Engineered MEVs for photoreceptor-targeted delivery of USP25 to alleviate diabetic retinopathy

Yaoxiang Sun; Shenyuan Chen; Xiaoyuan Qi; Yuntong Sun; Zhengmei Jiang; Jie Chang; Yongjun Ma; Jin Huang; Benshuai You; Fengtian Sun

doi:10.1186/s12951-025-03671-w

. 2025 Aug 20;23:575. doi: 10.1186/s12951-025-03671-w

Engineered MEVs for photoreceptor-targeted delivery of USP25 to alleviate diabetic retinopathy

Yaoxiang Sun ^1,^#, Shenyuan Chen ^1,^#, Xiaoyuan Qi ^1,^5,^#, Yuntong Sun ³, Zhengmei Jiang ^1,⁵, Jie Chang ³, Yongjun Ma ³, Jin Huang ^1,^4,^✉, Benshuai You ^2,^✉, Fengtian Sun ^3,^✉

PMCID: PMC12366390 PMID: 40830790

Abstract

Diabetic retinopathy (DR) is the major cause of vision decline in adults worldwide. Photoreceptor loss is considered a main pathogenesis of retinal dysfunction in DR. Recently, mesenchymal stem cell-derived extracellular vesicles (MEVs) treatment has been considered a promising cell-free approach for retinal disorders. However, the role and mechanism of MEVs in alleviating photoreceptor injury in DR remain unclear. In this study, MEV treatment improved retinal function and inhibited photoreceptor apoptosis in db/db mice. Mechanistically, the deubiquitinating enzyme ubiquitin-specific peptidase 25 (USP25) in MEVs was responsible for the MEV-mediated photoreceptor therapy by inhibiting hyperglycemia-induced αA-crystallin (CRYAA) ubiquitination. Moreover, USP25-enriched MEVs modified with the photoreceptor-targeting peptide MH42 (MEVs^MH42-USP25) were prepared by genetic engineering and surface conjugation. MEVs^MH42-USP25 exhibited elevated repairing efficiency to attenuate retinal dysfunction and photoreceptor loss in db/db mice. Our study develops an MEV-based nanocarrier for photoreceptor-targeted delivery and highlights the effectiveness of MEVs^MH42-USP25 as novel therapeutics for DR.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-025-03671-w.

Keywords: Photoreceptor, Extracellular vesicles, Diabetic retinopathy, CRYAA, USP25

Background

As a prevalent complication of diabetes, diabetic retinopathy (DR) is the leading cause of visual impairment in adults worldwide [1]. Hyperglycemia-induced photoreceptor loss is recognized as a key contributor to retinal dysfunction in DR [2]. Although current treatment options, such as laser therapy, antiangiogenic drug injection, and vitrectomy, have achieved encouraging progress, limited visual improvement and unfavorable side effects remain [3, 4]. Thus, the pathophysiology clarification of photoreceptor injury in DR and the development of effective therapeutic strategies are urgently needed.

Extracellular vesicles (EVs) are a type of nanoscale particles secreted by living cells [5]. By participating in intercellular communication, EVs are involved in various biological processes, including retinal homeostasis regulation [6, 7]. Increasing studies have suggested that the therapeutic effects of mesenchymal stem cell (MSC) transplantation are largely mediated by EV release [8, 9]. Recently, MSC-derived EVs (MEVs) have emerged as a novel approach for retinal disease therapy by delivering proteins, nucleic acids, and lipids [10]. For instance, MEVs exhibit neuroprotective and anti-inflammatory roles in retinal ischemic disorders [11]. Our previous study reveals that MEV-delivered NEDD4 mitigates hyperglycemia-induced retinal oxidative damage by regulating the PTEN/AKT/NRF2 axis [12]. However, whether MEVs can improve photoreceptor injury-induced retinal dysfunction in DR remains unclear.

Accumulating studies have demonstrated the potential of EVs as promising and secure carriers in tissue regeneration [13, 14]. Relative to synthetic nanotools, EVs exhibit many favorable characteristics, such as biocompatibility, stability, low immunogenicity, and biological barrier permeability [15, 16]. Owing to their natural origin, EVs can escape phagocytosis and exhibit an extended half-life [17]. For instance, EV-based inhalable vaccines show broader distribution and extended retention on the respiratory mucosa [18]. Moreover, the conjugation of membrane structures with functional ligands or peptides enables EVs with cell-specific targeting [19]. Through bioengineering, the characteristics of EVs can be optimized to enhance their therapeutic efficiency [20]. Nevertheless, the MEV-based strategy for photoreceptor-targeted therapy in DR has not yet been established.

In this study, we evaluated the therapeutic role of MEVs in retinal dysfunction in db/db mice. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) and gene knockout mice were used to explore the potential mechanism of MEV-mediated photoreceptor therapy. We found that MEVs alleviated photoreceptor injury by ubiquitin-specific peptidase 25 (USP25)-mediated αA-crystallin (CRYAA) deubiquitination. Based on lentivirus-mediated genetic manipulation and CP05 peptide-induced combination with the EV membrane, USP25-enriched MEVs modified with photoreceptor-targeting peptide MH42 (MEVs^MH42-USP25) were constructed. MEVs^MH42-USP25 allowed the efficient delivery of USP25 to photoreceptors, thereby ameliorating diabetic retinal dysfunction. Our findings establish an MEVs^MH42-USP25-based photoreceptor-targeted strategy for DR therapy.

Methods

Ethics

This research was approved by the Medical Ethics Committee of Affiliated Jinhua Hospital of Zhejiang University (JHYY202404).

Antibodies

The primary antibodies utilized were as follows: anti-rhodopsin (IF: 1:100, 30438-1-AP, Proteintech, China), anti-s-opsin (IF: 1:200, ab235274, Abcam, USA), anti-CRYAA (IF: 1:500, WB: 1:1000, ab181866, Abcam, USA), anti-USP25 (IF: 1:200, sc-398414, SANTA CRUZ, USA; WB: 1:2000, ab187156, Abcam, USA), anti-CD9 (WB: 1:2000, ET1601-9, HUABIO, China), anti-CD63 (WB: 1:1000, ET1607-2, HUABIO, China), anti-HSP90 (WB: 1:2000, ET1605-56, HUABIO, China), anti-TSG101 (WB: 1:2000, 28283-1-AP, Proteintech, China), anti-calnexin (WB: 1:2000, 10427-2-AP, Proteintech, China), anti-PCNA (WB: 1:1000, ab26, Abcam, USA), anti-Bcl-2 (WB: 1:1000, HA721235, HUABIO, China), anti-Bax (WB: 1:20000, ET1603-34, HUABIO, China), anti-β-actin (WB: 1:10000, 4970, CST, USA), anti-ubiquitin (WB: 1:2000, ET1609-21, HUABIO, China), anti-HA (WB: 1: 3000, AE036, ABclonal, China), anti-His (WB: 1: 5000, 66005-1-Ig, Proteintech, China), and anti-MYC (WB: 1:5000, 60003-2-Ig, Proteintech, China).

Characterization of MSCs

MSCs were separated from fresh umbilical cords as previously described [21]. After osteogenic and adipogenic inductions, MSCs were treated with Alizarin Red S and Oil Red O to evaluate their multiple differentiation ability. The MSC markers such as CD11b, CD34, CD45, CD73, CD105, and CD166 were determined by flow cytometry.

Extraction and characterization of MEVs

MEVs were extracted from the supernatants of MSCs maintained in serum-free medium (CM-SC01, Pricella, China) by ultracentrifugation as previously described [22]. Transmission electron microscopy (TEM) and nanoparticle flow cytometry (NanoFCM) were used to detect the structure and particle size of MEVs. Western blot was applied to determine the EV markers including CD9, CD63, HSP90, TSG101, and calnexin.

Cell culture, treatment and transfection

In vitro, 661 W cells were maintained in low glucose (LG) medium. To mimic diabetic damage, 661 W cells were seeded in each plate at a density of 5 × 10⁴ cells/mL and stimulated with high glucose (HG) medium (30 mM), followed by the treatment with MEVs (1 × 10⁶, 1 × 10⁷, or 1 × 10⁸ particles/mL) for 24 h.

661 W cells with 70% confluency were transfected with 100 pmol of siRNA or 2.5 µg of plasmid using 5 µL of Lipo8000 (C0533, Beyotime, China) in serum-free condition. After 5 h, 661 W cells were treated with complete HG medium. The siRNA sequences were as follows:

siUSP25-1: sense: AAUUCCAAGUUUCCAUUACUA.

siUSP25-1: antisense: GUAAUGGAAACUUGGAAUUAG.

siUSP25-2: sense: UCAAGAACUCUGCUAAUGGCU.

siUSP25-2: antisense: CCAUUAGCAGAGUUCUUGAAG.

siRNA NC: sense: UUCUCCGAACGUGUCACGUTT.

siRNA NC: antisense: ACGUGACACGUUCGGAGAATT.

USP25 knockdown in MSCs

Lentiviral vectors with the USP25 short hairpin RNA (shRNA) were synthesized by Sangon Biotech (Shanghai, China). MSCs were transfected with the recombinant lentivirus using Lipo8000 (C0533, Beyotime, China). MEVs^shUSP25 were purified from the supernatant of MSCs^shUSP25 by ultracentrifugation. The shRNA sequences were as follows:

shRNA NC: TTCTCCGAACGTGTCACGT.

USP25 shRNA: ACTTCTCCTGTTGACGATA.

Preparation of MEVs^MH42-USP25

HEK293T cells were transfected with the lentivirus encoding USP25 which was synthesized by Sangon Biotech (Shanghai, China). Subsequently, MSCs were transfected with the recombinant lentivirus. MEVs-USP25 were purified from the serum-free supernatant of MSCs-USP25 by ultracentrifugation. The MH42 peptide sequence was SPALHFLGGGSC. The MH42-CP05 fusion peptide sequence was SPALHFLGGGSCCRHSQMTVSRL. The scrambled peptide sequence was LGSGFPGLSHAC. The scrambled CP05 fusion peptide sequence was LGSGFPGLSHACCRHSQMTVSRL. MEVs-USP25 (10 µg) were mixed with the fusion peptides (40 µg) at 4 °C for 16 h. Unbound peptides were then collected by ultrafiltration. MEVs^MH42-USP25 and scrambled CP05 fusion peptide-conjugated MEVs-USP25 (MEVs^Scrbl-USP25) were extracted by ultracentrifugation, and dissolved in PBS. TEM, NanoFCM and Western blot were used for the characterization of MEVs^MH42-USP25. The protein expression of USP25 in MEVs, MEVs-USP25, MEVs^MH42, and MEVs^MH42-USP25 was verified by Western blot. An ELISA kit (ML-E-23571, Enzyme-Linked Biotechnology Co., Ltd., China) was used to quantify the total USP25 protein concentration after lysis. The USP25 protein level per particle was calculated using the following equation:

To assess conjugation efficiency, the fluorescence of mixtures containing FITC-labeled peptides and EVs was determined. After combination, the fluorescence of unbound FITC-labeled peptides was detected. The conjugation efficiency was quantified using the following equation:

One-way ANOVA with Tukey’s multiple comparison test was used to compare the conjugation efficiency among groups. The final solution was collected and analyzed by flow cytometry.

In vivo distribution of mevs, MEVs-USP25, MEVs^MH42, and MEVs^MH42-USP25

To evaluate the distribution of MEVs, MEVs-USP25, MEVs^MH42, and MEVs^MH42-USP25 in retinal tissues, peptides were labeled with FITC and EVs were labeled with PKH26. Following intravitreal injection, db/db mice were treated with MH42-CP05 fusion peptides, MEVs, MEVs-USP25, MEVs^MH42, and MEVs^MH42-USP25, respectively. After 24 h, the retinal tissues were snap-frozen, cut into 10 μm sections, incubated with anti-rhodopsin or anti-s-opsin antibody overnight, counterstained with DAPI solution for 15 min, and observed using a confocal microscope.

Animals

db/db and db/m mice were purchased from Cavens Laboratory (Changzhou, China). To evaluate the retinal therapeutic effects of MEVs, db/db mice were intravitreally injected with 1 µL of MEVs (1 × 10⁶, 1 × 10⁷, or 1 × 10⁸ particles) every 4 weeks. To investigate whether MEVs alleviate photoreceptor injury by delivering USP25, db/db mice were intravitreally injected with 1 µL of MEVs (1 × 10⁷ particles) and 1 µL of MEVs^shUSP25 (1 × 10⁷ particles) every 4 weeks, respectively. To explore the proper intervention dose of MEVs^MH42-USP25, db/db mice were intravitreally injected with 1 µL of MEVs^MH42-USP25 (1 × 10⁶, 1 × 10⁷, or 1 × 10⁸ particles) every 4 weeks. To evaluate the therapeutic efficiency of MEVs^MH42-USP25, db/db mice were intravitreally injected with 1 µL of MEVs, MEVs-USP25, MEVs^MH42, MEVs^MH42-USP25, and MEVs^Scrbl-USP25 (1 × 10⁷ particles) every 4 weeks, respectively. To compare the therapeutic effects of MEVs-USP25 and MEVs^MH42-USP25 containing equal USP25 protein content, db/db mice were intravitreally injected with 1 µL of MEVs-USP25 and MEVs^MH42-USP25 (100 pg of USP25 protein level) every 4 weeks, respectively. At week 24, all mice were sacrificed using 4% isoflurane, and retinal tissues were isolated.

Wild type (WT) and CRYAA knockout (CRYAA^KO) mice were purchased from Cyagen Bioscience (Suzhou, China) and induced to diabetic model by high-fat diet (HFD) feeding plus injection with streptozotocin (STZ; 30 mg/kg per day; Sigma-Aldrich, MO) intraperitoneally for 7 consecutive days, as in previous studies [23]. Two weeks after STZ injection, CRYAA^KO mice were treated with adeno-associated virus serotype 9 carrying CRYAA (AAV9-CRYAA) or AAV9-null via subretinal injection. Subsequently, WT and CRYAA^KO mice were intravitreally injected with 1 µL of MEVs (1 × 10⁷ particles) every 4 weeks. Twenty-four weeks after STZ administration, all mice were sacrificed with 4% isoflurane, and retinal tissues were isolated.

Electroretinogram (ERG) analysis

After dark-adaption overnight, db/m and db/db mice were intraperitoneally injected with ketamine (85 mg/kg) and xylazine (15 mg/kg) under dim red light. The corneas were treated with 0.5% proxymetacaine. The ground electrode was placed on the tail. The recording electrode was placed on the cornea. The reference electrode was placed on the mouth. A white flash (3.0 cd.s/m²) was applied for inducing this experiment.

Hematoxylin and Eosin (HE) staining

After enucleation, the eyeballs were fixed in 4% paraformaldehyde for 24 h and embedded in paraffin. The retinas were cut into 4 μm sections and treated with HE staining to detect retinal structure and outer nuclear layer (ONL) thickness.

Terminal deoxynucleotidyl transferase dUTP Nick end labeling (TUNEL) staining

The apoptotic cells in retinal tissues were detected using the TUNEL Kit (C1089, Beyotime, China) according to the manufacturers’ instructions. Following counterstaining with DAPI solution for 15 min, the retinal tissues were observed under a confocal microscope.

Immunofluorescence staining

After deparaffinization and antigen retrieval, the retinal sections were blocked with 5% bovine serum albumin (BSA) solution for 30 min and stained with the primary antibodies overnight at 4 °C. Following the treatment with FITC/Cy5-conjugated secondary antibodies for 2 h and DAPI solution for 20 min, the retinal tissues were photographed using a confocal microscope.

LC-MS/MS

The protein samples were isolated from the 661 W cells after MEV treatment, subjected to digestion and analyzed by LC-MS/MS. In addition, 661 W cells after MEV treatment were incubated with co-immunoprecipitation (Co-IP) solution and treated with CRYAA antibody (1:200, c-69687, SANTA CRUZ, USA), followed by the LC-MS/MS to detect the proteins interacted with CRYAA. The detailed LC-MS/MS analysis is described in Supplementary Methods.

Western blot

The proteins were extracted using radioimmunoprecipitation assay buffer (EA0002, SparkJade, China), exposed to SDS/PAGE and transferred onto polyvinylidene difluoride membranes. The membranes were blocked in 5% skim milk for 1 h and then treated with primary antibodies overnight at 4 °C. Following the incubation with HRP-conjugated secondary antibodies (ab6721/ab6728, Abcam, USA) for 1 h, the bands were detected by the enhanced chemiluminescence reagent.

Co-IP assay

After Co-IP solution (P0013, Beyotime, China) incubation, cells were treated with the CRYAA antibody (1:200, c-69687, SANTA CRUZ, USA) and IgG (negative control) with rotation for 24 h, respectively. After the combination with magnetic beads for 6 h, the samples were determined by Western blot.

EdU staining

661 W cells were maintained in fresh medium supplemented with EdU working reagent for 3 h and fixed in 4% paraformaldehyde for 25 min. After permeabilization with 0.1% Triton X-100 for 20 min, 661 W cells were treated with the EdU Assay Kit (ab219801, Abcam, USA) for 30 min, incubated with DAPI buffer for 15 min and photographed using a microscope.

Cell counting Kit-8 (CCK8) experiment

661 W cells were plated in 96-well plates (1500 cells/well) and incubated with 20 µL of CCK8 solution (ab228554, Abcam, USA) for 1 h. The absorbance was determined at 450 nm by an enzyme-linked immunosorbent plate assay reader every 24 h.

Safety evaluation in vivo

Aqueous humor and serum samples were collected from db/db mice at weeks 24 and 48. ELISA kits were applied to determine cytokine levels in the aqueous humor and serum, including IL-1β (E-EL-M0037, Elabscience, China), IL-6 (E-EL-M0044, Elabscience, China), IL-8 (SEKM-0046, Solarbio, China), and TNF-α (E-EL-M3063, Elabscience, China). The intraocular pressure of db/db mice at weeks 24 and 48 was measured using tonometers. Blood cell counts of db/db mice at weeks 24 and 48 were determined using hemocytometers. The histological characteristics of the heart, liver, spleen, lung and kidney were assessed after HE staining. MH42 peptides, commercial human CD63 protein (Ag19690, Proteintech, China), and human CD9 protein (ab152263, Abcam, USA) were diluted with coating solution, coated in 96-well EIA/RIA plates (Coning) overnight at 4 °C, and blocked with 200 µL of blocking solution for 2 h. Subsequently, serum samples were diluted 1:100 with 0.1% BSA, added into each well, and incubated for 1 h at 37 °C. After washing three times with PBST, the samples were treated with HRP-conjugated goat anti-mouse IgG antibody (1:2000) or HRP-conjugated goat anti-human IgG antibody (1:2000). TMB substrate was used for color development, which was then terminated using a termination solution. Absorbance was measured at 450 nm using an ELISA reader. Serum samples from untreated db/m mice were used as negative controls.

Statistical analysis

The statistical analysis was conducted using GraphPad Prism software (GraphPad, San Diego, USA). The results were shown as mean ± SEM. The two-tailed unpaired Student’s t-test was performed for comparison between two groups. One-way ANOVA with Tukey’s multiple comparison test was performed to determine the significance among multiple groups. P value < 0.05 was considered statistically significant.

Results

MEVs alleviate photoreceptor injury in db/db mice

MSCs were isolated from human umbilical cords using an established separation method. Microscopic images demonstrated the spindle-like morphology of MSCs (Supplementary Figure S1 A and B). After osteogenic and adipogenic medium induction, Alizarin Red S and Oil Red O staining confirmed the multiple differentiation potential of MSCs (Supplementary Figure S1 C and D). Flow cytometric analysis showed that MSCs were positive for the expression of CD73, CD105, and CD166, and negative for the expression of CD11b, CD34, and CD45 (Supplementary Figure S1 E). Subsequently, MEVs were extracted from the supernatant of MSCs by ultracentrifugation. TEM and NanoFCM results indicated that MEVs displayed a cup-shaped structure with an average particle size of 79.2 nm (Fig. 1A and B). Western blot demonstrated the presence of EV markers including CD9, CD63, HSP90, and TSG101 in MEVs (Fig. 1C). We then evaluated the distribution of MEVs after intravitreal injection. Biodistribution analysis revealed the localization of PKH26-labeled MEVs in various retinal layers for up to 3 weeks (Supplementary Figure S2), which was consistent with the previous study [24]. These results confirm the successful extraction of MEVs.

To investigate the role of MEVs in retinal dysfunction, db/db mice were treated with MEVs by intravitreal injection (Fig. 1D). ERG analysis was performed to detect retinal electrophysiological functions, including photopic and scotopic responses. Compared with db/m mice, db/db mice displayed impaired retinal function, as evidenced by the reduced amplitudes of photopic b-waves and scotopic a/b-waves. Notably, MEV treatment recovered the photoreceptor function of db/db mice (Fig. 1E-I). Histological analysis demonstrated that long-term hyperglycemia led to retinal degeneration and photoreceptor-nuclei-residing ONL loss, while MEVs preserved retinal integrity and improved ONL thickness (Fig. 1J and K). TUNEL staining revealed that MEVs significantly reduced photoreceptor apoptosis rates in the ONL (Fig. 1L and M). Consistently, immunofluorescence staining showed the loss of rhodopsin⁺ rods and s-opsin⁺ cones in db/db mice, whereas MEVs increased the number of rod and cone photoreceptors (Fig. 1N-Q). Notably, the retinal therapeutic efficiency of MEVs with 1 × 10⁷ particles was markedly higher than that of MEVs with 1 × 10⁶ particles, whereas no significant improvement was observed in photoreceptor therapy when the dose of MEVs was increased to 1 × 10⁸ particles, implying that 1 × 10⁷ particles is the optimal dose of MEVs. These results indicate that MEVs alleviate retinal dysfunction by inhibiting photoreceptor loss in DR.

MEVs inhibit HG-induced apoptosis of 661 W cells

Next, we assessed the role of MEVs in HG medium-treated 661 W cells. In vitro tracing experiments confirmed the internalization of PKH26-labeled MEVs by 661 W cells (Fig. 2A). CCK8 assay and EdU staining presented that the HG medium inhibited the proliferative ability of 661 W cells, whereas MEV treatment promoted the proliferation of 661 W cells (Fig. 2B-D). TUNEL staining demonstrated that MEVs effectively reduced the apoptotic rate of 661 W cells under HG conditions (Fig. 2E and F). Western blot analysis also indicated that MEVs upregulated PCNA and Bcl-2 expressions and downregulated Bax protein level in 661 W cells cultured in HG medium (Fig. 2G and H). These findings reveal that MEVs protect 661 W cells from HG stimulation.

Fig. 2 — Anti-apoptosis potential of MEVs in 661 W cells. (A) Internalization of PKH26-labeled MEVs by 661 W cells after co-incubation for 24 h. Scale bars, 50 μm. (B) CCK8 assay for the proliferation of 661 W cells after MEV treatment (n = 4). (C, D) EdU staining for the proliferation of 661 W cells after MEV treatment and the analysis of EdU positive rate (n = 5). Scale bars, 100 μm. (E, F) TUNEL staining for the apoptosis of 661 W cells after MEV treatment and the analysis of TUNEL positive rate (n = 5). Scale bars, 100 μm. (G, H) Western blot analysis for PCNA, Bcl-2 and Bax expression in 661 W cells after MEV treatment and the corresponding quantification of protein levels compared with the LG group (n = 3). No data were excluded from the analyses. Data collection and analysis were performed by investigators blinded to the experimental groups. n represents independent biological replicates. Data in (B), (D), (F), and (H) are presented as mean ± SEM and were compared using one-way ANOVA with Tukey’s multiple comparison test. ns, not significant, *P < 0.05, **P < 0.01, and ***P < 0.001

MEVs ameliorate hyperglycemia-induced photoreceptor injury by upregulating CRYAA expression

To explore the therapeutic mechanism of MEVs, we detected the differentially expressed proteins in untreated and MEV-treated 661 W cells under HG conditions using LC-MS/MS. The results indicated that MEV treatment markedly elevated CRYAA protein level (Fig. 3A). Previous studies have reported the retinal therapeutic role of CRYAA in experimental uveitis, DR, and retinal stress responses [25–27]. Therefore, we hypothesized that MEVs attenuate photoreceptor injury in DR by upregulating CRYAA expression. Immunostaining and Western blot confirmed that MEVs promoted CRYAA expression in 661 W cells (Fig. 3B-E). In vivo experiments further verified that MEV injection caused the upregulation of CRYAA in the ONL (Fig. 3F-I), suggesting that MEVs reverse the hyperglycemia-induced CRYAA inhibition in photoreceptors.

Fig. 3 — MEVs alleviate hyperglycemia-induced photoreceptor injury by upregulating CRYAA expression. (A) The proteins in 661 W cells cultured in HG medium with or without MEV treatment were detected using LC-MS/MS. (B, C) Representative CRYAA immunostaining images of 661 W cells after MEV treatment and the quantitative analysis of fluorescent intensity compared with the LG group on the basis of three sections per group with three images counted in each section. Scale bars, 25 μm. (D, E) Western blot analysis for the CRYAA protein expression in 661 W cells after MEV treatment and the corresponding quantification of protein level compared with the LG group (n = 3). (F, G) Representative retinal CRYAA immunostaining images of db/m and db/db mice after MEV treatment and the quantitative analysis of fluorescent intensity compared with the db/m mice group on the basis of three mice per group with three images counted in each mouse. Scale bars, 100 μm. (H, I) Western blot analysis for the retinal CRYAA protein expression in db/m and db/db mice after MEV treatment and the corresponding quantification of protein level compared with the db/m mice group (n = 3). (J) Schematic diagram displaying the experimental design of WT and CRYAA^KO mice. (K) Western blot to verify CRYAA knockout in the retinal tissues of CRYAA^KO mice. (L) Fasting blood glucose level testing of WT and CRYAA^KO mice at the indicated time points (n = 6). (M, N) Retinal HE staining and ONL thickness measurement in WT and CRYAA^KO mice after STZ and MEV treatment (n = 3). Scale bars, 100 μm. (O, P) TUNEL staining of retinas and the analysis of TUNEL⁺ cell percentage of total ONL cells in WT and CRYAA^KO mice after STZ and MEV treatment on the basis of three mice per group with three images counted in each mouse. Scale bars, 100 μm. (Q, R) Representative retinal rhodopsin immunostaining images of WT and CRYAA^KO mice after STZ and MEV treatment and the quantitative analysis of fluorescent intensity compared with the WT + STZ group on the basis of three mice per group with three images counted in each mouse. Scale bars, 100 μm. (S, T) Representative retinal s-opsin immunostaining images of WT and CRYAA^KO mice after STZ and MEV treatment and the quantitative analysis of fluorescent intensity compared with the WT + STZ group on the basis of three mice per group with three images counted in each mouse. Scale bars, 100 μm. No data were excluded from the analyses. Data collection and analysis were performed by investigators blinded to the experimental groups. n represents independent biological replicates. Data in (C), (E), (G), (I), (L), (N), (P), (R), and (T) are presented as mean ± SEM and were compared using one-way ANOVA with Tukey’s multiple comparison test. ns, not significant, **P < 0.01, and ***P < 0.001. GCL: ganglion cell layer; INL: inner nuclear layer; ONL: outer nuclear layer

To evaluate the role of CRYAA in photoreceptor damage in DR, 661 W cells were exposed to HG medium and treated with CRYAA-overexpressing plasmids. Western blot confirmed that CRYAA expression was markedly upregulated in 661 W cells after transfection (Supplementary Figure S3A). CCK8 assay and EdU staining demonstrated that CRYAA overexpression markedly elevated the proliferative ability of 661 W cells (Supplementary Figure S3B-D). Compared with the control group, transfection with CRYAA overexpression plasmids reduced the apoptotic rate of 661 W cells under HG condition (Supplementary Figure S3E and F). Previous studies have indicated that CRYAA can prevent cell death by inhibiting oxidative stress [28, 29]. Therefore, we evaluated retinal oxidative markers following CRYAA-overexpressing plasmid transfection and found that CRYAA overexpression efficiently reduced the levels of reactive oxygen species (ROS) and malondialdehyde (MDA) and elevated the levels of glutathione (GSH) and superoxide dismutase (SOD) in 661 W cells (Supplementary Figure S4A-E). These results indicate that CRYAA is an essential therapeutic target for photoreceptor injury in DR.

Subsequently, CRYAA^KO mice were used to determine whether MEVs alleviate photoreceptor damage through CRYAA upregulation. WT and CRYAA^KO mice were fed with HFD and intraperitoneally injected with STZ, followed by the intravitreal injection with MEVs (Fig. 3J). Western blot confirmed CRYAA knockout in the retinas of CRYAA^KO mice (Fig. 3K). Continuous blood glucose tests showed that WT and CRYAA^KO mice remained hyperglycemic throughout the experiment (Fig. 3L). HE staining images presented that MEVs alleviated hyperglycemia-induced retinal degeneration and elevated ONL thickness. In contrast, MEVs did not significantly increase ONL thickness in CRYAA^KO mice (Fig. 3M and N). TUNEL staining further demonstrated that CRYAA knockout impaired the anti-apoptotic effects of MEVs on photoreceptors (Fig. 3O and P). Moreover, MEVs exerted limited effects to upregulate the expression of rhodopsin and s-opsin in STZ-treated CRYAA^KO mice (Fig. 3Q-T). To further confirm the essential role of CRYAA in MEV-mediated photoreceptor therapy, we used AAV9-CRYAA to induce CRYAA re-expression in STZ-treated CRYAA^KO mice by subretinal injection, followed by MEV administration (Supplementary Figure S5A). Western blot confirmed CRYAA re-expression in the retinal tissues of CRYAA^KO mice after AAV9-CRYAA injection (Supplementary Figure S5B). We found that CRYAA re-expression restored MEV-mediated retinal therapeutic roles in DR, including ONL thickness enhancement and photoreceptor apoptosis inhibition, whereas AAV9-null injection had limited effects on the functional recovery of MEVs (Supplementary Figure S5C-K). These findings indicate that MEVs alleviate photoreceptor injury in DR by upregulating CRYAA expression.

MEVs transfer USP25 to inhibit CRYAA ubiquitination

Next, we investigated the molecular mechanism underlying MEV-mediated CRYAA upregulation in photoreceptors. We incubated 661 W cells with cycloheximide, a protein translation inhibitor, and measured CRYAA protein levels. Compared with the HG group, MEV treatment remarkably enhanced the half-life of CRYAA (Fig. 4A). Co-IP assay also presented that MEVs reversed HG-induced CRYAA ubiquitination (Fig. 4B), indicating that MEVs promote CRYAA expression through a ubiquitin proteasome mechanism. Subsequently, we conducted immunoprecipitation and mass spectrometry analyses to detect potential interacting proteins and obtained 66 candidate proteins. Compared with the HG group, CRYAA protein combined more deubiquitinating enzyme USP25 in 661 W cells after MEV treatment (Fig. 4C and D). Therefore, we detected USP25 expression in MEVs. Western bolt confirmed that USP25 was enriched in MEVs and MEVs upregulated USP25 expression in 661 W cells under HG condition (Fig. 4E and F). Moreover, intravitreal MEV injection promoted USP25 expression in the ONL of db/db mice (Fig. 4G and H). To explore whether USP25 deubiquitinates and stabilizes CRYAA, we evaluated CRYAA ubiquitination in 661 W cells treated with USP25 overexpression plasmids or siRNAs. The Co-IP assay confirmed that CRYAA was the partner of USP25 (Fig. 4I). Moreover, CRYAA ubiquitination was markedly downregulated in USP25-overexpressing cells (Fig. 4J). In contrast, USP25 knockdown promoted CRYAA ubiquitination in 661 W cells (Fig. 4K). Rescuing USP25 knockdown cells with USP25^WT reduced CRYAA ubiquitination, whereas treatment with the USP25^C178S mutant, which lacks USP25 deubiquitinating enzyme activity, exerted no significant effect (Supplementary Figure S6A and B). To explore the lysine site on CRYAA involved in USP25 deubiquitination regulation, we constructed single-site mutants at all CRYAA lysine sites and observed that K86R abolished USP25 knockdown-induced CRYAA ubiquitination (Supplementary Figure S7A and B), indicating that K86 is the key site for USP25-mediated CRYAA deubiquitination. Furthermore, we inhibited USP25 expression in MSCs after shRNA transfection and purified MEVs^shUSP25 from the conditioned medium of MSCs^shUSP25. Western blot confirmed the inhibition of USP25 in MSCs^shUSP25 and MEVs^shUSP25 compared with the control groups (Fig. 4L and M). USP25 knockdown significantly abolished the MEV-mediated CRYAA deubiquitination in 661 W cells (Fig. 4N). Taken together, these results indicate that MEVs transfer USP25 to deubiquitinate and stabilize CRYAA.

Fig. 4 — MEVs transfer USP25 to inhibit CRYAA ubiquitination. (A) CHX experiment to detect the half-time of CRYAA in 661 W cells after MEV treatment and the corresponding quantification of CRYAA protein level at the indicated time points compared with the HG group (n = 3). (B) Ubiquitination assay of CRYAA protein in 661 W cells treated with MEVs and the corresponding quantification of CRYAA protein level and CRYAA ubiquitination level compared with the HG group (n = 3). (C, D) LC-MS/MS analysis for CRYAA binding proteins in MEV-treated 661 W cells. (E) Western blot for USP25 expression in MEVs. (F) Western blot to detect USP25 expression in 661 W cells after MEV treatment and the corresponding quantification of USP25 protein level compared with the LG group (n = 3). (G) Western blot for the retinal USP25 protein expression in db/m and db/db mice after MEV administration and the corresponding quantification of USP25 protein level compared with the db/m mice group (n = 3). (H) Representative immunofluorescence staining images of USP25 in retinal tissues of db/m and db/db mice after MEV injection and the quantitative analysis of fluorescent intensity compared with the db/m mice group on the basis of three mice per group with three images counted in each mouse. Scale bars, 100 μm. (I) Co-IP experiment for the combination of USP25 and CRYAA in 661 W cells. (J) Ubiquitination assay of CRYAA protein in 661 W cells transfected with NC or USP25 overexpression plasmids and the corresponding quantification of CRYAA protein level and CRYAA ubiquitination level compared with the NC group (n = 3). (K) Ubiquitination assay of CRYAA protein in 661 W cells transfected with NC or si-USP25 and the corresponding quantification of CRYAA protein level and CRYAA ubiquitination level compared with the NC group (n = 3). (L) Western blot for the validation of USP25 knockdown in MSCs after USP25 shRNA transfection and the corresponding quantification of USP25 protein level compared with the MSCs group (n = 3). (M) Western blot to detect the USP25 protein expression in MEVs and MEVs^shUSP25 and the corresponding quantification of USP25 protein level compared with the MEVs group (n = 3). (N) Ubiquitination assay of CRYAA protein in 661 W cells after MEVs or MEVs^shUSP25 treatment and the corresponding quantification of CRYAA protein level and CRYAA ubiquitination level compared with the HG group (n = 3). No data were excluded from the analyses. Data collection and analysis were performed by investigators blinded to the experimental groups. n represents independent biological replicates. Data in (A), (B), (F), (G), (H), (J), (K), (L), (M), and (N) are presented as mean ± SEM and were compared using two-tailed unpaired Student’s t-test (A, B, J, L, M) or one-way ANOVA with Tukey’s multiple comparison test (F, G, H, K, N). ns, not significant, *P < 0.05, **P < 0.01, and ***P < 0.001. GCL: ganglion cell layer; INL: inner nuclear layer; ONL: outer nuclear layer

MEVs alleviate photoreceptor injury in db/db mice by delivering USP25

To verify the key role of USP25 in MEV-induced photoreceptor therapy, db/db mice were treated with MEVs and MEVs^shUSP25 by intravitreal injection, respectively (Fig. 5A). ERG analysis showed that USP25 knockdown abolished the MEV-induced improvement in retinal function, as evidenced by reduced amplitudes of scotopic a/b-waves and photopic b-waves compared with the MEVs group (Fig. 5B-F). Retinal HE staining images presented that MEVs^shUSP25 displayed a limited ability to preserve retinal structure and increase ONL thickness (Fig. 5G and H). TUNEL staining further indicated that reduced USP25 level in MEVs remarkably impaired the anti-apoptotic effects of MEVs, as evidenced by the increased number of TUNEL⁺ cells in the ONL of db/db mice treated with MEVs^shUSP25 relative to the MEV-treated group (Fig. 5I and J). In addition, USP25 knockdown reversed PCNA and Bcl-2 upregulation and Bax downregulation in the retinas of db/db mice after MEV injection (Fig. 5K and L). Moreover, immunostaining presented that MEVs, but not MEVs^shUSP25, upregulated the retinal expression of rhodopsin and s-opsin in db/db mice (Fig. 5M-P). These findings suggest that USP25 is required for the MEV-induced therapeutic potential in hyperglycemia-triggered photoreceptor damage.

Fig. 5 — USP25 knockdown impairs MEV-mediated photoreceptor therapy in db/db mice. (A) Schematic diagram showing the experimental design of db/db mice treated with MEVs or MEVs^shUSP25. (B-D) Scotopic ERG analysis and a/b-wave amplitude changes in db/m and db/db mice treated with MEVs or MEVs^shUSP25 (n = 6). (E, F) Photopic ERG analysis and b-wave amplitude changes in db/m and db/db mice treated with MEVs or MEVs^shUSP25 (n = 6). (G, H) Retinal HE staining and ONL thickness measurement in db/m and db/db mice treated with MEVs or MEVs^shUSP25 (n = 3). Scale bars, 100 μm. (I, J) TUNEL staining of retinas and the analysis of TUNEL⁺ cell percentage of total ONL cells in db/m and db/db mice treated with MEVs or MEVs^shUSP25 on the basis of three mice per group with three images counted in each mouse. Scale bars, 100 μm. (K, L) Western blot analysis for PCNA, Bcl-2 and Bax expression in db/m and db/db mice treated with MEVs or MEVs^shUSP25 and the corresponding quantification of protein levels compared with the db/m mice group (n = 3). (M, N) Representative retinal rhodopsin immunostaining images of db/m and db/db mice treated with MEVs or MEVs^shUSP25 and the quantitative analysis of fluorescent intensity compared with the db/m mice group on the basis of three mice per group with three images counted in each mouse. Scale bars, 100 μm. (O, P) Representative retinal s-opsin immunostaining images of db/m and db/db mice treated with MEVs or MEVs^shUSP25 and the quantitative analysis of fluorescent intensity compared with the db/m mice group on the basis of three mice per group with three images counted in each mouse. Scale bars, 100 μm. No data were excluded from the analyses. Data collection and analysis were performed by investigators blinded to the experimental groups. n represents independent biological replicates. Data in (C), (D), (F), (H), (J), (L), (N), and (P) are presented as mean ± SEM and were compared using one-way ANOVA with Tukey’s multiple comparison test. ns, not significant, **P < 0.01, and ***P < 0.001. GCL: ganglion cell layer; INL: inner nuclear layer; ONL: outer nuclear layer

Preparation of MEVs^MH42-USP25 for photoreceptor-targeted delivery of USP25

To further enhance the therapeutic efficiency of MEVs against retinal dysfunction in DR, we intended to construct engineered MEVs with increased photoreceptor targeting and USP25 expression. As shown in Fig. 6A, we transfected MSCs with USP25 lentivirus, isolated MEVs-USP25 from the supernatant of MSCs-USP25, and conjugated MH42 onto the membrane of MEVs-USP25 using the CP05 peptide. TEM demonstrated that MEVs, MEVs-USP25, MEVs^MH42, and MEVs^MH42-USP25 displayed typical cup-shaped morphologies (Fig. 6B). NanoFCM results presented that the average size of MEVs^MH42-USP25 was slightly larger than that of MEVs (Fig. 6C). The EV markers such as CD9, CD63, HSP90, and TSG101 were expressed in MEVs, MEVs-USP25, MEVs^MH42, and MEVs^MH42-USP25 (Fig. 6D). Moreover, we detected USP25 expression in six batches of MEVs, MEVs-USP25, MEVs^MH42, and MEVs^MH42-USP25, and found that USP25 protein level was significantly increased in MEVs-USP25 and MEVs^MH42-USP25 compared with the MEVs and MEVs^MH42 groups (Fig. 6E, F and Supplementary Figure S8A, B). After lysis, USP25 was quantified to be 173 ± 15 pg per 10⁸ particles in MEVs^MH42-USP25 (Supplementary Figure S8C). Subsequently, we measured the conjugation efficacy of MH42 peptide and MH42-CP05 fusion peptide. The results showed that the conjugation efficiency of MH42-CP05 fusion peptide was significantly higher than that of MH42 peptide (Fig. 6G). Flow cytometry analysis further confirmed the effective binding of the MH42-CP05 fusion peptide to the surface of MEVs^MH42 and MEVs^MH42-USP25 (Supplementary Figure S8D). Biodistribution analysis verified that MEVs and MEVs-USP25 were located in the GCL, INL, and ONL after intravitreal injection, whereas conjugation of the MH42-CP05 fusion peptide contributed to the location of MEVs^MH42 and MEVs^MH42-USP25 in the ONL (Fig. 6H and Supplementary Figure S8E). In addition, immunofluorescence staining results showed that MEVs^MH42 and MEVs^MH42-USP25 displayed significantly increased co-localization with rhodopsin⁺ rods and s-opsin⁺ cones after intravitreal injection (Supplementary Figure S9A-D). The above results suggest that the MH42 peptide enables MEVs^MH42-USP25 to target photoreceptors for USP25 delivery.

Fig. 6 — Construction and characterization of MEVs^MH42-USP25. (A) Schematic diagram of MEVs^MH42-USP25 preparation. (B) TEM images of MEVs, MEVs-USP25, MEVs^MH42, and MEVs^MH42-USP25. Scale bars, 50 nm. (C) NanoFCM for the particle diameter of MEVs, MEVs-USP25, MEVs^MH42, and MEVs^MH42-USP25. (D) Western blot to detect the EV markers in MEVs, MEVs-USP25, MEVs^MH42, and MEVs^MH42-USP25. (E, F) Western blot analysis for USP25 protein expression in three batches of MEVs, MEVs-USP25, MEVs^MH42, and MEVs^MH42-USP25 and the corresponding quantification of protein level compared with the MEVs group (n = 3). (G) The conjugation efficacy of FITC-labeled MH42 peptides or FITC-labeled MH42-CP05 fusion peptides with MEVs or MEVs-USP25 (n = 3). (H) Biodistribution of FITC-labeled MH42-CP05 fusion peptides and PKH26-labeled MEVs, MEVs-USP25, MEVs^MH42, or MEVs^MH42-USP25 in the retinas after intravitreal injection for 1 day. Scale bars, 100 μm. No data were excluded from the analyses. Data collection and analysis were performed by investigators blinded to the experimental groups. n represents independent biological replicates. Data in (F) and (G) are presented as mean ± SEM and were compared using one-way ANOVA with Tukey’s multiple comparison test. ns, not significant, ***P < 0.001. GCL: ganglion cell layer; INL: inner nuclear layer; ONL: outer nuclear layer

Enhanced efficiency of MEVs^MH42-USP25 in photoreceptor therapy

We then explored the optimal in vivo dose of MEVs^MH42-USP25. As shown in Supplementary Figure S10A-N, MEVs^MH42-USP25 at a dose of 1 × 10⁷ particles significantly ameliorated hyperglycemia-induced retinal dysfunction and photoreceptor damage compared with the 1 × 10⁶ particle group. Notably, no significant difference in therapeutic efficacy was observed between the 1 × 10⁷ particle and 1 × 10⁸ particle groups. Therefore, we selected 1 × 10⁷ particles as the dose for comparing the therapeutic potential of MEVs, MEVs-USP25, MEVs^MH42, and MEVs^MH42-USP25 in DR (Fig. 7A). The results of Western blot demonstrated that MH42 conjugation and USP25 loading enhanced the ability of MEVs to upregulate retinal USP25 expression in db/db mice (Figure S11A and B). Photopic and scotopic ERG analyses indicated that intravitreal delivery of MEVs^MH42-USP25 more effectively alleviated retinal dysfunction in db/db mice compared to MEVs, MEVs-USP25, and MEVs^MH42, as evidenced by the higher amplitudes of photopic b-waves and scotopic a/b-waves (Fig. 7B-F). Retinal HE staining demonstrated that MEV-treated db/db mice exhibited improved retinal structure and increased ONL thickness, while MEVs^MH42-USP25 further mitigated diabetes-induced retinal degeneration and photoreceptor loss (Fig. 7G and H). Consistently, TUNEL staining presented that MEVs^MH42-USP25 treatment resulted in a lower rate of photoreceptor apoptosis in the ONL compared with MEVs, MEVs-USP25, and MEVs^MH42 groups (Fig. 7I and J). As verified by rhodopsin and s-opsin staining, the hyperglycemia-induced loss of rods and cones was further reversed by MEVs^MH42-USP25 treatment (Fig. 7K-N).

Fig. 7 — Elevated therapeutic efficacy of MEVs^MH42-USP25 in vivo. (A) Schematic diagram showing the experimental design to investigate the repairing role of MEVs^MH42-USP25 in vivo. (B, C) Photopic ERG analysis and b-wave amplitude changes in db/m and db/db mice treated with MEVs, MEVs-USP25, MEVs^MH42, or MEVs^MH42-USP25 (n = 6). (D-F) Scotopic ERG analysis and a/b-wave amplitude changes in db/m and db/db mice treated with MEVs, MEVs-USP25, MEVs^MH42, or MEVs^MH42-USP25 (n = 6). (G, H) Retinal HE staining and ONL thickness measurement in db/m and db/db mice treated with MEVs, MEVs-USP25, MEVs^MH42, or MEVs^MH42-USP25 (n = 3). Scale bars, 100 μm. (I, J) TUNEL staining of retinas and the analysis of TUNEL⁺ cell percentage of total ONL cells in db/m and db/db mice treated with MEVs, MEVs-USP25, MEVs^MH42, or MEVs^MH42-USP25 on the basis of three mice per group with three images counted in each mouse. Scale bars, 100 μm. (K, L) Representative retinal rhodopsin immunostaining images of db/m and db/db mice treated with MEVs, MEVs-USP25, MEVs^MH42, or MEVs^MH42-USP25 and the quantitative analysis of fluorescent intensity compared with the db/m mice group on the basis of three mice per group with three images counted in each mouse. Scale bars, 100 μm. (M, N) Representative retinal s-opsin immunostaining images of db/m and db/db mice treated with MEVs, MEVs-USP25, MEVs^MH42, or MEVs^MH42-USP25 and the quantitative analysis of fluorescent intensity compared with the db/m mice group on the basis of three mice per group with three images counted in each mouse. Scale bars, 100 μm. No data were excluded from the analyses. Data collection and analysis were performed by investigators blinded to the experimental groups. n represents independent biological replicates. Data in (C), (E), (F), (H), (J), (L), and (N) are presented as mean ± SEM and were compared using one-way ANOVA with Tukey’s multiple comparison test. ns, not significant, *P < 0.05, **P < 0.01, and ***P < 0.001. GCL: ganglion cell layer; INL: inner nuclear layer; ONL: outer nuclear layer

In addition, MEVs^MH42-USP25 and MEVs-USP25 were normalized to equal USP25 protein concentrations and intravitreally injected into db/db mice (Supplementary Figure S12A). Compared with the MEVs-USP25 group, MEVs^MH42-USP25 with the same USP25 protein level displayed superior therapeutic effects to improve retinal function, maintain ONL thickness, inhibit photoreceptor apoptosis, and upregulate rhodopsin and s-opsin expression (Supplementary Figure S12B-N). Moreover, a comparison between the retinal therapeutic effects of MEVs^MH42-USP25 and MEVs^Scrbl-USP25 revealed that MEVs^MH42-USP25 more effectively alleviated hyperglycemia-induced retinal dysfunction and photoreceptor loss in db/db mice compared with the MEVs^Scrbl-USP25 group (Supplementary Figure S13A-P). These findings suggest that both MH42 peptide conjugation and USP25 protein loading synergistically contribute to MEVs^MH42-USP25-mediated effective therapy for photoreceptors in DR.

In addition to the retinal therapeutic outcomes, MEVs^MH42-USP25 treatment also displayed biosafety profiles. Few abnormalities were observed in ocular inflammation, intraocular pressure, serum cytokine levels, anti-MH42 and anti-MEV antibody production, blood cell counts, and histological features of the heart, liver, lung, spleen, and kidney in MEVs^MH42-USP25-treated db/db mice at weeks 24 and 48 (Supplementary Figure S14A-G and Supplementary Figure S15A-G). Collectively, these findings suggest that MEVs^MH42-USP25 exhibit enhanced therapeutic efficacy and good safety profiles in alleviating diabetic retinal dysfunction and photoreceptor damage.

Discussion

Photoreceptor loss is an essential cause of retinal dysfunction in DR [30]. In this study, we demonstrated that MEVs improved retinal function in db/db mice by delivering USP25 to inhibit CRYAA ubiquitination and degradation in photoreceptors. Importantly, we developed a novel MEV-based therapeutic system conjugated with the photoreceptor-targeting peptide MH42 and enriched with USP25 protein to achieve a targeted therapy for photoreceptors in DR.

It is recognized that MSCs transplantation maintains tissue homeostasis and repairs injured cells mainly through the paracrine pathway [31]. As a major cargo of paracrine effect, MEVs contribute to tissue regeneration by delivering proteins, nucleic acids, and lipids [32]. Recently, MEVs have emerged as a novel strategy for the treatment of ocular diseases [33]. Kaur et al. have revealed the immunomodulatory potency of MEVs in inhibiting the infiltration of retinal reactive T cells into the eyes, thus alleviating autoimmune uveitis progression [34]. In dry age-related macular degeneration, MEVs prevent oxidative damage in retinal pigment epithelium cells by regulating the Nrf2/Kepa1 pathway [35]. Notably, several studies have reported the therapeutic potential of MEVs in DR. Chen et al. have shown that MEVs suppress Müller gliosis by activating NRF2-mediated antioxidant effects in DR [36]. Furthermore, Sun et al. have revealed that MEVs inhibit diabetic retinal neovascularization [37]. However, whether and how MEVs protect photoreceptors against hyperglycemic injury remain unclear. In this study, intravitreal MEV administration restored retinal function by alleviating photoreceptor loss in db/db mice. In vitro results also verified that MEVs alleviated HG-induced apoptosis of 661 W cell apoptosis, providing a foundation for establishing novel therapies for retinal dysfunction in DR.

CRYAA is a member of the small shock protein family and serves as a molecular chaperone [38]. Recent studies have mainly focused on the benefits of CRYAA in protecting eye lens function and transparency. Zhou et al. have reported that CRYAA is epigenetically suppressed in the lens epithelium of age-related nuclear cataracts [39]. Here, we performed LC-MS/MS to investigate the differentially expressed proteins in 661 W cells following MEV treatment. MEV-mediated CRYAA protein upregulation in photoreceptors was observed both in vivo and in vitro. Under chronic stress, CRYAA serves as an essential photoreceptor regulator [40]. Ruebsam et al. have demonstrated the retinal neuroprotective potential of CRYAA in diabetes [26]. However, the role of CRYAA in hyperglycemia-induced photoreceptor loss remains unclear. Here, we transfected 661 W cells with CRYAA overexpression plasmids and observed that CRYAA activation markedly inhibited the apoptosis of 661 W cells under HG stimulation. Moreover, the genetic loss of CRYAA in diabetic mice abolished the therapeutic effects of MEVs to improve retinal function and inhibit photoreceptor apoptosis. AAV9-CRYAA-mediated CRYAA re-expression could recover the role of MEVs to alleviate photoreceptor injury, suggesting that MEVs protect photoreceptors by upregulating CRYAA expression.

MEVs transport various cargos to regulate the recipient cell functions. Shuttling of proteins is considered the major mechanism underlying MEV-mediated therapeutic effects in retinal disorders. Sun et al. have reported that hypoxia-preconditioned MEVs ameliorate retinal degeneration by delivering GAP43 [41]. BDNF-enriched MEVs accelerate Müller cell survival by regulating Wnt signaling [42]. In this study, we revealed that MEVs upregulated CRYAA protein level by delivering USP25 to inhibit CRYAA ubiquitination and degradation. Although CRYAA has been demonstrated as a substrate of the ubiquitin proteasome pathway [43], the interaction between the deubiquitinating enzyme USP25 and CRYAA remains unclear. Hence, these results provide novel insights into MEV-induced photoreceptor therapy for DR.

Over the last few years, EVs have gained increasing attention as promising nanocarriers for therapeutic cargo [44]. The membrane structure and nanoscale size of EVs can protect encapsulated contents from degradation and allow them to cross impermeable biological barriers [45]. Due to their retinal therapeutic effects, good biocompatibility, and low immunogenicity, we selected MEVs as the delivery platform in this study. Considering that MEVs inhibit hyperglycemia-induced photoreceptor loss by USP25-induced CRYAA deubiquitination, we introduced USP25 protein into MSCs by lentiviral transfection, followed by the release of MEVs-USP25. Recent evidence has revealed that MEVs exhibit low retinal cell targeting after intravitreal injection, resulting in reduced repairing efficacy [46]. The non-specific MEV distribution in retinal cells is a significant challenge in MEV-mediated DR therapy. Increasing evidence suggests that functionalizing the membrane structures of MEVs can enhance their targeting potential [47]. Strategies involving minimal procedures to modify MEVs are highly desirable for clinical applications [48]. Recently, Gao et al. have revealed that CP05 anchor peptides specifically bind to CD63 on EV membranes [49]. Xu et al. have modified EVs with TBP-CP05 fusion peptides to transfer miR-214 to osteoclasts [50]. Thus, the use of CP05 peptides to engineer MEVs may be a promising strategy for DR-targeted therapy. The MH42 peptide was recently identified as a photoreceptor-targeting ligand [51]. MH42 peptide modification may promote MEV accumulation in photoreceptors. In this study, we conjugated the MH42-CP05 fusion peptide to the surface of MEVs-USP25 based on the interaction between the CP05 peptide and the EV marker CD63. After intravitreal injection, MEVs^MH42-USP25 were predominantly localized in the photoreceptors and further improved retinal function in db/db mice compared with natural MEVs, suggesting that MEVs^MH42-USP25 are effective nanotherapeutics for photoreceptor injury in DR.

However, the utilization of rodent models for DR study has potential limitations in the replication of chronic neurovascular damage and immune involvement in patients with DR. Thus, primate DR models are needed to investigate the intervention dosage and timing of MEVs^MH42-USP25. Although MH42 peptides have been shown to guide lipid nanoparticles to photoreceptors for mRNA delivery in mice and rhesus macaques [51], whether MH42 peptides can effectively target human photoreceptors requires further investigation. Moreover, as a tetraspanin enriched on the surface of EVs, CD63 is considered an EV marker. The CP05 peptide has recently been found to specifically bind to the second extracellular loop of CD63, a structure is highly conserved across species [52]. Increasing studies have shown that EVs from various human, mouse and rat cells can be directly modified via CP05 conjugation [53–55], indicating that the CP05-CD63 interaction remains stable across species. Therefore, coupling MH42 peptides to EVs derived from different species sources using CP05 as an anchor peptide is a feasible strategy. In addition, the scalability of lentiviral modification and MEV purification under good manufacturing practice (GMP) conditions is essential for clinical translation. Optimizing the design of lentiviral vectors is crucial to achieve stable gene integration into MSCs. Recent advancements in 3D culture technology have facilitated the large-scale cultivation of MSCs transfected with lentiviruses to increase the production of engineered MEVs [56]. However, traditional EV separation techniques such as ultracentrifugation, size exclusion chromatography, immune-affinity capture, and polymer precipitation often fail to simultaneously ensure both high purity and yield. Moreover, some of these methods are time-consuming and lack reproducibility, limiting their suitability for clinical application [57]. Future studies should consider novel purification strategies, such as tangential flow filtration, affinity chromatography, and microfluidic technology, for efficient and automated purification of engineered MEVs with high purity [58]. Finally, the characteristics, safety, batch consistency, and stability of engineered MEVs should be further tested to meet the GMP requirements for their clinical applicability.

Conclusion

In conclusion, we have established a photoreceptor-targeted transportation system as a novel therapeutic approach for DR by delivering USP25 protein. The findings of this study demonstrated that MEVs ameliorated long-term hyperglycemia-induced retinal dysfunction and photoreceptor loss by USP25-mediated CRYAA deubiquitination. MEVs^MH42-USP25 decorated with MH42 peptides efficiently delivered the deubiquitinating enzyme USP25 to the photoreceptors. Consequently, increased USP25 protein level further prevented photoreceptors from diabetic stimulation in db/db mice. Our study reveals the effectiveness and biosafety of MEVs^MH42-USP25 in photoreceptor therapy and provides a promising cell-free therapeutic strategy for DR.

Supplementary Information

Supplementary Material 1.^{(40MB, docx)}

Acknowledgements

Not applicable.

Abbreviations

AAV9: Adeno-associated virus serotype 9
BSA: Bovine serum albumin
CCK8: Cell Counting Kit-8
Co-IP: Co-immunoprecipitation
CRYAA: αA-crystallin
CRYAA^KO: CRYAA knockout
DR: Diabetic retinopathy
EVs: Extracellular vesicles
GMP: Good manufacturing practice
GSH: Glutathione
HE: Hematoxylin and eosin
HFD: High-fat diet
HG: High glucose
LC-MS/MS: Liquid chromatography-tandem mass spectrometry
LG: Low glucose
MDA: Malondialdehyde
MEVs: Mesenchymal stem cell-derived extracellular vesicles
MEVs^MH42-USP25: MEVs conjugated with the photoreceptor-binding peptide MH42 and internally loaded with USP25 protein
MSCs: Mesenchymal stem cells
NanoFCM: Nanoparticle flow cytometry
ONL: Outer nuclear layer
ROS: Reactive oxygen species
SOD: Superoxide dismutase
STZ: Streptozotocin
TEM: Transmission electron microscopy
TUNEL: Terminal deoxynucleotidyl transferase dUTP nick end labeling
USP25: Ubiquitin-specific peptidase 25
WT: Wild type

Author contributions

Y.S., J.H., B.Y. and F.S. designed the study. Y.S., S.C., X.Q., Y.S., Z.J., J.C. and Y.M. designed the experimental protocol, handled the experiments, and performed the statistical analysis. Y.S., S.C. and X.Q. wrote the manuscript. J.H., B.Y. and F.S. revised the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (82402495, 82402487, 82202618, 82401900), Zhejiang Provincial Natural Science Foundation of China (LQN25H180001, LQ23H260006), Natural Science Foundation of Jiangsu Province (BK20240512), Jinhua Science and Technology Program (2024-3-068, 2024-4-063, JYZDXK-2023-09), Top Talent Support Program for young and middle-aged people of Wuxi Health Committee (BJ2023104).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This research was approved by the Medical Ethics Committee of Affiliated Jinhua Hospital of Zhejiang University (JHYY202404).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Yaoxiang Sun, Shenyuan Chen and Xiaoyuan Qi contributed equally to this work.

Contributor Information

Jin Huang, Email: 13914298422@163.com.

Benshuai You, Email: youbenshuai@163.com.

Fengtian Sun, Email: jsdxsft@163.com.

References

1.Vos S, Portillo JC, Hubal A, Bapputty R, Pfaff A, Aaron R, et al. CD40 induces unfolded protein response, upregulation of VEGF, and vascular leakage in diabetic retinopathy. Diabetes. 2025;74(5):798–811. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Wen Y, Chen X, Feng H, Wang X, Kang X, Zhao P, et al. Kdm6a deficiency in microglia/macrophages epigenetically silences Lcn2 expression and reduces photoreceptor dysfunction in diabetic retinopathy. Metabolism. 2022;136:155293. [DOI] [PubMed] [Google Scholar]
3.Wang S, Yang H, Zheng J, Tong A, Mu S, Wang D, et al. Recent advances and prospects of nanoparticle-based drug delivery for diabetic ocular complications. Theranostics. 2025;15(8):3551–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Cheng Y, Zhang M, Li C, Su L, Fu L, Wu S, et al. Endothelial AGGF1 promotes retinal angiogenesis by coordinating TNFSF12/FN14 signalling. Nat Commun. 2025;16(1):1332. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Pinheiro AAS, Torrecilhas AC, Souza BSF, Cruz FF, Guedes HLM, Ramos TD, et al. Potential of extracellular vesicles in the pathogenesis, diagnosis and therapy for parasitic diseases. J Extracell Vesicles. 2024;13(8):e12496. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Manai F, Smedowski A, Kaarniranta K, Comincini S, Amadio M. Extracellular vesicles in degenerative retinal diseases: a new therapeutic paradigm. J Control Release. 2024;365:448–68. [DOI] [PubMed] [Google Scholar]
7.Park HS, Park JH, Oh MK, Yu F, Yu KR. Advancements in 3D bioprinting for precision medicine: enhancing patient-derived organoids and extracellular vesicle applications in inflammatory diseases. Int J Bioprint. 2024;10: 4054. [Google Scholar]
8.Weng Z, Zhang B, Wu C, Yu F, Han B, Li B, et al. Therapeutic roles of mesenchymal stem cell-derived extracellular vesicles in cancer. J Hematol Oncol. 2021;14(1):136. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Yoo D, Jung SY, Go D, Park JY, You DG, Jung WK, et al. Functionalized extracellular vesicles of mesenchymal stem cells for regenerative medicine. J Nanobiotechnol. 2025;23(1):219. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Zhu J, Huang J, Sun Y, Xu W, Qian H. Emerging role of extracellular vesicles in diabetic retinopathy. Theranostics. 2024;14(4):1631–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Mathew B, Acha LG, Torres LA, Huang CC, Liu A, Kalinin S, et al. MicroRNA-based engineering of mesenchymal stem cell extracellular vesicles for treatment of retinal ischemic disorders: engineered extracellular vesiclesand retinal ischemia. Acta Biomater. 2023;158:782–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Sun F, Sun Y, Zhu J, Wang X, Ji C, Zhang J, et al. Mesenchymal stem cells-derived small extracellular vesicles alleviate diabetic retinopathy by delivering NEDD4. Stem Cell Res Ther. 2022;13(1):293. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Long Y, Yang B, Lei Q, Gao F, Chen L, Chen W, et al. Targeting senescent alveolar epithelial cells using engineered mesenchymal stem cell-derived extracellular vesicles to treat pulmonary fibrosis. ACS Nano. 2024;18(9):7046–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Yang J, Hu H, Guo Q, Chen X, Li S, Yin G, et al. Electrospun polylactic acid-glycolic acid composite hydrogel scaffold loaded with 3D extracellular vesicles for nasal septal cartilage defect repair. Int J Bioprint. 2024;10:4118. [Google Scholar]
15.Ding JY, Chen MJ, Wu LF, Shu GF, Fang SJ, Li ZY, et al. Mesenchymal stem cell-derived extracellular vesicles in skin wound healing: roles, opportunities and challenges. Mil Med Res. 2023;10(1): 36. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Zeng ZL, Xie H. Mesenchymal stem cell-derived extracellular vesicles: a possible therapeutic strategy for orthopaedic diseases: a narrative review. Biomater Transl. 2022;3(3):175–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.van Griensven M, Balmayor ER. Extracellular vesicles are key players in mesenchymal stem cells’ dual potential to regenerate and modulate the immune system. Adv Drug Deliv Rev. 2024;207: 115203. [DOI] [PubMed] [Google Scholar]
18.Fan L, Wang L, Wang X, Zhang H. Exosomes-based particles as inhalable COVID-19 vaccines. Biomed Tech (Berl). 2023;4:24–7. [Google Scholar]
19.Phan J, Kumar P, Hao D, Gao K, Farmer D, Wang A. Engineering mesenchymal stem cells to improve their exosome efficacy and yield for cell-free therapy. J Extracell Vesicles. 2018;7(1): 1522236. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Hu Z, Qian S, Zhao Q, Lu B, Lu Q, Wang Y, et al. Engineering strategies for apoptotic bodies. Smart Medicine. 2024;3(3): e20240005. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Ji C, Zhang J, Shi L, Shi H, Xu W, Jin J, et al. Engineered extracellular vesicle-encapsulated CHIP as novel nanotherapeutics for treatment of renal fibrosis. NPJ Regen Med. 2024;9(1): 3. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Guo D, Sun Y, Wu J, Ding L, Jiang Y, Xue Y, et al. Photoreceptor-targeted extracellular vesicles-mediated delivery of Cul7 siRNA for retinal degeneration therapy. Theranostics. 2024;14(13):4916–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Xie SY, Liu SQ, Zhang T, Shi WK, Xing Y, Fang WX, et al. USP28 serves as a key suppressor of mitochondrial morphofunctional defects and cardiac dysfunction in the diabetic heart. Circulation. 2024;149(9):684–706. [DOI] [PubMed] [Google Scholar]
24.Mathew B, Ravindran S, Liu X, Torres L, Chennakesavalu M, Huang CC, et al. Mesenchymal stem cell-derived extracellular vesicles and retinal ischemia-reperfusion. Biomaterials. 2019;197:146–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Rao NA, Saraswathy S, Pararajasegaram G, Bhat SP. Small heat shock protein αA-crystallin prevents photoreceptor degeneration in experimental autoimmune uveitis. PLoS One. 2012;7(3):e33582. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Ruebsam A, Dulle JE, Myers AM, Sakrikar D, Green KM, Khan NW, et al. A specific phosphorylation regulates the protective role of αA-crystallin in diabetes. JCI Insight. 2018;3(4):e97919. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Yang W, Lu B, Chen Q, Wang J, Zhou J, Li R, et al. CRYAA activates the SIRT1-pi3K/AKT signaling pathway by suppressing mir-155-5p to protect the RPE. Arch Biochem Biophys. 2025;770: 110435. [DOI] [PubMed] [Google Scholar]
28.Christopher KL, Pedler MG, Shieh B, Ammar DA, Petrash JM, Mueller NH. Alpha-crystallin-mediated protection of lens cells against heat and oxidative stress-induced cell death. Biochim Biophys Acta. 2014;1843(2):309–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Liu X, Liu B, Zhou M, Fan F, Yu M, Gao C, et al. Circular RNA HIPK3 regulates human lens epithelial cells proliferation and apoptosis by targeting the miR-193a/CRYAA axis. Biochem Biophys Res Commun. 2018;503(4):2277–85. [DOI] [PubMed] [Google Scholar]
30.Tonade D, Kern TS. Photoreceptor cells and RPE contribute to the development of diabetic retinopathy. Prog Retin Eye Res. 2021;83: 100919. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Wu R, Fan X, Wang Y, Shen M, Zheng Y, Zhao S, et al. Mesenchymal stem cell-derived extracellular vesicles in liver immunity and therapy. Front Immunol. 2022;13:833878. [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Song Y, You Y, Xu X, Lu J, Huang X, Zhang J, et al. Adipose-derived mesenchymal stem cell-derived exosomes biopotentiated extracellular matrix hydrogels accelerate diabetic wound healing and skin regeneration. Adv Sci. 2023;10(30):e2304023. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Zhang Z, Liang X, Zhou J, Meng M, Gao Y, Yi G, et al. Exosomes in the pathogenesis and treatment of ocular diseases. Exp Eye Res. 2021;209: 108626. [DOI] [PubMed] [Google Scholar]
34.Kaur G, Bae EH, Zhang Y, Ciacciofera N, Jung KM, Barreda H, et al. Biopotency and surrogate assays to validate the immunomodulatory potency of extracellular vesicles derived from mesenchymal stem/stromal cells for the treatment of experimental autoimmune uveitis. J Extracell Vesicles. 2024;13(8): e12497. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Tang Y, Kang Y, Zhang X, Cheng C. Mesenchymal stem cell exosomes as nanotherapeutics for dry age-related macular degeneration. J Control Release. 2023;357:356–70. [DOI] [PubMed] [Google Scholar]
36.Chen Y, Tong J, Liu C, He C, Xiang J, Yao G, et al. MSC-derived small extracellular vesicles mitigate diabetic retinopathy by stabilizing Nrf2 through miR-143-3p-mediated inhibition of neddylation. Free Radic Biol Med. 2024;219:76–87. [DOI] [PubMed] [Google Scholar]
37.Sun F, Sun Y, Wang X, Zhu J, Chen S, Yu Y, et al. Engineered mesenchymal stem cell-derived small extracellular vesicles for diabetic retinopathy therapy through HIF-1α/EZH2/PGC-1α pathway. Bioact Mater. 2024;33:444–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Jia ZK, Fu CX, Wang AL, Yao K, Chen XJ. Cataract-causing allele in CRYAA (Y118D) proceeds through endoplasmic reticulum stress in mouse model. Zool Res. 2021;42(3):300–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Zhou P, Luo Y, Liu X, Fan L, Lu Y. Down-regulation and CpG island hypermethylation of CRYAA in age-related nuclear cataract. FASEB J. 2012;26(12):4897–902. [DOI] [PubMed] [Google Scholar]
40.Besirli CG, Nath M, Yao J, Pawar M, Myers AM, Zacks D, et al. HSPB4/CRYAA protect photoreceptors during retinal detachment in part through FAIM2 regulation. Neurol Int. 2024;16(5):905–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Sun Y, Sun Y, Chen S, Yu Y, Ma Y, Sun F. Hypoxic preconditioned MSCs-derived small extracellular vesicles for photoreceptor protection in retinal degeneration. J Nanobiotechnol. 2023;21(1):449. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Gao Y, Li H, Qin C, Yang B, Ke Y. Embryonic stem cells-derived exosomes enhance retrodifferentiation of retinal Müller cells by delivering BDNF protein to activate Wnt pathway. Immunobiology. 2022;227(3):152211. [DOI] [PubMed] [Google Scholar]
43.Yang H, Ping X, Zhou J, Ailifeire H, Wu J, Nadal-Nicolás FM, et al. Reversible cold-induced lens opacity in a hibernator reveals a molecular target for treating cataracts. J Clin Invest. 2024;134(18): e169666. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Widjaja G, Jalil AT, Budi HS, Abdelbasset WK, Efendi S, Suksatan W, et al. Mesenchymal stromal/stem cells and their exosomes application in the treatment of intervertebral disc disease: a promising frontier. Int Immunopharmacol. 2022;105: 108537. [DOI] [PubMed] [Google Scholar]
45.Zhang X, Zhang H, Gu J, Zhang J, Shi H, Qian H, et al. Engineered extracellular vesicles for cancer therapy. Adv Mater. 2021;33(14): e2005709. [DOI] [PubMed] [Google Scholar]
46.Mathew B, Torres LA, Gamboa Acha L, Tran S, Liu A, Patel R, et al. Uptake and distribution of administered bone marrow mesenchymal stem cell extracellular vesicles in retina. Cells. 2021;10(4):730. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Wu P, Zhang B, Ocansey DKW, Xu W, Qian H. Extracellular vesicles: a bright star of nanomedicine. Biomaterials. 2021;269:120467. [DOI] [PubMed] [Google Scholar]
48.Tang TT, Wang B, Li ZL, Wen Y, Feng ST, Wu M, et al. Kim-1 targeted extracellular vesicles: a new therapeutic platform for RNAi to treat AKI. J Am Soc Nephrol. 2021;32(10):2467–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Gao X, Ran N, Dong X, Zuo B, Yang R, Zhou Q, et al. Anchor peptide captures, targets, and loads exosomes of diverse origins for diagnostics and therapy. Sci Transl Med. 2018;10(444):eaaw0534. [DOI] [PubMed] [Google Scholar]
50.Xu L, Xu X, Liang Y, Wen C, Ouyang K, Huang J, et al. Osteoclast-targeted delivery of anti-miRNA oligonucleotides by red blood cell extracellular vesicles. J Control Release. 2023;358:259–72. [DOI] [PubMed] [Google Scholar]
51.Herrera-Barrera M, Ryals RC, Gautam M, Jozic A, Landry M, Korzun T, et al. Peptide-guided lipid nanoparticles deliver mRNA to the neural retina of rodents and nonhuman primates. Sci Adv. 2023;9(2): eadd4623. [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Zou Y, Li L, Li Y, Chen S, Xie X, Jin X, et al. Restoring cardiac functions after myocardial infarction-ischemia/reperfusion via an exosome anchoring conductive hydrogel. ACS Appl Mater Interfaces. 2021;13(48):56892–908. [DOI] [PubMed] [Google Scholar]
53.Yu S, Liao R, Bai L, Guo M, Zhang Y, Zhang Y, et al. Anticancer effect of hUC-MSC-derived exosome-mediated delivery of PMO-miR-146b-5p in colorectal cancer. Drug Deliv Transl Res. 2024;14(5):1352–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Han G, Zhang Y, Zhong L, Wang B, Qiu S, Song J, et al. Generalizable anchor aptamer strategy for loading nucleic acid therapeutics on exosomes. EMBO Mol Med. 2024;16(4):1027–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Zha Y, Li Y, Lin T, Chen J, Zhang S, Wang J. Progenitor cell-derived exosomes endowed with VEGF plasmids enhance osteogenic induction and vascular remodeling in large segmental bone defects. Theranostics. 2021;11(1):397–409. [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Xu Y, Wei C, Ma L, Zhao L, Li D, Lin Y, et al. 3D mesenchymal stem cell exosome-functionalized hydrogels for corneal wound healing. J Control Release. 2025;380:630–46. [DOI] [PubMed] [Google Scholar]
57.Jia Y, Yu L, Ma T, Xu W, Qian H, Sun Y, et al. Small extracellular vesicles isolation and separation: current techniques, pending questions and clinical applications. Theranostics. 2022;12(15):6548–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Fu P, Zhang J, Li H, Mak M, Xu W, Tao Z. Extracellular vesicles as delivery systems at nano-/micro-scale. Adv Drug Deliv Rev. 2021;179: 113910. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1.^{(40MB, docx)}

Data Availability Statement

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

[CR1] 1.Vos S, Portillo JC, Hubal A, Bapputty R, Pfaff A, Aaron R, et al. CD40 induces unfolded protein response, upregulation of VEGF, and vascular leakage in diabetic retinopathy. Diabetes. 2025;74(5):798–811. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR2] 2.Wen Y, Chen X, Feng H, Wang X, Kang X, Zhao P, et al. Kdm6a deficiency in microglia/macrophages epigenetically silences Lcn2 expression and reduces photoreceptor dysfunction in diabetic retinopathy. Metabolism. 2022;136:155293. [DOI] [PubMed] [Google Scholar]

[CR3] 3.Wang S, Yang H, Zheng J, Tong A, Mu S, Wang D, et al. Recent advances and prospects of nanoparticle-based drug delivery for diabetic ocular complications. Theranostics. 2025;15(8):3551–70. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR4] 4.Cheng Y, Zhang M, Li C, Su L, Fu L, Wu S, et al. Endothelial AGGF1 promotes retinal angiogenesis by coordinating TNFSF12/FN14 signalling. Nat Commun. 2025;16(1):1332. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR5] 5.Pinheiro AAS, Torrecilhas AC, Souza BSF, Cruz FF, Guedes HLM, Ramos TD, et al. Potential of extracellular vesicles in the pathogenesis, diagnosis and therapy for parasitic diseases. J Extracell Vesicles. 2024;13(8):e12496. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR6] 6.Manai F, Smedowski A, Kaarniranta K, Comincini S, Amadio M. Extracellular vesicles in degenerative retinal diseases: a new therapeutic paradigm. J Control Release. 2024;365:448–68. [DOI] [PubMed] [Google Scholar]

[CR7] 7.Park HS, Park JH, Oh MK, Yu F, Yu KR. Advancements in 3D bioprinting for precision medicine: enhancing patient-derived organoids and extracellular vesicle applications in inflammatory diseases. Int J Bioprint. 2024;10: 4054. [Google Scholar]

[CR8] 8.Weng Z, Zhang B, Wu C, Yu F, Han B, Li B, et al. Therapeutic roles of mesenchymal stem cell-derived extracellular vesicles in cancer. J Hematol Oncol. 2021;14(1):136. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR9] 9.Yoo D, Jung SY, Go D, Park JY, You DG, Jung WK, et al. Functionalized extracellular vesicles of mesenchymal stem cells for regenerative medicine. J Nanobiotechnol. 2025;23(1):219. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR10] 10.Zhu J, Huang J, Sun Y, Xu W, Qian H. Emerging role of extracellular vesicles in diabetic retinopathy. Theranostics. 2024;14(4):1631–46. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR11] 11.Mathew B, Acha LG, Torres LA, Huang CC, Liu A, Kalinin S, et al. MicroRNA-based engineering of mesenchymal stem cell extracellular vesicles for treatment of retinal ischemic disorders: engineered extracellular vesiclesand retinal ischemia. Acta Biomater. 2023;158:782–97. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Sun F, Sun Y, Zhu J, Wang X, Ji C, Zhang J, et al. Mesenchymal stem cells-derived small extracellular vesicles alleviate diabetic retinopathy by delivering NEDD4. Stem Cell Res Ther. 2022;13(1):293. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR13] 13.Long Y, Yang B, Lei Q, Gao F, Chen L, Chen W, et al. Targeting senescent alveolar epithelial cells using engineered mesenchymal stem cell-derived extracellular vesicles to treat pulmonary fibrosis. ACS Nano. 2024;18(9):7046–63. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR14] 14.Yang J, Hu H, Guo Q, Chen X, Li S, Yin G, et al. Electrospun polylactic acid-glycolic acid composite hydrogel scaffold loaded with 3D extracellular vesicles for nasal septal cartilage defect repair. Int J Bioprint. 2024;10:4118. [Google Scholar]

[CR15] 15.Ding JY, Chen MJ, Wu LF, Shu GF, Fang SJ, Li ZY, et al. Mesenchymal stem cell-derived extracellular vesicles in skin wound healing: roles, opportunities and challenges. Mil Med Res. 2023;10(1): 36. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR16] 16.Zeng ZL, Xie H. Mesenchymal stem cell-derived extracellular vesicles: a possible therapeutic strategy for orthopaedic diseases: a narrative review. Biomater Transl. 2022;3(3):175–87. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR17] 17.van Griensven M, Balmayor ER. Extracellular vesicles are key players in mesenchymal stem cells’ dual potential to regenerate and modulate the immune system. Adv Drug Deliv Rev. 2024;207: 115203. [DOI] [PubMed] [Google Scholar]

[CR18] 18.Fan L, Wang L, Wang X, Zhang H. Exosomes-based particles as inhalable COVID-19 vaccines. Biomed Tech (Berl). 2023;4:24–7. [Google Scholar]

[CR19] 19.Phan J, Kumar P, Hao D, Gao K, Farmer D, Wang A. Engineering mesenchymal stem cells to improve their exosome efficacy and yield for cell-free therapy. J Extracell Vesicles. 2018;7(1): 1522236. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Hu Z, Qian S, Zhao Q, Lu B, Lu Q, Wang Y, et al. Engineering strategies for apoptotic bodies. Smart Medicine. 2024;3(3): e20240005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Ji C, Zhang J, Shi L, Shi H, Xu W, Jin J, et al. Engineered extracellular vesicle-encapsulated CHIP as novel nanotherapeutics for treatment of renal fibrosis. NPJ Regen Med. 2024;9(1): 3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Guo D, Sun Y, Wu J, Ding L, Jiang Y, Xue Y, et al. Photoreceptor-targeted extracellular vesicles-mediated delivery of Cul7 siRNA for retinal degeneration therapy. Theranostics. 2024;14(13):4916–32. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR23] 23.Xie SY, Liu SQ, Zhang T, Shi WK, Xing Y, Fang WX, et al. USP28 serves as a key suppressor of mitochondrial morphofunctional defects and cardiac dysfunction in the diabetic heart. Circulation. 2024;149(9):684–706. [DOI] [PubMed] [Google Scholar]

[CR24] 24.Mathew B, Ravindran S, Liu X, Torres L, Chennakesavalu M, Huang CC, et al. Mesenchymal stem cell-derived extracellular vesicles and retinal ischemia-reperfusion. Biomaterials. 2019;197:146–60. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Rao NA, Saraswathy S, Pararajasegaram G, Bhat SP. Small heat shock protein αA-crystallin prevents photoreceptor degeneration in experimental autoimmune uveitis. PLoS One. 2012;7(3):e33582. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR26] 26.Ruebsam A, Dulle JE, Myers AM, Sakrikar D, Green KM, Khan NW, et al. A specific phosphorylation regulates the protective role of αA-crystallin in diabetes. JCI Insight. 2018;3(4):e97919. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR27] 27.Yang W, Lu B, Chen Q, Wang J, Zhou J, Li R, et al. CRYAA activates the SIRT1-pi3K/AKT signaling pathway by suppressing mir-155-5p to protect the RPE. Arch Biochem Biophys. 2025;770: 110435. [DOI] [PubMed] [Google Scholar]

[CR28] 28.Christopher KL, Pedler MG, Shieh B, Ammar DA, Petrash JM, Mueller NH. Alpha-crystallin-mediated protection of lens cells against heat and oxidative stress-induced cell death. Biochim Biophys Acta. 2014;1843(2):309–15. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR29] 29.Liu X, Liu B, Zhou M, Fan F, Yu M, Gao C, et al. Circular RNA HIPK3 regulates human lens epithelial cells proliferation and apoptosis by targeting the miR-193a/CRYAA axis. Biochem Biophys Res Commun. 2018;503(4):2277–85. [DOI] [PubMed] [Google Scholar]

[CR30] 30.Tonade D, Kern TS. Photoreceptor cells and RPE contribute to the development of diabetic retinopathy. Prog Retin Eye Res. 2021;83: 100919. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR31] 31.Wu R, Fan X, Wang Y, Shen M, Zheng Y, Zhao S, et al. Mesenchymal stem cell-derived extracellular vesicles in liver immunity and therapy. Front Immunol. 2022;13:833878. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Song Y, You Y, Xu X, Lu J, Huang X, Zhang J, et al. Adipose-derived mesenchymal stem cell-derived exosomes biopotentiated extracellular matrix hydrogels accelerate diabetic wound healing and skin regeneration. Adv Sci. 2023;10(30):e2304023. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR33] 33.Zhang Z, Liang X, Zhou J, Meng M, Gao Y, Yi G, et al. Exosomes in the pathogenesis and treatment of ocular diseases. Exp Eye Res. 2021;209: 108626. [DOI] [PubMed] [Google Scholar]

[CR34] 34.Kaur G, Bae EH, Zhang Y, Ciacciofera N, Jung KM, Barreda H, et al. Biopotency and surrogate assays to validate the immunomodulatory potency of extracellular vesicles derived from mesenchymal stem/stromal cells for the treatment of experimental autoimmune uveitis. J Extracell Vesicles. 2024;13(8): e12497. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Tang Y, Kang Y, Zhang X, Cheng C. Mesenchymal stem cell exosomes as nanotherapeutics for dry age-related macular degeneration. J Control Release. 2023;357:356–70. [DOI] [PubMed] [Google Scholar]

[CR36] 36.Chen Y, Tong J, Liu C, He C, Xiang J, Yao G, et al. MSC-derived small extracellular vesicles mitigate diabetic retinopathy by stabilizing Nrf2 through miR-143-3p-mediated inhibition of neddylation. Free Radic Biol Med. 2024;219:76–87. [DOI] [PubMed] [Google Scholar]

[CR37] 37.Sun F, Sun Y, Wang X, Zhu J, Chen S, Yu Y, et al. Engineered mesenchymal stem cell-derived small extracellular vesicles for diabetic retinopathy therapy through HIF-1α/EZH2/PGC-1α pathway. Bioact Mater. 2024;33:444–59. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.Jia ZK, Fu CX, Wang AL, Yao K, Chen XJ. Cataract-causing allele in CRYAA (Y118D) proceeds through endoplasmic reticulum stress in mouse model. Zool Res. 2021;42(3):300–9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR39] 39.Zhou P, Luo Y, Liu X, Fan L, Lu Y. Down-regulation and CpG island hypermethylation of CRYAA in age-related nuclear cataract. FASEB J. 2012;26(12):4897–902. [DOI] [PubMed] [Google Scholar]

[CR40] 40.Besirli CG, Nath M, Yao J, Pawar M, Myers AM, Zacks D, et al. HSPB4/CRYAA protect photoreceptors during retinal detachment in part through FAIM2 regulation. Neurol Int. 2024;16(5):905–17. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR41] 41.Sun Y, Sun Y, Chen S, Yu Y, Ma Y, Sun F. Hypoxic preconditioned MSCs-derived small extracellular vesicles for photoreceptor protection in retinal degeneration. J Nanobiotechnol. 2023;21(1):449. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Gao Y, Li H, Qin C, Yang B, Ke Y. Embryonic stem cells-derived exosomes enhance retrodifferentiation of retinal Müller cells by delivering BDNF protein to activate Wnt pathway. Immunobiology. 2022;227(3):152211. [DOI] [PubMed] [Google Scholar]

[CR43] 43.Yang H, Ping X, Zhou J, Ailifeire H, Wu J, Nadal-Nicolás FM, et al. Reversible cold-induced lens opacity in a hibernator reveals a molecular target for treating cataracts. J Clin Invest. 2024;134(18): e169666. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Widjaja G, Jalil AT, Budi HS, Abdelbasset WK, Efendi S, Suksatan W, et al. Mesenchymal stromal/stem cells and their exosomes application in the treatment of intervertebral disc disease: a promising frontier. Int Immunopharmacol. 2022;105: 108537. [DOI] [PubMed] [Google Scholar]

[CR45] 45.Zhang X, Zhang H, Gu J, Zhang J, Shi H, Qian H, et al. Engineered extracellular vesicles for cancer therapy. Adv Mater. 2021;33(14): e2005709. [DOI] [PubMed] [Google Scholar]

[CR46] 46.Mathew B, Torres LA, Gamboa Acha L, Tran S, Liu A, Patel R, et al. Uptake and distribution of administered bone marrow mesenchymal stem cell extracellular vesicles in retina. Cells. 2021;10(4):730. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR47] 47.Wu P, Zhang B, Ocansey DKW, Xu W, Qian H. Extracellular vesicles: a bright star of nanomedicine. Biomaterials. 2021;269:120467. [DOI] [PubMed] [Google Scholar]

[CR48] 48.Tang TT, Wang B, Li ZL, Wen Y, Feng ST, Wu M, et al. Kim-1 targeted extracellular vesicles: a new therapeutic platform for RNAi to treat AKI. J Am Soc Nephrol. 2021;32(10):2467–83. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR49] 49.Gao X, Ran N, Dong X, Zuo B, Yang R, Zhou Q, et al. Anchor peptide captures, targets, and loads exosomes of diverse origins for diagnostics and therapy. Sci Transl Med. 2018;10(444):eaaw0534. [DOI] [PubMed] [Google Scholar]

[CR50] 50.Xu L, Xu X, Liang Y, Wen C, Ouyang K, Huang J, et al. Osteoclast-targeted delivery of anti-miRNA oligonucleotides by red blood cell extracellular vesicles. J Control Release. 2023;358:259–72. [DOI] [PubMed] [Google Scholar]

[CR51] 51.Herrera-Barrera M, Ryals RC, Gautam M, Jozic A, Landry M, Korzun T, et al. Peptide-guided lipid nanoparticles deliver mRNA to the neural retina of rodents and nonhuman primates. Sci Adv. 2023;9(2): eadd4623. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR52] 52.Zou Y, Li L, Li Y, Chen S, Xie X, Jin X, et al. Restoring cardiac functions after myocardial infarction-ischemia/reperfusion via an exosome anchoring conductive hydrogel. ACS Appl Mater Interfaces. 2021;13(48):56892–908. [DOI] [PubMed] [Google Scholar]

[CR53] 53.Yu S, Liao R, Bai L, Guo M, Zhang Y, Zhang Y, et al. Anticancer effect of hUC-MSC-derived exosome-mediated delivery of PMO-miR-146b-5p in colorectal cancer. Drug Deliv Transl Res. 2024;14(5):1352–69. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR54] 54.Han G, Zhang Y, Zhong L, Wang B, Qiu S, Song J, et al. Generalizable anchor aptamer strategy for loading nucleic acid therapeutics on exosomes. EMBO Mol Med. 2024;16(4):1027–45. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR55] 55.Zha Y, Li Y, Lin T, Chen J, Zhang S, Wang J. Progenitor cell-derived exosomes endowed with VEGF plasmids enhance osteogenic induction and vascular remodeling in large segmental bone defects. Theranostics. 2021;11(1):397–409. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR56] 56.Xu Y, Wei C, Ma L, Zhao L, Li D, Lin Y, et al. 3D mesenchymal stem cell exosome-functionalized hydrogels for corneal wound healing. J Control Release. 2025;380:630–46. [DOI] [PubMed] [Google Scholar]

[CR57] 57.Jia Y, Yu L, Ma T, Xu W, Qian H, Sun Y, et al. Small extracellular vesicles isolation and separation: current techniques, pending questions and clinical applications. Theranostics. 2022;12(15):6548–75. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR58] 58.Fu P, Zhang J, Li H, Mak M, Xu W, Tao Z. Extracellular vesicles as delivery systems at nano-/micro-scale. Adv Drug Deliv Rev. 2021;179: 113910. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Engineered MEVs for photoreceptor-targeted delivery of USP25 to alleviate diabetic retinopathy

Yaoxiang Sun

Shenyuan Chen

Xiaoyuan Qi

Yuntong Sun

Zhengmei Jiang

Jie Chang

Yongjun Ma

Jin Huang

Benshuai You

Fengtian Sun

Abstract

Graphical abstract

Supplementary Information

Background

Methods

Ethics

Antibodies

Characterization of MSCs

Extraction and characterization of MEVs

Cell culture, treatment and transfection

USP25 knockdown in MSCs

Preparation of MEVsMH42-USP25

In vivo distribution of mevs, MEVs-USP25, MEVsMH42, and MEVsMH42-USP25

Animals

Electroretinogram (ERG) analysis

Hematoxylin and Eosin (HE) staining

Terminal deoxynucleotidyl transferase dUTP Nick end labeling (TUNEL) staining

Immunofluorescence staining

LC-MS/MS

Western blot

Co-IP assay

EdU staining

Cell counting Kit-8 (CCK8) experiment

Safety evaluation in vivo

Statistical analysis

Results

MEVs alleviate photoreceptor injury in db/db mice

Fig. 1.

MEVs inhibit HG-induced apoptosis of 661 W cells

Fig. 2.

MEVs ameliorate hyperglycemia-induced photoreceptor injury by upregulating CRYAA expression

Fig. 3.

MEVs transfer USP25 to inhibit CRYAA ubiquitination

Fig. 4.

MEVs alleviate photoreceptor injury in db/db mice by delivering USP25

Fig. 5.

Preparation of MEVsMH42-USP25 for photoreceptor-targeted delivery of USP25

Fig. 6.

Enhanced efficiency of MEVsMH42-USP25 in photoreceptor therapy

Fig. 7.

Discussion

Conclusion

Supplementary Information

Acknowledgements

Abbreviations

Author contributions

Funding

Data availability

Declarations

Ethics approval and consent to participate

Consent for publication

Competing interests

Footnotes

Contributor Information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Preparation of MEVs^MH42-USP25

In vivo distribution of mevs, MEVs-USP25, MEVs^MH42, and MEVs^MH42-USP25

Preparation of MEVs^MH42-USP25 for photoreceptor-targeted delivery of USP25

Enhanced efficiency of MEVs^MH42-USP25 in photoreceptor therapy