Human eye tissue
Human retinas were collected from 2 patients with end-stage glaucoma and 1 deceased healthy donor after eye enucleation (Table S1). Retinas were routinely fixed, dehydrated, embedded, and sectioned (5 µm) for later analysis of HE staining, immunofluorescent/ immunohistochemical staining or transmission electron microscopy (TEM). All human retinal tissues were obtained from the Eye Center, The Second Affiliated Hospital, School of Medicine, Zhejiang University between February 2021 and September 2021. Acquisitions and experiments involving human tissue were performed in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki) and were approved by the Human Research Ethics Committee of the Second Affiliated Hospital, School of Medicine, Zhejiang University (No. 120211118).
Animals
Male SD rats weighing 200 ~ 250g were purchased from Vital River Laboratory Animal Technology Co., Ltd (Beijing, PRC) and housed in illumination (12: 12 h light: dark cycle) and temperature-controlled and pathogen-free facilities with free access to clean water and food. All animal care and experiments were approved by the Laboratory Animal Ethics Committee of the Second Affiliated Hospital, School of Medicine, Zhejiang University (No.2023-004) and conducted in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research.
Purification of primary RGCs
Primary RGCs were purified and cultured in accordance with our previous study(27). Briefly, the retina isolated from P7 mice (C57BL/6, purchased from the Charles River Laboratories) was dissected under a microscope and digested to single-cell by papain. For purification, the cell suspension was processed through three-step panning: (i) incubation in a negative panning plate coated by the lectin for 30 min (shake the dish gently per 15 min); (ii) incubation in another lectin-coated dish for 10 min; (iii) incubation in a positive panning plate coated with anti-mouse-Thy1.2 antibody (Bio-Rad Laboratories, USA) for 45 min (shake the dish per 15 min). Cells attached to the plate were digested with trypsin and collected. After resuspension, purified RGCs were seeded on glass coverslips coated with mouse Laminin (R&D Systems, USA) and Poly-D-Lysine (Sigma-Aldrich, USA) and cultured by RGC growth medium (equivalent Neurobasal Medium and DMEM containing supplemental factors). We maintained RGCs in a humidified incubator containing 5% CO2 and 95% air at 37 ℃ and performed following interventions 2 days later.
Rat model of chronic ocular hypertension
Polystyrene beads (Polysciences, USA) were adequately shaken, dropped out and centrifuged at 2,000 rpm for 1 min. After removing the supernatant, a sufficient volume of 75% ethanol was added for vortexing, and then the mixture was centrifuged and repeatedly washed 2 times. The 6 µm and 10 µm diameter beads were dispersed in normal saline (NS) containing 0.1% carbomer (molecular weight 4 mDa) at a concentration of 1×107/mL. Before injection, 5 µL of 6 µm diameter beads, 5 µL of 10 µm diameter beads, and 2 µL of air were taken up into a microsyringe (Hamilton, Switzerland) through a 33-gauge needle in turn.
After general anesthesia (1% sodium pentobarbital, 5 mL/kg), proparacaine eye drops (Alcon, USA) were applied for local anesthesia. Normal sterilization and anterior chamber (AC) paracentesis were conducted with a 32-gauge needle before the microbeads were quickly injected into the central AC with the prepared microsyringe through the canal. Finally, the bubbles were adjusted to the inner mouth of the canal, and the needle was withdrawn 1 min later. Levofloxacin eye drops (Santen, Japan) were applied for disinfection. TonoLab (Icare, Finland) was used as an IOP monitor without anesthesia. The microbead injection was repeated under nonideal IOP elevation at Day 7. Eyes with endophthalmitis, intraocular bleeding, corneal perforation, or obvious anterior chamber leakage were excluded.
Oxygen/glucose deprivation (OGD)
For Müller cells (MIO-M1 cell line), 0.3g/L glucose was formulated with glucose and DMEM (glucose free, Thermo Fisher Scientific, USA), and applied to Müller cells in a 0.5% O2, 5% CO2 atmosphere for 22 h. For RGCs or Müller cell-RGC co-cultured system, 0.3g/L glucose was formulated with RGC growth medium and DMEM, and cultured in a 0.5% O2, 5% CO2 atmosphere for 4 h. Cells were incubated with fluorescent probes for living image or changed into the environment with proper glucose concentration and atmosphere for Transwell or scratch wounding assays. The cell line was tested to be free of mycoplasma contamination and was authenticated via STR profiling.
Live/dead assay
Using Viability/Cytotoxicity Assay Kit (US EVERBRIGHT, PRC), cells were gently washed with PBS once and stained with Calcein AM (2 µM) and Ethidium homodimer-1 (EthD-I, 4 µM) for 15 min at 37°C. Finally, the images of live (green) and dead (red) cells were observed under a fluorescence microscope (Leica DMi8, Germany).
ROS measurement
ROS was detected with dihydroethidium (DHE) (Beyotime, PRC). RGCs were loaded with 5 mM DHE for 30 min at 37 ℃ while frozen sections were loaded with 200mM DHE for 30 min at room temperature (RT) following 3 times PBS wash and DAPI Fluoromount-G™ (Yeasen Biotech, PRC) staining. RGCs or sections were observed under a fluorescence microscope (Leica DMi8 for RGCs or DM6B for sections).
Scratch wounding migration.
Müller cells were seeded in a 6-well plate at a density of 5 × 105 cells/well in 2 mL of medium overnight. The confluent cell monolayers were scratched by a sterile 200-µL micropipette tip and washed with PBS. After OGD, the medium was changed to a complete medium without NAC or NADPH. Cell migration was examined using a light microscope (Leica DMi1) at scratching or 24 h after OGD, and the scratch widths were quantified by ImageJ software (NIH, USA).
Transwell assay
The second day of RGC purification, Müller cells (1 × 104) were seeded on an 8 µm pore diameter-Transwell and cultured overnight upon RGCs (inner Transwell, 100mL, outer Transwell, 300mL). After OGD, supernatants from RGCs were half removed, resulting in 150mL remaining, and 150mL DMEM with extra glucose was added to outer Transwell forming a 1g/L glucose concentration. Inner Transwell was treated similarly to build a 1g/L glucose concentration. Additionally, 3 µL fetal bovine serum (FBS) was pipetted to outer Transwell in order to accelerate the migration process. Migration after OGD lasts 24 h under 5% CO2 and 95% air. Furthemore, Müller cells without pre-OGD treatment were treated by supernatants from 6 h re-oxygenated RGCs and performed this migration exam again. Müller cells were fixed with 4% paraformaldehyde (PFA) for 15 min washed with PBS and stained with crystal violet (Beyotime). Cells inner Transwell were abrased with cotton swabs and migrated cells were imaged by a Leica DMi1 light microscope. Cells were quantified by ImageJ software.
Rat eyeballs processing for morphological examination
After being sacrificed, eyes or optic nerves were enucleated under a stereomicroscope (Leica S9i) and fixed with 4% PFA following eradication of cornea, iris, lens and extraocular muscle tissue reserving posterior segments (eye cups). For the paraffin section, eye cups or optic nerves were normal dehydration (Leica ASP300S) and paraffin embedded. 5-µm thickness paraffin sections were cut parallel with the optic axis through the optic disks or traverse optic nerves 1 mm from eyeballs. Sections were stained with hematoxylin and eosin (H&E) or dewaxed, rehydrated and boiling citrate buffer incubated for antigen retrieval. For the frozen section, eye cups were dehydrated in 5%, 12.5% and 20% sucrose (m/v, in PBS) and embedded in an equivalent OCT compound and 20% sucrose. 8-µm thickness frozen sections were cut paralleled with the optic axis through the optic disks and used for staining immediately. For retinal mounts, retinas were gently torn against sclera tissues and cut into four-leaf clovers unfolding on slides. Then cold methanol was dropped for degreasing and flat mounts were immersed and reserved in methanol at -20 ℃.
In the H&E-stained section, 4 standard areas of each sample were used for cell quantification in RGCL. Every standard area contained a 200 µm-length retina which was located at 1mm from the optic disk and 2 researchers blind to rat grouping calculated the cell number in RGCL independently.
In retinal flat mount, fluorescent signal was imaged once every 0.5 µm step on z-stack, 3D structure reconstruction and signal colocalization were conducted with Imaris 9.0.1 (Oxford Instrument, UK) and MATLAB Compiler Runtime V8.4 (Mathworks, USA).
Immunohistochemistry (IHC)
After deparaffinization and antigen retrieval, sections were washed with PBS once and immersed in 3% H2O2 for 10 min at RT and blocked with normal goat serum (NGS) for 1 h at RT. Thereafter, sections were incubated at 4°C overnight with the following primary antibodies. Sections were then washed with PBS and incubated with horseradish peroxidase (HRP)–conjugated goat anti-rabbit IgG polyclonal antibodies for 23 min at RT. The sections were stained with 3,3’- diaminobenzidine and counterstained with hematoxylin. Then, sections were washed with water and mounted in neutral balsam. All slides were imaged with a light microscope (Leica DM2500).
Immunofluorescent (IF) staining
The frozen sections or RGCs cultured on coverslips were fixed with 4% PFA for 10 min and permeabilized with 0.3% Triton X-100 and 0.2% bovine serum albumin (BSA) in PBS for 1 h at RT. The retinal mounts and antigen-retrieved paraffin sections were washed with PBS once and permeabilized the same as frozen sections. After rinsed 3 times in PBS, they were blocked in buffer (5% NGS, 2% BSA, and 0.1% Tween-20 in PBS) for 1 h at RT. Then they were incubated with primary antibody (Table S2) at 4°C overnight and species-specific secondary fluorescent antibodies for 4 h (retina mounts) or 1 h (RGCs and sections) at RT. After being washed with PBS, samples were stained with DAPI and imaged with a Leica DM6B fluorescence microscope or a Leica SP8 confocal microscope.
TUNEL assay
Terminal transferase deoxyuridine triphosphate nick end labeling (TUNEL) assay was performed using frozen sections according to the manufacturer’s methods of One-Step TUNEL Apoptosis Assay Kit (Beyotime, China). Finally, the sections were stained with DAPI and observed under a Leica DMi8 microscope. TUNEL-positive cells were labeled in a standard area same as RGCL cell counting.
Western blot (WB)
Total protein was extracted by Total Protein Extraction Kits (Solarbio, PRC) with lysis buffer 60 µL/well (6-well plate) or 300 µL/rat retina. Retina should be 270 W ultrasonicated for 2 s twice by ultrasonic disruptor (Scientz, PRC) when lysis buffer was added. Protein concentration was quantified by a Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific, Massachusetts, USA) and diluted with lysis buffer to equalization and the final resolution was mixed with loading buffer (Sangon Biotech, PRC).
Equal amounts of protein were electrophoresed on 12% or 4%~20% SDS PAGE Gels (ACE Biotechnology, PRC) and transferred to PVDF membranes (Millipore, USA). Membranes were then blocked with Protein Free Rapid Blocking Buffer (YaMei, China) for 1 h and probed with primary antibodies at 4°C overnight and secondary antibodies for 1 h at RT. The blots were cut and separated into sections for different antibody probing (Table S2). Protein expression was detected in a chemiluminescence solution (Bio-Rad, USA) by Chemiluminescence imaging system (Bio-Rad) and analyzed via densitometry with ImageLab software. To calculate the relative expression of a specific protein, the nonphosphorylated form or GAPDH served as a reference.
Transmission electronic microscopy (TEM)
After fixing in 2.5% glutaraldehyde in a 0.1 M phosphate buffer for 4 h at RT and overnight at 4°C, rat or human retinas were fixed with 1% OsO4 and washed with ddH2O, followed by dehydration with graded alcohol and embedding in acetone, acetone with 33% epoxy resin, and epoxy resin successively. Ultrathin sections (50 nm) were cut by a Leica UC7 slicer and placed on copper grids, stained with uranyl acetate and lead citrate, and examined under a 120 kV electron microscope (Tecnai G2 Spirit, FEI, USA).
Optical coherence tomography (OCT) imaging
After general anesthesia (1% sodium pentobarbital, 5mL/kg), tropicamide phenylephrine eye drops (Santen) were applied for pupillary dilation. Carbomer gel was applied to the cornea to prevent dryness. Images were scanned under the ring scan mode centered by the optic nerve head using a Phoenix Micron IV retinal imaging microscope (Phoenix, USA), and 20 images at the same position were averaged to eliminate artifacts. The Phoenix software traced the edge of retinas around the optic nerve head.
Electroretinography (ERG)
After general anesthesia (1% sodium pentobarbital, 4mL/kg) and 10 min light adaption, rats were performed the full-field ERG test under a Ganzfeld Q450C stimulator (Roland, Germany). Recording electrodes were attached to the central cornea, and reference electrodes were placed hypodermically on the central forehead while grounding electrodes were inserted hypodermically into the root of the tail. To evaluate photopic negative response (PhNR), a period of 2,000 µs red light stimulation was performed at 0.4 cd.s/m2 against a royal blue background of 25 cd/m2 twice and signals were averaged to eliminate artifacts. The curves of PhNR illustrated in this article were drawn with Origin 2019b (OriginLab, USA), and Savitzky-Golay method was applied with setting points of window at 5 ~ 10. The PhNR amplitude was noted with the original data and defined as the potential between the trough following the b-wave and the baseline without signal processing(28).
Formulation and delivery of NAC and NADPH
NAC (A8199, Sigma-Aldrich, USA) and NADPH tetrasodium salt (HY-F0003, MedChemExpress, USA) were preserved under conditions as manufacturers suggested. For in vitro experiments and intravitreal injection, powders were dissolved to 10 mM in NS and diluted with medium (to different concentrations mentioned in the Results part) for OGD or NS (to a concentration of 0.5 mM NAC, 1 mM NADPH and their equivoluminal mixture) for intravitreal injection, following sterilization by filtration. The entrance of intravitreal injection was 2mm posterior to the corneoscleral limbus and 5 µL solution with 1% fluorescein sodium was gently injected. Facilities and anesthetics/antibiotics of intravitreal injection were same as AC injection and retinal hemorrhage or obvious leakage individuals were excluded. For in vivo intraperitoneal injection, various doses of NAC (300 mg/kg), NADPH (10 mg/kg) and mixture (150 mg/kg NAC & 5 mg/kg NADPH) were dissolved in NS just before injection according to body weight.
Statistics
All data were reported as mean ± standard deviation (SD) and all statistical tests were two-tailed. Results of p < 0.05 were considered statistically significant. All experiments were carried out at least three times. The repeated times of the experiments were given in corresponding captions. Comparisons between 2 groups were analyzed by Student's t-test while multiple groups were assessed by one-way analysis of variance (ANOVA) followed by Bonferroni post hoc test.