Laboratory animals

All animal experiments were conducted in accordance with the EU Animal Experimentation Directive 2010/63/EU and the Association for Research in Vision and Ophthalmology (ARVO) guidelines. In this study, the implemented animal protocol was scrutinized and approved by the government agency responsible for animal welfare in the state of North Rhine-Westphalia (Landesamt für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen, Germany). The experiments were performed in C57BL/6J male mice (8–9 weeks old), Nox2 −/− (gp91 phox−/−) and age-matched wildtype (WT) mice. The standard mouse housing conditions were: 12 h light/dark cycle, 22 ± 2 °C, 55 ± 10% humidity, and free access to food and water.

In vivo episcleral vein occlusion model (H-IOP)

All animals received episcleral vein occlusion surgeries in their right eyes (H-IOP) and sham surgeries in their left eyes (Sham) (Fig. 1A). The glaucoma mouse model was induced by occlusion of the three episcleral veins in their right eyes [19, 23]. Briefly, mice were anesthetized using a ketamine (100 mg/kg) and xylazine (10 mg/kg) solution via intraperitoneal injection. We also applied one drop of the 4 mg/mL oxybuprocain hydrochloride (Novesine® 0.4% Eyedrops, OmniVision®, OmniVision GmbH, Puchheim, Germany) onto the scathe ocular surface for local anesthesia. A cut was made across the conjunctiva and Tenon’s capsule at the limbal edge of the right eye under an operating microscope. Two relaxing incisions were made at the edges of the initial incisions, and the tissue was recessed posteriorly to expose the underlying extraocular superior and lateral rectus muscles. These muscles were gently pulled aside with a suture, bringing the episcleral veins into view. After surgical isolation, microforceps were positioned under the episcleral veins adjacent to the lateral and superior rectus and to the superior oblique muscles. A hand-held ophthalmic thermal cautery device (Fine Science Tools GmbH, Heidelberg, Germany) was employed to cauterize each vein until venous congestion was noticeable, indicating blockage without any leakage (Fig. 1B, the Image from Ruiz-Ederra et al., 2005). Great care was taken to minimize blood loss and avoid damage to the conjunctiva and the underlying sclera. Finally, the conjunctiva was put back to its original location, and ofloxacin ointment was given onto the ocular surface to prevent inflammation. Sham surgery was conducted similarly but without causing damage to the episcleral veins. Euthanasia via cervical dislocation was performed on the mice at three distinct post-operative intervals: 4 days, 8 days, and 2 weeks.

Fig. 1

Schematic diagram of the construction of mouse glaucoma model and analysis of retinal samples. (A) The overview of in vivo H-IOP mouse model construction. (B) The schematic diagram of episcleral vein cautery. Image from Ruiz-Ederra et al., 2005 [23]. The illustration of the mouse eye depicts episcleral veins (red) relative to extraocular muscles, with three veins marked by cauterization interruptions. Pre- and post-cauterization (EVC-treated) images of these veins are provided, with arrows pointing to cauterization sites. (C) The overview of ex vivo retinal explants Pressure 60 mmHg model construction and inhibition of gp91ds-tat. (D) The overview of the retina sample preparation and the imaging and analysis of the retinal flat mount and cryosection. Abbreviations: EVC, episcleral vein occlusion; IO, inferior oblique; IR, inferior rectus; LR, lateral rectus; MR, medial rectus; SO, superior oblique; SR, superior rectus

IOP monitoring

The IOP of the mice was monitored using a TonoLab rebound tonometer (iCare, Vantaa, Finland) in animals once every two days [24, 25]. All measurements were conducted between 9:00 a.m. to 12:00 p.m. for comparability. After six consecutive measurements, the tonometer generated a mean IOP. We took 5 sets of measurements and averaged these IOP values.

Ex vivo elevated hydrostatic pressure model (Pressure 60 mmHg)

The C57BL/6J, Nox2 −/−, and WT mice were euthanized by cervical dislocation. Eyes were immediately enucleated and transferred to Petri dishes containing ice-cold phosphate-buffered saline (PBS) (Carl Roth,1108.1), followed by peeling out the intact retina and dissecting the vitreous humor. The retinal explants were divided equally into four segments, ensuring that the ganglion cells were facing upward, and placed in Mixed Cellulose Ester (MCE) Membrane Filters (Advantec, A045R047Z-P). The retinal explants were transferred to lumox culture dishes 35 (Sarstedt, Nümbrecht, Germany). Retinal tissues were cultured in the standard culture medium, Dulbecco’s Modified Eagle’s Medium/Nutrient Mixture F-12 (DMEM/F12; Gibco BRL, Eggenstein, Germany) supplemented with 10 µg/mL porcine insulin, 100 U/mL penicillin, 100 µg/mL streptomycin. To minimize variability, retinal tissues were randomly assigned to the control and Pressure 60 mmHg groups.

For the Pressure 60 mmHg model, retinal explants undergo incubation for 8,16,24 h in a tailor-made pressure incubator chamber at a hydrostatic pressure of 60 mmHg, which simulates intraocular conditions under abnormally elevated IOP. Pressure incubation chambers with lids and non-directional valves allow access to 5% CO2 in the incubator at 37 °C for constant monitoring of internal air pressure employing manometers. Control retinas were cultured in the incubator with humidified 5% CO2 at 37 °C under normal atmospheric pressure (Fig. 1C).

gp91ds-tat pharmacological intervention for experimental glaucoma models

The peptide gp91ds-tat (YGRKKRRQRRRCSTRIRRQL-NH2), a NOX2-specific inhibitor, interferes with NADPH oxidase assembly by targeting a sequence essential for binding gp91phox with p47phox [26, 27]. HIV-tat peptide, an amino acid sequence internalized by all cells, is linked to the gp91phox sequence to facilitate cell entry [28]. For the Pressure 60 mmHg model, the retinal explants from C57BL/6J mice were incubated by adding different concentrations of gp91ds-tat (Anaspec, San Jose, CA, USA) in the standard culture medium for 24 h (Fig. 1C). Control or Pressure 60 mmHg retinas were incubated in the standard culture medium (Vehicle). For the in vivo H-IOP model, the different concentrations of gp91ds-tat (dissolved in normal saline) were injected intraperitoneally (10 ml/kg) 30 min after the H-IOP surgery and once every two days until the end of the experiment. Sham and H-IOP mice were injected with normal saline (Vehicle) (10 ml/kg). The mice were euthanized by cervical dislocation after two weeks. The preparation of retinal explants was performed as described above.

Immunostaining of retinal flat mounts and cryosections

Each retina was divided into four equal sections along the dotted line in the schematic of Fig. 1D-1. Then, each quarter of the retinal explants from different mice served as an independent sample of each group, which was subjected to immunostaining for different markers and preparation of frozen sections. Retinal explants were fixed in 4% paraformaldehyde (PFA) (Histofix, Roth, Karlsruhe, Germany) for 30 min after rinsing with PBS. Subsequently, the retinas were rinsed twice with PBS for 10 min and dehydrated in 30% sucrose solution at 4 °C for 24 h to be analyzed further. After washing 3 times (10 min each) in PBS, the retinas were incubated in a buffer solution containing 1% BSA and 0.3% Triton-X-100 in PBS for 2 h at room temperature. Then, the retinas were incubated overnight at 4 °C with the primary antibody (Table 1). Next, the retinas were washed with PBS and placed with the secondary antibody in 1% BSA and 0.3% Triton-X-100 in PBS for 2 h at room temperature. Afterward, the retinas were rinsed with PBS three times and mounted on slides.

After fixation, the retinas were embedded in an optimal cutting temperature (OCT) compound (Sakura Finetek, Torrance, CA, USA) for cryostat sectioning. Retinal explants were cut vertically to a thickness of 12 μm by a Leica CM3050S cryostat (Leica Microsystems, Buffalo Grove, IL), collected onto gelatin-coated slides, and stored frozen until immunohistochemical processing. The retinal sections were washed with PBS for 5 min over 3 times. Then, the retina was blocked with 3% Normal goat serum and 0.1% Triton X-100 in PBS for 60 min, followed by primary antibodies (Table 1) at 4 °C for overnight incubation. After washing with PBS for 5 min 3 times, ensure that all procedures are performed in the dark and apply the secondary antibody 1:1000 in PBS for 2 h. Sections were rinsed with PBS. Finally, the retinal flat mounts and cryosection slides were mounted with VECTASHIELD® mounting medium with DAPI (BIOZOL Diagnostica Vertrieb GmbH, Eching, Germany) and covered with a coverslip.

Table 1 Antibodies used for histological analysesImaging and analysis of retinal flat mounts and cryosections

Images were captured using a Zeiss Imager M.2 equipped with an Apotome.2 (Carl Zeiss; Jena, Germany). As detailed in Fig. 1D-2, three areas—central, middle, and peripheral—were imaged under a 20X magnification for each quarter of the retinal flat mount. Similarly, for retinal cryosections, the Zeiss Imager M.2 with an Apotome.2 was employed to photograph the central, middle, and peripheral regions of three consecutive sections from each sample, using a 20X objective.

Fig. 2

Pathologically high intraocular pressure-induced glaucomatous RGC loss and neurodegeneration. (A) Representative images of flat mount retina immunostained with Brn3a after Pressure 60 mmHg. Scale bar, 50 μm. (B) Analysis of the number of Brn3a-positive RGCs at different times after Pressure 60 mmHg. Data are shown as mean ± SEM (n = 6 in each group, one-way ANOVA with Tukey’s multiple comparisons test, *p < 0.05,**p < 0.01, ***p < 0.001, ****p < 0.0001). (C) Time course of IOP before as well as 14 days after H-IOP surgery. Data are shown as mean ± SEM (n = 6 in each group, two-way ANOVA with Šídák’s multiple comparisons test, ****p < 0.0001). (D) Representative images of flat mount retina immunostained with Brn3a and PPD-stained ON after H-IOP. Scale bar, 50 μm and 5 μm. (E) Analysis of the number of Brn3a-positive RGCs at different times after H-IOP. (F) Analysis of the number of axons and the percent of degenerating axons in the different groups. Data are shown as mean ± SEM (n = 6 in each group, one-way ANOVA with Tukey’s multiple comparisons test, *p < 0.05,**p < 0.01, ***p < 0.001, ****p < 0.0001). (G) Messenger RNA expression of Nox enzymes (Nox2, p47phox, Nox1, Nox4) in H-IOP retinas. Data are presented as the fold-change after H-IOP versus control. Data are shown as mean ± SEM (n = 6 in each group, Unpaired T-test, *p < 0.05,**p < 0.01, ***p < 0.001, ****p < 0.0001). (H) Representative images of retinal cross-sections immunostained with NOX2 after Pressure 60 mmHg (Scale bar, 50 μm) and the analysis of NOX2 fluorescence intensity in the retinal vessels and different retinal layers. (I) Representative images of retinal cross-sections immunostained with DHE staining after Pressure 60 mmHg. Scale bar, 50 μm. Analysis of DHE fluorescence intensity in the retinal vessels and different retinal layers. The white arrows point to cross-sections of retinal blood vessels. Data are presented as the percent fluorescence intensity of the Pressure 60 mmHg versus control. Data are shown as mean ± SEM (n = 6 in each group, Unpaired T-test, **p < 0.01, ****p < 0.0001)

Glia-vascular unit interactions in the retinal flat mounts were examined using a Leica SP8 confocal laser-scanning microscope (Leica, Wetzlar, Germany) at 20X magnification. Additionally, CD31 immunolabeled retinal flat mounts were observed with a Leica SP8 confocal microscope. Z-stack images, comprising 20–25 layers for each area (Fig. 1D-3), were compiled along the Z-axis to produce a 2D representation of the retinal vasculature.

Fig. 3

gp91ds-tat treatment effectively modulates the expression of major immune-related genes in H-IOP injured retina. (A) Principal component analysis shows the clustering of proteins in different samples. (B) Volcano plot showing differentially expressed proteins between H-IOP and control group, H-IOP + gp91ds-tat and H-IOP, H-IOP + gp91ds-tat and control group. (C) The Venn diagram was created by the differentially expressed data of proteins. (D) The most enriched KEGG Level 2 pathways for the 635 co-differentially expressed proteins. Hierarchical clustering illustrates distinct expression differences of immune genes between the three groups and homogeneity between groups. (E) The 10 most enriched KEGG pathways for the differentially expressed proteins. The size of the symbol represents the number of genes, and the colors represent the p-value. (F) The 10 most enriched gene ontology terms for the parental genes of the differentially expressed immune genes. Enriched GO terms are on the vertical axis, and the number of annotated differentially expressed genes associated with each GO term is indicated on the horizontal axis

Image analysis was conducted using ImageJ2 2.3.0 (http://rsb.info.nih.gov/ij/), NIH, Bethesda, MD, USA. Quantitative assessments of various markers, including RGCs count et al., were performed. The mean count from the three regions of a quarter-section of the retina provided a measure of overall retinal changes, which was then utilized for further statistical evaluation.

Culture and stimulation of primary microglia

As described previously [29], primary microglia were isolated from C57BL/6J neonatal mouse pups (P0). Cell suspension was plated on cell culture dishes (Nunc) coated with Poly-D-Lysine (#A003E stock 1 mg/ml (use 50ng/ml in PBS)) in PM Medium (Neurobasal-A, 10% horse serum, B27, P/S/G). Primary Microglia at passage 2 or 3 were used for experiments. Cells were stimulated with 100nM ET-1 (E7764, Sigma-Aldrich) and inhibited with 20 μm MEK (PD98059) for 20 min and 24 h starting at 3 days in vitro (DIV). For Immunostaining, primary microglia grown on slides were treated and then washed with PBS, followed by fixation with 4% PFA for 30 min. The immunostaining protocol for primary microglia follows the same procedure as used for the retinal flat-mount.

Quantification of cytosolic ROS

Retinas were also homogenized in lysis buffer as previously described [30]. Lysates were centrifuged at 1500 g for 3 min, and the supernatant was exposed to 20 µM 2,7-dichlorofluorescein derivative 6-carboxy-2,7-dichlorodihydrofluorescein diacetate, di(acetoxymethyl ester) (referred to as DCF throughout this manuscript) for 30 min at 37 °C. Fluorescence was measured using a plate reader (Tecan infinite 200Pro) (ex/em = 485/535 nm).

Quantification of superoxide (O2−)

O2- levels were measured in frozen sections of unfixed retinas by staining with the fluorescent dye dihydroethidium (DHE) [19, 31,32,33]. Immediately after thawing, tissue sections were incubated with 1 µM of DHE for 30 min at 37 °C. The slides were mounted with VectaShield mounting medium (Vector Laboratories, Burlingame, CA) and covered with a coverslip. Photographs of retinal cross-sections were taken using a Leica SP8 confocal laser scanning microscope (Leica, Wetzlar, Germany). As described previously, the staining intensity of blood vessels and individual retinal layers was measured using ImageJ software (NIH, http://rsb.info.nih.gov/ij/).

Assessment of optic nerve degeneration

Optic nerve (ON) degeneration was assessed by p-phenylenediamine (PPD)-stained ON cross sections, as previously described [34, 35]. The semi-thin cross sections of ON were taken from 1.0 mm posterior to the eyeballs. In brief, the ON was separated from the eyeball and fixed overnight in a phosphate-buffered 3% glutaraldehyde/paraformaldehyde mixture at 4 °C. Following overnight treatment in 1% osmium tetroxide at 4 °C, ON were rinsed in 0.1 M phosphate buffer and 0.1 M sodium-acetate buffer, then dehydrated in graded ethanol concentrations. After embedding ON in resin (Eponate-12; Ted Pella), 1 μm sections were cut and stained in 1% PPD for 10 min. The sections were imaged under a Zeiss Axio Imager Z1 Microscope (Zeiss, Oberkochen, Germany) with a Zeiss Plan-ACHROMAT 100×Lens (Zeiss, Oberkochen, Germany). Image Acquisition was performed using a Canon EOS 6D Mk II camera (Canon, Krefeld, Germany) and CanonEOS Utility Software (Canon, Krefeld, Germany). Following a protocol similar to that of retinal cryosection imaging, three fields (including one central and two peripheral) were imaged from three sequential cross-sections for each ON, utilizing a 100× magnification. ON axon numbers were counted manually with ImageJ software (http://rsb.info.nih.gov/ij/, NIH, Bethesda, MD, USA), and the average numbers were calculated as axon density per square millimeter of ON and then multiplied by cross-sectional area to calculate the total number of axons per ON [31].

Western blotting

Tissues were processed for Western blotting as previously described [36]. In brief, the retinal tissues and primary microglia were washed twice with cold PBS and incubated in 100 µL lysis buffer (T-PER Tissue Protein Extraction Reagent + 2% protease inhibitor) per whole retina explants on ice. After 15 min of incubation, the retina and cell lysates were dissected and sonicated for 1 min in an ultrasonic bath on ice and then centrifuged at 1000 g for 8 min. The supernatants were collected for further analysis. The BCA Protein Assay Kit (Pierce, Rockford, IL, USA) determined each lysate’s protein concentration per the manufacturer’s instructions. From each retina lysate, 20 µg of protein was loaded onto a Novex NuPAGE 12% Bis-Tris polyacrylamide gel (Thermo Fisher, USA). The gel electrophoresis was run using NuPAGE running buffer MES at room temperature with a voltage of 130 V for 60 min. After electrophoresis, the proteins were transferred to polyvinylidene fluoride membranes (Bio-Rad, Gladesville, Australia) using a dry transfer system (Bio-Rad, Hercules, CA, USA), and a standard transfer buffer (with 20% methanol), a voltage of 20 V was applied for 7 min. For immunoblotting, membranes were incubated with the appropriate antisera ET-1 (1:1000, NB300-526, Novus Biologicals, Littleton, USA), p47phox (1:1000, sc-17,845, Santa Cruz Biotechnology), gp91phox (1:1000, sc-130,543, Santa Cruz Biotechnology), polyclonal antibodies recognizing phospho-JNK (Thr183/185) (1:1000; catalog no. 9251), phospho-ERK1/2 (Thr202/204) (1:1000; catalog no. 9101) and phospho-p38 MAPK (Thr180/182) (1:1000; catalog no. 4511) overnight at 4 °C, and labeling was carried out using a multi-step detection procedure. First, appropriate biotinylated secondary antibodies were reacted with membranes, and then streptavidin-peroxidase conjugates were applied. Blots were developed with a 0.016% solution of 3-amino-9-ethyl carbazole in 50 mM sodium acetate (pH 5) containing 0.05% Tween-20 and 0.03% H2O2. Images were acquired from labeled blots and analyzed for densitometry using the software program ImageJ2 2.3.0. Densitometry values were then normalized for β-actin (1:10000, ab6276, Abcam).

MS measurement

Proteins were extracted from the retinal tissue of these individuals using the detergent sodium dodecyl sulfate (SDS), which is the most efficient reagent for lysing tissue and cells to achieve complete protein extraction [37]. Subsequently, DIA-MS proteomics analysis was performed with a high-resolution, high-quality precision LTQ-Orbitrap Elite mass spectrometer. The continuum MS data were collected by an ESI-LTQ Orbitrap XL-MS system (Thermo Scientific, Bremen, Germany) and searched against the UniProt database with MaxQuant software version 1.5.3.30 (Max Planck Gesellschaft, Germany). Due to the random nature of “Birdshot” label-free quantitative proteomics, protein identification or abundance data are sometimes missing in some samples [38]. 40 A target-decoy-based false discovery rate (FDR) was set to 0.01 for the identification of peptides and proteins, the minimum peptide length was 6 amino acids, and the minimum unique peptides were set at 2. Fold changes of the label-free quantitation (LFQ) intensities were calculated to identify the significantly differentially expressed proteins (DEPs). We performed differential expression analysis on the quantitative proteomics data targeting the H-IOP relative to the control |Student’s T-test Difference| ≥ 0.5 change threshold and -Log P-value > 1.3. To outline the DEPs profiles of the transcripts, generating volcano maps via the R ggplot2 package and performing hierarchical cluster analysis via the Manhattan distance metric and the Ward minimum variance method from the heatmap package in R. Principal component analysis (PCA), implemented in the prcomp function of R, was conducted to abstract the main characteristics of the data, which served as an indicator of the overall state of the data.

Gene ontology and kyoto encyclopedia of genes and genomes pathway analysis

Gene Ontology (GO; http://geneontology.org/) analysis is based on three annotated ontologies, including the exploration of molecular function (MF), cellular components (CC), and biological processes (BP), which originate from DEPs targeting genes. Simultaneously, the Kyoto Encyclopedia of Genes and Genomes (KEGG; http://www.kegg.jp/) analysis was performed to assess the biological functions and enrichment pathways of the DE genes. GO and KEGG analyses were performed by assigning R packages based on hypergeometric distributions.

Enzyme-linked immunosorbent assay (ELISA)

The concentration of cytokines in total retinal and primary microglia lysates was measured by ELISA. Tissue samples were sonicated in PBS supplemented with protease and phosphatase inhibitors (Complete protease inhibitor cocktail, Roche). Interleukin 1beta (IL-1β) (DY401), tumor necrosis factor-alpha (TNF-α) (DY410), and interleukin 6 (IL-6) (DY406) Quantikine® ELISA′s were purchased from R&D Systems.

Quantification of gene expression by quantitative PCR

Messenger RNA for the pro-oxidant enzymes, the NOX enzymes (Nox1, Nox2, p47 phox, Nox4), for the vascular endothelial growth factor-A (Vegf-a), for the antioxidant redox enzymes, heme oxygenase 1 (Ho-1), and glutathione peroxidases 1 (Gpx1), for the cytokines, Tnf-α, IL-1β, superoxide dismutase type 2 (Sod2), and for the nitric oxide synthase (NOS) isoforms, eNos, iNos, and nNos, was quantified in the whole retinal explants as described before [39]. Tissue samples were homogenized (FastPrep; MP Biomedicals, Illkirch, France). RNA was isolated using peqGOLD TriFast™ (PEQLAB), and cDNA was generated with the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Darmstadt, Germany). Quantitative real-time RT-PCR (qPCR) reactions were performed on a StepOnePlus™ Real-Time PCR System (Applied Biosystems) using SYBR® Green JumpStart™ Taq ReadyMix™ (Sigma-Aldrich, Munich, Germany) and 20 ng cDNA. Relative mRNA levels of target genes were quantified using the comparative threshold (CT) normalized to the housekeeping gene TATA-binding protein (Tbp). The qPCR primer sequences are shown in Table 2.

Table 2 Primer sequences used for quantitative PCR analysisStatistics

Details of the statistical test used for each experiment are in figure legends, along with n and p-value. All data is represented as mean ± SEM. Statistical analysis was performed using GraphPad Prism 9 (GraphPad Software, La Jolla, California, USA).