Remote Preconditioning Provides Protection Against Retinal Cell Death From Retinal Detachment

Abstract

Purpose: Remote preconditioning involves injury to a tissue that results in protection to a subsequent injury in a distal tissue. Here, we investigated the impact of remote preconditioning on retinal detachment (RD) injury, hypothesizing that a previous contralateral RD would protect the fellow retina against inflammation and cell death following its detachment.

Methods: RD was created in adult C57BL/6J mice with subretinal sodium hyaluronate injection. Preconditioning involved RD in the right eye at 1, 3, 7, or 28 days before left eye detachment, whereas the control group only received RD to the left eye. Retinas were harvested 24 hours post-left eye detachment in both groups. Cell death was assessed using Cell Death Detection ELISA and mRNA expression was evaluated via qRT-PCR.

Results: Contralateral RD promoted a transient protection against retinal cell death from 1 to 3 days and waned by 7 days compared with control RD retinas with intact fellow retinas. Contralateral RD significantly protected against post-RD cell death (P = 0.0002) and caspase 3 cleavage (P = 0.0449), compared with control RD retinas with intact fellow retinas 1-day post-RD. Detached fellow retinas from the preconditioning group expressed significantly less Tnfa (P = 0.0066), Cxcl10 (P = 0.0099), and Fas (P = 0.0223) mRNAs, compared with the detached retinas of the control group. In contrast, upregulation of type-I-IFN pathway genes, including Irf7 (P = 0.0106) and Ifit1 (P = 0.0740), following RD was higher in the preconditioning group.

Conclusions: RD in one eye produces a transient remote preconditioning effect that protects the fellow retina against retinal cell death following subsequent RD.

Retinal detachment (RD), defined as the physical separation of the neurosensory retina from the underlying retinal pigment epithelium (RPE) and choroid, results in outer retinal hypoxia, inflammation, and photoreceptor (PR) degeneration.¹ This ocular emergency affects 5 to 20 people per 100,000 individuals every year, and can lead to permanent vision impairment if left untreated.² The primary factor contributing to permanent vision loss post-RD is the death of PR cells, which has been demonstrated to occur predominantly via caspase-dependent apoptosis.^3–8 

Following RD, the retina mounts an innate immune response, marked by the upregulation of proinflammatory genes and cytokines, which further exacerbate tissue damage.^9,10 In the human vitreous, for example, from a panel of 32 analytes, expression of 28 analytes were significantly upregulated following RD.¹⁰ Analytes included interferons (IFNs)-α2 and -γ, monocyte chemoattractant protein-1 (MCP-1/CCL2), interferon γ-induced protein 10 kDa (IP-10/ CXCL10), growth-regulated oncogene (GRO/CXCL1), among others, highlighting the proinflammatory nature of this condition.¹⁰ Exacerbated inflammation contributes to the development of proliferative vitreoretinopathy (PVR), which has been shown to limit the success rates of reattachment surgeries.^9,11,12 Thus, strategies to limit inflammation and mitigate PR cell death post-RD are crucial. 

Preconditioning is a phenomenon where an exposure to a stressful stimulus initiates an adaptative response that mitigates the extent of injury induced by a subsequent stressor.¹³ In remote preconditioning, an injury to a tissue, such as hindlimb ischemia, triggers protection from a subsequent injury in a distal tissue, such as cerebral stroke.¹⁴ Although extensively studied in cardiovascular diseases,¹⁴ the retina similarly shows a robust adaptative plasticity, in which multiple systemic and local physiological and pharmacologic stimuli, such as brief ischemia, intravitreal lipopolysaccharide (LPS) injection, and hypoxia-inducible factor (HIF) stabilizers, have been shown to protect the retina against a subsequent injury.¹⁵ 

Although the retina has shown the ability to be preconditioned in multiple injury models,¹⁵ the potential of remote preconditioning to limit injury induced by RD and improve cell survival have not been explored. In this work, we hypothesized that detachment of one retina would remotely precondition the fellow retina, providing protection against cell death and inflammation following its detachment. We tested this hypothesis using our in vivo mouse model of experimental RD. 

Methods

Mouse RD Model

Adult (male/female) C57BL/6J mice of 8 to 10 weeks of age were procured from the Jackson Laboratory for this study. The mice were housed in a temperature-controlled environment with a 12-hour light-dark cycle. Mice were anesthetized and underwent RD as previously detailed.^10,16 In the control group, RDs were performed to the left eyes, and detached retinas and fellow-eye (right) retinas were preserved intact and harvested 1 day later, as shown in Figure 1. In the preconditioning group, the right eye underwent RD, followed by detachment of the left eye at 1, 3, 7, or 28 days later. Both retinas were harvested 1 day later. 

Figure 1.

 

Experimental design. Two groups were studied: the control group, which underwent a single RD on their left eyes; and the preconditioning group, which had their right eyes detached at 1, 3, 7, or 28 days prior to detachment of the left eye. Eyes were enucleated 1 day (24 hours) after detachment of the left eye and harvested for further processing in both groups.

Ethics Statement

Animal experiments were conducted in accordance with the guidelines established by the Institutional Animal Care and Use Committee (IACUC) of the University of Michigan and the Association for Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Visual Research. 

Sample Harvesting

Animals were euthanized, and retina samples were harvested and snap-frozen in liquid nitrogen, and stored at −80°C until analysis. RPE was isolated for RNA extraction using RNAprotect Cell Reagent (Qiagen, cat # 76526) based on a previously published methodology,¹⁷ which has been recently validated and detailed by our group.¹⁶ 

DNA Fragmentation

DNA fragmentation was evaluated using the Cell Death Detection ELISA kit (Roche, cat # 11774425001), according to a previously described methodology.¹⁸ 

Terminal Deoxynucleotidyl Transferase-Mediated dUTP Nick-End Labeling Assay

Apoptosis in retinal cryosections was assessed using the DeadEnd Fluorometric TUNEL System (Promega, cat # G3250) 3 days post-detachment of the left eyes. The assay was performed using a protocol previously detailed by our group.^19,20 Images were acquired utilizing the Leica STELLARIS 8 FALCON Confocal Microscope (Leica Microsystems, Wetzlar, Germany). 

RNA Extraction, cDNA Preparation, and qRT-PCR

Total RNA was extracted from the retinas and RPE using the RNeasy Plus Micro kit (Qiagen, cat # 74034), according to the supplier’s instructions. Complementary DNA (cDNA) was subsequently synthesized using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher, cat # 4368813), in accordance with the manufacturer’s protocol. The qRT-PCR was conducted using CFX384 Touch (Bio-Rad). Expressions of Ccl2,²¹ Ccl3,²² Cd14,²³ Cxcl10,²² Fasl, Ifnar1, Irf-7, Irf-9,²³ Mlkl, Nlrp3,²⁴ Oas1a, Stat1,²³ and Tlr4²⁵ were evaluated using Pum1²³ as a housekeeping gene, in 10 µL of a solution containing Fast SYBRTM Green Master Mix (Applied Biosystems, cat # 4385612). Information on primer pair sequences is provided in Supplementary Table S1. Additionally, expression of Ccl9, Cxcl1, Fas, Hmox, Il1b, Ifit1, Lcn2, and Tnfa were assessed using duplex TaqMan gene expression assays, with VIC-labeled Pum1 TaqMan assay as an internal control, in a solution containing Fast Advanced Master Mix (Thermo Fisher, cat # 4444554). Primer-probe assay information is provided in Supplementary Table S2. Gene expression levels were estimated using the ΔΔCt method, using Pum1 as the endogenous housekeeping gene.²⁶ 

Western Blot Analysis

For protein isolation, retinas were lysed by sonication on ice in 70 µL of a RIPA buffer solution (Sigma, cat # R0278) containing phosphatase (Thermo Scientific, cat # A32957) and protease (Thermo Scientific, cat # A32955) inhibitors, and then cleared by centrifugation at 16,000 × g for 10 minutes (4°C). Protein concentrations were assessed using the RC DC Protein Assay Kit (Bio-Rad, cat # 5000120). Protein samples were denatured in NuPAGE LDS Sample Buffer (Thermo Scientific, cat # NP0007) with Reducing Agent (Thermo Scientific, cat # NP0009), 20 µg of protein was separated on a 4% to 15% SDS-PAGE gel (Mini-PROTEAN TGX; Bio-Rad, cat # 4561086) and electrophoretically transferred to a polyvinylidene fluoride (PVDF) membrane (Bio-Rad, cat # 1620177). After blocking in a 5% bovine serum albumin (BSA; Sigma, cat # A7906) in Tris-buffered saline (TBS; Bio-Rad, cat # 1706435), the membranes were incubated overnight with the primary antibodies to caspase-3 (Novus, cat # 56113, 1:1000), LC3B (CST, cat #2775S, 1:1000), phospho-MLKL (CST, cat #37333S, 1:1000), RIPK1 (BD, cat # 610458, 1:1000), RIPK3 (Abgent, cat # AP7819B-EV, 1:1000), or β-actin (Sigma, cat # A5316, 1:3000) as loading control and then incubated for 2 hours with secondary antibodies, anti-rabbit (CST, cat # 70745, 1:2000), anti-goat (Jackson Immune, cat # 805-035-180, 1:7500), or anti-mouse (Invitrogen, cat # 32430, 1:1000). Signals were detected using SuperSignal West Pico PLUS Chemiluminescent Substrate (Thermo Scientific, cat # 34580) and a fluorescence imager (c500; Azure Biosystems). Band intensities were determined using ImageJ (version 1.53m). Uncropped blots are available in the Supplementary Materials. 

Statistical Analysis

Data in figures are presented as the mean value ± standard error of the mean (SEM). Statistical evaluations were performed using GraphPad Prism version 10 for Windows (GraphPad Software, San Diego, CA, USA, www.graphpad.com). 

Results

Retinal Detachment in One Eye Induces a Transitory Remote Preconditioning Effect

Our group and others have demonstrated that RD induces a significant and predominantly caspase-dependent apoptotic PR cell death as early as 8 hours after detachment, peaking approximately 1 to 3 days post RD and diminishing over time.^3–6 We hypothesized that this injury could produce a remote preconditioning in the contralateral eye. An experimental protocol to examine the potential for remote preconditioning due to RD of the contralateral eye is shown in Figure 1. The amount of DNA fragmentation in the potentially preconditioned retina is compared to that obtained in a control detached retina from a mouse with no prior RD of the contralateral eye. As a quantitative measure of retinal cell death, we used a cytoplasmic nucleosome capture assay¹⁸ reflecting retinal cell DNA fragmentation, a hallmark of apoptosis.²⁷ We first examined the time course of DNA fragmentation after RD with no preconditioning. Figure 2A illustrates approximately 5.2-fold elevated levels of DNA fragmentation in RD retinas, compared with attached retinas in their respective fellow eyes at 1 day post RD (P < 0.0001) and 3 days post RD (P < 0.0001). Importantly, Figure 2A does not include preconditioning. Instead, it presents the time-course of DNA fragmentation following RD in the detached eye compared to the intact, non-injured fellow eye. 

Figure 2.

 

Temporal course of DNA-fragmentation post-retinal detachment in C57BL/6 J mice and the effects of remote preconditioning between eyes. (A) Time-course of retinal cell death after retinal detachment (RD). DNA fragmentation levels in RD and fellow retinas were examined at 1, 3, and 7 days post retinal detachment. Evaluation of the impact of preconditioning times on DNA fragmentation (B) 1, (C) 3, (D) 7, and (E) 28 days. The gray bars represent samples from the control group, where one eye underwent detachment while the fellow eye was preserved; the blue bar represents samples from the detached eye of previously preconditioned mice. (F) Detection of apoptotic cell death in retinal sections using the TUNEL assay. Green indicates TUNEL-positive cells; and blue indicates DAPI. Scale bar = 100 µm. Bar charts depict the average values along with the standard error of the mean (SEM). Statistical evaluation was conducted using 1-way ANOVA with repeated measures, followed by Tukey’s post hoc examination. Significance levels were denoted as * for P < 0.05, ** P < 0.01; *** P < 0.001, and **** P < 0.0001. O.D., optical density; ONL, outer nuclear layer; INL, inner nuclear layer; RPE, retinal pigment epithelium.

Cell death waned by 7 days post RD (P = 0.3058). Thus, we selected 1 day post RD as the time to evaluate preconditioning protection in subsequent experiments. 

Previous studies of remote preconditioning established that transient distal ischemia to the hindlimb can produce a preconditioning effect in the brain that lasts for approximately 3 days.²⁸ However, in our mouse model, retinal detachment is not transient and therefore the resulting ischemia is continuous. We therefore examined if a detachment in one eye at several times (1, 3, 7, and 28 days) before RD of the experimental eye would confer protection against RD-induced cell death in the contralateral eye. We found that previous RD in the contralateral eye 1 day prior to RD in the experimental eye induced a significant protection to DNA fragmentation in the other eye (P = 0.0005; Fig. 2B). The level of DNA fragmentation in the preconditioned retina was approximately 45% without preconditioning and was not significantly elevated relative to the undetached retina in the control mouse (P = 0.2619). A similar 47% protection by preconditioning (P = 0.0005) was obtained when the contralateral retina was detached 3 days prior to detachment of the experimental retina (Fig. 2C). In contrast, at 7 and 28 days after the contralateral RD, the preconditioning effect on DNA fragmentation was lost (Figs. 2D, 2E). Short-term retinal preservation through the reduction of apoptosis was assessed in retinal cryosections using the TUNEL assay and corroborated the DNA fragmentation assay results, demonstrating a statistically significant reduction in preconditioned samples (P = 0.0244; Fig. 2F). Thus, RD in one eye promotes a transient protection against retinal cell death in the contralateral eye, which waned by 7 days following preconditioning. 

Preconditioning Provides Retinal Cell Death Protection Following RD

To corroborate apoptotic cell death indicated by the DNA fragmentation assay, we assessed protein levels of caspase-3, which is produced by proteolytic cleavage of pre-caspase-3 during apoptosis.²⁹ As shown in Figure 3A, cleavage of pro-caspase-3 was abrogated in the preconditioned retina at 1 day after RD of the contralateral eye (P = 0.0448), indicating suppression of apoptotic cell death by preconditioning. To assess if preconditioning skews death toward other modes of cell death that do not involve DNA fragmentation, levels of RIPK1, RIPK3, and phosphorylated MLKL proteins were assessed as indicators of necroptosis pathway activation. In addition, conversion of LC3-I to LC3-II was evaluated to determine if autophagy was altered. No significant differences were observed between the detached eye from the control group and the preconditioned detached eye. 

Figure 3.

 

Preconditioning provides protection from cell death following retinal detachment (RD). Representative immunoblots depicting the expression levels of (A) pro-caspase-3 and cleaved-caspase-3 (B) necroptotic pathway markers RIPK1, RIPK3, and phospho-MLKL; (C) autophagy marker LC3B; and, normalized to β-actin in C57BL/6 J mice, 1 day post retinal detachment. The gray bars represent samples from the control group, where one eye underwent detachment while the fellow eye was preserved; and the blue bars represent samples from the detached eye of previously preconditioned mice. Blots are representative of n = 4 (A) and n = 5 (B, C) independent animals in each condition. Bar charts depict the average values along with the standard error of the mean (SEM). Statistical evaluation was conducted using 1-way ANOVA with repeated measures, followed by Tukey’s post hoc examination. Significance levels were denoted as * for P < 0.05. ns, nonsignificant.

Preconditioning Modifies mRNA Expression Toward a Cytoprotective Profile

To evaluate the effect of preconditioning on key markers of stress and inflammation in pathways previously implicated in RD^10,30 or shown to associate with preconditioning,³¹ mRNA levels were assessed by qRT-PCR. The results (Figs. 4A–F) show that at 1 day post RD the detached retinas subjected to 1 day of preconditioning exhibited significantly less expression of mRNAs encoding the proinflammatory cytokines Il1b (fold change [FC] = 0.41, P = 0.072), Tnfa (FC = 0.30, P = 0.0066), Cxcl1 (FC = 0.40, P = 0.0459), and Cxcl10 (FC = 0.59, P = 0.0099), as did the control detached retinas without preconditioning. Although trending down, the effect of preconditioning on Ccl2 (FC = 0.83, P = 0.8147) and Ccl3 (FC = 0.63, P = 0.113) mRNA upregulation was not significant. 

Figure 4.

 

Preconditioning alters the retinal gene expression response at 1 day post retinal detachment. Relative mRNA expression of (A) Il1b; (B) Tnfa; (C) Cxcl1; (D) Cxcl10; (E) Ccl2; (F) Ccl3; (G) Ifnar; (H) Stat1; (I) Irf7; (J) Irf9; (K) Oas1; (L) Ifit1; (M) Cd14; (N) Nlrp3; (O) Tlr4; (P) Hmox; (Q) Fas; (R) Fasl; and (S) Mlkl; in C57BL/6J mice, 1 day post retinal detachment, as compared to the 1 day preconditioning group. The gray bars represent samples from the control group, where one eye underwent detachment while the fellow eye was preserved; the blue bars represent samples from the detached eye of previously preconditioned mice. Pum-1 was used as a housekeeping control gene for normalization. Bar charts depict the average values along with the standard error of the mean (SEM). Statistical evaluation was conducted using 1-way ANOVA with repeated measures, followed by Tukey’s post hoc examination. Significance levels were denoted as *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001. ns, nonsignificant.

We also evaluated mRNAs levels of interferon signaling genes (ISGs) due to their well-described role of in preconditioning.^31,32 We observed a significant downregulation of mRNAs that code for the interferon alpha receptor Ifnar (FC = 0.44, P = 0.0005), Stat1 (FC = 0.50, P = 0.0007), as well as Irf9 (FC = 0.49, P = 0.0117). Conversely, we observed increased mRNA levels of Irf7 (FC = 2.24, P = 0.0106), Oas1 (FC = 2.48, P = 0.0185), and a trend for increased expression of Ifit1 mRNA levels (FC = 1.57, P = 0.074) with preconditioning (Figs. 4G–L). We also evaluated additional innate immune signaling pathways that have been shown to associate with preconditioning.³¹ We observed decreased expression of mRNAs coding for the TLR-modulating co-receptor Cd14 (FC = 0.43, P < 0.0001) and Nlrp3 (FC = 0.56, P = 0.0164), whereas no significant effect of precondition was observed for Tlr4 (FC = 0.86, P = 0.2523; Figs. 4M–O). Thus, preconditioning diminished the upregulation of expression of many, but not all cytokines, whereas increasing the upregulation of several ISGs after RD. 

Additionally, we sought to evaluate the effect of preconditioning on genes associated with cell stress and retinal cell death. We observed that preconditioning significantly diminished upregulation of Hmox (FC = 0.37, P = 0.0242), reduced the expression of Fas receptor mRNA, Fas (FC = 0.36, P = 0.0223), produced a nearly significant reduction in Fas ligand mRNA upregulation, Fasl (FC = 0.40, P = 0.0626), but did not affect Mlkl (FC = 1.01, P = 0.999) mRNA expression (Figs. 4P–S). Altogether, results indicate a cytoprotective shift in mRNA expression profile upon preconditioning. 

Based on the time course of the protective effect of preconditioning on DNA fragmentation, where the effect waned after 7 days, we evaluated if the effects on mRNA expression after RD exhibited a similar time course after detachment. Figure 5 shows the transient nature of the cytoprotective profile in mRNA expression with preconditioning. The upregulation of the mRNAs encoding cytokines Ccl2, Ccl3, and Cxcl10, as well as Fas were significantly reduced at 1 day after RD of the contralateral retina, but the preconditioning effect was lost at 7 and 28 days post RD (see Figs. 5A–F). In contrast, preconditioning had significant inhibitory effects on upregulation of Hmox after RD at both 1 and 7 days after detachment of the contralateral eye. Surprisingly, preconditioning increased the upregulation of Ifit1 mRNA after RD to increase even at 28 days following detachment of the contralateral eye. Thus, the effects of preconditioning on inflammatory gene expression exhibited a similar time course to that of protection from cell death. In contrast, upregulation of the ISG Ifit1 was increased at even 4 weeks after preconditioning, suggesting that sustained detachment of the contralateral retina is continuously impacting the experimental retina in some way. 

Figure 5.

 

Preconditioning modifies retinal mRNA expression towards a transient cytoprotective profile that wanes by 7 days. Relative mRNA expression of (A) Fas; (B) Ccl2; (C) Cxcl10; (D) C; (E) Ccl2; (F) Ccl3; in C57BL/6J mice, 1 day post retinal detachment, as compared to the 3, 7, and 28 day preconditioning group. The gray bars represent samples from the control group, where one eye underwent detachment while the fellow eye was preserved; and the blue bars represent samples from the detached eye of previously preconditioned mice. Pum-1 was used as a housekeeping control gene for normalization. Bar charts depict the average values along with the standard error of the mean (SEM). Statistical evaluation was conducted using 1-way ANOVA with repeated measures, followed by Tukey’s post hoc examination. Significance levels were denoted as *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001.

We recently demonstrated that RD causes a profound transient transcriptomic response in the RPE, resembling an innate immune defense response.¹⁶ Therefore, in addition to studying the retinal response to preconditioning, we evaluated the effect of preconditioning on RPE mRNA expression. Genes were selected based on our recent evaluation of transcriptome modifications in the RPE following RD.¹⁶ As depicted in Figure 6, we found no significant effect of preconditioning on RPE upregulation of mRNAs encoding cytokines (Ccl2 and Ccl9) or ISG (Irf7, Oas1, and Ifit1), or of an oxidative stress marker (Hmox). We did, however, find a significant decrease in the upregulation of Mlkl mRNA in the preconditioned RPE (FC = 0.60, P = 0.0175), whereas upregulation of Lcn2 mRNA was greatly increased by preconditioning (FC = 3.44, P = 0.0013). 

Figure 6.

 

Preconditioning effect on the RPE gene expression response at 1 day post retinal detachment. Relative mRNA expression of (A) Ccl2; (B) Ccl9; (C) Irf7; (D) Oas1; (E) Ifit1; (F) Hmox; (G) Mlkl; (H) Lcn2; in C57BL/6J mice, 1 day post retinal detachment, as compared to the 1 day preconditioning group. The gray bars represent samples from the control group, where one eye underwent detachment while the fellow eye was preserved; and the orange bars represent retinal pigment epithelium (RPE) samples from the detached eye of previously preconditioned mice. Pum-1 was used as a housekeeping control gene for normalization. Bar charts depict the average values along with the standard error of the mean (SEM). Statistical evaluation was conducted using 1-way ANOVA with repeated measures, followed by Tukey’s post hoc examination. Significance levels were denoted as *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001. ns, nonsignificant.

Discussion

In this study, we demonstrate for the first time that contralateral RD produces a transient protective phenotype lasting from 1 to 3 days. This protective effect prevented post-detachment retinal DNA fragmentation and caspase 3 cleavage compared to control RD retinas with intact fellow-eye retinas. Our findings build upon previous investigations where our group demonstrated increased mRNA expression of Fas and Fasl subsequent to RD,⁷ with inhibition of Fas signaling mitigating RD-induced PR apoptosis.³³ Here, we extend these observations by demonstrating that contralateral RD at 1 and 3 days prior to RD can suppress mRNA expression of Fas and Fasl in mouse retinas. This reduction correlates with decreased retinal DNA fragmentation, which is a characteristic feature of apoptotic cell death.²⁷ Previous studies had demonstrated that inhibition of apoptosis could promote a significant increase in necroptotic PR death.³⁴ Thus, to confirm if apoptosis inhibition would not skew cell death toward a necroptotic mode of cell death, Mlkl mRNA expression was evaluated by qRT-PCR and levels of RIPK1, RIPK3, and phosphorylated MLKL proteins, as indicators of necroptotic pathway activation, were assessed by Western blotting. The results indicated that preconditioning prevents cell apoptosis, without promoting necroptosis. Additionally, we demonstrated that autophagy was also not significantly different between the study groups, indicating an overall prevention of retinal cell death by preconditioning. 

The window of time that retinal remote preconditioning produces a protective effect is transient, with protection at 1 and 3 days after detachment of the contralateral retina that wanes at 7 days. This matches the duration of protection seen in other paradigms of remote ischemic conditions (RICs), which lasts for up to 3 days after transient ischemia of a distal tissue.²⁸ Because the subretinal hyaluronic acid injection model of RD produces a continued ischemic insult rather than a transient one, it seemed possible that the window of protection after RD would be extended. The protection coincides very closely with the time course of DNA fragmentation increase observed in the detached retina (see Fig. 2A). This suggests that retinal cell death may be the initiator of the preconditioning signal that impinges upon the fellow retina. This hypothesis requires testing. 

Furthermore, our study reveals a significant suppression of upregulation of mRNAs coding for proinflammatory cytokines (Il1b, Tnfa, Cxcl1, and Cxcl10), suggesting a dampening of the inflammatory response to RD, which can also contribute to reduction of tissue damage and retinal cell death.³⁵ It is well documented that although multiple preconditioning stimuli can promote a remote protective response which suppresses inflammation and cell death, most protective paradigms converge on the IFN-signaling response.^31,32 Here, we observe an upregulation of ISG (Irf7, Oas1, and Ifit1) with 1 day of preconditioning. This upregulation suggests activation of the interferon response pathway, which could play a role in priming the cells for survival and defense mechanisms.^32,36 It has been previously described, for example, that IRF7 is required to promote tolerance to ischemic brain injury³² and IFIT1 is required for preconditioning in stroke³⁷; additionally, differential expression of Oas1 has been reported in retinal ischemia.³⁸ Given that ISGs have been implicated in ischemic preconditioning, the observed increased response may be relevant to the fellow eye preconditioning mechanism. 

Our study revealed a discrepancy between the reduction mRNAs that code for proinflammatory cytokines in the retina and the lack of a similar pattern in RPE. This distinct response may be in part due to the blood-retinal barrier function of the RPE layer, which mounts an innate immune response to locally protect the choriocapillaris and retina.^16,39 Notably, preconditioning causes the upregulation of LCN2 mRNA in the RPE to be significantly exaggerated after RD. Lipocalin 2 is known for its anti-inflammatory and barrier protective roles in other barrier tissues, such as the intestines.^40,41 Therefore, a role for Lipocalin-2 expressed in the RPE in protection of the retina after RD should be considered. Additionally, our findings revealed a significant decrease in the upregulation of Mlkl mRNA in the RPE caused by preconditioning. MLKL is involved in the necroptotic pathway and implicated in RPE cell death under stress conditions.⁴² Future studies will explore alternative conditioning stimuli, elucidate underlying mechanisms, and assess post-conditioning strategies⁴³ in the context of RD. 

In conclusion, we demonstrate for the first time that RD in one eye produces a transient protective phenotype that protects the fellow eye against retinal cell death and inflammatory gene expression following a subsequent RD. Future study will help dissect the mechanisms involved in this effect and evaluate if these protective pathways can be leveraged therapeutically for preconditioning or postconditioning following RD. 

Acknowledgments

Supported by the NIH R01EY020823 to D.N.Z. and S.F.A. S.F.A. additionally received support from NIH R01EY029349, NIH R01EY031961, and Research to Prevent Blindness (RPB). B.M.S. was supported by Training Grant T32AR07080 from the National Institute of Arthritis and Musculoskeletal and Skin Diseases. The research was further supported by an NEI Vision Core Research Grant NIH P30EY007003. The authors are solely responsible for the content, and it does not necessarily reflect the official views of the National Institutes of Health. 

Disclosure: B.M. Scavuzzi, None; S. Shanmugam, None; M. Yang, None; J. Yao, None; H. Hager, None; B. Kaur, None; L. Jia, None; S.F. Abcouwer, None; D.N. Zacks, None