Efficacy of extracellular vesicles derived from mesenchymal stromal cells in regulating senescence: In vitro and in vivo insights

Abstract

BACKGROUND

Extracellular vesicles derived from mesenchymal stromal cells (MSC-EVs) can be used for anti-aging therapy and treating various aging-related diseases. However, the clinical application of MSC-EVs is still limited, mainly due to insufficient information on the preparation process, quality, and mechanism of action of MSC-EVs.

AIM

To study the biological effects of MSC-EVs in regulating cellular senescence.

METHODS

In this study, we developed a clinical-grade production process for MSC-EVs and defined the release criteria for products suitable for human use. To support the clinical use of our product as a therapeutic agent, we performed efficacy assays to evaluate the anti-aging capacity of MSC-EVs in vitro and in vivo.

RESULTS

The functional analysis results revealed that MSC-EVs significantly reduced the levels of senescence-associated β-galactosidase, matrix metallopeptidase 1, P21, and interleukin-1β and increased the level of collagen I in a naturally aged cell model of human dermal fibroblasts. Similarly, treatment with MSC-EVs effectively improved D-gal-induced subacute aging in mice, aging-related histopathological changes, oxidative stress, and aging-related gene expression.

CONCLUSION

These findings indicate that MSC-EVs can partially alleviate D-gal-induced senescence by reducing oxidative stress and regulating metabolism. Overall, these findings strongly suggest that MSC-EVs hold promise for aging therapy.

Key Words: Senescence; Anti-aging; Mesenchymal stromal cells; Extracellular vesicles; Transcriptomics; Metabolomics

Core Tip: This study addresses the practical challenges of using mesenchymal stromal cell-derived extracellular vesicles (EVs) for anti-aging therapy. We developed a scalable production method and quality control standards for clinical use. Tests in cells and animals have shown that these EVs effectively decrease the expression of aging markers, repair age-related damage, reduce oxidative stress, and normalize gene activity. The key mechanisms involved in fighting oxidative stress and fine-tuning metabolism. These findings position EVs as promising, ready-to-use anti-aging treatments.

INTRODUCTION

Aging is a dynamic and complex process characterized by loss of homeostasis in tissues and decreased regenerative capacity[1]. The population of people over 65 will exceed 1.6 billion by the middle of the century[2]. With an increase in the size of the elderly population, the demand for healthcare will increase substantially, which will challenge the ability of the existing healthcare system to manage such patients efficiently[3,4]. The field of senescence has also encouraged scientific research institutions and enterprises to put more effort into anti-aging and disease research and develop more effective treatment approaches to improve the quality of life.

For decades, researchers have investigated the mechanisms of aging and limited lifespan and elucidated the mechanisms underlying the progression of aging and the onset of age-associated pathologies[5-7]. Researchers have proposed several anti-aging interventions, including drug therapy, caloric restriction[8,9], and the elimination of senescent cells[10]. However, the efficacy of these approaches in delaying senescence or ameliorating senescence-associated pathologies is suboptimal. López-Otín et al[7] reviewed 12 tentative hallmarks that serve as universal attributes of aging across various species, focusing on the relevance of cellular senescence to the aging process, pointing to stem cell depletion as a key mechanism contributing to senescence[11]. Thus, stem cell-based therapy, especially treatment with mesenchymal stromal cells (MSCs), has become an innovative anti-aging approach[12,13]. A phase I/II double-blind and placebo-controlled study showed that the application of intravenous exogenous allogenic MSCs can reverse the symptoms of frailty in elderly individuals, significantly improving quality of life, physical performance, and reducing chronic inflammation[14]. However, using MSCs in therapeutic applications poses several challenges, including the risk of cellular rejection, tumorigenesis, and problems related to cell delivery and engraftment. These concerns have led researchers to assess alternative strategies for using MSCs for treatment while mitigating the risks related to their application. One such promising strategy involves using extracellular vesicles (EVs) derived from MSCs (MSC-EVs).

The cargo of MSC-EVs consists of various cytokines, growth factors, bioactive lipids, and regulatory microRNAs (miRNAs) that can participate in cell-to-cell communication and cell signaling and alter the metabolism of cells or tissues at short or long distances in vivo. A recently published study provided further insights into applying MSC-EVs as powerful cell-free regenerative medicine[15]. These vesicles have the therapeutic ability of MSCs and can influence tissue response to injury, infection, and disease[16]. Lei et al[17] showed that EVs derived from umbilical cord-derived MSCs (UC-MSCs) can delay the aging of naturally aged mice throughout the body and significantly alter the degenerative functions of various tissues and organs. Many preclinical studies have shown that multiple sources of EVs[18-23], especially those derived from UC-MSCs, are prospective cell-free therapeutic agents for aging therapy. However, key parameters, including quality, reproducibility, and potency, determine the development of therapies based on EVs. Large-scale production of EVs faces multiple challenges, including low yield, heterogeneity, targeted delivery, storage stability, and the lack of standardized protocols to ensure quality, safety, and consistency. Current isolation techniques, such as ultracentrifugation and density gradient methods, suffer from limited yield and insufficient purity, making them inadequate for clinical-scale applications. Further research and optimization of exosome isolation, characterization, and functional evaluation strategies are essential to advance their clinical translation[24,25]. Standardized production processes and rigorous quality control will be critical to achieving this goal.

This study established a highly efficient technique for extracting and characterizing MSC-EVs. Additionally, we identified and implemented crucial quality control checkpoints for MSC-EVs. These measures were taken to ensure consistent yield, quality, and reproducibility of the MSC-EVs, rendering them suitable for clinical use. Next, we conducted several experiments to determine the effects of MSC-EVs on senescence in senescent cells and aged murine models. We found that MSC-EVs inhibited the aging-related secretory phenotype at the cellular level and reduced the attenuation of age-associated degenerative changes in multiple organs. Moreover, integrated metabolomics and transcriptomics analyses were performed, and the results confirmed the anti-aging mechanism of MSC-EVs. Our study highlighted the critical factors to consider when developing clinical intervention strategies using MSC-EVs for treating age-related conditions.

MATERIALS AND METHODS

Establishment and identification of the UC-MSC bank

Clinical-grade EVs were generated exclusively from UC-MSCs isolated and cultured in a current good manufacturing practice (cGMP) facility. MSCs were collected from the umbilical cord of healthy human donors after informed consent was provided, and the study received ethical clearance. Briefly, Wharton’s jelly isolated from the umbilical cord was homogenized and centrifuged to separate the primary cells (passage 0). Next, a master cell bank (passage 2) was established in the second generation (passage 2), and a post-production cell bank (PPCB) was established in the fourth generation (passage 4). The identity, safety, and functional activity of UC-MSCs in master cell bank and PPCB were extensively evaluated and verified following the methods described in other studies[26,27].

UC-MSCs culture and conditioned medium collection

UC-MSCs frozen from PPCB were thawed and seeded at a density of 8000 cells/cm² in the cell factory system, propagated in basal medium formulated explicitly for MSCs, and augmented with a serum-free supplement (Yocon, China) at 37 °C under 5% CO₂. For the cell culture experiments, the basal medium, serum-free replacement medium, and phosphate buffered saline (PBS) were subjected to overnight ultracentrifugation at 160000 × g to remove vesicle-like particles before application. When the cell confluence reached 80%-90%, the culture medium was aspirated and cryopreserved at -80 °C for subsequent experiments. Each aliquot of the conditioned cell culture supernatant underwent rigorous sterility testing; the samples yielded negative results for anaerobic and aerobic bacterial contamination and were validated for low endotoxin levels (< 1 endotoxin unit per milliliter), as was the absence of mycoplasma contamination via polymerase chain reaction (PCR) analysis. The individual harvests were gently thawed at 4 °C overnight, pooled, and filtered through a filter membrane (5 μm pore size) to remove large vesicles and particles.

Enrichment and identification of EVs

After filtering, the conditioned medium underwent tangential flow filtration (TFF) using a hollow fiber membrane with a molecular weight cutoff of 100 kDa, resulting in a 20-fold concentration enhancement[27]. The medium was subsequently centrifuged at 300 × g for 30 minutes at 4 °C, after which the supernatant was centrifuged at 16500 × g for 30 minutes at 4 °C to remove cellular debris. Next, the clear fraction was transferred to fresh centrifuge tubes and ultracentrifuged at 100000 × g for 70 minutes at 4 °C, facilitating the sedimentation of EVs[28]. The EVs were subsequently solubilized in PBS and stored at -80 °C for further study. The EVs were subsequently identified via high-sensitivity flow cytometry (nano flow cytometry), transmission electron microscopy, and western blotting following protocols described in another study[29].

Cell proliferation assay

Human dermal fibroblasts (HDFs, Pricella, China) were cultured in DMEM supplemented with 10% (v/v) fetal bovine serum and penicillin/streptomycin, and the culture mixture was incubated at 37 °C in an atmosphere containing 5% CO₂. To induce natural senescence, HDFs were subcultured for more than 30 passages. Cell proliferation was measured via a WST-8 Cell Counting Kit-8 (Beyotime, China) assay. Briefly, 100 μL of the cell suspension (2000 cells/well) was added to 96-well plates. After cell adhesion, different concentrations of MSC-EVs ranging from (0.001-100) × 10⁸ particles/mL were added. After incubation for 72 hours, 10 μL of WST-8 solution was added to each well, followed by further incubation for 2 hours. Subsequently, the optical density was measured at 450 nm via a microplate reader.

Migration assay

Transwell migration and scratch wound assays were performed to investigate the effects of MSC-EVs on cell migration. HDFs were seeded in the upper compartments of Transwell inserts (Corning, NY, United States) and incubated for 24 hours in complete DMEM. The medium was subsequently replaced with foetal bovine serum-depleted DMEM containing MSC-EVs at a concentration of 1 × 10⁸ particles/mL, and the cells were incubated for 24 hours. The cells that migrated to the underside were preserved with 4% paraformaldehyde and stained with 0.5% crystal violet for visual inspection and enumeration via a light microscope. To conduct the scratch wound assay, HDFs were cultured in 24-well plates in complete DMEM until the cells reached 90% confluence. A sterile pipette tip was used to scratch the cell monolayer. The cells were then incubated in DMEM (without foetal bovine serum) supplemented with MSC-EVs (1 × 10⁸ particles/mL) at 37 °C. Images of the scratch were captured at 0, 6, and 22 hours, and the images were quantitatively analyzed using ImageJ software. For the control condition, HDFs were incubated with a comparable volume of PBS instead of MSC-EVs.

Detection of senescence-associated β-galactosidase activity

The HDFs (passage 30) were plated in six-well culture dishes at a seeding density of 5000 cells/cm² and grown in complete DMEM. The wells assigned to the experimental group received MSC-EVs at 1 × 10⁸ particles, whereas those assigned to the control group received an equivalent volume of PBS. When 90% confluence was reached, the assay for senescence-associated β-galactosidase (SA-β-gal) activity was conducted following the manufacturer’s protocol (Beyotime, China). For comparison, younger HDFs (passage 5) were treated identically and served as an additional control.

RNA extraction and real-time quantitative PCR analysis

HDFs were grown in a 24-well plate at a seeding density of 6000 cells/cm². Next, EVs (1 × 10⁸ particles) were added to the wells, and the cells were incubated at 37 °C for another 48 hours. We assessed the expression levels of several genes, including matrix metallopeptidase 1 (MMP-1), collagen type I (COL-1), P21, interleukin-1β (IL-1β), and GAPDH, across various samples via real-time quantitative PCR (qPCR) (Applied Biosystems, CA, United States). The sequences of the primers used in this study are listed in Supplementary Table 1.

Visualization of the distribution of MSC-EVs

An imaging system was used to determine the biodistribution of MSC-EVs. To obtain DiD-labeled MSC-EVs (DiD-EVs), 1 × 10¹¹ particles/mL were incubated with 10 μmol/L DiD (Yeasen, China) according to the manufacturer’s instructions. In the control setup, PBS was incubated with 10 μmol/L DiD (Yeasen, China). Nano flow cytometry analysis was subsequently conducted to assess the labeling efficiency. Then, 2 × 10⁹ DiD-EVs were administered intravenously to BALB/c nude mice, whereas PBS was injected into the mice following the same protocol. The mice were anesthetized with 1% sodium phenobarbital at 4, 24, 48, and 72 hours to determine which tissue the signal originated from. The heart, liver, kidney, lung, spleen, and brain were collected and imaged with an IVIS spectral imaging system (Perkin Elmer, MA, United States).

Animals and treatment

Juvenile BALB/c mice (n = 24; weight: 20 ± 2 g) were obtained from Vital River Laboratory Animal Technology Co., Ltd., Beijing. The Scientific Ethics Special Committee of the Academic Committee of Ocean University of China approved the protocols, adhering to international standards for animal research. After the mice were acclimated to the laboratory environment, a subacute aging mouse model was induced by daily intraperitoneal injection of D-galactose (D-gal, 500 mg/kg, Sigma, Japan) for 10 weeks. For the control groups, PBS was processed similarly. Four weeks after continuous intraperitoneal injection of D-gal, the mice were divided into one of three groups, each comprising eight individuals: A control group, a D-gal-induced aging model group, and a group treated with MSC-EVs. The mice in the MSC-EV group were administered MSC-EVs once every 3 days via the tail vein at a dose of 5 × 10¹⁰ particles/kg. The mice were administered PBS once every 3 days via the tail vein for the D-gal-induced aging model group and the control group. The injection volume was the same as that of the MSC-EV group. After six weeks of treatment with MSC-EVs, the mice were euthanized by the combined method. First, the mice were anesthetized. The injection dose was 0.2 mL per 10 g of body weight, and tribromoethanol (purchased from Nanjing Aibei Biotechnology Co., Ltd., China) was injected intraperitoneally. After the anesthetic injection was complete, the mice were then sacrificed by cervical dislocation. After the sacrifice operation, the toes of the mice were pressed by hand or with tweezers. If there was no response from the mice, they could be confirmed dead. Finally, liver tissue samples were harvested for subsequent hematoxylin and eosin staining. The work has been reported in line with the ARRIVE guidelines 2.0.

Measurement of the malondialdehyde content and oxidation-associated factors in the mouse liver

The levels of malondialdehyde (MDA), superoxide dismutase (SOD), and catalase (CAT) in the mouse liver were determined using relevant kits. Briefly, hepatic samples (approximately 50 mg) were homogenized in 0.5 mL of ice-cold PBS via an ice-chilled homogenizer. The resulting homogenate was centrifuged at 5000 × g for 15 minutes at 4 °C to obtain the supernatant. The MDA content, which indicates lipid peroxidation levels, was quantitatively assessed using an MDA assay kit (Beyotime, China). The enzymatic activities of SOD and CAT were quantified by the chemiluminescence method. Commercial assay kits for SOD and MDA were obtained from Nanjing Jiancheng Bioengineering Institute.

Detection of age-related and oxidation-associated genes in murine hepatic tissue

Total RNA was isolated from hepatic tissue and subsequently reverse-transcribed. Next, the gene expression profiles of age-related genes, such as P21 and IL-1β, and oxidation-associated genes, including Cu/Zn-SOD, Mn-SOD, and CAT, were quantitatively analyzed across the different groups via qPCR analysis. Quantitative data analysis was conducted by the 2^-∆∆Ct method, and the 18S rRNA expression levels were used for normalization. The sequences of the primers used for the analysis are listed in Supplementary Table 1.

Metabolomic and transcriptomic analyses of the mouse liver

In this study, liver samples from the control, D-gal-induced aging, and MSC-EV-treated groups (16 samples in total) were used for metabolomic and transcriptomic analyses to gain deeper insight into the anti-senescence properties of MSC-EVs. Liver samples were collected and randomly placed in one of two subsets: Training data and validation data. Untargeted metabolomics was performed to identify EV-related metabolic signatures. Transcriptome sequencing was performed to analyze the effects of MSC-EVs on the liver transcriptome of D-gal-induced aging mice. The metabolomics and transcriptomics analyses were performed using BMKCloud (http://www.biocloud.net). Transcriptome and metabolome data were analyzed by principal component analysis (PCA), orthogonal partial least squares discriminant analysis (OPLS-DA) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, respectively.

After normalization of the original metabolome data, follow-up analysis was carried out. PCA and Spearman correlation analysis were used to judge the repeatability of the samples within groups and the quality of the control samples. The identified compounds were searched for classification and pathway information in the KEGG, Human Metabolome Database, and LipidMaps databases. According to the grouping information, the difference multiples were calculated and compared, and a t-test was used to calculate the significance P-value of each compound. The R language package tools was used to perform OPLS-DA modeling, and 200 permutation tests were performed to verify the reliability of the model. The VIP value of the model was calculated via multiple cross-validations. The method of combining the difference multiple, the P-value and the VIP value of the OPLS-DA model was adopted to screen the differentially abundant metabolites. The screening criteria were fold change > 1, P-value < 0.01 and VIP > 1. The differentially abundant metabolites associated with KEGG pathway enrichment significance were calculated using a hypergeometric distribution test. For transcriptome analysis, the bioinformatics analysis platform BMKCloud was used for the following tasks: Quality control, sequence comparison, gene function annotation, gene expression quantification, and differential expression analysis. In this study, differential expression analysis of two conditions/groups (DESeq2) was used to analyze the differential expression of the two groups. DESeq2 provides statistical routines for determining differential expression in digital gene expression data via a model based on the negative binomial distribution. The resulting P-values were adjusted using Benjamini and Hochberg’s approach to control the false discovery rate. Genes with an adjusted P-value < 0.01 and a fold change ≥ 2 according to DESeq2 were considered differentially expressed. KEGG pathway enrichment analysis of the above differentially expressed genes was subsequently performed. Finally, Pearson correlation coefficients were calculated to integrate the metabolomics and transcriptomics data, and correlation coefficients with R² values greater than 0.9 were selected. The relationship between metabolomics and transcriptomics was visualized via Cytoscape (version 2.8.2).

Statistical analysis

All the assays were performed in triplicate, with at least three independent repeats. The data are presented as the mean ± SD. Differences with statistical significance were defined by an independent samples t-test or one-way analysis of variance (ANOVA) by SPSS 18.0 software. All the results were considered statistically significant at P < 0.05.

RESULTS

Key quality assurance checkpoints for clinical-grade MSC-EVs

To transition from preclinical results to clinical application, we selected the production of EVs from UC-MSCs that adhere to cGMP standards. Most UC-MSCs have fibroblast-like morphological characteristics. In contrast, a few appear to be spindle-shaped or irregularly triangular (Figure 1A). These cells can differentiate into adipogenic, osteogenic, and chondrogenic lineages (Figure 1B-D) and are positive for mesenchymal markers (Figure 1E).

Figure 1 The characteristics of umbilical cord-derived mesenchymal stromal cells and derived extracellular vesicles derived from mesenchymal stromal cells. A: Morphological examination of umbilical cord-derived mesenchymal stromal cells (UC-MSCs) (scale bar indicates 100 μm); B-D: Multilineage differentiation of UC-MSCs. B: Osteogenic differentiation characterized by calcium phosphate deposits, visualized by Alizarin Red staining (scale bar indicates 100 μm); C: Adipogenic differentiation confirmed by the accumulation of lipid vacuoles, highlighted by Oil Red O staining (scale bar indicates 20 μm); D: Chondrogenic differentiation indicated by the synthesis of proteoglycans stained with Alcian Blue (scale bar indicates 100 μm); E: Surface marker expression profile of UC-MSCs. These cells were positive for the mesenchymal markers CD73 (98.0%), CD105 (97.3%), CD44 (95.7%), and CD90 (98.6%) but negative for CD45 (0.041%), CD34 (0.12%), CD11b (0.0%), and HLA-DR (0.031%); F: The ultrastructure of extracellular vesicles (EVs) derived from mesenchymal stromal cells (MSC-EVs) was examined by transmission electron microscopy (the scale bar indicates 100 nm); G: Immunoblotting was performed for the biomarkers of UC-MSCs and MSC-EVs. Full-length blots are presented in Supplementary Figure 4; H: The histogram illustrates the particle size distribution of MSC-EVs. To evaluate the concentration and dimensional spread of the particles, the suspension containing EVs was diluted at a 1:1000 ratio with phosphate buffered saline; I: The CD9, CD63, and CD81 proteins expressed on the surface of MSC-EVs were detected by nano flow cytometry. For immunofluorescence staining, fluorescein isothiocyanate-labeled antibodies against CD9, CD63, or CD81 were added to the EV suspension. All antibodies were purchased from BD Biosciences; J: Comparative particle quantification of MSC-EVs was performed before and after treatment with Triton X-100. To assess purity, a 5 μL aliquot of 10% Triton X-100 (purchased from Sigma-Aldrich, Japan) was mixed with a 45 μL suspension of EVs. The nano flow cytometry platform was used to analyze the characteristic side scatter burst profiles of the EVs. Purity = (pre-disruption particle count - post-disruption particle count)/pre-disruption particle count × 100%, higher values indicate greater purity. MSCs: Mesenchymal stromal cells; EVs: Extracellular vesicles; TSG101: Tumor susceptibility gene 101.

TFF and ultracentrifugation enriched the EVs, and the number of EVs ranged between 5000 billion and 9000 billion per harvest. Most EVs obtained via our protocol had saucer-like or cup-like structures (Figure 1F). Western blotting analysis confirmed that our particles expressed small EV markers (CD9, CD81, and tumor susceptibility gene 101) but did not express GM130 (Figure 1G). The median diameter of the MSC-EVs was 73.20 nm. In comparison, the average particle size was 80.30 nm (Figure 1H). The nano flow cytometry results indicated that the percentages of EVs expressing CD9, CD63, and CD81 were 60.5%, 14.1%, and 19.8%, respectively (Figure 1I). Overall, the presence of CD9, CD81, and tumor susceptibility gene 101 in the MSC-EV samples confirmed the vesicular nature of the isolated particles, whereas the absence of GM130 suggested a low level of cellular impurities. Furthermore, the purity of the EV particles was also assessed using Triton X-100 treatment. Triton X-100 is a non-ionic surfactant that can lyse the phospholipid membranes of EVs without affecting impurity protein particles. In this study, nano flow cytometry was used to detect changes in the quantity of EV particles before and after treatment with Triton X-100. The purity of the EVs reached 79% (Figure 1J). Sterility assays of the MSC-EVs confirmed that anaerobic and aerobic bacterial contamination was absent. Endotoxin levels were below 0.5 EU/mL, and the results of the qPCR analyses confirmed that mycoplasma was absent.

MSC-EVs ameliorated senescence in vitro

Cellular senescence was initially considered to cause cells to exit the cell cycle due to the limited proliferative capacity of cultured HDFs[30,31]. With increasing cellular senescence, the proliferative potential and migratory ability of fibroblasts decrease[31]. In this study, we first evaluated the influence of exosomal interventions on the proliferation and migration of HDFs. The results of the assay indicated that MSC-EVs enhanced the proliferation of HDFs in a concentration-dependent manner, peaking at 1 × 10⁸ particles/mL, above which the effect plateaued (Figure 2A). The results of the Transwell migration assay and scratch wound healing assay indicated that MSC-EVs significantly promoted HDF migration at a concentration of 1 × 10⁸ particles/mL (Figure 2B and C). Next, the potency of the EVs was tested by using a naturally senescent HDF model, which is considered a classical senescent cell model. The expression of SA-β-gal, a widely recognized marker of cellular senescence, was evaluated across different experimental groups. At passage 30, senescent HDFs presented increased expression of SA-β-gal compared with their younger counterparts at passage 5 (Figure 2D). Real time qPCR was performed to examine the effects of MSC-EVs on the expression of the age-related genes P21 and IL-1β in passage 30 HDFs. As shown in Figure 2, treatment with MSC-EVs significantly decreased the P21 (Figure 2E) and IL-1β (Figure 2F) transcript levels in senescent HDFs. Additionally, the expression of matrix-degrading enzymes (MMP-1) was lower in HDFs treated with MSC-EVs than in HDFs from the untreated group (Figure 2G). In contrast, the expression of COL-1, a key gene responsible for the tensile strength of skin tissue, was greater in HDFs treated with MSC-EVs. These results indicated that MSC-EVs can restore disturbances in mRNA expression in naturally senescent dermal fibroblasts (Figure 2H).

Figure 2 Extracellular vesicle derived from mesenchymal stromal cell treatment rejuvenated human dermal fibroblasts. A: The results of the Cell Counting Kit-8 assay demonstrated the proliferation of human dermal fibroblasts (HDFs) after treatment with extracellular vesicles derived from mesenchymal stromal cells (MSC-EVs) for 72 hours; B: Transwell migration assays (left) and subsequent quantitative analysis were performed using ImageJ software (right) for HDFs treated with MSC-EVs; C: Scratch assays provided qualitative (left) and quantitative (right) evidence for an increase in the migration of HDFs after treatment with MSC-EVs. In vitro wound healing assays revealed that the migration of HDFs increased after treatment with MSC-EVs at 6 and 22 hours after administration. Specifically, at the 22-hour mark, the wound healing rate of the MSC-EV-treated cells (85.00%) was significantly greater than that of the phosphate buffered saline-treated control cells (57.33%); D: Representative images of senescence-associated β-galactosidase-stained samples and quantification of HDFs. Quantitative assessments revealed that the percentage of senescence-associated β-galactosidase-positive cells after treatment with MSC-EVs (30%) was considerably lower than that recorded in the untreated control group of the passage 30 cohort (49%); E-H: Real time quantitative polymerase chain reaction analyses of the expression of the senescence and inflammatory markers P21, interleukin-1β, the extracellular matrix remodeling enzyme matrix metallopeptidase 1, and collagen type I. The data represent the mean ± SD of three replicates. ^aP < 0.05, ^bP < 0.01. For the control condition, human dermal fibroblasts were incubated with a comparable volume of phosphate buffered saline instead of extracellular vesicles derived from mesenchymal stromal cells. The images were quantitatively analyzed using ImageJ software. MSC-EVs: Extracellular vesicles derived from mesenchymal stromal cells; SA-β-gal: Senescence-associated β-galactosidase; EVs: Extracellular vesicles; IL-1β: Interleukin-1β; MMP-1: Matrix metallopeptidase 1; COL-1: Collagen type I.

Intravenous administration of MSC-EVs alleviated aging-related phenotypes in mice

To show that our product can be used as a therapeutic agent in the clinical setting, comprehensive in vivo studies must be conducted to validate its anti-aging properties. Small-molecule fluorophores exhibit strong and stable fluorescence in the membranes of EVs, allowing the in vivo tracking of EVs in animal models[32,33]. First, we investigated the in vivo trafficking of MSC-EVs by injecting DiD-EVs intravenously into BALB/c nude mice through the tail vein. Nano flow cytometry revealed that 46.5% of the MSC-EVs were stained with DiD (Figure 3A). The median diameter of the DID-EVs was 73.25 nm, with a mean particle size of 85.27 nm (Figure 3B). After the MSC-EVs were injected, a specific signal indicating the accumulation of MSC-EVs was detected in the liver (Figure 3C). Our results demonstrated that the liver is a preferential target organ of MSC-EVs when they are administered via the tail vein, indicating that MSC-EVs may be promising agents for treating liver disease.

Figure 3 Biodistribution of administered extracellular vesicles derived from mesenchymal stromal cells in mice. A: Nano flow cytometry analysis for determining the efficiency of labeling extracellular vesicles derived from mesenchymal stromal cells with DiD (DiD-labeled MSC-EVs); B: The particle size distribution of DiD-labeled MSC-EVs; C: The relative biodistribution of DiD-labeled MSC-EVs in nude mice. PBS: Phosphate buffered saline; EVs: Extracellular vesicles.

Several studies have shown that D-gal can accelerate aging in mice for more than eight weeks, decrease the rate of weight gain, and reduce body weight and the hepatic mass-body weight ratio[34,35]. To evaluate the potential effects of MSC-EVs on age-associated pathophysiology, D-gal-induced aged mice were intravenously injected (via the tail vein) with MSC-EVs twice a week for six weeks. Poisoning and death did not occur during the experiment, which lasted 10 weeks. At the end of the experiment, the final liver index of the model group was significantly lower than that of the EV-treated group (Figure 4A). The images of hematoxylin and eosin-stained tissues revealed circular vacuoles in the cytoplasm in the model group. Occasionally, small areas of focal necrosis, nuclear fragmentation with lymphocyte infiltration, and small areas of venous congestion were observed. After the administration of MSC-EVs, the hepatic anomalies detected in the liver images of the D-gal-induced subacute aging model mice were significantly ameliorated (Figure 4B).

Figure 4 Intravenous administration of extracellular vesicles derived from mesenchymal stromal cells alleviated various age-related phenotypic changes in mice. A: The liver indices of the mice; B: Representative images of hematoxylin and eosin-stained liver tissue in the control, model, and mice were treated with extracellular vesicles derived from mesenchymal stromal cells after induction with D-galactose groups (n = 3). Black arrow: Hydropic degeneration in hepatocytes; green arrow: Congestion in the central vein; yellow arrow: Cytoplasmic vacuoles of different sizes, indicating degeneration of hepatocyte vacuoles; blue arrow: Nuclear fragmentation with scant lymphocytic infiltration and focal necrosis. The scale bar of the hematoxylin and eosin scan = 200 μm; the scale bar of the enlarged image = 50 μm; C: Malondialdehyde levels in the liver; D: Total superoxide dismutase (SOD) levels in the liver; E: Catalase levels in the liver; F-H: Quantitative polymerase chain reaction was performed to determine the expression of the oxidation-related genes Mn-SOD (F), Cu/Zn-SOD (G), and catalase (H) in the three groups; I and J: Quantitative polymerase chain reaction was performed to determine the expression of the aging-related genes interleukin-1β (I) and P21 (J) in the three groups. Control: Untreated mice served as the normal baseline; model: D-galactose-induced aging mice; MSC-EVs: Mice were treated with extracellular vesicles derived from mesenchymal stromal cells after induction with D-galactose. ^aP < 0.05, ^bP < 0.01, ^cP < 0.001. EVs: Extracellular vesicles; MSC-EVs: Extracellular vesicles derived from mesenchymal stromal cells; MDA: Malondialdehyde; T-SOD: Total superoxide dismutase; CAT: Catalase; IL-1β: Interleukin-1β.

D-gal is metabolized in the liver, and under the catalysis of galactose oxidase, high levels of D-gal can be converted into aldose and hydrogen peroxide to produce reactive oxygen species (ROS). An increase in ROS levels may lead to oxidative stress, inflammation, abnormal mitochondrial function, and apoptosis[36]. MDA, SOD, and CAT serve as established indicators for assessing oxidative stress[37]. MDA is the end product of fatty acid β-oxidation, while SOD and CAT can prevent the excessive accumulation of ROS in cells. In this study, we examined the levels of oxidative stress and antioxidants in the livers of mice in different groups. Compared with those in the model group, the MDA levels were lower, and the concentrations of SOD and CAT were greater in the MSC-EV treatment group (Figure 4C-E). The results of the qPCR assays also revealed that MSC-EVs upregulated the expression of MnSOD, Cu/Zn-SOD, and CAT (Figure 4F-H). Thus, we next examined the expression of P21 and IL-1β and found that the expression of P21 (Figure 4I) and IL-1β (Figure 4J) in the model group was considerably greater than that in the control group. Taken together, these results showed that treatment with MSC-EVs significantly decreased the level of MDA in the liver tissue of D-gal-induced aged mice, increased the viability and level of SOD and CAT gene expression, and decreased the transcript levels of P21 and IL-1β. These results suggest that MSC-EVs have anti-senescence effects in D-gal-induced aging mouse models.

Integrated metabolomic and transcriptomic profiling revealed the anti-aging effect of MSC-EVs on D-gal-induced aged mice

To elucidate the underlying anti-aging mechanisms of MSC-EVs, comprehensive metabolomic and transcriptomic profiling was conducted on liver specimens obtained from the D-gal-induced aging model group (D-gal group) and the MSC-EV-treated group (MSC-EVs group). Association analysis of the metabolome and transcriptome can overcome the limitations of a single omics study and more comprehensively explain the metabolic processes and mechanisms underlying changes.

PCA (Figure 5A) and OPLS-DA (Supplementary Figure 1A) revealed separation between the D-gal and MSC-EV groups, indicating that MSC-EVs altered the metabolic profile of D-gal-induced senescent mice. In this study, 4416 metabolites were identified. Among these metabolites, lipids and lipid-like molecules were the most abundant, accounting for 28.74%, whereas organic acids and their derivatives accounted for 14.27% (Figure 5B). The volcano plot (Supplementary Figure 1B) and hierarchical clustering heatmap (Figure 5C) revealed that among the differentially expressed metabolites, 160 were upregulated and 54 were downregulated. The metabolites were assigned to KEGG pathways belonging to 13 main categories (Supplementary Figure 1C).

Figure 5 Integrated transcriptomic and metabolomic analysis. A: A score plot of the principal component analysis representing metabolomic differences between the model group (in blue) and the mice were treated with extracellular vesicles derived from mesenchymal stromal cells (MSC-EVs) after induction with D-galactose group (in red). The cross-validation of the predicted model showed a PC1 value of 35.89% for the first component and a PC2 value of 18.34% for the second component for all patterns (positive and negative total ions); B: Metabolite categorization. Classification of the 4416 identified metabolites, providing an overview of the metabolic diversity affected by treatment with MSC-EVs; C: A clustered heatmap of differentially abundant metabolites was constructed to illustrate the patterns of metabolite expression across samples, showing how these patterns are altered after treatment with MSC-EVs; D: A score plot of principal component analysis representing transcriptomic differences between the model group (in blue) and the MSC-EV-treated group (in red); E: Hierarchical clustering analysis of all differentially expressed genes. The heatmap with dendrograms represents the patterns of gene expression and clustering of genes with similar expression profiles; F: A bar chart of the top 30 Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways significantly enriched for differentially expressed genes and metabolites, ranked according to the level of enrichment; G: A KEGG pathway analysis (kgml) map. A visual representation of the pathways, where different shapes and colors denote distinct types. Red circles represent KEGG pathways, blue squares represent metabolites, and yellow triangles represent genes. MSC-EVs: Extracellular vesicles derived from mesenchymal stromal cells; PCA: Principal component analysis.

Transcriptional profiling of liver tissues from the model and MSC-EV groups was performed to assess upstream regulatory changes. PCA revealed that the model and MSC-EV groups formed distinct clusters (Figure 5D). A total of 137 genes were subsequently differentially expressed in the MSC-EV group relative to the model group. A volcano plot was constructed to visualize the 79 significantly upregulated genes and 58 downregulated genes between the two groups (Supplementary Figure 2A). A heatmap generated based on results of hierarchical clustering analysis was used to visualize the expression patterns of these differentially expressed genes in both groups (Figure 5E). KEGG pathway enrichment analysis classified the 137 differentially expressed genes into five principal categories: Cellular processes, environmental information processing, human diseases, metabolism, and organismal systems (Supplementary Figure 2B).

Network analysis of transcriptomic and metabolomic data was performed to assess the mechanism underlying the anti-aging effects of MSC-EVs on D-gal-induced aged mice. A Venn diagram was generated by comparing pathways associated with genes in the transcriptomic data and metabolites in the metabolomic data. The figure shows the overlap between pathways affected by differentially expressed genes and metabolites and highlights the shared and unique pathways; 22 shared pathways were identified (Supplementary Figure 3). The KEGG pathways associated with differentially expressed genes and metabolites identified in this study were subsequently visualized (Figure 5F). We found a complex interrelationship between four pathways, which included antifolate resistance, the prolactin signaling pathway, central carbon metabolism in cancer, and nicotinate and nicotinamide metabolism (Figure 5G). The concentrations of L-aspartate, L-tyrosine, L-methionine, and L-phenylalanine in the livers of EV-treated mice were significantly greater than those in the livers of D-gal-induced aged mice (Figure 5G). These amino acids are not just the basic building blocks of proteins; they also interact with central carbon metabolism through various metabolic pathways; provide energy, cofactors, and signaling molecules; and participate in cell growth, maintenance, and functional regulation[38]. An increase in these amino acid levels suggests alterations in metabolic pathways related to neurotransmitter synthesis, protein synthesis, or other metabolic processes[39]. Methionine is also involved in the synthesis of glutathione, a potent antioxidant. An increase in methionine levels indicates a response to oxidative stress, which is often associated with aging[40]. Increasing these amino acids might be part of a compensatory mechanism to counteract oxidative damage. An increase in the levels of amino acids such as phenylalanine and methionine indicates increased protein turnover and tissue repair processes[41]. Thus, an increase in the levels of these amino acids might suggest that the treatment facilitates tissue repair mechanisms in the livers of aging mice.

In this study, we also found that synthesis of cytochrome c oxidase 2 (Sco2) expression in the liver was greater in the EV-treated group than in the model group. The Sco2 gene assembles the cytochrome c oxidase complex, a key component of the mitochondrial electron transport chain. The cytochrome c oxidase complex is responsible for the final step in mitochondrial respiration, where electrons are transferred to oxygen and enable ATP synthesis[42]. Increased Sco2 expression might promote mitochondrial biogenesis or increase mitochondrial function, which often decreases with age. Increasing mitochondrial function can increase energy production and ameliorate some age-related cellular deficits.

DISCUSSION

Many researchers have investigated aging for several decades and developed interventions to counteract age-related health deterioration. However, no satisfactory solution has been found to rejuvenate or improve aging[43,44]. The use of EVs obtained from MSCs is currently under investigation for the treatment of age-associated conditions, largely owing to the ability of these EVs to regulate immunity and promote tissue regeneration[45]. However, only a few studies have used MSC-EVs to treat aging-related diseases in clinical practice. A limitation of EV-based therapy is the lack of standardized preparation protocols, and the quality criteria and mechanism of action for clinical-grade MSC-EV products are also unclear[46]. In this study, we produced clinical-grade EVs from human UC-MSCs. We subsequently investigated the roles and possible mechanisms underlying the effects of MSC-EVs on aging and aging-related changes in vitro and in vivo. To develop MSC-EVs as therapeutic products, the production and purification methods must be compatible with GMP’s and can be industrialized[43]. The isolation and purification of EVs are critical steps for subsequent research and clinical applications. Currently, the most commonly used separation techniques include ultracentrifugation, ultrafiltration, immunoaffinity capture, size-exclusion chromatography, and precipitation, each with unique advantages and limitations. As the gold standard for exosome isolation, ultracentrifugation achieves separation of EVs with varying densities, sizes, and morphologies through progressively increasing centrifugal force. While simple to perform, this method suffers from low purity, a lengthy processing time, and a limited yield. Although sucrose or iodixanol density gradient centrifugation can improve purity, prolonged exposure to high concentrations of sucrose may compromise the structural integrity of EVs. Ultrafiltration employs semipermeable membranes with defined pore sizes for separation, making it suitable for preliminary purification but with limited resolution. Immunoaffinity capture, which is based on antigen-antibody interactions, offers high purity and resolution, although residual capture antibodies may interfere with downstream functional studies. Size-exclusion chromatography represents a gentle isolation method that yields highly active EVs. Yet, its resolution decreases for particles approaching or exceeding the maximum pore size limit, often necessitating a combination with other techniques. Precipitation is straightforward and requires no specialized equipment, but impurity contamination affects purity[47,48].

Emerging technologies such as immunocapture, microfluidics, and EXODUS systems show great promise, although they currently face challenges, including high costs and difficulty scaling up[49]. The optimization of combined separation methods may enable the preparation of EVs with both high purity and yield, advancing industrial-scale production. Notably, TFF has significant advantages in processing large-volume samples, potentially overcoming the limitations of ultracentrifugation[50]. Our results suggest that the combined application of TFF and ultracentrifugation can be used as a protocol for UC-MSC manufacturing. In this study, we successfully developed an integrated approach combining TFF and umbilical cord for isolating highly pure small EVs from UC-MSC culture supernatants. This method effectively overcomes the limitations of traditional ultracentrifugation, significantly improving the vesicle recovery rate and purity while maintaining their structural integrity and biological activity. Notably, our protocol sets standards for the clinical application of the product and incorporates essential quality control checkpoints focused on safety, biological activity, and characterization analysis (Table 1). These criteria included the size distribution and particle concentration of the EVs prepared by nano flow cytometry; the presence of EV markers; and morphology, purity, and safety profile assays. The criteria identified in this study were highly consistent with authoritative group standards, such as the “Minimal information for studies of extracellular vesicles 2023” guidelines[51] and the requirement for small EVs of human pluripotent stem cells[52].

Table 1 The critical quality control points of clinical-grade extracellular vesicles derived from mesenchymal stromal cells.

	Parameter	Method	Release criteria
MSCs	Security
	Sterility	BacT/ALERT	Negative
	Endotoxin (EU/mL)	According to Chinese pharmaceutics	< 0.5
	Mycoplasma	qPCR	Negative by qPCR
	Biological activity and characterization
	Cell viability (%)	Trypan blue staining	≥ 90%
	Marker profile	Flow cytometry	CD90, CD105, CD73, CD 44 ≥ 95%; CD45, CD34, CD11b, CD HLA-DR ≤ 2%
	Trilineal differentiation	Osteogenesis, lipogenesis and chondrogenesis	Positive
EVs	Security
	Sterility	BacT/ALERT	Negative
	Endotoxin (EU/mL)	According to Chinese pharmaceutics	< 0.5
	Mycoplasma	qPCR	Negative by qPCR
	Biological activity and characterization
	Morphology	Transmission electron microscopy	Nonagglomerated saucer-like or cup-like structures with clear edges and membrane structure
	Particle size	NanoFCM	Distributed in a range of < 200 nm with a size peak in this range
	Marker protein	Western blotting	Presence of the surface biomarkers (CD9 and CD81) and interior biomarker TSG101; absence of the negative biomarker GM130
	Purity	Triton X-100/NanoFCM	The phospholipid membranes of EVs can be lysed by Triton X-100

MSCs: Mesenchymal stromal cells; qPCR: Quantitative polymerase chain reaction; NanoFCM: Nano flow cytometry; EVs: Extracellular vesicles; TSG101: Tumor susceptibility gene 101.

To confirm the effectiveness of our product in clinical settings, an efficacy assay was performed to evaluate the anti-aging ability of MSC-EVs in vitro and in vivo. HDFs are the most common cell type in the skin and play essential roles in maintaining its structure and function. Owing to their easy access and culture characteristics and their plasticity in cell experiments in vitro, they are widely used to study cell senescence, cell biology, wound healing, and various skin diseases. In this study, MSC-EVs significantly reduced the expression of SA-β-gal, MMP-1, P21, and IL-1β but increased the expression of COL-1 in passage 30 HDFs. Thus, HDFs can be used as a cellular model to evaluate the anti-aging function of EVs. They are effective experimental systems with biological significance and are convenient for studying EVs.

Choosing and confirming an appropriate experimental model is also an important part of evaluating the effectiveness of MSC-EVs. D-gal was first used in studies on aging in 1962 and has been widely used[53]. Owing to cost and time constraints, simulated aging models are generally better than natural aging models. The results of this study indicated that a mouse model of aging was successfully created by administering D-gal via intraperitoneal injection. Next, MSC-EVs were administered to model mice to conduct in vivo studies and validate the anti-aging effect of MSC-EVs. In the in vivo study, MSC-EVs attenuated hepatic damage caused by D-gal, which was characterized by hepatocyte swelling, abnormal shapes, disrupted organization, vacuolar degeneration, and a reduction in binucleated hepatocytes. It also helps restore the normal structure of the liver. Our findings demonstrated that the transplantation of MSC-EVs decreased MDA levels and ameliorated the oxidative milieu in the mitochondria of aged mice. These observations support the hypothesis that a decrease in oxidative stress is a potential mechanism underlying the anti-aging effects of MSC-EVs.

We integrated metabolomics and transcriptomics to analyze and identify potential pathways involved to determine the mechanism underlying the MSC-EV-mediated recovery of D-gal-induced aged mice. In this study, the concentrations of L-aspartate, L-tyrosine, L-methionine, and L-phenylalanine in the livers of EV-treated mice were significantly greater than those in the livers of D-gal-induced aged mice. These amino acids can be transformed and used in multiple metabolic pathways, which are integral units of proteins and are involved in energy metabolism, biosynthesis, and signaling[38,39]. Mitochondrial function decreases with age, and mitochondrial dysfunction may increase oxidative stress that accelerates cellular senescence. Sco2 is a key enzyme in the mitochondrial electron transport chain and plays an important role in maintaining mitochondrial function[54,55]. In this study, the expression of Sco2 was upregulated in the exosome-treated group compared with the model group. Our results suggest that the anti-aging efficacy of MSC-EVs predominantly involves a decrease in oxidative stress and the maintenance of mitochondrial function.

Oxidative stress refers to the phenomenon in which free radicals or nonradical species damage living cells by extracting electrons from biomolecules, triggering chain reactions that ultimately destroy cellular structures. ROS play a central biological role. When the levels of oxygen and its derived free radicals exceed the intrinsic antioxidant defense capacity of a cell, a state of oxidative stress damage occurs. The exacerbation and persistence of oxidative stress have been confirmed as key contributors to various clinically prevalent diseases[56,57]. According to the free radical theory, oxidative stress-induced cellular damage is a core mechanism underlying numerous age-related diseases, and this persistent impairment progressively disrupts the structure and function of organisms[58]. Excessive accumulation of ROS compromised macromolecular function and membrane systems, ultimately accelerating the progression of aging-related diseases and organismal senescence in an irreversible manner[59,60].

As multifunctional carriers, EVs not only effectively suppress ROS generation but also remodel aberrant intercellular communication and enhance stem cell function and quality, thereby delaying aging-associated stem cell exhaustion. Studies have demonstrated that the treatment of senescent MSCs with EVs derived from young MSCs induces glycolytic oxidative phosphorylation, reduces the activity of the senescence marker SA-β-gal, and activates the expression of pluripotency factors (octamer-binding protein 4, sex-determining region Y-box 2, Kruppel-like factor 4, and cellular myelocytomatosis oncogene)[61]. MSC-EVs play a pivotal role in regulating energy metabolism and oxidative stress by transferring miRNAs to modulate mitochondrial metabolism[62]. Research has shown that MSC-EVs carry specific miRNAs (e.g., miR-24-3p and miR-96-5p) and functional proteins to regulate hepatic lipid metabolism and inflammatory responses. For example, human umbilical cord-derived MSC-EVs inhibit lipid deposition and inflammation by targeting Keap-1 and alleviating oxidative stress via the nuclear factor erythroid-2-related factor 2/NAD(P)H quinone oxidoreductase 1 antioxidant pathway[63]. Additionally, MSC-EVs carrying miR-302b activate hypoxia-inducible factor-1 alpha, triggering a cascade that modulates multiple signaling pathways to delay premature aging, improve stem cell function, and reprogram energy metabolism toward a glycolytic mode[64]. These groundbreaking findings suggest that MSC-EVs can act as natural antioxidants, effectively ameliorating mitochondrial dysfunction and mitigating oxidative stress damage by regulating the expression of specific miRNAs. These findings provide an innovative therapeutic strategy for treating aging and other oxidative stress-related diseases. Future research should focus on elucidating the biogenesis, targeted delivery mechanisms, and functional validation of exosomal miRNAs to advance the clinical translation of this approach. To comprehensively elucidate the intricate molecular mechanisms and pathways mediating the geroprotective effect of MSC-EVs, further empirical studies need to be conducted.

In conventional anti-aging approaches, senolytics - emerging senescent cell-clearing compounds - combat age-related diseases by systematically eliminating senescent cells through the inhibition of senescent cell antiapoptotic pathways[65]. However, they suffer from off-target toxicity and regenerative deficiencies. Caloric restriction (CR) enhances autophagy via adenosine 5’-monophosphate-activated protein kinase/silent information regulator 1 pathway activation but negatively impacts wound healing, bone health, cognition, reproductive function, and infection resistance[66].

In contrast, MSC-EVs exert anti-aging effects through pleiotropic mechanisms, such as suppressing senescence signals (downregulating the nuclear factor-kappa B pathway), promoting tissue regeneration (transforming growth factor-β1/Smad3 activation), and reprogramming immune phenotypes (inducing macrophage M2 polarization) via bioactive cargo delivery[67]. This multifactorial synergy demonstrates superior therapeutic potential over single-mechanism therapies (e.g., senolytics) for complex pathologies. In terms of translational feasibility, MSC-EVs support versatile administration routes (intravenous, nebulized inhalation, and local injection). Their acellular nature eliminates the tumorigenic risks and immune rejection associated with MSC transplantation, whereas senolytics are limited to oral delivery, and CR requires unsustainable long-term dietary control. Nevertheless, MSC-EVs face challenges in terms of heterogeneity control, scalable production, and long-term safety validation.

To address these limitations, a hierarchical framework should be developed: Temporal synergy (senolytic clearance followed by EV-mediated regeneration), metabolic priming (CR preconditioning to increase EV efficacy), and engineering optimization (bone-targeting aptamer-modified EVs combined with low-dose senolytics). This strategy will advance precision anti-aging interventions through multimechanism collaboration.

CONCLUSION

We developed a protocol for producing MSC-EVs that met cGMP standards and defined release criteria for products suitable for human application. These criteria included the size distribution and particle concentration of the EVs prepared; EV markers; and morphology, purity, and safety profile assays. Our pharmacokinetic, in vitro, and in vivo pharmacodynamic results confirmed the anti-aging effect of MSC-EVs. Our findings also suggested that MSC-EVs can decrease senescence induced by D-gal by reducing oxidative stress and modulating metabolism. Thus, MSC-EVs may be promising therapeutic agents for preventing and managing senescence and related disorders.