Quercetin Delivered by Mesenchymal Stem Cell-Derived Exosomes Improves Liver Fibrosis via the PI3K/Akt Signaling Pathway Authors: Dan Fang, Jing Zhang, Qingqing Zhao, Yuting Fan, Xiaojin Lin, Dan Liang, Lianjun Zhong, Yingchun Zhang, Simeng Yan, Tao Shen, Yingkun Meng, Xing Zhao, Shi Zuo Published: 2026 ABSTRACT Liver fibrosis represents a critical stage in the progression of chronic liver diseases to cirrhosis and hepatocellular carcinoma; however, effective therapeutic options remain limited. Although quercetin, a natural flavonol, possesses potent antifibrotic properties, its clinical utility is severely hindered by poor aqueous solubility and low bioavailability. To address this limitation, we developed a nanoparticle-based drug delivery system using quercetin-loaded human umbilical cord mesenchymal stem cell (hUC-MSC)-derived exosomes (hUC-MSC-exo-Que). Leveraging the innate biocompatibility and targeting capability of exosomes, this strategy aims to improve the pharmacokinetic limitations of quercetin and amplify its therapeutic efficacy. Our results demonstrate that hUC-MSC-exo-Que significantly attenuates liver fibrosis in a carbon tetrachloride-induced mouse model, outperforming free quercetin at the equivalent dose. This enhanced efficacy is attributed to the superior inhibition of hepatic stellate cell activation, as confirmed by in vitro studies. The engineered exosomes exhibited a sustained drug release profile (up to 48 h) and maintained excellent stability for at least 1 week. Integrating network pharmacology with experimental validation, we identify the antifibrotic mechanism involving potent inhibition of the PI3K/Akt signaling pathway, with hUC-MSC-exo-Que achieving markedly greater pathway suppression than free quercetin. By successfully transforming a potent but poorly bioavailable phytochemical into a targeted nanotherapeutic, we present a promising preclinical strategy for liver fibrosis treatment and demonstrate a proof-of-concept platform for hydrophobic drug delivery. Introduction Liver fibrosis (LF) is a major global health concern and a common pathological stage in the progression of chronic liver injury , as well as a necessary precursor to cirrhosis. Cirrhosis-associated complications account for approximately one million deaths annually, representing nearly 50% of liver disease-related mortality. Liver fibrosis results from chronic hepatic inflammation secondary to multiple etiologies, including viral hepatitis, alcoholic and nonalcoholic fatty liver disease, drug toxicity, cholestasis, and autoimmune liver disorders. However, effective therapeutic strategies for liver fibrosis or cirrhosis remain limited. , Existing treatments frequently lack sufficient specificity and efficacy, highlighting the need for targeted therapeutic approaches. , Liver fibrosis is closely associated with the activation of hepatic stellate cells (HSCs), , characterized by upregulation of α-smooth muscle actin (α-SMA) and collagen type I (Col I). Given that HSC activation is the pivotal event in fibrogenesis, strategies aimed at inhibiting or eliminating activated HSCs may alleviate fibrosis and potentially reverse disease progression. , Quercetin (3,3′,4′,5,7-pentahydroxyflavone) is a plant-derived flavonol and one of the most abundant dietary polyphenols, exhibiting antioxidant and anti-inflammatory characteristics. ,,− It has been shown to improve liver function and attenuates fibrosis , with similar antifibrotic effects has been reported in the myocardium, lung, kidney, and skeletal muscle. Despite its therapeutic potential, clinical translation of quercetin faces significant pharmacological challenges including intrinsically poor water solubility, a short plasma half-life, and notably low oral bioavailability (approximately 17% in rats and 1% in humans), which severely limits its systemic exposure and efficacy. ,− Exosomes are 30–150 nm extracellular vesicles. ,, Mesenchymal stem cell-derived exosomes offer additional advantages by combining natural hepatoprotective bioactivity with carrier capacity, making them particularly attractive for liver-targeted therapies. Collectively, these features position cell-derived exosomes as next-generation, clinically translatable delivery tools for proteins, nucleic acids and small-molecule drugs. , Encapsulation of quercetin within exosomes may overcome its pharmacokinetic limitations and improve preclinical efficacy, as suggested by previous studies. ,− However, several gaps remain: (i) the optimal loading method for incorporating quercetin into hUC-MSC-exo has not been systematically compared; and (ii) the antifibrotic efficacy of quercetin-loaded hUC-MSC-exo has not been rigorously evaluated against free quercetin under matched conditions. The present study aims to address these gaps by comparing four loading methods, identifying the optimal one (extrusion), and providing a head-to-head evaluation of hUC-MSC-exo-Que versus free quercetin in both in vitro and in vivo models of liver fibrosis. While the overall concept builds on prior work, our study provides a preliminary evaluation of hUC-MSC-exosome-based delivery for quercetin in the context of liver fibrosis. These findings may help clarify the potential benefits of this approach. Experimental Section Materials Quercetin (purity: >97.00%) was purchased from Glpbio Technology Inc. (Montclair, CA, USA GN10266). The human HSC line LX-2 and immortalized normal human hepatocyte line LO2 were obtained from Procell Life Science and Technology Co., Ltd. (Wuhan, China); TGF-β1 (0333129PD19) was obtained from Novoprotein Scientific (Shanghai) Inc., and CCl4 (purity >99%) was obtained from Rionlon Bohua Pharmaceutical and Chemical (Tianjin, China). Male C57BL/6 mice, aged 6–8 weeks and weighing 22 ± 3 g were obtained from Tianqin Biotechnology (Changsha, China). All experimental animals in this study were handled in accordance with the animal experimental guidelines approved by the Animal Ethics Committee of Guizhou Chinese Medical University (animal ethics approval number: 2503172). Antibodies against α-SMA and GAPDH were purchased from Huabio (China), and those against Akt and p-Akt were obtained from Selleck. Exosome Isolation and Characterization Human umbilical cords were obtained from the affiliated hospital of Guizhou Medical University with the approval from the Ethics Committee of the Affiliated Hospital of Guizhou Medical University (Approval No. 2026-38). Human umbilical cord mesenchymal stem cells (hUC-MSCs) were seeded at a density of 1 × 105 cells/mL and cultured in exosome-depleted medium for 24 h. The conditioned medium was collected, and hUC-MSC-exo were isolated using a series of ultracentrifugation steps as previously described. The morphology was detected using transmission electron microscopy (TEM). Particle size distribution and zeta potential were determined using a nanoparticle tracking analyzer (Resuntech, Shenzhen, China). The protein concentration of the exosomes was quantified with a BCA assay kit (Beyotime, China) following the manufacturer’s instructions. The expression of exosomal markers (CD63, CD81, TSG101) and the negative marker calnexin was confirmed by Western blotting. Preparation of hUC-MSC-exo-Que Quercetin was loaded into hUC-MSC-exo using the following four methods: Co-incubation: The mixture of hUC-MSC-exo and quercetin was incubated at 37 °C for 1 h. Electroporation: Loading was performed using a Celetrix electroporation system (Celetrix, USA) with parameters set at 450 V for 120 ms. After electroporation, the mixture was incubated at 37 °C for 30 min to allow pore closure. Extrusion: The mixture was passed 20 times through a 200 nm polycarbonate membrane using a liposome extruder, followed by incubation at 37 °C for 30 min. Sonication: The mixture was sonicated on ice using an ultrasonic disruptor (BioSafer 650-92, China) with a 20% amplitude in cycles of 36 s on and 4 s off, for six cycles, before incubating on ice for 2 min and then at 37 °C for 1 h. Determination of the % Encapsulation Rate (% ER) for hUC-MSC-exo-Que A spectral scan was performed using different concentrations of quercetin. A standard curve was first established using known quercetin concentrations by measuring its absorbance at 370 nm using a microplate reader (Synergy H1, BioTek Instruments, USA). To determine the encapsulation efficiency (EE), the hUC-MSC-exo-Que were separated from free (unencapsulated) quercetin by centrifugation at 100,000 × g for 70 min at 4 °C. The absorbance of the supernatant, which contained the free quercetin, was measured at 370 nm. The EE was calculated using the following formula: EE(%)=[(Totalamountofquercetinadded−Amountoffreequercetininsupernatant)/Totalamountofquercetinadded]×100% In Vitro Drug Release The in vitro release of quercetin from hUC-MSC-exo-Que was assessed via dialysis. Quercetin was loaded into exosomes at a 1:1 (m/m) ratio by extrusion. The mixture was placed in a dialysis bag (molecular weight cutoff: 12–14 kDa) and immersed in 30 mL of release medium (phosphate-buffered saline [PBS] with 1% Tween 80, pH 7.4). The tubes were incubated at 37 °C with shaking (120 rpm) in the dark. Aliquots of the external medium were collected and replaced with fresh medium at predetermined intervals to maintain sink conditions. The amount of quercetin released was quantified by measuring the absorbance at 370 nm using a microplate reader. The experiments were performed in triplicate. Storage Stability of hUC-MSC-exo-Que The stability of the hUC-MSC-exo-Que formulation was evaluated over 1 week. The particle size distribution and zeta potential (as an indicator of surface charge) were monitored using a nanoparticle tracking analyzer (Resuntech, Shenzhen, China). Internalization of hUC-MSC-exo-Que To assess cellular internalization, hUC-MSC-exo-Que were labeled with the fluorescent dye DiO. Free dye was removed by centrifugation using a 100-kDa ultrafiltration tube. Subsequently, LX-2 cells were cocultured with the DiO-labeled hUC-MSC-exo-Que for 24 h. After incubation, cells were washed with PBS, fixed with 4% paraformaldehyde, and counterstained with DAPI (10 μg/mL). Cellular internalization was visualized using a laser scanning confocal microscope (Nikon AX). The DiO signal (green) from the internalized exosomes was detected at an excitation wavelength of 485 nm, and the DAPI signal (blue) from the nuclei was detected at 430 nm. Biodistribution of hUC-MSC-exo-Que To evaluate the in vivo biodistribution, hUC-MSC-exo-Que were labeled with the near-infrared fluorescent dye DIR by incubation at 37 °C for 30 min, following the manufacturer’s protocol. The labeled nanoparticle system (hUC-MSC-exo-Que-DIR) was purified via ultrafiltration to remove unbound dye and resuspended in PBS. C57BL/6 mice were intravenously injected via the tail vein with a dose of 100 μg of hUC-MSC-exo-Que-DIR. At 24 h postinjection, the mice were euthanized, and major organs (heart, liver, spleen, lung, and kidney) were harvested for ex vivo fluorescence imaging using an in vivo imaging system to quantify exosome accumulation. Bioinformatics Analysis The PharmMapper, SwissTargetPrediction database and DrugBank were also used to predict and identify possible quercetin targets. Genes closely related to liver fibrosis were obtained from the GeneCards database. An intersection analysis was conducted to identify genes between the two gene sets. Protein–Protein Interaction (PPI) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed to identify the biological pathways of these genes. Animal Procedures and Treatments Male C57BL/6 mice (6–8 weeks old; 22 ± 3 g) were obtained from Tianqin Biotechnology (Changsha, China). All animals were housed under standard conditions (55%–60% humidity, 22 °C–24 °C, 12-h light/dark cycle) for 7 days before beginning the experiments. All procedures were approved by the Animal Ethics Committee of Guizhou Medical University (Approval No. 2503172). Mice were randomly assigned to five groups (n = 5–8 per group): olive oil, CCl4 + NS, CCl4 + quercetin, CCl4 + hUC-MSC-exo, and CCl4 + hUC-MSC-exo-Que. Liver fibrosis was induced by intraperitoneal injection (i.p.) of CCl4 (1.6 mL/kg, 15%–20% in olive oil) twice weekly for 6 weeks. After 2 weeks of CCl4 administration, the treatment groups received tail vein injections of quercetin (1 mg/kg), hUC-MSC-exo, or hUC-MSC-exo-Que (an equivalent amount in saline) once every 3 days. The olive oil and CCl4 + NS groups received saline alone. Following the final administration, the mice were weighed and anesthetized with sodium pentobarbital (50 mg/kg i.p.). Blood was collected via orbital puncture, and the livers were rapidly excised, frozen in liquid nitrogen and stored at −80 °C or in 10% paraformaldehyde for subsequent analysis. Serum alanine aminotransferase (ALT), and aspartate aminotransferase (AST) levels were measured. Statistical Analysis Statistical analyses were performed using GraphPad Prism 8.0 (San Diego, CA, USA). The Shapiro-Wilk test was used to assess normality, and Levene’s test was used to assess homogeneity of variances. The Studentʼs t-test was used for pairwise comparisons between the groups. A one-way analysis of variance (ANOVA) with Tukey’s post hoc test was also performed to compare multiple groups. P-values <0.05 were considered statistically significant. Data are presented as the mean ± SD. All experiments were conducted with at least three independent biological replicates unless otherwise stated. *p < 0.05; **p < 0.01; ***p < 0.001. Results and Discussion Isolation and Characterization of huc-MSCs Figure A shows the extraction process of hUC-MSCs. hUC-MSCs exhibited a characteristic adherent, fibroblast-like, spindle-shaped morphology (Figure B). Immunophenotyping demonstrated that the cells were positive for the hUC-MSC surface marker CD73 (Figure C) but negative for the hematopoietic markers CD31 and CD34 (Figure D–E). Isolation and characterization of hUC-MSC. (A) Schematic of hUC-MSC isolation. (B) Phase-contrast inverted light microscopy showing the morphological characteristics of hUC-MSCs (×40; scale bar represents 200 μm). (C–E) Flow cytometric analysis of cultured cells expressing the mesenchymal stem cell-positive markers CD73 (96%), CD31 (0%), and CD34 (0%) (n = 3). Isolation and Characterization of hUC-MSC-exo hUC-MSC-exo were isolated by ultracentrifugation (Figure A). Transmission electron microscopy (TEM) revealed spherical vesicles with a characteristic lipid bilayer (Figure B). The average particle size was approximately 82 ± 28.8 nm (Figure C), and the zeta potential was approximately −6.18 ± 3.415 mV (Figure D), consistent with reported exosomal characteristics. Western blot confirmed the presence of exospecific markers CD63, TSG101 and CD81, whereas calnexin, a negative marker was detected only in hUC-MSC cell lysates (Figure E). Ultracentrifugation, a widely used method for exosome isolation, was used to prepare standardized hUC-MSC-exo. Isolation and characterization of hUC-MSC-exo. (A) Schematic diagram of the isolation of hUC-MSC-exo. (B) TEM image of hUC-MSC-exo. (C) Particle size distribution of hUC-MSC-exo indicated using a Nanocoulter counter. (D) Zeta potential distribution of hUC-MSC-exo indicated using a Nanocoulter counter. (E) Western blot analysis of hUC-MSC-exo and hUC-MSC (n = 3). Functional Evaluation of Quercetin Treatment with 10 ng/mL TGF-β1 for 24 h significantly upregulated the mRNA levels of α-SMA and COL1A1 in LX-2 cells. , Consistent with these findings, TGF-β1 stimulation induced LX-2 activation in our study, as evidenced by increased expression of α-SMA and COL1A1 (Figure A–B) (p < 0.001). All primer sequences used in this study are listed in Table S1. Subsequently, varying concentrations of quercetin were applied to TGF-β1-induced LX-2 cells and were found to inhibit their activation (Figure C–D), consistent with previous reports. CCK-8 assays showed that quercetin at concentrations 20 μM exhibited cytotoxicity in LO2 cells (Figure E–F). Therefore, 10 μM quercetin was used in all subsequent in vitro experiments. In vitro functional evaluation of quercetin. (A–B) qRT-PCR analysis of α-SMA and COL1α1 mRNA expression in TGF-β1-induced LX-2 cells. (C–D) qRT-PCR analysis of α-SMA and COL1α1 mRNA expression in TGF-β1-induced LX-2 cells treated with varying concentrations of quercetin. (E–F) CCK-8 assay for the viability of LX-2 and LO2 cells treated with varying quercetin concentrations. Data are presented as the mean ± SD of three independent biological samples n = 3. *p < 0.05, **p < 0.01, ***p < 0.001. Evaluation of Four Loading Strategies for hUC-MSC-exo-Que Quercetin exhibits limited therapeutic efficacy due to poor solubility, short half-life, and low oral bioavailability. ,− Advanced delivery systems, particularly exosome-based delivery platforms, have emerged as a promising strategy to overcome these limitations. , Owing to their endosomal origin, nanoscale size, and lipid-rich bilayer, exosomes exhibit high physicochemical stability and prolonged circulation, making them attractive drug carriers. They can be efficiently loaded with miRNAs, anticancer drugs or other bioactive molecules, thereby increasing therapeutic efficacy and reducing systemic side effects. Compared with exosomes derived from other donor cells, MSC-derived exosomes have higher stability and lower immunogenicity, making them an optimal delivery vehicle. Drug loading into exosomes is generally achieved via endogenous or exogenous methods. The endogenous approach achieves drug loading through pretreatment of parent cells, during which free drugs are internalized and subsequently packaged into exosomes via the multivesicular bodies (MVBs) pathway. It has been reported that endogenous loading of quercetin into extracellular vesicles (EVs) can be obtained by pretreating bone marrow-derived mesenchymal stem cells with quercetin. Bone marrow-derived MSC-exo-Que are isolated after preconditioning with quercetin. In contrast, exogenous drug loading involves introducing drugs into preisolated exosomes through membrane permeabilization techniques. This approach is typically encompasses coincubation, sonication, electroporation, surfactant-assisted permeation, freeze–thaw cycles, and extrusion method. While endogenous loading methods are generally more complex, exogenous approaches are more widely used due to their practicality and simplicity. Here, we compared four exogenous methods: coincubation, electroporation, extrusion, and sonication. Co-incubation relies on passive diffusion and preserves exosomal integrity but is limited to specific drugs. Electroporation involves the application of an electric field to create transient membrane pores, facilitating drug diffusion. ,, Extrusion mixes exosomes and drug by passing through a microporous membrane (100–400 nm), enabling high loading efficiency and uniform particle size. , Sonication uses probe-generated pores to allow small hydrophilic molecules to enter. , As shown in Figure A–B, the sonication and extrusion methods yielded particle sizes (80 ± 24.2 and 82 ± 21.0 nm, respectively) closest to native exosomes (82 ± 28.8 nm). Quercetin exhibited a characteristic absorbance peak at 370 nm (Figure S1), and a standard curve was established (Figure S2). Evaluation of the effects of different treatment regimens on the viability of LO2 cells revealed that quercetin alone, MSC-exo, and the four drug-loading systems did not significantly affect cell viability (Figure C). Consistent with previous reports, quercetin alone did not appear to promote LO2 cell proliferation. Based on our experimental results, 10 μM was selected as the working concentration, at which free quercetin neither promoted nor inhibited LO2 cell proliferation. Regarding the exosome vehicle, preclinical studies have demonstrated that MSC-exos promote liver regeneration, including enhanced recovery after partial hepatectomy in rats. However, one study found that under normal conditions, MSC-exos (0–10 μg/mL) did not induce proliferation in TAMH, THLE-2, or HuH-7 cells, but at 0.1 μg/mL provided protection against injury. Similarly, stem cells derived from human exfoliated deciduous teeth (SHED)-exo showed no effect on AML12 cells at 0–20 μg/mL within 24 h, but promoted proliferation at 2 μg/mL after 48 h. Thus, the absence of a proliferative response with MSC-exo-Que is consistent with the observation that neither free quercetin (10 μM) nor empty MSC-exosomes (3 μg/mL) directly promote healthy LO2 cell proliferation under normal conditions. The pro-regenerative function of MSC-exosomes may require longer exposure (≥48 h) or an injured microenvironment. Therefore, the absence of effect observed in our experiment is likely attributable to short incubation time (24 h) and lack of a stressed microenvironment, rather than inefficacy of MSC-exo-Que. Further studies using injury models and extended observation periods are warranted. hUC-MSC-exo loaded with quercetin. (A–B) Particle size and zeta potential measurements after drug loading by coincubation, electroporation, extrusion and sonication. (C) CCK-8 assay of LO2 cell viability under different treatment conditions. (D) Encapsulation efficiency after drug loading by coincubation, electroporation, extrusion and sonication. (E–F) Particle size and zeta potential of hUC-MSC-exo-Que complexes at varying loading ratios via extrusion method. (G–H) qRT-PCR analysis of α-SMA and COL1α1 mRNA expression in TGF-β1-activated LX-2 cells treated with hUC-MSC-exo-Que by extrusion complexes at different ratios. (I) Rate of quercetin release after encapsulation by extrusion. (J–K) Changes in the particle size and zeta potential of extrusion-loaded quercetin at different storage temperatures over time (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001. Encapsulation efficiency analysis revealed that extrusion achieved the highest efficiency (55.8%) (Figure D), likely due to enhanced membrane disruption and mixing during processing, which led to a higher drug-loading capacity. Accordingly, extrusion was identified as the optimal loading method, consistent with a previous report. However, this approach is widely used for initial assessment of loading efficiency, we acknowledge that it cannot definitively distinguish between drug truly encapsulated within the exosome lumen and drug merely adsorbed onto the exosome surface. This represents a recognized limitation in the field, therefore, the reported encapsulation efficiency (55.8%) likely includes both populations. Future studies are warranted to quantitatively differentiate between these populations. Quercetin was loaded into hUC-MSC-exo at various ratios using the extrusion method. Particle size and zeta potential remained comparable among different drug-loading ratios (Figure E–F). However, compared to free quercetin, the 1:1 formulation demonstrated the strongest inhibitory effect, resulting a 5-fold reduction in α-SMA mRNA (p < 0.001, Figure G) and a 2-fold reduction in COL1A1 mRNA (p < 0.001, Figure H). Although hUC-MSC-exo alone (100–200 μg/mL) has been reported to suppress LX-2 activation, increasing exosome concentration in our system (3–9 μg/mL) did not enhance the inhibitory effect. This finding suggests that quercetin is the primary contributor to the observed antifibrotic activity, and that at the low exosome concentrations used (3–9 μg/mL), the exosomes likely function mainly as carriers rather than exerting intrinsic inhibitory activity. The drug release profile of the optimized 1:1 hUC-MSC-exo-Que formulation demonstrated sustained release pattern up to 48 h (Figure I). Storage stability studies showed no significant changes in particle size or zeta potential over 7 days at 4, −20, or −80 °C (Figure J–K). Liver-Targeting Delivery and HSC Uptake of hUC-MSC-exo-Que To evaluate in vivo biodistribution, DiR-labeled hUC-MSC-exo-Que was administered via tail vein injection. After 24 h, fluorescence signals were primarily detected in the liver, followed by the spleen (Figure A), consistent with previously reported exosome distribution patterns of DiR-labeled exosomes. These findings indicate that the exosomal carrier retains its intrinsic homing capability, facilitating hepatic targeting of quercetin. Cellular uptake was assessed using confocal microscopy. DiO-labeled hUC-MSC-exo-Que (green) was observed within LX-2 cells (DAPI-stained nuclei, blue) after 24 h (Figure B), consistent with previous reports. , This demonstrates that loading quercetin into hUC-MSC-exo preserves the natural tropism of LX-2 cells and enables effective intracellular drug delivery. In vivo biodistribution and cellular uptake of hUC-MSC-exo-Que. (A) In vivo distribution of hUC-MSC-exo-Que detected by a live small animal imaging system. (B) Cellular internalization of hUC-MSC-exo-Que by LX-2 cells was visualized using confocal microscopy. Scale bars = 20 μm. Representative fluorescence images from three independent experiments are shown. hUC-MSC-exo-Que Enhances Quercetin’s Inhibitory Effect on LX-2 Cell Activation Immunofluorescence and quantitative analysis showed that hUC-MSC-exo-Que suppressed α-SMA expression more effectively than the same dose of free quercetin (p < 0.05, Figure A). Consistently, qRT-PCR revealed significant downregulation of α-SMA (3.16-fold, p < 0.001) and COL1A1 mRNA (2.08-fold, p < 0.001) (Figure B–C), and Western blot confirmed reduced α-SMA protein levels (1.22-fold, p < 0.05) (Figure D). All differences were statistically significant, indicating that exosome-loaded quercetin exhibits enhanced in vitro anti-HSC activation efficacy. These findings align with previous reports that quercetin-loaded exosomes increased the potency of doxorubicin-treated cells, demonstrating the potential of exosomes in improving drug bioavailability in vitro. In vitro functional evaluation of quercetin encapsulated by extrusion. (A) Representative fluorescence images (bar = 50 μm) and quantitative analysis of fluorescence intensity. (B–C) qRT-PCR analysis of α-SMA mRNA and COL1A1 mRNA expression suppression by encapsulated versus free quercetin treatment. (D) Western blot analysis of α-SMA protein expression suppression by encapsulated versus free quercetin n = 3. *p < 0.05, **p < 0.01, ***p < 0.001. hUC-MSC-exo-Que Inhibits LX-2 Activation: Potential Involvement of PI3K/Akt Signaling in Part Quercetin targets and liver fibrosis-related genes were intersected, yielding 80 common genes after deduplication (Figure A). PPI analysis of these genes was performed (Figure B, R = 0.9), followed by KEGG pathway enrichment analysis of 50 core genes from the PPI network. The results highlighted enrichment of the PI3K/Akt signaling pathway (Figure C), with a well-established role in alleviating liver fibrosis. , Activation of this pathway involves phosphorylation of (p-Akt), and functions as a key effector of PI3K signaling. To validate this, we assessed Akt and p-Akt levels using Western blotting. TGF-β1 stimulation activated the PI3K/Akt pathway in vitro, which was markedly suppressed by hUC-MSC-exo-Que treatment. Notably, hUC-MSC-exo-Que exerted a stronger inhibitory effect than free quercetin alone (Figure D). These findings suggest that the superior suppression of the PI3K/Akt pathway may contribute to the improved antifibrotic efficacy of the exosome-based delivery system. hUC-MSC-exo-Que inhibits LX-2 activation via the PI3K/Akt signaling pathway. (A) Venn diagram of genes closely related to LF and targets for quercetin. (B) PPI network of the overlapping DEGs (R = 0.9). (C) KEGG pathway enrichment analysis of the hub genes identified from the PPI network. (D) Western blotting of the protein expression levels of p-Akt, Akt and α-SMA. hUC-MSC-exo-Que Reduces CCl4-Induced Liver Fibrosis in Mice The hemolytic potential of Que, MSC-exo, and MSC-exo-Que was assessed using human and mouse whole blood as well as washed erythrocytes. As illustrated in Figure A, incubation of Que, MSC-exo, and MSC-exo-Que with human and mouse whole blood and washed erythrocytes at 37 °C for 60 min, the absorbance of the supernatant was measured at 405 nm for washed erythrocyte samples and at 570 nm for whole blood samples, relative to the negative control (PBS, 0% hemolysis) and positive control (10% Triton X-100, 100% hemolysis), all treatment groups exhibited hemolysis rates below 1% (Tables and ), indicating that Que, MSC-exo, and MSC-exo-Que have good biocompatibility. In vivo functional assessment of hUC-MSC-exo-Que. (A) Hemolysis of human and mouse whole blood and washed erythrocytes by Que, MSC-exo, and MSC-exo-Que (n = 3). (B) Representative macroscopic morphology of the liver and photographs of H&E and Sirius Red staining images. The images are representative of three independent experiments (scale bar = 100 μm). (C–D) Serum AST and ALT levels (n = 5–8). (E) Liver/Body Weight Index (n = 5–8). (F–G) qRT-PCR analysis of α-SMA mRNA and COL1A1 mRNA expression suppression by encapsulated versus free quercetin (n = 3). (H) Western blot analysis of α-SMA protein expression suppression by encapsulated versus free quercetin in the liver (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001. 1Calculated Hemolysis Ratios (in %) When Using Mouse and Human Erythrocyte Solution SpeciesQueMSC-exoMSC-exo-QueHuman0 ± 0.120.47 ± 0.070.50 ± 0.12Mouse0.35 ± 0.130.09 ± 0.040.14 ± 0.04aNormalization was performed on data averaged from three experimental replicates using 10% Triton X-100 and PBS samples as positive and negative controls, respectively. Calculated Hemolysis Ratios (in %) When Using Mouse and Human Erythrocyte Solution Normalization was performed on data averaged from three experimental replicates using 10% Triton X-100 and PBS samples as positive and negative controls, respectively. 2Calculated Hemolysis Ratios (in %) When Using Mouse and Human Whole Blood Solution SpeciesQueMSC-exoMSC-exo-QueHuman0 ± 0.030 ± 0.010 ± 0.03Mouse0.45 ± 0.350.30 ± 0.080.59 ± 0.14aNormalization was performed on data averaged from three experimental replicates using 10% Triton X-100 and PBS samples as positive and negative controls, respectively. Calculated Hemolysis Ratios (in %) When Using Mouse and Human Whole Blood Solution Normalization was performed on data averaged from three experimental replicates using 10% Triton X-100 and PBS samples as positive and negative controls, respectively. Repeated CCl4 administration is a well-established model for inducing liver fibrosis in vivo. Therefore, we used a CCl4-induced liver fibrosis model to evaluate the efficacy of hUC-MSC-exo-Que. As shown in Figure B, hUC-MSC-exo-Que markedly reduced CCl4-induced necrosis, inflammatory infiltration, and liver fibrosis compared to free quercetin. Serum aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels were significantly decreased following hUC-MSC-exo-Que treatment than free quercetin (1.63-fold, p < 0.05; 1.50-fold, p < 0.01) (Figure C–D). The hUC-MSC-exo-Que-treated group showed a lower liver index than the CCl4 + NS model group; however, the difference was not significant compared with the group treated with an equivalent dose of quercetin (Figure E). Notably, hUC-MSC-exo-Que treatment led to a significantly greater reduction in α-SMA (1.63-fold, p < 0.001) and COL1A1 (1.17-fold, p < 0.01) mRNA expression in liver tissue (Figure F–G). This antifibrotic effect was further confirmed at the protein level by Western blot, which showed a significant reduction in α-SMA expression (1.63-fold, p < 0.05) (Figure H), consistent with the observed transcriptional changes. These results demonstrate that hUC-MSC-exo-Que significantly ameliorates liver injury and fibrosis in CCl4-induced mice, with superior efficacy compared with free quercetin. This enhanced therapeutic effect aligns with previous reports showing that exosome-delivered quercetin exhibits improved potency in Alzheimer’s disease. Conclusions In this study, quercetin-loaded exosomes (hUC-MSC-exo-Que) were prepared and evaluated in a CCl4-induced mouse model of liver fibrosis. The engineered hUC-MSC-exo-Que nanosystem offers several advantages, including enhanced tissue penetration, sustained drug release, a favorable safety profile, and good biocompatibility. Our findings demonstrate that hUC-MSC-exo-Que alleviates liver fibrosis by inhibiting the activation of LX-2 cells. We acknowledge that exosome-based delivery of natural compounds has been previously reported; however, our study provides a preliminary evaluation of an optimized loading method (extrusion) in the specific context of liver fibrosis. Nevertheless, several limitations should be acknowledged. First, the mechanistic evidence presented is primarily correlative, and further studies are needed to establish causality. Second, although the CCl4-induced mouse model is a well-established tool, it does not fully recapitulate the complex and multifactorial etiology of human liver fibrosis. Third, the reliance on a single cell line (LX-2) and a limited set of markers (α-SMA and COL1A1) may restrict the generalizability of our findings. Despite these limitations, the hUC-MSC-exo-Que system shows potential as a promising approach for liver fibrosis treatment, although further validation is needed. Supplementary Material