Purified sNgR-Fc was prepared as described previously. ¹⁷ This protein comprises a 310 amino acid fragment of rat NgR1 fused to a rat IgG1-Fc fragment (Biogen Idec, Inc., Cambridge, MA). Rat IgG1 (control protein) was purchased from Protos Immunoresearch (San Francisco, CA). 

Animal experiments were approved by the University of Hong Kong Faculty of Medicine Committee on the Use of Live Animals in Teaching and Research: The effect of Nogo receptor complex-related antagonists on the survival of retinal ganglion cells in a rat glaucoma model (No. 1120-05). All efforts were made to minimize the number of animals used and their suffering, in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Adult female Sprague-Dawley (SD) rats (starting weight, 250 to 280 g) were used for the study. A total of 62 rats were used in evaluating RGC loss in the ocular hypertension model (n = 10 for each group) and the optic nerve transection model (n = 6 for each group). Forty-two rats were used for c-Fos staining in this study (n = 6 for each group). Ten rats were used for immunohistochemistry and 20 rats were used for Western blot analysis for NgR1. The animals were anesthetized with intraperitoneal injection of ketamine (80 mg/kg) and xylazine (8 mg/kg) during the experiments and were euthanized at the end of the experiment with an overdose of pentobarbital sodium (150 mg/kg). Alcaine 0.5% (Alcon-Couvreur, Puurs, Belgium) was applied to the eyes before all operations and antiseptic eye drops (Tobres; tobramycin sulfate 0.3%, Alcon-Couvreur) were used to prevent infection after the surgery. Carprofen (Rimadyl; Pfizer, New York, NY) (0.025 mg/mL) in drinking water was used as analgesic for 7 days after the surgeries. All drug dosing experiments were performed double-blinded. 

Argon laser (Ultima 2000SE Argon Laser; Coherent, Palo Alto, CA) was used to coagulate the limbal and three episcleral veins in the right eye of the rats as described previously. ²⁸ The untreated left eyes served as controls for IOP elevation. A second identical laser treatment was applied 7 days later to block the neovascular flow. The IOP of the eyes were measured using a tonometer (Tono-Pen XL; Mentor, Norwell, MA) at different time points after the laser operation. An average of 10 measurements was obtained for each eye. Fluorogold (FG) labeling of RGCs was performed 7 days before euthanization. ^28,29 Both superior colliculi (SC) were exposed after removing a small piece of skull and cortex, and a piece of absorbable gelatin sponge (Gelfoam; Pharmacia & Upjohn, Kalamozoo, MI) soaked with FG (6% in distilled H₂O; Fluorochrome, Denver, CO) was placed on the surface of SC. FG retrogradely labeled intact RGCs. Rats were euthanized with an overdose of anesthesia 2 or 4 weeks after the first laser exposure for RGC counting. Ten animals were used in each group for RGC counting. 

The optic nerve was exposed by incising the conjunctiva of the eye followed by separation of the retractor bulbi muscle, then piercing and dissecting the meninges with blunt forceps. The vascular integrity of the retina was verified by fundus examination, and animals whose retina showed ischemic damage were excluded from the study. The right optic nerve was completely transected at 1.5 mm from the optic disc at Day 0, and the left optic nerve at 2 days before euthanization as the control. Surviving RGCs were retrogradely labeled by placing a piece of absorbable gelatin sponge (Gelfoam; Pharmacia & Upjohn) soaked with 6% FG at the ocular stump 2 days before euthanization. Animals were euthanized 7 days after surgery. Six animals were used for each group. 

After the first laser treatment or optic nerve transection, the right eyes immediately received an intravitreal injection of 2 μg sNgR-Fc in 2 μL phosphate-buffered saline (PBS), using a 26-gauge sterile microsyringe (10 μL, Hamilton #80300; Reno, NV), in the 5-day, 2-week, 4-week ocular hypertension model or 1-week optic nerve transection model. The injection site was just below the limbus of the cornea to incur minimal injury to the retina. After the injection, 0.3% trobramycin sulfate ophthalmic ointment (Alcon-Couvreur) was smeared around the wound to prevent infection. The protein was administered once a week in the 4-week ocular hypertension group. The rats in the control groups received intravitreal injection of 2 μL PBS or 2 μg control protein (rat IgG1) in 2 μL PBS. 

At predefined time, both eyes of each animal were enucleated and fixed in 4% paraformaldehyde (PFA) for 60 minutes after transcardial perfusion with 0.9% saline. Retinas were prepared as flat mounts and the FG-labeled RGCs were counted under fluorescence microscope using an ultraviolet filter (excitation wave length = 330 to 380 nm) as described. ^28,29 The RGCs were quantified under an eyepiece grid of 200 × 200 μm² along the median line of each quadrant, starting from the optic disc to the border at 500 μm intervals. Eight microscopic fields for each quadrant and a total of 32 per retina for four quadrants were counted. Percent loss of RGCs was used to examine the survival effects of the treatments. The data were expressed as the density of cells (number of cells/mm²) and also analyzed in terms of relative percent RGC loss in the injured right eye compared with the contralateral left intact eye from the same animal. 

To determine whether sNgR-Fc can affect the synaptic input after the induction of ocular hypertension, the c-Fos protein which was transiently expressed in neurons after synaptic stimulation was used as the marker of neuronal connectivity. ^{23 –27} The expression of c-Fos protein in immunohistochemically stained RGC indicates the intactness of the neural synaptic pathways. A total of 42 animals were used for c-Fos staining (n = 6 for each group). FG labeling of RGCs was performed 7 days before euthanization. The animals treated with sNgR-Fc or PBS were exposed to 30 minutes of ambient light before euthanization and were euthanized with an overdose of anesthesia at 5 days, and 2 and 4 weeks after the first laser coagulation. ²⁶ The eyes were enucleated after transcardial perfusion with 0.9% saline and were fixed in 4% PFA for 1 hour. After removing the corneas and lens, the eyecups were fixed further in PFA for 4 hours and then transferred to 30% sucrose solution at 4°C for 16 hours. Ten-micron-thick frozen retinal consecutive sections were cut and placed onto different slides. The sections containing the optic nerve stump were selected from a 1 in 5 series for immunostaining of c-Fos protein. After washing in 0.01 M PBS, the retinal sections were incubated in 0.01 M sodium citrate with 0.5% Tween-20 (pH 6.0) at 85°C–90°C for 30 minutes. After cooling to room temperature, the sections were washed and incubated in 0.5% Triton/PBS for 10 minutes, and then blocked with 10% normal goat serum for 1 hour. Incubation with rabbit anti-c-Fos antibody (1:100, CalBioChem, San Diego, CA) was performed at 4°C for 16 hours. Sections then were washed with PBS three times and incubated with AlexaFluor 568 goat anti-rabbit second antibody (1:400, Molecular Probes, Eugene, OR) at room temperature for 2 hours. The sections were analyzed under fluorescent microscopes. 

The eyes were enucleated at 0 and 4 weeks after injury for NgR1 immunohistochemistry. RGCs were retrogradely labeled with FG 7 days before euthanization. Retinal sections were prepared as above except that antigen retrieval was not performed. The sections were incubated with mouse anti-NgR1 (1:50; Biogen Idec, Inc.) and AlexaFluor 568 goat anti-mouse second antibody (1:400, Molecular Probes). After washing, the sections were mounted and analyzed under a confocal microscope (LSM 510 META; Carl Zeiss, Jena, Germany). The sections for NgR1 expression were also processed for immunoperoxidase staining. The sections were incubated with biotin-conjugated secondary antibody (goat anti-mouse; DakoCytomation, Glostrup, Denmark) and visualized with diaminobenzidine (DAB; 0.03% DAB, 0.003% H₂O₂, pH 7.2). The nuclei were counterstained by hematoxylin. The number of animals for each group is 5. 

The number of total FG-labeled RGCs and RGCs exhibiting colocalization of FG and c-Fos in each retinal section was counted. Three retinal sections from each eye were counted to generate the average value. The number of labeled cells in each retinal section was therefore used as an estimation of the labeled cell density in this study. FG-labeled RGCs were examined using an ultraviolet filter (excitation wavelength = 330–380 nm). Texas red-stained cells were examined using filters with excitation wavelengths of 510–560 nm (red). 

To measure NgR1 in the retina, the animals were euthanized at 0 and 4 weeks after laser coagulation and treatments with PBS or sNgR-Fc. The retinas were dissected and homogenized in lysis buffer (10 mM Tris pH 7.4, 150 mM NaCl, 1 mM EDTA, 1mM EGTA) supplemented with 10% protease inhibitor cocktail and 1% phosphatase inhibitor cocktail from Sigma (St. Louis, MO). After centrifugation at 13,000 rpm for 30 minutes to remove cell debris, the protein concentration of the supernatant was measured using a protein assay kit (Bio-Rad DC; Bio-Rad Laboratories, Hercules, CA). A 40 to 80 μg aliquot of proteins from each sample was subjected to 6% (for NgR1) or 10% (for β-actin) SDS-polyacrylamide gel electrophoresis and transferred onto PVDF membrane. The membranes were blocked with blocking buffer (Odyssey; LI-COR, Inc., Lincoln, NE) for 1 hour at room temperature. Incubation with mouse anti-NgR1 (1:100; Biogen Idec, Inc.) antibody or mouse antibody against actin (1:2000; Chemicon, Hofheim, Germany) was performed for 16 hours at 4°C. After washing, the membranes were incubated with goat anti-mouse secondary antibody 1:2000 (IRDye 800CW; LI-COR, Inc.) in blocking buffer (Odyssey; LI-COR, Inc.) for 1 hour at room temperature. After washing, the membranes were scanned and analyzed using an infrared imaging system (Odyssey Infrared Imaging System; LI-COR, Inc.). All experiments for Western blot analysis were performed with 5 animals in each group and repeated 2 to 3 times. 

All data are expressed as mean ± SEM. Statistical analysis was evaluated by unpaired Student's t-test for comparisons of two groups. One-way analysis of variance (ANOVA) followed by post hoc Tukey's multiple comparison test was used to compare the NgR1 level in the different groups. Paired Student's t-test was used to compare the RGC density between the right and contralateral left eyes in the same group. Data were analyzed statistically with commercially-available software (SPSS 12.0; SPSS Inc., Chicago, IL). 

Previous studies have demonstrated that normal RGCs express NgR1 mRNA in the adult mouse. ³⁰ Here we further examined NgR1 protein expression in normal and injured rat retinas using immunohistochemistry. NgR1 positive staining was visible mainly in the ganglion cell layer (GCL) and the nerve fiber layer (NFL) in the normal retina (Figs. 1A, 1B). Normal RGCs expressed moderate staining for NgR1 but stronger immunoreactivity in the density of labeling for NgR1 occurred in RGCs 4 weeks after laser coagulation (Figs. 1A, 1B). This finding was confirmed by Western blot analysis (Fig. 1C). The increased IOP had a significant effect on NgR1 levels (one-way ANOVA, F _(3,16) = 6.82; P < 0.01). NgR1 expression was low in normal retina and increased to 1.8-fold at 4 weeks after the injury (P < 0.05). The expression of NgR1 was similar for sNgR-Fc treatment versus control PBS group. 

To explore the efficacy of sNgR-Fc on RGC survival, we induced an ocular hypertension induced glaucoma model in adult rats. The episcleral and limbic veins on the eye were sealed by two consecutive argon laser treatments (7 days apart) to elevate IOP in the eye. Similar to those previously described, ^28,31 the IOP of the glaucomatous eyes were found to be elevated by approximately 1.7 times after two laser treatments (data not shown). In general, after 2 weeks of IOP elevation, there would be approximately 13% RGC loss in this model. Single intravitreous injection of sNgR-Fc almost completely rescued RGC loss as compared with the PBS and control protein-injected control eyes in the 2-week model (P < 0.001; Fig. 2A). To evaluate its efficacy in the chronic glaucoma model where RGC loss is maximal after 4 weeks, ³² weekly injection of sNgR-Fc to the injured eye for 4 weeks significantly reduced RGC loss (n = 10; P < 0.001; Fig. 2A) and maintained the RGC density in the retina (Fig. 2B). Similarly, sNgR-Fc treatments improved RGC survival when compared with the PBS control treatment (P < 0.001) 1 week after optic nerve transection in the rats (n = 10; Fig. 3). These demonstrate the neuroprotection function of sNgR-Fc in chronic glaucoma and acute injury. 

We previously noted that there is synaptic abnormality in the retina on imposing ocular hypertension. ⁷ We hypothesized that sNgR-Fc would prevent synaptic deafferentation while maintaining the efferent connectivity intact, thus preventing the injured neurons from dying. The c-Fos protein has been established as a neuronal connectivity marker in the retina and was employed in this study to indicate synaptic connectivity. ²⁷ The c-Fos was expressed in the nuclei of RGCs retrogradely labeled with FG in the normal retina (Fig. 4A). However, there were fewer c-Fos-positive RGCs after the induction of ocular hypertension (Fig. 4B). The ratio of c-Fos containing RGCs to the number of surviving RGCs was lower at 5 days (P < 0.05), 2 weeks (P < 0.01), and 4 weeks (P < 0.01) post laser-induced injury (PBS group, n = 6 for each group; Fig. 5A). The ratio was restored after sNgR-Fc treatment (P < 0.05 for 5 days and P < 0.001 for 2 and 4 weeks). Similar data were obtained by evaluating the number of c-Fos positive RGCs or total RGCs in the retina (Fig. 5B). Here we found that there was no significant RGC loss at 5 days after the first laser coagulation (P > 0.05; Fig. 5B). It suggests that synaptic disconnection before the loss of RGCs can be detected under the stress of ocular hypertension. Some surviving RGCs (FG-labeled) were not positive for c-Fos suggesting that there is continuous synaptic deafferentation under the elevated IOP. These results suggest a synaptotoxic mechanism, similar to that thought to occur in Alzheimer's disease, ³³ may operate in experimental glaucoma. RGC death may be a secondary event and inhibiting synaptic injury using sNgR-Fc may delay the progression of glaucoma. 

Here we demonstrate that the NgR1 antagonist, soluble Nogo-66 receptor, attenuated synaptic degeneration as measured by c-Fos in the retina and in addition enhanced RGC survival in a rat chronic glaucoma model. The use of sNgR-Fc may be a novel approach for inhibiting synaptic injury and promoting RGC survival in glaucoma. 

Glaucoma is a common eye disease that can cause irreversible loss of vision if left undiagnosed and untreated. Recent studies showed that in a mouse model of glaucoma, RGCs manifest growing dysfunction via somatic shrinkage, retrograde labeling deficits, and downregulation of RGC-specific genetic programs before death. ⁶ Our data also suggest that RGCs undergo synaptic degeneration before cell loss under ocular hypertension. The synaptic degeneration may play a crucial role in the subsequent RGC death. ⁷ The visual field changes in glaucoma may be caused by both synaptic degeneration and cell loss. Other than the primary degeneration, the self-perpetuation process of secondary degeneration may occur during which RGC bodies gradually succumb to the noxious environment in glaucoma. ³⁴ These results suggest inhibiting synaptic injury may provide a new insight for early intervention in delaying the progression of glaucoma. 

NgR1 is associated with axon regeneration and neurite growth in the CNS. We found that there was NgR1 expression in the RGCs and upregulated after the induction of ocular hypertension. Soluble NgR1 protein can inhibit the activity of NgR1 and promotes axon regeneration after spinal cord injury, ^17,22 stroke, ¹⁶ or dorsal root rhizotomy. ^18,22 Inhibiting the function of NgR by RNA interference ¹⁹ or transfecting the dominant-negative form of NgR1 ²⁰ can stimulate optic nerve axon regrowth. Knockout of NgR1 is effective for enhancing axonal regeneration after optic nerve crush. ²¹ However, there are no reports about the effect of sNgR-Fc on neuron survival and synaptic function. Using a model that mimics the etiology and progression of chronic glaucoma, we found that sNgR-Fc prevented RGC loss associated with an increase of intraocular pressure. These agents prevented RGC loss even over a long-term 4-week course of treatment. Because the RGC loss becomes stable after 4 weeks in this ocular hypertension model, ³² the evident neuroprotection of sNgR-Fc for a 4-week course indicates a long-term prevention of RGC damage in this ocular hypertension model. More importantly, we found that there was no significant RGC loss but the number of RGCs expressing c-Fos was significantly lower in the experimental animals at 5 days. However, the treatment of sNgR-Fc promoted synaptic connection at 5 days even up to 4 weeks. It suggests that sNgR-Fc may promote RGC survival by prevention of the synaptic disconnection in the early stage of the RGCs, and inhibiting NgR1 activity can slow the progress of glaucoma by rescuing both the synapses and RGCs. We also tested the efficacy of sNgR-Fc in the extreme conditions of optic nerve transection. Even in that stringent environment, NgR1 antagonist provided a significant level of RGC protection for up to a week. 

A member of NgR1/p75 neurotrophin receptor and NgR1/TROY (an orphan receptor in the tumor necrosis factor TNF family) signaling complexes that prevent axonal regeneration in the presence of three myelin inhibitors is leucine-rich repeat (LRR) and Ig domain-containing Nogo receptor-interacting protein-1 (LINGO-1). ³⁵ We previously found that blocking LINGO-1 function with a soluble version of the extracellular domain of LINGO-1 or an antibody against LINGO-1 promotes survival of damaged RGCs in a chronic adult rat ocular hypertension model. ²⁸ Combination of brain-derived neurotrophic factor (BDNF) with LINGO-1 antagonist can provide long-term protection for RGCs. BDNF receptor, tyrosine kinase B (TrkB) may be the predominant mediator of this neuroprotection. ^29,31 As a glycosyl phosphatidyl inositol (GPI) anchored protein, NgR1 directly binds the myelin proteins, and then LINGO-1 and/or p75/TROY transduce the signal into the cells. ^36,37 Whether sNgR-Fc exerted neuroprotection in this study via the interaction with LINGO-1 and/or p75/TROY should be further investigated. 

In this study, we found that there was RGC death after the induction of chronic ocular hypertension and that synapse degeneration precedes RGC death. This would suggest that there may indeed be an opportunity for the rescue of RGCs that are undergoing synapse alteration, but possibly not yet committed to die. NgR1 antagonism by sNgR-Fc prevents synaptic dysfunction and promotes RGC survival which may represent a novel therapy for treating glaucoma.