2.1 Synthesis and Characterisation of tFNAs-DJ-1-saRNA

The tFNAs-DJ-1-saRNA was synthesised at a concentration of 1 μM using three single-stranded DNA sequences (S2, S3 and S4; Genscript, Nanjing, China; Table 1), a single-stranded DNA modified to include a DJ-1-saRNA cis-sense strand (S1-saRNA sense; Table 1), and one saRNA antisense strand (Table 1). The synthetic method was similar to that of tFNAs. Equal concentrations of the S1-saRNA sense, saRNA antisense strand, S2, S3 and S4 were combined in TM buffer (50 nM MgCl₂, 10 nM Tris–HCl, pH = 8.0), mixed thoroughly and then heated to 95°C for 10 min. The mixture was cooled to 4°C for 20 min. The successful synthesis of tFNAs and tFNAs-DJ-1-saRNA was confirmed by polyacrylamide gel electrophoresis (PAGE), high-performance capillary electrophoresis (HPCE), dynamic light scattering (DLS) and transmission electron microscope (TEM).

2.2 Cell Culture and Treatment

R28 cells, a rat retinal precursor cell line with differentiation potential, are widely used in the study of retinal neurons and can mimic the physiological and pathological responses of retinal neurons under OGD/R conditions. R28 cells were obtained from ATCC and cultured under OGD/R conditions using Dulbecco's Modified Eagle Medium (DMEM) medium supplemented with 10% Fetal Bovine Serum (FBS) and 1% penicillin and streptomycin (P/S). The cells were maintained at 37°C in an atmosphere containing 95% air and 5% CO₂ for 24 h. The OGD/R protocol involved culturing the cells in the DMEM medium (95% nitrogen and 5% CO₂) for 8 h, followed by a medium change to DMEM supplemented with 10% FBS and 1% P/S complete medium, after which the cells were reoxygenated for 24 h at 37°C in an environment of 5% CO₂ and 95% air. To determine the optimal concentration of tFNAs-DJ-1-saRNA, the cells were co-cultured with tFNAs-DJ-1-saRNA at concentrations of 125, 250 and 375 nM along with 250 nM tFNA and saRNA. The optimal concentration of tFNAs-DJ-1-saRNA was identified as 250 nM for subsequent studies.

2.3 Cellular Uptake

To evaluate the uptake of tFNAs and tFNAs-DJ-1-saRNA by the R28 cells, single-stranded DNA was fluorescently labelled with Cy5. R28 cells were incubated with Cy5-tFNAs-DJ-1-saRNA, Cy5-tFNAs and Cy5-DJ-1-saRNA for 8 h, respectively. Subsequently, the samples were collected and the uptake efficiency was assessed by measuring the fluorescence intensity of Cy5 within the cells using flow cytometry (Attune NxT, Thermo Fisher Scientific, USA).

2.4 Cell Viability

Cell viability was assessed using the Cell Counting Kit-8 (CCK-8; Yeasen Biotechnology, China), which facilitated the determination of the optimal conditions for modelling and the ideal concentration of tFNAs-DJ-1-saRNA. The cells were treated according to standard cell culture and treatment protocols. Following the instructions provided by CCK-8, 10 μL of the CCK-8 reagent was added to each well. The relative cell viability for each group was calculated by absorbance.

2.5 Cellular Immunofluorescence Analysis

Cells were plated on dishes and incubated with Cy5-tFNAs-DJ-1-saRNA, Cy5-tFNAs or Cy5-DJ-1-saRNA for 16 h. Cells were washed with phosphate-buffered saline (PBS) and fixed with paraformaldehyde, followed by permeabilisation using Triton X-100 and incubation with fluorescein isothiocyanate (FITC)-phalloidine and 4′,6-diamidino-2-phenylindole for cytoskeleton staining and nucleus staining. Finally, the immunofluorescence of the samples was observed using laser scanning confocal microscopy to assess the cellular uptake.

2.6 Western Blotting

The complete retinal tissue was carefully separated from the eyeball under a stereoscopic microscope and was subsequently placed into a 1.5-mL centrifuge tube containing 200 μL of protein extraction buffer (RIPA protein lysis buffer + PMSF protease inhibitor, 1:100). The retinal tissue was homogenised using an ultrasonic fragmentation device until no tissue fragments were visible in the test tube. The centrifuge tubes containing fully lysed retinal tissue proteins were subjected to high-speed centrifugation at 12000 rpm for 10 min at 4°C. After centrifugation, the supernatant was transferred to a new tube for protein quantification. The extracted proteins were quantified and their concentrations were equilibrated using a BCA Protein Quantification Kit (KeyGEN, Nanjing, China). The supernatant containing 20 μg of protein was heated to 100°C and maintained for 10 min, followed by separation using a 10% SDS-PAGE gel. The separated proteins were transferred onto a polyvinylidene difluoride (PVDF) membrane (Millipore, Merck, Germany). Subsequently, the PVDF membrane was blocked with a 5% bovine serum albumin (BSA) solution for 1 h at room temperature. The diluted primary antibodies were incubated overnight at 4°C, including PARK7/DJ-1 (1:1000, ab18257, Abcam), GPX4 (1:5000, ET1706-45, huabio), xCT (1:1000, HA721868, huabio) and RBPMS (1:1000, abs152188, absin). β-actin (1:1000, 4970S, CST) served as an internal control. The membrane was incubated with antirabbit IgG and horseradish peroxidase-conjugated antibody (1:5000, ab6721, Abcam) at ambient temperature for 60 min and visualised using an enhanced chemiluminescence system (ProteinSimple, USA).

2.7 Real-Time Fluorescence Quantitative Polymerase Chain Reaction Analysis

The total RNA was isolated and purified using a FastPure Cell/Tissue total RNA isolation kit (Vazyme Biotechnology, China). Reverse transcription was performed using HiScript III reverse transcriptase (Vazyme Biotechnology, China). All target mRNAs were amplified via quantitative real-time PCR (qRT-PCR) using SYBR Green I PCR Master Mix. All BLAST search designs used GAPDH amplification as a control. The corresponding primers are presented in Table 2.

2.8 Lipid ROS

Lipid ROS serves as a reliable marker for detecting ferroptosis in cells. The levels of lipid ROS were quantified using a BODIPY 581/591 C11 fluorescent probe (Thermo Fisher Scientific, USA). Briefly, the cells were incubated with 10 μM BODIPY 581/591 C11 for 30 min at 37°C, followed by three washes with PBS. The level of lipid ROS was assessed by the intensity of red-green fluorescence observed using fluorescence microscopy, where red fluorescence indicated unoxidised cells and green fluorescence indicated oxidised cells.

2.9 Fe²⁺ Detection

The intracellular Fe²⁺ levels were measured using a FerroOrange fluorescent probe (Cell Signalling Technology). The cells were incubated with a working solution of FerroOrange (1 μM) for 30 min at 37°C. Without washing, the intensity of red fluorescence was directly observed using a fluorescence microscope to determine the intracellular Fe²⁺ levels.

2.10 Transmission Electron Microscopy

The ultrastructure of the R28 cells in each experimental group was examined using TEM. The cells were collected following standard protocols, prefixed in 3% glutaraldehyde, postfixed in 1% osmium tetroxide, dehydrated using a series of acetone washes and subsequently embedded. The resulting cell-fixed blocks were sectioned into thin slices measuring 70–90 nm. These sections were then placed on TEM grids and analysed using a TEM JEM-1400FLASH.

2.11 Mitochondrial Membrane Potential

The mitochondrial membrane potential (Δψm) was assessed using JC-10 mitochondrial staining (Yeasen Biotechnology, China). Briefly, the cells were incubated with JC-10 (20 μM) for 30 min at 37°C and subsequently washed three times with PBS. Flow cytometry was used for analysis, with R5 representing the population of the cells exhibiting normal mitochondrial membrane potential and R6 representing the population of the cells with decreased or lost mitochondrial membrane potential. The ratio of R5 to R6 in each region was calculated to evaluate the extent of mitochondrial membrane potential damage.

2.12 ROS-Level Detection

Flow cytometry was used to detect the ROS levels in R28 cells subjected to OGD/R. The intracellular ROS levels were assessed using a DCFH-DA assay kit (Yeasen Biotechnology, China). Following incubation with 10 mM DCFH-DA in DMEM for 30 min at 37°C, the cells were harvested, washed three times with PBS, suspended in PBS and subsequently analysed using flow cytometry.

2.13 Animal Models and Treatments

All experimental procedures were approved by the Ethics Committee of Medical Laboratory Animal Welfare at the Central Theater Command General Hospital (licence number: 2023034). Male Sprague–Dawley (SD) rats (weighing 250–300 g) were obtained from the Animal Experimental Center of the General Hospital of the Central Theater Command. The rats were housed in a controlled environment with sufficient food and water, maintained at a temperature of 22°C ± 1°C and a humidity level of 55% ± 5%. They were subjected to a 12-h light/dark cycle.

The rats were divided into four groups: control, I/R 1 day group, I/R 3 day group and I/R 7 day group, with three rats in each group (n = 3 rats/6 eyes) to assess the retinal damage at different time points following I/R. In subsequent experiments, the rats were categorised into five groups: control, I/R, I/R + saRNA, I/R + tFNAs and I/R + tFNAs-DJ-1-saRNA, with three rats in each group (n = 3 rats/6 eyes) to assess the therapeutic effects of different drugs. The therapeutic concentrations of saRNA, tFNAs and tFNAs-DJ-1-saRNA were set at 1 μM, with intravitreal injections administered one day prior to the RI/RI. The RI/RI was induced by the compression of the anterior chamber [26, 27]. The primary procedure involved deeply anaesthetising the rats with a 2% sodium pentobarbital solution (60 mg/kg). A 30G needle was then inserted into the anterior chamber, and the intraocular pressure (IOP) was progressively elevated from 60 to 110 mmHg using an anterior chamber pressor device. This elevated pressure was maintained for 60 min before gradually decreasing back to 60 mmHg.

2.14 Retinal Haematoxylin–Eosin Staining and RGCs Counts

All eyeballs were enucleated and the optic nerve was severed with a blade positioned 1 mm posterior to the bulbar wall. The eye cups were fixed, dehydrated and embedded. The eye cup specimens were sectioned into three to four consecutive 4-μm cryosections through the optic nerve head. Some sections were subsequently baked and stained with haematoxylin—eosin (HE) to facilitate the observation of the histological changes in the retina under a light microscope. RGCs were quantified by examining each section at a high power (10 × 40) across five random, nonrepetitive fields and the average number of RGCs in each field was calculated.

2.15 Retinal Immunofluorescence Staining

After incubation with the diluted primary antibody overnight at 4°C, the stained samples were treated with histochemical staining, RBPMS (1:500, ab152188, Absin) and Dapi (1:8, ab104139, Abcam).

2.16 Ultra-Widefield Swept-Source Optical Coherence Tomography/Optical Coherence Tomography Angiography Imaging

To assess the changes in retinal thickness and blood flow across the different groups, the rat retinas were examined using ultra-widefield swept-source optical coherence tomography angiography (UWF-SS OCTA) (Towardpi Medical Technology Co., China). Following the manufacturer's instructions, the OCTA reference mode was configured as the small-animal mode. After administering intraperitoneal anaesthesia and performing routine mydriasis, blood flow images covering a 24 × 20 mm² area and B-scan images depicting various retinal layers were obtained using an imaging system [28, 29].

2.17 Statistical Analysis

All data were analysed using GraphPad Prism 9.5 software and presented as mean ± standard deviation (SD). One-way analysis of variance (ANOVA) was used to compare multiple groups. p values less than 0.05 were set as statistically significant.

3.1 Synthesis and Characterisation of tFNAs-DJ-1-saRNA

Four distinct DJ-1-targeted saRNAs were designed (Table 3), and their effects on enhancing DJ-1 expression in R28 cells were verified by RT-qPCR. After transfecting the four different saRNAs using a transfection reagent for 24 h, saRNA3 was selected for the synthesis of tFNAs-DJ-1-saRNA, as confirmed by RT-qPCR (Figure S1). Previous studies demonstrated that tFNAs can readily self-assemble from four specially designed ssDNA strands (S1, S2, S3 and S4; Table 1) via straightforward annealing. In this study, we modified the DJ-1-saRNA sense at the 3′ end of S1, in accordance with prior research, to create a new S1 strand (S1-saRNA sense, Table 1). The synthesis of tFNAs-DJ-1-saRNA was conducted following the same procedure as that for tFNAs. A schematic of the synthesis of tFNAs-DJ-1-saRNA is shown in Figure 1A.

8% PAGE was employed to verify the successful synthesis of tFNAs-DJ-1-saRNA. tFNAs-DJ-1-saRNA (lane 7) exhibited the slowest migration compared to S1-saRNA sense (lane 1), saRNA antisense (lane 2), ssDNAs (lanes 3–5) and tFNAs (lane 6) (Figure 1B). HPCE revealed that the size of tFNAs-DJ-1-saRNA was 193 bp, confirming its successful synthesis (Figure 1C). The TEM images demonstrated that both tFNAs and tFNAs-DJ-1-saRNA were triangular in the 2D plane (Figure 1D,E), which is consistent with the results of previous studies [20, 26]. Additionally, particle size and zeta potential measurements obtained through DLS indicated that the particle size of tFNAs was approximately 12.71 nm, whereas that of tFNAs-DJ-1-saRNA was approximately 17.00 nm (Figure 1D,E). The zeta potential of tFNAs-DJ-1-saRNA (approximately −6.35 mV) was lower than that of tFNAs (approximately −1.98 mV) (Figure 1D,E). These results suggest that DJ-1-saRNA was successfully loaded onto tFNAs.

3.2 Cellular Uptake of tFNAs-DJ-1-saRNA and DJ-1 Expression

To investigate the potential protective effect of tFNAs-DJ-1-saRNA on RI/RI, R28 cells were used for in vitro cellular experiments. We initially assessed the intracellular uptake of tFNAs-DJ-1-saRNA using immunofluorescence and flow cytometry. Confocal laser scanning microscopy revealed that the majority of the cells incubated with Cy5-tFNAs-DJ-1-saRNA and Cy5-tFNAs exhibited red fluorescence, whereas no significant red fluorescence was detected with Cy5-saRNA (Figure 1F,H). The flow cytometry results indicated that over 60% of the cells were able to uptake Cy5-tFNAs-DJ-1-saRNA or Cy5-tFNAs after 12 h of incubation (Figure 1G,I). In contrast, the rate of uptake of Cy5-saRNA was relatively low, approximately 10%. These findings suggest that tFNAs can enhance the uptake of DJ-1-saRNA in R28 cells.

Subsequently, we employed RT-PCR and WB analyses to assess DJ-1 expression following incubation with tFNAs-DJ-1-saRNA (Figure 1J–L), revealing that both mRNA and the protein levels of DJ-1 in R28 cells were significantly elevated after treatment with 125 nM tFNAs-DJ-1-saRNA. In contrast, DJ-1-saRNA did not significantly affect DJ-1 expression. Therefore, tFNAs can be utilised as an effective delivery system for DJ-1-saRNA, facilitating the upregulation of the target gene expression and promoting the associated biological effects.

3.3 Antioxidant Effect of tFNAs-DJ-1-saRNA

To investigate the antioxidant effect of tFNAs-DJ-1-saRNA, R28 cells were subjected to various OGD/R conditions (OGD 2, 4, 6, 8, 10, 12 and 24 h; R 24 h). The CCK-8 assay indicated that cell viability gradually decreased with OGD duration. Under OGD 8 h/R 24 h conditions, the cell viability decreased to approximately 60% (Figure S2). The effects of the different concentrations of tFNAs-DJ-1-saRNA on the R28 cells were also examined, revealing that tFNAs-DJ-1-saRNA at concentrations ranging from 0 to 500 nM (0, 62.5, 125, 250 and 500 nM) did not significantly affect the cell viability (Figure S3). In subsequent experiments, the R28 cells were cultured under OGD 8 h/R 24 h conditions with the addition of tFNAs-DJ-1-saRNA to evaluate their potential to enhance cell viability. tFNAs-DJ-1-saRNA (125, 250 and 375 nM) significantly improved the viability of the R28 cells, with 250 nM tFNA-DJ-1-saRNA exhibiting the most pronounced effect. In contrast, DJ-1-saRNA alone did not affect the viability of the R28 cells (Figure S4).

These results aligned with the antioxidant effects of tFNAs and tFNAs-DJ-1-saRNA reported previously [20, 26]. Ferroptosis is characterised by oxidative damage. To determine whether ferroptosis occurs in R28 cells under OGD/R conditions, we measured the ROS levels and assessed ferroptosis following treatment with various drugs (saRNA, tFNAs and tFNAs-DJ-1-saRNA) at a concentration of 250 nM. A detailed examination of the ultrastructural changes in the cells was conducted via TEM, revealing cytoplasmic and organelle swelling, as well as mitochondrial atrophy in R28 cells subjected to OGD/R injury (Figure 2A). The alterations in mitochondria within the OGD/R group were particularly pronounced, exhibiting smaller and darker mitochondria, destruction of mitochondrial cristae and disruption of the mitochondrial outer membrane. These changes are similar to those observed in the mitochondria of the ferroptosis agonist-treated (erastin) group. All these morphological changes are indicators of ferroptosis. Treatment with tFNAs-DJ-1-saRNA significantly mitigated the pathological changes in R28 cells, indicating an improvement in mitochondrial morphology comparable to that observed in the ferroptosis inhibitor-treated (Fer-1) group (Figure 2A).

Lipid peroxidation serves as the gold standard for detecting ferroptosis via the measurement of the intensity of red-green fluorescence using fluorescence microscopy. Red fluorescence indicates unoxidised cells and green fluorescence indicates oxidised cells. We found that green fluorescence was significantly increased and red fluorescence was significantly decreased in OGD/R, saRNA and erastin groups compared with the control group. Conversely, after treatment with tFNAs, tFNAs-DJ-1-saRNA and Fer-1, we observed an increase in red fluorescence and a decrease in green fluorescence (Figure 2B). Intracellular Fe²⁺ is a typical marker of ferroptosis. Therefore, the intracellular Fe²⁺ levels were further investigated. The Fe²⁺ levels (red fluorescence) were significantly elevated in the OGD/R and saRNA groups after OGD/R injury. However, after treatment with tFNAs, tFNAs-DJ-1-saRNA and Fer-1, the Fe²⁺ levels (red fluorescence) were significantly reduced (Figure 2C). Notably, tFNAs and tFNAs-DJ-1-saRNA significantly inhibited OGD/R-induced ferroptosis.

Given the significant changes in mitochondrial morphology observed in the R28 cells following OGD/R injury, we further evaluated the alterations in mitochondrial function. Mitochondria are crucial organelles responsible for ATP production, redox signalling and cell death. The JC-10 probe was employed to assess the mitochondrial membrane potential. Flow cytometric analysis revealed that the ratio of healthy to damaged mitochondria (R5/R6) significantly decreased after OGD/R injury. However, this ratio notably improved following tFNAs-DJ-1-saRNA treatment (R5/R6). Furthermore, tFNAs-DJ-1-saRNA effectively mitigated OGD/R-induced cell stress and cytotoxicity (Figure 3A,B).

The ROS levels were assessed using flow cytometry, revealing a significant increase in R28 cells following OGD/R injury. Moreover, treatment with tFNAs-DJ-1-saRNA effectively reduced the ROS levels in the R28 cells (Figure 3C,D).

3.4 In Vivo Therapeutic Effect of tFNAs-DJ-1-saRNA

Section 3.3 presented the protective effects of tFNAs-DJ-1-saRNA on OGD/R-injured R28 cells in vitro. To further investigate the changes in the retina following RI/RI and the therapeutic effect of tFNAs-DJ-1-saRNA in vivo, we established a rat model of RI/RI by inducing high IOP through anterior chamber perfusion. We employed HE staining and optical coherence tomography (OCT) to assess the retinal thickness 1, 3 and 7 days post-RI/RI, alongside OCTA to evaluate alterations in the retinal blood vessels. HE staining revealed that after 1 d of I/R, the thickness of IPL increased and the number of RGCs decreased. At 3 and 7 d postmodelling, the thickness of the inner retina decreased, the number of RGCs continued to decline, IPL became thinner and the inner nuclear layer exhibited disorganisation (Figure 4A,D–F). The OCT findings corroborated these results, showing increased retinal thickness, oedema in the inner retina and ischaemic changes in the outer retina 1 d after I/R. Notably, atrophy of the inner retina and a progressive decrease in the total retinal thickness were observed 3 and 7 d postmodelling (Figure 4C,G). The OCTA results indicated that the large retinal vein was tortuous and dilated, with an enlarged vein diameter 1 d after I/R. The tortuosity of the retinal vein was alleviated by day 3 postmodelling. On day 7, the diameter of the venous lumen gradually returned to normal levels (Figure 4B).

Due to the dilution of the vitreous cavity itself and the presence of metabolic enzymes within the tissue, a higher concentration of intravitreal drugs is necessary to enhance drug stability and efficacy in vivo. Previous studies reported a drug concentration of 1000 nM²⁶. SaRNA, tFNAs and tFNAs-DJ-1-saRNA were administered into the vitreous cavity prior to modelling, and the effects of these drugs were evaluated 7 d postmodelling. HE staining revealed a significant reduction in the number of RGCs as well as in the thickness of IPL and the inner retina 7 d after model establishment. Treatment with tFNAs and tFNAs-DJ-1-saRNA ameliorated the loss of RGCs and mitigated atrophy of IPL and the inner retina (Figure 5A–D). OCT was employed to further assess the retinal condition of the rats in vivo, revealing a significant reduction in the total retinal thickness and observable retinal atrophy 7 d after modelling. Although treatment with saRNA alone did not yield significant changes, administration of tFNAs and tFNAs-DJ-1-saRNA effectively improved this pathological state (Figure 5E,F).

RBPMS is a critical protein in neuronal survival, synapse formation and repair, and serves as the active protein in RGCs. Therefore, RBPMS expression provides insights into the activity of RGCs [30, 31]. Immunofluorescence staining revealed that the expression of the RBPMS protein (indicated by red fluorescence) in the ganglion cell layer following RI/RI was significantly lower than that observed in the normal group. Furthermore, RBPMS protein expression in the saRNA group was comparable to that in the RI/RI group, whereas the expression levels increased following treatment with tFNAs and tFNAs-DJ-1-saRNA. These findings suggest that drug treatment enhanced the activity of RGCs and improved their function (Figure 5G,H).

To further investigate the effect of tFNAs-DJ-1-saRNA on DJ-1 protein expression in the context of RI/RI and explore its potential therapeutic mechanisms, we conducted WB experiments. A significant decrease in DJ-1 protein expression was observed after RI/RI. However, the DJ-1 protein levels were elevated following treatment with tFNA and tFNAs-DJ-1-saRNA, with the tFNAs-DJ-1-saRNA group exhibiting higher expression than the tFNA group. Additionally, we observed a significant reduction in RBPMS expression after RI/RI, which was restored following tFNA and tFNAs-DJ-1-saRNA administration, corroborating the findings of the immunofluorescence experiments. Furthermore, we found that tFNAs-DJ-1-saRNA enhanced the expression of the ferroptosis-protective proteins GPX4 and xCT in the retinal tissue, thereby mitigating ferroptosis-related damage. This suggests that tFNAs-DJ-1-saRNA may safeguard the retinal tissue by inhibiting ferroptosis, ultimately reducing ganglion cell damage during RI/RI (Figure 5I,J).