Blue Light Induces Retinal Ganglion Cell Damage by Stimulating Drp1-Dependent Mitochondrial Fission and Activating NF-κB/NOX4 Axis Authors: Xin-Hao Han, Tao Wang, Gui-Feng Sun, Xin Xie, Kun Zhang, Xiao-Yu Wang, Zhen Zhang, Xing Wang, Xin-Hui Qu, Xiao-Jian Han Published: 2026 ABSTRACT PurposeThis study aimed to investigate the mechanisms of blue light-induced neurotoxicity in retinal ganglion cells (RGCs), focusing on the roles of mitochondrial dynamics and oxidative stress.MethodsThe impact of blue light exposure was assessed in vitro and in vivo. Key molecular changes were analyzed, and the effects of pharmacological inhibition of Drp1 and/or NOX4 were evaluated on mitochondrial function and RGC apoptosis.ResultsBlue light triggered mitochondrial fission by upregulating Drp1 and downregulating MFN2. This disruption promoted the nuclear translocation and phosphorylation of p65, which subsequently enhanced NOX4 transcription and increased mitochondrial reactive oxygen species (ROS) production. Inhibiting either Drp1 or p65 suppressed NOX4 expression and ROS generation. Furthermore, combined inhibition of Drp1 and NOX4 effectively restored mitochondrial function and reduced RGC apoptosis.ConclusionsBoth Drp1 and NOX4 contribute to blue light-induced RGC damage, and our findings highlight the importance of the Drp1/mitochondrial fission/p65/NOX4 signaling axis in this process, leading to oxidative stress. Targeting this signaling axis represents a promising therapeutic strategy for preventing blue light-induced retinal injury. Light-emitting diodes (LEDs) are extensively used in various applications, such as lighting, displays, and electronics, because of their energy efficiency, longevity, and compact form factor. Nonetheless, the emission spectra of LED lighting are characterized by a pronounced blue light component (400-490 nm), which can traverse the cornea and lens to directly reach the retina. This exposure is associated with ocular conditions such as dry eye, cataracts, and age-related macular degeneration. Prolonged exposure to blue light leads to the generation of reactive oxygen species (ROS) and mitochondrial damage within retinal cells, thereby exacerbating photochemical reactions that result in cellular dysfunction or apoptosis.1,2 Despite these findings, the precise molecular mechanisms underlying blue light-induced retinal damage remain inadequately understood, posing challenges to the development of effective therapeutic strategies. Excessive exposure to blue light results in significant photochemical injury across various retinal cell types, including photoreceptors, RPE, and retinal ganglion cells (RGCs). Research indicates that RGCs are particularly vulnerable to blue light-induced damage because of their lack of myelination within the retina. Exposure to blue light on the retina influences the mitochondrial chromophore, resulting in neuronal death.3,4 Within the ocular environment, RGC axons are unmyelinated and possess mitochondria-rich axonal varicosities,5,6 which are the primary targets of blue light before it penetrates deeper retinal layers. Mitochondrial chromophores, such as cytochrome c oxidase, cytochrome P450, and flavin components, absorb a substantial portion of blue light within the 400 to 490 nm spectrum.7 Previous research indicates that light exposure modifies mitochondrial metabolic states by affecting the activities of electron transport chain complexes I-V and dehydrogenases.8 Blue light exposure leads to structural remodeling and disrupts mitochondrial membrane potential in RGC-5 cells, while also inducing cell death through the activation of mitochondrion-dependent apoptotic pathways.9 Specifically, blue light disrupts mitochondrial photoreceptor components, including flavins and cytochromes, when incident on the retina.10 These findings highlight mitochondria within RGCs as critical targets for blue light–induced phototoxicity. Despite advances in understanding, the regulatory mechanisms by which blue light influence mitochondrial dynamics and function, as well as their specific roles in retinal light-induced injury, remain insufficiently elucidated. Mitochondria are highly dynamic organelles that ubiquitously undergo processes of growth, division, and fusion across nearly all eukaryotic cell types. The dynamic balance between fusion and fission facilitates efficient mitochondrial transport, promotes mitochondrial homogenization, and thereby supports enhanced oxidative phosphorylation.11 Mitochondrial fission and fusion are primarily mediated by three dynamin-related GTPases: dynamin-related protein 1 (Drp1) is responsible for fission, whereas mitofusin 1/2 (MFN1/2) and optic atrophy 1 (OPA1) facilitate the fusion of the mitochondrial outer and inner membranes, respectively, in mammalian cells.12 Mitochondrial fragmentation is frequently observed in aging and various disease pathologies and is widely recognized as a hallmark of mitochondrial dysfunction.13 An imbalance in mitochondrial fusion-fission homeostasis results in mitochondrial damage, leading to increased production of mitochondrial reactive oxygen species (mtROS).14 Intracellular ROS primarily originate from mitochondrial and enzymatic pathways.15 As second-messenger molecules, ROS play a crucial role in modulating essential physiological processes such as inflammation, proliferation, and differentiation, thereby contributing to the maintenance of cellular homeostasis.16 However, mitochondrial dysfunction in conjunction with enzymatic sources, can lead to increased ROS production and a diminished capacity for ROS scavenging, resulting in pathologically elevated intracellular ROS levels.15,17 Excessive ROS can oxidize polyunsaturated fatty acids, leading to the irreversible opening of the mitochondrial permeability transition pore. This event triggers the release of cytochrome c, which subsequently activates caspase-3, inducing apoptosis.18,19 Elevated ROS levels are implicated in driving oxidative stress, inflammatory responses, and neovascularization in various ophthalmic pathologies, including chemical corneal burns,20 diabetic retinopathy,21 and age-related macular degeneration.22 Despite these insights, the mechanistic relationship between mitochondrial fission-fusion dynamics and ROS generation in blue light-induced retinal injury remains inadequately understood. NADPH oxidases (NOXs) and lipoxygenases are significant contributors to ROS production.23 The NOX family consists of seven members: NOX1-5, DUOX1, and DUOX2.24 Within tissues, the expression levels of NOX4 are positively correlated with its ability to mediate ROS generation.15,25 NOX4 is a predominant source of ROS in the kidney,26 whereas in the heart, ROS derived from NOX4 exacerbate mitochondrial oxidative stress and ROS-induced ROS Release (RIRR), thereby promoting mitochondrial dysfunction and cardiomyocyte apoptosis, which ultimately resulting in cardiac dysfunction.27 The genetic ablation of NOX4 markedly reduces ROS production in glioblastoma cells.28 The mammalian NF-κB family comprises five structurally related transcription factors: p50, p52, p65 (also known as RelA), c-Rel, and RelB. Among these, p65, RelB, and c-Rel possess C-terminal transcriptional activation domains that are crucial for the positive regulation of gene expression. Under basal conditions in most cells, NF-κB dimers reside in the cytoplasm as inactive complexes bound to inhibitory IκB proteins. Upon activation of canonical signaling pathways, the proteasomal degradation of IκB proteins releases NF-κB dimers, allowing their translocation to the nucleus where they exert potent transcriptional activation functions. Previous research has demonstrated that NF-κB regulates NOXs in human aortic vascular smooth muscle cells, thereby facilitating the production of NOX-derived ROS.29 Inhibition of the NF-κB signaling pathway has been shown to impede TGF-β and TNF-α-induced NOX4 expression.30 These findings suggest that NF-κB functions as a critical transcriptional regulator of NOXs. In the present study, we demonstrate that exposure to blue light leads to an upregulation of Drp1 expression and a downregulation of MFN2 expression, resulting in mitochondrial fission. This sequence of events induces the nuclear translocation of p65, which subsequently activates NOX4 transcription. The resultant increase in NOX4 expression precipitates significant ROS production, causing damage to RGC and retinal injury. Importantly, inhibition of Drp1 or p65 was found to suppress NOX4 expression and ROS production. Furthermore, the combined pharmacological inhibition of Drp1 and NOX4 was effective in restoring mitochondrial function and reducing RGC apoptosis both in vitro and in vivo. Collectively, these findings elucidate the molecular mechanisms underlying blue light–induced RGC damage and identify potential therapeutic targets. Material and Methods Animal All animal experiments performed in this study were approved by the Animal Experiment Ethics Committee of Jiangxi Provincial People's Hospital, China. The animal studies were conducted in accordance with the ARVO statement for the use of animals in ophthalmic and vision research. Male Sprague-Dawley (SpD) rats, aged eight to 12 weeks, were anesthetized using 1.0% pentobarbital sodium (5 mg/100g body weight, intraperitoneally), with efforts made to minimize animal suffering. Cell Culture Rat precursor cells (R28) were provided by Guo-Tong Xu (Tongji Eye Institute, Tongji University School of Medicine). R28 cells were cultured in low glucose DMEM supplemented with 10% fetal bovine serum and maintained at 37°C with 5% CO2 in a humidified environment. RNA Interference R28 cells were transfected with siRNA targeting DRP1 (RiboBio) or a scrambled siRNA (RiboBio) as a control. The targeting sequences were as follows: Drp1-si-1: 5′-GGAGCCAACUGGACAUUAAdTdT-3′ (sense) and 5′-UUAAUGUCCAGUUGGCUCCdTdT-3′ (antisense), Drp1-si-2: 5′- GUGGGCUAAUGAACAAUAAdTdT-3′ (sense) and 5′- UUAUUGUUCAUUAGCCCACdTdT-3′ (antisense), Drp1-si-3: 5′- GCUGAUCCCGGUCAUCAAUdTdT-3′ (sense) and 5′- AUUGAUGACCGGGAUCAGCdTdT-3′ (antisense). SiRNAs were transfected into R28 cells with ribo-FECT CP Transfection Kit (#C10511-05, RiboBio) following the manufacturer's instructions. Light Exposure System Cells were seeded in a 35-mm dish and subjected to blue light (450 nm, 1000 lx) or red light (630 nm, 1000 lx) within a temperature-controlled incubator set at 37°C with 5% CO₂. Cells maintained in complete darkness served as the control group. Prior to the initiation of all light exposure experiments, the culture medium was strictly replaced with phenol red-free medium to eliminate any potential phototoxic effects or ROS generation caused by the photoactivation of phenol red. Adult SpD rats were housed in light-tight cabinets and exposed to blue light (450 nm, 1500 lx) or red light (630 nm, 1500 lx) for specified durations, while control animals were kept in dark conditions. Western Blot Analysis Total protein was separated using SDS-PAGE and subsequently transferred onto a polyvinylidene difluoride membrane. The membranes were blocked with Tris-buffered saline solution with Tween-20 containing 5% skim milk for two hours and then incubated overnight at 4°C with the following primary antibodies: OPA1 (no. 80471; Cell Signaling Technology, Danvers, MA, USA), Drp1 (no. 8570; Cell Signaling Technology), Phospho-Drp1 (no. 4494; Cell Signaling Technology), MFN2 (no. 9482; Cell Signaling Technology), p65 (no. 8242; Cell Signaling Technology), Phospho-p65 (no. 3033; Cell Signaling Technology), VDAC (no. 4661; Cell Signaling Technology), NOX4 (Ab133303; Abcam, Cambridge, MA, USA), GAPDH (no. 10494-1-AP; Proteintech, Rosemont, IL, USA), β-actin (no. AC038; ABclonal Technology, Woburn, MA, USA). After three washes with Tris-buffered saline solution with Tween-20, the membranes were incubated with secondary antibodies (Goat Anti-Rabbit IgG, no. SA00001-1; Proteintech; Goat Anti-Mouse IgG, no. SA00001-1; Proteintech) for two hours. The bands were visualized using a chemiluminescence detection kit (no. 20350ES; Yeasen, Shanghai, China), and images were captured using a BioRad ChemiDoc MP digital gel image analysis system (Bio-Rad Laboratories, Ann Arbor, MI, USA). Mitochondrial Imaging and Analysis R28 cells were prepared in confocal Petri dishes. The mitochondria of live cells were stained with a 100 nM MitoTracker Red CMXRos (no. M7512; Thermo Fisher Scientific, Waltham, MA, USA) working solution at 37°C for 30 minutes. Mitochondrial imaging was achieved by detection of fluorescence using confocal microscope (FV3000; Olympus, Tokyo, Japan), and mitochondrial length was measured using ImageJ software. Quantitative Real-Time PCR Total RNA was extracted from R28 cells using an RNA extraction kit (no. DP451; Tiangen Biotech, Beijing, China). The RNA was reverse transcribed into cDNA using the M5 Sprint qPCR RT kit with gDNA remover (no. MF949-01; Mei5 Biotechnology Co., Ltd., Beijing, China). Quantitative real-time PCR was performed using the M5 Hiper Realtime Super Mix with High Rox (no. MF013-505; Mei5 Biotechnology Co.). The primer sequences used are listed as follows: NOX4 forward: 5′-ACACCTCTGTCTGCTTGTTTGG-3′, reverse: 5′-TTCTCCTGCTAGGGACCTTCTG-3′. Drp 1 forward: 5′-CAGGAACTGTTACGGTTCCCTAA, reverse: 5′-CCTGAATTAACTTGTCTCGCGA. β-actin forward: 5′-TCGTGCGTGACATTAAAGAGAA-3′, reverse: 5′-AGGAAGGAAGGCTGGAAGAGAG-3′. Immunofluorescence Cells were fixed with 4% paraformaldehyde for 15 minutes at room temperature, followed by permeabilization using 0.1% Triton X-100 (no. T8200; Solarbio Life Science, Beijing, China) for 10 minutes. Subsequently, the cells were blocked with 5% bovine serum albumin (BSA, #36105ES, YEASEN) for 30 minutes. Primary antibodies, diluted in the blocking buffer, were incubated with the samples overnight at 4°C. Following three washes of five minutes each with PBS containing 0.1% Tween-20 (PBST), the samples were incubated with Alexa Fluor-conjugated secondary antibodies (no. 8889; Cell Signaling Technology) for one hour at room temperature, protected from light. Nuclear counterstaining was performed using 4’,6-diamidino-2-phenylindole (DAPI, no. C0065; Solarbio Life Science) for five minutes. Apoptosis Assay Cells were harvested via gentle trypsinization (excluding EDTA), washed twice with cold PBS, and resuspended in 1 × Annexin V binding buffer at a concentration of 1 × 10⁶ cells/mL. The cells were then stained with 5 µL of Annexin V-FITC and 5 µL of propidium iodide (PI) using the Annexin V-FITC/PI Apoptosis Kit (no. E-CK-A211; Elabscience, Wuhan, China) for 15 minutes at room temperature in the dark. Unstained and single-stained controls were processed concurrently for compensation. The samples were analyzed within one hour using a BD FACS Canto II flow cytometer. Mitochondrial Membrane Potential Assay Cells were incubated with 50 nM tetramethylrhodamine ethyl ester (TMRE; no. T669; Invitrogen, Carlsbad, CA, USA) at 37°C for 30 minutes, followed by washing with PBS. Images were acquired using an inverted fluorescence microscope (Eclipse Ti2-U; Nikon Inc., Melville, NY, USA), and fluorescence intensity was quantified via flow cytometry (FACS Canto II; Becton Dickinson, Franklin Lakes, NJ, USA). Measurement of Intracellular ROS Level Cells were washed with PBS and incubated with dihydroethidium (DHE; no. 309800; Sigma-Aldrich Corp., St. Louis, MO, USA) at a final concentration of 10 µM in the dark for 30 minutes at room temperature. Images were acquired using an inverted fluorescence microscope (Eclipse Ti2-U; Nikon Inc.). Measurement of Mitochondrial ROS Level Cells were incubated with 5 µM mitoSOX (no. M36008; Invitrogen) at 37°C for 30 minutes, followed by washing with PBS. Images were acquired using an inverted fluorescence microscope (Eclipse Ti2-U; Nikon Inc.) and fluorescence intensity was quantified using flow cytometry (FACS Canto II; Becton Dickinson). Hematoxylin-Eosin (H&E) Staining and Immunohistochemistry (IHC) Staining Retinal tissues from SpD rats underwent graded ethanol dehydration, xylene clearing, and paraffin embedding, followed by sagittal sectioning a thickness of 5 µm. For H&E staining, the deparaffinized sections were processed using an H&E staining kit (no. C0105S; Beyotime Institute of Biotechnology, Jiangsu, China) in accordance with established protocols. For IHC staining, the paraffin sections were deparaffinized, rehydrated, and subjected to antigen retrieval utilizing a citrate buffer. Endogenous peroxidase activity was inhibited with 3% H₂O₂. The sections were incubated with a primary antibody overnight at 4°C, followed by incubation with a HRP-conjugated secondary antibody. DAB was used for chromogen detection, and nuclei were counterstained with hematoxylin. TUNEL Staining Apoptotic cells were detected in 4% paraformaldehyde-fixed, paraffin-embedded retinal sections using a one-step TUNEL apoptosis detection kit (no. E-CK-A320; Elabscience) according to the manufacturer's protocol. Fluorescent signals were visualized using a Nikon fluorescence microscope (Eclipse Ti2-U). Chromatin Immunoprecipitation (ChIP) and PCR Assay ChIP was performed using the MAGnify Chromatin Immunoprecipitation System (no. 492024; Thermo Fisher Scientific) in accordance with the manufacturer's instructions. Briefly, cells were cross-linked with 1% formaldehyde, lysed, and subjected to sonication to shear the DNA. Chromatin was immunoprecipitated using either control IgG or an anti-p65 antibody. Potential p65-binding sites within the proximal promoter region of the NOX4 gene were amplified via PCR with the following primers: Forward: 5’-CAGGCTCTAGGTTTGGAGGAGC-3’ and Reverse: 5’-CTGCATCCTCAAACTGAACCCTTA-3’. The PCR products were visualized through agarose gel electrophoresis, and images were captured using a BioRad ChemiDoc MP digital gel image analysis system. Transmission Electron Microscopy Rat retinas were fixed in 2.5% glutaraldehyde at 4°C for 24 hours. The tissues were post-fixed in 1% osmium tetroxide for one hour, dehydrated through a graded acetone series (range, 30% to 100%), and embedded in Epox 812 resin. Semi-thin sections (1 µm) were stained with methylene blue for orientation purposes. Ultrathin sections (70 nm) were stained with uranyl acetate for 15 minutes and lead citrate for 5 minutes, followed by imaging with a JEM-1400FLASH TEM (JEOL Ltd., Tokyo, Japan) operating at 80 kV. Statistical Analysis All analyses were performed using GraphPad Prism (v8.0). Data derived from three or more independent experiments are presented as the mean ± SD. Comparisons between groups were performed using unpaired, two-tailed Student's t-tests or one-way ANOVA. Statistical significance was defined as P < 0.05. Result Blue Light Exposure Induces Mitochondrial Fission in RGCs To validate the cellular phenotype of R28 cells used in this study, low-passage cells (passages ≤25) were assessed for retinal lineage-specific markers. Immunofluorescence confirmed robust expression of RGC markers Brn3a and RBPMS (Supplementary Figs. S1A, S1B). Western blot analysis further verified RBPMS protein expression in R28 lysates, whereas photoreceptor-specific markers Recoverin and Rhodopsin were undetectable; in contrast, all three proteins were readily observed in adult rat retina lysates (Supplementary Fig. S1C). These results demonstrate that the R28 cells used herein exhibit a molecular signature characteristic of RGC lineage commitment without detectable photoreceptor contamination, supporting their appropriateness as a reproducible in vitro model for mechanistic studies of RGC-relevant pathways. To explore the impact of blue light on RGC mitochondria, we developed specialized exposure setups for both cellular (Figs. 1A, 1B) and animal (Figs. 1C, 1D) studies, using high-performance, stable LED panels (Fig. 1E). These panels were connected to a DC-regulated power supply, allowing for precise control of light intensity through voltage adjustments. To ensure stable operation and prevent overheating, the LED backplates were equipped with recirculating water-cooling systems (Figs. 1F, 1G). The spatial uniformity of illumination was validated by measuring irradiance at multiple locations across the panels (Figs. 1H–J) using a calibrated digital spectral radiometer (TASI-8720; TASI Electronics, Suzhou, China), with results confirming high uniformity (Figs. 1K, 1L). Light exposure setup. (A) Setup for cellular exposure to blue light. (B) Setup for cellular exposure to red light. (C) Animal cabinet for exposure to blue light. (D) Animal cabinet for exposure to red light. (E) A representative LED light panel. (F) Recirculating water-cooling for the cellular exposure setup. (G) Recirculating water-cooling for the animal exposure cabinet. (H) Layout of the illuminance sampled points for the cellular exposure setup. (I) Layout of the illuminance sampled points for the animal blue-light exposure cabinet. (J) Layout of the illuminance sampled points for the animal red-light exposure cabinet. (K) Blue light irradiance at sampled points. (L) Red light irradiance at sampled points. Subsequently, we developed a model of blue light-induced injury and analyzed proteins associated with mitochondrial dynamics following exposure. Our findings revealed a significant upregulation in both mRNA and protein expression of Drp1, a key mitochondrial fission protein (Figs. 2A, 2B, 2D). Recognizing the pivotal role of phosphorylation at serine 616 (Ser616) in modulating Drp1’s GTPase activity and mitochondrial fission,31 we further investigated the phosphorylation status of Drp1 at this site following blue light exposure. Our findings indicated a substantial increase in Drp1 phosphorylation at Ser616 due to blue light exposure (Figs. 2A, 2C). Additionally, the expression of the mitochondrial outer membrane fusion protein MFN2 was notably downregulated post-exposure (Figs. 2E, 2F), whereas the expression of the inner membrane fusion protein OPA1 remained unchanged (Figs. 2G, 2H). As a control, R28 cells were subjected to the same experimental conditions with red light exposure, which demonstrated no alterations in mitochondrial morphology or in the expression of proteins related to mitochondrial dynamics (Supplementary Figs. S2A–G). To investigate the morphological changes in mitochondria induced by blue light exposure, R28 cells were incubated with MitoTracker Red for 30 minutes to label the mitochondria. Fluorescence microscopy analysis revealed a punctate and fragmented mitochondrial morphology with a significantly reduced average length following blue light exposure (Figs. 2I, 2J). In contrast, red light exposure did not result in such alterations (Supplementary Figs. S2H, S2I). Blue light induces mitochondrial hyper-fission and alters the expression of mitochondrial dynamics-related proteins in R28 cells. (A) Western blot analysis of p-Drp1 and total Drp1 expression in R28 cells after blue light exposure for the indicated times. BL, blue light. (B, C) Quantification of Western blot results shown in (A). (D) QRT-PCR analysis of Drp1 mRNA expression levels in R28 cells following blue light exposure at the indicated time points. Data are presented as mean ± SD; n = 4. (E) Western blot analysis of MFN2 expression in R28 cells after blue light exposure for the indicated times. (F) Quantification of Western blot results were shown in (E). (G) Western blot analysis of OPA1 expression in R28 cells after blue light exposure for the indicated times. (H) Quantification of Western blot results shown in (G). (I) After blue light exposure, R28 cells mitochondria were labeled with Mito Tracker Red and mitochondrial morphology images were captured by confocal microscopy. Scale bars: 20 µm. The below panels are the amplification of white dotted boxes. Scale bars: 5 µm. (J) The relative length of mitochondria in each group was analyzed. 300 mitochondria/group from three different random fields were measured. Data were presented as mean ± SD of three independent experiments. ns P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001. To further validate these effects in vivo, a rat retinal injury model induced by blue light was developed, followed by H&E and IHC staining of retinal tissues. The findings demonstrated a significant decrease in the number of RGCs within the ganglion cell layer (GCL) after exposure to blue light (Figs. 3A, 3B). In contrast, no such changes were observed following exposure to red light (Supplementary Figs. S3A, S3B). Subsequent IHC analysis revealed an upregulation of Drp1 expression and a downregulation of MFN2 expression in the GCL post-blue light exposure. OPA1 immunoreactivity in the GCL appeared relatively preserved compared to Drp1 and MFN2 (Figs. 3C–E), although subtle changes cannot be excluded because of the lack of quantitative measurement and possible cell loss. Our in vitro and in vivo data demonstrate that blue light promotes mitochondrial fission in RGCs-primarily within the GCL-by upregulating Drp1 and downregulating MFN2, suggesting a key role for Drp1 in this process, although effects on other retinal layers cannot be entirely ruled out. Blue light exposure induced rat RGC death and altered the expression of mitochondrial dynamics-related proteins in vivo. (A) H&E staining of retinal sections from SpD rats exposed to blue light for indicated times. Scale bars: 50 µm. The below panels are the amplification of black dotted boxes. Black arrows indicate the RGCs. Scale bars: 20 µm. n = 6. (B) Quantitative analysis of the number of RGCs across groups. (C–E) IHC analysis of Drp1, MFN2, and OPA1 expression in RGCs from retinae dissected after 24-hour blue light exposure. Scale bars: 50 µm. The below panels are the amplification of black dotted boxes. Scale bars: 20 µm. n = 5. Data are presented as mean ± SD. ***P < 0.001. Inhibition of Drp1 Reduces Mitochondrial Fission and NOX4 Levels in RGCs Under Blue Light Exposure To further explore the influence of Drp1 on mitochondrial fission triggered by blue light exposure, R28 cells were pre-treated with either Drp1 siRNA (Supplementary Fig. S4) or the Drp1 inhibitor Mdivi-1 prior to exposure. The results indicated that inhibition of Drp1 significantly mitigated mitochondrial fission induced by blue light (Figs. 4A, 4B). Given that Drp1-Ser616 phosphorylation and Drp1 translocation to the mitochondria can result in substantial ROS accumulation,32 and considering that NOX4 is a primary source of ROS,26,27 we investigated whether Drp1 might regulate NOX4 expression during blue light-induced mitochondrial fission. Our findings revealed a significant upregulation of NOX4 expression following blue light exposure (Figs. 4C, 4D), whereas exposure to red light did not affect NOX4 expression (Supplementary Fig. S5). NOX4 is predominantly localized to the endoplasmic reticulum and mitochondria, where it interacts with mitochondrial complex I.33 This interaction compromises mitochondrial function and facilitates the production of ROS. We quantified NOX4 expression in both cytoplasmic and mitochondrial fractions. Mitochondrial and cytoplasmic fractions were isolated from R28 cells following a 12-hour exposure to blue light. Subsequent western blot analysis demonstrated a significant elevation of NOX4 expression in both compartments (Figs. 4E–G). Moreover, inhibition of Drp1 attenuated the blue light-induced upregulation of NOX4 expression at both the mRNA and protein levels. (Figs. 4H–J). These findings suggest that Drp1 may mediate mitochondrial fission induced by blue light exposure through the regulation of NOX4 in RGCs. Inhibition of Drp1 reduces mitochondrial fission and NOX4 levels in RGCs under blue light exposure. (A) Mitochondrial morphology in R28 cells were labeled with Mito Tracker Red and images were captured by confocal microscopy after 12-hour blue light exposure. Cells were pre-treated with either Drp1-targeting siRNA (48 hours) or the Drp1 inhibitor Mdivi-1 (5 µM; six hours). Scale bars: 20 µm. The below panels are the amplification of white dotted boxes. Scale bars: 5 µm. (B) The relative length of mitochondria in each group was analyzed. 300 mitochondria/group from four different random fields were measured. (C) Western blot analysis of NOX4 expression in R28 cells after exposure to blue light for the indicated times. n = 6. (D) Quantification of the Western blot results shown in (C). (E) Western blot analysis of NOX4 expression in cytoplasmic and mitochondrial fractions isolated from R28 cells after 12 hours of blue light exposure. n = 4. (F, G) Quantification of the Western blot results shown in (E). Data are presented as mean ± SD. (H) R28 cells were pre-treated with either Drp1-targeted siRNA (48 hours) or the Drp1 inhibitor Mdivi-1 (5 µM; six hours) and subjected to blue light exposure. NOX4 protein expression was analyzed by Western blot. n = 4. (I) Quantification of Western blot results in (H). (J) R28 cells were pretreated with either Drp1-targeted siRNA (48 hours) or the Drp1 inhibitor Mdivi-1 (5 µM; six hours) followed by blue light exposure. NOX4 mRNA expression levels were quantified by qRT-PCR. Data are presented as mean ± SD; n = 4. *P < 0.05, **P < 0.01, ***P < 0.001. Drp1-Mediated Mitochondrial Fission Modulates NOX4 Expression Through the NF-κB Signaling Pathway in RGCs Subjected to Blue Light Exposure Next, we investigated the underlying mechanism by which DRP1 influences NOX4 levels to mediate mitochondrial differentiation induced by blue light exposure. Previous studies has have shown that the transcription factor NF-κB can regulate NOXs expression in human aortic vascular smooth muscle cells,29 prompting us to assess whether NF-κB similarly regulates NOX4 expression in RGCs following blue light exposure. To address this, we initially examined NF-κB activation in response to blue light. Western blot analysis indicated a significant increase in the phosphorylation levels of p65 post-exposure, without alterations in its total protein expression (Figs. 5A, 5B). Immunofluorescence analysis further confirmed the nuclear translocation of p65 after blue light exposure (Fig. 5C). Then, to validate the role of NF-κB in regulating NOX4 transcription, we constructed a series of truncated NOX4 promoter fragments (−2000 to −1 bp, −1709 to −1 bp, −1504 to −1 bp, −1411 to −1 bp, −1306 to −1 bp, and −1034 to −1 bp) and cloned them upstream of a luciferase reporter gene. Dual-luciferase assays revealed a significant decrease in promoter activity following the deletion of the region between −1411 and −1306 bp (Fig. 5D). Bioinformatic analysis using the JASPAR database predicted a potential p65-binding motif (ATTCCCT) within this region (Fig. 5E). Chromatin immunoprecipitation (ChIP)-PCR confirmed the direct binding of p65 to the predicted site on the NOX4 promoter (Fig. 5F). Collectively, these findings indicate that p65 directly binds to the NOX4 promoter and activates its transcription. NF-κB regulates NOX4 transcription in RGCs subjected to blue light exposure. (A) Western blot analysis of p-p65 and total p65 expression in R28 cells after exposure to blue light for the indicated times. n = 3. (B) Quantification of the Western blot results were shown in (A). (C) Immunofluorescence (IF) detection of p65 cellular location in R28 cells after 12 hours of blue light exposure. Scale bars: 10 µm. (D) Relative luciferase activity of truncated NOX4 promoter fragments. The 293T cells were transiently transfected with serial deletion constructs (−2000/−1 bp, −1709/−1 bp, −1504/−1 bp, −1411/−1 bp, −1306/−1 bp, −1034/−1 bp), followed by dual-luciferase assay after 48 hours. (E) Predicted NF-κB binding site within the NOX4 promoter region (−1411 to −1306 bp), identified using the JASPAR database. (F) NOX4 promoter occupancy by p65 analyzed by ChIP-PCR. Data were presented as mean ± SD of three independent experiments. ns P > 0.05, *P < 0.05, ***P < 0.001. Additionally, we investigated whether DRP1 regulates NOX4 levels to mediate mitochondrial differentiation induced by blue light exposure through NF-κB signaling. The application of Bay 11-7082, a p65 inhibitor, reversed the blue light-induced phosphorylation of p65 (Figs. 6A, 6B). Intriguingly, Bay 11-7082 administration was found to effectively suppress both protein (Figs. 6C, 6D) and mRNA (Fig. 6E) expression of NOX4 induced by blue light exposure, indicating that p65 transcriptionally regulates NOX4 expression. To further elucidate the relationship between Drp1 and the activation of p65, R28 cells were pre-treated with either Drp1 siRNA or the Drp1 inhibitor Mdivi-1 before blue light exposure. The findings demonstrated that inhibition of Drp1 effectively abrogated the blue light-triggered phosphorylation of p65 (Figs. 6F, 6G). Collectively, these results suggest that Drp1-mediated mitochondrial fission modulates NOX4 expression through the NF-κB signaling pathway in response to blue light exposure. Drp1-mediated mitochondrial fission modulates NOX4 expression through the NF-κB signaling pathway in RGCs subjected to blue light exposure. (A) R28 cells pretreated with Bay 11-7082 (10 µM) for six hours followed by 12 hours’ blue light exposure. Western blot analysis of p-p65 and total p65 expression. (B) Quantification based on results shown above. (C) NOX4 protein expression analyzed by Western blot in R28 cells pre-treated with the p65 inhibitor Bay 11-7082 (six hours) followed by blue light exposure (12 hours). BAY = Bay 11-7082. n = 5. (D) Quantification of Western blot results in (C). (E) NOX4 mRNA levels measured by RT-qPCR under treatment conditions identical to (C). (F) Western blot analysis of p65 and p-p65 expression in R28 cells after 12-hour blue light exposure, after pretreatment with Drp1 siRNA (48 hours) or Mdivi-1 (5 µM; six hours). (G) Quantification of Western blot results shown in (F). Data were presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. The Inhibition of Drp1 or NOX4 Synergistically Attenuates Blue Light-Induced Oxidative Damage and Apoptosis in RGCs Both In Vitro and In Vivo To further investigate the roles of Drp1 and NOX4 in blue light-induced oxidative stress, R28 cells were pretreated with the NOX inhibitor apocynin, the Drp1 inhibitor Mdivi-1, or a combination of both agents. TMRE staining indicated that treatment with either apocynin or Mdivi-1 alone restored the mitochondrial membrane potential depolarization induced by blue light, with the combination therapy producing a more pronounced effect (Figs. 7A–C). The assessment of mtROS using MitoSOX Red revealed that both inhibitors reduced blue light-induced mtROS production, with co-treatment significantly enhancing this effect (Figs. 7D–F). Additionally, the evaluation of cytosolic ROS using DHE staining confirmed the suppression of excessive intracellular ROS generation by either inhibitor, with maximal inhibition observed on co-administration (Fig. 8A). Flow cytometric analysis of Annexin V/PI staining further demonstrated that blue light induced apoptosis in R28 cells, whereas treatment with apocynin or Mdivi-1 ameliorated blue light-induced apoptosis, showing greater efficacy when used in combination (Figs. 8B, 8C). These findings collectively suggest that exposure to blue light induces excessive ROS production and apoptosis in RGCs through the involvement of Drp1 and NOX4. Notably, the combination of Mdivi-1 and Apocynin provided enhanced protective effects compared to either treatment alone, suggesting that although Drp1 regulates NOX4, these two molecules may also exert their pro-apoptotic effects through complementary or parallel pathways Inhibition of Drp1 and NOX4 blocks blue light-induced loss of mitochondrial membrane potential and mitochondrial ROS overproduction. (A) R28 cells were pretreated with 10 µM Apocynin, 5 µM Mdivi-1, or the combination of both for two hours, and subjected to blue light exposure. Mitochondrial membrane potential was assessed by TMRE staining. Images captured by fluorescence microscopy. Scale bars: 50 µm. (B) Flow cytometry quantification of TMRE fluorescence intensity under conditions identical to (A). (C) Quantification of flow cytometry results from (B). (D) Mitochondrial ROS levels detected by MitoSOX Red staining (red) under treatments described in (A). Images captured by fluorescence microscopy. Scale bars: 50 µm. (E) Flow cytometry analysis of MitoSOX Red fluorescence intensity under conditions identical to (D). (F) Quantification of flow cytometry results from (E). Apo, apocynin. Data are presented as mean ± SD of three independent experiments. **P < 0.01, ***P < 0.001. Inhibition of DRP1 and NOX4 mitigates excessive intracellular ROS production and apoptosis induced by blue light exposure. R28 cells were pretreated with 10 µM Apocynin, 5 µM Mdivi-1, or the combination of both for two hours, and subjected to blue light exposure. (A) Intracellular ROS detected by DHE staining (red) and visualized by fluorescence microscopy. Scale bars: 50 µm. (B) Apoptosis quantified by flow cytometry after Annexin V/PI co-staining. (C) Quantification of the ratio of apoptotic cells. Data were presented as mean ± SD of three independent experiments. ***P < 0.001. Subsequently, SpD rats received intravitreal injections of 1 nmol Mdivi-1, 2 nmol Apocynin, or a combination of both, administered 24 hours before exposure to blue light. Transmission electron microscopy analysis indicated that mitochondria in the control group displayed an elongated morphology with smooth outer membranes and well-defined cristae. In contrast, exposure to blue light resulted in mitochondrial damage, characterized by a rounded morphology, disrupted cristae, and vacuolization. Notably, pretreatment with the drugs mitigated these structural alterations, preserving the integrity of the cristae (Fig. 9A). IHC staining further demonstrated that Mdivi-1, Apocynin, or their combination significantly suppressed blue light-induced upregulation of NOX4 in RGCs (Fig. 9B). Consistent with in vitro findings, TUNEL staining revealed that both agents, whether used individually or in combination, attenuated blue light-induced apoptosis in RGCs (Fig. 9C). Collectively, these results underscore the roles of Drp1 and NOX4 as central mediators of blue light-induced mitochondrial damage and apoptosis in RGCs, providing novel therapeutic insights into the prevention of blue light-induced retinal injury. Targeting Drp1-mediated fission or NADPH oxidase blocks blue light-induced RGC damage in vivo. SpD rats were intravitreally injected with 1 nmol Mdivi-1, 2 nmol Apocynin, or the combination of both for 24 hours, and subjected to blue light exposure. Retina were dissected after exposure. (A) Mitochondrial morphology in RGCs was assessed by transmission electron microscopy. Scale bars: 200 nm. n = 3. (B) Immunohistochemical analysis of NOX4 expression in RGCs. Scale bars: 50 µm. The below panels are the amplification of black dotted boxes. Scale bars: 20 µm. n = 5. (C) RGC apoptosis was detected by TUNEL staining. Scale bars: 50 µm. n = 5. Discussion The pervasive adoption of LED lighting has resulted in increased exposure to blue light, both voluntarily and involuntarily, thereby raising substantial public health concerns related to blue light-induced visual hazards.2,34 Compared to other wavelengths within the visible spectrum, blue light is characterized by higher energy levels.35 This high-energy light is capable of penetrating the cornea and lens, reaching the retina directly and causing photochemical damage.36 The retina, with its intricate cellular composition and layered structure, is a vital component of the central nervous system.37 RGCs are located in the inner retinal layer and, along with their axons, constitute the optic nerve, which is responsible for transmitting visual signals. Mitochondria are predominantly concentrated within regularly spaced axonal varicosities along the unmyelinated segments of RGC axons.6,38 The elongated axons of RGCs are particularly susceptible to energy deficiencies, making them highly reliant on mitochondrial function and especially vulnerable to mitochondrial dysfunction and oxidative stress.39 The inner mitochondrial membrane is critically important as it houses respiratory complexes that contain key chromophores, such as flavins and porphyrins, which have absorption peaks within the blue light spectrum.40 Studies have shown that the activation of photosensitive chromophores by blue light in monocytes leads to the generation of intracellular ROS.41 Similarly, when blue light interacts with mitochondrial chromophores, it can induce an oxidative stress response through photochemical or photomechanical mechanisms, thereby contributing to RGC damage.7 Excessive exposure to blue light has also been shown to significantly increase ROS production in retinal pigment epithelial cells, which can result in photoreceptor loss, lipid peroxidation, and apoptosis.1 Despite these findings, the regulatory mechanisms by which blue light affects mitochondrial morphology and function, and how these changes contribute to retinal phototoxicity, are not yet fully understood. In this study, we demonstrate that exposure to blue light induces mitochondrial hyper-fission in RGCs, characterized by fragmented tubular or spherical morphologies and a marked reduction in average mitochondrial length. This perturbation was associated with dysregulated expression of proteins related to mitochondrial dynamics, characterized by increased levels of Drp1 and decreased levels of MFN2. Additionally, in vivo exposure of rats to blue light replicated mitochondrial damage in RGCs, evidenced by rounded mitochondrial shapes, cristae fragmentation, and significant vacuolization. Collectively, these findings identify blue light as a potent inducer of mitochondrial structural and functional compromise in RGCs. Mitochondrial fusion and fission are crucial for regulating organelle morphology, facilitating content exchange, maintaining mitochondrial DNA (mtDNA) integrity, enabling mitochondrial inheritance, and segregating damaged mitochondria.11 Impaired fusion leading to mitochondrial fragmentation results in severe transport deficits.42 Although fragmented mitochondria exhibit increased motility, they also show elevated levels of mtROS and reduced ATP production.43 Mitochondrial damage further enhances mtROS generation and the release of immunogenic factors.44 Notably, Drp1 phosphorylation and mitochondrial translocation, facilitated by Stat2, promote excessive mitochondrial fission, which shifts cellular metabolism from ATP synthesis to ROS production, thereby inducing NF-κB-dependent polarization in inflammatory macrophages.45 Our findings indicate that exposure to blue light significantly elevates mtROS levels in RGCs. Intracellular ROS can originate from various sources, with mitochondrial and enzymatic pathways being the primary contributors. A notable example of the synergistic interaction between different ROS sources is the phenomenon of RIRR.46 RIRR is a fundamental feedforward mechanism in which localized ROS generation initiates an amplification cascade, leading to a rapid increase in intracellular ROS accumulation, which subsequently drives tissue inflammation and cell death.47 Our data suggest that the significant increase in intracellular and mitochondrial ROS levels in RGCs upon blue light exposure may implicate mtROS-triggered RIRR as a contributing factor to blue light-induced phototoxicity. The inhibition of Drp1 mitigated both excessive mitochondrial fission and mitochondrial damage induced by blue light exposure, significantly reducing intracellular ROS levels and apoptosis. These results suggest that blue light exposure triggers Drp1 overexpression, which in turn leads to mitochondrial damage and mtROS production. The resultant cellular injury may be further intensified by the large-scale ROS production facilitated by the RIRR mechanism. The enzymatic systems predominantly responsible for ROS production include NOXs, lipoxygenases, xanthine oxidase, and cyclooxygenases, with NOXs being the principal source of ROS generation across various cell types.24 Among the NOXs family, NOX4 is unique in its constitutive activity and mitochondrial localization, with its tissue expression level directly correlating with its ROS-generating capacity.25 In our study, we observed that blue light exposure upregulated NOX4 expression; however, this upregulation was diminished upon Drp1 inhibition. Additionally, treatment with a NOX inhibitor effectively suppressed blue light-induced intracellular ROS production, mitochondrial damage, membrane potential disruption, and apoptosis. The findings of this study indicate that exposure to blue light results in the upregulation of NOX4, which subsequently enhances ROS production and leads to cellular damage. This process is intricately linked to increased expression of Drp1. Nonetheless, the exact mechanisms underlying these observations have yet to be fully elucidated. Existing evidence suggests that mtROS can stimulate macrophage inflammatory responses by activating NF-κB and the NLRP3 inflammasome.48,49 Additionally, hydrogen peroxide (H₂O₂) is known to facilitate the rapid activation of NF-κB through the induction of ubiquitination and degradation of IκBα via phosphorylation at Ser32 and Ser36, or through direct phosphorylation at Tyr42.50 Moreover, research by Manea et al.29 has shown that NF-κB transcriptionally regulates the NOX subunit p22phox in human aortic vascular smooth muscle cells, thereby enhancing NADPH oxidase-mediated ROS production. In light of these findings, we explored the involvement of NF-κB in blue light-induced damage to RGCs. Our results reveal that blue light exposure increases phosphorylated p65 (p-p65) levels in RGCs, an effect that is mitigated by the inhibition of Drp1. Importantly, pharmacological inhibition of NF-κB signaling was found to suppress the blue light-induced upregulation of both NOX4 protein and mRNA. Complementary evidence from dual-luciferase reporter assays and ChIP-PCR has further corroborated the role of NF-κB as a direct transcriptional regulator of NOX4. Although our mechanistic data demonstrated a direct regulatory link between Drp1/p65 and NOX4 transcription, the additive neuroprotective effects observed with pharmacological inhibitors indicate that the Drp1/p65/NOX4 axis is not the sole pathway involved. It is likely that DRP1 and NOX4 also contribute to RGC death through independent mechanisms under blue light stress. Conclusions Our findings indicate that exposure to blue light results in the upregulation of Drp1 expression and excessive mitochondrial fission, which in turn leads to an increased production of mitochondrial-derived ROS. This process activates the NF-κB signaling pathway, thereby enhancing the expression of NOX4 and further amplifying ROS production through the RIRR mechanism, ultimately contributing to retinal damage. These insights offer a mechanistic foundation for the development of preventive and therapeutic strategies against blue light-induced ocular injury. Supplementary Material Supplement 1