Intravitreal Injection of Hydrogen Peroxide Induces Acute Retinal Degeneration, Apoptosis, and Oxidative Stress in Mice

Purpose Oxidative stress is a common pathological condition for multiple retinal diseases. Hydrogen peroxide (H2O2) has been applied as an oxidative stress inducer for the in vitro studies. Here, we report the in vivo effect of H2O2 exposure to the mouse retina and its underlying mechanism. Methods The H2O2 or saline solution was intravitreally injected into the eyes of female C57BL/6J mice for two consecutive days. The retinal structure was evaluated by in vivo imaging using spectral domain optical coherence tomography (OCT) and validated by histological assessment as well as retinal marker expression. In addition, retinal stress, cell apoptosis, and antioxidant enzyme expression were also determined. Results Retinal and outer nuclear layer thickness thinning was observed at days 7 and 14 by OCT imaging with the treatment of 10 μg H2O2, which was confirmed by the histopathological analysis. The expressions of photoreceptor (Rho, Rora, Rorb, and Rcvrn), bipolar cell (Chat and Calb2), and retinal pigment epithelial (Rpe65) markers were reduced in the H2O2-treated group, whereas the expression of retinal ganglion cell marker (Tubb3) was increased. TUNEL-positive cells were obviously found in the outer nuclear layer and inner nuclear layer of H2O2-treated mice but sparely found in the ganglion cell layer. Coherently, apoptotic gene expressions (Casp3, Casp9, Bax, and Parp8) were significantly increased in the retina with increasing dosages of H2O2, while Bcl2 expression was mildly decreased. In addition, the expressions of Gfap and antioxidant enzyme genes (Txn2, Sod2, and Gpx4) were significantly upregulated in the retina after the H2O2 treatment, compared to the vehicle control group. Conclusions This study revealed that intravitreal injection of H2O2 induces acute retinal damage by increasing oxidative stress and cell apoptosis in the retina. This acute retinal degeneration mouse model could provide a platform for drug screening against oxidative stress and retinal diseases.


Introduction
Retinal diseases, including age-related macular degeneration (AMD), glaucoma, diabetic retinopathy (DR), and retinitis pigmentosa (RP), are the leading cause of irreversible blindness and visual impairment in most developed countries [1,2], affecting more than 300 million people worldwide. Although intraocular pressure lowering can slow down the progression of glaucoma and photodynamic therapy and antivascular endothelial growth factor treatments are effective in neovascular AMD and DR, the retinal diseases still cannot be cured. Elucidating the disease mechanisms can facilitate the development of new treatments against the retinal diseases [3].
The common pathology in the retinal diseases is the retinal degeneration mediated by cell apoptosis [4,5]. Antiapoptotic treatments have been proven to prevent the retinal cells from degeneration [6,7]. Reactive oxygen species (ROS) induces oxidative stress through lipid peroxidation, disruption of normal mitochondrial function, and DNA damage, all of which can initiate the caspase-mediated apoptosis pathway [8]. Hydrogen peroxide (H 2 O 2 ), as one of the ROS, has widely been used to induce cellular oxidative stress in different cell lines, including retinal pigment epithelial (RPE) and 611W cells [9,10]. The in vitro cell culture can mimic the oxidative disease mechanisms for initial high-throughput drug screening; yet, the in vivo model would be more suitable for the development of antioxidative treatments before clinical trials [11,12]. Intracameral injection of H 2 O 2 has been shown to cause edematous ciliary process edema and deterioration as well as corneal endothelial damage [13]. However, the in vivo effect and mechanism of H 2 O 2 on mammalian retina in experimental models have yet to be determined.
In the current study, we aimed to investigate the effect and mechanism of intravitreal injection of H 2 O 2 in mice so as to establish an in vivo platform for drug screening against oxidative stress-related retinal diseases. The retinal structure and cell integrity were evaluated by in vivo imaging, histological assessment, and gene expression analysis. In addition, cell apoptosis and oxidative stress status in the H 2 O 2 -treated retina were also determined.  MO) was diluted with saline to the final concentration of 5, 8, and 10 μg/μl. The mice were anesthetized with 1 : 1 mixture (1.5 ml/kg) of ketamine (100 mg/ml) and xylazine (20 mg/ml), and intravitreal injection with a posterior approach behind the corneoscleral limbus of the eyeball was conducted under a stereomicroscope (MZ 9.5; Leica, Germany) without damaging the lens. A prepulled glass pipette connected to a Hamilton syringe (Reno, NV) and prefilled with mineral oil (Sigma-Aldrich) was used for the H 2 O 2 injection. On the stating day (day 0) and the day after (day 1), 1 μl per eyeball of H 2 O 2 solution or equal volume of saline control was intravitreally injected with a period over 2 minutes. At postinjection days 2, 7, and 14, the mice were sacrificed for further histological and gene expression analyses.

Material and Methods
2.3. In Vivo Imaging. The combination of confocal scanning laser ophthalmoscope (cSLO) with spectral domain optical coherence tomography (OCT; RETImap animal, Roland, Germany) was used for in vivo imaging of the retina before and after the H 2 O 2 injection. A superluminescent diode with a wavelength of 830 ± 50 nm was used as the laser source, and a maximum scan speed of 25,000 A-scan/sec was applied. A micrometer-resolution, three-dimensional imaging with infrared cSLO provides a planar visualization of the retina. The digital image depth of the cSLO was 16 frames per second with software module eye-tracking activated. The fundus photographs and OCT images were simultaneously captured on the exact retinal locus in a 30°circle surrounding the optic nerve head. These images were averaged automatically by the built-in software to augment the signal-to-noise ratio. The thicknesses of the retina, defined as the distance between the inner limiting membrane and Bruch's membrane, and the outer nuclear layer (ONL) were manually measured from each averaged OCT image at 600 μm away from the edge of the optic disc. Two images were measured for each experimental mouse.

Histological
Assessment of the Retina. The H 2 O 2 -treated mice were anesthetized and sacrificed by the perfusion with 0.9% saline followed by 4.0% paraformaldehyde in 0.1 M Na 2 HPO 4 /NaH 2 PO 4 buffer (pH 7.4). The eyeballs were enucleated, post-fixed in 4.0% paraformaldehyde for 4 hours, and cryoprotected with 30% sucrose/PBS for 2 days. The eyeball slices (10 μm) were sectioned using the vibratome (Leica), stained with hematoxylin and eosin, and imaged with the pupil-optic nerve position using a light microscope (Nikon, Japan). At 600 μm from the edge of the optic nerve cup, the thicknesses of the retina, ONL, inner nuclear layer (INL), and inner plexiform layer (IPL) were measured, and the cell densities of each retinal layer were also counted [14]. Six sections were measured for each experimental mouse.

Retinal Stress and Apoptosis Analyses.
Retinal stress was evaluated by the expression of glial fibrillary acidic protein (Gfap) by the immunofluorescence analysis with rabbit anti-Gfap antibody (1 : 400; Abcam, the United Kingdom) on the vibratome-sectioned eyeball slices (10 μm). Apoptosis was evaluated using the TUNEL method coupled with fluorescein (DeadEnd™ Fluorometric TUNEL System kit, Promega, Madison, WI) and DAPI counterstain (1 : 2000; Sigma-Aldrich). Apoptotic cells were visualized under the Leica TCS SP5-II fluorescence confocal microscope and counted at 600 μm from the edge of the optic nerve cup. Six sections were counted for each experimental mouse.
2.6. Gene Expression Analysis. The retina was dissected from the eyeball in the RNAlater (Invitrogen, Carlsbad, CA) immediately after enucleation. Total RNA was extracted from the retina using the TRIzol® reagent (Invitrogen) and reverse transcribed into complementary DNA with random primers using the Superscript First-Strand Synthesis System (TaKaRa, Japan). Gene expression analysis was performed using the Power SYBR Green PCR Master Mix (TaKaRa) with specific primers (Supplementary Table 1). The Gapdh gene was used as housekeeping gene for normalization.

Statistical
Analysis. The data was represented as mean ± standard error of mean (SEM) and analyzed using Stata 14.
The data distribution was analyzed by the Kolmogorov-Smirnov test. The means were compared using the independent t-test or one-way or repeated two-way analysis of variance (ANOVA) with Bonferroni post hoc test. Statistical significance was defined as p < 0 05.

Hydrogen Peroxide Induced Oxidative Stress in the
Retina. In addition to cell apoptosis, the expression of antioxidant genes was determined. Compared to the saline control, the expression of thioredoxin-2 (Txn2) gene was significantly upregulated with increasing doses of H 2 O 2 (8 μg: 1.53 ± 0.11 folds, p < 0 01 and 10 μg: 1.78 ± 0.10 folds, p < 0 001; Figure 6(a)). In addition, the application of 8 μg and 10 μg H 2 O 2 also showed a significant elevation in superoxide dismutase 2 (Sod2) gene expression by 1.34 ± 0.09 folds (p < 0 01) and 10 μg: 1.32 ± 0.02 folds (p < 0 05), respectively, compared to the saline control ( Figure 6(b)). In contrast, the expression of glutathione peroxidase 4 (Gpx4) gene did not show statistically significant changes after the H 2 O 2 application (Figure 6(c)). These suggested that the increased expression of antioxidant genes could be responded to the H 2 O 2 -induced oxidative stress elevation in the retina.

Discussion
In the current study, our results showed that ( Retinal cell death is a key pathology of retinal degenerative diseases, including AMD, glaucoma, RP, and DR. The underlying molecular mechanisms still remain elusive. The transgenic models, such as rd1 or rd10, are useful to investigate the pathology and the disease mechanisms of retinal degeneration as they exhibit progressive retinal cell loss spontaneously [17]; yet, most of them have an early onset of degeneration, and the severity and the onset of the retinal degeneration cannot be modulated [3]. Instead, chemicalinduced models can also be applied for the study of disease mechanisms and for the screening of potential treatments. N-Methyl-D-aspartate (NMDA) has been applied to induce RGC apoptosis through the NMDA receptor [18]. In this study, we injected H 2 O 2 intravitreally to induce retinal degeneration in mice, and we found that H 2 O 2 induces degeneration in all retinal layers (Figures 1 and 2), which the photoreceptor cells are the most severely damaged, followed by the cells in the INL. Interestingly, the expression of βIII-tubulin gene (Tubb3) showed an increase in the 10 μg H 2 O 2 treatment (Figure 3), in which cell apoptosis and   oxidative stress are significantly found. This could be explained by previous reports that overexpression of βIIItubulin represents a major mechanism of drug resistance to microtubule-interacting agents, such as taxanes and Vinca alkaloids, and βIII-tubulin is conditionally expressed in A2780 cells after hypoxia [19]. Nevertheless, we demonstrated a dose-dependent retinal damage for the H 2 O 2induced model (Figures 3 and 4), indicating that the severity of retinal degeneration can be modulated and controlled. Besides, the reduction in IPL thickness and the increased   Gfap expression (Figure 2) could resemble retinal stress and ischemia [20], which could be induced by the H 2 O 2mediated oxidative stress. Previous studies reported that H 2 O 2 induces apoptosis in RGCs in vitro and inhibits the phosphorylation of p38 and extracellular signal-regulated kinases1/2 [21]. The rabbit study also showed that intracameral injection of H 2 O 2 causes morphological changes in the ciliary processes with upregulation of 3-aminotriazole, an inhibitor of catalase [13]. H 2 O 2 has been found to be toxic to both the lens and cornea in high concentrations [22]. However, the in vivo effect of H 2 O 2 exposure to the retina is still poorly understood. This study, for the first time, reported the effect of intravitreal injection of H 2 O 2 in mouse retina. Based on the in vivo imaging and histological assessment, we confirmed that H 2 O 2 induces acute retinal degeneration in mice, in which cell apoptosis begins right after the H 2 O 2 application although the retina structure is still maintained (Figures 4 and 5).
Mitochondrial ROS triggers the release of cytochrome c and other proapoptotic proteins, which can trigger caspase activation and apoptosis [23]. H 2 O 2 has been found to induce apoptosis in rat nucleus pulposus cells as well as PC12 cells through the mitochondria-mediated pathway [24,25]. Coherently, we found that H 2 O 2 exposure enhances the expression of the mitochondrial antioxidant genes (Txn2 and Sod2) in the retina ( Figure 6). Simultaneously, the mitochondrial apoptosis-related genes (Casp3, Bax, and Parp8) are upregulated in the H 2 O 2 -treated retina ( Figure 5). Herein, we postulate that H 2 O 2 induces mitochondrial oxidative stress, which in turn triggers the mitochondrial apoptotic pathway and leads to the retinal cell death and the structural disruption. Further investigations could focus on the mitochondrial oxidative stress-targeted pharmacological agents against the H 2 O 2 -induced retinal degeneration.
In summary, this study revealed that intravitreal injection of H 2 O 2 induces acute retinal degeneration in mice by increasing oxidative stress and cell apoptosis. This acute retinal degeneration mouse model could provide a drug screening platform for oxidative stress and retinal diseases.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
The authors indicated no potential conflicts of interest.