between the prebiotic MOS carrier and the encapsulated drug. However, a limitation of the current experimental design is the lack of a standalone MOS or MOS-LA blank NP group. Although our in vitro assays confirmed the prebiotic function of the MOS-LA/LA-Q NPs, the absence of this in vivo control precludes a definitive quantification of the independent contribution of MOS to the overall gut microbiota remodeling and SCFA changes observed. Future studies incorporating this control will be crucial to fully dissect the mechanism of action and validate the proposed synergy.

3.7. Effect of MOS-LA/LA-Q NPs on metabolite levels, distribution, oxidative stress, inflammatory factor burden, and correlation with gut microbiota in AD model mice

SCFAs are key metabolic products generated by gut microbiota, and disruptions in microbial balance often leads to altered SCFA production and distribution [35]. SCFAs can reduce the release of inflammatory factors by inhibiting microglial activation, enhance blood-brain barrier integrity, and thereby limit the penetration of peripheral inflammatory factors and toxic substances (e.g., β-amyloid) into the brain. MCT-1 serves as the primary transporter for SCFAs [36], making it important to assess not only the SCFA levels across tissues but also MCT-1 expression in both brain and colon tissues. Here, we used LC-MS to analyze SCFA levels in brain tissue, plasma, and feces across the experimental groups. The fecal SCFA levels were significantly lower in the AD model group than in the control group (P < 0.05). However, MOS-LA/LA-Q NPs intervention significantly increased fecal SCFA levels (P < 0.01), as well as the plasma and brain SCFA levels (P < 0.05, P < 0.01) (Fig. 7A). These results suggest that MOS-LA/LA-Q NPs not only remodeled the gut microbiota but also restored SCFAs levels and their systemic distribution. To further investigate SCFAs transport, we examined MCT-1 expression. Western blot results showed that MCT-1 expression in both colon and brain tissues was significantly higher in the MOS-LA/LA-Q NPs group than in the model group (P < 0.05, P < 0.01) (Figs. 7B and C). These findings suggest that MOS-LA/LA-Q NPs enhanced SCFA bioavailability through MCT-1-mediated transport from the colon into circulation and across the blood-brain barrier into the brain.

Fig. 7.

Effect of mannan-oligosaccharides-linoleic acid/linoleic acid-quercetin nanoparticles (MOS-LA/LA-Q NPs) on metabolite levels, distribution, oxidative stress, inflammatory factor burden, and correlation with gut microbiota in Alzheimer's disease (AD) model mice. (A) Short-chain fatty acids (SCFAs) and lactic acid concentrations in feces, plasma, and brain tissue measured using liquid chromatography-mass spectrometry (LC-MS). (B) Western blot analysis of monocarboxylate transporter 1 (MCT-1) protein expression in brain and colon tissues, with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as the internal control. (C) Quantification of MCT-1 protein expression levels in brain and colon tissues. (D) Levels of catalase (CAT) activity, superoxide dismutase (SOD) activity, malondialdehyde (MDA), and inflammatory cytokines tumor necrosis factor alpha (TNF-α), interleukin-1β (IL-1β), interleukin-6 (IL-6) concentration in brain tissue. (E) Spearman correlation heatmap showing the relationships between specific gut microbiota at genera, SOD and CAT levels, MDA concentration, and TNF-α, IL-1β, IL-6 in the brain. Data are presented as mean ± standard deviation (SD). In Figs. 7A–D, n = 6 for each group. g: genus; f: family; Q: quercetin.

In AD pathology, glucose metabolism in the brain is significantly impaired, and lactate serves as an important alternative energy source, facilitating astrocyte-neuron coupling to maintain mitochondrial respiration and synaptic activity [37]. Elevated activity of indoleamine-2,3-dioxygenase 1 (IDO1) in astrocytes of patients with AD inhibits glycolysis and lactate production, leading to neuronal energy deficits. Notably, IDO1 inhibition (e.g., using PF-068) can restore lactate production in astrocytes and promote its transport to neurons via MCT-1/2, ultimately improving synaptic plasticity and cognitive function [38]. Here, lactate levels were significantly lower in both feces and brain tissues of the AD model group than in those of the control group (P < 0.05, P < 0.01). However, MOS-LA/LA-Q NPs intervention significantly reversed this reduction, restoring lactate levels in both feces and brain tissues (P < 0.01 for both) (Fig. 7C). This suggests that the increased lactate observed in brain tissue may be attributed to the direct regulatory effects of MOS-LA/LA-Q NPs on astrocytes, potentially via attenuation of neuroinflammation and oxidative stress. Conversely, elevated feces lactate and colon MCT-1 expression imply that gut-derived lactate production and subsequent transport to the brain via MCT-1 may also contribute to maintaining neuronal energy supply in AD mice. In conclusion, MOS-LA/LA-Q NPs alleviated AD-related pathological changes by remodeling the gut microbiota, restoring metabolite balance, and enhancing the levels and distribution of key metabolites such as SCFAs and lactate. This dual mechanism, providing antioxidant and anti-inflammatory effects from SCFAs and sustaining neuronal energy through lactate transport, highlights the therapeutic potential of MOS-LA/LA-Q NPs in mitigating cognitive decline in AD model mice.

Oxidative stress is one of the core mechanisms in AD pathology. Aβ accumulation promotes free radical generation, damages cell membranes, and triggers neurodegeneration. Decreased activities of antioxidant enzymes (e.g., SOD and CAT) further exacerbate oxidative damage, while elevated MDA levels serve as a hallmark of lipid peroxidation and oxidative injury. Moreover, pro-inflammatory cytokines (e.g., TNF-α, IL-6, and IL-1β) amplify AD inflammatory responses by activating microglia and astrocytes, thereby worsening Aβ accumulation and tau pathology. To evaluate the MOS-LA/LA-Q NPs regulatory effects on oxidative stress and neuroinflammation, we examined antioxidant enzyme levels and inflammatory factor concentrations in the brain tissue of AD model mice. Results showed that MOS-LA/LA-Q NPs significantly enhanced CAT and SOD activity (P < 0.01 and P < 0.05, respectively), while significantly decreased TNF-α, IL-6, IL-1β, and MDA levels (P < 0.01, P < 0.01, P < 0.01, and P < 0.05, respectively) (Fig. 7D). These findings suggest that MOS-LA/LA-Q NPs exert anti-inflammatory and antioxidant effects, contributing to their neuroprotective potential. We hypothesized that MOS-LA/LA-Q NPs could alleviate oxidative stress and inflammation not only through direct pathways but also via gut microbiota remodeling and enhanced SCFAs production.

To explore this connection, we performed correlation analyses between gut microbial genera and environmental factors using Spearman's correlation coefficients. The analysis revealed that antioxidant enzymes (CAT and SOD) were positively correlated with genera such as Faecalibaculum, Allobaculum and Dubosiella, while showing negative correlations with Frisingicoccus, Turicibacter and Proteus. Conversely, MDA levels were positively correlated with Frisingicoccus, Turicibacter, and Proteus, and negatively with beneficial genera such as Bifidobacterium, Faecalibaculum and Mucispirillum. Similarly, pro-inflammatory cytokines (IL-1β, IL-6, and TNF-α) exhibited negative correlations with Mucispirillum, Parvibacter and Lactobacillus, while being positively correlated with Frisingicoccus, Turicibacter and Proteus (as shown in Fig. 7E). Frisingicoccus has been significantly associated with anxiety in a Parkinson's mouse model, where its modulation via acupuncture reduced inflammation [39]. Proteus is a key intestinal inflammation inducer in Crohn's disease and other inflammatory conditions [40]. Moreover, Turicibacter has been implicated as a potential biomarker in autism spectrum disorders. In contrast, Faecalibaculum, Dubosiella, and Allobaculum are major SCFA producers, which play a crucial role in modulating oxidative stress and neuroinflammation [41]. Supplementation with Bifidobacterium also exerts anti-inflammatory effects by enhancing SCFA production and supporting the proliferation of other beneficial bacteria. Furthermore, Mucispirillum helps maintain redox balance by scavenging oxygen and ROS during inflammatory conditions [42]. Collectively, these studies confirm that MOS-LA/LA-Q NPs reduce oxidative stress and neuroinflammation in AD model mice by restoring gut microbial balance and enhancing SCFAs production.

3.8. MOS-LA/LA-Q NPs restore the integrity of the colon and blood-brain barrier in AD model mice

The intestinal barrier and the blood-brain barrier (BBB) are two critical physiological barriers in the body, and their dysfunction is closely linked to AD progression [43]. Gut microbiota and their metabolites play essential roles in regulating the integrity of both the intestinal barrier and the BBB, thus influencing key AD pathological features, including neuroinflammation and Aβ deposition [44]. Intestinal barrier disruption, characterized by abnormally high expression of claudin-2 and downregulation of occludin and MUC2, facilitates the translocation of pathogens and LPS, which in turn activates inflammatory responses [45]. Our immunohistochemical analysis revealed that MOS-LA/LA-Q NPs intervention significantly reduced the elevated expression of colonic claudin-2 and markedly increased the expression of both MUC2 and occludin in AD model mice compared to the corresponding levels in the model group. Notably, the MOS-LA/LA-Q NPs group showed significantly higher expression of colonic occludin and MUC2 than the Q group (Figs. 8A and B). In order to further prove the health of the mouse intestinal barrier, FD4 fluorescence imaging was used to verify the intestinal peristalsis and intestinal permeability (Figs. S7A and B). The results showed that the fluorescence intensity of the MOS-LA/LA-Q NPs group was significantly lower than that of the model group at 4 h after gavage, and the overall fluorescence was concentrated in the colon stage, indicating that the mice had good intestinal peristalsis. Furthermore, by analyzing the fluorescence intensity of FD4 in the serum of the mice, the serum fluorescence intensity of the MOS-LA/LA-Q NPs group was significantly lower than that of the model group, which confirmed that the intestinal barrier of MOS-LA/LA-Q NPs was intact and less FD4 leaked into the blood, which was consistent with our immunohistochemical results. These findings suggest that the improvement of colon integrity by MOS-LA/LA-Q NPs may be associated with their gut microbiota effects, as previously reported for MOS. Here, restoration of colon barrier function may contribute to the mitigation of leaky gut-induced AD pathology.

Fig. 8.

Mannan-oligosaccharides-linoleic acid/linoleic acid-quercetin nanoparticles (MOS-LA/LA-Q NPs) restore the integrity of the colon and blood-brain barrier in Alzheimer's disease (AD) model mice. (A,B) Quantitative analysis of mucin 2 (MUC2), occludin and claudin-2 protein expression in colon tissues (A), and representative immunohistochemical images (B). (C) Western blot analysis and grayscale quantification of low-density lipoprotein receptor-related protein 1 (LRP-1) and receptor for advanced glycation endproducts (RAGE). (D) Western blot analysis and grayscale quantification of occludin and claudin-1 (CLDN1) in brain tissue. Data are presented as mean ± SD. In Figs. 8A–C and D, n = 6 for each group. IOD: integrated optical density; GAPDH: glyceraldehyde-3-phosphate dehydrogenase; Q: quercetin.

Moreover, we assessed the impact of MOS-LA/LA-Q NPs on BBB integrity. Results showed that MOS-LA/LA-Q NPs significantly increased the expression of occludin and CLDN1 in brain tissues compared to that in model mice (Figs. 8C and D), suggesting effective reconstruction of BBB integrity. This protective effect on the BBB is likely mediated by the remodeling of gut microbiota and the increased production of beneficial metabolites such as SCFAs [46]. BBB dysfunction is a well-established contributor to AD pathogenesis, leading to impaired clearance of Aβ from the brain into peripheral circulation. Specifically, decreased expression of low-density lipoprotein receptor-related protein 1 (LRP-1) and elevated expression of RAGE hinder Aβ transport, exacerbating its accumulation in the brain. Here, MOS-LA/LA-Q NPs intervention significantly reduced RAGE expression while increasing LRP-1 expression in AD model mice. This dual regulation of Aβ transporters may directly contribute to the observed reduction in brain Aβ1-42 levels following MOS-LA/LA-Q NPs treatment. In conclusion, MOS-LA/LA-Q NPs effectively repair both the intestinal barrier and the BBB, demonstrating strong potential in alleviating BBB dysfunction and improving AD.

3.9. MOS-LA/LA-Q NPs reduce glial cell activation by regulating the TLR4/NF-κB pathway and decreasing pro-inflammatory cytokine release

AD is characterized by the dysregulation of microglia and astrocytes, the resident brain immune cells, which become activated in response to Aβ aggregation [47]. Microglia activation triggers neuroinflammation, contributing to neuronal death. Previous studies have demonstrated that AD-related microglial inflammation can be attenuated through modulation of the TLR4/MYD88/NF-κB signaling pathway [48]. To investigate the regulatory effects of MOS-LA/LA-Q NPs on neuroinflammation, we examined their role in modulating microglia and astrocyte activation as well as their impact on the NF-κB pathway. The results showed that, compared to the model group, MOS-LA/LA-Q NPs significantly reduced the expression of the glial cell activation markers IBA-1 and GFAP (Figs. 9A and B). As activated microglia and astrocytes secrete high levels of pro-inflammatory cytokines (such as IL-1β, IL-6 and TNF-α), we conducted an ELISA experiment to assess the inflammation, and the results are shown in Fig. 8. The MOS-LA/LA-Q NPs group exhibited a significant reduction in the levels of these cytokines, with effects more pronounced than those observed in the Q group. These results suggest that MOS-LA/LA-Q NPs effectively inhibit glial cell hyperactivation and normalize inflammatory factor expression in the brains of AD model mice.

Fig. 9.

Mannan-oligosaccharides-linoleic acid/linoleic acid-quercetin nanoparticles (MOS-LA/LA-Q NPs) reduce glial cell activation and decrease pro-inflammatory cytokine release by modulating the toll-like receptor 4 (TLR4)/nuclear factor kappa-B (NF-κB) pathway. (A) Representative immunofluorescence images showing the expression of 4′,6-diamidino-2-phenylindole (DAPI) (blue), glial fibrillary acidic protein (GFAP) (green), and ionized calcium-binding adaptor molecule 1 (IBA-1) (red) in the cerebral cortex and hippocampus across different groups. (B) Quantitative fluorescence intensity analysis of GFAP and IBA-1 in brain sections. (C) Western blot images showing the expression of NF-κB, TLR4 and GFAP proteins in brain tissue, with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) used as a loading control. (D) Grayscale value analysis of NF-κB, TLR4, and GFAP protein expression levels based on Western blot results. Data are presented as mean ± standard deviation (SD). In Figs. 9B and D, n = 6 for each group. Q: quercetin.

To further elucidate the underlying mechanism, we analyzed the expression of key proteins in the TLR4/NF-κB inflammatory signaling pathway using Western blotting. TLR4 and NF-κB protein levels were significantly elevated in the model group, confirming activation of this pathway and its role in glial cell activation and pro-inflammatory cytokine production. Notably, MOS-LA/LA-Q NPs intervention markedly reversed the upregulation of both TLR4 and NF-κB (P < 0.01 for both) (Figs. 9C and D). Although the Q group also showed some downregulation of these proteins (P > 0.05 and P < 0.05, respectively), the extent of suppression was less pronounced than that achieved in the MOS-LA/LA-Q NPs group.

Taken together, these results suggest that the potent anti-inflammatory effects of MOS-LA/LA-Q NPs stem from the dual functionality of their nanoparticle structure and composition. On one hand, the NPs may directly impair NF-κB activation by blocking TLR4 receptor dimerization or inhibiting downstream myeloid differentiation primary response protein 88 (MyD88)/TIR domain-containing adaptor protein inducing interferon-β (TRIF) signaling. On the other hand, the intrinsic Q anti-inflammatory properties (e.g., its ability to inhibit IKKβ phosphorylation) likely synergize with the gut microbiota remodeling effects of MOS, enhancing immunomodulation. This combined action further suppresses NF-κB nuclear translocation and pro-inflammatory gene transcription. In contrast, the Q group alone lacked the multi-target synergy and optimized delivery system provided by the nanoparticle design, resulting in comparatively limited efficacy in blocking the inflammatory cascade. In conclusion, MOS-LA/LA-Q NPs effectively ameliorated microglia and astrocyte dysregulation and modulated the TLR4/NF-κB pathway, leading to significant inflammatory effects in AD model mice. In our future work, we will further investigate their ability to accumulate at sites of brain inflammation using additional methods, such as small-animal live imaging to validate this hypothesis.

3.10. MOS-LA/LA-Q NPs reduce oxidative stress damage by regulating the Keap1/Nrf2/heme oxygenase-1 (HO-1) pathway

Oxidative stress plays a pivotal role in AD pathogenesis and is widely recognized as a key contributing factor [49]. Nrf2, a crucial oxidative stress response protein, maintains intracellular ROS balance and protects cells from free radical-induced damage. Previous studies have reported reduced Nrf2 expression in the brains of AD mouse models, with downregulation of Nrf2 accelerating cognitive impairment. In contrast, Nrf2 activation enhances antioxidant defense mechanisms, mitigating oxidative damage and alleviating cognitive impairment in AD model mice [50]. Q, a potent antioxidant, improves cognitive deficits in AD model mice by reducing oxidative stress. In the material characterization section of our study, we demonstrated that MOS-LA/LA-Q NPs improve the solubility of Q and protect it from degradation in the gastrointestinal environment, thereby enhancing its release and bioavailability. Based on these findings, we hypothesized that encapsulating quercetin within MOS-LA/LA-Q NPs could enhance its regulatory effects on the Nrf2 pathway. Results showed that compared with the model group, MOS-LA/LA-Q NPs intervention significantly decreased Keap1 expression and increased Nrf2 expression levels (Figs. 10A and B). Moreover, MOS-LA/LA-Q NPs promoted significant nuclear translocation and aggregation of Nrf2, suggesting that these NPs effectively inhibit Nrf2 degradation by either degrading Keap1 or blocking its binding to Nrf2. This allows Nrf2 to accumulate in the nucleus and activate downstream antioxidant pathways. In contrast, although the Q intervention group slightly increased Nrf2 expression and partially inhibited the Keap1-Nrf2 interaction, it failed to effectively regulate Keap1 levels or promote Nrf2 nuclear translocation, resulting in limited antioxidant capacity. These findings were further validated by WB analysis, which confirmed the immunofluorescence results (Figs. 10C and D).

Fig. 10.

Mannan-oligosaccharides-linoleic acid/linoleic acid-quercetin nanoparticles (MOS-LA/LA-Q NPs) reduce oxidative stress injury by regulating the kelch-like ECH-associated protein 1 (Keap1)/nuclear factor erythroid 2-related factor 2 (Nrf2)/heme oxygenase-1 (HO-1) pathway. (A) Representative immunofluorescence images of Keap1, Nrf2 and HO-1 expression in brain tissue. (B) Quantitative fluorescence intensity analysis of Keap1, Nrf2 and HO-1 in brain tissue. (C) Representative Western blot images of Nrf2 and Keap1 protein expression in brain tissue. (D) Grayscale value analysis of Nrf2 and Keap1 protein bands from Western blot results. Data are presented as mean ± standard deviation (SD). In Figs. 10B and D, n = 6 for each group. DAPI: 4′,6-diamidino-2-phenylindole; Q: quercetin.

To further assess the activation of the Keap1/Nrf2 pathway and its downstream target gene HO-1, we evaluated HO-1 expression levels using immunofluorescence. Results showed that HO-1 expression closely corresponded to the activation state of the Keap1/Nrf2 pathway. Among the four groups, the MOS-LA/LA-Q NPs group exhibited the highest HO-1 expression level (Figs. 10A and B), where significant intranuclear Nrf2 aggregation strongly promoted HO-1 transcription. This suggests that MOS-LA/LA-Q NPs facilitated Nrf2 release by targeting Keap1 inhibition, thus enabling effective Nrf2 nuclear translocation, thereby enhancing robust activation of HO-1 antioxidant gene expression. This mechanism thereby enhances the neuronal defense against oxidative damage. In comparison, although HO-1 expression in the Q group was significantly higher than that in the model group (P < 0.05), it remained significantly lower than that in the MOS-LA/LA-Q NPs group (P < 0.05). This result indicates that Q alone only partially stabilized Nrf2, without sufficiently promoting its nuclear translocation, thus limiting HO-1 transcriptional activation. These findings suggest that Q's regulatory effect on the Keap1-Nrf2 pathway remains primarily at the level of inhibiting Nrf2 degradation, failing to overcome the nuclear translocation bottleneck and resulting in a weaker antioxidant response. Our results also suggest that MOS-LA/LA-Q NPs may more effectively cross the BBB due to their smaller particle size, thereby exerting stronger antioxidant effects. Moreover, the encapsulation of Q within the NP system not only enhances its solubility and protects it from degradation in the gastrointestinal tract but also improves its bioavailability, which likely contributes to more effective modulation of the Keap1-Nrf2 axis observed. Furthermore, the gut microbiota remodeling effects of MOS-LA/LA-Q NPs and the associated production of metabolites such as SCFAs may further support oxidative stress alleviation. Probiotics and their metabolites can reduce oxidative stress in middle-aged and elderly mice, aligning with our findings. In conclusion, MOS-LA/LA-Q NPs significantly reduced oxidative stress and alleviated the AD-related pathology via activation of the Keap1/Nrf2/HO-1 pathway. However, further studies using Nrf2 knockout models are needed to fully validate the reliability and specificity of this pathway.

4. Conclusion

Here, we successfully designed and prepared lipid bilayer MOS-LA/LA-Q NPs through the self-assembly of Q and MOS bridged by LA. Comprehensive characterization demonstrated that these NPs exhibited high Q EE, excellent aqueous stability, and an innovative structural design. The outer layer, composed of MOS, effectively protected the inner Q core from degradation in the gastrointestinal environment, enabling a sustained and slow release of LA-Q triggered specifically by microbial metabolic enzymes in an intestinal environment. This design realizes a novel microbe-specific triggered release. Regarding biological efficacy, MOS-LA/LA-Q NPs significantly promoted gut microbiota diversity in AD model mice, enhanced intestinal flora restoration, suppressed the proliferation of pathogenic bacteria, and regulated the SCFAs and lactic acid levels. Compared with quercetin alone, MOS-LA/LA-Q NPs more effectively alleviated cognitive impairment in AD model mice, restored gut-brain axis functionality, and significantly reduced hallmark pathological changes associated with AD by modulating multiple signaling pathways. Overall, this work offers a new strategy to address the multifactorial challenges of AD, providing a multi-pathway approach for its prevention and treatment. It also presents a novel perspective on the development and application of MOS and highlights a promising strategy to enhance the neuroprotective effects of Q.

CRediT authorship contribution statement

Guifeng Zhang: Writing – review & editing, Writing – original draft, Methodology. Hongjiang Li: Formal analysis. Wei Zhang: Visualization, Investigation. Dan Pu: Visualization, Software. Longxing Wang: Data curation. Xiaofei Han: Supervision, Methodology, Funding acquisition, Conceptualization. Ju Wu: Methodology, Conceptualization.

Data availability

Data will be made available on request.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article.