CREG Protects Retinal Ganglion Cells loss and Retinal Function Impairment Against ischemia-reperfusion Injury in mice via Akt Signaling Pathway

The irreversible death of retinal ganglion cells (RGCs) plays an important role in the pathogenesis of glaucoma. Cellular repressor of E1A-stimulated genes (CREG), a secreted glycoprotein involved in cellular proliferation and differentiation, has been shown to protect against myocardial and renal ischemia‐reperfusion damage. However, the role of CREG in retinal ischemia-reperfusion injury (RIRI) remains unknown. In this study, we aimed to explore the effect of CREG on RGCs apoptosis after RIRI. We used male C57BL/6J mice to establish the RIRI model. Recombinant CREG was injected at 1 day before RIRI. The expression and distribution of CREG were examined by immunofluorescence staining and western blotting. RGCs survival was assessed by immunofluorescence staining of flat-mounted retinas. Retinal apoptosis was measured by the staining of TdT-mediated dUTP nick-end labeling and cleaved caspase-3. Electroretinogram (ERG) analysis and optomotor response were conducted to evaluate retinal function and visual acuity. The expressions of Akt, phospho-Akt (p-Akt), Bax, and Bcl-2 were analyzed by western blotting to determine the signaling pathways of CREG. We found that CREG expression was decreased after RIRI, and intravitreal injection of CREG attenuated RGCs loss and retinal apoptosis. Besides, the amplitudes of a-wave, b-wave, and photopic negative response (PhNR) in ERG, as well as visual function, were significantly restored after treatment with CERG. Furthermore, intravitreal injection of CREG upregulated p-Akt and Bcl-2 expression and downregulated Bax expression. Our results demonstrated that CREG protected RGCs from RIRI and alleviated retinal apoptosis by activating Akt signaling. In addition, CREG also improved retinal function and visual acuity.


Introduction
Glaucoma is the most common cause of irreversible blindness and visual impairment globally, afflicting about 3-5% of people over the age of 40. As the rapid growth of the world's geriatric population, the number of people with glaucoma is expected to double by 2040 [1,2]. It's a progressive optic neuropathy characterized by the degeneration and loss of retinal ganglion cells (RGCs) located in the inner retina [3]. RGCs are the only output neurons in the central nervous system that transmit visual information from the retina to the brain, and their axons, along with the retinal blood vessels, cross the lamina cribrosa to form the optic nerve [4]. Although glaucoma is a multifactorial disease process, many studies have demonstrated that elevated Guojing Lu 15827088571@163.com intraocular pressure (IOP) levels represent the main risk factor for glaucoma [5]. Under high IOP, the pressure gradient of the lamina cribrosa increases, leading to a decrease in blood supply and impaired axoplasmic transport of optic nerve fibers. Due to the lack of neurotrophic support, RGCs are more likely to occur irreversible loss [6,7]. A proportion of patients with glaucoma show progressive RGCs death and visual field defects despite adequately controlled IOP in the physiological range with drugs or surgical procedures. The death of RGCs results from multiple mechanisms, including apoptosis, oxidative stress, inflammation, autophagy, ischemia, and hypoxia. Therefore, protecting RGCs from damage is a potential target for new neuroprotective and IOP-independent therapies to halt or slow glaucoma progression [8].
Cellular repressor of E1A-stimulated genes (CREG), originally identified in a yeast two-hybrid screen, acts as a transcriptional repressor and suppresses both transcriptional activation and cellular transformation caused by the adenovirus E1A oncoprotein in transfection assays [9]. Lately, CREG was found to be a secreted glycoprotein located in the perinuclear region of the cell that regulates the process of cell growth, migration, and differentiation [10]. Previous studies have suggested that CREG overexpression inhibited vascular smooth muscle cell apoptosis, whereas CREG knockdown significantly increased apoptosis. Similarly, Liu et al. [11] indicated that the upregulation of CREG reduced human umbilical vein endothelial cells apoptosis induced by high glucose and high palmitate. Besides, a recent study found that intravitreal administration of recombinant CREG prevented renal cell apoptosis and protected renal tissue against kidney injury and fibrosis with a high-salt diet [12]. Most recently, accumulating evidence revealed a critical role of CREG in ischemia-reperfusion injury [13]. Several findings have demonstrated that CREG could increase bone marrow-derived mesenchymal stem cell survival and postinfarct perfusion of myocardial infarction and ischemic stroke [14,15]. A new research reported that exogenous CREG inhibited cardiomyocyte death, mitigated infarcted areas, and improved cardiac function by modulation of lysosomal autophagy in myocardial ischemia-reperfusion injury [16]. Additionally, Yang et al. [13] have demonstrated that CREG overexpression alleviated hepatocytes death and inflammatory responses, whereas CREG deletion accelerated hepatic damage during the progression of ischemia-reperfusion injury. The above evidence indicated the potential protective effect of CREG against retinal ischemia-reperfusion injury (RIRI). RIRI is an important pathophysiological basis of glaucoma, triggered by transient intraocular hypertension, which induces neuronal damage and impaired vision. Previous studies have proved a high sensitivity of RGCs to an ischemic injury. The RIRI model simulates the pathogenesis of acute glaucoma and is a classic animal model for studying RGCs damage after ischemic insult [6,17]. In this study, we investigated whether CREG could inhibit RIRI-stimulated RGCs apoptosis and explore the related mechanisms.

Animals
Wild-type C57BL/6J mice (male, 8-week-old, weight: 20-22 g) provided by the Laboratory Animal Center of Wuhan University, were bred in an air-conditioned barrier system with a 12 h light-dark cycle and free access to food and water. All animal procedures were approved by the requirements of the National Institutes of Health Guide for the Care and Use of Laboratory Animals and conducted according to the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Visual Research.

The Model of RIRI
The RIRI model was replicated following the method described in previous literature [18]. The mice were placed on a temperature-controlled heating blanket during the surgical procedure to prevent hypothermia caused by anesthesia. Firstly, the mice were anesthetized by an intraperitoneal injection of 1% pentobarbital (35 mg/kg). The corneal surface anesthesia was achieved using 0.5% tetracaine hydrochloride, and the pupils were dilated using 1% tropicamide. The anterior chamber of the right eye was then cannulated with a 30-gauge needle connected to a saline infusion system, and the height was adjusted to sustain an IOP of 80 mmHg for 1 h. During the retinal ischemia, iris whitening and the disappearance of retinal red flex were observed. The left eye was treated with the same procedure but normal IOP as the control. After removing the infusion needle, levofloxacin hydrochloride eye gel was applied to prevent infection. Mice were euthanized 0,1,3, and 7 days after RIRI model replication for further study.

Intravitreal Injection of CREG
The accurate intravitreal injection was achieved according to the method of Yang et al. [19]. The injection needle was prepared by a dual-stage glass micropipette puller (Narishige, Japan) and installed in the programmable nanoliter injector (Nanoject III, Drummond, USA). After calibrating the needle and setting up the program, 1 µl of CREG protein or phosphate buffered saline (PBS) was injected into the posterior vitreous chamber by inserting the micropipette 1-2 mm behind the corneoscleral limbus. To prevent drug leakage, leave the needle in place for 1 min and remove it slowly when finishing the injection. After the operation, levofloxacin hydrochloride eye gel was applied to the binocular conjunctival sac. Recombinant mouse CREG protein (R&D Systems, MN, USA) was dissolved in PBS. The RIRI-CREG group and control-CREG group were injected with CREG protein at 1 day before RIRI, while the RIRI group and control group received PBS as above.

RGCs Labeling and Quantification
Mice eyes were removed and fixed in 4% PFA for 1 h. The retinas were carefully isolated, blocked overnight at 4 ℃ with 5% BSAT, then incubated with a Brn3a antibody (Synaptic Systems, 411001, 1:1000) at 4 ℃ for 48 h. After rinsing in PBS, the retinas were exposed in Alexa Fluor 594 antibody (Jackson ImmonoResearch Laboratories, #711-585-152, 1:500) at 4 ℃ overnight to label RGCs. The petalshaped retinal flat-mounts were made by scissoring the retina with four radial incisions, then observed with a fluorescent microscope (IX51, Olympus, Tokyo, Japan). RGCs were counted in two microscopic fields per retinal petal at 0.8-1.2 mm and 1.8-2.2 mm from the optic nerve head. Surviving RGCs were quantified by ImageJ software (National Institutes of Health, USA).

TdT-mediated dUTP nick-end Labeling (TUNEL) Staining Analysis
Retinal apoptosis was assessed using a one-step TUNEL apoptosis assay kit (Beyotime Biotechnology, Shanghai, China) by labeling the 3' -end of the fragmented DNA of the apoptotic cells [20]. The above-mentioned retina slices were permeabilized with 0.5% Triton X-100 in PBS at room temperature for 5 min, and incubated with TUNEL working solution in darkness at 37 ℃ for 1 h. Then, the slices were rinsed with PBS and stained with DAPI (Servicebio, Wuhan, China). Finally, the TUNEL-positive cells with green fluorescence were observed with a fluorescent microscope (BX53, Olympus, Tokyo, Japan). For quantitation of apoptotic cells, six fields located at distances of 1/6, 1/2, and 5/6 radii from the optic disc were calculated using ImageJ software.

Electroretinogram (ERG) Analysis
After dark adaption overnight, mice were anesthetized intraperitoneally, and their pupils were dilated with 1% tropicamide in the darkroom. Signals were measured using a ground electrode in the tail, a reference electrode needle placed subcutaneously between the ears, and two gold electrodes on the corneal surface. For dark-adapted ERG, white flashes of 0.001, 0.003, 0.010, 0.030, 0.100, 0.300, 1.000, 3.000, 10.000 cd.s/m 2 with 1 min intervals in between each flash were delivered using the RetiMINER-C visual electrophysiological system (IRC Medical Equipment Co., Ltd, Chongqing, China). Measurements of a-and b-wave amplitudes at 3.000 cd.s/m 2 were taken for analysis. For photopic negative response (PhNR), mice were light adapted for 10 min with a green background (25.000 cd.s/m 2 ), and responses to flash stimulus of 10.000 cd.s/m 2 were recorded. The amplitude of PhNR was the first negative deflection following the b-wave relative to the baseline.

Optomotor Response (OMR)
As described previously, a MATLAB software (MathWorks, USA)-based virtual optomotor system was used to evaluate OMR in mice [21]. After dark adaptation, mice were placed individually on an upright cylinder allowed to move freely, and they were surrounded by four LCD computer monitors displaying the rotating vertical gratings. When a grating perceptible to mice was displayed on the monitors, mice would halt the movements and rotated their heads to follow the grating. The head movements of mice were captured by a camera placed directly above the platform. The visual acuity was quantified as the maximum spatial frequency of the grating (cyc/deg).
analyses were performed using one-way analysis of variance (ANOVA) followed by Tukey's post hoc test for multiple comparisons. A value of P < 0.05 was considered statistically significant.

The Expression of CREG was Decreased After RIRI
The expression and distribution of CREG in retinal tissues were observed 3 and 7 days after RIRI. Immunofluorescence staining showed that CREG was diffusely distributed in the ganglion cell layer (GCL), the inner plexiform layer (IPL) and the inner nuclear layer (INL) (Fig. 1A). Compared with the control group, CREG immunoreactivity of the retina, especially GCL, was significantly decreased after ischemic injury (Fig. 1A). Meanwhile, western blotting analysis confirmed that the protein levels of CREG were downregulated in the RIRI group (P < 0.05, Fig. 1B). These findings identified the potential involvement of CREG in RIRI. Therefore, we injected CREG into the vitreous body of mice and investigated its effect on the retina.

Intravitreal Injection of CREG Reduced RGCs loss After RIRI
To assess the ability of CREG to reduce RGCs loss, immunofluorescence staining of flat-mounted retinas was conducted to quantify the surviving RGCs 7 days after RIRI ( Fig. 2A-D). The statistical analysis indicated the density

Statistical Analysis
All data were analyzed using GraphPad Prism 9.0 (Graph-Pad Software, San Diego, USA). Data were presented as the mean ± standard error of mean (SEM), and statistical

CREG Attenuated Apoptosis and Protected Against RIRI via Akt Pathway
To explore the protective mechanisms of CREG against RIRI, we assessed the levels of the apoptotic proteins Akt, phospho-Akt (p-Akt), Bax, and Bcl-2. Previous studies have demonstrated that ischemia-induced apoptosis is closely related to Akt pathway [22]. As demonstrated by western blot analysis, upregulation of Bax and downregulation of p-Akt persisted from 3 to 7 days after RIRI (P < 0.05, Fig. 4C and E), which was significantly suppressed by CREG treatment on days 3 post-ischemia damage (P < 0.05, Fig. 4D and F). In addition, there was no significant change in the relative expression level of Akt and Bcl-2 after RIRI ( Fig. 4A and G). Notably, a significant increase in Bcl-2 expression was detected in the RIRI-CREG group (P < 0.05, Fig. 4H). These data suggested that CREG mitigated retinal ischemic-induced apoptosis via Akt pathway, together with downstream apoptotic regulators.

Intravitreal Injection of CREG Facilitated Retinal Function and Visual Acuity
Next, we performed ERG to monitor the effect of CREG on impaired retinal function, and the waves response to 3 cd.s/m 2 were selected ( Fig. 5A and D). An evident of surviving RGCs was significantly decreased by 41% in the RIRI group (P < 0.0001, Fig. 2E), whereas pretreatment with CREG markedly improved RGCs survival (P < 0.01). Moreover, intravitreal injection of CREG in the control group had no effect on the density of RGCs. All these data suggested that CREG ameliorated RGCs loss induced by retinal damage.

Intravitreal Injection of CREG Inhibited Retinal Apoptosis After RIRI
We examined the protective effect of CREG against apoptosis on the retina 1 day after RIRI using TUNEL staining. In comparison with the control mice, a great quantity of TUNEL-positive cells was observed after ischemic insult and mainly located in the GCL, INL, and outer nuclear layer (ONL) (Fig. 3A). However, intravitreal injection of CREG significantly diminished retinal positive cells after RIRI (P<0.001, Fig. 3B), especially in the GCL (P<0.05, Fig. 3C). Immunofluorescent staining further corroborated these data, and strong fluorescent signals of cleaved caspase-3 were obtained in the GCL, INL, OPL, and ONL after retinal damage. When mice were pretreated with CREG, the immunoreactivity of cleaved caspase-3 was remarkably decreased (Fig. 3D). Overall, these results suggested that CREG could effectively inhibit ischemia-induced apoptosis. (n = 4 to 6, **P < 0.01, ****P < 0.0001). All data are expressed as the mean ± SEM acuity, as measured by OMR, which was aligned with those obtained from ERG (Fig. 5I). The statistical analysis of mice movements showed that the visual acuity was reduced after RIRI compared to the normal mice (P < 0.0001, Fig. 5J). In contrast, intravitreal injection of CREG significantly promoted the recovery of visual acuity (P < 0.05). Notably, there was no significant difference in retinal function and visual acuity in the control-CREG group compared to the normal eyes. These data suggested that pretreatment with reduction in the a-and b-wave amplitudes was observed on days 3 and 7 after ischemic injury (P < 0.0001, Fig. 5B and C). However, the RIRI-induced decreased amplitudes were redressed in mice with CREG injection (P < 0.05, Fig. 5F and G). In addition, the amplitude of PhNR was used to evaluate RGCs function and our results showed a significant decrease in the RIRI group compared with the control group (P < 0.0001, Fig. 5E and H). However, CREG treatment restored the amplitude of PhNR (P < 0.01). Moreover, we further explored the role of CREG in improving visual Fig. 3 Intravitreal injection of CREG attenuated of retinal apoptosis after RIRI. (A) Immunofluorescent staining of retinal slices was stained with TUNEL (green) and DAPI (blue) in the control group, the control + CREG group, the 1d group and the 1d + CREG group. (B) Quantitation of TUNEL-labeled cells per specific area of retinal slices in each group (n = 3, ***P < 0.001, ****P < 0.0001). (C) Quantitation of TUNEL-labeled cells in the GCL per specific area of retinal slices in each group (n = 3, *P < 0.05, ***P < 0.001). (D) Immunofluorescent staining of retinal slices was stained with cleaved caspase-3 (green) and DAPI (blue) in the control group, the control + CREG group, the 3d group and the 3d + CREG group. All data are expressed as the mean ± SEM them more susceptible to an ischemic insult [24]. Due to the lack of retinal blood flow, the oxygenation capacity and supply of nutrients are reduced, and then breakdown products accumulate, triggering oxidative stress, inflammation, and retinal damage that ultimately ends with RGCs death [6,25]. In the present study, we found that CREG was diffusely expressed in the GCL, IPL, INL, and OPL, and downregulated after RIRI, especially in RGCs, which indicated that decreased expression of CREG was associated with retinal dysfunction and CREG may play a crucial role in ischemic injury. To investigate whether CREG possessed protective effects on the retina in RIRI, we injected CREG into the vitreous chamber of mice with RIRI and observed that the density of remaining RGCs was significantly increased compared to the RIRI group. Our data demonstrated that CREG alleviated retinal dysfunction and visual impairment after RIRI.

Discussion
Retinal ischemia is a common underlying pathogenesis involved in the progress of ophthalmic diseases, including glaucoma, age-related macular degeneration, diabetic retinopathy, and central retinal artery occlusion, which may result in visual impairment and even blindness [23]. The RIRI model, a pressure-induced mode model, is defined as transient retinal ischemia caused by sharply increased IOP, which simulates the ischemic insult occurs in glaucoma. RGCs, being highly metabolic for their specific neuronal function, are sensitive to oxygen deficiency, thus making  D, F, H) The retinal expression of Akt, p-Akt, Bax and Bcl-2 with CREG treatment was detected by western blotting (n = 3, *P < 0.05). All data are expressed as the mean ± SEM We found that intravitreal injection of CREG inhibited retinal cell apoptosis. Intrinsic and extrinsic stress signals such as hypoxia, ischemia, and reperfusion in RIRI induce mitochondrial injury. RGCs require higher energy to maintain metabolic balance, particularly vulnerable to mitochondrial dysfunction resulting from stress insults [28]. Damaged mitochondria can be fixed through a complex mechanism that regulates the synthesis of mitochondrial DNA (mtDNA), CREG had effective protection against RGCs loss after ischemia-induced damage.
It has been well established that the death of RGCs occurs by apoptosis in clinical glaucoma and animal models, which can be triggered by a variety of molecular pathways [26]. Quantification of TUNEL-positive cells further substantiated that the apoptosis of RGCs accounted for the damage to the retina, consistent with current literature [27].  C) The quantification of the a-and b-wave amplitudes after RIRI (n = 10, ***P < 0.001 ****P < 0.0001). (D) Representative waveforms of ERG at 3.000 cd.s/m 2 in the control group, the control + CREG group, the 3d group and the 3d + CREG group. (E) Representative waveforms of PhNR at 10.000 cd.s/m 2 in the control group, the control + CREG group, the 3d group and the 3d + CREG group. (F,  G, H) The quantification of the a-wave, b-wave and PhNR amplitudes with CREG treatment (n = 10, *P < 0.05, **P < 0.01, ****P < 0.0001). (I) OMR was detected by the virtual optomotor system. (J) The quantification of visual acuity in the control group, the control + CREG group, the 7d group and the 7d + CREG group. (n = 6, *P < 0.05, ****P < 0.0001). All data are expressed as the mean ± SEM a significant recovery of the a-and b-wave amplitudes with intravitreal CREG. Notably, a recent report [41] elucidated that CREG prevented retinal photoreceptor apoptosis from light exposure by activating PI3K/Akt signaling pathway and inhibiting p38/JNK signaling pathway. Furthermore, the amplitudes of PhNR were significantly lessened following RIRI, thought to reflect the damage to the function of RGCs. Besides, current evidence suggested that the reduction in PhNR amplitudes was positively correlated with the loss of RGCs, associated with the damage of RGCs electrophysiological function [27,42]. The PhNR is believed to originate from the inner retina and be related to the electrophysiological function of RGCs, representing the spiking potentials of RGCs activity [43]. Interestingly, CREG was able to partially correct the loss in amplitude. As described above, CREG exerted a protective effect on retinal electrophysiological function, especially RGCs and photoreceptors, for which the specific underlying mechanism needs to be further investigated.
Previous studies have emphasized the relationship between RGCs death and visual pathways impairment during glaucoma occurrence and development [44]. Therefore, we further used OMR to assess the visual function in RIRI models of mice. OMR, a rapid and non-invasive test widely used to measure visual acuity, is strongly reliant on functioning RGCs which dominate the response to the movement of visual stimulus [45]. Feola et al [46] found that the visual acuity of the experimental glaucoma model was decreased, which is consistent with the present study. Furthermore, our results demostrated a remarkably increased visual acuity of ischemia-induced mice after treatment with CERG, indicating that CREG contributed to the amelioration of visual function in mice after RIRI.
In summary, our study provided evidence that CREG protected RGCs from RIRI and alleviated retinal apoptosis by activating Akt signaling and inhibiting mitochondrial apoptotic pathways. In addition, CREG also improved retinal function and visual acuity to some extent. This study is the first to reveal the neuroprotective potential to prevent RGCs death and improve vision preservation during RIRI. Although further exploration is essential, our present findings may provide a promising therapeutic approach for application in ophthalmic studies of glaucoma.
proteins, and lipids to restore mitochondrial function [29]. Mitochondrial membrane permeabilization is enhanced when mtDNA damage becomes excessive, often followed by mitochondrial dysfunction. Ultimately, these alterations promote the progressive death of RGCs through apoptosis [30]. We analyzed several critical proteins in the mitochondrial apoptosis pathway to further investigate the underlying protective mechanisms of CREG. Akt, a member of serine/ threonine kinase family, acts as a central regulator of downstream molecules involved in cell survival and proliferation when activated by phosphorylation [22]. A growing number of reports have indicated that the phosphorylated form of Akt plays a critical role in mitochondrial apoptotic pathway through regulating the members of Bcl-2 family, including anti-apoptotic Bcl-2 and the pro-apoptotic Bax [31,32]. The balance between Bax and Bcl-2 stabilizes the permeability of the mitochondrial membrane [33]. When apoptosis is initiated, activated Bax promotes the release of cytochrome C from the outer mitochondria membrane space into cytosol, leading to the formation of the apoptosome, followed by apoptotic executor caspase-3 cleavage and activation [34,35]. In contrast, Bcl-2 can inhibit the release of cytochrome c and improve cell survival [22]. In our study, we detected Bax and cleaved caspase-3 expression increased, and p-Akt expression decreased after RIRI, which is consistent with the previous study [36]. In addition, the present data suggested that intravitreal injection of CREG upregulated p-Akt and Bcl-2 expression, and downregulated Bax and cleaved caspase-3 expression. Analogously, a recent study reported that CREG overexpression suppressed cell death by modulating Bax/Bcl-2/caspase-3 signaling during ischemia-induced liver insult. Moreover, Chen et al. [37] demonstrated that CREG ameliorated apoptosis of hippocampal neurons exposed to ischemia-reperfusion injury, which was attributed to its effects on regulating Akt signaling and downstream events, while CREG knockdown accelerated apoptosis. Taken together, our results confirmed that CREG protected RGCs against RIRI and inhibited apoptosis via the activation of Akt signaling pathway, which regulated downstream mitochondrial apoptotic pathways.
ERG is a mass electrical activity of the retina evoked by a series of brief flash stimuli, typically used to evaluate the function of retinal neurons. Under dark-adapted conditions, the a-wave is primarily generated by photoreceptors, while the b-wave is generated by Müller cells and bipolar cells [38]. Given that retinal neurons are very sensitive to ischemia, ERG is well-suited for assessing retinal function in glaucoma [39]. Retinal hypoperfusion and increased levels of excitatory amino acids attributed to RIRI interfere with cellular homeostatic modulation, thus attenuating the a-and b-wave amplitudes, which is consistent with our findings [40]. Meanwhile, compared with the RIRI group, there was