Free
Retina  |   March 2014
Subretinal Delivery of AAV2-Mediated Human Erythropoietin Gene Is Protective and Safe in Experimental Diabetic Retinopathy
Author Affiliations & Notes
  • Hua Xu
    Department of Ophthalmology of Shanghai Tenth People's Hospital, and Tongji Eye Institute, Tongji University School of Medicine, Shanghai, China
  • Limei Zhang
    Department of Regenerative Medicine and Stem Cell Research Center, Tongji University School of Medicine, Shanghai, China
  • Limin Gu
    Department of Ophthalmology of Shanghai Tenth People's Hospital, and Tongji Eye Institute, Tongji University School of Medicine, Shanghai, China
  • Lixia Lu
    Department of Ophthalmology of Shanghai Tenth People's Hospital, and Tongji Eye Institute, Tongji University School of Medicine, Shanghai, China
    Department of Regenerative Medicine and Stem Cell Research Center, Tongji University School of Medicine, Shanghai, China
  • Guangping Gao
    Gene Therapy Center, University of Massachusetts Medical School, Boston, Massachusetts
  • Weiye Li
    Department of Ophthalmology of Shanghai Tenth People's Hospital, and Tongji Eye Institute, Tongji University School of Medicine, Shanghai, China
    Department of Regenerative Medicine and Stem Cell Research Center, Tongji University School of Medicine, Shanghai, China
    Department of Ophthalmology, Drexel University College of Medicine, Philadelphia, Pennsylvania
  • Guoxu Xu
    Department of Ophthalmology, Second Affiliated Hospital of Soochow University, Suzhou, China
  • Juan Wang
    Department of Ophthalmology of Shanghai Tenth People's Hospital, and Tongji Eye Institute, Tongji University School of Medicine, Shanghai, China
    Department of Regenerative Medicine and Stem Cell Research Center, Tongji University School of Medicine, Shanghai, China
  • Furong Gao
    Department of Ophthalmology of Shanghai Tenth People's Hospital, and Tongji Eye Institute, Tongji University School of Medicine, Shanghai, China
    Department of Regenerative Medicine and Stem Cell Research Center, Tongji University School of Medicine, Shanghai, China
  • Jing-Ying Xu
    Department of Ophthalmology of Shanghai Tenth People's Hospital, and Tongji Eye Institute, Tongji University School of Medicine, Shanghai, China
    Department of Regenerative Medicine and Stem Cell Research Center, Tongji University School of Medicine, Shanghai, China
  • Jun Yao
    Department of Ophthalmology of Shanghai Tenth People's Hospital, and Tongji Eye Institute, Tongji University School of Medicine, Shanghai, China
  • Fang Wang
    Department of Ophthalmology of Shanghai Tenth People's Hospital, and Tongji Eye Institute, Tongji University School of Medicine, Shanghai, China
  • Jingfa Zhang
    Department of Ophthalmology of Shanghai Tenth People's Hospital, and Tongji Eye Institute, Tongji University School of Medicine, Shanghai, China
    Department of Regenerative Medicine and Stem Cell Research Center, Tongji University School of Medicine, Shanghai, China
  • Guo-Tong Xu
    Department of Ophthalmology of Shanghai Tenth People's Hospital, and Tongji Eye Institute, Tongji University School of Medicine, Shanghai, China
    Department of Regenerative Medicine and Stem Cell Research Center, Tongji University School of Medicine, Shanghai, China
    Institute for Nutritional Sciences, Tongji University, Shanghai, China
  • Correspondence: Guo-Tong Xu, Department of Ophthalmology of Shanghai Tenth People's Hospital, Tongji Eye Institute, and Department of Regenerative Medicine, Tongji University School of Medicine, 1239 Siping Road, Medical School Building, Room 521, Shanghai, 200092 China; [email protected]
  • Jingfa Zhang, Department of Ophthalmology of Shanghai Tenth People's Hospital, Tongji Eye Institute, and Department of Regenerative Medicine, Tongji University School of Medicine, 1239 Siping Road, Medical School Building, Room 708, Shanghai, 200092 China; [email protected]
  • Fang Wang, Department of Ophthalmology of Shanghai Tenth People's Hospital, and Tongji Eye Institute, Tongji University School of Medicine, 301 Middle Yanchang Road, Shanghai, 200072 China; [email protected]
Investigative Ophthalmology & Visual Science March 2014, Vol.55, 1519-1530. doi:https://doi.org/10.1167/iovs.13-13155
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Hua Xu, Limei Zhang, Limin Gu, Lixia Lu, Guangping Gao, Weiye Li, Guoxu Xu, Juan Wang, Furong Gao, Jing-Ying Xu, Jun Yao, Fang Wang, Jingfa Zhang, Guo-Tong Xu; Subretinal Delivery of AAV2-Mediated Human Erythropoietin Gene Is Protective and Safe in Experimental Diabetic Retinopathy. Invest. Ophthalmol. Vis. Sci. 2014;55(3):1519-1530. https://doi.org/10.1167/iovs.13-13155.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose.: We studied and developed a gene-based intraocular erythropoietin (EPO) therapy for diabetic retinopathy (DR), by which the applicability of neuroprotective therapy with favorable safety profile is attempted.

Methods.: Hematocrit (Hct) was measured in C57BL/6 mice after intramuscular injection of AAV2-CMV-hEPO virus. Diabetes was induced by intraperitoneal injection of streptozotocin in Sprague-Dawley (SD) rats. Subretinal or intravitreal injection was performed in SD rats and Dark Agouti (DA) rats. The human EPO (hEPO) concentration was measured with ELISA. Blood–retinal barrier (BRB) breakdown was measured with Evans blue permeation. Retinal function was evaluated with electroretinography (ERG). Retinal cell apoptosis was detected with TUNEL. Retinal thickness and cell counts were examined by light microscopy. Retinal vascular changes were evaluated with fundus fluorescein angiography (FFA) and confocal microscopy.

Results.: The serum hEPO was elevated 2 weeks after AAV2-CMV-hEPO virus injection, and Hct began to increase after 4 weeks. After subretinal injection, hEPO expressions in aqueous humor, vitreous, and retina followed a dose- and time-dependent manner. In the AAV2-CMV-hEPO–treated diabetic group, BRB was maintained, and retinal cell apoptosis was significantly reduced. The ERG results showed that the retinal function remained unchanged for at least one year after subretinal injection of AAV2-CMV-hEPO virus. Long-term expression of hEPO following subretinal injection of AAV2-CMV-hEPO virus did not induce neovascularization in retina and choroid.

Conclusions.: The AAV2-CMV-hEPO gene therapy is safe, and it exerts long-term protective effects on diabetic retinas. Thus, the gene therapy by using AAV2-CMV-hEPO for DR is feasible.

Introduction
Diabetic retinopathy (DR) is one of the leading causes of legal blindness in the working population, which is defined as neurodegenerative and microvascular disease. 1 Despite significant advances in DR treatment, the main therapies still are laser photocoagulation and vitrectomy, and both are helpful or applicable in the late stage of DR. 24 Recently, anti-VEGF reagents are used widely in experimental and clinical studies, with proved benefit for proliferative diabetic retinopathy (PDR), diabetic macular edema (DME), and wet age-related macular degeneration (wet AMD). 47 However, anti-VEGF agents normally are not suitable for early DR, because the retinal ischemia is absent at this stage. 8 Retinal vasculature abnormality and neuronal cell death are the characteristics of early DR. 1,9,10 The collective evidences in humans and experimental animals exist that alterations in neuronal function and viability, together with functional changes in the vasculature, may contribute to the pathogenic mechanisms of DR beginning shortly after the onset of diabetes. 9,11 Therefore, the cytoprotection for early DR is a reasonable approach. 
Erythropoietin (EPO) is best known as a hematopoietic cytokine produced in fetal liver and adult kidney. Studies showed that EPO and its receptor (EPOR) also were expressed in the human central nervous system, 12 including retina. 13,14 In diabetic rats, EPOR expression is upregulated in diabetic retina. 15 The EPO, administered via intraperitoneal 16 or intravitreal 15 injection, was reported to maintain BRB integrity, and protect retinal vascular and neuronal cells in the early stage of diabetes. Single intravitreal injection of EPO is a safe procedure even with a high dose of EPO (1000 U) in rabbits. 17 A recent study demonstrated that the visual acuity of the patients with refractory DME was improved, and macular edema was reduced following 3 doses of intravitreal human EPO (hEPO, 5 U/eye) injections every 6 weeks. 18 However, repeated intraocular injections might cause several side effects, such as endophthalmitis, vitreous hemorrhage, and retinal detachment. 19 On the other hand, constant hEPO expression achieved with gene therapy through a subretinal approach might be able to overcome the disadvantages of repeated intravitreal injection. 20 Since EPO per se is an angiogenic factor, 2123 we must study whether a longstanding high concentration of EPO in the eye as a consequence of gene therapy will cause pathologic angiogenesis in the retina. 
Adeno-associated viral (AAV) vectors are becoming popular in the field of ocular gene therapy in recent years for their minimal immune responses, 24 high safety, 25 and long-term expressions. 26 Although neutralizing antibody against the AAV capsid was detected following a high dose intraocular treatment, no sufficient attenuation to the transgenic expression was observed. 27 To date, a number of studies have demonstrated the efficacy of gene therapy using AAV vectors to treat ocular diseases. 28,29 For example, AAV-mediated sFlt-1 significantly reduced ocular neovascularization in mice and monkeys. 30 In 2008, some clinical studies demonstrated that subretinal administration of the AAV2-mediated retinal pigment epithelium 65 (RPE 65) gene was safe and effective in treating Leber's congenital amaurosis (LCA). 3133 The results showed that visual acuity and light sensitivity were significantly improved in treated eyes, and no side effects were observed. 25 Prolonged expression of hEPO by AAV2 could be achieved in SD rats 34 and nonhuman primates 35 following a single subretinal injection, which laid down the foundation for AAV2-mediated hEPO expression to treat DR. 
Based on the knowledge of protective effect of EPO, 3640 especially on DR 15,16,18,41 and the availability of AAV vectors in ocular gene therapy, we studied the efficacy and safety of AAV2-CMV-hEPO virus in diabetic and normal rats. 
Methods
Reagents
Evans blue (30 mg/mL) and streptozotocin (STZ, pH 4.5) were purchased from Sigma-Aldrich (St. Louis, MO), and Beijing Superior Chemicals and Instruments Co., Ltd. (Beijing, China). Isolectin GS-IB4 (I21411) was purchased from Invitrogen (Carlsbad, CA). Fluorescein sodium (20%) was purchased from Guangzhou Mingxing Pharmaceutical Co., Ltd. (Guangzhou, China). Cell-viability assay kit (In situ Cell Death Detection Kit, catalog # 11958100) was purchased from Roche (Shanghai, China) and the EPO ELISA kit (Human Erythropoietin Immunoassay ELISA kit, catalog # 271203) was purchased from R&D (Shanghai, China). The AAV2-CMV-EGFP virus (original concentration 1 × 1012 GC/mL) was from Visualization and Graphics Technical Community (VGTC, Beijing, China). Both AAV2-CMV-hEPO virus and AAV2-CMV-hVEGF virus were made at the same original concentration of 1 × 1013 GC/mL. 
Cell Culture
The HEK 293T cells and R28 cells (rat retinal neuronal cell line) were maintained in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum. The AAV2-CMV-hEPO virus (1 × 1013 GC/mL, 10 μL) was added into the medium for infection; 48 hours later, the supernatant was collected for hEPO concentration measurement with the ELISA kit. The cells without virus treatment served as control. Supernatant from R28 cells was collected for the following colony-forming unit-erythroid (CFU-E) assay in vitro. 
Animals Studies
Male Sprague-Dawley (SD) rats, Dark Agouti (DA) rats, and C57BL/6 mice were purchased from SLAC Laboratory Animal Co. Ltd. (Shanghai, China), and housed with a 12/12-hour light/dark cycle. The animals were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 
To study infection efficiency, male adult SD rats were injected either intravitreally or subretinally with AAV2-CMV-EGFP virus (1 × 1012 GC/mL, 3 μL/eye, n = 12/group). The eyes failed for injections and with intraocular hemorrhage were excluded. At 2, 4, 6, and 8 weeks after injection, the rats were killed, and the eyes were enucleated for EGFP examination in the retinas with cryosections and in the RPE cells with the flat mount of RPE-Bruch's membrane-choriocapillaris complex (RBCC). 
To confirm the biological activity of hEPO, C57BL/6 mice (±25 g, n = 20) were injected intramuscularly with AAV2-CMV-hEPO virus (1 × 1013 GC/mL, 6 μL). The age-matched control mice (n = 20) were injected with equivalent volume of PBS (6 μL). The blood (1 mL) was withdrawn at the following time points: 1, 2, 4, 8, and 12 weeks after injection, for hematocrit (Hct) assay with Wintrobe's method. The serum hEPO concentration was measured with an ELISA kit. 
For dose-dependent expressions, the rats were injected subretinally with 6 doses of AAV2-CMV-hEPO virus; that is, 3 × 1010 GC/eye, 9 × 109 GC/eye, 3 × 109 GC/eye, 3 × 108 GC/eye, 3 × 107 GC/eye, and 3 × 106 GC/eye (n = 6). The rats with PBS injection (n = 6, 3 μL/eye) served as controls. One month later, serum, aqueous humor, vitreous, and retina were collected separately for ELISA measurement of hEPO concentration. 
To study the long-term expression of hEPO in the eye, the rats were injected subretinally with AAV2-CMV-hEPO virus (1 × 1013 GC/mL, 3 μL/eye, n = 6) and the aqueous humor was collected from 7 days to 2 years after injection for ELISA assay. 
To examine the protective effect of AAV2-CMV-hEPO, diabetes was induced by intraperitoneal injection of STZ (60 mg/kg body weight); and diabetes conditions were confirmed with 3 consecutive blood glucose measurements (>250 mg/dL). The rats that failed to develop diabetes were excluded. Two weeks after diabetes onset, the rats were divided randomly into 6 groups: normal control (N), diabetic control (D), diabetic rats with intravitreal hEPO protein injection (E0, positive control), and three doses of AAV2-CMV-hEPO virus–treated diabetic groups (E1, E2, and E3). For the N and D groups, rats were injected subretinally with PBS (3 μL/eye); for the E0 group, rats were injected intravitreally with hEPO protein (8 mU/μL, 2 μL/eye); for the E1 to E3 groups, the rats were injected subretinally with 3 different doses of AAV2-CMV-hEPO virus (3 μL/eye); that is, 3 × 1010 GC/eye (E1), 9 × 109 GC/eye (E2), and 3 × 109 GC/eye (E3). One month after intraocular administration, the rats were studied for blood–retinal barrier (BRB) permeability assay, TUNEL assay, and histologic evaluation. 
For the safety assay, DA rats (n = 5) and SD rats (n = 5) were injected subretinally with AAV2-CMV-hEPO (1 × 1013 GC/mL, 3 μL/eye). Equivalent volume of PBS or AAV2-CMV-hVEGF (1 × 1013 GC/mL) was administered subretinally into the eyes as negative or positive control, respectively. Two and 4 weeks later, fundus fluorescein angiography (FFA) was performed to detect the leakage of the retinal blood vessels. 
Six months after subretinal injection, neurosensory retina and RBCC flat mounts in SD rats were dissected and stained with isolectin to evaluate the vascular changes. 
Intravitreal and Subretinal Injection
Animals were anesthetized with 2% sodium pentobarbital (1 mL/400 g body weight). The pupils were fully dilated with 0.5% tropicamide (Wuxi Shanhe Group, Jiangsu, China). Topical anesthesia with 0.4% oxybuprocaine hydrochloride eye drops (Eisai Co. Ltd., Tokyo, Japan) was applied to reduce the animals' discomfort. Intravitreal injection was performed with a 30-gauge, 0.5-inch needle (BD Biosciences, Franklin Lakes, NJ) on a microsyringe (Lot # 395629; Hamilton, Reno, NV), using a temporal approach, 2 mm behind limbus. For subretinal injection, a tunnel was first performed with a 30-gauge, 0.5-inch needle, 2 mm posterior and parallel to the limbus, and then AAV2-CMV-hEPO virus was delivered into subretinal space with a 33-gauge needle (Lot # 440602; Hamilton) on a microsyringe (P/N: 7634-01/00; Hamilton). The formation of a bleb immediately after subretinal injection indicated successful operation. 
Detection of hEPO in Serum, Aqueous Humor, Vitreous, and Retina
One month after subretinal injection, the rats were anesthetized with 2% sodium pentobarbital. The pupils were fully dilated and the eyes were topically anesthetized. The blood (500 μL) was collected from tail and centrifuged at 12,000g, 4°C for 2 minutes and supernatant serum was collected. Under a dissecting microscope, the center of the cornea was penetrated with a 30-gauge, 0.5-inch needle (BD Biosciences) on a microsyringe (Lot # 401083, 100 μL; Hamilton) for aqueous humor collection. Then, the eyes were enucleated and the anterior parts were removed, and the vitreous was aspirated carefully with an Eppendorf pipette (0.5–10 μL). After that, the retina was isolated carefully under the dissecting microscope. The samples were stored at −80°C until assay. 
The hEPO concentrations in serum, aqueous, vitreous, and retina were evaluated by ELISA kit according to the manufacturer's instructions. Before detection, single retina was homogenized in radio-immunoprecipitation assay (RIPA) buffer (150 μL/retina). The concentration of total protein was determined with a bicinchoninic acid (BCA) protein assay kit (Pierce Biotechnology, Rockford, IL). The samples of serum, aqueous humor, and vitreous were diluted by ×1, ×10, and ×12, respectively. 
CFU-E Assay
Seven- to 9-week-old C57BL/6 mice were euthanized, and their bone marrow was obtained by flushing of mouse femurs and tibias with RPMI1640 medium supplemented with 10% fetal bovine serum. The bone marrow cells in methylcellulose medium (MethoCult GF M3534; Stemcell Technologies, Vancouver, British Columbia, Canada) were diluted to 1 × 105 cells/mL for CFU-E assay. 
The aqueous humors from SD rats as well as the supernatant from R28 cells treated with or without AAV2-CMV-hEPO virus were collected in advance. The hEPO concentrations in these samples were measured with ELISA. Then, the samples were diluted to 10% by the methylcellulose medium and the suspension of bone marrow cells for the CFU-E colony assay. Equivalent concentrations of commercial rHuEPO (EPIAO; Shenyang Sunshine Pharmaceutical Co., Ltd., Shenyang, China) served as positive control (PC) with two medium concentrations (PC 0.25, 0.25 U/mL; PC 0.5, 0.5 U/mL) and one higher concentration (PC 1, 1 U/mL). Colony numbers were enumerated using a light microscope under the magnification of ×200. Bone marrow cells without any treatment served as normal control (NC); aqueous from normal SD rats (A-NC) and supernatant from normal R28 cells (S-NC) also were used as controls. The erythropoietic function of the aqueous humor from SD rats with AAV2-CMV-hEPO subretinal injection (A-AE) and the supernatant from R28 cells infected with AAV2-CMV-hEPO (S-AE) were studied. 
BRB Permeability Examination
The BRB permeability was measured with the method described previously. 15 Briefly, the rats were deeply anesthetized, and Evans blue solution (30 mg/mL) was injected intravenously. Then 0.1 mL blood was drawn at 15-minute intervals up to 2 hours to obtain the time averaged Evans blue plasma concentration. After the dye had circulated for 2 hours, the chest cavity was opened, and rats were perfused via left ventricle with approximately 200 mL perfusion buffer (1% paraformaldehyde dissolved in 0.05 M citric acid, pH 3.5, 37°C). Immediately after the perfusion, both eyes were enucleated. Two retinas from the same animal were pooled together and dried in a Speed-Vac (37°C). After the dry weight was determined, the retinas were incubated in 300 μL formamide for 18 hours at 70°C. The extracts were centrifuged through 30,000 Nominal Molecular Weight Limit (NMWL) centrifuge filter (Microcon; Millipore Corporation, Billerica, MA) at 3,000g, 4°C, for 45 minutes. Blood samples and retinal extracts were measured with a spectrophotometer (Beckman DU800; Beckman Coulter, Inc., Pasadena, CA). Evans blue permeation (EBP), as a function of BRB permeability, was calculated and expressed in (μL plasma × g retinal dry wt−1·h−1). 
Histology
Cryosections of retina were prepared according to the method published previously. 15 The enucleated eyeballs were fixed in PBS-buffered 4% paraformaldehyde for 24 hours and then opened along the ora serrata. The posterior eyecups were dehydrated through a gradient concentration of sucrose from 10% to 30%. After dehydration, the eyecups were embedded in OCT compound (Tissue Tek; Sakura Finetek USA, Inc., Torrance, CA). Serial sections (10 μm) were cut on a cryostat microtome (Micron HM 525; Leica Microsystems, Inc., Wetzlar, Germany) for TUNEL assay. For paraffin sections, eyes were fixed in 10% formalin-glacial acetic acid-ethanol fixative solution (FAA) for 10 minutes, then in PBS-buffered 4% paraformaldehyde overnight. The posterior eyecups were dehydrated through a gradient concentration of ethanol from 50% to 100% (each for 30 minutes), and then in xylene (twice, 10 minutes each), soaked in paraffin wax (twice, 1 hour each). After that, the eyecups were embedded into paraffin blocks (EM-400 Embedding Medium Paraffin; Leica Microsystems, Inc.). The paraffin sections (5 μm) were cut on a paraffin microtome (RM2126RT; Leica Microsystems, Inc.), and stained with hematoxylin and eosin (H&E) for retinal thickness measurement and cell counts. 
In Situ Detection of Cell Death in the Retina by TUNEL Assay
The TUNEL assay was performed with a kit according to the manufacturer's instructions (In Situ Cell Death Detection Kit; Roche, China, Ltd.). Positive controls were retinal sections that had been treated with grade I DNase-I for 10 minutes at room temperature before the labeling procedure. Negative controls were the retinal sections treated with 10 μL label solution, but incubated in the absence of terminal transferase. The sections, rinsed three times with PBS after incubation, were analyzed by fluorescence microscopy (Nikon, Yokohama, Japan), with an excitation wavelength in the range of 450 to 490 nm. 
Measurements of the Retinal Thickness and Cell Counts
The thickness of different retinal layers was measured under light microscopy (×400), including: outer limiting membrane to inner limiting membrane (OLM-ILM), outer plexiform layer to ganglion cell layer (OPL-GCL), outer nuclear layer (ONL), inner nuclear layer (INL), and outer plexiform layer to inner nuclear layer (OPL-INL). Two measurements were taken on each section, at two reference lines, which were 1 mm away from the optic nerve head on the nasal and temporal sides. 
The numbers of cells in the ONL and INL were counted in the same regions as the thickness measured under the ×400 magnification. The cell nuclei within a fixed 25-μm column, centered with the 1 mm reference lines, were counted. The cell density then was expressed as the cell counts per millimeter width of retina in the different layers. 
Neurosensory Retina and RBCC Isolation, and Immunostaining With Isolectin
The detailed methods for isolation and immunostaining of neurosensory retina and RBCC flat mounts were described by Zhang et al. 42 Briefly, under a dissecting microscope, RBCC and neurosensory retina were isolated carefully. For immunostaining, the retina and RBCC were incubated with Isolectin GS-IB4 (1:100; Invitrogen, I21411; Alexa Fluor 488 Conjugate; Life Technologies, Carlsbad, CA) in blocking buffer overnight at 4°C. After washing thoroughly in PBS for 15 minutes, the retina and RBCC then were mounted with mounting medium (Dako Cytomation Fluorescent Mounting Medium, S3023; Dako North America, Inc., Carpinteria, CA). Images were taken with the fluorescence microscope or confocal laser-scanning microscope (TCS SPE; Leica Microsystems, Inc.). 
Fundus Fluorescein Angiography (FFA)
The DA rats were anesthetized with 2% sodium pentobarbital (1 mL/400 g body weight), and the pupils were dilated. Soon after intraperitoneal injection of fluorescein sodium (10%, 75 mg/kg), the rats were placed on a stereotactic platform, and dynamic fundus photographs were taken. 
Electroretinogram (ERG)
The SD rats were dark adapted for a 12-hour period overnight, and ERGs were recorded in the dark room. Briefly, animals were anesthetized with intraperitoneal injection of 2% sodium pentobarbital (1 mL/400 g body weight), pupils were dilated with 0.5% tropicamide, and corneas were anesthetized with a drop of 0.4% oxybuprocaine hydrochloride. Recordings were made with an electrode placed on the cornea. The reference electrode was placed subcutaneously in the anterior scalp between the eyes, and the ground electrode was inserted into the tail. Three bright flashes were used: 6.325 × e−4 cd × s/m2, 6.325 × e−3 cd × s/m2, and 6.325 × e−2 cd × s/m2
Statistical Analysis
Data are expressed as mean ± SE. The statistical analysis was carried out using the Student's t-test. A P value of 0.05 or less was considered statistically significant. 
Results
Infection Efficiency of AAV2 Vectors to RPE Cells and Photoreceptors Following Different Injection Routes
The infection efficiency of AAV2 in RPE cells and photoreceptors was determined in SD rat eyes following either intravitreal or subretinal injection. The AAV2-mediated enhanced green fluorescent protein (AAV2-CMV-EGFP) was used as a reporter. After subretinal injection, robust EGFP expression was detected in RPE cells and photoreceptors from 2 to 8 weeks after injection (Supplementary Fig. S1, right panels). However, in the eyes with intravitreal injection, the EGFP expression began to be detectable beyond 6 weeks after injection. Only few EGFP can be observed in the ganglion cell layer (GCL) and INL, while nonexpression is observed in the RPE cells (Supplementary Fig. S1, left panels). 
We also used RBCC flat mount to test AAV2 vector infection efficiency in RPE cells with those two different injection methods (Supplementary Fig. S2). From 2 to 8 weeks, no EGFP expression was detected in RPE cells with intravitreal injection (Supplementary Fig. S2, left panels). However, for subretinal injection, EGFP expression began to be detectable in a few RPE cells 4 days after subretinal injection (data not shown), and became more robust and pervasive from 2 weeks (Supplementary Fig. S2, right panels). These results indicated that, with the same dosage of the AAV2 vectors, the infection efficiency in RPE cells and photoreceptors with intravitreal injection is much lower than that with subretinal injection. Therefore, subretinal injection was used in the following experiments. 
Biological Function Assay of hEPO in Mice Following Intramuscular Injection of AAV2-CMV-hEPO Virus
Before subretinal injection, we verified the expression and biological activity of hEPO. High level of hEPO (5174.8 mU/mL) was detected in the supernatant of HEK 293T cells 48 hours after AAV2-CMV-hEPO virus addition (Fig. 1A). Then, we intramuscularly injected AAV2-CMV-hEPO virus to C57BL/6 mice to detect its erythropoietic function. The result showed that, compared to normal control mice, the Hct of AAV2-CMV-hEPO injected mice began to increase at 4 weeks after injection (increased by 72%), and was persistent from 8 weeks (increased by 84%) to 12 weeks (increased by 85%, P < 0.05, n = 4, Fig. 1B). Simultaneously, the serum hEPO level also was elevated. As early as 2 weeks, the serum hEPO concentration began to increase (3.4 ± 1.3 mU/mL), and was persistent from 4 weeks (3.7 ± 1.6 mU/mL) to 8 weeks (6.8 ± 1.6 mU/mL), and to 12 weeks (7.9 ± 1.6 mU/mL, P < 0.05, n = 4, Fig. 1C). These results indicated that the hEPO protein expressed by AAV2-CMV-hEPO virus has the function of promoting red blood cell production. 
Figure 1
 
The biological activity assay of hEPO in vitro and in vivo. (A) Human EPO concentration was significantly increased 48 hours after infection to HEK 293T cells with AAV2-CMV-hEPO virus, which reached 5174.8 mU/mL (averaged hEPO concentration of 4 independent experiments). (B) The Hct was increased from 4 to 12 weeks after intramuscular injection of AAV2-CMV-hEPO in C57BL/6 mice (1 × 1013 GC/mL, 6 μL/mouse, n = 4/time point). (C) The increased Hct in AAV2-CMV-hEPO–injected mice was correlated with the increased serum hEPO protein (n = 4/time point). Each datum was expressed as mean ± SE. *P < 0.05 when compared to that in normal control.
Figure 1
 
The biological activity assay of hEPO in vitro and in vivo. (A) Human EPO concentration was significantly increased 48 hours after infection to HEK 293T cells with AAV2-CMV-hEPO virus, which reached 5174.8 mU/mL (averaged hEPO concentration of 4 independent experiments). (B) The Hct was increased from 4 to 12 weeks after intramuscular injection of AAV2-CMV-hEPO in C57BL/6 mice (1 × 1013 GC/mL, 6 μL/mouse, n = 4/time point). (C) The increased Hct in AAV2-CMV-hEPO–injected mice was correlated with the increased serum hEPO protein (n = 4/time point). Each datum was expressed as mean ± SE. *P < 0.05 when compared to that in normal control.
Colony Formation Assay for the Biological Activity of hEPO Expressed by AAV2-CMV-hEPO Virus
To further confirm the biological activity of expressed hEPO by AAV2-CMV-hEPO, CFU-E colony formation assay was applied in vitro. The hEPO concentrations of aqueous humors from SD rats, as well as the supernatant from R28 cells treated with or without AAV2-CMV-hEPO virus were quantified with ELISA. No hEPO was detected in nontreated samples; that is, neither in aqueous humor from control SD rats nor in supernatant from control R28 cells, but the concentration of hEPO was approximately 2.5 U/mL in aqueous humor from AAV2-CMV-hEPO virus–treated SD rats, and approximately 5 U/mL in the supernatant from R28 cells infected with AAV2-CMV-hEPO virus. The data showed that there were approximately 2 CFU-E colonies per ×200 visual field in the NC, A-NC, and S-NC groups, while there were approximately 5 CFU-E colonies in the aqueous humor from SD rats with subretinal injection of AAV2-CMV-hEPO (A-AE) and 7 CFU-E colonies in the supernatant from R28 cells infected with AAV2-CMV-hEPO (S-AE, Supplementary Fig. S3A). The numbers of CFU-E colonies in the A-AE and S-AE groups were similar to those in the corresponding PC groups. The representative images of CFU-E in different groups showed that there were more red strawberry-like colonies (CFU-E) in the positive control, A-AE, and S-AE groups (Supplementary Fig. S3B), indicating the erythropoietic function of the expressed hEPO by AAV2-CMV-hEPO virus. 
The Dose- and Time-Dependent Expression Pattern of AAV2-CMV-hEPO
The hEPO concentrations in aqueous humor, vitreous, and retina followed a dose-dependent manner (Fig. 2). The data showed that, 4 weeks after subretinal injection, hEPO concentrations in aqueous humor samples in the eyes treated with different doses of AAV2-CMV-hEPO virus, from 3 × 1010 to 3 × 106 GC, were 1065.3, 653.8, 282.6, 78.8, 0, and 0 mU/mL, and those in vitreous were 2763.1, 2286.9, 1178.1, 599.9, 62.7, and 12.2 mU/mL, and in the retina they were 2.31, 1.03, 0.84, 0, 0, and 0 mU/mg protein. It should be noted that no hEPO was detected in aqueous humor (Fig. 2A) and retina (Fig. 2C) when lower doses of AAV2-CMV-hEPO virus (3 × 107 and 3 × 106 GC) were administered, and the concentration of hEPO in vitreous (Fig. 2B) was higher than that in aqueous humor in the same infected eye. 
Figure 2
 
Human EPO concentrations in aqueous humor (A), vitreous (B), and retina (C) 4 weeks after subretinal injection of AAV2-CMV-hEPO in SD rats (3 μL/eye, n = 6). The doses of the virus vectors in 7 groups are as followings: (0) PBS control, (1) 3 × 1010 GC, (2) 9 × 109 GC, (3) 3 × 109 GC, (4) 3 × 108 GC, (5) 3 × 107 GC, and (6) 3 × 106 GC. Each datum was expressed as mean ± SE.
Figure 2
 
Human EPO concentrations in aqueous humor (A), vitreous (B), and retina (C) 4 weeks after subretinal injection of AAV2-CMV-hEPO in SD rats (3 μL/eye, n = 6). The doses of the virus vectors in 7 groups are as followings: (0) PBS control, (1) 3 × 1010 GC, (2) 9 × 109 GC, (3) 3 × 109 GC, (4) 3 × 108 GC, (5) 3 × 107 GC, and (6) 3 × 106 GC. Each datum was expressed as mean ± SE.
Figure 3 showed that long-term expression of hEPO could be achieved with a single subretinal injection. Human EPO in aqueous humor was detected as early as 7 days after injection, and its expression sustained for more than 2 years (Fig. 3). The hEPO concentration was 99.6 mU/mL at 7 days (n = 6); and it increased to 1572.8 mU/mL at 28 days, and this high level was maintained for approximately 1 year. At 18 months, the hEPO concentration was decreased to 645.4 mU/mL and at approximately 2 years it was 449.2 mU/mL. Serum hEPO concentrations at corresponding time points were undetectable, while Hct was increased (in 3 of 7 rats) 18 weeks after subretinal injection, and maintained in the following examination, indicating the leakage of hEPO protein (data not shown). 
Figure 3
 
Aqueous hEPO expression follows a time-dependent manner after subretinal injection of AAV2-CMV-hEPO. The concentrations of hEPO in the aqueous were detected with ELISA at different time points after subretinal injection of AAV2-CMV-hEPO (1 × 1013 GC/mL, 3 μL, n = 6/group). Each datum was expressed as mean ± SE.
Figure 3
 
Aqueous hEPO expression follows a time-dependent manner after subretinal injection of AAV2-CMV-hEPO. The concentrations of hEPO in the aqueous were detected with ELISA at different time points after subretinal injection of AAV2-CMV-hEPO (1 × 1013 GC/mL, 3 μL, n = 6/group). Each datum was expressed as mean ± SE.
BRB Maintenance by AAV2-CMV-hEPO in Diabetic Rats After Subretinal Injection
We used Evans blue permeation quantitation method to test the BRB maintenance effect of AAV2-CMV-hEPO on experimental DR rats with single subretinal injection of three gradient diluted concentrations of the virus (E1–E3: E1, 1 × 1013 GC/mL; E2, 3 × 1012 GC/mL; E3, 1 × 1012 GC/mL, 3 μL/eye) 2 weeks after diabetes onset. The results showed that the leakage of Evans blue into the retina of 6-week diabetic rat was significantly increased (8.6 ± 1.8 μL plasma × g retinal dry wt−1 · h−1), and approximately 4.4-fold of that in normal control (1.6 ± 0.2 μL plasma × g retinal dry wt−1 · h−1, P < 0.05, n = 8); 1 month after virus injection, the leakage of Evans blue was dramatically reduced, which was decreased by 66.3% (E1, 2.9 ± 0.4 μL plasma × g retinal dry wt−1 · h−1), 75.6% (E2, 2.1 ± 0.4 μL plasma × g retinal dry wt−1 · h−1), and 61.6% (E3, 3.3 ± 0.9 μL plasma × g retinal dry wt−1 · h−1), respectively, compared to diabetic control (P < 0.05, n = 8). In the hEPO-treated group (E0), which served as a positive control, BRB breakdown was reduced by 67.4% (2.8 ± 0.2 μL plasma × g retinal dry wt−1 · h−1, P < 0.05, n = 8). There was no significant difference among the E0 group and 3 AAV2-CMV-hEPO–treated groups (P > 0.05, Fig. 4). 
Figure 4
 
Evans blue-albumin permeation measurement treated with or without AAV2-CMV-hEPO. N, normal control; D, diabetic rats; E1 to E3, diabetic rats injected with 3 doses of AAV2-CMV-hEPO virus (E1, 1 × 1013 GC/mL; E2, 3 × 1012 GC/mL; and E3, 1 × 1012 GC/mL, 3 μL/eye and n = 8); E0, diabetic rats injected with hEPO protein (16 mU/eye). Each datum was expressed as mean ± SE. *P < 0.05 when compared to the diabetic group.
Figure 4
 
Evans blue-albumin permeation measurement treated with or without AAV2-CMV-hEPO. N, normal control; D, diabetic rats; E1 to E3, diabetic rats injected with 3 doses of AAV2-CMV-hEPO virus (E1, 1 × 1013 GC/mL; E2, 3 × 1012 GC/mL; and E3, 1 × 1012 GC/mL, 3 μL/eye and n = 8); E0, diabetic rats injected with hEPO protein (16 mU/eye). Each datum was expressed as mean ± SE. *P < 0.05 when compared to the diabetic group.
To further confirm the efficacy of AAV2-CMV-hEPO on experimental DR, we also administered AAV2-CMV-hEPO at other time points after diabetes onset. For instance, when we injected AAV2-CMV-hEPO 1 or 2 months after diabetes onset, the similar protection on BRB was confirmed (data not shown). 
Neuronal Protection by Subretinal AAV2-CMV-hEPO in Diabetic Rats
To test if subretinal administration of AAV2-CMV-hEPO could protect retinal neurons from diabetes insult, we performed TUNEL assay and histologic analysis. Compared to normal control, TUNEL-positive cells were increased significantly in the diabetic group, mainly in ONL, and few in INL (Fig. 5). After treatment with AAV2-CMV-hEPO (E2) and hEPO protein (E0), TUNEL-positive cells were significantly reduced (Fig. 5). The other two doses of AAV2-CMV-hEPO (E1 and E2) also effectively reduced the cell death of retinal neurons (Supplementary Fig. S4). 
Figure 5
 
TUNEL assay in 1.5-month diabetic rat retina treated with or without AAV2-CMV-hEPO. N, normal control; D, diabetic rats; E2, diabetic rats injected with AAV2-CMV-hEPO (3 × 1012 GC/mL, 3 μL/eye); E0, diabetic rats injected with hEPO protein (16 mU/eye); NC, negative control; PC, positive control. n = 6/group. Scale bar: 50 μm.
Figure 5
 
TUNEL assay in 1.5-month diabetic rat retina treated with or without AAV2-CMV-hEPO. N, normal control; D, diabetic rats; E2, diabetic rats injected with AAV2-CMV-hEPO (3 × 1012 GC/mL, 3 μL/eye); E0, diabetic rats injected with hEPO protein (16 mU/eye); NC, negative control; PC, positive control. n = 6/group. Scale bar: 50 μm.
Morphometric examination in H&E-stained retinal paraffin sections demonstrated that the retinal thickness and cell counts were greatly reduced in diabetic rats when compared to those in normal control, while subretinal administration of E1 to E3 could significantly restore those to nearly the normal level (Fig. 6). In 6-week diabetic rat retinas, the total retinal thickness (OLM-ILM) was 163 ± 1.2 μm compared to that in normal control (261 ± 1.1 μm, n = 4, P < 0.05). In AAV2-CMV-hEPO–treated groups, total retinal thickness was 182 ± 6.4 μm (E1), 242 ± 5.2 μm (E2), and 182 ± 7.0 μm (E3), respectively (n = 4, P < 0.05 when compared to the diabetic group, Fig. 6A). For the E0 group, the total retinal thickness was 192 ± 3.5 μm (n = 4, P < 0.05 when compared to the diabetic group, Fig. 6A). For other parameters (OPL-GCL, ONL, INL, and OPL-INL), the same trends also were demonstrated, with the most obvious change in ONL. The thickness of ONL in diabetic retina was reduced significantly when compared to that in normal control (37 ± 1.0 μm vs. 62 ± 0.7 μm, n = 4 and P < 0.05). In three AAV2-CMV-hEPO–treated groups, the thickness of ONL was well restored with 42 ± 1.4 μm (E1), 63 ± 1.6 μm (E2), and 51 ± 0.7 μm (E3, n = 4, P < 0.05 when compared to the diabetic group). In the E0 group, the thickness of ONL was 57 ± 2.3 μm (n = 4, P < 0.05 when compared to the diabetic group, Fig. 6A). To further confirm the above changes, retinal thickness in each group was studied in cryosections stained with DAPI (Supplementary Fig. S5). A significant decrease in retinal thickness was detected in the diabetic group (133 ± 4.2 μm, n = 5), when compared to that in normal control (199 ± 1.4 μm, n = 5, P < 0.01), the E2 group (180 ± 3.2 μm, n = 5, P < 0.01), and the E0 group (186 ± 3.9 μm, n = 5, P < 0.01, Supplementary Fig. S5). The thickness of ONL in diabetic retina also was significantly reduced when compared to that in normal control (44 ± 1.1 μm vs. 68 ± 1.1 μm, n = 5 and P < 0.01), and it was well protected in the E2 (62 ± 2.7 μm, n = 5, P < 0.01 compared to the diabetic group) and E0 (61 ± 0.8 μm, n = 5, P < 0.01 compared to the diabetic group) groups (Supplementary Fig. S5). 
Figure 6
 
Protecting retinal thickness (A) and preventing loss of retinal neurons (B) in STZ-diabetic rat by AAV2-CMV-hEPO administration. N, normal control; D, diabetic rats; E1 to E3, diabetic rats injected with AAV2-CMV-hEPO (E1, 1 × 1013 GC/mL; E2, 3 × 1012 GC/mL; and E3, 1 × 1012 GC/mL, 3 μL/eye and n = 4); E0, diabetic rats injected with hEPO protein (16 mU/eye). Each datum was expressed as mean ± SE. *P < 0.05 when compared to the diabetic group.
Figure 6
 
Protecting retinal thickness (A) and preventing loss of retinal neurons (B) in STZ-diabetic rat by AAV2-CMV-hEPO administration. N, normal control; D, diabetic rats; E1 to E3, diabetic rats injected with AAV2-CMV-hEPO (E1, 1 × 1013 GC/mL; E2, 3 × 1012 GC/mL; and E3, 1 × 1012 GC/mL, 3 μL/eye and n = 4); E0, diabetic rats injected with hEPO protein (16 mU/eye). Each datum was expressed as mean ± SE. *P < 0.05 when compared to the diabetic group.
Meanwhile, the cell counts in ONL in diabetic rats also were significantly reduced (1529 ± 45 vs. 2526 ± 32 cells/mm retina, n = 4 and P < 0.05). In contrast, the cell counts in AAV2-CMV-hEPO–treated retinas were 1791 ± 86 (E1), 2154 ± 26 (E2), and 1801 ± 93 (E3) cells/mm retina, respectively (n = 4, P < 0.05 when compared to the diabetic group). In the E0 group, the cell counts in ONL were 1899 ± 52 cells/mm retina (n = 4, P < 0.05 when compared to the diabetic group, Fig. 6B). For the cell counts in INL in diabetic rats treated with or without AAV2-CMV-hEPO, the same trend also was observed. The cell counts were significantly reduced in diabetic rats when compared to those in normal control (456 ± 22 vs. 630 ± 15 cells/mm retina, n = 4, P < 0.05). The numbers were increased to 488 ± 21 (E1), 573 ± 18 (E2, n = 4, P < 0.05), and 497 ± 13 (E3) cells/mm retina following AAV2-CMV-hEPO treatment. In the E0 group, the cell counts in INL were 522 ± 45 cells/mm retina (n = 4, Fig. 6B). 
No Retinal Vascular and Neuronal Changes Following AAV2-CMV-hEPO Injection
To test whether constant expression of hEPO by AAV2-CMV-hEPO was angiogenic, we performed FFA examination. Figure 7 showed the characteristic leakage of retinal vessels in the AAV2-CMV-hVEGF–treated group 2 and 4 weeks after subretinal injection, which served as a positive control, while in the PBS and AAV2-CMV-hEPO groups no leakage was detected, indicating the integrity of retinal blood vessels. 
Figure 7
 
Fluorescein angiography in DA rats 2 and 4 weeks after subretinal injection in three groups. There was significant leakage from the retinal vessels in the AAV2-CMV-hVEGF–treated group; while in PBS- and AAV2-CMV-hEPO–treated groups, there was no leakage. The arrows indicated the leakage areas (n = 5).
Figure 7
 
Fluorescein angiography in DA rats 2 and 4 weeks after subretinal injection in three groups. There was significant leakage from the retinal vessels in the AAV2-CMV-hVEGF–treated group; while in PBS- and AAV2-CMV-hEPO–treated groups, there was no leakage. The arrows indicated the leakage areas (n = 5).
Retinal flat mounts also were stained with isolectin to confirm the retinal vascular changes (Fig. 8). The abnormalities of retinal blood vessels were found only in the AAV2-CMV-hVEGF group, which were located mainly around the virus-exposed region. To specify the regions, we used confocal microscopy to stratify the retinal blood vessels into 3 layers. As data showed in Figure 8, no obvious change was detected in the innermost vascular plexus in 3 groups, but in the intermediate and outermost vasculatures, the density of retinal blood vessels in the AAV2-CMV-hVEGF group was higher than that in the other two groups, and the vessels also were distorted and irregular. However, in the AAV2-CMV-hEPO group, the retinal vessels appeared normal and no ectopic vessels were found in any retinal vascular plexus (Figs. 8G–I). 
Figure 8
 
Vascular changes at different levels of the retinal vascular plexuses with confocal microscope in three groups. Six months after subretinal injection, the rats were killed and the retinas were stained with isolectin. (A, D, G) The innermost vascular plexus. (B, E, H) The intermediate vascular plexus. (C, F, I) The outermost vascular plexus. (AC) PBS group (n = 10). (DF) AAV2-CMV-hVEGF group (n = 10). (GI) AAV2-CMV-hEPO group (n = 8). Scale bar: 100 μm.
Figure 8
 
Vascular changes at different levels of the retinal vascular plexuses with confocal microscope in three groups. Six months after subretinal injection, the rats were killed and the retinas were stained with isolectin. (A, D, G) The innermost vascular plexus. (B, E, H) The intermediate vascular plexus. (C, F, I) The outermost vascular plexus. (AC) PBS group (n = 10). (DF) AAV2-CMV-hVEGF group (n = 10). (GI) AAV2-CMV-hEPO group (n = 8). Scale bar: 100 μm.
We also quantitated the occurrence rate of retinal neovascularization (RNV) and choroidal neovascularization (CNV) in 3 groups (see Table). The data showed that in the AAV2-CMV-hVEGF group, RNV was detected in all the retinal flat mounts (n = 10, 100%), while no RNV was detected in the PBS-injected (n = 10) and AAV2-CMV-hEPO–injected (n = 8) groups. Since the trauma-induced CNV is inevitable, the occurrence of CNV is approximately 20%, 90%, and 12.5%, respectively, in the PBS, AAV2-CMV-hVEGF, and AAV2-CMV-hEPO groups. 
Table
 
Quantitation of RNV and CNV in Three Groups 6 Months After Subretinal Injection
Table
 
Quantitation of RNV and CNV in Three Groups 6 Months After Subretinal Injection
Groups Eyes, n RNV, n CNV, n RNV, % CNV, %
PBS 10 0 2 0 20
AAV2-CMV-hVEGF 10 10 9 100 90
AAV2-CMV-hEPO 8 0 1 0 12.5
To test the influence of sustained hEPO expression on retinal neuronal functions, we performed ERG examination 1 year after subretinal administration. The data showed no significant difference among three groups in terms of the changes of a-wave or b-wave amplitudes with three different bright flashes (Fig. 9); however, the amplitude of b-wave in the AAV2-CMV-hVEGF group seemed to be lower than that in the other two groups. This result indicated that long-term expression of hEPO by AAV2-CMV-hEPO did not affect the electrophysiologic function of retina to the light response. 
Figure 9
 
The ERG evaluation of the rats one year after subretinal injection. (A) The a-wave amplitude of ERG in 3 groups. (B) The b-wave amplitude of ERG in 3 groups. (C) The representative results of ERG in 3 groups. Three bright flashes were used for stimulation, that is, 6.325 × e−4 cd × s/m2, 6.325 × e−3 cd × s/m2, and 6.325 × e−2 cd × s/m2. The PBS group (n = 12 eyes), AAV2-CMV-hVEGF group (n = 4 eyes), and AAV2-CMV-hEPO group (n = 10 eyes).
Figure 9
 
The ERG evaluation of the rats one year after subretinal injection. (A) The a-wave amplitude of ERG in 3 groups. (B) The b-wave amplitude of ERG in 3 groups. (C) The representative results of ERG in 3 groups. Three bright flashes were used for stimulation, that is, 6.325 × e−4 cd × s/m2, 6.325 × e−3 cd × s/m2, and 6.325 × e−2 cd × s/m2. The PBS group (n = 12 eyes), AAV2-CMV-hVEGF group (n = 4 eyes), and AAV2-CMV-hEPO group (n = 10 eyes).
Discussion
Diabetic retinopathy (DR), affecting the microvasculature and neurons of the retina, is the most severe ocular complication of diabetes. 1 Previous studies have elucidated that intraocular administration of hEPO is an effective and safe procedure for treating DR. 15,17,18,41 The development of a therapy that can improve patient compliance, avoid the potential complications due to repeated intraocular injections, and provide a long-term, stable expression of hEPO without proangiogenic side effects in the eye, is of the vital importance. The AAV2 is particularly promising for retinal gene therapy because of its sustained, long-term expression of the transgene in various species. 43 To date, no toxicity and inflammatory response with AAV2 administration was reported in clinical 25 and experimental research. 44 Besides, in the eye, AAV vectors also could achieve similar prolonged transgene (such as EPO) expressions following subretinal injection. 34,4547 A series of studies on intravitreal hEPO in our laboratory has laid down the foundation for exploring AAV2-CMV-hEPO gene therapy to treat experimental DR. 15,17,18,4851 In this study, we confirmed the biological function of hEPO protein expressed by AAV2-CMV-hEPO virus, and studied the dose- and time-dependent expressions, the protective effects on diabetic retina, as well as the safety of AAV2-CMV-hEPO after subretinal injection. 
The AAV-mediated gene expression in the retina was affected by multiple factors, such as promoters, serotypes of AAVs, and injection routes, and so forth. 44 Previous studies have reported that AAV2-CMV-EGFP expression was detected as early as 2 weeks after subretinal injection (1 × 109 GC), becoming more robust at 4 and 6 weeks, and remained stable for over 180 days. 44,52,53 The AAV2 mainly targets photoreceptors and a small number of RPE cells after subretinal injection, but when administered intravitreally, AAV2 could efficiently infect retinal ganglion cells (RGCs), Müller cells, the ciliary body, and cells in the INL, presumably amacrine cells. 44,52 In the present study, we used AAV2-CMV-EGFP as a reporter vector and compared the infection efficiency with 2 delivery routes (intravitreal injection versus subretinal injection). Our results showed that EGFP could be detected in RPE cells as early as 4 days after subretinal injection (3 × 109 GC, data not shown), but not in the intravitreal injection group (Supplementary Fig. S2). The cryosections of the retina also revealed that EGFP expression was much more prominent in the subretinal injection group than that in the intravitreal group (Supplementary Fig. S1). This indicated that subretinal administration can achieve higher infection efficiency in RPE than intravitreal administration. Lebherz et al. 44 also reported that subretinal injection could result in 5- to 10-fold higher expression levels in retina than intravitreal injection. The relative lower infection efficiency with intravitreal injection might be due to the dilution of the injected vectors by vitreous, and the barrier function of inner limiting membrane (ILM) to the vectors. 
The EPO, with a structure of glycoprotein that acts as a major regulator of erythropoiesis, was reported recently to have a potent neuroprotective effect in retina. Therefore, erythropoietic function may be used as a biological marker of the epo transgene technique. Before examination of AAV2-CMV-hEPO efficacy in rats, we tested the erythropoietic function of this vector in C57BL/6 mice after intramuscular injection. The result showed that Hct in AAV2-CMV-hEPO–injected mice were significantly increased by 72%, 84%, and 85% at 4, 8, and 12 weeks after intramuscular injection, respectively (6 × 1010 GC, Fig. 1). The increased Hct was accompanied by the upregulation of serum hEPO protein, which indicated that the secreted hEPO protein by AAV2-CMV-hEPO has the biological function, that is, the production of red blood cells. Johnston et al. 53 reported that 4 weeks after intramuscular administration of AAV-CMV-mEPO (5 × 1010 GC), Hct was increased by 66% in mice. In our study, we used a slightly higher dose of AAV-CMV-hEPO (6 × 1010 GC); Hct was increased by 72% at the same time point. To further confirm the biological activity of expressed hEPO by AAV2-CMV-hEPO, CFU-E colony formation assay was performed in vitro. The CFU-Es are precursory cells to generate mature erythrocytes and their marked feature is responding specifically to EPO. 54 The in vitro results showed that hEPO in rat aqueous humor and in the supernatant of rat R28 cells, when treated with AAV2-CMV-hEPO virus, significantly increased the numbers of the CFU-E colonies (Supplementary Fig. S3), indicating the expressed hEPO in our system has the biological function, that is, erythropoiesis. These results indicated that hEPO produced by AAV-CMV-hEPO has the erythropoietic function, which laid down the foundation for its efficacy study in diabetic rats. 
In this study, we observed a long-term aqueous hEPO expression after subretinal injection (Fig. 3). Until the end point of observation, we detected the high levels of aqueous hEPO for more than 2 years in rats. Previous studies also reported long-term EPO level detected in aqueous humor after subretinal injection in nonhuman primates. 26,35 When we injected different doses of the AAV2-CMV-hEPO virus, a dose-dependent response was noticed, that is, the higher dose of vector achieved the higher expression level of hEPO protein (Fig. 2). In our previous study, the optimal dose of hEPO protein to treat experimental rat DR is 50 to 200 ng/eye (equivalent to 500–2000 mU/mL). 15 In our recent clinical study, we used 1000 mU/mL hEPO, which could effectively reduce the macular edema and improve the visual acuity in patients with DME. 18 In this study, the mean vitreous hEPO concentration of the first three groups (E1–E3) can reach 1178 mU/mL. So we chose these three doses of vectors to treat DR in SD rats. 
The condition of DR is characterized by BRB breakdown and neuronal degeneration. 1,9,10 In this study, we detected an elevated Evans blue permeation in diabetic rat retina, indicating the breakdown of the inner BRB (Fig. 4). The leakage of the Evans blue was significantly reduced when diabetic rats were treated with subretinal AAV2-CMV-hEPO virus for 4 weeks (Fig. 4). Meanwhile, increased apoptotic neurons in ONL were observed in diabetic rat retinas, and they were significantly reduced after treatment with AAV2-CMV-hEPO virus or hEPO protein (Fig. 5). The similar effects also were observed in retinal thickness measurement and cell counts (Fig. 6). Compared to other groups (E1, E3, E0), the E2 group seems to have relatively less leakage of Evans blue, less TUNEL-positive cells, higher retinal thickness, and more cell numbers, though there was no significant difference between each other. We also used this dose (E2) to test its effect on retinal function in DR rats, and ERG results showed that the retinal function (b-wave amplitude) was slightly restored to approximately 70% of that in normal rats (data not shown). The above data indicated that AAV2-CMV-hEPO gene therapy has the protective effect on DR In our previous study, we reported that hEPO might exert its protection on BRB function and retinal neurons in diabetic rats through a negative feedback mechanism in the HIF1α pathway, which down-regulated mRNA and protein levels of HIF-1α, VEGF-A, and endogenous EPO. 48 A recent report also showed that intravitreal injection of EPO protects against retinal vascular regression at the early stage of DR in STZ-induced diabetic rats through its downregulations on EPO receptor, VEGF, and VEGF receptor. 41 Combining these data, it was extrapolated that sustained expression of hEPO by AAV2-CMV-hEPO to maintain BRB integrity and protect retinal neurons might also be through the downregulation of VEGF. 
Since the high concentration of EPO was found in patients with PDR, 23 EPO has been considered as an independent angiogenic factor in the development of PDR, 23 with some experimental evidence. For instance, EPO could induce retinal endothelial cell angiogenesis in vitro, 55 and intraocular injections of soluble EPO receptor reduced retinal neovascularization in 19-day-old mice with retinopathy of prematurity (ROP). 23 In addition, the T allele of SNP rs1617640 in the promoter of the epo gene was reported to be significantly associated with PDR. 56 All these data lead to the concern that EPO is an angiogenic factor and it might be harmful when used in the eye. However, we and others reported that to rabbit eyes, intravitreal hEPO protein did not cause side effects, such as retinal vasculature (angiogenesis or vessel leakage), retinal anatomy, or ERG function, even with a high dose up to 1000 U over 6 months. 17,57 There is a report that a single EPO injection of 2000 U to three patients with acute vascular occlusion of the posterior pole did not induce obvious side effect. 58 Evidence exists that EPO was safe and effective in treating patients with chronic and progressive DME. 18 In the current study, we also examined whether or not the constant expression of hEPO by AAV2-CMV-hEPO virus, which generated an equivalent level of EPO protein in the vitreous of patients with PDR, has a pathologically angiogenic effect in the eye. We used AAV2-CMV-hVEGF as a positive control, because previous reports showed that subretinal injection of adenovirus-mediated VEGF 59 or AAV-mediated VEGF 60 induced CNV and RNV in experimental animals. In AAV2-CMV-hVEGF–treated eyes, we detected significant fluorescein leakage (Fig. 7) and obvious retinal vessel abnormalities in intermediate and outermost retinal vascular plexuses (Figs. 8E, 8F) with high occurrence of VEGF-induced RNV (100%) and CNV (90%, see Table). In contrast, RNV and CNV occurrence was essentially none/less in the PBS- and AAV2-CMV-hEPO–injected groups, indicating that the constant expression of hEPO protein by AAV2-CMV-hEPO does not induce pathologic angiogenesis in the eye under the current condition. Based on a previous report, the median EPO level in vitreous of patients with PDR was 464.0 mU/mL (36.5 mU/mL in control patients with nondiabetic ocular diseases). 23 The sustained and high expression level of hEPO (2763.1 mU/mL in the vitreous) in the present study, which is approximately 6 times that in patients with PDR (Fig. 3), still is safe and effective. Therefore, we proposed that the safety of constantly high EPO in eyes is DR stage-dependent, in other words, a dependency of the hypoxic status of retina. The similar concept was introduced in the EPO regimen for ROP. 61,62  
In summary, our data supported the notion that AAV2-CMV-hEPO gene therapy is effective for the long-term protection of BRB and retinal neurons in experimental DR rats. The subretinal administration is a safe procedure, which is an effective method for prolonged expression of hEPO. Based on these findings, we propose that AAV2 vector carrying EPO gene or other specific genes could be a potential therapeutic modality for long-term preventive or adjunctive therapy for DR or other retinal degenerative diseases. 
Supplementary Materials
Acknowledgments
Supported by the following research grants: the National Key Basic Research Program of China (2011CB965102, 2012CBA01308, and 2013CB967501), National Natural Science Foundation of China (31171419 and 81000383), National High Technology Research and Development Program of China (2012AA020906), International Collaboration Fund (2011DFB30010), Research Fund for the Doctoral Program of Higher Education of China (20100072120051), Science and Technology Commission of Shanghai (11DZ1920904), the Fundamental Research Funds for the Central Universities (2009KJ109), and Program of Tongji University School of Medicine (2010QH04 and 2010YF02). 
Disclosure: H. Xu, None; L. Zhang, None; L. Gu, None; L. Lu, None; G. Gao, None; W. Li, None; G. Xu, None; J. Wang, None; F. Gao, None; J.-Y. Xu, None; J. Yao, None; F. Wang, None; J. Zhang, None; G.-T. Xu, None 
References
Gardner TW Antonetti DA Barber AJ LaNoue KF Levison SW. Diabetic retinopathy: more than meets the eye. Surv Ophthalmol . 2002; 47 (suppl 2): S253–S262. [CrossRef] [PubMed]
Kollias AN Ulbig MW. Diabetic retinopathy: early diagnosis and effective treatment. Dtsch Arztebl Int . 2010; 107: 75–83, quiz 84. [PubMed]
Simo R Hernandez C. Advances in the medical treatment of diabetic retinopathy. Diabetes Care . 2009; 32: 1556–1562. [CrossRef] [PubMed]
Yasukawa T Tabata Y Kimura H Ogura Y. Recent advances in intraocular drug delivery systems. Recent Pat Drug Deliv Formul . 2011; 5: 1–10. [CrossRef] [PubMed]
Chung EJ Kang SJ Koo JS Choi YJ Grossniklaus HE Koh HJ. Effect of intravitreal bevacizumab on vascular endothelial growth factor expression in patients with proliferative diabetic retinopathy. Yonsei Med J . 2011; 52: 151–157. [CrossRef] [PubMed]
Nicholson BP Schachat AP. A review of clinical trials of anti-VEGF agents for diabetic retinopathy. Graefes Arch Clin Exp Ophthalmol . 2010; 248: 915–930. [CrossRef] [PubMed]
Mitchell P. A systematic review of the efficacy and safety outcomes of anti-VEGF agents used for treating neovascular age-related macular degeneration: comparison of ranibizumab and bevacizumab. Curr Med Res Opin . 2011; 27: 1465–1475. [CrossRef] [PubMed]
Cheung N Mitchell P Wong TY. Diabetic retinopathy. Lancet . 2010; 376: 124–136. [CrossRef] [PubMed]
Lieth E Gardner TW Barber AJ Antonetti DA. Retinal neurodegeneration: early pathology in diabetes. Clin Experiment Ophthalmol . 2000; 28: 3–8. [CrossRef] [PubMed]
Barber AJ. A new view of diabetic retinopathy: a neurodegenerative disease of the eye. Prog Neuropsychopharmacol Biol Psychiatry . 2003; 27: 283–290. [CrossRef] [PubMed]
Whitmire W Al-Gayyar MM Abdelsaid M Yousufzai BK El-Remessy AB. Alteration of growth factors and neuronal death in diabetic retinopathy: what we have learned so far. Mol Vis . 2011; 17: 300–308. [PubMed]
Marti HH Wenger RH Rivas LA Erythropoietin gene expression in human, monkey and murine brain. Eur J Neurosci . 1996; 8: 666–676. [CrossRef] [PubMed]
Hernandez C Fonollosa A Garcia-Ramirez M Erythropoietin is expressed in the human retina and it is highly elevated in the vitreous fluid of patients with diabetic macular edema. Diabetes Care . 2006; 29: 2028–2033. [CrossRef] [PubMed]
Juul SE Yachnis AT Christensen RD. Tissue distribution of erythropoietin and erythropoietin receptor in the developing human fetus. Early Hum Dev . 1998; 52: 235–249. [CrossRef] [PubMed]
Zhang J Wu Y Jin Y Intravitreal injection of erythropoietin protects both retinal vascular and neuronal cells in early diabetes. Invest Ophthalmol Vis Sci . 2008; 49: 732–742. [CrossRef] [PubMed]
Zhu B Wang W Gu Q Xu X. Erythropoietin protects retinal neurons and glial cells in early-stage streptozotocin-induced diabetic rats. Exp Eye Res . 2008; 86: 375–382. [CrossRef] [PubMed]
Zhang JF Wu YL Xu JY Pharmacokinetic and toxicity study of intravitreal erythropoietin in rabbits. Acta Pharmacol Sin . 2008; 29: 1383–1390. [CrossRef] [PubMed]
Li W Sinclair SH Xu GT. Effects of intravitreal erythropoietin therapy for patients with chronic and progressive diabetic macular edema. Ophthalmic Surg Lasers Imaging . 2010; 41: 18–25. [CrossRef] [PubMed]
Jager RD Aiello LP Patel SC Cunningham ET Jr. Risks of intravitreous injection: a comprehensive review. Retina . 2004; 24: 676–698. [CrossRef] [PubMed]
Stieger K Mendes-Madeira A Meur GL Oral administration of doxycycline allows tight control of transgene expression: a key step towards gene therapy of retinal diseases. Gene Ther . 2007; 14: 1668–1673. [CrossRef] [PubMed]
Ribatti D Vacca A Roccaro AM Crivellato E Presta M. Erythropoietin as an angiogenic factor. Eur J Clin Invest . 2003; 33: 891–896. [CrossRef] [PubMed]
Watanabe D. [Erythropoietin as a retinal angiogenic factor in proliferative diabetic retinopathy]. Nihon Ganka Gakkai Zasshi . 2007; 111: 892–898. [PubMed]
Watanabe D Suzuma K Matsui S Erythropoietin as a retinal angiogenic factor in proliferative diabetic retinopathy. N Engl J Med . 2005; 353: 782–792. [CrossRef] [PubMed]
Amado D Mingozzi F Hui D Safety and efficacy of subretinal readministration of a viral vector in large animals to treat congenital blindness. Sci Transl Med . 2010; 2: 21ra 16. [CrossRef]
Simonelli F Maguire AM Testa F Gene therapy for Leber's congenital amaurosis is safe and effective through 1.5 years after vector administration. Mol Ther . 2010; 18: 643–650. [CrossRef] [PubMed]
Lebherz C Auricchio A Maguire AM Long-term inducible gene expression in the eye via adeno-associated virus gene transfer in nonhuman primates. Hum Gene Ther . 2005; 16: 178–186. [CrossRef] [PubMed]
Buch PK Bainbridge JW Ali RR. AAV-mediated gene therapy for retinal disorders: from mouse to man. Gene Ther . 2008; 15: 849–857. [CrossRef] [PubMed]
Roy K Stein L Kaushal S. Ocular gene therapy: an evaluation of recombinant adeno-associated virus-mediated gene therapy interventions for the treatment of ocular disease. Hum Gene Ther . 2010; 21: 915–927. [CrossRef] [PubMed]
Colella P Auricchio A. AAV-mediated gene supply for treatment of degenerative and neovascular retinal diseases. Curr Gene Ther . 2010; 10: 371–380. [CrossRef] [PubMed]
Lai CM Shen WY Brankov M Long-term evaluation of AAV-mediated sFlt-1 gene therapy for ocular neovascularization in mice and monkeys. Mol Ther . 2005; 12: 659–668. [CrossRef] [PubMed]
Bainbridge JW Smith AJ Barker SS Effect of gene therapy on visual function in Leber's congenital amaurosis. N Engl J Med . 2008; 358: 2231–2239. [CrossRef] [PubMed]
Cideciyan AV Aleman TS Boye SL Human gene therapy for RPE65 isomerase deficiency activates the retinoid cycle of vision but with slow rod kinetics. Proc Natl Acad Sci U S A . 2008; 105: 15112–15117. [CrossRef] [PubMed]
Maguire AM Simonelli F Pierce EA Safety and efficacy of gene transfer for Leber's congenital amaurosis. N Engl J Med . 2008; 358: 2240–2248. [CrossRef] [PubMed]
Auricchio A Rivera VM Clackson T Pharmacological regulation of protein expression from adeno-associated viral vectors in the eye. Mol Ther . 2002; 6: 238–242. [CrossRef] [PubMed]
Stieger K Le Meur G Lasne F Long-term doxycycline-regulated transgene expression in the retina of nonhuman primates following subretinal injection of recombinant AAV vectors. Mol Ther . 2006; 13: 967–975. [CrossRef] [PubMed]
Buemi M Caccamo C Nostro L Cavallaro E Floccari F Grasso G. Brain and cancer: the protective role of erythropoietin. Med Res Rev . 2005; 25: 245–259. [CrossRef] [PubMed]
Latini R Brines M Fiordaliso F. Do non-hemopoietic effects of erythropoietin play a beneficial role in heart failure? Heart Fail Rev . 2008; 13: 415–423. [CrossRef] [PubMed]
Monge Naldi A, Gassmann M, Bodmer D. Erythropoietin but not VEGF has a protective effect on auditory hair cells in the inner ear. Cell Mol Life Sci . 2009; 66: 3595–3599. [CrossRef] [PubMed]
Moore EM Bellomo R Nichol AD. Erythropoietin as a novel brain and kidney protective agent. Anaesth Intensive Care . 2011; 39: 356–372. [PubMed]
Chateauvieux S Grigorakaki C Morceau F Dicato M Diederich M. Erythropoietin, erythropoiesis and beyond. Biochem Pharmacol . 2011; 82: 1291–1303. [CrossRef] [PubMed]
Mitsuhashi J Morikawa S Shimizu K Ezaki T Yasuda Y Hori S. Intravitreal injection of erythropoietin protects against retinal vascular regression at the early stage of diabetic retinopathy in streptozotocin-induced diabetic rats. Exp Eye Res . 2013; 106: 64–73. [CrossRef] [PubMed]
Zhang J Xu G Zhang L A modified histoimmunochemistry-assisted method for in situ RPE evaluation. Front Biosci (Elite Ed) . 2012; 4: 1571–1581. [CrossRef] [PubMed]
McClements ME MacLaren RE. Gene therapy for retinal disease. Transl Res . 2013; 161: 241–254. [CrossRef] [PubMed]
Lebherz C Maguire A Tang W Bennett J Wilson JM. Novel AAV serotypes for improved ocular gene transfer. J Gene Med . 2008; 10: 375–382. [CrossRef] [PubMed]
Igarashi T Miyake K Masuda I Takahashi H Shimada T. Adeno-associated vector (type 8)-mediated expression of soluble Flt-1 efficiently inhibits neovascularization in a murine choroidal neovascularization model. Hum Gene Ther . 2010; 21: 631–637. [CrossRef] [PubMed]
Yu H Chen L Jiang J. Administration of pigment epithelium-derived factor delivered by adeno-associated virus inhibits blood-retinal barrier breakdown in diabetic rats. Mol Vis . 2011; 16: 2384–2394.
Lai CC Wu WC Chen SL Recombinant adeno-associated virus vector expressing angiostatin inhibits preretinal neovascularization in adult rats. Ophthalmic Res . 2005; 37: 50–56. [CrossRef] [PubMed]
Zhang J Hu LM Xu G Anti-VEGF effects of intravitreal erythropoietin in early diabetic retinopathy. Front Biosci (Elite Ed) . 2010; 2: 912–927. [CrossRef] [PubMed]
Shen J Wu Y Xu JY ERK- and Akt-dependent neuroprotection by erythropoietin (EPO) against glyoxal-AGEs via modulation of Bcl-xL, Bax, and BAD. Invest Ophthalmol Vis Sci . 2010; 51: 35–46. [CrossRef] [PubMed]
Chu Q Zhang J Wu Y Differential gene expression pattern of diabetic rat retinas after intravitreal injection of erythropoietin. Clin Experiment Ophthalmol . 2011; 39: 142–151. [CrossRef] [PubMed]
Lei X Zhang J Shen J EPO attenuates inflammatory cytokines by Muller cells in diabetic retinopathy. Front Biosci (Elite Ed) . 2011; 3: 201–211. [CrossRef] [PubMed]
Auricchio A Kobinger G Anand V Exchange of surface proteins impacts on viral vector cellular specificity and transduction characteristics: the retina as a model. Hum Mol Genet . 2001; 10: 3075–3081. [CrossRef] [PubMed]
Rolling F. Recombinant AAV-mediated gene transfer to the retina: gene therapy perspectives. Gene Ther . 2004; 11 (suppl 1): S26–S32. [CrossRef] [PubMed]
Adamson JW. Regulation of red blood cell production. Am J Med . 1996; 101: 4S–6S. [CrossRef] [PubMed]
Haribalaganesh R Sheikpranbabu S Banumathi E Gurunathan S. Pigment epithelium-derived factor inhibits erythropoietin-induced retinal endothelial cell angiogenesis by suppression of PI3K/Akt pathway. Exp Eye Res . 2010; 90: 726–733. [CrossRef] [PubMed]
Tong Z Yang Z Patel S Promoter polymorphism of the erythropoietin gene in severe diabetic eye and kidney complications. Proc Natl Acad Sci U S A . 2008; 105: 6998–7003. [CrossRef] [PubMed]
Song BJ Cai H Tsai JC Chang S Forbes M Del Priore LV. Intravitreal recombinant human erythropoietin: a safety study in rabbits. Curr Eye Res . 2008; 33: 750–760. [CrossRef] [PubMed]
Lagreze WA Feltgen N Bach M Jehle T. Feasibility of intravitreal erythropoietin injections in humans. Br J Ophthalmol . 2009; 93: 1667–1671. [CrossRef] [PubMed]
Baffi J Byrnes G Chan CC Csaky KG. Choroidal neovascularization in the rat induced by adenovirus mediated expression of vascular endothelial growth factor. Invest Ophthalmol Vis Sci . 2000; 41: 3582–3589. [PubMed]
Wang F Rendahl KG Manning WC Quiroz D Coyne M Miller SS. AAV-mediated expression of vascular endothelial growth factor induces choroidal neovascularization in rat. Invest Ophthalmol Vis Sci . 2003; 44: 781–790. [CrossRef] [PubMed]
Chen J Connor KM Aderman CM Smith LE. Erythropoietin deficiency decreases vascular stability in mice. J Clin Invest . 2008; 118: 526–533. [PubMed]
Chen J Connor KM Aderman CM Willett KL Aspegren OP Smith LE. Suppression of retinal neovascularization by erythropoietin siRNA in a mouse model of proliferative retinopathy. Invest Ophthalmol Vis Sci . 2009; 50: 1329–1335. [CrossRef] [PubMed]
Footnotes
 G-TX, JZ, and FW contributed equally to the work presented here and should therefore be regarded as equivalent authors.
Figure 1
 
The biological activity assay of hEPO in vitro and in vivo. (A) Human EPO concentration was significantly increased 48 hours after infection to HEK 293T cells with AAV2-CMV-hEPO virus, which reached 5174.8 mU/mL (averaged hEPO concentration of 4 independent experiments). (B) The Hct was increased from 4 to 12 weeks after intramuscular injection of AAV2-CMV-hEPO in C57BL/6 mice (1 × 1013 GC/mL, 6 μL/mouse, n = 4/time point). (C) The increased Hct in AAV2-CMV-hEPO–injected mice was correlated with the increased serum hEPO protein (n = 4/time point). Each datum was expressed as mean ± SE. *P < 0.05 when compared to that in normal control.
Figure 1
 
The biological activity assay of hEPO in vitro and in vivo. (A) Human EPO concentration was significantly increased 48 hours after infection to HEK 293T cells with AAV2-CMV-hEPO virus, which reached 5174.8 mU/mL (averaged hEPO concentration of 4 independent experiments). (B) The Hct was increased from 4 to 12 weeks after intramuscular injection of AAV2-CMV-hEPO in C57BL/6 mice (1 × 1013 GC/mL, 6 μL/mouse, n = 4/time point). (C) The increased Hct in AAV2-CMV-hEPO–injected mice was correlated with the increased serum hEPO protein (n = 4/time point). Each datum was expressed as mean ± SE. *P < 0.05 when compared to that in normal control.
Figure 2
 
Human EPO concentrations in aqueous humor (A), vitreous (B), and retina (C) 4 weeks after subretinal injection of AAV2-CMV-hEPO in SD rats (3 μL/eye, n = 6). The doses of the virus vectors in 7 groups are as followings: (0) PBS control, (1) 3 × 1010 GC, (2) 9 × 109 GC, (3) 3 × 109 GC, (4) 3 × 108 GC, (5) 3 × 107 GC, and (6) 3 × 106 GC. Each datum was expressed as mean ± SE.
Figure 2
 
Human EPO concentrations in aqueous humor (A), vitreous (B), and retina (C) 4 weeks after subretinal injection of AAV2-CMV-hEPO in SD rats (3 μL/eye, n = 6). The doses of the virus vectors in 7 groups are as followings: (0) PBS control, (1) 3 × 1010 GC, (2) 9 × 109 GC, (3) 3 × 109 GC, (4) 3 × 108 GC, (5) 3 × 107 GC, and (6) 3 × 106 GC. Each datum was expressed as mean ± SE.
Figure 3
 
Aqueous hEPO expression follows a time-dependent manner after subretinal injection of AAV2-CMV-hEPO. The concentrations of hEPO in the aqueous were detected with ELISA at different time points after subretinal injection of AAV2-CMV-hEPO (1 × 1013 GC/mL, 3 μL, n = 6/group). Each datum was expressed as mean ± SE.
Figure 3
 
Aqueous hEPO expression follows a time-dependent manner after subretinal injection of AAV2-CMV-hEPO. The concentrations of hEPO in the aqueous were detected with ELISA at different time points after subretinal injection of AAV2-CMV-hEPO (1 × 1013 GC/mL, 3 μL, n = 6/group). Each datum was expressed as mean ± SE.
Figure 4
 
Evans blue-albumin permeation measurement treated with or without AAV2-CMV-hEPO. N, normal control; D, diabetic rats; E1 to E3, diabetic rats injected with 3 doses of AAV2-CMV-hEPO virus (E1, 1 × 1013 GC/mL; E2, 3 × 1012 GC/mL; and E3, 1 × 1012 GC/mL, 3 μL/eye and n = 8); E0, diabetic rats injected with hEPO protein (16 mU/eye). Each datum was expressed as mean ± SE. *P < 0.05 when compared to the diabetic group.
Figure 4
 
Evans blue-albumin permeation measurement treated with or without AAV2-CMV-hEPO. N, normal control; D, diabetic rats; E1 to E3, diabetic rats injected with 3 doses of AAV2-CMV-hEPO virus (E1, 1 × 1013 GC/mL; E2, 3 × 1012 GC/mL; and E3, 1 × 1012 GC/mL, 3 μL/eye and n = 8); E0, diabetic rats injected with hEPO protein (16 mU/eye). Each datum was expressed as mean ± SE. *P < 0.05 when compared to the diabetic group.
Figure 5
 
TUNEL assay in 1.5-month diabetic rat retina treated with or without AAV2-CMV-hEPO. N, normal control; D, diabetic rats; E2, diabetic rats injected with AAV2-CMV-hEPO (3 × 1012 GC/mL, 3 μL/eye); E0, diabetic rats injected with hEPO protein (16 mU/eye); NC, negative control; PC, positive control. n = 6/group. Scale bar: 50 μm.
Figure 5
 
TUNEL assay in 1.5-month diabetic rat retina treated with or without AAV2-CMV-hEPO. N, normal control; D, diabetic rats; E2, diabetic rats injected with AAV2-CMV-hEPO (3 × 1012 GC/mL, 3 μL/eye); E0, diabetic rats injected with hEPO protein (16 mU/eye); NC, negative control; PC, positive control. n = 6/group. Scale bar: 50 μm.
Figure 6
 
Protecting retinal thickness (A) and preventing loss of retinal neurons (B) in STZ-diabetic rat by AAV2-CMV-hEPO administration. N, normal control; D, diabetic rats; E1 to E3, diabetic rats injected with AAV2-CMV-hEPO (E1, 1 × 1013 GC/mL; E2, 3 × 1012 GC/mL; and E3, 1 × 1012 GC/mL, 3 μL/eye and n = 4); E0, diabetic rats injected with hEPO protein (16 mU/eye). Each datum was expressed as mean ± SE. *P < 0.05 when compared to the diabetic group.
Figure 6
 
Protecting retinal thickness (A) and preventing loss of retinal neurons (B) in STZ-diabetic rat by AAV2-CMV-hEPO administration. N, normal control; D, diabetic rats; E1 to E3, diabetic rats injected with AAV2-CMV-hEPO (E1, 1 × 1013 GC/mL; E2, 3 × 1012 GC/mL; and E3, 1 × 1012 GC/mL, 3 μL/eye and n = 4); E0, diabetic rats injected with hEPO protein (16 mU/eye). Each datum was expressed as mean ± SE. *P < 0.05 when compared to the diabetic group.
Figure 7
 
Fluorescein angiography in DA rats 2 and 4 weeks after subretinal injection in three groups. There was significant leakage from the retinal vessels in the AAV2-CMV-hVEGF–treated group; while in PBS- and AAV2-CMV-hEPO–treated groups, there was no leakage. The arrows indicated the leakage areas (n = 5).
Figure 7
 
Fluorescein angiography in DA rats 2 and 4 weeks after subretinal injection in three groups. There was significant leakage from the retinal vessels in the AAV2-CMV-hVEGF–treated group; while in PBS- and AAV2-CMV-hEPO–treated groups, there was no leakage. The arrows indicated the leakage areas (n = 5).
Figure 8
 
Vascular changes at different levels of the retinal vascular plexuses with confocal microscope in three groups. Six months after subretinal injection, the rats were killed and the retinas were stained with isolectin. (A, D, G) The innermost vascular plexus. (B, E, H) The intermediate vascular plexus. (C, F, I) The outermost vascular plexus. (AC) PBS group (n = 10). (DF) AAV2-CMV-hVEGF group (n = 10). (GI) AAV2-CMV-hEPO group (n = 8). Scale bar: 100 μm.
Figure 8
 
Vascular changes at different levels of the retinal vascular plexuses with confocal microscope in three groups. Six months after subretinal injection, the rats were killed and the retinas were stained with isolectin. (A, D, G) The innermost vascular plexus. (B, E, H) The intermediate vascular plexus. (C, F, I) The outermost vascular plexus. (AC) PBS group (n = 10). (DF) AAV2-CMV-hVEGF group (n = 10). (GI) AAV2-CMV-hEPO group (n = 8). Scale bar: 100 μm.
Figure 9
 
The ERG evaluation of the rats one year after subretinal injection. (A) The a-wave amplitude of ERG in 3 groups. (B) The b-wave amplitude of ERG in 3 groups. (C) The representative results of ERG in 3 groups. Three bright flashes were used for stimulation, that is, 6.325 × e−4 cd × s/m2, 6.325 × e−3 cd × s/m2, and 6.325 × e−2 cd × s/m2. The PBS group (n = 12 eyes), AAV2-CMV-hVEGF group (n = 4 eyes), and AAV2-CMV-hEPO group (n = 10 eyes).
Figure 9
 
The ERG evaluation of the rats one year after subretinal injection. (A) The a-wave amplitude of ERG in 3 groups. (B) The b-wave amplitude of ERG in 3 groups. (C) The representative results of ERG in 3 groups. Three bright flashes were used for stimulation, that is, 6.325 × e−4 cd × s/m2, 6.325 × e−3 cd × s/m2, and 6.325 × e−2 cd × s/m2. The PBS group (n = 12 eyes), AAV2-CMV-hVEGF group (n = 4 eyes), and AAV2-CMV-hEPO group (n = 10 eyes).
Table
 
Quantitation of RNV and CNV in Three Groups 6 Months After Subretinal Injection
Table
 
Quantitation of RNV and CNV in Three Groups 6 Months After Subretinal Injection
Groups Eyes, n RNV, n CNV, n RNV, % CNV, %
PBS 10 0 2 0 20
AAV2-CMV-hVEGF 10 10 9 100 90
AAV2-CMV-hEPO 8 0 1 0 12.5
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×