Antioxidant and Cytoprotective Potential of Erythropoietin in Mitigating Oxidative Stress-Induced Changes in the Retinal Pigment Epithelium

Erythropoietin (EPO) protects cells by inhibiting apoptosis, oxidative stress and inflammation in several models of retinal degeneration. In this study, we demonstrate the effects of recombinant Adeno Associated Virus (AAV) vector-mediated delivery of a modified form of erythropoietin (EPO-R76E) in an established mouse model of dry-AMD in which retinal degeneration is induced by RPE oxidative stress. Experimental vector AAV-EPO-R76E and control vector AAV-GFP were packaged into serotype-1 (AAV1) to enable RPE selective expression. RPE oxidative stress-mediated retinal degeneration was induced by exon specific deletion of the protective enzyme MnSOD (encoded by Sod2) by cre/lox mechanism. Experimental mice received subretinal injection of AAV-EPO-R76E in the right eye and AAV-GFP in the left eye. Western blotting of RPE/Choroid protein samples from AAV-EPO-R76E injected eyes showed RPE specific exogenous protein expression. Retinal degeneration was monitored by electroretinography (ERG). EPO-R76E over-expression in RPE delayed the progressive retinal degeneration as measured by light microscopy in RPE specific Sod2 knockout mice. Delivery of EPO-R76E vector can be used as a tool to prevent retinal degeneration induced by RPE oxidative stress as seen in this mouse model.

(mtDNA) in the neural retina and RPE cells resulting in reactive oxygen species (ROS) overproduction [3]. High mitochondrial ROS production imbalances antioxidant and cytoprotective systems in the RPE and play a pivotal role in AMD pathogenesis. Antioxidants, growth factors, and neurotrophic factors are widely proposed to protect RPE cells from oxidative damage-associated changes. [4].
Erythropoietin (EPO), a secreted cytokine, is FDA approved for the treatment of anemia. EPO has been shown to act as a novel agent in vascular protection against acute lung injury by promoting angiogenesis [5]. EPO provides neuroprotective effects in several animal models, as it blocks apoptotic pathways and indirectly induces endogenous antioxidants in neurons [6,7]. Intravenous EPO delivery improved visual acuity and color vision in patients following indirect traumatic neuropathy [8]. Studies show that systemic or retinal delivery of EPO or EPO-R76E, a modified form of EPO with reduced erythropoietic activity, can improve the function of retinal ganglion cells and photoreceptors cells [9][10][11][12][13][14][15][16][17][18][19].
We reported Sod2 flox/flox -VMD2 cre mice as an animal model of dry AMD by conditional genetic deletion of manganese superoxide dismutase (MnSOD, encoded by Sod2), a mitochondrial antioxidant enzyme in the RPE [20]. Loss of MnSOD in RPE leads to the induction of oxidative stress, further promoting progressive retinal degeneration seen as early as 4-6 months [21]. We have used this animal model to test various drugs, antioxidant genes, and nutritional supplements to improve the function of RPE and neural retina [22][23][24][25][26].
Eerythropoietin and the erythropoietin receptor (EPOR) are widely expressed within retinal cells, and several groups have tested the ability of exogenous EPO to ameliorate retinal degeneration associated in animal models of diabetic retinopathy, retinitis pigmentosa and other forms of retinal degeneration [17,27]. However, the potential of EPO to limit retinal degeneration associated with age-related macular degeneration (AMD) has not explored. Chronic oxidative stress in the RPE plays an important role in RPE loss in dry-AMD [2,3,28,29]. In response to sustained oxidative stress, RPE cells die by necroptosis [30,31]. We hypothesize that sustained expression of EPO-R76E in the RPE using an AAV vector will improve the health and survival of RPE and retinal photoreceptors. Thus, we evaluated the efficacy of the modified form of EPO in protecting RPE from oxidative stressinduced changes in our mouse model of dry-AMD. We show that EPO-R76E improves retinal function and preserves retinal thickness affected due to chronic oxidative stress in the RPE.

Study Design
Ten-to twelve-week-old mice of both sexes without were used to test the vector in vivo. The experimental vectors were delivered to the mice eyes by subretinal injection. The mice were analyzed for transgene expression by Western blotting six weeks following injection. The retinal function was recorded by scotopic ERG at 6 and 9 months of age (3 & 6 months after injection). At the end of the experiments (nine months of age), the retinal tissues were analyzed by histology. A schematic diagram of the experimental design is shown in Fig-1.

Cell culture and vector production
AAV plasmid having a modified form of EPO (EPO-R76E) cDNA were provided by Dr. Tonia S. Rex Transfection medium with DNA & PEI was replaced with complete growth media after 24 hours and further incubated for another 24 hours to allow transgene expression. Next day, cells were checked for GFP fluorescence.
After that, cells were dislodged using cold phosphate buffered saline (PBS). The cells were pelleted at 14,000 × g for 3 min at 4 0 C and kept at -80 0 C for western blotting. Around 1mg each of AAV-GFP and AAV-EPO-R76E plasmids were given to the Vector Core of the Center for Vision Research at the University of Florida to package into AAV1 serotype capsids. The packaged virus was purified using Iodixanol gradients and anion exchange chromatography [32] and achieved a stock concentration of 1 X 10 12 viral genome copies per milliliter (VG/ml).

Animals and Injection procedures
All the procedures involving animals in this study followed the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and the protocols were approved by Institutional Animal Care and Use Committee (IACUC) at the University of South Florida and University of Florida. Breeding pairs of RPE specific Sod2 deleted mice (Sod2 flox/flox -VMD2-cre) on C57BL/6 background were set up to generate the mice for this project.
These transgenic mice have the VMD2 promoter driving inducible cre transgene [33] and loxP sites surrounding exon 3 of Sod2 [34]. The cre transgene was induced by providing Dox food to the mother of newborn pups from P0-P14 that deleted Sod2 as described in detail by Mao et.al [20]. The genotype of the transgenic pups showing mutations of rd1, rd8 and rd10 were removed from the study. Intraperitoneal injection of mixture of ketamine (95 mg/kg) and xylazine (8 mg/kg) was used to anesthetize the mice for in vivo procedures such as fundus imaging,

P0
P14 P18 2-3 Month 4 month 9 month electroretinography (ERG), and retinal injections. The procedures from one of our previous publications were followed for eye dilation and local anesthesia [23]. Mice received subretinal injection of 1 μL of 10 12 VG/mL of AAV1-EPO-R76E (i.e., 1 × 10 9 total viral particles) in one eye and an equal dose of AAV1-GFP in the contralateral eye [23,35]. Any eyes with retinal detachment or any structural defect due to injection were excluded from further analysis.

Western blot analysis
Proteins from cells and retinal tissues were analyzed by western blotting to determine the expression of Odyssey instrument that showed two different colors for two different protein bands depending upon size.

In vivo Fundus imaging.
GFP fluorescence fundus imaging was performed to check the spread and expression of control vector ( Fig.   2C ) using Phoenix Micron 3 fundus camera. For this, pupils of the mice were dilated once with 1% atropine and twice with 2.5 phenylephrine, then mice were anesthetized, the cornea was lubricated by one drop of artificial tears (GenTeal, Alcon). The eyes of the mice were positioned to face the fundus camera and images were receorded keeping the optic nerve at the center using GFP filter and normal bright field filter.

Scotpoic Electroretinography (ERG)
To monitor retinal function, the scotopic (dark adapted) ERG response was measured using the Espion full-field ERG system Espion ColorDome Ganzfeld ERG system (Diagnosys, Inc., Lowell, MA) according to an established protocol [23,24] . For this, mice were dark-adapted overnight, and pupils were dilated with one drop each of 1% atropine, 2.5% phenylephrine, and a local anesthetic (proparacaine) was applied. After that, mice were anesthesiszed with ketamine/xyalzine as described for ocular injections. ERG a-wave, b-wave and c-wave amplitudes from both the eyes recorded at a flash intensity of 20dB . The results were compared between control and experimental vector injected eyes at 3 months and 6 months following subretinal injction. We used intraperitoneal injection of EUTHASOL® euthanasia solution (pentobarbital sodium and phenytoin sodium) to euthanize the experiemental mice. We followed the procedures for perfusions and tissue processing, embedding and sectioning as previously described [20]. Semithin cross-sections of 0.5 μm from resin embedded retinal tissue were cut through the optic nerves and mounted on glass slides. These sections were stained with 1% toluidine blue and 2% borate in distilled water. Stained sections were examined at 4X, 20X and 100X by light microscopy using Keyence All-in-One Fluorescence Microscope BZ-X800 (Itasca, IL 60143, U.S.A)

2.8.Statistical analysis.
GraphPad Prism 5.0 was used to illustrate the graph for ERG and light microscopy measurement data. Twotailed student t tests were used to test the statistical significance of differences in results. All the data were represented as mean ± SEM unless otherwise indicated. A p-value of <0.05 was considered significant.

EPO expression in RPE-specific Sod2 deleted mice.
Even though AAV1 transduces both Mueller glia and RPE following intravitreal injection, Mueller glia expression is much less than RPE [36]. In order to restrict transgene expression to RPE cells, we injected an AAV1 vector containing human modified EPO (EPO-R76E) into the subretinal space of 3-month-old RPE specific Sod2 deleted mice [37]. The contralateral eyes from the same animals were injected with AAV1 expressing humanized GFP as a control to evaluate the impact of subretinal injection or virus induced effects. By using fluorescence fundus imaging, we observed GFP expression over 50-70% of the retina (Fig-2C) that suggested .the efficiency of subretinal viral delivery. To detect and quantify exogenous transgene expression, the level of AAV-delivered EPO protein expression was examined one month following subretinal injection using an EPO antibody. The control and experimental vector injected eyes were harvested from a cohort of mice 1 month following injection. The retina and RPE/choroid from each eye were collected separately for protein analysis. EPO antibody detected exogenous expression in RPE/choroid samples injected with AAV-EPO (Fig-2D) as we see a 37KD protein band.
As expected, we found negligible expression of EPO in retina (Fig-2E) confirming the RPE specific tropism of AAV1 [38].

Improved Retinal and RPE function.
Under dark-adapted conditions, the ERG amplitudes of Sod2 deleted mice are lower than control mice. Three months following EPO treatment, a-wave and b-wave ERG amplitudes were significantly different between eyes treated with the experimental and the control vectors (Fig-3). At 6 months of age (3 months following injection), the eyes treated with AAV-EPO vector showed 46% improvement in a-wave response and 39% increase in b-wave response compared to contralateral eyes injected with AAV-GFP vector (Fig-3A and 3B). At 9 months of age (6 months following injection), the loss in a-and b-wave response were prevented (Fig-3D and 3E). We found 50% improvement of a-wave response and 32% improvement of b-wave response. The c-wave ERG response reflects the health of RPE. We recorded a 44% (Fig-3C) and 38% (Fig-3H) improvement in c-wave ERG responses at 6 months of age (3 months following injection) and at 9 months of age (6 months following injection) respectively in eyes injected with AAV-EPO vector compared to untreated eyes injected with AAV-GFP vector. It should be noted that long-term expression of GFP in the RPE of rodents does not affect the ERG response [39,40].

Improvement in retinal structure after treatment with AAV-EPO-R76E is revealed by light microscopy.
Previously, we have reported the decrease in retinal thickness in RPE-specific Sod2 knock out mice as the age progresses [20]. The effects of Sod2 deleted changes in the retina were visible by light microscopy as progressive RPE and photoreceptor cell degeneration in all AAV-GFP injected eyes (Fig-4A and B). We recorded around 32% preservation of retinal thickness in AAV-EPO injected eyes compred to control eyes (Fig-4C) . In control eyes injected with AAV-GFP vector, the RPE monolayer thinning along with irregular melanin pigment distribution was noticed (Fig-5B). Thinning of the RPE monolayer are indicative of RPE loss and impaired RPE integrity. The changes in response to AAV-EPO-R76E was indicated by thicker RPE implying better structural Integrity (Fig-5A).

Discussion
Clinical and experimental evidence for both dry and wet form of age-related macular degeneration (AMD) demonstrates disruption of the structural and functional integrity of the RPE in addition to loss of photoreceptors [2,[41][42][43]. We have shown that the deletion of mitochondrial form of superoxide dismutase (MnSOD or Sod2), an antioxidant gene especially in the RPE of an animal model impairs retinal structure and function. Oxidative stress is one of the key contributors to age-related retinal degeneration, particularly in dry-AMD [2,43]. Therefore, efforts are in progress to develop a therapeutic that can prevent further loss of structural and functional integrity of the RPE induced by oxidative damage. Growth factors offer the potential to prevent cell loss and degeneration of retinal cells from oxidative stress if they can be delivered to the specific cells. Therefore, using a cell-specific gene therapeutic approach, we have shown that subretinal delivery of serotype 1 AAV (AAV1) driving recombinant erythropoietin (EPO-R76E) can restrict EPO transgene expression in the RPE. This protects both the structural and functional integrity of RPE and retina impaired by RPE-specific oxidative stress.
Erythropoietin (EPO) is a hormone produced primarily by the kidneys, with small amounts made by the liver. EPO plays a key role in the production of red blood cells (RBCs), which carry oxygen from the lungs to the rest of the body. EPO is also expressed locally in the retina under the control of hypoxia inducible factor (HIF-1) [44]. EPO is present in considerably higher concentrations in eyes with diabetic macular edema than in eyes with exudative AMD or normal eyes [45]. The results from several studies indicate that systemic delivery of erythropoietic EPO is therapeutic for a broad range of neurodegenerative diseases [46][47][48][49]. The safe use of EPO is demonstrated in the clinic with other diseases as it can traverse the intact blood-brain and blood-retina barriers in therapeutic concentrations [44,50]. It was previously reported that the systemic delivery of EPO-R76E was able to provide successful preservation of retinal ganglion cells and visual function without significantly increasing hematocrit, unlike regular EPO [11]. Tao and colleagues recently demonstrated that pre-treament of mice by subretinal injection of AAV2-EPO, protected the retina from acute N-Methyl-N-Nitrosourea (MNU) toxicity [51]. There is a need, however, to demonstrate the best strategies for developing and delivering EPO or erythropoietic stimulating agents for the treatment of patients with atrophic or dry-AMD [52]. Adenoassociated virus (AAV) vectors can transduce a wide range of dividing and non-dividing cell types, which has made these vectors an important tool for ophthalmic gene therapy. A major advantage of AAV vectors is the long-term expression of therapeutic gene as episomes within cells that can be obtained after in vivo gene delivery [53]. AAV-EPO gene therapy vector offers the advantages of delivering and stably expressing EPO gene (or its protein product) to the physiologically relevant target tissues such as RPE using specific AAV serotypes (AAV1) or promoters (e.g. VMD2).
Our results indicate that stable expression of EPO-R76E in RPE cells protected the RPE and its nearby photoreceptors under the conditions of oxidative stress. EPO can protect the retina by acting directly on the RPE or by acting in a paracrine fashion on photoreceptors and Mueller glial. EPO has been shown to help maintain the barrier properties of the RPE, and this may contribute to protective role [27,54]. EPO protected RPE cells barrier integrity disrupted by oxidative stress by reducing intracellular ROS and restoring cellular antioxidant potential [55]. These authors also reported that there was a reduction in the secretion of inflammatory cytokines Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 12 April 2021 doi:10.20944/preprints202104.0273.v1 (TNFα, and IL1-1β) and a decrease in caspase-3 activity under oxidative stress in response to EPO treatment.
Since EPO is secreted, it may also protect retinal structure and function by acting directly on photoreceptors.
Exogenous EPO could directly interact with the photoreceptors allowing them to maintain the metabolic activity despite increased oxidative stress-related effects. This may also activate a signal transduction cascade in the photoreceptors [56][57][58][59]. The eyes from Sod2 flox/flox -VMD2 cre mice evidenced an increase in oxidative stress as early as two months of age. Protection from oxidative stress could be one of the reasons whereby EPO is permitting increased survival and prolonged function of photoreceptors. Another mechanism could be interactions with surrounding cells such as Mueller glia, which, in turn, can release proteins that support the survival of photoreceptors [60][61][62].
In AMD patients, the RPE and photoreceptors are comprimized in the macular region, causing loss of central vision. For treatment of macular degeneration, therefore, protection of cone photoreceptors is essential, because they are entiched in the maclula and are critical for visual acuity. Treatment with RPE-specific EPO-R76E using gene therapy may allow patients to have a useful vision for a longer period due to extendedexpression of EPO in the RPE, thus further preventing loss of vision. We observed protection of scotopic a, b, and c wave full-field ERG response signifying the protection of photoreceptors and RPE. Histological analysis at nine months of age that clearly demonstrated visible preservation of the photoreceptor and RPE layers (Fig-4A and 5A) compared to control treated eyes (Fig-4B and 5B). In the future, we aim to perform molecular analysis of RPE/ choroid and photoreceptors to evaluate changes in protective and inflammatory gene and protein expression as shown in other studies [63]. As Sod2 deletion is related to mitochondrial dysfunction in RPE, it will be interesting to see whether supplementation of EPO-R76E can rescue mitochondrial dysfunction and improve bioenergetics, as noticed in RPE cells derived from AMD patients [64].
EPO signaling increases choroidal macrophages and cytokine expression and exacerbates choroidal neovascularization, conditions associated with the advanced wet-form of AMD [65]. EPO receptor signaling supports retinal function after vascular injury [66], but its pro-angiogenic properties may limit the usefulness of unregulated EPO expression as a therapy for dry AMD. We plan to determine if EPO-R76E stimulated choroidal neovasuclarization (CNV) using the laser-induced CNV model.
This pilot study showed the protective effect of EPO in preserving retinal structure and function while maintaining stable expression in the RPE. Even though we did not see any harmful effect of EPO-R76E in our animal model of dry-AMD, we must also study the impact of prolonged EPO expression in normal mice. In our study, we did not measure the level of EPO expression in retinal tissue by AAV-EPO. We aim to compare the expression levels using intravitreal or systemic injection of other EPO activating compounds or clinically approved EPO protein in further studies. Given that cone photoreceptor loss is prominent in dry-AMD, it will be necessary to perform focal ERG and/or optokinetic responses.to measure cone function and also include spectral domain optical coherence tomography (SD-OCT) to monitor progessive preservation of retinal layers in vivo [23]. Other than that, we also aim to check the contribution of different cell types (Mueller glia, photoreceptors, astrocytes, microglia, endothelial cells, ganglion cells, etc.) in protecting the retina by analyzing transcriptional landscape at a single-cell level. We predict that local stable EPO expression can impact the proteome changes in RPE under conditions of oxidative stress. Proteomics studies can supplement learning the proteome changes by EPO while protecting the cellular microenvironment.
Institutional Review Board Statement: All the procedures involving animals in this study followed the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and the protocols were approved by Institutional Animal Care and Use Committee (IACUC) at the University of South Florida (USF) and University of Florida (UF).
Informed Consent Statement: Not applicable." Data Availability Statement: All the data is available through this manuscript.
Acknowledgments: Authors would like to thank Dr. Tonia S. Rex (Vanderbilt University, Nashville, TN USA), for providing the pAAV.EpoR76E, James H Thomas (University of Florida, Gainesville, FL, USA) for providing CsCl2 grade plasmid DNA and Vince A. Chiodo (University of Florida, Gainesville, FL, USA) for the endotoxin free AAV packaging. The authors are also grateful to Lindsey Buzz for helping in genotyping.

Conflicts of Interest:
The authors declare no conflict of interest." The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results".