Stress Granule Induction in Rat Retinas Damaged by Constant LED Light

Abstract

Purpose

Stress granules (SGs) are cytoplasmic biocondensates formed in response to various cellular stressors, contributing to cell survival. Although implicated in diverse pathologies, their role in retinal degeneration (RD) remains unclear. We aimed to investigate SG formation in the retina and its induction by excessive LED light in an RD model.

Methods

Rat retinas were immunohistochemically analyzed for SG markers G3BP1 and eIF3, and SGs were also visualized by RNA fluorescence in situ hybridization. Additionally, SGs were induced in primary retinal cell and eyeball cultures using sodium arsenite. Light exposure experiments used LED lamps with a color temperature of 5500 K and 200 lux intensity for short-term or two- to eight-day exposures.

Results

SGs were predominantly detected in retinal ganglion cells (RGCs) and inner nuclear layer (INL) cells, with arsenite-induction verified in RGCs. SG abundance was higher in animals exposed to light for 2-8 days compared to light/dark cycle controls. RGCs consistently exhibited more SGs than INL cells, and INL cells more than outer nuclear layer (ONL) cells (Scheirer-Ray-Hare test: H = 13.2, P = 0.0103 for light condition, and H = 278.2, P < 0.00001 for retinal layer). These observations were consistent across four independent experiments, each with three animals per light condition.

Conclusions

This study characterizes SGs in the mammalian retina for the first time, with increased prevalence after excessive LED light exposure. RGCs and INL cells showed heightened SG formation, suggesting a potential protective mechanism against photodamage. Further investigations are warranted to elucidate the role of SGs in shielding against light stress and their implications in retinopathies.

Retinal degeneration (RD) is a neurodegenerative disease with numerous contributing factors, encompassing processes like remodeling, photoreceptor death, and the deterioration of both the structure and function of this tissue.¹ The eye has evolved protective mechanisms to safeguard against excessive light exposure, which can be detrimental to the retina.^2,3 Artificial light alters the natural illumination, leading to adverse effects on retinal function. Light-emitting diodes (LEDs) have become the primary source of both household and public lighting. They are also integral to modern technologies like computers, TVs, tablets, cell phones, and gaming consoles. As a result, our visual system is expose to LED light in excess. Despite the cost-effectiveness and energy efficiency of LEDs, they emit a significant amount of blue light, with wavelengths ranging from 460 to 500 nm.⁴ This blue light can potentially have detrimental effects on human vision.⁵

The retina is the most energetically demanding tissue, known for its abundant oxygen supply and rich content of polyunsaturated fatty acids. It also contains elevated levels of photosensitizers, further rendering its susceptibility to oxidative stress.^6–8 The production of reactive oxygen species (ROS) in the context of oxidative stress is a crucial component of the common pathway leading to neural damage in various acute and chronic neurological eye disorders.⁹ Although ROS play important roles in cell signaling and regulation, their excessive production can result in damage to cellular macromolecules, including DNA, proteins, and lipids.¹⁰

We have previously established a model of RD induced by continuous exposure to low-intensity LED light (LL) in Wistar albino rats.¹¹ Our research has revealed an elevation of ROS production within the outer nuclear layer (ONL) of retinas from animals exposed to LL for five days. Concurrently, there has been a decline in docosahexaenoic acid, a crucial component of rod cell external segment membranes, likely because of oxidative processes.¹² Additionally, significant photoreceptor cell loss has been observe after seven days of LL exposure, accompanied by heightened rhodopsin phosphorylation early in the exposure period. Notably, most retinal ganglion cells (RGCs) and inner nuclear layer (INL) cells remain viable, albeit with alterations in the expression and localization of photopigments like melanopsin (OPN4) and neuropsin (OPN5).¹³ These findings underscore the varied susceptibility of distinct retinal cell types to damage induced by excessive LED light exposure. This heterogeneity is anticipated given the diverse array and distribution of photopigments across retinal cells, along with alternative phototransduction pathways. Moreover, the observed differences may also stem from the presence of distinct defense mechanisms against photodamage among these cell types.

Cellular stress triggers an adaptive program known as the integrated stress response enables to endure and survive. Nevertheless, if cells are unable to overcome it, it can also activate the cell death mechanism to eliminate damaged cells.¹⁴ Depending on the intensity and duration of the stress, one of these two responses prevails. A prominent feature of the integrated stress response is its ability to inhibit global protein synthesis promoting the accumulation of translationally stalled 48S complexes that undergo liquid-liquid phase separation. These complexes form biocondensates of RNA and proteins known as stress granules (SGs).^15,16 SGs are enriched in polyadenylated mRNA, small ribosomal subunits, translation initiation factors, RNA-binding proteins (RBPs), and other factors.^17–19 The function of SGs has been linked to the regulation of translation, stability, and storage of cytoplasmic mRNA; however, this is not completely elucidated and is a matter of controversy.²⁰ Considering that many molecules linked to signaling pathways are concentrated in the SG, it has been proposed that they act as signaling hubs.²¹ Numerous factors linked to apoptosis are concentrated in SG, inhibiting several pro-apoptotic signaling pathways and favoring cell survival.^22–27 The formation of SGs also diminishes the accumulation of ROS.²⁸ Given the oxidative stress observed in our RD model, we hypothesized that excessive light exposure could induce SG formation in the retina. The objective of this study was to characterize, for the first time, SG formation along the different cell types of the mammalian retina and to examine its potential induction by light exposure.

Material and Methods

Animals

All procedures conducted adhered to the guidelines outlined in the ARVO statement for the use of animals. Additionally, all protocols were approved by the local animal committee (School of Chemistry, UNC, Exp. EXP#2023-00453889-UNC-ME#FCQ). Male albino Wistar rats, aged 12 to 15 weeks, bred in our laboratory for five years, were housed under a 12-hour light/12-hour dark (LD) cycle. White fluorescent light of ∼50 lux intensity was provided. Throughout the experiment, the rats had access to food and water as desired.

Immunohistochemistry (IHC)

IHC was conducted as before.¹¹ Rat eyes were fixed in 4% (w/v) paraformaldehyde in PBS overnight (ON) at 4°C, cryoprotected in sucrose 30% (w/v) and mounted in an optimal cutting temperature compound (OCT) (Tissue-Tek; Sakura Finetek USA, Inc., Torrance, CA, USA). Then, 20 µm retinal sections were cut along the horizontal meridian (nasal-temporal) by cryostat (HM525 NX; Thermo Fisher Scientific, Waltham, MA, USA). Sections were washed in PBS and permeabilized with PBS-T (PBS with 0.5% Triton X-100), for 90 minutes at room temperature (RT). Then, they were blocked with blocking buffer [PBS supplemented with 3% (w/v) BSA, 0.1% (v/v) Tween 20, 1% (v/v) Glycine, and 0.02% (w/v) Sodium Azide] for 2.5 hours at RT and incubated with anti-eIF3 (1:300, sc-16377; Santa Cruz Biotechnology, Dallas, TX, USA) or anti-G3BP1 (1:1000, A302-033A; Bethyl Labs, Montgomery, TX, USA), two robust SGs markers,^15,29–31 diluted in blocking buffer, ON at 4°C in a humidified chamber. Samples were then rinsed in PBS-TW (0.1% (v/v) Tween 20 in PBS) and incubated with goat anti-rabbit IgG Alexa Fluor 488 or 549 (1:1000; Thermo Fisher Scientific), respectively, and DAPI (3 µM), for one hour at RT. Finally, they were washed three times in PBS and mounted with Mowiol (Sigma-Aldrich Corp., St. Louis, MO, USA).

Fluorescence In Situ hybridization (FISH) and Immunofluorescence (IF)

Samples were obtained and fixed as detailed in the preceding section. FISH-IF was carried out according to Meyer and colleagues.³² Fixed samples were permeabilized with TBS-T (0.01 M Tris buffer pH 7.4, 0.1 M NaCl, 0.2% [v/v] Triton X-100), 20 minutes at RT, washed five minutes with TBS, and prehybridized using hybridization buffer (H-7140, Sigma-Aldrich Corp.) for one minute at RT. Subsequently, they were hybridized with Cy3-Oligo(dT)₃₀ (Sigma-Aldrich Corp.) diluted in hybridization buffer, at a final concentration of 20 nM, ON at 40°C in a humidified chamber. Then, samples were rinsed once with Washing Buffer 1 (50% [v/v] formamide 12.6 M, 0.25 M NaCl, 0.075 M Tris Buffer [pH 8.5], and 0.1% [v/v] Tween 20) five minutes at RT in constant shaking, four times with washing buffer 2 (0.05 M NaCl, 0.075 M Tris Buffer [pH 8.5] and 0.1% [v/v] Tween 20) for 15 minutes each in constant shaking and with TBS 5 min. Then, they were blocked with blocking buffer for 2.5 h at RT with continuous shaking and incubated overnight with anti-eIF3 (sc-16377; Santa Cruz Biotechnology) diluted in blocking buffer at 4°C in a humidified chamber. Samples were then rinsed in TBS and incubated with goat anti-rabbit IgG Alexa Fluor 488 (1:1000) and DAPI (3 µM), for one hour at RT. Finally, they were washed three times with TBS-T for five minutes each, once with TBS for five minutes, and mounted in Mowiol.

Primary Retinal Cell Cultures

Primary retinal cell cultures were obtained from rats at postnatal day 7.³³ Retinas were manually dissected with gentle up-and-down passes in ice-cold Ca⁺²–Mg⁺²–free Tyrode's buffer and treated with papain (P3125; Sigma-Aldrich Corp.) for 20 minutes and deoxyribonuclease I (18047-019; Invitrogen, Carlsbad, CA, USA) for 10 minutes at 37°C under constant air flow in a humid atmosphere. Cells were precipitated at 5000g for 20 minutes at 4°C and resuspended in 10% fetal bovine serum (FBS; Gibco, Thermo Fisher Scientific)–DMEM. Cells were seeded in coverslips treated with poly-L-lysine (10 µg/mL) and grown for six days in neurobasal medium (Gibco, Thermo Fisher Scientific) supplemented with 0.05% (v/v) amphotericin B, 0.1% (v/v) forskolin (Sigma Aldrich), 0.02% (v/v) Recombinant Human BDNF (R&D Systems), 2% (v/v) B27 (GIBCO), and 1% (v/v) L-glutamine (Glutamax; Gibco, Thermo Fisher Scientific).

SG Induction by Sodium Arsenite

To induce SG formation in cell cultures, sodium arsenite (NaAsO₂, S7400; Sigma-Aldrich Corp.) was added to the media at final concentrations of 250 or 500 µM and incubated for 30 minutes. For ex vivo induction of retinal SGs, eyeballs were dissected and subjected to intravitreal injection using a 30-gauge needle to deliver 2 µL of either 50 mM sodium arsenite or vehicle (sterile Milli-Q H₂O). After injection, the eyeballs were incubated at 37°C in 10% FBS-DMEM for 30 minutes, rinsed with PBS, and subsequently fixed and processed for IHC after the protocol described above.

Immunocytochemistry

Cells were washed with PBS, fixed 15 minutes in 4% PFA and permeabilized 10 minutes with −20°C methanol at RT, according to Kedersha and Anderson.³¹ After 3X-washes with cold PBS for five minutes, they were incubated for one hour with blocking buffer. Then samples were incubated with anti-G3BP1 (1:1000, A302-033A; Bethyl Labs) and anti-DM1A (1:1000; Sigma-Aldrich Corp.) diluted in blocking buffer, ON at 4°C in a humidified chamber. Samples were then rinsed in PBS-T and incubated with goat anti-rabbit IgG Alexa Fluor 488 or 549 (1:1000), respectively, and DAPI (3 µM), for one hour at RT. Finally, they were washed three times with PBS and mounted with Mowiol.

Light Exposure Protocols and Induction of Retinal Degeneration (RD)

RD was induced according Contín et al.¹¹ Animals were exposed to constant light (LL) for varying durations: 54 hours (LL2, ∼2 days), 102 h (LL4, ∼4 days), 150 h (LL6, ∼6 days), and 198 h (LL8, ∼8 days). LED lamps (T-13/4 3294-15/T2C9-1HMB, color temperature of 5500 K; Everlight Electronic Co., Ltd., Taipei, Taiwan) were installed in specially designed boxes, with the temperature-controlled at 24° ± 1°C. Light intensity at the level of the rats’ eyes was measured at 200 lux using a light meter (model 401036; Extech Instruments Corp., Waltham, MA, USA). Control retinas were obtained from animals maintained in an LD cycle (50 lux white fluorescent light), dissected six hours after the lights were turned on (Fig. 4). For short-duration light exposure experiments, identical stimulation boxes were utilized; animals were kept in darkness for 18 hours before being exposed to light for intervals of zero to 12 hours (Fig. 3). Euthanasia was performed using a CO₂ chamber.

Figure 4.

Induction of SGs in the inner retina by excessive low-intensity LED light exposure in a model of retinal degeneration. Animals were exposed to constant LED light (200 lux, color temperature of 5500 K) for 54 hours (LL2, ∼2 days), 102 hours (LL4, ∼4 days), 150 hours (LL6, ∼6 days), and 198 hours (LL8, ∼8 days), in temperature-controlled chambers at 24° ± 1°C. Control retinas were obtained from animals maintained in a LD cycle (white fluorescent light <50 lux) and dissected six hours after the lights were turned on. (A) Representative images of retinas analyzed by IHC using an anti-G3BP1 antibody. LD untreated and LD represent the same image before and after processing for quantification (See Material and Methods). All other images are shown after processing. Scale bar: 10 µm. (B) Quantification of SGs per cell in different layers of the retina in each light condition. Data represent results from four independent experiments, each using three animals per group, with four to six images acquired per case (n = 16–24 images per group). The Scheirer-Ray-Hare test was used for analyzing all light conditions and retinal layers together (H = 13.2, P = 0.01034 for light conditions; H = 278.2, P < 0.00001 for retinal layer; H = 3.301, P = 0.91408 for interactions between the two factors; see Table and Supplementary Table S1). The results from LL2, LL6, and LL8 were higher than LD (Dunn's test with a Bonferroni correction). (C–E) Data from the GCL (C), INL (D), and ONL (E) are visualized for comparison within each retinal layer. GCL was analyzed by ANOVA test followed by Tukey's post hoc, whereas INL and ONL were analyzed by Kruskal-Wallis rank sum test followed by Dunn's test with a Bonferroni correction. Significance is indicated by asterisks (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

Figure 3.

SGs in retinas exposed to short periods of low-intensity LED light. Rats, adapted to darkness for 18 hours, were subjected to light exposure experiments. Rats were exposed to 200 lux intensity LED light for zero (control), four, six, or 12 hours. After exposure, retinas were dissected and subjected to IHC using an anti-G3BP1 antibody to visualize SGs. Representative images of the retina are shown. For the zero hour condition, the raw image is displayed on the left, whereas the processed image obtained after segmentation (as described in the Material and Methods section) to identify and quantify SGs is shown on the right. For all other time points, only the processed images are displayed. Nuclei were stained with DAPI, whereas SGs were labeled with anti-G3BP1 antibody. Scale bar: 10 µm.

Image Acquisition and Quantification of SGs

Confocal imaging was conducted using a FluoView FV1200 confocal microscope (Olympus, Tokyo, Japan). A 60×/1.3 silicone immersion objective (UPLSAPO60XS; Olympus, Tokyo, Japan) captured images at a resolution of 2048 × 720 pixels. Image Z stacks were obtained with a pinhole of adjusted for 2 µm optical slices and 1.4 zoom, Kalman 0. ImageJ software was used for image processing.

For SG quantification, each slice within the Z stacks underwent the following procedures: (1) Background subtraction (value = 70), outlier removal (radius = 1, threshold = 200), background rolling ball radius subtraction (radius = 3), and gamma adjustment (value = 2). The stack was duplicated, and Gaussian filters with radii of 1 (g1) and 3 (g3) were applied, followed by subtraction (g1–g3). A maximum intensity projection was created from the resulting stack. (2) Manual selection of each retinal layer area. (3) Counting SGs within each ROI corresponding to the layers using Process > Find Maxima > Prominence = 100–200. Quantification of GCL and INL nuclei was manual. ONL nuclei quantification used Process > Find Maxima > using a prominence value to select each nucleus only once.

Statistical Analysis

The analysis was conducted with RStudio. Normality and homogeneity of the variance assumptions were evaluated using the Shapiro-Wilks and the Bartlett tests, respectively. The non-parametric Scheirer-Ray-Hare test was used to verify the significance difference between the days of light treatment and between layers of the retina. In case there was a significant difference, the Dunn`s ad hoc test with a Bonferroni correction were employed. We also compared the number of SGs formed in the different retinal layers throughout the light treatment. The Kruskal-Wallis test was used to verify if there was a statistically significant difference between treatments in INL and ONL, followed by Dunn’s test with a Bonferroni correction. GCL was analyzed using an ANOVA test followed by Tukey's post hoc test. Similarly, for each light treatment condition, a non-parametric Kruskal-Wallis test followed by an ad hoc Dunn’s test with a Bonferroni correction were performed to determine whether there was a difference between the layers of the retina.

Results

Detection and Characterization of SGs in Rat Retina

To characterize SGs within the retina, we initially performed IHC using antibodies targeting the established SG marker G3BP1.^15,30,31 Figures 1A illustrates the G3BP1 immunolabeling in retinas kept in darkness for 18 hours or exposed for 48 hours to 200 lux LED light (LL2). Figure 1B shows RGCs in LL2 with higher resolution. Numerous distinct spots resembling SGs are evident, consistent with their typical appearance. Given the protein's localization within both these biocondensates and the cytoplasm, a background signal is also observable. Predominantly, these foci are observed within the RGCs, with some presence in the INL and rare occurrence in the ONL.

Figure 1.

Visualization of SGs in rat retina. (A) IHC staining of retinas from rats kept in darkness for 18 hours (Dark) or exposed to constant LED light (200 lux intensity) for 2 days (LL2). Images acquired by confocal fluorescence microscopy depict the SG marker G3BP1 detected using specific antibodies. Nuclei stained with DAPI (blue); G3BP1 visualized in green. Scale bar: 10 µm. (B) Higher-resolution images of retinal ganglion cells (RGCs) from experiments conducted similarly to panel A in LL2 conditions. (C) Simultaneous analysis of RGCs using poly(A)+ RNA fluorescence in situ hybridization (Poly[A]+ RNA-FISH) and IF. Cy3-Oligo(dT)30 probe binds to the 3′-poly(A) end of polyadenylated mRNAs concentrated within SGs. Anti-eIF3 antibody, another SG marker, was used. Nuclei stained with DAPI (blue); eIF3 in green; Poly(A)+ RNA in red. Scale bar: 5 µm. (See Material and Methods for experimental details).

To further confirm the presence of SGs in the retina, we conducted dual-labeling of SGs using Poly(A)+ RNA Fluorescent In Situ Hybridization (Poly(A)+ RNA-FISH) in combination with immunofluorescence using an antibody targeting eIF3, another established SG marker.^15,29,31 Cy3-Oligo(dT)₃₀, a probe specifically binding to the poly(A) tail of mRNA concentrated within SGs, was employed. Figure 1C shows the co-localization of both signals, Poly(A)+ RNA and eIF3, within bright spots, providing evidence for the presence of SGs within retinal tissue, primarily concentrated in RGCs.

Induction of Retinal SGs by Sodium Arsenite

To validate the identity of the structures identified, we conducted experiments to determine whether sodium arsenite, a well-established inducer of SGs,^31,34,35 could effectively augment the quantity or size of the observed bright spots detected by IHC. In Figure 2A, immunocytochemistry demonstrates G3BP1 labeling in primary cell cultures of the total rat retina. Nontreated cells (control) display a diffuse cytoplasmic labeling of G3BP1. In contrast, cells treated with NaAsO₂ exhibit distinct dot-shaped structures characteristic of G3BP1 concentration within SGs. Notably, a higher number of these distinct dot-shaped structures were visible at the higher arsenite concentration utilized.

Figure 2.

Induction of SGs with arsenite in cultured cells and retinal tissue. (A) Primary retinal cultures from rats at postnatal day 7 were treated with the SG inducer sodium arsenite at concentrations of 250 or 500 µM for 30 minutes or left untreated as control. The left panels display α-Tubulin and DAPI signals, whereas the right panels show G3BP1 staining. Scale bar: 5 µm. (B) Ocular globes were dissected and immediately intravitreally injected with 2 µL of 50 mM sodium arsenite or vehicle solution (sterile MQ H₂O), followed by a 30-minute incubation at 37°C in 10% FBS-DMEM. Subsequently, the eyes were fixed and processed for IHC. The left panels show in RGCs the signal from G3BP1, and the right panels show G3BP1 and DAPI staining. Scale bar: 10 µm.

Subsequently, we investigated the impact of arsenite on ex vivo retinal tissue. Eyecups were dissected, subjected to intravitreal injection with arsenite, and cultured for 30 minutes before IHC analysis. In Figure 2B, a few discernible bright spots are evident in nontreated RGCs (control); however, a marked rise in SG-like dots is observed in RGCs from retinas treated with NaAsO₂. Remarkably, the highest count of bright spots was observed in RGCs. These findings substantiate that treatment with NaAsO₂ induces the formation of SGs, both in cell culture and ex vivo retinal experiments.

SGs in Retinas Exposed to Short-Term Low-Intensity LED Light

In Figure 1, we detected SGs in animals exposed to constant light for 48 hours, we aimed to ascertain whether shorter durations (0–12 hours) of low-intensity LED light exposure (200 lux) could induce the formation of these cellular biocondensates. Figure 3 and the Table illustrate the presence of these granules in both control conditions and at three distinct time points after light exposure. The Scheirer-Ray-Hare test revealed a significant effect of the retinal layer on the number of SGs per cell (H = 131.34, P < 0.00001), while no significant effects were observed for the light treatment (H = 2.01, P = 0.5696) or the interaction between light treatment and retinal layer (H = 1.36, P = 0.9680). Additionally, we confirmed through Kruskal-Wallis rank sum tests that there are no statistically significant differences in the number of SGs per cell across light treatments within each retinal cell layer (ONL: χ² = 6.738, P = 0.08074; INL: χ² = 6.482, P = 0.09038; GCL: χ² = 4.628, P = 0.20120).

Table.

Number of SGs Per Cell Under Different Light Treatments

Light Exposure Duration*	GCL (Mean ± SD)	INL (Mean ± SD)	ONL (Mean ± SD)
Short-term exposure (Fig. 3)
0 hours†	2.90 ± 1.98	0.22 ± 0.21	0.015 ± 0.016
4 hours	2.05 ± 1.46	0.45 ± 0.39	0.035 ± 0.035
6 hours	2.88 ± 2.24	0.37 ± 0.31	0.012 ± 0.012
12 hours	1.63 ± 1.34	0.22 ± 0.24	0.022 ± 0.015
Long-term exposure (Fig. 4)
LD‡	1.73 ± 1.51	0.11 ± 0.09	0.003 ± 0.004
2 days (LL2)	3.99 ± 1.38	0.24 ± 0.13	0.010 ± 0.009
4 days (LL4)	2.89 ± 1.33	0.20 ± 0.11	0.005 ± 0.007
6 days (LL6)	3.49 ± 1.17	0.25 ± 0.17	0.009 ± 0.010
8 days (LL8)	3.56 ± 1.29	0.31 ± 0.18	0.008 ± 0.010

^*Light exposure at 200 lux of white LED light.

^†Animals were kept in darkness for 18 hours before being exposed to light for zero, four, six, or 12 hours.

^‡LD: 12-hour light (<50 lux, white fluorescent)/12-hour dark cycle, sampled six hours after lights on.

SG Formation in Rat Retinas Exposed to Prolonged Low-Intensity LED Light

To examine SGs within our RD model—a condition where animals are continually exposed to 200 lux LED light for two to eight days—we assessed SG counts across various retinal layers in control conditions and during the progression of RD. Our objective was to pinpoint the particular retinal cell types exhibiting heightened SG formation in response to prolonged light exposure. Figure 4A shows representative retinal images in which SGs were labeled with anti-G3BP1. To quantify SGs as a function of light exposure time, we analyzed retinas of animals exposed to two to eight days of LL. Four independent experiments were carried out, in each of them 3 animals per group were used and five to seven images were taken for each lighting condition (n = 20–28/group). Figures 4B and 4C and the Table show the results of the four experiments analyzed together. To analyze the combined effect of the studied factors, we used the Scheirer-Ray-Hare test. Both light exposure time (H = 13.2; P = 0.01034) and retinal layer (H = 278.2; P < 0.00001) exhibited significant changes in SG numbers when normalized by the number of cells (total nuclei of each layer). No combined effect of both factors was observed (H = 3.301; P = 0.91408). Post-hoc analysis (Dunn's test with Bonferroni correction) revealed lower SG counts per cell in control retinas compared to those of animals maintained for two, six, or eight days in LL (Fig. 4B). In all cases except LD controls, SG numbers were significantly higher in RGCs than in INL cells, and higher in INL cells than in ONL cells, where SGs were rare (Fig. 4; Table, and Supplementary Material S1). Similar results were obtained when each factor was analyzed separately using the Kruskal-Wallis rank sum test (Figs. 4C–E, Supplementary Material S1). Considering the variability in cell number and size within each layer, we further normalized the data by area (Supplementary Material 2). The results remained consistent, with a greater number of SGs in light-treated retinas, and RGCs harboring more SGs than INL cells, which, in turn, contained more SGs than ONL cells.

Discussion

In this study, we used a multifaceted approach, using IHC with antibodies targeting two SG markers, eIF3 and G3BP1, along with dual-labeling through Poly(A)+ RNA-FISH and immunofluorescence with an eIF3-specific antibody. Through these methodologies, we successfully identified SGs in the rat retina. To validate the authenticity of our SG detection, we conducted experiments using sodium arsenite, a well-established inducer of SG formation. Remarkably, retinas or retinal cell cultures treated with arsenite exhibited a significant increase in SG-like dots compared to control retinas, which showed minimal SG presence. Furthermore, we investigated SG development in a model of RD induced by prolonged exposure to constant low-intensity LED light. After 48 hours of light exposure, a notable increase in SGs was observed in both RGCs and the INL, indicating that excessive light exposure indeed induces SG formation. Notably, the layers of the retina that demonstrated higher SG quantities were the ones that exhibited greater resilience in the face of RD, highlighting a potential link between SG formation and cellular survival in this degenerative model.

The majority of investigations into SGs have been conducted in cell cultures, often induced under non-physiological conditions. While SG presence has been documented in neurons, such as in mouse primary cortical neurons,³⁶ we observe a dearth of prior descriptions regarding SGs in the retina. Our search yielded only one account of G3BP1-containing cytoplasmic granules in the rat retinal ganglion cell line RGC-5.³⁷ Notably, uncertainties persist concerning the origin and nature of this cell line.^38–41 Although SGs have been identified in cultured retinal pigment epithelium cells,⁴² our study constitutes the inaugural documentation of SGs in the retina and their response to a environmental stimulus, such as light.

Considering the pro-survival role attributed to SGs and our finding that they are induced in the inner retina in response to prolonged exposure to low-intensity LED light, it can be hypothesized that these granules play a role in the survival of RGCs and INL cells. In these cells, the number of SGs increases, whereas in cones and rods, the neurons that undergo cell death, are scarcely distinguishable. In this study, we observed a correlation, and future experiments are required to determine the validity of this hypothesis.

Conversely, SGs have also been associated with neurodegenerative diseases such as amyotrophic lateral sclerosis, frontotemporal dementias, and Alzheimer's disease.¹⁶ Some proteins associated with SGs have been linked to the formation of pathological protein aggregates (e.g., TDP-43 and FUS proteins). In many cases, the formation of these aggregates is associated with mutations in proteins normally found in SGs, which, when modified, impact the disassembly and clearance of the granules.^16,43 Also, it has been observed that under chronic stress conditions, such as nutrient starvation, noncanonical SGs acquire a different composition and may promote apoptosis.⁴⁴ Additionally, chronic stress can precondition neurons to not form SGs.⁴⁵

In our approach, we cannot distinguish whether the detected SGs favor cell survival or, conversely, in the context of continuous light exposure, chronically promote the possibility of producing pathological aggregates or apoptosis. In the majority of experimental paradigms used to study SGs, their induction occurs under conditions of acute stress, typically confined to a short duration of minutes or a few hours. In the model we employed, the stress is chronic (2–8 days). We are unaware of the dynamics of SG formation and clearance. For instance, SGs could initially form rapidly in photoreceptor cells, which are most susceptible to light stress, but unable to survive, triggering death processes that do not involve SG formation.

Our study reveals, for the first time, the presence of SGs in the mammalian retina, with their numbers increasing in response to excessive LED light exposure. Significantly, their prevalence is notably higher in inner retinal cells, which exhibit a remarkable resistance to light-induced damage. Further investigations are necessary to determine whether SGs play a pivotal role in shielding against light stress or potentially contribute to other retinopathies.