DNA hydrogel-delivered hypoxic exosomes alleviate radiation-induced skin injury by stabilizing AKT

Abstract

Background

Radiation-induced skin injury (RISI) is a common and refractory complication during tumor radiotherapy, characterized by radiation stress-induced impairment of keratinocyte viability and regenerative signaling, ultimately leading to delayed skin repair. Recently, hypoxia-preconditioned adipose-derived stem cell-derived exosomes (A-Hexos), a cell-free therapeutic strategy with favorable biocompatibility, has considerable potential in skin injury repair. However, the underlying mechanisms of A-Hexos in the treatment of RISI have not been fully elucidated, and their delivery efficiency and retention capacity at skin injury sites remain suboptimal. In this study, we aimed to enhance the targeted delivery efficiency of exosomes in skin tissues, systematically evaluate the therapeutic effects of A-Hexos in RISI repair, and elucidate the underlying molecular mechanisms.

Results

In vitro experiments demonstrated that A-Hexos significantly restored the viability, proliferative capacity, and migratory behavior of irradiated keratinocytes (HaCaT). To enhance the local delivery efficiency and tissue retention of exosomes at injured sites in vivo, we constructed a DNA hydrogel-based delivery system loaded with A-Hexos (Gel@A-Hexos), which was engineered through the specific interaction between an exosome membrane protein-specific aptamer (Apt CD63) designed on long DNA strands and the exosomal membrane protein, CD63. In a murine RISI model, Gel@A-Hexos markedly alleviated radiation-induced injury and promoted skin tissue structural reconstruction. Mechanistically, miRNA sequencing revealed that miR-486-5p was significantly enriched in A-Hexos, while proteomic analysis further indicated that Gel@A-Hexos treatment markedly downregulated RNF213 expression and concomitantly upregulated AKT expression in injured skin tissues. Functionally, miR-486-5p delivered by A-Hexos stabilized AKT protein levels by inhibiting RNF213-mediated ubiquitin-dependent AKT degradation, thereby alleviating RISI and promoting tissue repair.

Conclusion

Collectively, this study proposes a long DNA strand-based exosome local delivery hydrogel and elucidates the molecular mechanism by which A-Hexos exert reparative effects on RISI via the miR-486-5p/RNF213/AKT signaling axis, providing new theoretical insights and a potential translational strategy for the precise treatment of RISI.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04653-2.

Background

Radiation-induced skin injury (RISI) is one of the most common complications during tumor radiotherapy, with approximately 95% of patients receiving radiotherapy experiencing varying degrees of skin damage during treatment [1]. While ionizing radiation effectively eradicates tumor cells, it inevitably damages normal skin tissue, triggering persistent inflammatory responses, impaired cellular regenerative capacity, and destruction of skin structure, thereby markedly delaying the wound healing process [2, 3]. Therefore, the occurrence of RISI is not merely a direct consequence of radiation exposure but also closely associated with cellular functional imbalance within the skin microenvironment, making it an urgent clinical therapeutic challenge.

Recently, adipose-derived stem cells (ADSCs) have attracted widespread attention owing to their marked tissue regenerative potential. However, direct cell transplantation is associated with safety concerns, including immune rejection, low survival rate, and potential tumorigenicity. Accumulating evidence indicates that the reparative effects of stem cells are primarily mediated by their paracrine activity [4]. Exosomes exhibit superior biosafety, lower immunogenicity, and greater clinical translational potential than cell-based therapies [5]. In addition, exosomes can cross biological barriers and efficiently deliver bioactive molecules to recipient cells. However, several limitations persist, including limited targeting specificity and the lack of standardized methods for isolation, purification, and large-scale production, which hinder clinical translation.

ADSC-derived exosomes, as key mediators of paracrine signaling, are nanosized extracellular vesicles (30–150 nm) enriched with diverse bioactive cargos, including miRNAs, lncRNAs, proteins, and lipids. Among these, miRNAs are considered critical functional effectors, with several well-characterized molecules such as miR-21 [6], miR-146a [7], and miR-125b [8]. In addition, lncRNAs such as H19 [9] and MALAT1 [10] have been shown to regulate ADSC exosome-mediated cellular repair processes. These bioactive cargos collectively contribute to the anti-inflammatory, cytoprotective, and pro-angiogenic properties of ADSC-derived exosomes. These properties underlie the potential of ADSC-derived exosomes to mitigate radiation-induced cellular damage and promote tissue repair. Accumulating evidence support the therapeutic potential of ADSC-derived exosomes in radiation-related injuries. For instance, ADSC-derived exosomes have been reported to alleviate acute radiation-induced dermatitis by upregulating hyaluronic acid synthase 1 expression [11]. Moreover, ADSC-based therapies have shown promising efficacy in mitigating radiation-induced fibrosis [12].

Hypoxia is an important physiological stimulus capable of inducing adaptive cellular responses [13]. In vivo, ADSCs are typically exposed to a hypoxic microenvironment of approximately 2% O₂, and moderate hypoxia is considered a critical physiological feature for maintaining their stemness and reparative potential [14, 15]. Appropriate hypoxic conditions can upregulate the expression of multiple factors associated with cell survival, proliferation, and migration in ADSCs, thereby enhancing their metabolic adaptability and tissue regenerative capacity [16, 17]. Hypoxic preconditioning can markedly remodel the miRNA composition and protein profiles of ADSC exosomes, thereby strengthening their ability to regulate target cell responses to adverse microenvironmental stimuli [18, 19]. Additionally, hypoxia-derived ADSC exosomes exert superior pro-regenerative effects compared with those in normoxia-derived exosomes in bone repair [20] and diabetic wound healing [21]. However, the therapeutic potential and molecular mechanisms of hypoxia-preconditioned ADSC exosomes (A-Hexos) in RISI is unclear.

Although ADSC exosomes have demonstrated promising therapeutic potential in various tissue injury models, their reparative efficacy is still constrained by two critical issues. The molecular mechanisms underlying the core reparative functions of exosomes have not yet been systematically elucidated, and the limited in vivo stability and insufficient local retention of exosomes markedly restrict their therapeutic efficiency at injury sites. In the context of the unique stress environment of RISI, enhancing biological effects of exosomes and achieving effective delivery to injured regions remains unresolved challenges.

An emerging class of programmable biomaterials, DNA hydrogels, exhibit unique advantages in drug delivery and tissue engineering owing to their precise sequence design capability, favorable biocompatibility, and controllable degradability [22]. Unlike most conventional hydrogels that lack “programmability,” DNA hydrogels incorporate nucleic acid aptamers capable of recognizing the exosomal transmembrane protein CD63 (Apt CD63) through rational sequence design [23], thereby enabling active capture and efficient loading of exosomes, stable immobilization, and sustained release within the local microenvironment. Moreover, the self-assembly property of DNA molecules based on complementary base pairing circumvents the need for chemical crosslinkers, such as glutaraldehyde or photo-crosslinking agents commonly required in conventional systems. This eliminated residual toxicity, and confers excellent biocompatibility and suitable mechanical properties, which provide a three-dimensional supportive microenvironment for exosomes and prolong their subcutaneous retention time [24].

We constructed a DNA hydrogel delivery system loaded with A-Hexos (Gel@A-Hexos), systematically evaluated its therapeutic efficacy in RISI repair, and further elucidated its key molecular mechanisms(Scheme 1), providing novel biomaterial-based strategies and theoretical evidence for the precise treatment of RISI.

Scheme 1.

Schematic diagram of DNA hydrogel-delivered hypoxic exosomes application in radiation-induced skin injury treatment

Methods

Cell culture

Human ADSCs and HaCaT cells were purchased from Procell Life Science & Technology Co. Ltd. (Wuhan, China). ADSCs were maintained in MSCM medium (ScienCell, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, Waltham, MA, USA), 1% penicillin-streptomycin (P/S), and mesenchymal stem cell growth supplement. HaCaT cells were cultured in RPMI-1640 medium (Gibco) containing 10% FBS and 1% P/S. Cells were incubated at 37 °C in a humidified atmosphere with 5% CO₂ and only early-passage cells were used for experiments.

Hypoxic treatment of ADSCs

When ADSCs reached 70–80% confluence, the medium was replaced with exosome-depleted FBS medium (Gibco). Cells were transferred into a hypoxia chamber (2% O₂, 5% CO₂; Panasonic, Japan) and cultured for 48 h. Control ADSCs were maintained under normoxic conditions (21% O₂). Conditioned media were collected for subsequent exosome isolation.

Exosome isolation

Exosomes were isolated from ADSC culture supernatants using differential ultracentrifugation. Supernatants were sequentially centrifuged at 300×g (10 min), 2,000×g (20 min), and 10,000×g (30 min) to remove cells and debris. The clarified solution was ultracentrifuged at 100,000×g for 70 min at 4 °C using a Beckman Optima L-100XP system(Brea, CA, USA). Pellets were washed with PBS and ultracentrifuged again at 100,000×g for 70 min. Purified exosomes were resuspended in PBS and stored at − 80 °C.

Exosome characterization

Exosome morphology was examined using transmission electron microscopy (Hitachi, Japan) after negative staining. Particle size and concentration were measured using nanoparticle tracking analysis (Particle Metrix, Germany). Protein expression of exosomal markers, CD63, CD81, and TSG101 (Abcam, UK), was assessed using western blotting. Total protein content was quantified using a BCA assay kit (Beyotime, China).

Cellular uptake of exosomes

Exosomes were labeled with PKH67 dye (Sigma-Aldrich, USA) following the manufacturer’s protocol. Excess dye was removed using ultracentrifugation. Labeled exosomes were co-incubated with HaCaT cells for 6 h, followed by fixation with 4% paraformaldehyde. Cells were stained with phalloidin (Solarbio Life Sciences, Beijing, China) and DAPI, and fluorescence images were captured using a Nikon confocal microscope(Tokyo, Japan).

CCK-8 cell viability assay

Cell viability was measured using a CCK-8 kit (Yeasen, China). Cells were seeded in 96-well plates and treated with normoxic or hypoxic A-Exos (10 µg/mL). At designated time points, CCK-8 solution was added and incubated for 2 h. Absorbance at 450 nm was recorded using a microplate reader (Bio-Rad Laboratories, Hercules, CA, USA).

LDH cytotoxicity assay

Cytotoxicity was assessed using an LDH assay kit (Beyotime, C0016). HaCaT cells were seeded in 96-well plates, treated as indicated, and cell-free supernatants were collected. LDH working solution was added and incubated for 30 min, and absorbance was measured at 490 nm.

Wound healing migration assay

Cells were seeded into six-well plates and allowed to form a monolayer. Straight scratches were created using a sterile pipette tip and debris was removed by PBS washes. Cells were cultured with exosomes (10 µg/mL), and wound areas were photographed at 0, 12, and 24 h. Migration rates were quantified using ImageJ(v6.0, Build 25, NIH, USA).

EdU proliferation assay

Cell proliferation was evaluated using an EdU assay kit (Beyotime). Cells were incubated with EdU working solution for 2 h, fixed with 4% paraformaldehyde, permeabilized, and stained according to the kit protocol. Fluorescence images were obtained with a Nikon inverted microscope.

ELISA

The concentrations of IL-6 and IL-1β in culture supernatants were measured using ELISA kits (Beyotime). Samples and standards were added to antibody-coated plates, followed by a detection antibody, HRP conjugate, and TMB substrate. Absorbance at 450 nm was recorded using a microplate reader.

Preparation of DNA hydrogel

All DNA oligonucleotides used in this study were custom-synthesized by Sangon Biotech (Shanghai, China). To prepare circular DNA (circ-DNA), phosphorylated linear single-stranded DNA templates were mixed with primers at a 1:1 molar ratio. The mixture was heated at 95 °C for 2 min and 65 °C for 2 min, followed by gradual cooling from 60 °C to 20 °C at − 0.5 °C per 30 s using a PCR thermal cycler (Bio-Rad Laboratories). After annealing, 2 U of T4 DNA ligase (New England Biolabs, Ipswich, MA, USA) was added and incubated at 22 °C for 12 h. Circularization was confirmed using 12% native PAGE performed at 20 V/cm for 2 h on ice. For rolling circle amplification (RCA), circ-DNA templates (circ-DNA-1 and circ-DNA-2, 50 nM each) were mixed with dNTPs (1 mM), BSA (0.2 mg/mL; Solarbio Life Sciences), NaCl (80 mM), and phi29 DNA polymerase (200 U/mL; New England Biolabs) in RCA buffer (50 mM Tris-HCl, 10 mM MgCl₂, 10 mM (NH₄)₂SO₄, 4 mM DTT, pH 7.5; Biometa Life Science Co, Ltd., Ningbo, China). The reaction was incubated at 37 °C with shaking at 450 rpm, followed by enzyme inactivation at 65 °C for 10 min. The two RCA products (DNA-chain-1 and DNA-chain-2) were then mixed at equal volumes and incubated at 37 °C and 450 rpm to self-assemble into the DNA hydrogel.

Structural and rheological analysis of hydrogel

Microstructures of DNA hydrogel were observed using a scanning electron microscope (Hitachi SU8100, Japan). Rheological properties were assessed using a TA Instruments AR-G2 rheometer with a 20-mm parallel plate, under 1% strain and 1 Hz frequency at 25 °C.

Exosome encapsulation in DNA hydrogel

CM-DiI-labeled exosomes were mixed with SYBR-stained DNA-chain-1 at 37 °C for 30 min, followed by addition of DNA-chain-2 to complete gelation. Fluorescence images were obtained to visualize encapsulated exosomes.

RISI model

C57BL/6J mice (6 weeks old, ~ 25 g) were housed in a specific pathogen-free animal facility with free access to food and water. All animal experiments were approved by The Animal Policy and Welfare Committee of Chengdu Medical College (Approval No: CMC-IACUC-2021024).

C57BL/6J mice were anesthetized with 2% isoflurane and hair on the right hind limb was removed using depilatory cream. Mice were placed on a polystyrene board, and non-target regions (head, trunk, and forelimbs) were shielded with 5–10 mm lead plates, leaving only the hind limb exposed. Irradiation was performed using an X-ray irradiator (Precision X-Ray Inc., North Branford, CT, USA) at a dose rate of approximately 1–2 Gy/min to deliver a single 30 Gy exposure (225 kV, 13 mA, 0.5 mm Cu filter, SSD 30 cm). After irradiation, the mice were allowed to recover on a warming pad and then euthanized under deep anesthesia on day 14 post-irradiation. Subsequently, full-thickness skin tissues from the irradiated site were harvested for histological analysis.

RISI was evaluated using a standardized scoring system. Briefly, skin reactions were graded on a scale of 1.0 to 5.5 based on severity: 1.0, no visible effect; 1.5, minimal erythema with mild dry skin; 2.0, moderate erythema and dry skin; 2.5, marked desquamation, minimal dry crusting; 3.0, dry desquamation, minimal dry crusting; 3.5, dry desquamation, dry crusting, superficial minimal scabbing; 4.0, patchy moist desquamation, moderate scabbing; 4.5, confluent moist desquamation, ulcers, large deep scabs; 5, open wound, full-thickness loss; 5.5, necrosis [11, 25].

Hematoxylin and eosin (H&E), Masson, immunohistochemistry (IHC), and TUNEL staining

Skin tissue paraffin sections were processed for H&E staining, Masson trichrome staining, immunohistochemistry (IL-6 and TNF-α; Abcam), and TUNEL staining (Solarbio Life Sciences) following standard protocols. Images were captured using a light or fluorescence microscope.

Western blotting

Total protein was extracted using RIPA buffer (Beyotime) with PMSF. Protein samples were separated using 12% SDS-PAGE, transferred to PVDF membranes (Millipore, Burlington, MA, USA), blocked, and incubated with primary and HRP-conjugated secondary antibodies. Signals were detected using ECL reagent (Yeasen).

RT-qPCR

Total RNA was extracted using an RNA extraction kit (Bio-Rad Laboratories). cDNA was synthesized using a reverse transcription kit (Bio-Rad Laboratories)and amplified using SYBR Green mix on a Bio-Rad CFX96 system. For miRNA analysis, reverse transcription was performed using the stem-loop method with the miRNA cDNA First Strand Synthesis Kit (Accurate Biotechnology, China). Primers used for miRNA qRT-PCR are listed in Additional file, Table 3. GAPDH and U6 served as internal controls. The primer sequences for RNF213 were as follows: forward (F), AACCTCAGATGCCAGCTCAC; reverse (R), CCATACAGCAGTGCAGGACA.

Cell transfection experiments

MiR-486-5p mimic/inhibitor (Jiman Biotech, China), siRNA, and RNF213 expression plasmids were transfected using Lipofectamine RNAiMAX or Lipofectamine 3000 (Thermo Fisher Scientific, Waltham, MA, USA). Transfection efficiency was evaluated using RT-qPCR or western blotting.

Dual-luciferase reporter assay

HaCaT cells were co-transfected with pGL3-RNF213 reporter plasmids (wild-type or mutated, 200 ng) and pRL-TK Renilla plasmid (20 ng), together with 50 nM miR-486-5p mimic or control using Lipofectamine 3000. After 24 h, cells were lysed and luciferase activities were quantified using the Dual-Luciferase Reporter Assay System (Promega, E1910, Madison, WI, USA). Firefly signals were normalized to Renilla activity, and data were expressed relative to the WT + NC group. All assays were performed in triplicate with at least three independent experiments.

Molecular docking analysis

The three-dimensional structural model of RNF213 was obtained using SWISS-MODEL (https://swissmodel.expasy.org/repository/), while the structure of AKT1 was retrieved from the AlphaFold Protein Structure Database (https://alphafold.ebi.ac.uk). Subsequently, protein–protein blind docking was performed using HDOCK and the plausible interactions were visualized and mapped using PyMOL.

Statistical analysis

Data are presented as mean ± SD. Statistical analyses were performed using GraphPad Prism 9.0. One-way ANOVA with Tukey’s post-hoc test or Student’s t-test was applied where appropriate. A p-value < 0.05 was considered statistically significant.

Results

Characterization of A-Exos and A-Hexos

As shown in Fig. 1A, human ADSCs were cultured under normoxic and hypoxic conditions, and the culture supernatants were collected for exosome isolation. The exosomal markers CD63, CD81, and TSG101 were all positively detected (Fig. 1B). Transmission electron microscopy revealed that the vesicles displayed a round, membrane-bound morphology, and nanoparticle tracking analysis showed that A-Exos (130.7 nm) and A-Hexos (116.4 nm) exhibited size distributions consistent with that of typical exosomes (Fig. 1C).

Fig. 1.

Characterization of A-Exos and A-Hexos. (A) Schematic illustration of the isolation of A-Exos and A-Hexos. Created with BioGDP.com. (B) Representative western blot bands of exosomal markers CD63, CD81, and TSG101. (C) Transmission electron microscopy images showing the morphology of A-Exos and A-Hexos (scale bar = 100 nm) And Nanoparticle tracking analysis of A-Exos and A-Hexos. (D) Volcano plot of differentially expressed miRNAs, with blue and red indicating downregulated and upregulated expression, respectively. (E) Heatmap of all significantly differentially expressed miRNAs between A-Exos and A-Hexos. (F, G) Functional enrichment analysis of differentially expressed proteins: (F) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment; (G) Gene Ontology (GO) enrichment, including biological process, molecular function, and cellular component. Exos, exosomes; Hexos, hypoxia-preconditioned exosomes

MiRNAs carried by exosomes play essential roles in regulating gene expression, and are involved in diverse physiological and pathological processes. To investigate the impact of hypoxic stimulation on the miRNA profiles of exosomes, we performed high-throughput sequencing to compare miRNAs between A-Exos and A-Hexos. Hypoxia markedly altered the miRNA composition of A-Exos and the volcano plot identified 13 miRNAs with significant differential expression (Fig. 1D). Hierarchical clustering based on these miRNAs revealed a clear separation between the two groups (Fig. 1E). Functional annotation indicated that the target genes of these miRNAs were enriched in multiple terms related to cellular viability and proliferation, including protein kinase activity and intracellular signal transduction (Fig. 1F). Consistently, KEGG pathway enrichment showed significant involvement of MAPK, mTOR, and AMPK pathways (Fig. 1G), which are closely associated with the regulation of cell survival and metabolic processes, suggesting that these miRNAs may cooperatively modulate several signaling pathways to enhance the viability, proliferation, and migratory potential of injured cells.

A-Hexos attenuate radiation-induced injury in HaCaT cells

After confirming that hypoxic conditions markedly altered the miRNA profile of ADSC-derived exosomes, the effects of A-Exos and A-Hexos on radiation-induced cellular responses were further evaluated. HaCaT cells were exposed to 20 Gy of X-ray irradiation to establish an in vitro radiation injury model, and an exosome concentration of 10 µg/mL was used for subsequent experiments.

To examine cellular uptake, exosomes were labeled with PKH67 and co-incubated with HaCaT cells. Confocal laser scanning microscopy revealed abundant green fluorescence signals distributed throughout the cytoplasm (Fig. 2A), indicating efficient internalization of both normoxic and hypoxic exosomes. Cell viability assays demonstrated that 20 Gy irradiation reduced HaCaT cell viability to approximately 65% (Fig. 2B). Treatment with A-Exos or A-Hexos increased cell viability to approximately 71% and 78%, respectively. Consistently, LDH release assays (Fig. 2C) and 7-AAD staining (Fig. 2D, E) showed that A-Hexos treatment significantly reduced radiation-induced membrane damage, with a ~ 24% decrease in LDH release and markedly lower proportion of 7-AAD-positive cells compared with those in the irradiated group. Cell proliferation was further evaluated using EdU proliferation assays (Fig. 2F, G), which revealed a significantly higher proportion of EdU-positive cells in the A-Hexos group than those in the A-Exos group. Clonogenic survival assays (Fig. 2H, I) showed that A-Hexos treatment increased colony formation by approximately 1.8-fold relative to irradiated controls. The effects of exosomes on cell migration were assessed using scratch wound assays (Fig. 2J, K). At 24 h post-scratch, the migration rate of irradiated cells was approximately 41%, which increased to 55% following A-Exos treatment and further to approximately 62% in the A-Hexos group.

Fig. 2.

A-Hexos attenuate radiation-induced injury in HaCaT cells. (A) Representative laser confocal microscopy images showing the uptake of PKH67-labeled exosomes (green) by HaCaT cells. Nuclei were stained with DAPI (blue). Scale bar = 10 μm. (B) CCK-8 assay assessing HaCaT cell viability after irradiation and treatment with A-Exos or A-Hexos. (C) LDH release assay evaluating the degree of plasma membrane damage in HaCaT cells. (D, E) Analysis of cell death by 7-AAD staining and flow cytometry: (D) representative live/dead cell distribution; (E) quantification of 7-AAD–positive cells. (F, G) EdU assay detecting HaCaT cell proliferation: (F) representative fluorescence images; (G) statistical analysis of EdU-positive cell ratio. Scale bar = 50 μm. (H, I) Colony formation assay evaluating long-term proliferative capacity: (H) representative dish images; (I) quantification of colony number. (J, K) Wound healing assay assessing cell migration: (J) representative images at 0, 12, and 24 h; (K) quantification of migration rate. Scale bar = 200 μm. (L, M) ELISA quantification of inflammatory cytokines IL-1β (L) and IL-6 (M) in the culture supernatant of HaCaT cells. HaCaT cells were exposed to 20-Gy radiation and treated with A-Exos or A-Hexos (10 µg/mL). Data are presented as mean ± SD (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001

To evaluate inflammatory responses, the levels of IL-1β and IL-6 in the culture supernatants were measured (Fig. 2L, M). Irradiation markedly increased the secretion of both cytokines, whereas exosome treatment significantly attenuated this response. Notably, A-Hexos exhibited a more pronounced reduction in IL-1β and IL-6 levels. Collectively, these results demonstrate that hypoxia-preconditioned ADSC-derived exosomes more effectively mitigate radiation-induced cellular injury by enhancing cell survival, proliferation, and migration while suppressing inflammatory responses.

Synthesis and characterization of exosome-loaded DNA hydrogel

The DNA hydrogel was assembled based on complementary hybridization between two ultralong single-stranded DNA chains (DNA-Chain-1 and DNA-Chain-2). Upon mixing the two chains in a 1.5-mL tube and gently shaking at 37 °C for 0.5 h, a semi-transparent hydrogel was formed (Fig. 3A). Circular templates (circ-DNA-1 and circ-DNA-2) were prepared by primer-annealing-mediated cyclization and served as templates for RCA. Non-denaturing PAGE confirmed successful cyclization, as both circ-DNA-1 and circ-DNA-2 exhibited slower migration compared with that of their linear precursors (Fig. 3B). Scanning electron microscopy revealed that the exosome-loaded DNA hydrogel retained a three-dimensional porous architecture, indicating that exosome incorporation did not markedly alter the network morphology (Fig. 3C). Rheological measurements showed that the storage modulus (G′) was substantially higher than the loss modulus (G″) within the 0–180 s time sweep, demonstrating the formation of an elastic hydrogel network (Fig. 3D).

Fig. 3.

Synthesis and characterization of exosome-loaded DNA hydrogel. (A) Bright-field image of the DNA hydrogel. (B) Non-denaturing PAGE verifying the successful circularization of RCA templates circ-DNA-1 and circ-DNA-2. Compared with their linear counterparts. (C) Scanning electron microscopy images of the blank DNA hydrogel and exosome-loaded DNA hydrogel. (D) Rheological characterization of DNA hydrogels showing storage modulus (G′) and loss modulus (G″). (E) Cumulative protein release curves over 15 days from hydrogels loaded with different exosome concentrations (Gel@Exo1, 1 mg/mL; Gel@Exo2, 2 mg/mL; Gel@Exo3, 3 mg/mL) and control hydrogels without exosomes. (F) Confocal fluorescence microscopy images showing the colocalization of SYBR Green I-labeled DNA hydrogels (green) and PHK26-labeled exosomes (red). (G) Photograph showing the injectability of the hydrogel through a syringe needle. (H, I) Representative images (H) and quantitative degradation curves (I) of Evans blue-labeled hydrogels incubated in DNase I (1 U/mL) solution for 12 h. (J, K) Representative images (J) and quantitative hemolysis analysis (K) of red blood cells incubated with hydrogels at different concentrations. (L) CCK-8 assay evaluating cell viability after 24 h co-culture with DNA hydrogels at different concentrations. Data are presented as mean ± SD (n = 3)

To evaluate the release behavior after exosome encapsulation, samples loaded with different exosome concentrations (1–3 mg/mL; designated Gel@Exo1, Gel@Exo2, and Gel@Exo3) and a blank DNA hydrogel (Gel@Blank) were quantitatively analyzed. The blank hydrogel exhibited negligible protein release over 15 d, whereas exosome-loaded samples displayed a sustained release profile, with the release rate positively correlated with the initial exosome concentration (Fig. 3E). Fluorescence colocalization imaging showed substantial overlap between the SYBR Green I-labeled DNA framework (green) and PHK26-stained exosomes (red), confirming efficient retention of exosomes within the hydrogel network (Fig. 3F). The DNA hydrogel exhibited good injectability and could be readily extruded through a syringe (Fig. 3G). In degradation assays, Evans blue-stained hydrogels incubated with 1 U/mL DNase I for 12 h showed over 80% mass loss (Fig. 3H, I), indicating enzyme-responsive degradability. Hemocompatibility tests demonstrated hemolysis rates below 3% across all hydrogel groups (Fig. 3J, K). CCK-8 assays further showed stable cell viability after co-culture with hydrogels of different concentrations, with no significant decrease compared with that in controls, indicating favorable cytocompatibility in vitro (Fig. 3L).

Gel@A-hexos significantly alleviates RISI

We evaluated the therapeutic potential of Gel@A-Hexos in promoting post-radiation skin repair using a murine RISI model. A single dose of 30 Gy X-ray was applied to establish the injury model (Fig. 4A). In the treatment groups (Gel@Exos and Gel@A-Hexos), DNA hydrogels loaded with 30 µg exosomes were subcutaneously injected at multiple points within the damaged area on day 0 and 7 post-irradiation, whereas the hydrogel-only group received an equivalent volume of blank DNA hydrogel. Skin appearance was recorded on days 0, 7, and 14, and mice were sacrificed on day 14 for histological and immunological analyses of skin tissues.

Fig. 4.

Gel@A-Hexos significantly alleviates RISI. (A) Schematic illustration of the experimental timeline (Control, IR, IR + Gel@A-Exos, IR + Gel@A-Hexos). (B) Evaluation criteria for RISI scores among different groups. (C) Representative macroscopic images of skin injury on mouse hind limbs at days 0, 7, and 14 after irradiation. (D) Hematoxylin and eosin staining of skin tissue collected on day 14 post-irradiation. Scale bar = 100 μm. (E) Masson’s trichrome staining of skin sections showing collagen deposition. Scale bar = 100 μm. (F) Immunohistochemical staining of proinflammatory cytokines, IL-6 and IL-1β, in skin tissue. Scale bars = 100 μm. (G) TUNEL staining (green) of skin tissue; the nuclei were stained with DAPI (blue). Scale bars = 100 μm. Mice were exposed to 30-Gy radiation and treated with Gel@A-Exos or Gel@A-Hexos via local administration

Typical RISI symptoms, including impaired hair regrowth, edema, desquamation, and erythema, were observed in the blank DNA hydrogel group (Fig. 4B). RISI was semi-quantitatively evaluated on days 0, 7, and 14 post-irradiation using the RISI scoring system (Fig. 4C). H&E staining (Fig. 4D) revealed pronounced pathological alterations in irradiated skin, including epidermal hyperplasia, dermal disorganization, increased eosinophilia, and extensive inflammatory cell infiltration. In contrast, these pathological changes were markedly alleviated in the treatment groups, with the Gel@A-Hexos group exhibiting skin architecture most closely resembling that of the normal control.

Collagen deposition and organization are critical indicators of wound healing; moderate and orderly deposition promotes tissue repair, whereas excessive or disordered accumulation may lead to hypertrophic scarring and impede normal healing. Masson’s trichrome staining demonstrated abundant subcutaneous collagen deposition in the blank DNA hydrogel group, indicating pronounced radiation-induced skin fibrosis. Compared with the Gel@A-Exos group, collagen fibers in the Gel@A-Hexos group were more loosely arranged and orderly, with markedly reduced deposition, suggesting that Gel@A-Hexos more effectively mitigates radiation-induced fibrosis and facilitates tissue reconstruction (Fig. 4E). Immunohistochemical analysis of IL-6 and IL-1β expression revealed that Gel@A-Hexos significantly reduced inflammatory cytokine levels in both the epidermis and dermis (Fig. 4F). TUNEL staining further demonstrated that the Gel@A-Hexos composite system effectively inhibited radiation-induced apoptosis, preserving tissue cell integrity and accelerating skin repair (Fig. 4G).

Furthermore, to assess the in vivo toxicity and safety of the DNA hydrogel, histopathological and biochemical analyses were conducted. On day 14 after DNA hydrogel injection, we harvested the heart, liver, spleen, lung, and kidney of mice for evaluation. H&E staining of these major organs revealed no significant inflammatory responses or structural abnormalities, with tissue morphology largely comparable to that of the control group (Fig. 5A). Serum biochemical analyses further confirmed the absence of systemic toxicity; liver function indicators (ALT and AST), renal function indicators (CREA and BUN), cardiac injury marker (CK), and lipid metabolism parameter (T-CHO) showed no significant differences between the DNA hydrogel-treated and control groups (Fig. 5B–G). In addition to this long-term assessment, acute toxicity was evaluated at 24 and 48 h following injection, and the results presented in Additional file, Figure S1. Collectively, at the evaluated dosage, the DNA hydrogel exhibited no apparent toxicity to major organs or adverse effects on systemic biochemical markers, underscoring its favorable safety profile and supporting its potential application in further studies.

Fig. 5.

Potential toxicity and safety evaluation of DNA hydrogel. (A) Representative hematoxylin and eosin (H&E)-stained images of the heart, liver, spleen, kidney, and intestinal tissues from mice on day 14 after DNA hydrogel treatment. Scale bar: 100 μm. (B–G) Biochemical analyses of mouse serum on day 14 after DNA hydrogel treatment, including AST, ALT, CK, CREA, BUN, and T-CHO levels. Data are presented as mean ± SD (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001

MiR-486-5p is upregulated in a-hexos and targets RNF213

To elucidate the mechanism by which A-Hexos protect against RISI, we further analyzed our previous miRNA sequencing results and identified eight miRNAs that were markedly upregulated in A-Hexos (Fig. 1D). Among them, miR-486-5p showed the most pronounced increase upon further validation (Fig. 6A). In addition, we performed proteomic analysis on mouse skin tissues treated with Gel@A-Exos and Gel@A-Hexos. The volcano plots revealed numerous differentially expressed proteins between the two groups (Fig. 6C), and hierarchical clustering heatmaps further demonstrated clearly distinct protein expression profiles (Fig. 6B), suggesting that A-Hexos activated signaling pathways distinct from those induced by A-Exos. GO enrichment analysis indicated that these differential proteins were mainly associated with biological processes, such as keratinization and epidermal differentiation, and significantly enriched in extracellular matrix structural components and immune response-related molecular functions (Fig. 6D), implying substantial alterations in barrier reconstruction and matrix remodeling following A-Hexos treatment. KEGG enrichment analysis further revealed that these differential proteins were broadly involved in chemokine signaling, and other immune defense and inflammation-related pathways (Fig. 6E), suggesting that they may regulate the immune microenvironment and facilitate extracellular matrix reconstruction to enhance skin repair after radiation injury. To identify the functional targets of miR-486-5p, we intersected the genes encoding significantly downregulated proteins in the proteomic dataset with the predicted targets of miR-486-5p (Fig. 6F). We identified RNF213 as one of the potential targets of miR-486-5p (Fig. 6G). Subsequent validation showed that radiation increased RNF213 protein levels in skin tissues, whereas Gel@A-Hexos treatment markedly reduced its expression (Fig. 6H). Using luciferase reporter constructs containing the wild-type or mutated RNF213 3′-UTR, we found that miR-486-5p significantly suppressed luciferase activity of the wild-type reporter, whereas mutation of the 3′-UTR completely abolished this effect (Fig. 6I), confirming RNF213 as a downstream target of miR-486-5p.

Fig. 6.

MiR-486-5p is upregulated in A-Hexos and targets RNF213. (A) qPCR validation of candidate miRNAs. (B) Volcano plot of differentially expressed proteins between Gel@A-Exos and Gel@A-Hexos groups; blue and red dots represent upregulated and downregulated proteins, respectively. (C) Hierarchical clustering heatmap of differentially expressed proteins in skin tissues from the Gel@A-Exos and Gel@A-Hexos groups. (D, E) Functional enrichment analysis of differentially expressed proteins: (D) Gene Ontology (GO) enrichment analysis, including biological processes, molecular functions, and cellular components; (E) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. (F) Venn diagram showing the overlap between the predicted target genes of miR-486-5p and the significantly downregulated proteins in mouse skin following Gel@A-Exos and Gel@A-Hexos treatment. (G) Schematic illustration of the complementary binding sites between miR-486-5p and the 3′-UTR region of RNF213 mRNA. (H) Western blot analysis of RNF213 protein expression in mouse skin tissues after irradiation and following Gel@A-Hexos treatment. (I) Dual-luciferase reporter assay showing the relative luciferase activity of wild-type (WT) and mutant (MUT) RNF213 3′-UTR constructs co-transfected with miR-486-5p mimic. (J, K) qPCR analysis of miR-486-5p expression in HaCaT cells after transfection with (J) miR-486-5p inhibitor or (K) miR-486-5p mimic. (L) qPCR analysis of RNF213 mRNA expression in HaCaT cells transfected with an miR-486-5p mimic or inhibitor. (M) Western blot analysis of RNF213 protein expression in HaCaT cells under the corresponding treatment conditions. All data are presented as mean ± standard deviation (SD) (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001. Exos, exosomes; Hexos, hypoxia-preconditioned exosomes

Next, cells were transfected with miR-486-5p mimic or inhibitor to achieve its overexpression or inhibition, respectively (Fig. 6J, K). qPCR analysis showed that RNF213 mRNA levels were significantly reduced in the miR-486-5p overexpression group (Fig. 6L), whereas they were markedly increased in the inhibition group. Western blotting further confirmed that changes in RNF213 protein abundance were consistent with the mRNA trends (Fig. 6M). Collectively, these results indicate that miR-486-5p negatively regulates the expression of RNF213.

A-Hexos alleviate radiation-induced cellular Injury through miR-486-5p

To investigate the essential role of miR-486-5p in A-Exos-mediated protection against radiation-induced cellular injury, we first downregulated endogenous miR-486-5p expression in ADSCs by transfecting a miR-486-5p inhibitor, and subsequently collected the corresponding exosomes for intervention in the cellular radiation injury model. Compared with untreated exosomes, exosomes derived from the miR-486-5p inhibitor-treated ADSCs exhibited markedly attenuated protective effects on irradiated cells, as evidenced by significantly reduced cell viability (Fig. 7A), decreased EdU positivity (Fig. 7B, C), and diminished colony-forming capacity (Fig. 7D, E). Furthermore, scratch-wound healing assays demonstrated that cells treated with miR-486-5p-downregulated exosomes displayed substantially impaired migration and repair ability (Fig. 7F, G), indicating that miR-486-5p is a key functional component required for A-Hexos to enhance cellular repair and confer protection against radiation-induced damage.

Fig. 7.

A-Hexos alleviate radiation-induced cellular injury through miR-486-5p. (A) Viability of irradiated HaCaT cells treated with A-Hexos or exosomes derived from ADSCs transfected with an miR-486-5p inhibitor, assessed using CCK-8 assay. (B, C) Assessment of cell proliferative capacity following treatment with different exosomes using EdU staining. (D, E) Analysis of long-term proliferative potential by colony formation assays after intervention with distinct exosome preparations. (F, G) Comparison of cell migration and wound-healing capacity using scratch assays in response to different exosome treatments. Data are presented as mean ± SD (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001

To validate the direct role of miR-486-5p within recipient cells, we further overexpressed miR-486-5p in HaCaT cells and assessed their functional response under the same irradiation conditions. MiR-486-5p overexpression markedly counteracted radiation-induced damage in HaCaT cells, not only restoring cell viability but also recovering proliferative capacity, thereby further confirming the central regulatory role of miR-486-5p in the cellular injury repair process (Additional file, Figure S2).

RNF213 silencing ameliorates radiation-induced injury by inhibiting AKT ubiquitination and degradation

To clarify the role of targeted inhibition of RNF213 in radiation-induced cellular injury, we transfected cells with siRNAs specifically targeting RNF213. All three designed siRNAs effectively suppressed RNF213 mRNA and protein expression, among which si#3 exhibited the strongest inhibitory efficiency and was therefore selected for subsequent functional studies (Fig. 8A, B). Functional assays demonstrated that, compared with the control group, RNF213-silenced cells displayed significantly enhanced post-irradiation cell survival (Fig. 8C), increased EdU-positive proliferative rates (Fig. 8D), and markedly higher number of colonies (Fig. 8E). In addition, scratch-wound healing assays revealed that downregulation of RNF213 substantially improved cellular migration and wound-closure capacity (Fig. 8F), indicating that RNF213 interference effectively mitigates radiation-induced cellular injury and promotes the repair process.

Fig. 8.

RNF213 silencing alleviates radiation-induced cellular injury and promotes functional recovery. (A, B) qPCR (A) and western blotting (B) analyses of RNF213 expression in HaCaT cells transfected with three siRNAs targeting RNF213 (si#1–si#3) or negative control siRNA (siNC). si#3 was selected for subsequent experiments owing to its highest silencing efficiency. (C) CCK-8 assay assessing the viability of irradiated HaCaT cells transfected with siNC or siRNF213. (D) Representative EdU-stained images showing proliferative activity in irradiated HaCaT cells following transfection with siNC or siRNF213. (E) Colony formation assay assessing the long-term proliferative capacity of irradiated HaCaT cells transfected with siNC or siRNF213. (F) Representative images of wound healing assay showing migration ability of irradiated HaCaT cells transfected with siNC or siRNF213 at 12 and 24 h. Scale bar = 200 μm.Scale bar = 200 μm.(G–I) Quantitative analysis of EdU-positive cells (G), colony numbers (H), and wound-closure percentages (I) corresponding to panels (D–F). All data are presented as mean ± SD (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001

RNF213, a protein with E3 ubiquitin ligase activity, typically exerts its effects by mediating the ubiquitination and degradation of its downstream targets. Our proteomic analysis revealed multiple differentially upregulated proteins in mouse skin tissues following Gel@A-Hexos treatment (Fig. 6C). Among them, AKT1, well known for its critical role in promoting cell proliferation, migration, and injury repair, drew particular attention. Importantly, AKT1 was identified as an upregulated protein following Gel@A-Hexos treatment, whereas RNF213 was simultaneously downregulated, as revealed by proteomic analysis. This inverse expression pattern implied a potential regulatory relationship between RNF213 and AKT1. We hypothesized that the increased AKT protein expression may be associated with RNF213 suppression. Cellular assays showed that RNF213 knockdown markedly elevated AKT protein levels, whereas RNF213 overexpression reduced AKT abundance (Fig. 9A, B). To evaluate the potential interaction between RNF213 and AKT1, we performed molecular docking analysis. The docking model showed that RNF213 and AKT1 form multiple predicted contact residue pairs at the interface through hydrogen bonds (Additional files, Figure S3 and Table S1), indicating a potential interaction between RNF213 and AKT1. Furthermore, co-immunoprecipitation (Fig. 9C) and immunofluorescence colocalization analyses (Fig. 9D) demonstrated a physical interaction between RNF213 and AKT. Cycloheximide-based protein stability assays further confirmed that, under inhibited protein synthesis, loss of RNF213 significantly prolonged the half-life of AKT, indicating a substantially reduced degradation rate of AKT when RNF213 is suppressed (Fig. 9E). Additionally, we observed a positive correlation between RNF213 expression levels and the degree of AKT ubiquitination (Fig. 9F, G). These results suggest that RNF213 likely functions as an E3 ligase that promotes AKT ubiquitination and accelerates its proteasomal degradation.

Fig. 9.

RNF213 interacts with AKT and facilitates its ubiquitination. (A) qPCR analysis of RNF213 mRNA levels in HaCaT cells following RNF213 knockdown (siRNF213) or overexpression (RNF213 plasmid), compared with the corresponding control groups. (B) Western blot analysis of RNF213 and AKT protein levels in HaCaT cells after RNF213 interference or overexpression. (C) Immunoprecipitation assay to examine the interaction between RNF213 and AKT. (D) Confocal microscopy showing the colocalization of AKT (red) and RNF213 (green) in HaCaT cells. Scale bar = 100 μm. (E) CHX chase assay assessing AKT protein stability in HaCaT cells transfected with siNC or siRNF213. Cells were treated with cycloheximide (CHX, 50 µg/mL) and harvested at the indicated time points (6 h) for western blot analysis. (F, G) HaCaT cells were transfected with siRNF213 (F) or RNF213 plasmids (G). After treatment with MG132 (20 µM) for 6 h, cells were collected and subjected to ubiquitination assays. Data are presented as mean ± SD (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001. CHX, cycloheximide

To further validate the causal role of the miR-486-5p/RNF213/AKT axis, rescue experiments were performed in irradiated cells by inducing miR-486-5p overexpression with or without RNF213 overexpression (Additional file, Figure S4). The western blot analysis showed that miR-486-5p overexpression markedly reduced RNF213 expression while restoring AKT protein level following irradiation. However, co-introduction of RNF213 significantly attenuated the miR-486-5p-induced upregulation of AKT expression. Consistently, the functional assays demonstrated that miR-486-5p overexpression significantly improved cell viability and proliferative capacity, as evidenced from results of the CCK-8, EdU incorporation, colony formation, and wound healing assays. Notably, these protective effects were partially reversed upon RNF213 overexpression. These findings further support that miR-486-5p exerts its protective effects against radiation-induced cellular injury, at least partially, by targeting RNF213 and modulating downstream AKT signaling.

Collectively, these findings indicate that RNF213 interference can block its mediated AKT ubiquitination-degradation pathway, thereby stabilizing AKT protein levels, enhancing cellular adaptation to hypoxic and radiation stress, and ultimately alleviating radiation-induced cellular injury.

Discussion

Radiotherapy is an indispensable modality in cancer treatment; however, RISI, one of its most common complications, continues to severely compromise patients’ quality of life and the smooth administration of radiotherapy. Radiation stress disrupts keratinocyte function, inflammatory homeostasis, and regenerative signaling, thereby delaying skin repair which involves the coordinated action of multiple cell types [26]. However, effective interventional strategies for RISI remain limited, highlighting the urgent need to develop novel therapeutic approaches that combine safety and biological efficacy. This study addresses the key question of “how to enhance exosome reparative efficacy in the radiation-injured microenvironment while achieving effective delivery.” We established a DNA hydrogel–based local exosome delivery system, providing a new therapeutic strategy for RISI. Our results demonstrate that this DNA hydrogel exhibits excellent exosome capture capacity and local retention performance, along with favorable biosafety. Moreover, we identified that A-Hexos alleviate RISI through the miR-486-5p/RNF213/AKT signaling axis.

ADSCs and their exosomes have shown promising therapeutic potential in various refractory skin injuries, including diabetic wounds, burns, and ulcers [27, 28]. However, significant functional variability exists among ADSC-derived exosomes from different sources or physiological states and strategies to further enhance their efficacy remain a critical challenge. Hypoxic preconditioning can markedly remodel the molecular composition of stem cell-derived exosomes, thereby enhancing their regulatory capacity over target cell proliferation, migration, and angiogenesis. Previous studies have shown that hypoxic preconditioning enhances the therapeutic efficacy of stem cell–derived exosomes in multiple disease models [29, 30]. The present study further confirms that hypoxic preconditioning similarly enhances the protective effects of A-Exos on keratinocytes under radiation stress, suggesting that hypoxia-induced functional remodeling of exosomes is a crucial strategy for improving their therapeutic potential.

Although the potential of exosomes in tissue repair is well-recognized, their clinical translation is limited by poor in vivo stability, short circulation half-life, and insufficient retention at injury sites [31]. To overcome this critical limitation, we developed a DNA hydrogel delivery system with exosome-specific capture capability. By incorporating a nucleic acid aptamer targeting the exosomal membrane protein, CD63, this system enables active capture and local immobilization of exosomes, significantly prolonging their retention at the injury site and enhancing sustained release efficiency. Compared with previously reported exosome-loaded hydrogel systems, Gel@A-Hexos exhibits several distinct advantages. Conventional hydrogels used for exosome delivery are primarily based on natural or synthetic polymers such as collagen, gelatin, chitosan, and hyaluronic acid [32], which typically rely on physical encapsulation or chemical crosslinking to retain exosomes within the matrix. Although these systems can ensure sustained release, their interaction with exosomes is often nonspecific, and it is difficult to precisely control loading efficiency and release kinetics. In addition, unlike conventional hydrogels that rely on physical adsorption or non-specific chemical crosslinking, the DNA hydrogel self-assembles via base-pair complementarity without the need for chemical or photo-crosslinkers, avoiding potential residual toxicity and providing excellent biocompatibility and injectability. This design renders it particularly suitable for local therapy of sensitive tissues, such as those affected by RISI. In a murine RISI model, Gel@A-Hexos markedly alleviated radiation-induced damage and promoted skin structural reconstruction, validating the effectiveness of this delivery strategy.

Accumulating evidence suggests that the repair of RISI is a multifactorial process involving the coordinated regulation of cell proliferation, oxidative stress, angiogenesis, fibrosis, and inflammation [1]. Among these mechanisms, the proliferative capacity of keratinocytes is considered a central determinant of RISI repair [33]. As the primary functional cells of the epidermis, keratinocytes form mechanical and moisture barriers through a coordinated proliferation–differentiation process, thereby playing a pivotal role in maintaining skin barrier integrity and facilitating wound repair [34]. These cellular processes are tightly regulated by complex molecular networks, with microRNAs emerging as critical modulators. miR-486-5p is a highly conserved microRNA reported to regulate cell proliferation, migration, angiogenesis, and fibrosis [35, 36]. In our research, through miRNA sequencing and functional validation, miR-486-5p is significantly enriched in A-Hexos and plays a critical role in mitigating radiation-induced keratinocyte injury. Further proteomic analysis and bioinformatic prediction identified RNF213 as a potential downstream target of miR-486-5p. RNF213, a multifunctional protein with E3 ubiquitin ligase activity, participates in inflammation regulation, hypoxic response, and angiogenesis [37, 38]. Notably, proteomic profiling further revealed that downregulation of RNF213 expression was accompanied by an increasing trend in AKT expression, indicating a potential regulatory relationship between RNF213 and AKT. Given the well-established role of AKT in promoting cell proliferation, migration, and tissue repair, AKT was a central focus of our investigation. Previous studies have demonstrated that the activation of AKT can attenuate radiation-induced injury through multiple mechanisms, including suppression of pro-apoptotic factors such as Bax [39], enhancement of anti-apoptotic protein levels such as Bcl-2 [40], and inhibition of inflammatory signaling. In this context, maintaining AKT expression level is critical for effective tissue repair following irradiation. In the present study, we identified a miR-486-5p/RNF213/AKT regulatory axis that functionally integrates into this established protective network. Specifically, our study confirmed that miR-486-5p delivered by A-Hexos directly targets RNF213, an E3 ubiquitin ligase, thereby inhibiting RNF213-mediated ubiquitination and degradation of AKT, maintaining AKT protein stability, and enhancing its signaling activity. Sustained activation of this pathway promotes cell survival, proliferation, and migration while suppressing excessive inflammatory responses, collectively facilitating the repair of RISI.

Despite these findings, certain limitations of this study should be acknowledged. First, although molecular docking analysis suggested a potential interaction between RNF213 and AKT1, the precise binding interface was not experimentally validated, warranting further investigation. Second, although we demonstrated that the DNA hydrogel enables localized and sustained exosome delivery, its current form is not suitable for topical application, which may limit its convenience and broader clinical applicability for superficial skin injuries. Thus, further optimization of the hydrogel formulation and delivery mode is required. Finally, the experimental model used in this study may not fully recapitulate the complexity of clinical RISI, where radiation injury is often accompanied by additional factors such as excisional wounds. Therefore, more clinically relevant and combinational RISI models should be explored in future studies.

In summary, from the perspectives of exosome functional enhancement, optimization of delivery strategies, and molecular mechanisms, this study proposes a novel DNA hydrogel–based local exosome delivery strategy and systematically elucidates how A-Hexos promote the repair of RISI via the miR-486-5p/RNF213/AKT signaling axis, thereby providing a theoretically grounded and potentially translatable solution for precise RISI therapy.

Conclusion

In this study, we demonstrated that A-Hexos exert significant protective effects against RISI by enhancing cellular survival, proliferation, and migration. To improve their in vivo therapeutic performance, we further developed a biocompatible DNA hydrogel–based delivery system capable of efficiently encapsulating A-Hexos, thereby enhancing their local retention and sustained release at the injury site and ultimately improving therapeutic efficacy in vivo. High-throughput proteomic and miRNA analyses identified miR-486-5p as a key functional cargo enriched in hypoxic exosomes, and mechanistic investigations revealed RNF213 as a direct target of miR-486-5p. Suppression of RNF213 attenuated radiation-induced cellular damage by stabilizing AKT through reducing its ubiquitination-mediated degradation, thereby preserving downstream pro-survival signaling.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ADSC

Adipose-derived stem cell

CHX

Cycloheximide

Circ-DNA

Circular DNA

Exos

Exosomes

FBS

Fetal bovine serum

GO

Gene Ontology

Hexos

Hypoxia-preconditioned exosomes

KEGG

Kyoto Encyclopedia of Genes and Genomes

P/S

Penicillin-streptomycin

ALT

Alanine aminotransferase

AST

Aspartate aminotransferase

CREA

Creatinine

BUN

Blood urea nitrogen

CK

Creatine kinase

T-CHO

Total cholesterol

Author contributions

QY: Formal analysis, Methodology, Data curation, Writing - Original Draft, Writing - Review & Editing. SD: Formal analysis, Methodology, investigation, Resources, Writing - Review & Editing. YL: Conceptualization, Methodology, Software, Writing - Review & Editing. HL: Formal analysis, Data curation, Writing - Review & Editing. SY: Methodology, investigation. YF: Investigation, Resources. GX: Methodology, Formal analysis, Writing - Review & Editing. TL: Validation, Formal analysis. TZ: Investigation, Resources, Formal analysis. ZJ: Investigation, Resources, Formal analysis. KP: Conceptualization, Investigation, Supervision, Writing - Review & Editing. DW: Conceptualization, Investigation, Supervision, Writing - Review & Editing. YX: Conceptualization, Investigation, Supervision, Writing - Review & Editing. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (82273574, 82273433, and 82203157), Science and Technology Department of Sichuan Province (2024ZYD0027), Health Commission of Chengdu (2025498 and WXLHCXJJ 25 − 19), Chengdu Medical College Excellent-talent Program (2024bjGzn01 and 2024yxGzn01), Irradiation Preservation and Effect Key Laboratory of Sichuan Province (2026FZBCK01), Chengdu Medical College (CYYZZ24-02, 2022LHTD-01, and 2022LHJYZD-01), Chengdu Medical College (CYYZZ24-02, 2022LHTD-01, and CYZYB25-02).

Data availability

The data will be made available on reasonable request.

Declarations

Ethics approval and consent to participate

All animal experiments were approved by The Animal Policy and Welfare Committee of Chengdu Medical College (Approval No: CMC-IACUC-2021024).

Competing interests

The authors declare no competing interests.