Extracellular Vesicles Derived from Adipose-Derived Mesenchymal Stem Cells Alleviate Apoptosis and Oxidative Stress of Retinal Pigment Epithelial Cells Through Activation of Nrf2 Signaling Pathway

Abstract

Purpose:

To examine the potential protective effects of adipose-derived mesenchymal stem cell-derived extracellular vesicles (ASC-EVs) on ARPE-19 cells exposed to hydrogen peroxide (H₂O₂) stress and to evaluate their ability to delay retinal degeneration in Royal College of Surgeons (RCS) rats.

Methods:

ARPE-19 cells were pre-treated with ASC-EVs for 24 h, followed by exposure to 200 μM H₂O₂ for an additional 24 h. RCS rats received an intravitreal injection of phosphate-buffered saline in one eye and ASC-EVs in the other eye.

Results:

ASC-EV pretreatment significantly protected against H₂O₂ in the Cell Counting Kit-8 assay and was also effective in the lactate dehydrogenase-release assay. It notably reduced early apoptosis (Annexin V-fluorescein isothiocyanate/propidium iodide assay) and late apoptosis (Terminal Deoxynucleotidyl Transferase dUTP Nick End Labeling assay), while significantly decreasing intracellular reactive oxygen species, glutathione levels, and superoxide dismutase activity. NFE2L2, HMOX1, and NQO1 mRNA levels, along with Nrf2, HO-1, and NQO1 protein levels, were significantly elevated with ASC-EV pretreatment. Compared with ARPE-19-derived EVs, 11 miRNAs were upregulated and 34 were downregulated in ASC-EVs. In RCS rats, intravitreal injections of ASC-EVs led to significant preservation of the outer nuclear layer and photoreceptor segments, along with increased nuclear Nrf2 expression and elevated HO-1 and NQO1 levels in the inner retina. Eyes that received intravitreal injections of ASC-EVs demonstrated significantly preserved electroretinography a- and b-wave amplitudes at 1 week post-injection, though this effect faded by 2 weeks.

Conclusions:

ASC-EVs mitigated apoptosis and oxidative stress in ARPE-19 cells subjected to H₂O₂ exposure and temporarily slowed retinal degeneration in RCS rats via Nrf2 pathway activation by miRNAs.

Introduction

Retinal degeneration encompasses a range of diseases that lead to blindness, characterized by the progressive loss of several cell types within the neurosensory retina, such as ganglion cells, photoreceptor cells, and retinal pigment epithelium (RPE) cells. ¹ This condition can result from numerous diseases, including age-related macular degeneration (AMD), retinitis pigmentosa (RP), Stargardt disease, diabetic retinopathy (DR), uveitis, and glaucoma. Like other cells in the nervous system, retinal cells exhibit a limited capacity to divide and regenerate. ² Consequently, once these cells are lost, there are currently no methods available to replace or regenerate them. Thus, the primary goal of treating retinal degeneration is to prevent further loss of retinal cells, thereby slowing the progression of the disease.

Mesenchymal stem cells (MSCs), including adipose-derived mesenchymal stem cells (ASCs), hold potential for treating retinal degeneration owing to their diverse bioprotective properties. These properties include the paracrine secretion of growth and neurotrophic factors, anti-inflammatory effects, anti-apoptotic characteristics, antioxidant capacity, and the potential for mesodermal differentiation. ^3,4 However, the exact mechanisms underlying their therapeutic effects remain largely unclear, though they are believed to involve the paracrine mechanism. ⁵

Extracellular vesicles (EVs) are membrane-bound vesicles enclosed by a phospholipid bilayer that contain nucleic acids, cytosolic and cell surface proteins, lipids, and metabolites. ⁶ EVs are essential for intercellular communication and serve as therapeutic mediators of MSCs via their paracrine actions. ⁷ MSC-derived EVs, such as ASC-derived extracellular vesicles (ASC-EVs), are gaining increasing interest in the field of biological therapy compared with MSCs themselves. This is primarily due to their excellent safety profile, ability to traverse the blood–brain barrier, and high abundance. ⁸ However, the potential of ASC-EVs for treating retinal degeneration remains unclear.

Herein, we aimed to assess the protective effects of ASC-EVs against hydrogen peroxide (H₂O₂)-induced apoptosis and oxidative stress in RPE cells and to elucidate the underlying molecular mechanism. Additionally, we used an animal model to evaluate the impact of intravitreal injection of ASC-EVs on retinal degeneration.

Materials and Methods

Isolation and cultivation of ASCs

Human adipose tissue was procured from a 30-year-old male donor through a liposuction procedure. The tissue was enzymatically digested with 0.075% (w/v) Type II collagenase (Sigma-Aldrich, St. Louis, MO) with gentle agitation for 40 min at 37°C. The isolated ASCs were cultured in an expansion medium containing α-minimum essential medium (α-MEM; Welgene, Gyeongsan, Gyeongsang, Republic of Korea), 10% fetal bovine serum (FBS; Capricorn Scientific GmbH, Ebsdorfergrund, Germany), and 5 ng/mL basic fibroblast growth factor (R&D Systems, Minneapolis, MN). Subsequently, the ASCs were cultured in expansion media and subcultured upon reaching 80% confluence. Approval for this study was obtained from the Institutional Review Board (IRB) of Hallym University Medical Center (IRB no. 2017–118), and informed consent was provided by the patient.

Preparation of ASC-conditioned media

ASCs at passage 6 were seeded at a density of 2.5 × 10³ cells/cm² and cultured in expansion media for 3 days. Upon reaching 80% confluence, the cells were washed once with phosphate-buffered saline (PBS) and then cultured in α-MEM for 24 h. Subsequently, the ASCs were cultured in α-MEM supplemented with 10% particle-depleted serum replacement (Xeno Free CTS; Gibco Life Technologies, Carlsbad, CA) for 2 days. The conditioned media, now containing EVs, was collected and filtered using a 0.22-µm bottle-top vacuum filter (Corning, Inc., Corning, NY).

Isolation of EVs from ASC-conditioned media

A tangential flow filtration (TFF) system with a 500-kDa cut-off hollow fiber membrane (EMD Millipore Corp., Billerica, MA), was utilized to isolate EVs from the ASC-conditioned media. A volume of 1–2 L of conditioned media was processed through the TFF system. The system’s flow rate was maintained at 180 rpm, ensuring that the pressure did not exceed 1.5 psi. To enhance EVs yield, 12 L of PBS was subsequently used to wash the filter, removing any residual EVs. The collected EVs were then filtered through a 0.22-µm bottle-top vacuum filter (Corning, Inc.).

Cell culture and treatments

The ARPE-19 human RPE cell line was obtained from the American Type Culture Collection (Manassas, VA) at passage 19. For the experiments, only cells at passage 25 or below were used. ARPE-19 cells were seeded at a density of 1.5 × 10⁴ cells/cm². All cultures were maintained at 37°C in a humidified atmosphere supplemented with 5% CO₂. Seeded cells were maintained for 24 h in DMEM/F12 medium (Gibco, Gaithersburg, MD) enriched with 10% FBS (Gibco) and 100 U/mL penicillin (Sigma-Aldrich, St. Louis, MO), and then starved in serum-free DMEM/F12 for 24 h.

Subsequently, the cells were treated with different concentrations of ASC-EVs for 24 h: 2 × 10⁵, 2 × 10⁴, and 2 × 10³ EVs per seeded ARPE-19 cell. The control group comprised cell culture in medium without ASC-EVs. We then added 400 µM of H₂O₂ to the cells to achieve a final concentration of 200 µM H₂O₂, followed by incubation for 24 h. H₂O₂ dilutions were prepared from a 30% stock solution (Sigma-Aldrich) using phenol red-free medium.

Cell viability and cytotoxicity assays

Cell viability in ARPE-19 cells after exposure to H₂O₂ and treatment with ASC-EVs was assessed using a Cell Counting Kit-8 (CCK-8; Dojindo Laboratories, Kumamoto, Japan). The absorbance at 450 nm was measured using a Synergy HTX multimode microplate reader (Agilent Technologies, Santa Clara, CA).

The cytotoxic effects of H₂O₂ and ASC-EVs on ARPE-19 cells were determined using a lactate dehydrogenase cytotoxicity detection kit (MK401; Takara Bio, Otsu, Japan). The assay used the culture medium collected after the cells were cultured as described previously. The absorbance at 490 nm was measured using the Synergy HTX microplate reader.

Apoptosis analysis

Early and late-stage apoptotic cell death was evaluated using an Annexin V–fluorescein isothiocyanate (FITC)/propidium iodide (PI) apoptosis detection kit (BD Biosciences, San Diego, CA). The CytoFLEX flow cytometer (Beckman Coulter, Brea, CA) was used to perform flow cytometry. Cells negative for PI and FITC were categorized as alive; cells negative for PI but positive for FITC were identified as early apoptotic; cells positive for both PI and FITC were labeled as late apoptotic or dead; and cells positive for PI but negative for FITC were designated as dead.

The presence of apoptotic DNA fragmentation, a marker of late-stage apoptosis, was detected using the Terminal Deoxynucleotidyl Transferase dUTP Nick End Labeling (TUNEL) assay employing the DeadEnd Fluorometric TUNEL System (G3250; Promega, Madison, WI).

Measurement of intracellular reactive oxygen species and antioxidant enzymes

To assess intracellular reactive oxygen species (ROS) levels, we employed the dichloro-dihydro-fluorescein diacetate/H2DCFDA-Cellular ROS Assay Kit (ab113851; Abcam, Cambridge, MA). The proportion of fluorescence-positive cells was quantified using a CytoFLEX flow cytometer using excitation and emission wavelengths of 488 and 525 nm, respectively.

The intracellular glutathione (GSH) level was determined using the GSH-Glo Glutathione Assay Kit (V6911; Promega), ensuring that an equal number of cells were used from each group. A standard curve was generated using GSH standards ranging from 0 to 5 µM, and the GSH concentration in each sample was extrapolated from this curve. Luminescence was assessed using the Synergy HTX microplate reader.

The Superoxide Dismutase Activity Kit (MAK379; Sigma-Aldrich) was used to quantify total superoxide dismutase (SOD) activity, ensuring that an equal number of cells were used from each group. SOD activity is expressed as a percentage inhibition rate, reflecting the extent to which SOD inhibits the chemical reaction that transforms superoxide radicals into H₂O₂ and oxygen. The measurement was conducted by quantifying the reduction in absorbance at 450 nm using the Synergy HTX microplate reader.

Real-time reverse transcription-polymerase chain reaction

RNA was isolated using the ReliaPrep RNA Miniprep System (Promega). Subsequently, first-strand cDNA was synthesized from 200 ng of oligonucleotide primers utilizing the GoScript Reverse Transcription System (Promega). Transcript quantification was performed in real-time using the AccuPower 2X GreenStar qPCR Master Mix (Bioneer) with a LightCycler 96 (Roche Life Science, Mannheim, Germany). Real-time reverse transcription-polymerase chain reaction (RT-PCR) results are presented as fold changes relative to the control group. Relative expression levels were determined using the comparative CT (ΔΔCT) method, normalizing the expression of target genes to GAPDH. Supplementary Table S1 displays the primer sequences used.

Western blotting

Total cellular proteins were extracted using a radioimmunoprecipitation assay buffer (Biosesang, Seoul, Republic of Korea) supplemented with protease inhibitors (p8340; Sigma-Aldrich) and phosphatase inhibitors (PhosSTOP; Roche, Basel, Switzerland). The primary antibodies used were rabbit anti-human antibodies: anti-nuclear factor erythroid-2-related factor 2 (Nrf2; 1:500 dilution, ab62352; Abcam), anti-heme oxygenase-1 (HO-1; 1:1,000 dilution, ab13243; Abcam), anti-NAD(P)H quinone dehydrogenase-1 (NQO1; 1:7,000 dilution, ab80588; Abcam), and anti-GAPDH (1:5,000 dilution, LF-PA0212; AbFrontier, Dallas, TX).

Immunoreactive bands were visualized using a horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin G secondary antibody (1:5,000 dilution, ab205718; Abcam) and a WEST-Queen RTS Western Blot Detection Kit (iNtRON Biotechnology, Seongnam, Republic of Korea). Densitometric analysis of Western blots was performed using ImageJ software (National Institutes of Health, Bethesda, MD).

Cell transfection with Nrf2 siRNA

To examine the impact of silencing NFE2L2 on Nrf2 signaling pathway, cells were transfected with siRNA targeting NFE2L2, while scramble siRNA served as a negative control. Both AccuTarget Predesigned siRNAs specific for human NFE2L2 and scramble siRNAs were purchased from Bioneer (Daejeon, Republic of Korea). The wells in the 96-well plates were treated with 3.3 pmol of siRNA, while those in the 12-well plates were treated with 100 pmol of siRNA for 12 h using the Lipofectamine RNAiMAX Transfection Reagent kit (Thermo Fisher Scientific, Waltham, MA).

For Western blotting assay, the medium was replaced with DMEM/F12 supplemented with 10% FBS, and the cells were further incubated for 48 h. For the CCK assay, the medium was replaced with ASC-EVs at a concentration of 2 × 10⁵ per seeded ARPE-19 cell, or with medium without ASC-EVs, and the cells were cultured for 24 h. Following this, the medium was replaced with 600 µM H₂O₂ and incubated for 2 h.

Isolation and sequencing of EV microRNA

Next-generation sequencing (NGS) was performed by Macrogen (Seoul, Republic of Korea). EVs were isolated from the conditioned medium of ARPE-19 cells using the Exo2D-for-RNA assay kit (EP-RU10; Exosomeplus, Seoul, Republic of Korea). MicroRNAs (miRNAs) in EVs were extracted using the miRNeasy Serum/Plasma Kit (217184; QIAGEN, Hilden, Germany). For library preparation, 10 ng of EV-RNA was utilized as the input for each library. The SMARTer smRNA-Seq Kit for Illumina (Takara Bio Inc., Kusatsu, Japan) was employed to construct small RNA libraries through polyadenylation, cDNA synthesis, and PCR amplification. Sequencing of the constructed cDNA fragments was performed on the Illumina platform using a sequence-by-synthesis method to determine read length.

MicroRNA expression level analysis

After sequencing, the quality of the raw sequence reads was assessed and filtered. Subsequently, the reads were processed to eliminate adapter sequences and exclude ribosomal RNA. These processed reads were aligned with miRBase 22.1, a reference database for miRNAs, and RNAcentral 14.0, a non-coding RNA database, to identify known miRNAs and other RNA types, such as transfer RNA, small nuclear RNAs, and small nucleolar RNAs. The raw data were normalized using the trimmed mean of the M-values technique via EdgeR. ⁹ To enable log2 transformation and the generation of a correlation plot, one unit was added to the normalized counts of the filtered miRNAs. Calculations for log counts per million and log-fold changes for each miRNA were conducted and compared between groups.

A statistical hypothesis test was employed using the ExactTest function in EdgeR to analyze the variance between the two groups. MiRNAs with differential expression across the groups were identified based on a fold-change threshold ≥2 and a false discovery rate (FDR)-adjusted p-value <0.05, using the Benjamin-Hochberg method.

Target gene prediction and functional analysis of miRNAs

A study of miRNA–target interactions (MTIs) was performed using miRTarBase 9.0, which provides experimentally validated data. ¹⁰ MTIs were considered strongly supported when validated by western blotting, quantitative PCR, or reporter assays. Conversely, MTIs confirmed using alternative experimental techniques, such as microarray, NGS, pulsed stable isotope labeling with amino acids in cell culture, and cross-linking and immunoprecipitation followed by NGS, were considered to be weakly supported.

Functional and pathway enrichment analyses for differentially expressed miRNAs were performed using DIANA-miRPath v3.0. ¹¹ A FDR threshold of 0.05 was applied to the p values to ensure the accuracy and relevance of the findings.

Intravitreal EVs delivery

The Institutional Animal Care and Use Committee (IACUC) of the Clinical Research Institute of Hallym University Medical Center approved all procedures involving animal experiments (IACUC no. 2021-3-0315-05). These practices adhered to the guidelines stipulated by the Association for Research in Vision and Ophthalmology for the ethical use of animals in vision and ophthalmical research.

Ten 21-day-old Royal College of Surgeons (RCS) rats were used as recipients. The right eyes of all rats received an intravitreal injection of ASC-EVs, while the left eyes received an intravitreal injection of sterile PBS as an internal control. The rats were anesthetized with an intraperitoneal injection of a mixture of 10 mg/kg tiletamine/zolazepam hypochloride (Zoletil; Virbac laboratories, France) and 20 mg/kg xylazine (Rompun; Bayer Korea, Republic of Korea). The corneas of the rats were numbed with 0.5% proparacaine hydrochloride (Paracaine; Hanmi Pharm, Seoul, Republic of Korea), and their pupils were dilated using a combination of 0.5% tropicamide and 0.5% phenylephrine (Tropherine; Hanmi Pharm). Before injecting ASC-EVs or PBS into the vitreous cavity, a corneal puncture was performed to reduce intraocular pressure, thereby minimizing the outflow of the injected substance. The pars plana was penetrated using a 33-gauge syringe needle, and 10 µL of sterile PBS with or without 9.3 × 10⁸ ASC-EVs were slowly injected into the vitreous cavity. The needle was left in situ for 30 s before being retracted to minimize backflow.

Electroretinogram

Full-field electroretinograms (ERGs) were obtained in all rats using a UTAS-E2000 Visual Electrodiagnostic System (LKC Technologies, Gaithersburg, MD) both pre-injection and at 1- and 2-weeks post-intravitreal injection. The rats underwent overnight dark adaptation. Following anesthesia, the rats’ pupils were dilated, and their eyes were given topical anesthesia according to the same protocol used to perform the intravitreal injection. An electrode composed of a polymethylmethacrylate lens with a gold metal contact was placed on the rat’s cornea with a drop of 2% hydroxypropyl methylcellulose gel (Hycell solution; Samil Pharm Co., Ltd., Seoul, Republic of Korea). Needle electrodes were inserted subcutaneously, with those placed in the forehead serving as the reference electrode and those in the tail serving as the ground electrode. The rats were then placed in a Ganzfeld illumination dome.

Full-field scotopic ERGs were simultaneously recorded from both eyes using a standard flash stimulus of 3.0 photopic cd·s·m⁻² with a scotopic strength of 7.5 scotopic cd·s·m⁻². ¹² On average, five recordings per flash intensity were obtained. The a-wave amplitude was determined by measuring the first negative trough from the baseline, while the b-wave amplitude was calculated as the difference between the a-wave trough and the subsequent maximum positive deflection.

Histology

The rats were anesthetized for enucleation as previously described, either 7 or 14 days after the injection. They were then transcardially perfused with PBS followed by a 4% formaldehyde solution. The eyes were enucleated and immediately fixed in Davidson’s solution (32% ethanol, 2.2% neutral buffered formalin, and 11% glacial acetic acid in distilled water) for 24 h. Following fixation, the eyes were dehydrated through a series of ethanol concentrations (50%, 70%, 95%, and 100%) and then stored in PBS at 4°C.

For cryosectioning, the eyes were embedded in 20% sucrose in optimal cutting temperature compound (Sakura Finetek, Torrance, CA), snap-frozen, and stored at −80°C until use. For paraffin sectioning, the eyes were embedded in paraffin, cut into 5-µm-thick sections, and stained with hematoxylin and eosin.

Immunohistochemical staining

Frozen tissue sections were used for immunohistochemical staining. These tissues were sliced into 5-µm-thick sections and fixed in 4% paraformaldehyde. Following fixation, the tissue sections were hydrated and subjected to antigen retrieval using a microwave oven with citrate buffer (10 mM, pH 6.0). To prevent non-specific binding, 1% bovine serum albumin was used as a blocking agent. The primary antibodies employed were rabbit anti-human antibodies: anti-NRF2 (1:100 dilution, ab62352; Abcam), anti-HO-1 (1:100 dilution, ab13243; Abcam), and anti-NQO1 (1:100 dilution, ab80588; Abcam). For secondary antibodies, Alexa Fluor 488-conjugated goat anti-rabbit IgG (1:100 dilution, A11008; Invitrogen, Carlsbad, CA) was used. Subsequently, nuclei were stained with Hoechst 33342 (Sigma-Aldrich; 10 μg/ml in 0.1% Triton X-100/PBS). Coverslips were then mounted onto glass slides using Vectashield antifade mounting medium (Vector Laboratories, Burlingame, CA) and sealed with nail polish. The slides were examined using a fluorescent microscope (DMi8; Leica Microsystems, Wetzlar, Germany).

Statistical analysis

All data are expressed as the mean ± standard error of the mean (SEM) from at least three independent experiments. The Wilcoxon signed-rank test was used to determine the effects of ASC-EV treatment. Statistical analyses were conducted using SPSS software (version 20.0; IBM SPSS Inc., Chicago, IL), whereas graphs were generated using GraphPad Prism (version 6.0; GraphPad Inc., San Diego, CA). Statistical significance was established at a p value <0.05.

Results

ASC-EVs protect RPE cells from H₂O₂-induced cytotoxicity

Exposure to 200 μM of H₂O₂ for 24 h resulted in approximately a 50% reduction in ARPE-19 cell viability (p < 0.001, Fig. 1A). A 24 h pretreatment with ASC-EVs at doses of 2 × 10³, 2 × 10⁴, and 2 × 10⁵ EVs per seeded ARPE-19 cell markedly enhanced ARPE-19 cell viability following H₂O₂ exposure (p = 0.019, 0.001, and 0.002, respectively).

FIG. 1.

Effects of ASC-EV pretreatment on H₂O₂-induced cytotoxicity in ARPE-19 cells. (A) ARPE-19 cell viability was analyzed using the Cell Counting Kit-8 assay. (B) H₂O₂ cytotoxicity was analyzed using a lactate dehydrogenase-release assay. ASC-EVs, extracellular vesicles derived from adipose-derived mesenchymal stem cells; EVs, extracellular vesicles; H₂O_2, hydrogen peroxide. ∗∗p < 0.01 the control vs. the H₂O₂-treated group. ^# p < 0.05 the H₂O₂-treated group vs. the ASC-EV-pretreated group at each concentration ^## p < 0.01 the H₂O₂-treated group vs. the ASC-EV-pretreated group at each concentration.

The cytotoxicity of 200 μM H₂O₂ in ARPE-19 cells was significantly reduced only in the groups pre-treated with 2 × 10⁵ ASC-EVs per cell (p = 0.008, Fig. 1B). Taken together, these findings suggest that pretreatment with ASC-EVs effectively inhibits H₂O₂-induced cytotoxicity in ARPE-19 cells.

ASC-EVs protect RPE cells from H₂O₂-induced apoptosis

Two different approaches were employed to demonstrate the protective effect of ASC-EV pretreatment on H₂O₂-induced apoptosis in ARPE-19 cells: Annexin V-FITC/PI staining assay and TUNEL assay. Figure 2A presents a typical result obtained though the Annexin V-FITC/PI staining assay. Compared with the control group, the percentage of alive cells was notably decreased in the H₂O₂-treated group (p = 0.003). This reduction was primarily due to an increase in the percentage of early (p = 0.002) and late apoptotic cells (p = 0.02) in the H₂O₂-treated group. Pretreatment with ASC-EVs at a dose of 2 × 10⁵ EVs per cell significantly enhanced cell viability and significantly decreased early apoptosis after exposure to 200 μM of H₂O₂ (p = 0.015, both; Fig. 2B).

FIG. 2.

Effect of ASC-EV pretreatment on H₂O₂-induced apoptosis in ARPE-19 cells. (A) H₂O₂-induced apoptosis in ARPE-19 cells treated with various ASC-EV concentrations was detected using the Annexin V-FITC/PI double-staining assay. (B) Graphs illustrating the proportions of alive cells, early apoptosis, and late apoptotic cells as determined by the Annexin V-FITC/PI staining assay. (C) Percentage of TUNEL-positive cells in each group. ASC-EVs, extracellular vesicles derived from adipose-derived mesenchymal stem cells; EVs, extracellular vesicles; H₂O_2, hydrogen peroxide. ∗p < 0.05 the control vs. the H₂O₂-treated group, ∗∗p < 0.01 the control vs. the H₂O₂-treated group. ^# p < 0.05 the H₂O₂-treated group vs. the ASC-EV-pretreated group at each concentration.

TUNEL assay indicated that DNA fragmentation mainly occurred during late apoptosis (Fig. 2C). The H₂O₂-treated group exhibited an increase in late apoptosis (p = 0.023). Pretreatment with ASC-EVs at a dose of 2 × 10⁵ EVs per seeded cell significantly decreased late apoptosis after exposure to 200 μM of H₂O₂ (p = 0.005, Fig. 2C).

Overall, these findings suggest that ASC-EVs at a dose of 2 × 10⁵ EVs per cell effectively suppress H₂O₂-induced apoptosis in ARPE-19 cells.

ASC-EVs reduce intracellular ROS levels in RPE cells by enhancing endogenous antioxidant production

Further studies employed 2 × 10⁵ EVs per cell, demonstrating a significant reduction in H₂O₂-induced cytotoxicity and apoptosis. To investigate the antioxidant effects of ASC-EVs, we evaluated changes in intracellular ROS levels as well as GSH and SOD enzyme activities. Exposure to 200 μM of H₂O₂ significantly increased intracellular ROS levels (p = 0.028) and decreased GSH (p = 0.043) and SOD levels (p = 0.028) relative to the control group (Fig. 3). In contrast, ROS levels were notably reduced in the group pretreated with ASC-EVs before exposure to H₂O₂ (p = 0.028, Fig. 3A) compared with the group without pretreatment. Intracellular GSH levels (p = 0.043, Fig. 3B) and SOD activity (p = 0.028, Fig. 3C) were significantly increased in the group pretreated with ASC-EVs prior to exposure to H₂O₂. These results show that ASC-EV pretreatment effectively alleviated oxidative stress in ARPE-19 cells following exposure to H₂O₂.

FIG. 3.

Effect of ASC-EV pretreatment on intracellular levels of (A) reactive oxygen species, (B) GSH, and (C) SOD enzyme activity in ARPE-19 cells after oxidative stress injury. ASC-EVs, extracellular vesicles derived from adipose-derived mesenchymal stem cells; EVs, extracellular vesicles; H₂O_2, hydrogen peroxide; GSH, glutathione; SOD, superoxide dismutase. ∗p < 0.05 the control vs. the H₂O₂-trea ted group. ^# p < 0.05 the H₂O₂-treated group vs. the ASC-EV-pretreated group.

ASC-EVs activate the Nrf2 signaling pathway

We investigated whether ASC-EV pretreatment can increase endogenous antioxidants in ARPE-19 cells under H₂O₂-mediated oxidative stress through the activation of the Nrf2 signaling pathway. Exposure to H₂O₂ did not significantly alter the expression levels of NFE2L2 mRNA and NRF2 protein (p = 0.500, both; Fig. 4A). In contrast, H₂O₂ exposure led to increased expression levels of NQO1 mRNA (p = 0.046, Fig. 4C), HO-1 (p = 0.043, Fig. 4B), and NQO1 protein (p = 0.043, Fig. 4C). However, mRNA expression levels of NFE2L2 (p = 0.043, Fig. 4A), HMOX1 (p = 0.043, Fig. 4B), and NQO1 (p = 0.028, Fig. 4C), as well as protein expression levels of Nrf2 (p = 0.043, Fig. 4A), HO-1 (p = 0.043, Fig. 4B), and NQO1 (p = 0.043, Fig. 4C), were significantly increased in the group pretreated with ASC-EVs before H₂O₂ exposure compared with the group without pretreatment. These findings imply that the activation of the Nrf2 signaling pathway could serve as a vital mechanism through which ASC-EV pretreatment protects ARPE-19 cells against oxidative stress.

FIG. 4.

Effect of ASC-EV pretreatment on the Nrf2 signaling pathway. Quantitative analysis of relative mRNA expression and protein expression levels of (A) NFE2L2 and NRF2, (B) HMOX1 and HO-1, and (C) NQO1 and NQO1. (D) Representative Western blots for Nrf2, HO-1, and NQO1. ASC-EVs, extracellular vesicles derived from adipose-derived mesenchymal stem cells; EVs, extracellular vesicles; H₂O_2, hydrogen peroxide. *p < 0.05 the control vs. the H₂O₂-treated group. ^# p < 0.05 the H₂O₂-treated group vs. the ASC-EV-pretreated group.

To further confirm the role of the Nrf2 signaling pathway in increasing endogenous antioxidant levels in ARPE-19 cells following ASC-EV treatment, we investigated the effect of NFE2L2-siRNA on H₂O₂-induced cytotoxicity. NFE2L2-siRNA efficiently silenced Nrf2 protein expression (Fig. 5A). As shown in Fig. 5B, transfection of ARPE-19 cells with NFE2L2-siRNA prevented the enhanced cell viability against H₂O₂-induced cytotoxicity that was induced by pre-treatment with ASC-EVs (p = 0.114, compared with the H₂O₂-treated group). In contrast, transfection with scramble RNA did not influence the protective effect of ASC-EVs on ARPE-19 cells exposed to H₂O₂-induced cytotoxicity (p = 0.010, compared with the H₂O₂-treated group).

FIG. 5.

Impact of Nrf2 silencing on the protective effects of ASC-EV pretreatment against H₂O₂-induced cytotoxicity in ARPE-19 cells. (A) Western blot analysis of Nrf2 expression levels post-transfection with Nrf2 siRNA. (B) ARPE-19 cell viability post-transfection with NFE2L2-siRNA, assessed using the Cell Counting Kit-8 assay. **p < 0.01 the control vs. the H₂O₂-treated group. ^# p < 0.05 the H₂O₂-treated group vs. the ASC-EV pretreated group. ^† p < 0.05, the ASC-EV-pretreated group vs. the ASC-EV-pretreated group post-transfection with Nrf2 siRNA. ASC-EVs, extracellular vesicles derived from adipose-derived mesenchymal stem cells; EVs, extracellular vesicles; H₂O_2, hydrogen peroxide.

MicroRNAs present in ASC-EVs

Table 1 and Fig. 6 present the differential expression of miRNAs between ASC-EVs and ARPE-19-derived EVs. In ASC-EVs, the relative expression levels of 11 miRNAs were increased, whereas the levels of 34 miRNAs were decreased compared with those in ARPE-19-derived EVs.

Table 1.

Identification of Differentially Expressed miRNAs in ASC-EVs Compared with ARPE-19-Derived EVs Using RNA-Sequencing

miRNA	False discovery rate-adjusted p value	Fold change	Regulation
hsa-miR-196a-5p	5.412 × 10−6	127.695	Up
hsa-miR-145-5p	1.123 × 10−4	50.391	Up
hsa-miR-8485	2.261 × 10−6	41.584	Up
hsa-miR-23a-5p	0.027	41.530	Up
hsa-miR-1273c	0.027	41.530	Up
hsa-miR-424-5p	4.584 × 10−4	28.867	Up
hsa-miR-4644	0.013	18.941	Up
hsa-miR-1228-5p	0.050	18.330	Up
hsa-miR-6126	2.809 × 10−4	12.943	Up
hsa-miR-1260a	0.008	12.672	Up
hsa-miR-1306-5p	0.029	7.962	Up
hsa-let-7c-5p	0.024	−4.938	Down
hsa-miR-92a-3p	0.026	−4.971	Down
hsa-miR-222-3p	0.050	−5.026	Down
hsa-miR-99b-5p	0.021	−6.358	Down
hsa-miR-19b-3p	0.011	−6.374	Down
hsa-miR-224-5p	0.003	−9.949	Down
hsa-miR-125a-5p	7.376 × 10−4	−11.519	Down
hsa-miR-19a-3p	8.304 × 10−4	−12.614	Down
hsa-miR-29b-3p	7.350 × 10−4	−13.541	Down
hsa-miR-100−5p	9.357 × 10-4	−14.088	Down
hsa-miR-27b-3p	0.001	−14.805	Down
hsa-miR-877-5p	0.002	−16.573	Down
hsa-let-7b-5p	5.064 × 10−5	−18.210	Down
hsa-miR-574-3p	5.692 × 10−5	−18.236	Down
hsa-miR-191-5p	6.414 × 10−5	−20.585	Down
hsa-miR-342-3p	2.173 × 10−5	−31.251	Down
hsa-let-7d-5p	3.431 × 10−5	−33.189	Down
hsa-miR-423-3p	3.287 × 10−4	−34.324	Down
hsa-let-7a-5p	1.985 × 10−6	−35.210	Down
hsa-let-7e-5p	3.683 × 10−6	35.220	Down
hsa-miR-186-5p	2.716 × 10−4	−40.009	Down
hsa-miR-23b-3p	7.915 × 10−7	−46.185	Down
hsa-miR-130a-3p	1.273 × 10−4	−46.464	Down
hsa-let-7a-3p	2.261 × 10−6	−68.499	Down
hsa-miR-29a-3p	2.400 × 10−7	−76.306	Down
hsa-miR-335-5p	2.629 × 10−7	−85.763	Down
hsa-let-7i-5p	5.412 × 10−6	−87.029	Down
hsa-miR-23a-3p	1.409 × 10−8	−91.927	Down
hsa-miR-25-3p	2.822 × 10−8	−98.659	Down
hsa-miR-30e-3p	1.214 × 10−7	−101.670	Down
hsa-let-7g-5p	2.209 × 10−9	−159.702	Down
hsa-let-7f-5p	4.360 × 10−10	−210.653	Down
hsa-miR-21-5p	6.472 × 10−10	−211.721	Down
hsa-miR-24-3p	4.360 × 10−10	−411.862	Down

ASC-EVs, extracellular vesicles derived from adipose-derived mesenchymal stem cells.

FIG. 6.

Volcano plot showing the fold change and false discovery rate-adjusted p values for miRNA levels in ASC-derived EVs compared with ARPE-19-derived EVs. Blue dots indicate downregulated miRNAs and red dots indicate upregulated miRNAs. Only miRNAs with p values <0.05 and |log2 FC| ≥ 1 are highlighted as blue or red dots. ASC-EVs, extracellular vesicles derived from adipose-derived mesenchymal stem cells; EVs, extracellular vesicles.

Supplementary Table S2 presents a comprehensive inventory of experimentally validated MTIs for miRNAs that exhibit elevated expression in ASC-EVs relative to ARPE-19-derived EVs, focusing exclusively on those with robust experimental evidence. The interaction between miR-196a-5p, the most strongly upregulated miRNA in ASC-EVs, and the expression of HMOX1 was evidenced through various techniques, including reporter assays, We stern blotting, and quantitative PCR. ¹³ Table 2 and Supplementary Table S3 present the results of the functional enrichment analysis for miRNAs that are overexpressed and underexpressed in ASC-EVs compared with ARPE-19-derived EVs. Among these pathways, the “pathways in cancer” (hsa05200) pathway was found to be involved in the Nrf2 signaling pathway.

Table 2.

Kyoto Encyclopedia of Genes and Genomes Pathways Regulated by miRNAs Overexpressed in ASC-EVs Relative to ARPE-19-Derived EVs

Pathway	miRNAs	p value
Fatty acid biosynthesis (hsa00061)	hsa-miR-424-5p, hsa-miR-1306-5p	<1 × 10−325
ECM-receptor interaction (hsa04512)	hsa-miR-145-5p, hsa-miR-23a-5p	<1 × 10−325
Proteoglycans in cancer (hsa05205)	hsa-miR-196a-5p, hsa-miR-145-5p, hsa-miR-424-5p	4.743 × 10−13
Lysine degradation (hsa00310)	hsa-miR-196a-5p, hsa-miR-424-5p, hsa-miR-1260a, hsa-miR-1306-5p	1.304 × 10−10
Adherens junction (hsa04520)	hsa-miR-196a-5p, hsa-miR-145-5p, hsa-miR-424-5p, hsa-miR-1260a	6.570 × 10−9
Fatty acid metabolism (hsa01212)	hsa-miR-424-5p, hsa-miR-1306-5p	5.628 × 10−7
TGF-beta signaling pathway (hsa04350)	hsa-miR-196a-5p, hsa-miR-145-5p, hsa-miR-23a-5p, hsa-miR-424-5p	1.627 × 10−5
Hippo signaling pathway (hsa04390)	hsa-miR-196a-5p, hsa-miR-145-5p, hsa-miR-424-5p, hsa-miR-1260a	2.952 × 10−5
Pathways in cancer (hsa05200)	hsa-miR-196a-5p, hsa-miR-145-5p, hsa-miR-424-5p	9.953 × 10−5
Steroid biosynthesis (hsa00100)	hsa-miR-23a-5p	6.290 × 10−4
Colorectal cancer (hsa05210)	hsa-miR-196a-5p, hsa-miR-145-5p, hsa-miR-424-5p	7.320 × 10−4
Transcriptional misregulation in cancer (hsa05202)	hsa-miR-1306-5p, hsa-miR-145-5p	0.008
Signaling pathways regulating pluripotency of stem cells (hsa04550)	hsa-miR-145-5p, hsa-miR-424-5p	0.011
Focal adhesion (hsa04510)	hsa-miR-145-5p	0.013
p53 signaling pathway (hsa04115)	hsa-miR-196a-5p, hsa-miR-424-5p	0.025
Chronic myeloid leukemia (hsa05220)	hsa-miR-196a-5p, hsa-miR-424-5p	0.033
Glioma (hsa05214)	hsa-miR-145-5p, hsa-miR-424-5p, hsa-miR-196a-5p	0.047

ASC-EVs, extracellular vesicles derived from adipose-derived mesenchymal stem cells; EVs, extracellular vesicles.

These findings suggest that the differentially expressed miRNAs in ASC-EVs modify target mRNAs involved in the activation of the Nrf2 signaling pathway. This modulation may help alleviate apoptosis and oxidative stress in ARPE-19 cells.

Intravitreal injection of ASC-EVs delays retinal degeneration in RCS rats

Based on histological and ERG studies, we investigated the potential of ASC-EVs to protect retinal cells in RCS rats, a well-established animal model for retinal degeneration. ¹⁴ Figure 7A illustrates the scotopic ERG response at 3.15 cd·s/m² in a representative case. Before the injection, no significant differences were noted between the two eyes. As the retina degenerated, the amplitude was decreased in both eyes 7 days post-injection; however, the ASC-EV-injected eyes showed a more preserved waveform. Fourteen days post-injection, the ERG amplitudes in both eyes were reduced, and the waveforms were almost undetectable.

FIG. 7.

Scotopic 3.15 ERG responses before, 7 days, and 14 days after intravitreal injection of PBS or ASC-derived EVs in RCS rats. (A) Representative ERG waveforms. Graphs show the (B) a-wave and (C) b-wave amplitudes of ERG responses. ASC-EVs, extracellular vesicles derived from adipose-derived mesenchymal stem cells; ERG, electroretinography; PBS, phosphate-buffered saline; RCS, Royal College of Surgeons.

Figure 7B and C shows the ERG a- and b-wave amplitudes for both groups, demonstrating a marked reduction in a- and b-wave amplitudes in the PBS-injected eyes. In contrast, the a- and b-wave amplitudes were significantly preserved in the ASC-EV-injected eyes at 7 days post-injection (p = 0.028 and 0.021, respectively), but this protective effect diminished by 14 days (p = 0.624 and 0.401, respectively).

Figure 8A and B presents representative images of retinal histology comparing both eyes of a rat 14 days after intravitreal injection. One eye received PBS, while the other eye was treated with ASC-EVs. In the retina of the eye treated with ASC-EVs, the outer nuclear layer (ONL) and the inner and outer segments of the photoreceptor layer (I/OS PR) were significantly preserved. Figure 8C shows the thickness measurements of the ONL and I/OS PR for both groups. It demonstrates that the ASC-EV-treated eye exhibited a significantly greater ONL thickness at distances of 0.25 and 1.25 mm from the optic nerve (p = 0.028 for both), as well as a significantly increased I/OS PR thickness at distances of 0.75 and 1.25 mm from the optic nerve (p = 0.028 for both).

FIG. 8.

Retinal histology analysis performed 14 days after intravitreal injection of PBS or ASC-derived EVs in RCS rats. (A) Representative image compar ing both eyes of a rat after intravitreal injection of PBS in one eye and ASC-EVs in the other. (B) Magnified image of the middle region of the retina. (B) Graphs depicting the thicknesses of the ONL and the I/OS PR. ASC-EVs, extracellular vesicles derived from adipose-derived mesenchymal stem cells; INL, inner nuclear layer; ONL, outer nuclear layer; PBS, phosphate-buffered saline; I/OS PR, inner and outer segments of photoreceptor layer; RCS, Royal College of Surgeons; RPE, retinal pigments epithelium.

This result illustrates that ASC-EVs contribute to the maintenance of photoreceptor viability over a defined time frame.

Activation of the Nrf2 signaling pathway in RCS rats following intravitreal injection of ASC-EVs

Figure 9 shows representative images of retinal immunohistochemistry from both eyes of a rat taken 7 days after intravitreal injection. One eye received PBS, while the other eye was treated with ASC-EVs. Figure 9A shows that in the inner nuclear layer (INL) of the PBS-injected eye, cytoplasmic Nrf2 expression is notably higher compared with nuclear Nrf2 expression. In the eye injected with ASC-EVs, there was a notable decrease in cytoplasmic Nrf2 expression in the INL and an increase in nuclear Nrf2 expression in both the INL and ONL compared with the PBS-injected eye. This pattern suggests a nuclear translocation of Nrf2 in response to ASC-EV treatment.

FIG. 9.

Representative images of retinal immunohistochemistry taken 7 days after intravitreal injection of PBS or ASC-derived EVs in RCS rats, comparing the eye injected with PBS to the eye treated with ASC-derived EVs. Retinal sections were stained with antibodies against (A) Nrf2 (green), (B) HO-1 (green), and (C) NQO1 (green), and all sections were counterstained with Hoechst 33342 (blue). ASC-EVs, extracellular vesicles derived from adipose-derived mesenchymal stem cells; GCL/NFL, ganglion cell layer/nerve fiber layer; INL, inner nuclear layer; IPL, inner plexiform layer; ONL, outer nuclear layer; PBS, phosphate-buffered saline; RCS, Royal College of Surgeons.

Figure 9B and C demonstrates elevated expression levels of HO-1 and NQO1 in the inner retina of the ASC-EV-treated eye compared with the PBS-injected eye. These findings suggest that intravitreal injection of ASC-EVs induces activation of the Nrf2 signaling pathway in the retina of RCS rats.

Discussion

This study provides new evidence on the protective effects of ASC-EVs against oxidative stress in ARPE-19 cells, demonstrating that these effects are mediated through the activation of the Nrf2 signaling pathway by EV-derived miRNAs. Additionally, ASC-EVs transiently preserved photoreceptor function and delayed photoreceptor cell loss in RCS rats by activating the Nrf2 signaling pathway.

Nrf2 regulates both the baseline and stress-induced expression of antioxidant response element (ARE)–dependent genes, thereby managing the physiological and pathological responses to oxidative stress. ¹⁵ Under normal conditions, Nrf2 is sequestered in the cytoplasm and maintained at low levels through its binding to Kelch-like ECH-associated protein 1 (Keap1). However, under conditions of increased oxidative stress, Nrf2 dissociates from Keap1 and translocates to the nucleus, where it initiates the transcription of ARE-dependent genes that encode antioxidant enzymes, such as HO-1, NQO1, SOD, and enzymes involved in GSH metabolism. ¹⁶ Nrf2 serves as a key regulator of the antioxidative defense system in the retina, playing an essential role in shielding retinal cells from ROS-induced damage, particularly in response to aging, sunlight exposure, and inflammation.

Exposure to high ROS levels causes excessive free radical accumulation, leading to the oxidation of lipids, nucleic acids, and proteins, which eventually results in cellular dysfunction and death of retinal cells, ultimately resulting in retinal degeneration. Thereby, dysfunction in Nrf2 signaling is closely associated with retinal degeneration. Nrf2 expression was decreased in dysmorphic RPE cells relative to morphologically normal RPE cells in AMD. ¹⁷ In DR, the increased binding of Nrf2 to Keap1 impairs its translocation to the nucleus, leading to decreased levels of GSH. ¹⁸ As a result, the Nrf2 signaling pathway is emerging as a promising new target for the treatment of retinal degeneration.

Several studies have shown that MSC-derived EVs, including ASC-EVs, exert antioxidant effects via the Nrf2 signaling pathway. ^{19 –27} Although few studies have elucidated the exact mechanism, it is hypothesized that MSC-derived EVs activate the Nrf2 signaling pathway in target cells through the transfer of miRNAs. ^{28 –30} Nonetheless, no previous study has identified specific miRNAs within MSC-derived EVs that directly regulate and activate the Nrf2 signaling pathway in retinal cells.

EV-mediated communication between RPE cells in damaged and healthy regions is considered a mechanism for the local spread of RPE pathology in retinal degeneration, known as the bystander effect. ³¹ Therefore, when establishing an in vivo model for retinal degeneration using ARPE-19 cells, it is essential to consider this EV-mediated bystander effect. ARPE-19 cells treated with ASC-derived EVs will be affected by both ASC-derived and ARPE-19-derived EVs, whereas control cells will be influenced solely by ARPE-19-derived EVs. Consequently, the differences in miRNA profiles between ASC-derived and ARPE-19-derived EVs are likely responsible for the therapeutic effects mediated by ASC-EVs.

Our study revealed that ASC-EVs exhibited the upregulation of 11 miRNAs, with miR-196a-5p being the most significantly upregulated, in comparison to ARPE-19-derived EVs. MiR-196a-5p suppresses ubiquitin-specific peptidase 15, ³² which negatively regulates Nrf2 by deubiquitinating Keap1, thereby indirectly upregulating Nrf2. ³³ Furthermore, miR-196a directly targets and inhibits Bach1, a transcriptional repressor of HMOX1, resulting in the indirect upregulation of HO-1. ¹³ Among the miRNAs in ASC-derived EVs, miR-145-5p, which is the second most significantly upregulated, inhibits the NF-κB and JNK pathways, thereby exerting neuroprotective effects. ³⁴ Inhibition of NF-κB mitigates ROS-induced damage in the RPE and photoreceptor cells. ³⁵ Additionally, miR-145-5p enhances SOD activity while decreasing ROS and malondialdehyde levels in H9C2 cells under oxygen- and glucose-deprivation/reperfusion conditions by inhibiting NADPH oxidase 1. ³⁶ These findings may be attributed to the protective effects of ASC-derived EVs against H₂O₂-induced oxidative stress in ARPE-19 cells and RCS rats, as demonstrated by our study.

RCS rats harbor a mutation in the MerTK gene, making them the most suitable animal model for studying RP caused by this mutation. However, RCS rats are among the oldest and most extensively studied animal models of retinal degeneration, ¹⁴ providing a well-characterized understanding of their genetic profiles, pathology, and disease progression. RCS rats exhibit impaired phagocytosis of shed outer segments from rod cells by RPE cells. This results in the accumulation of oxidative DNA damage in photoreceptor cells, especially within their nuclei, which leads to cell death. ³⁷ By week 3, the photoreceptor outer segments are disrupted. By week 4, degeneration of the ONL begins, with an increase in pyknotic cells. At week 7, ∼50% of the photoreceptor nuclei have degenerated. By week 12, the ONL is reduced to a single layer of photoreceptor cell bodies. ³⁸ Additionally, RCS rats offer a size advantage over mouse models, facilitating easier surgical procedures. The availability of immunocompromised strains, such as the Foxn1^mu immunocompromised RCS rat, further enhances their utility. In addition, as naturally occurring animal models, RCS rats exhibit consistent and predictable disease progression, offering significant advantages for research. These factors justify the choice of RCS rats for this study. However, future studies should verify the results of this research using various animal models of retinal degeneration.

Our retinal immunohistochemical analysis demonstrated that intravitreal injection of ASC-EVs activates the Nrf2 signaling pathway in the retina. In the eyes treated with ASC-EVs, there was an increased expression of nuclear Nrf2 in the ONL. Since the ONL contains the nuclei of photoreceptor cells, this may explain the mechanism behind the delayed degeneration of photoreceptor cells following intravitreal injection of ASC-EVs. This finding is consistent with our retinal histology results, which showed that eyes receiving ASC-EV treatment demonstrated significantly preserved ONL and I/OS PR structures compared with eyes receiving PBS injections. Moreover, our ERG results corroborate this finding, showing that the ERG a- and b-wave amplitudes were significantly preserved in the ASC-EV-treated eyes compared with those that received PBS injections. The a-wave reflects the activity of photoreceptor cells and is commonly used as an indicator of their functional status. In contrast, the b-wave response is primarily driven by bipolar cells. However, reduced signal transmission from photoreceptor cells to bipolar cells leads to attenuation of the b-wave, rendering it a functional indicator of photoreceptor cell health. ³⁹

Additionally, the eyes treated with ASC-EVs exhibited decreased cytoplasmic and increased nuclear Nrf2 expression in the INL. The INL comprises the cell bodies of horizontal cells, bipolar cells, amacrine cells, and Müller cells, which collectively constitute the inner retinal layer. This finding is thought to be associated with the increased expression of HO-1 and NQO1 observed in the inner retina in this study. This result is considered plausible because ASC-derived EVs were injected into the vitreous cavity, which is anatomically closer to the inner retina than to the photoreceptors. Consistent with this, the ERG b-wave, which is primarily driven by bipolar cells, was significantly preserved in the eyes injected with ASC-derived EVs compared with those receiving PBS injections. However, further research is needed to elucidate the mechanism by which the increased expression of HO-1 and NQO1 in these cells affects photoreceptor cell viability.

The transient effectiveness of ASC-derived EVs in RCS rats may be attributable to their rapid clearance from the vitreous cavity. ⁴⁰ Additionally, the bioactive components within EVs may exhibit a limited duration of activity, resulting in an initial therapeutic peak followed by a subsequent decline in efficacy. ⁴¹ To prolong the therapeutic benefits of intravitreal injections of ASC-EVs, implementing an optimized dosing regimen with more frequent injections may be necessary to sustain effective levels. Additionally, engineering EVs to include specific drug payloads may enhance their therapeutic effects and longevity.

Our study has some limitations. First, ARPE-19 cells are known to lose their specialized phenotype after multiple passages, which may limit their ability to accurately represent in vivo conditions. It has been documented that low-passage ARPE-19 cells can express genes specific to native human RPE cells when properly cultured and differentiated. ⁴² We obtained ARPE-19 cells at passage number 19 and utilized only those below passage 25 in our experiments to minimize the risk of losing their ability to differentiate. Nonetheless, to validate our findings and better simulate in vivo conditions, further experiments should be conducted using primary human RPE cells from donor eyes, RPE cells differentiated from stem cells, or ARPE-19 cells treated with nicotinamide. Second, this study focused solely on the impact of ASC-EVs on the Nrf2 pathway. However, the increased viability of ARPE-19 cells under oxidative stress and the temporary delay in retinal degeneration in RCS rats may involve other pathways. Future research should employ methods like bulk RNA sequencing to investigate the additional effects of ASC-EVs on retinal degeneration. Third, we were unable to evaluate the impact of intravitreal ASC-derived EV injection on RPE cells in RCS rats. This limitation should be addressed in future studies by employing ERG protocols specific to RPE, such as the c-wave response, as well as by conducting histological analyses or RT-qPCR directly on RPE tissue.

Conclusions

We presented novel findings regarding the protective effects of ASC-derived EVs against oxidative stress in ARPE-19 cells, which was facilitated by the activation of the Nrf2 signaling pathway by EV-miRNAs. Moreover, ASC-EVs delayed retinal degeneration in RCS rats by activating the Nrf2 signaling pathway. Our results offer new insights into the development of therapeutic strategies for retinal degeneration.

Authors’ Contributions

J.S.H.: Investigation, formal analysis, and writing—original draft. H.B.S.: Investigation, writing—original draft, and writing—review and editing. G.L.: Conceptualization, methodology, supervision, and writing—review and editing. S.J.: Conceptualization, methodology, supervision, and writing—review and editing. D.J.M.: Conceptualization, methodology, investigation, formal analysis, funding acquisition, writing—original draft, and writing—review and editing.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This research was supported by the Hallym University Research Fund 2022 (HURF-2022–15) (D.J.M.), the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (grant no. 2021R1G1A1093576) (D.J.M.), the U.S. National Institutes of Health grant (R35GM147513) (S.J.), and the Maryland Stem Cell Research Fund (MSCRFL-6109) (S.J.).