Decreased d-Serine Levels Prevent Retinal Ganglion Cell Apoptosis in a Glaucomatous Animal Model

Abstract

Purpose: The purpose of this study was to determine d-Serine and d-Serine synthetase serine racemase (SR) expression and whether decreased d-Serine expression has protective effects on retinal ganglion cells (RGCs) in a glaucomatous animal model.

Methods: The rat chronic intraocular hypertension (COH) model was generated as a glaucomatous animal model by cauterizing three episcleral veins. Quantitative analysis of RGC survival was determined by the counting of retrograde FluoroGold-labeled RGCs. The level of d-Serine in the retinas and aqueous humor was determined by Ultra High Performance Liquid Chromatography coupled to triple-quadrupole Mass Spectrometry (UHPLC-MS/MS). Retinal expression of serine racemase (SR) protein was determined by immunohistochemistry and Western blot analysis. The TUNEL assay was used to detect cell apoptosis.

Results: The content of d-Serine increased significantly in the glaucomatous retina of the COH model 2 weeks after surgery compared with the control retina. d-Serine synthetase SR expression in the right glaucomatous eye increased slightly after surgery compared with that in the left control eye and remained at this high level for 6 weeks after surgery. SR-positive cells were located mainly in the ganglion cell layer (GCL) of the retina. d-Amino acid oxidase (DAAO) treatment significantly increased RGC survival in the glaucomatous eyes, and the TUNEL assay was used to confirm that DAAO reduced the number of TUNEL-positive cells in glaucomatous eyes. However, excess d-Serine could not exacerbate RGC loss in the COH model.

Conclusions: Increased d-Serine and SR expressions in the retina of the COH model were detected. DAAO treatment significantly increased RGC survival in the glaucomatous eyes. These results suggest that decreased d-Serine expression has protective effects on RGCs.

Glaucoma comprises a group of diseases characterized by progressive retinal ganglion cell (RGC) death and optic nerve degeneration, resulting in irreversible blindness.¹ Elevated IOP is an important risk factor for RGC death in glaucoma. However, the pathophysiologic relationship between elevated IOP and RGC death remains poorly understood.² Many molecular pathophysiologic processes, such as axonal transport failure, neurotrophic factor deprivation, excitotoxic damage, and oxidative stress, act alone or in cooperation with other pathophysiologic processes to promote RGC death.³ Excitotoxicity caused by N-methyl-d-aspartate (NMDA) receptor overactivation has been shown to play a role in intraocular hypertension–induced RGC apoptosis.⁴ 

The NMDA receptor is essential for central nervous system function. The amino acid glutamate has been shown to act as a neurotoxin that exerts its toxic effects on RGCs predominantly through the NMDA subtype of the glutamate receptor.⁴ When the nervous system is injured, large amounts of glutamate are released, neurons become depolarized, and then abnormal increases in NMDA receptor activity occur. NMDA receptor overactivation results in excessive Ca²⁺ influx, which leads to damage-inducing free radical production and contributes to cell death.⁵ 

d-Serine, an unusual amino acid synthesized in the brain by serine racemase (SR) and degraded by d-amino acid oxidase (DAAO), is another endogenous NMDA receptor coagonist and serves as a ligand at the glycine site of the NMDA receptor.^6–8 d-Serine has a higher affinity for the NMDA receptor than glycine. Moreover, selective elimination of d-serine in brain tissues attenuates NMDA receptor function, indicating that d-serine plays an indispensable role in physiologically activating the glutamate receptor.⁹ 

In the retina, d-serine is also selectively distributed in Müller cells, astrocytes, and neuronal ganglion cells.¹⁰ d-Serine release is inhibited by extracellular calcium removal and augmented by increases in extracellular calcium levels.¹¹ d-Serine enhances glutamate-induced calcium responses in RGCs.¹² Consistent with this finding, endogenous d-serine degradation induced by treatment with DAAO attenuated the NMDA-induced calcium response, thereby rescuing RGCs from damage.¹³ Loss-of-function studies involving mouse models have shown that SR mutations significantly attenuate the retinal excitotoxicity induced by intravitreal injections of NMDA.¹⁴ Thus, it is believed that SR may be a new target in the treatment of diseases such as glaucoma that involve NMDA mediated excitotoxicity.¹⁵ 

The findings of all these studies indicate that d-serine is an endogenous coagonist of the NMDA receptor and thus plays an important role in the pathogenesis of glaucoma; however, this topic has not yet been fully explored. In this study, we measured d-serine expression in the retina in the rat chronic intraocular hypertension (COH) model as a glaucomatous animal model to study the relationship between d-serine levels and RGC apoptosis. We also intravitreally administered DAAO to reduce d-serine levels in the retina to investigate the protective effect of this treatment on RGCs. Our results may provide researchers with clues regarding the pathogenesis of glaucoma. Moreover, we may have unveiled a new drug target through which glaucoma can be clinically treated. 

Materials and Methods

All experimental protocols were approved by the Eye Institute of the Eye & ENT Hospital of Fudan University. 

Animals

All experiments and animal care procedures were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and the guidelines of the University of Fudan Institutional Animal Care and Use Committee. Male Wistar rats (200 to 250 g) were obtained from the SLAC Laboratory Animal Company. The animals were housed under a 12-hour light/dark cycle. Unilateral IOP elevation characteristic of the rat COH model was induced by episcleral vein cauterization, as described previously in detail.¹⁶ 

Rat COH Model

The Wistar rat COH model was reproduced as described previously in detail.^16,17 Briefly, the rats were anesthetized by an intraperitoneal injection of 10% chloral hydrate (0.36 mL/100 g), and their eyes were further anesthetized with 0.5% alcaine (Alcon-Couvreur, Puurs, Belgium) before surgery. Three episcleral veins in the right eye were carefully separated and cauterized. The left eye, which served as a sham-operated control eye, was simultaneously subjected to a similar surgery; however, its episcleral veins were not cauterized. 

IOP was measured in both eyes using a tonometer (TonoLab; Mentor, Norwell, MA, USA) while the animals were under general anesthesia, both before surgery and at the following times after surgery: 7, 14, 21, and 42 days. Rats with IOPs that returned to normal were excluded from the study. 

Intravitreal Drug Injections

We performed intravitreal injections using a 30-gauge needle connected to a Teflon tube with a 10-μL Hamilton syringe after dilating the rat pupil with 1% atropine sulfate. The tip of the needle was inserted through the dorsal limbus of the eye. 

A subset of rats received intravitreal injections of 3 μL 10 mU/μL DAAO (Sigma-Aldrich, St. Louis, MO, USA) or 5 μL 200 nM d-serine (Sigma-Aldrich) before surgery and every 4 days thereafter. These rats received intravitreal injections of the same volume saline in their contralateral eyes at the indicated time points.¹³ 

Retrograde Labeling and Counting of RGCs

One week before being killed, the anesthetized rats received bilateral microinjections of FluoroGold (2 μL/injection of 3%; Sigma-Aldrich), which was diluted in 10% DMSO saline, in their superior colliculi. The microinjections were administered in a stereotactic apparatus, as previously described.^18,19 

FluoroGold is taken up by retinal ganglion cell axon terminals and bilaterally transported retrogradely to the somata of the cells, where it persists for at least 3 weeks without significantly fading or leaking. 

The rats were euthanized at a predefined time, and their eyes were immediately enucleated. The eyes were subsequently fixed in 4% paraformaldehyde for 1 hour. After washing the retinas with saline, the retinas were dissected and prepared as flat mounts by making four radial incisions and then placing them on slides in gold antifade mountant (Thermo Fisher Scientific, Carlsbad, CA, USA). 

To quantify the labeled RGCs, we observed the retinal slides under a fluorescence microscope using UV excitation (330 to 385 nm). Each retina was visually divided into four quadrants, and each quadrant was divided further into central (1.5 mm from the optic disc) and peripheral regions (3 mm from the optic disc). Two fields in each region were counted. Sixteen microscopic fields in each retina were counted at a final magnification of ×200. The data were expressed as the percentage difference in the numbers of RGCs between the right and left eyes (left eye percentage, mean ± SEM).^16,20 

UHPLC-MS/MS Measurement of d-Serine

After anesthetizing the rats, a capillary tube was used to collect approximately 20 μL aqueous fluid from each eye. The aqueous fluid drawn from the two eyes was combined as one sample for analysis. The rats were euthanized, and the two retinas were also collected as one sample for analysis. For each retina, an aliquot of each individual sample was precisely weighed, extracted using 200 μL extraction solvent (precooled at −20°C; acetonitrile-methanol-water, 2:2:1), and cleared by centrifugation. For aqueous humor, a 40-μL aliquot of each individual sample was added to 120 μL methanol and collected by centrifugation. Derivatization was performed according to a previous report with a few modifications. Briefly, a 50-μL aliquot of the clear supernatant (or standard solution) was dried under a gentle nitrogen flow. The residual was reconstituted with 50 μL 10% methanol and then mixed with 50 μL of 5% Nα-(2,4-dinitro-5-fluorophenyl)-l-alaninamide (FDAA) in acetone and 10 μL 1 mol/L sodium bicarbonate. The reaction mixtures were incubated at 40°C for 60 minutes. After the reaction, 10 μL 2 mol/L hydrochloride was added. Each tube was mixed thoroughly and dried under a gentle nitrogen flow. The residual was reconstituted with 100 μL methanol and centrifuged at 12,000 rpm and 4°C for 15 minutes. The clear supernatant was subjected to Ultra High Performance Liquid Chromatography coupled to triple-quadrupole Mass Spectrometry (UHPLC-MS/MS) analysis. UHPLC separation was carried out using an Agilent 1290 Infinity II series UHPLC System (Agilent Technologies, Santa Clara, CA, USA) equipped with a Waters ACQUITY UPLC C18 column (100 × 2.1 mm, 1.7 μm; Waters, Milford, MA, USA). An Agilent 6460 triple quadrupole mass spectrometer (Agilent Technologies) equipped with an Agilent jet stream electrospray ionization (AJS-ESI) interface was applied for assay development. Agilent MassHunter Work Station Software (B.08.00; Agilent Technologies) was used for multiple reaction monitoring data acquisition and processing. 

Western Blot Analysis

Western blot analysis was performed as previously described.²¹ The anti-Glial Fibrillary Acidic Protein (GFAP) (mouse, 3670S) antibody was supplied by Cell Signaling Technology (Beverly, MA, USA), and the indicated anti-SR antibody (rabbit, ab45434) was provided by Abcam (Cambridge, UK). The immunoreactive proteins were visualized using a chemiluminescence detection kit (ECL Western Blotting Substrate; Thermo Fisher Scientific) and analyzed using Gelpro Analyzer Software (Version 4.0; Media Cybernetics, Rockville, MD, USA). 

Immunohistofluorescence

Paraformaldehyde-fixed (4%) and optimal cutting temperature compound (frozen section medium)-embedded rat eyes were sectioned at a thickness of 10 μm and stained with the following primary antibodies using standard methods¹⁶: anti-SR (rabbit, ab45434; Abcam), anti-GFAP (mouse, 3670S; Cell Signaling Technology, Danvers, MA, USA), and anti-Brn3A (goat, sc-31984; Santa Cruz Biotechnology, Dallas, TX, USA). The nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI). 

TUNEL Assay

A TUNEL assay was used to detect cell apoptosis. An In Situ Cell Death Detection Kit (TMR red; Roche, Mannheim, Germany) was used to process retinal cryosections to detect apoptotic cells, whereas Brn3A (goat, sc-31984; Santa Cruz Biotechnology) antibody was used as an RGC marker. Nuclei were stained with Hoechst 33258. 

For quantification, three to four sections from each retina were used, and on each section, four fields (each of 500-μm retinal length), comprising two fields from the periphery and two from the central regions, were analyzed. In each image, the number of TUNEL-positive cells in GCL was counted. 

Statistical Analysis

Data are expressed as the mean ± SEM. Paired data from two groups were compared using paired t-tests, and other unpaired data from two groups were compared using t-tests. One-way ANOVA was used to compare means among multiple groups. Multiple comparison between the groups was performed using Bonferroni test as a post hoc test. P < 0.05 was considered statistically significant. 

Results

RGC Loss in the COH Model

We generated the COH model by cauterizing three episcleral veins, as previously described.^16,17 To confirm that we had successfully developed the COH model, we measured IOP and assessed RGC loss. We measured IOP in all animals before surgery and 1, 2, 3, and 6 weeks after surgery with a TonoLab (Icare, Espoo, Finland). As expected, the IOP of the right experimental eye increased significantly after surgery compared with that of the left control eye (Fig. 1A). The IOP of the right experimental eye remained at a high level (∼17 mm Hg) for 6 weeks after surgery (16.39 ± 1.04 mm Hg, 1 week after surgery; 17.56 ± 1.20 mm Hg, 2 weeks after surgery; 17.50 ± 1.84 mm Hg, 3 weeks after surgery; 17.17 ± 0.95 mm Hg, 6 weeks after surgery), and the IOP of the left control eye remained at a lower level (∼11 mm Hg) during the same time period. 

Figure 1

 

RGC loss in the COH model. (A) Measurements of IOP before and 1, 2, 3, and 6 weeks after surgery. The IOP of the glaucomatous right eye increased significantly at all time points compared with that of the control left eye. The results are expressed in mm Hg and presented as the mean ± SEM; ***P < 0.001, **P < 0.01, *P < 0.05 indicate significant differences from control eyes (n = 6 for each group). (B–F) Representative enlarged images of the flat mounted retinas showing FluoroGold-labeled RGCs in control eyes and COH model eyes. Schematic diagram showing the different regions of the retinas (B). The flat mounts of the retinas from the glaucomatous right eyes (COH model eyes) (E, F) displayed significantly greater RGC survival than those from the control eyes (C, D) at 2 weeks after surgery. (G) Quantitative analysis of RGC survival. Values are the mean ± SEM; ***P < 0.001 comparing the COH group with the control group, paired t-test (n = 6 for each group).

To assess RGC survival in the rat COH model, we counted the numbers of FluoroGold-labeled RGCs in flat-mounted retinas. As shown in Figures 1B–1F, the right experimental eye contained less FluoroGold-labeled RGCs than the left control eye. Quantitative analysis showed that the RGC density in the left control eye was 2290 ± 70 cells/mm² in the central region of the eye and 2069 ± 29 cells/mm² in the peripheral region of the eye, which was comparable to the results of a previous report.²² The RGC density in the right experimental eye was 1824 ± 78 cells/mm² in the central region of the eye and 1638 ± 31 cells/mm² in the peripheral region of the eye at 2 weeks after surgery (Fig. 1G). 

Increased d-Serine and SR Expression in the Retina of the COH Model

Previous studies have shown that the excitotoxicity caused by NMDA receptor (NMDAR) overactivation plays a major role in RGC apoptosis.^4,23 Recent studies have demonstrated that d-serine functions as an NMDAR coagonist and thus plays a key role in NMDAR activity and NMDAR-mediated neurotoxicity. Therefore, we tested whether the content of d-serine increases in the eyes of COH model rats. We found that the content of d-serine increased significantly in the retina 2 weeks after surgery compared with the control retina, measured by UHPLC-MS/MS, although there was no significant difference between glaucomatous and control aqueous humor (Fig. 2). We also tested whether d-serine synthetase SR expression increases in the eyes of COH model rats. We tested retinal SR expression using Western blot and found that SR expression in the experimental right eye increased slightly after surgery compared with that in the left control eye and remained at this high level for 6 weeks after surgery (Figs. 3A, 3B). We also detected increased glial fibrillary acidic protein (GFAP) expression in Müller cells of the right experimental eye compared with the left control eye. Moreover, we observed continuous increases in GFAP immunoreactivity in response to RGC damage in the former eye compared with the latter eye (Figs. 3A, 3C). To determine the localization of SR expression in the retina, we stained tissue sections for SR and performed immunohistochemistry analysis. We noted an increased number of SR-positive cells in the ganglion cell layer (GCL) in the experimental retina compared with the corresponding layer in the control retina at 2 weeks after surgery. Immunohistochemistry analysis showed that SR immunoreactivity was reflective of the presence of both Müller cells (Fig. 3D; GFAP-immunoreactive cells) and RGCs (Fig. 4; Brn3A-immunoreactive nuclei) and that SR and Brn3A exhibited partial colocalization in the GCL. 

Figure 2

 

Increased d-serine in the retinas of glaucomatous rats determined by UHPLC-MS/MS. (A) Amino acid standards were separated by UHPLC-MS/MS: 1, D-Ser, T_R = 5.79 minutes; 2, glycine, T_R = 6.55 minutes. (B, C) Retina samples from control rats (B) and glaucomatous rats (C) at 2 weeks after surgery. (D, E) Aqueous humor samples from control rats (D) and glaucomatous rats (E) at 2 weeks after surgery. (F) Quantification of d-serine in the retinas from glaucomatous and control rats at 2 weeks after surgery. d-Serine in the retinas from glaucomatous eyes significantly different from that in control eyes (n = 3 for each group). (G) Quantification of d-serine in aqueous humor from glaucomatous rats and control rats at 2 weeks after surgery. The results shown are the mean ± SEM from triplicate experiments; *P < 0.05, unpaired t-test.

Figure 3

 

SR expression increased in the retinas of the COH model rats. (A) Western blotting analysis of protein expression in the entire retina before and 1, 2, 3, and 6 weeks after surgery. SR expression increased in COH model eyes compared with control eyes, and this change was accompanied by increases in GFAP expression in the former group compared with the latter group. The full-length blots are included in the Supplementary Information. (B) Quantitative analysis of SR expression. The results were normalized to β-actin expression and presented as fold differences in SR expression between the glaucomatous right eyes (COH model eyes) (R) and control left eyes (L). The results are presented as the mean ± SEM; ***P < 0.001, *P < 0.05 indicate significant differences from control eyes (n = 6 for each group) (1-way ANOVA). (C) Retinal sections from the control and glaucomatous eyes of the rats were immunostained for GFAP (green), SR (red), and DAPI (blue) and then imaged with a confocal microscope 2 weeks after surgery.

Figure 4

 

SR expression was localized in RGCs. Representative confocal images showing SR (red) immunolabeling in Brn3a (green)-positive RGCs. The nuclei were stained with DAPI (blue). The retinal sections were obtained from all the control eyes and glaucomatous eyes (COH model eyes) of the rats at 2 weeks after surgery. Representative enlarged and merged images confirming that SR expression was colocalized with the RGC marker Brn3a.

Decreases in d-Serine Expression Can Reduce RGC Loss in the COH Model

d-Serine is specifically degraded by DAAO. To investigate the neuroprotective effects of DAAO on RGC survival in the glaucomatous retina, we retrogradely labeled RGCs by injecting FluoroGold into the superior colliculi and analyzed RGC loss. As shown in Figures 5A–5H, DAAO treatment significantly increased RGC survival in DAAO-treated glaucomatous eyes compared with saline-treated glaucomatous eyes 2 weeks after surgery. Even 4 weeks after surgery, DAAO treatment still was able to increase RGC survival. (Figs. 5I–5L) The quantitative analysis of the FluoroGold-positive RGCs in the retinal mounts is shown in Figures 5M and 5N. Sham-operated retinas treated with DAAO displayed an RGC density similar to that of control retinas. RGC density in the glaucomatous eye treated with DAAO 2 weeks after surgery increased significantly to 2188 ± 43/mm² in central region of the eye and 1872 ± 56/mm² in the peripheral region of the eye, whereas RGC density in the glaucomatous eye treated with saline decreased to 1834 ± 60/mm² in the central region of the eye and 1529 ± 86/mm² in the peripheral region of the eye. Enhanced survival of RGCs was also found 4 weeks after surgery in the glaucomatous eyes treated with DAAO compared with those treated with saline, both in the central and peripheral regions. 

Figure 5

 

FluoroGold-labeled RGCs in glaucomatous eyes after DAAO treatment. (A–L) Representative fields in flat-mounted retinas showing RGC survival in control eyes and glaucomatous eyes. B, D, F, H, J, and L are enlarged images of A, C, E, G, I, and K. The flat-mounted retinas from the saline-treated normal eyes (A, B), DAAO-treated normal eyes (C, D), saline-treated glaucomatous eyes 2 weeks after surgery (E, F), DAAO-treated glaucomatous eyes 2 weeks after surgery (G, H), saline-treated glaucomatous eyes 4 weeks after surgery (I, J), and DAAO-treated glaucomatous eyes 4 weeks after surgery (K, L) were all labeled with FluoroGold. The control eyes were obtained from the rats at 2 weeks after surgery. (M) Quantitative analysis of RGC survival in central region. (N) Quantitative analysis of RGC survival in peripheral region. The saline group compared with the DAAO group, the COH+Saline-2W group compared with the COH+DAAO-2W group, and the COH+Saline-4W group compared with the COH+DAAO-4W group (1-way ANOVA followed by Bonferroni posttest; n = 8 to 10 for each group). Values are the mean ± SEM; **P < 0.01, ***P < 0.001.

We also performed TUNEL on retinal sections to evaluate retinal cell apoptosis. As shown in Figure 6A, TUNEL-positive cells were mainly in the GCL, partial colocalized with RGCs (Brn3A-immunoreactive nuclei). Quantification of TUNEL-positive cells is shown in Figure 6B, and DAAO reduced the number of TUNEL-positive cells in glaucomatous eyes compared with glaucomatous eyes treated with saline. These results indicate that DAAO can reduce RGC loss in treated eyes compared with untreated eyes in COH model rats. 

Figure 6

 

Prevention increase of TUNEL-positive cells after DAAO treatment in glaucomatous eyes. (A) Retinal sections were immunostained for Brn3a (green), TUNEL (red), and Hoechst (blue) and then imaged with a confocal microscope 2 weeks after surgery. The arrows indicate apoptotic cells. (B) Quantification of TUNEL-positive cells in the GCL (n = 5 for each group). Values are the mean ± SEM and represent the number of TUNEL-positive cells per 500 μm (1-way ANOVA test followed by Bonferroni posttest; ***P < 0.001, *P < 0.05).

Excess d-Serine Could Not Exacerbate RGC Loss in the COH Model

We also attempted to study whether d-serine exacerbates RGC loss in the COH model. We counted the number of FluoroGold-labeled RGCs to assess RGC loss. In the sham-operated retinas, we found that the density of the labeled RGCs in d-serine–treated eyes was 2231 ± 43/mm² in the central region of the eye and 1939 ± 34/mm² in the peripheral region of the eye. These densities were not significantly different from those of the control eye (Fig. 7). In glaucomatous eyes, the RGC density in the saline-treated eye decreased to 1875 ± 44/mm² in the central region of the eye and 1438 ± 78/mm² in the peripheral region of the eye, and the RGC density in the d-serine–treated eye was 1840 ± 43/mm² in the central region of the eye and 1423 ± 84/mm² in the peripheral region of the eye. Thus, RGC density was similar between the glaucomatous eyes treated with saline and those treated with d-serine. Figure 7 shows that RGC loss was similar between the glaucomatous eyes treated with d-serine and those treated with saline. These results indicate that saturating amounts of d-serine had no effect on RGC death in glaucomatous eyes. 

Figure 7

 

RGC survival in glaucomatous eyes after d-serine treatment. (A–H) Representative fields from flat-mounted retinas showing RGC survival in control eyes and glaucomatous eyes. B, D, F, and H are enlarged images of A, C, E, and G. The flat-mounted retinas from the saline-treated normal eyes (A, B), d-serine–treated normal eyes (C, D), saline-treated glaucomatous eyes (E, F), and d-serine–treated glaucomatous eyes (G, H) were all labeled with FluoroGold. The control eyes and glaucomatous eyes were obtained from the rats at 2 weeks after surgery. (I) Quantitative analysis of RGC survival. The Saline group compared with the d-serine group and the COH+Saline group compared with the COH+D-serine group. (1-way ANOVA test followed by Bonferroni posttest; n = 8 for each group). There were no significant differences between the groups.

Discussion

d-Serine has been reported to regulate NMDAR activity and act as a main NMDAR coagonist in the vertebrate retina.²⁴ It has been confirmed that d-serine rather than glycine is the dominant coagonist required for neuronal cell death elicited by NMDAR stimulation in the hippocampus.²⁵ Endogenous d-serine also participates in inducing NMDA-mediated neuronal death in rat cerebrocortical slices.²⁶ However, whether d-serine is involved in retinal neuronal death remains unclear. RGC loss is a hallmark of various retinal diseases, including glaucoma, retinal ischemia, and diabetic retinopathy, and NMDAR-mediated excitotoxicity is thought to be an important contributor to RGC death in these diseases.²⁷ Therefore, further research about the role of d-serine in glaucoma-induced RGC death is warranted. 

A previous study showed that NMDA-induced RGC death was enhanced by d-serine administration and decreased by DAAO administration.¹³ Moreover, less extensive NMDA-induced RGC death was noted in mice displaying severe SR protein depletion than in mice with higher SR protein expression levels.¹⁴ The results of these studies show that decreased d-serine levels can protect against NMDA-induced RGC loss. In our study, we investigated whether d-serine participates in RGC death in the rat COH model, a chronic glaucoma model characterized by prolonged and moderate intraocular hypertension. We found that d-serine and SR expression was increased in retinas with elevated IOP and that a greater density of SR-positive cells was found mainly in the GCL. These findings are consistent with those of a previous study showing that RGCs displayed positive staining for SR.²⁸ However, we found that the IOP of glaucomatous eyes was approximately 17 mm Hg, whereas the IOP of control eyes was 10 mm Hg (Fig. 1); these results contrast with those of previous reports, as the IOP elevation noted herein was more modest than that noted in other studies.^18,28,29 It is possible that the degree of IOP elevation is affected by the strain induced in the rats, as individual rats have different sensitivities to vein cauterization surgery. Moreover, whether IOP remains at a high level is influenced by age. The IOP of aged rats may remain at a high level because these rats recover more slowly than younger rats. 

Müller cells, which are specialized retinal glial cells, play an important role in retinal homeostasis. GFAP is minimally expressed in astrocytes and Müller cells in normal retinas and its expression is highly upregulated in activated Müller cells.³⁰ In our study, we also noted increased GFAP expression at all time points (Fig. 3). This result showed that increases in SR expression accompanied Müller cell activation. As d-serine is believed to potentiate NMDA responses in the retina following its release from Müller cells, we plan to further investigate the impact of the relationship between d-serine and Müller cells on the pathogenesis of glaucoma. 

Several researchers have sought to identify drugs that prevent excessive NMDAR activation to decrease excitotoxic damage while preserving normal neuronal function to avoid side effects. Uncompetitive NMDA receptor antagonists are required to block excess NMDA receptor activation.^23,31 Memantine, one of the most common noncompetitive NMDA receptor antagonists, has been used clinically for the treatment of Parkinson's disease and Alzheimer's disease and has displayed an excellent safety profile for over 20 years in Europe. Memantine has also been shown to exert neuroprotective effects in mouse, rat, and monkey models of glaucoma.¹⁸ However, human trials regarding the use of memantine for the treatment of open-angle glaucoma have yielded confusing and disappointing results.³² Thus, it is necessary to continue the search for new noncompetitive NMDA receptor antagonists. Antagonists that compete with d-serine may represent new drugs that can protect RGCs in patients with glaucoma. In our study, we investigated the neuroprotective effects of DAAO on RGC survival in glaucomatous retinas and found that DAAO can reduce RGC loss in the COH model. It is possible that the selective elimination of d-serine attenuates NMDAR activation. NMDAR downregulation may result in decreased entry of many factors, such as Ca²⁺. This hypothesis should be confirmed by further study. 

A previous study showed that NMDA-induced RGC death was enhanced by d-serine administration in vivo.¹³ In our study, we also investigated whether d-serine enhanced RGC loss in the COH model, and we found that RGC loss was similar between glaucomatous eyes treated with d-serine and glaucomatous eyes treated with saline (Fig. 7). It is possible that the elevated levels of d-serine in glaucomatous eyes produced maximal NMDA activation, which was not further accentuated by exogenous administration of d-serine. However, d-serine did not induce more RGC death than saline in the COH model. 

In conclusion, we demonstrated that increased SR expression, which increased d-serine levels, contributed to RGC death by exacerbating NMDA receptor activation and that DAAO had neuroprotective effects on RGC survival in the rat model of glaucoma. These results may provide clues regarding the pathogenesis of glaucoma and suggest that SR may be a new drug target for the clinical treatment of glaucoma. 

Acknowledgments

Supported by the National Natural Science Foundation of China (Grants 81400395, 81470624, 81470623, 81770925, and 81790641). 

Disclosure: X. Zhang, None; R. Zhang, None; X. Zhou, None; J. Wu, None