Extracellular Signal-Regulated Kinase 1/2 Pathway Is Insufficiently Involved in the Neuroprotective Effect by Hydrogen Sulfide Supplement in Experimental Glaucoma

Abstract

Purpose: Glaucoma is a neurodegenerative eye disease characterized by gradually impaired visual field and irreversible blindness due to retinal ganglion cell (RGC) loss. Our previous studies have confirmed that hydrogen sulfide (H₂S) takes part in the glaucomatous process and contributes to RGC protection. The present study aimed to further investigate the role of extracellular signal-regulated kinase 1/2 (ERK 1/2) pathway underlying the impact of H₂S, to better understand the mechanism through which H₂S exerts neuroprotection in glaucoma.

Methods: An established rat glaucoma model was used and 168 rats were qualified to undergo sodium hydrosulfide (NaHS, a H₂S donor)/PD98059 (an ERK inhibitor) treatment. Then the survival and apoptosis of RGC were evaluated through retrograde labeling and TUNEL staining, along with activity evaluations of ERK 1/2 pathway, intrinsic apoptotic pathway, glial activation, nuclear factor kappa B (NF-κB) pathway, nicotinamide adenine dinucleotide phosphate (NADPH) oxidase, autophagy, and TNF-α production through immunohistochemistry, Western blotting, and ELISA.

Results: The study demonstrated that NaHS suppressed ERK 1/2 pathway activity similarly to PD98059 in retinas of experimental glaucoma rats, while PD98059 also similarly suppressed glial activation, NF-κB pathway, NADPH oxidase, and TNF-α production. However, PD98059 did not affect RGC survival, apoptotic regulation, or autophagy as NaHS did.

Conclusions: Our study indicated that inhibition of ERK 1/2 pathway might partly contribute to the neuroprotection by H₂S in experimental glaucoma; however, it was insufficient to initiate the therapeutic effect on its own.

Recently hydrogen sulfide (H₂S) has been recognized as the third endogenous gaseous signaling molecule alongside carbon monoxide (CO) and nitric oxide (NO).^1–3 It is mainly synthesized by cystathionine beta-synthase (CBS), cystathionine gamma-lyase (CSE), and 3-mercaptopyruvate sulfurtransferase (3-MST) in various mammalian tissues.¹ As a potent reductant, H₂S regulates the conformations and activities of certain biological enzymes and signal channel receptors by reducing the disulfide bonds of cysteine residues, and it also enables the functional modulation of certain proteins by modifying cysteine residues with sulfhydration or trisulfide bridging.⁴ It has been reported that H₂S plays critical roles in multiple physiological and pathological processes, such as vasorelaxation, neuroendocrine regulation, and inflammation.^5–8 More interestingly, H₂S has shown promising therapeutic efficiency potential in situations of morbidity within multiple body systems. It works to alleviate inflammatory response, oxidative stress, and energy shortage under disease conditions like ischemia–reperfusion injury and neurodegenerative disorders, and has increasingly aroused the interests of pharmacologists, especially those devoted to the difficult enterprise of neuroprotection.^9,10 

As far as neuroprotection is concerned, research on the treatment of glaucomatous neuropathy has always been one of the most complicated tasks. As a kind of neurodegenerative eye disease, glaucoma is characterized by gradually impaired visual field and latent prognosis of irreversible blindness due to apoptosis of retinal ganglion cells (RGC), for which elevated intraocular pressure (IOP) is recognized as a primary risk factor.¹¹ Although the mechanism underlying the genesis and progress of RGC apoptosis in glaucoma has not been fully elucidated, abundant evidence has shown that glial activation, oxidative stress, inflammatory reaction, and autophagic response are closely involved.^12–15 To date, lowering IOP through pharmacological intervention, laser therapy, or filtration surgery has proven to be the only approach that has definite efficiency in treating glaucoma clinically.¹⁶ However, even achievement of satisfactory IOP reduction may not slow or stop the deterioration of visual function in the course of glaucoma.¹⁷ Thus there exists a strong demand for exploration of neuroprotective strategies independent of IOP lowering to ameliorate glaucomatous neuropathy, and in this respect, H₂S apparently tends to perform effectively, considering significant findings in the treatment of other neurodegenerative diseases in the central nervous system.^18–20 

In line with expectations, H₂S has shown profound involvement in various retinal neuropathy processes, while exogenous donors exhibited therapeutic potential in conditions of several retinal diseases. It has been confirmed that CBS, CSE, and 3-MST all existed in mammalian retina and positively correlated with the endogenous H₂S level in retina under different pathological situations.²¹ Exogenous H₂S supplement was found to improve the survival of RGC that suffered N-methyl-D-aspartic acid (NMDA)-induced excitatory neurotoxicity, diabetic retinopathy, and ischemia–reperfusion injury, and the neuroprotective effect of H₂S was found to be partly attributed to its capability of oxidation resistance, inflammation suppression, and energy–productivity restoration.^22–24 In the field of glaucoma, our research team first verified downregulation of retinal expressions of H₂S synthases, accompanied by a decrease of endogenous H₂S level in experimental glaucoma. We further observed that restoration of retinal H₂S content with exogenous donor (sodium hydrosulfide, NaHS) effectively reversed the loss of RGC.²⁵ Then Liu et al.²⁶ and Ozer et al.²⁷ also reported preservation of RGC from glaucomatous neuropathy with H₂S donors in vitro and in vivo. Afterward, we kept on with this subject and illustrated the antiapoptosis effect of H₂S on RGC as well as involvement of various pathological activities in experimental glaucoma.²⁸ These discoveries generally outline the role of H₂S in glaucoma and have increasingly drawn attention to the therapeutic potential of H₂S in glaucoma treatment. 

Since research that focuses on H₂S in connection with glaucoma has emerged only in the last several years, solid information concerning this subject is still sparse and remains to be thoroughly explored. Our previous work uncovered a trivial part of the whole veil and obtained results of some significance; however, the conclusions appeared to just scratch the surface and did not reveal the internal mechanism. Thus, in the present study, we repeated previous experimental procedures and investigated the role of extracellular signal-regulated kinase 1/2 (ERK 1/2), which is involved in many critical physiopathological processes as well as in glaucomatous neuropathy. We expected to better understand the mechanism through which H₂S exerts neuroprotection in experimental glaucoma. 

Materials and Methods

Animals

All animals enrolled in this study underwent procedures in accordance to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Commercially available male Sprague-Dawley rats (8 weeks of age and 200–250 g weight) were provided by Shanghai Lab Animal Research Center (Shanghai, China). The experimented rats were fed ad libitum and housed in a standardized room with conditioned temperature and humidity as well as day–night rhythm illumination. Every surgery required was performed under systemic and topical anesthesia to minimize the suffering of the rats. 

Induction of Glaucoma Model and IOP Measurement

The chronic ocular hypertension (COH) model was induced in rats to simulate experimental glaucoma as described in our previous work.²⁵ Only right eyes were operated on. Briefly, rats were anesthetized with an intraperitoneal injection of ketamine hydrochloride (25 mg/kg; Sigma-Aldrich Corp., St. Louis, MO, USA) and xylazine (10 mg/kg; Sigma-Aldrich Corp.) plus topical administration of 0.5% proparacaine hydrochloride eyedrops (Bausch & Lomb, Tampa, FL, USA). A Hystem Cell Culture Scaffold kit (HCCS; Sigma-Aldrich Corp.) was used to produce a kind of premixed in situ cross-linking hydrogel, and 7 μL was injected into the anterior chamber of each eye immediately after mixing, with the use of a Hamilton microsyringe equipped with a 31-gauge needle (Hamilton Bonaduz AG, Bonaduz, Switzerland). Finally, a drop of ofloxacin ophthalmic ointment (Sinqi Pharmaceutical Corp., Shenyang, China) was placed over cornea to prevent postoperation infection. 

IOP measurement was achieved with a rodent-customized rebound tonometer TonoLab (Icare, Vantaa, Finland). Measurements were performed between 10 AM and 2 PM under transient systemic anesthesia with isoflurane inhalation (2%–4%), just before the surgical operation and once weekly after the operation. Rats were euthanized 4 weeks after surgery, and samples were collected for the following procedures. The qualifying criterion for COH was an IOP elevation of at least 30% increase compared to the individual preoperation. 

Study Design and Pharmacological Administration

Rats were randomly divided into COH, COH + NaHS, COH + PD98059, and COH + NaHS + PD98059 groups with the random number generator of Microsoft Office Excel (Professional Plus 2013 Version; Microsoft, Redmond, WA, USA). The dose and frequency of NaHS administration followed the scheme determined in our previous work.²⁵ The NaHS solution was intraperitoneally injected (5.6 mg/kg; Sigma-Aldrich Corp.) into the COH + NaHS group and the COH + NaHS + PD98059 group, beginning 3 days before COH induction and maintenance of daily delivery until euthanasia, that is, for 4 weeks after surgery, while saline of an equivalent volume was synchronously intraperitoneally injected into the COH group and COH + PD98059 group. PD98059 is a specific inhibitor of mitogen-activated protein kinase (MAPK) activation by MAPK/ERK signal cascade and indirectly blocks the activation of the ERK 1/2 pathway.²⁹ In the present study, PD98059 was intravitreally injected as previously described with minor modifications.^30,31 Briefly, rats were systemically anesthetized and pupils dilated, and for the COH + PD98059 group and COH + NaHS + PD98059 group, 2 μL PD98059 (5 mmol/L; Sigma-Aldrich Corp.) dissolved in 10% dimethylsulfoxide (DMSO; Sigma-Aldrich Corp.) was injected into the vitreous cavity with a pulled glass micropipette, beginning 3 days before COH induction and maintaining weekly delivery until euthanasia, while 2 μL 10% DMSO was synchronously injected into the vitreous cavity of the COH group and COH + NaHS group. Ofloxacin ophthalmic ointment (Sinqi Pharmaceutical Corp.) was used at the end. 

Retrograde Labeling and Counting of RGC

To determine the survival status of RGC in experimented rats, retrograde labeling of RGC with a 1,1′-dioctadecyl-3,3,3′,3′-tetramethy-indocarbocyanin perchlorate (DiI; Life Technologies, Grand Island, NY, USA) injection into the superior colliculus and subsequent RGC counting were performed as described in our previous study.²⁵ Briefly, the labeling dye was prepared by dissolving 40 mg DiI in 1 mL DMSO. Following systemic anesthesia and headlock of the rat, the skin and muscles over the parietal bone were removed, and cranium piercing was performed at bilateral points 6.4 mm behind the fonticuli anterior and 1.5 mm apart from the midline. Then the needle of a Hamilton 10-μL microsyringe was vertically inserted 4.0 mm deep from the skull surface, and 1.5 μL labeling dye was gently injected at each point, respectively. Seven days later, the labeled rats were perfused transcardially with saline and 4% paraformaldehyde (PFA; Sigma-Aldrich Corp.) in sequence. After further postfixing in 4% PFA for 1 hour, the retinal neuroepithelium layers were detached and then whole mounted on slides. The RGC-labeled flat preparations were scanned and captured under a Zeiss Imager M1 microscope (Carl Zeiss Microscopy, Jena, Germany). RGC counting was performed in fields of view at the distances of approximately one-sixth, three-sixths, and five-sixths retinal radius from the optic disc in each quadrant, employing an image analysis system (Image Pro Plus Version 6.0; Media Cybermetics, Silver Spring, MD, USA). The average density of labeled RGC at the same eccentricity was considered to represent the number of viable RGC for the particular retinal position. 

Isolation of RGC From Whole Retinas

To evaluate the apoptotic status of RGC in experimented rats, we first isolated and collected RGC from whole retinas of distinct groups with a classical two-step immunopanning procedure as previously described with minor modifications.^25,32 Briefly, a 10-cm culture dish coated with anti-rat Thy 1.1 antibody (1:500 dilution, MCA47R; AbD Serotec, Kidlington, Oxford, UK) for positive selection and a 10-cm culture dish coated with anti-rat CD11b antibody (1:500 dilution, MCA275R; AbD Serotec) for negative selection were freshly prepared 1 day before panning. Six rats from a single group were euthanized by CO₂ and the retinal neuroepithelia were dissociated from the right eyes and incubated in papain solution (16.5 U/mL; Worthington Biochemical Corp., Lakewood, NJ, USA) containing DNase I (0.004%; Sigma-Aldrich Corp.) at 37°C for 30 minutes. After digestion, the tissues were repeatedly washed and triturated into separated and suspended cells in bovine serum albumin (BSA; Sigma-Aldrich Corp.)/ovomucoid (Worthington Biochemical Corp.) solutions of gradient concentrations. Then the cells were transferred to and incubated in the negative-selection dish and positive-selection dish successively, being incubated for 1 hour for each panning step. Finally, after washing away unattached cells in the positive-selection dish, the remaining cells that were bonded to Thy 1.1 antibody on the dish surface were regarded as RGC and were collected for further apoptosis assay. Panning of every six retinas yielded approximately 5 × 10⁵ isolated RGC, which were regarded as a single sample statistically and underwent further experimentation. Isolated RGC were seeded on adhering pretreated coverslips placed in a 24-well plate at a density of approximately 5 × 10⁵ cells/mL culture medium for each well. 

TUNEL Assay of RGC

Nuclear DNA fragmentations in apoptotic RGC were visualized with TUNEL, using an In Situ Cell Death Detection kit (Roche Diagnostics Corporation, Indianapolis, IN, USA). According to manufacturer's instructions, the adherent cells were fixed in 4% PFA for 30 minutes and subsequently permeated with cold 0.1% Triton X-100 (Sigma-Aldrich Corp.) for 10 minutes. Then the cells were incubated with TUNEL reaction reagent, consisting of 10 μL specific enzyme solution and 90 μL fluorescein-conjugated labeling solution, for 1 hour at 37°C. Afterward, the adherent cells were mounted with fluoromount medium (Life Technologies) incorporated with 4′,6-diamidino-2-phenylindole (DAPI). Fluorescence imaging was recorded with a Zeiss Imager M1 microscope and analyzed with the use of Image Pro Plus software (Version 6.0, Media Cybermetics). The rates of TUNEL-positive cells in six randomly selected fields of view on one coverslip were averaged to reflect the apoptotic status of RGC. 

Immunohistochemistry of Retinal Tissue

Rats were perfused transcardially with saline and 4% PFA in sequence. Right eyes were enucleated and postfixed in 4% PFA at 4°C overnight, and went through dehydration using sucrose solutions of gradient concentrations. Afterward, the anterior segments were discarded and the remaining eye cups were embedded in specialized medium, and were further frozen to be prepared for sectioning. Eye cups were sectioned nasotemporally across the optic disc at a thickness of 7 μm. Three slices that were randomly selected from one eye went through staining for a particular target protein. The slices were permeated in cold 0.3% Triton X-100 for 30 minutes and further blocked in 1% BSA (Sigma-Aldrich Corp.) for 1 hour. Then the slices were respectively subjected to incubation with primary antibodies to Bax (1:100 dilution, sc-526; Santa Cruz Biotechnology, Dallas, TX, USA), Bcl-2 (1:100 dilution, sc-783; Santa Cruz Biotechnology), cleaved caspase-3 (1:200 dilution, #9664; Cell Signaling Technology, Danvers, MA, USA), Iba-1 (1:1000 dilution, #019-19,741; Wako Pure Chemical Industries, Osaka, Japan), and GFAP (1:500 dilution, #12389; Cell Signaling Technology) at 4°C overnight. The slices were further incubated with fluorescein-conjugated secondary antibody at room temperature for 90 minutes, and were mounted with the DAPI-incorporated fluoromount medium and scanned with the Zeiss Imager M1 microscope. Images captured were analyzed with Image Pro Plus software (Version 6.0, Media Cybermetics). The method of image capture and analysis of fluorescent density were in accordance with our previous work.²⁵ Briefly, three randomly selected fields of view within 1-mm distance from each unilateral optic disc margin were outlined, and the mean optical density of fluorescent staining was calculated. 

Western Blotting

Protein contents were extracted from homogenized retinas and went through quantification of concentration. Samples of equivalent protein content were separated by electrophoresis with a Mini-Protean electrophoresis device (Bio-Rad, Hercules, CA, USA), and then proteins were transferred to polyvinylidene fluoride membranes in a Mini Trans-Blot electrophoretic transfer device (Bio-Rad). The membranes were first blocked with 5% skim milk for 2 hours and then incubated, respectively, with primary antibodies to Bax (1:200 dilution, sc-526; Santa Cruz Biotechnology), Bcl-2 (1:200 dilution, sc-783; Santa Cruz Biotechnology), cleaved caspase-3 (1:1000 dilution, #9664; Cell Signaling Technology), Iba-1 (1:1000 dilution, #019-19,741; Wako Pure Chemical Industries), GFAP (1:1000 dilution, #12389; Cell Signaling Technology), IκBα (1:1000 dilution, #9247; Cell Signaling Technology), phospho-IκBα (1:1000 dilution, #9246; Cell Signaling Technology), phospho-p65 (1:1000 dilution, #3033; Cell Signaling Technology), gp91^phox (1:1000 dilution, ab80508; Abcam, Cambridge, UK), Beclin-1 (1:1000 dilution, #3495; Cell Signaling Technology), LC3 (1:500 dilution, #4108; Cell Signaling Technology), ERK 1/2 (1:1000 dilution, #9107; Cell Signaling Technology), phospho-ERK 1/2 (1:500 dilution, ab76165; Abcam), and β-actin (1:1000 dilution, #3700; Cell Signaling Technology) at 4°C overnight. After the membranes were further incubated with horseradish peroxidase–conjugated secondary antibodies for 90 minutes, the target protein blots were visualized with an ImageQuant LAS 4000 mini system (GE Healthcare Bio-Sciences, Piscataway, NJ, USA). Finally, the relative density of the bands was calculated with the Image Pro Plus software (Version 6.0, Media Cybermetics). 

ELISA Assay of TNF-α

TNF-α level in retinal tissue was measured with the application of a specialized TNF-α ELISA Kit (R&D Systems, Minneapolis, MN, USA) according to the manufacturer's instructions. The concentrations of TNF-α in distinct samples were calculated against a standard curve, which was achieved by fitting values of standard samples from the kit. 

Statistical Analysis

All data obtained were analyzed with SPSS software (Version 19.0; IBM, Armonk, NY, USA). The normality of data distributions was sustained with Shapiro-Wilk test, while the variance homogeneity of data distributions was evaluated with Levene's test. One-way ANOVA, followed by Bonferroni test for condition of equal variances assumed or by Games-Howell test for condition of equal variances not assumed, was employed to evaluate significance statistically. An α level of 0.05 was considered significant. All data were expressed as mean ± SD. 

Results

General Summary of Animal Usage and Profile of IOP Measurements

A total of 242 rats were subjected to operations throughout this study, but 45 rats that failed to meet the qualifying criterion for COH as well as 29 rats that suffered severe inflammation or ocular infection were excluded. Ultimately 168 rats with an average distribution in animal number of each group were included, and the number of animals used in various procedures is specified in Table 1. 

Table 1

 

Number of Rats Used in Various Procedures

 

The IOP measurements of all the grouped rats at different time points are summarized in Table 2. There was no significant difference in IOP values between any two of the groups at each time point (n = 42, P > 0.1 between any two among the COH group, COH + NaHS group, COH + PD98050 group, and COH + NaHS + PD98059 group at each time point, respectively). The data present relatively stable and similar IOP elevation in each group, and indicate that drug administrations did not significantly alter the IOP level in this study. 

Table 2

 

Summary of IOP Measurements at the Various Time Points

 

ERK 1/2 Pathway Activity Was Suppressed by Both NaHS and PD98059 in COH Rats

ERK 1/2 are members of the MAPK family of serine/threonine protein kinases. Once activated, ERK 1/2 phosphorylates specific serines and threonines on target proteins and these act as late-stage messengers to convey external stimuli from cell surface to cellular targets, such as cytoskeletal proteins, to regulate protein translation and transcription factors. As shown by the results of Western blotting, the phosphorylation rate of ERK 1/2 in COH retinas decreased by 74.8% for NaHS treatment, 83.2% for PD98059 treatment, and 81.8% for NaHS + PD98059 treatment (n = 6, each P < 0.001 versus the COH group, respectively), while no significant disparity was observed between any two groups that received distinct pharmacological interventions (n = 6, each P > 0.99 between any two among the COH + NaHS group, COH + PD98050 group, and COH + NaHS + PD98059 group, respectively) (Fig. 1). Thus the results indicate that H₂S supplement suppressed the activity of ERK 1/2 pathway to a similar extent as specific ERK 1/2 inhibitor in experimental glaucoma, and there was neither a synergistic effect nor an antagonistic effect observed in the combination of the two pharmacological interventions. 

Figure 1

 

ERK 1/2 pathway activity was suppressed by both NaHS and PD98059 treatments in COH rats. (A) Representative Western blotting bands of retinal phospho-ERK 1/2 and total ERK 1/2 protein expression of different groups. (B) Semiquantitative analysis of Western blotting gray values (n = 6, *P < 0.05 versus the COH group). Bars represent means ± SD.

Impact of ERK 1/2 Inhibition on RGC Survival in COH Rats Treated With NaHS

The survival status of RGC in each group is intuitively illustrated by retrograde labeling with DiI (Fig. 2). As shown in Figure 2, NaHS treatment alone effectively improved the number of surviving RGC in COH rats (increased by 11.1%, 15.6%, 31.7% at distances of one-sixth, three-sixths, and five-sixths retinal radius, respectively; n = 6, each P < 0.001 versus the COH group), and a similar increment of surviving RGC was also observed in COH rats treated by the combination of NaHS and PD98059 (increased by 10.3%, 17.2%, 35.1% at distances of one-sixth, three-sixths, and five-sixths retinal radius, respectively; n = 6, P = 0.01, < 0.001, and < 0.001 versus the COH group, respectively). However, there was no significant difference in RGC amount between the group treated with NaHS alone and the group treated with a combination of the two drugs (n = 6, each P > 0.99 at corresponding retinal positions, respectively), and besides, PD98059 treatment alone did not achieve a positive effect in saving RGC from loss in COH rats (n = 6, each P > 0.99 at corresponding retinal positions versus the COH group, respectively). The results suggest that simple inhibition of ERK 1/2 pathway did not directly affect RGC survival in experimental glaucoma, nor did it alter the protective effect on RGC introduced by H₂S supplement. 

Figure 2

 

Inhibition of ERK 1/2 pathway did not directly affect RGC survival as H₂S supplement did in COH rats. (A) Representative images of retrograde-labeled RGC with DiI (red fluorescence) in whole-mounted retinas of different groups (magnification ×100, scale bar: 100 μm). (B) Quantitative analysis of survival RGC in distinct retinal positions (n = 6, *P < 0.05 versus the COH group, #P < 0.05 versus the COH + NaHS group, and P < 0.05 versus the COH + PD98059 group). Bars represent means ± SD.

Impact of ERK 1/2 Inhibition on RGC Apoptosis in COH Rats Treated With NaHS

As apoptosis of RGC appears to be the pathological basis of glaucomatous neuropathy, we used the methods of RGC isolation and TUNEL staining to help understand the relationship between ERK 1/2 pathway and the neuroprotection introduced by NaHS treatment in experimental glaucoma (Fig. 3). In line with the results of retrograde-labeled RGC counting, NaHS by itself remarkably reversed the proportion of apoptotic RGC in COH rats (decreased by 48.0%, n = 18, P = 0.003 versus the COH group), while joint administration of NaHS and PD98059 demonstrated a similar antiapoptosis effect on RGC in COH rats (decreased by 57.7%, n = 18, P = 0.001 versus the COH group; P > 0.99 versus the COH + NaHS group), and PD98059 alone did not significantly alter the status of RGC apoptosis in COH rats (n = 18, P = 0.98 versus the COH group). 

Figure 3

 

Inhibition of ERK 1/2 pathway did not directly affect RGC apoptosis as H₂S supplement did in COH rats. (A) Representative images of TUNEL staining (green fluorescence) for apoptosis evaluation of RGC isolated from different groups (magnification ×200, scale bar: 50 μm). (B) Quantitative analysis of TUNEL-positive RGC (n = 18, *P < 0.05 versus the COH group, #P < 0.05 versus the COH + NaHS group, and P < 0.05 versus the COH + PD98059 group). Bars represent means ± SD.

Intrinsic apoptotic pathway, which includes Bcl-2 as an antiapoptotic component and Bax as a proapoptotic component, plays an important role in the initiation of RGC apoptosis in glaucoma, for example, through regulating the activation of the caspase cascade. We further employed immunohistochemistry and Western blotting to confirm the activity variations of intrinsic apoptotic pathway and caspase cascade in situations under different pharmacological interventions. As Figure 4 presents, a simple treatment with NaHS effectively increased Bcl-2 staining intensity, and decreased staining intensity of Bax as well as of cleaved caspase-3 in COH rat retinas, especially within the ganglion cell layer (GCL) (n = 6, P < 0.001, = 0.004, and = 0.02 versus the COH group, respectively). Combined delivery of NaHS and PD98059 led to similar changes in the three proteins in the GCL as NaHS alone did (n = 6, P < 0.001, = 0.01, and = 0.001 versus the COH group, respectively; P > 0.99, > 0.99, and = 0.95 versus the COH + NaHS group, respectively). However, simple PD98059 administration failed to make any significant alteration in comparison with the COH group (n = 6, P > 0.99, > 0.99, and = 0.18 versus the COH group, respectively). Western blotting results (Fig. 5) reconfirmed the discovery from immunohistochemistry, as NaHS treatment actually upregulated the relative ratio of Bcl-2/Bax protein expression (n = 6, P < 0.001 versus the COH group) and downregulated the cleaved caspase-3 protein expression (n = 6, P < 0.001 versus the COH group) in COH rat retinas. And joint administration of NaHS and PD98059 resulted in an effect on Bcl-2/Bax and cleaved caspase-3 protein expression similar to that of NaHS alone (n = 6, each P < 0.001 versus the COH group, respectively ). Accordingly, PD98059 alone did not trigger any obvious variation of either the relative ratio of Bcl-2/Bax protein expression (n = 6, P = 0.67 versus the COH group) or the cleaved caspase-3 protein expression (n = 6, P > 0.99 versus the COH group) in COH rats. 

Figure 4

 

Inhibition of ERK 1/2 pathway did not directly affect intrinsic apoptotic pathway activity as H₂S supplement did in COH rats (immunohistochemistry demonstration). (A–C) Representative immunofluorescence staining images of Bcl-2, Bax, and cleaved caspase-3 proteins (arrowheads, red fluorescence) in retinal slices of different groups (magnification ×200, scale bar: 50 μm). (D–F) Quantitative analysis of fluorescence staining intensities of Bcl-2, Bax, and cleaved caspase-3 in GCL of different groups (n = 6, *P < 0.05 versus the COH group, #P < 0.05 versus the COH + NaHS group, and P < 0.05 versus the COH + PD98059 group). Bars represent means ± SD.

Figure 5

 

Inhibition of ERK 1/2 pathway did not directly affect intrinsic apoptotic pathway activity as H₂S supplement did in COH rats (Western blotting demonstration). (A) Representative Western blotting bands of retinal Bcl-2, Bax, cleaved caspase-3, and internal controlling β-actin protein expression of different groups. (B, C) Semiquantitative analysis of Western blotting gray values (n = 6, *P < 0.05 versus the COH group, #P < 0.05 versus the COH + NaHS group, and P < 0.05 versus the COH + PD98059 group). Bars represent means ± SD.

The above results revealed that simple inhibition of ERK 1/2 pathway was not enough to attenuate RGC apoptosis as H₂S supplement did in experimental glaucoma, and the antiapoptosis effect of H₂S supplement was not to be suppressed or enhanced while ERK 1/2 pathway was specifically inhibited. 

Impact of ERK 1/2 Inhibition on Glial Cell Activation in COH Rats Treated With NaHS

Activation of glial cells has been proven to play an essential role all through the process of glaucomatous neuropathy, and the neuroprotective effect of H₂S in glaucoma is supposed to be partly attributed to its attenuation of glial activation. We employed immunohistochemistry and Western blotting to focus on the variations of retinal protein expression of Iba-1 (a specific marker for microglia cells) and glial fibrillary acidic protein (GFAP, mainly expressed in Müller cells/astrocytes) in our study. As shown in Figures 6 and 7, the Iba-1 staining-positive cells, identified as microglia cells, significantly decreased in amount and altered in shape from bulky body and thick processes to a thinner appearance with NaHS treatment in COH rats, and quantitative analysis of Iba-1 staining/band optical density consistently presented an obvious decrease in both immunohistochemistry (n = 6, P = 0.002 versus the COH group) and Western blotting (n = 6, P < 0.001 versus the COH group). Interestingly, both administration of PD98059 alone (n = 6, both P < 0.001 versus the COH group in immunohistochemistry and Western blotting) and the combination of NaHS and PD98059 (n = 6, both P < 0.001 versus the COH group in immunohistochemistry and Western blotting) resulted in a similar decrease of Iba-1 protein expression and alteration of microglial cell shape as NaHS alone did in COH rats. Furthermore, there was no significant difference in Iba-1 protein expression between any two groups treated with NaHS, PD98059, and NaHS + PD98059 as demonstrated by immunohistochemistry (n = 6, P > 0.99, = 0.25, and > 0.99 for the COH + NaHS group versus the COH + PD98059 group; the COH + NaHS group versus the COH + NaHS + PD98059 group; and the COH + PD98059 group versus the COH + NaHS + PD98059 group, respectively) and Western blotting (n = 6, P > 0.99 between any two groups). On the other hand, NaHS treatment effectively downregulated retinal protein expression of GFAP in COH rats (n = 6, both P < 0.001 versus the COH group in immunohistochemistry and Western blotting); delivery of PD98059 alone (n = 6, both P < 0.001 versus the COH group in immunohistochemistry and Western blotting) or joint administration of NaHS and PD98059 (n = 6, both P < 0.001 versus the COH group in immunohistochemistry and Western blotting) also initiated similar downregulation of GFAP in COH rats as NaHS alone, and no significant difference was observed between any two of the three pharmacologically treated groups (n = 6, each P > 0.99 between any two groups in both immunohistochemistry and Western blotting, respectively) (Figs. 6, 7). 

Figure 6

 

Glial activation was suppressed by both H₂S supplement and ERK 1/2 pathway inhibition in COH rats (immunohistochemistry demonstration). (A, B) Representative immunofluorescence staining images of Iba-1 and GFAP (arrowheads, red fluorescence) in retinal slices of different groups (magnification ×200, scale bar: 50 μm). (C, D) Quantitative analysis of fluorescence staining intensities of Iba-1 and GFAP in retina of different groups (n = 6, *P < 0.05 versus the COH group). Bars represent means ± SD.

Figure 7

 

Glial activation was suppressed by both H₂S supplement and ERK 1/2 pathway inhibition in COH rats (Western blotting demonstration). (A) Representative Western blotting bands of retinal Iba-1, GFAP, and internal controlling β-actin protein expression of different groups. (B, C) Semiquantitative analysis of Western blotting gray values (n = 6, *P < 0.05 versus the COH group). Bars represent means ± SD.

The result above revealed that H₂S supplement introduced an effect of suppression on glial cell activation similar to that of simple blockade of ERK 1/2 pathway in experimental glaucoma, and there was no synergistic effect or antagonistic effect between the two solutions. 

Impact of ERK 1/2 Inhibition on Activity of Nuclear Factor Kappa B (NF-κB) Pathway in COH Rats Treated With NaHS

NF-κB pathway regulates transcription of various proinflammation genes, including interleukin-1 (IL-1) and tumor necrosis factor-α (TNF-α), which are believed to engage in the pathogenesis of glaucoma. In this study, we focused on the phosphorylation level of IκBα, a key inhibitor protein for NF-κB activation, and p65 subunit from the NF-κB family to evaluate the activity of NF-κB pathway in COH rats. As shown in Figure 8, quantitative analysis of Western blotting bands showed that NaHS treatment remarkably downregulated the phosphorylation levels of IκBα and p65 in COH rat retinas (n = 6, P < 0.001 and = 0.03 versus the COH group, respectively), while similar downregulation of phosphorylated IκBα and p65 was observed in the groups treated with PD98059 alone (n = 6, P < 0.001 and = 0.02 versus the COH group, respectively) and with NaHS + PD98059 (n = 6, both P < 0.001 versus the COH group). The differences in the phosphorylation levels of the two target proteins between any two of the three treated groups were negligible (for IκBα, n = 6, P = 0.06, > 0.99, and = 0.36 for the COH + NaHS group versus the COH + PD98059 group, the COH + NaHS group versus the COH + NaHS + PD98059 group, and the COH + PD98059 group versus the COH + NaHS + PD98059 group, respectively; for p65, n = 6, P > 0.99, = 0.32, and = 0.60 for the COH + NaHS group versus the COH + PD98059 group, the COH + NaHS group versus the COH + NaHS + PD98059 group, and the COH + PD98059 group versus the COH + NaHS + PD98059 group, respectively). The results indicated that the suppressive effect brought about by H₂S supplement on NF-κB pathway activity in experimental glaucoma resembled the effect caused by inhibition of ERK 1/2 pathway, and it seemed that there was no interactive effect in the situation of combined intervention. 

Figure 8

 

NF-κB pathway activity was suppressed by both H₂S supplement and ERK 1/2 pathway inhibition in COH rats. (A) Representative Western blotting bands of retinal IκBα, phospho-IκBα, phospho-p65, and internal controlling β-actin protein expression of different groups. (B, C) Semiquantitative analysis of Western blotting gray values (n = 6, *P < 0.05 versus the COH group). Bars represent means ± SD.

Impact of ERK 1/2 Inhibition on Activity of Nicotinamide Adenine Dinucleotide Phosphate (NADPH) Oxidase in COH Rats Treated With NaHS

It is acknowledged that oxidative stress, mostly due to excessive reactive oxygen species, significantly contributes to the neuronal lesions in the glaucomatous condition. Since NADPH oxidase is a predominant source of reactive oxygen species, we chose one of its subunits, gp91^phox, as a target for the present study. Western blotting showed that all of NaHS, PD98059, and NaHS + PD98059 treatments obviously decreased the retinal protein expression of gp91^phox (n = 6, each P < 0.001 versus the COH group) (Fig. 9). However, the decreasing degree of gp91^phox expression varied with disparate treatment: The joint administration of NaHS and PD98059 resulted in the greatest decrease (n = 6, P < 0.001 versus the COH + NaHS group), and treatment with PD98059 alone led to a minimal decrease (n = 6, P < 0.001 versus the COH + NaHS group). The synergistic effect of joint administration of NaHS and PD98059 indicated that inhibition of ERK 1/2 pathway might only partly contribute to the mechanism through which H₂S supplement suppressed the activity of NADPH oxidase in experimental glaucoma. 

Figure 9

 

NADPH oxidase activity was suppressed by both H₂S supplement and ERK 1/2 pathway inhibition in COH rats. (A) Representative Western blotting bands of retinal gp91^phox and internal controlling β-actin protein expression of different groups. (B) Semiquantitative analysis of Western blotting gray values (n = 6, *P < 0.05 versus the COH group, #P < 0.05 versus the COH + NaHS group, and P < 0.05 versus the COH + PD98059 group). Bars represent means ± SD.

Impact of ERK 1/2 Inhibition on Cellular Autophagy in COH Rats Treated With NaHS

Cellular autophagy has been proven to be involved crucially in the progression of glaucomatous neuropathy and to have a deeply interactive relationship with neuronal apoptosis. Beclin-1 and LC3 serve as two representative proteins that are usually tested for the evaluation of autophagic activity. As shown in Figure 10 on Western blotting analysis, the retinal protein expression level of Beclin-1 and transformation rate of LC3 II from LC3 I were dramatically decreased by NaHS treatment (n = 6, P = 0.003 and < 0.001 versus the COH group, respectively) as well as by combined treatment with NaHS and PD98059 (n = 6, P = 0.01 and < 0.001 versus the COH group, respectively) in COH rats, while no significant difference was observed between these two treated groups (n = 6, both P > 0.99). However, PD98059 alone failed to lead to any noticeable variation in Beclin-1 expression or LC3 II/LC3 I transformation rate (n = 6, P > 0.99 and = 0.42 versus the COH group, respectively). The result revealed that simple inhibition of ERK 1/2 pathway might not affect autophagic activity in the retina in experimental glaucoma, and that the suppressive effect of H₂S supplement on cellular autophagy was not to be affected by ERK 1/2 pathway blockade. 

Figure 10

 

Inhibition of ERK 1/2 pathway did not directly affect cellular autophagy as H₂S supplement did in COH rats. (A) Representative Western blotting bands of retinal Beclin-1, LC3, and internal controlling β-actin protein expression of different groups. (B, C) Semiquantitative analysis of Western blotting gray values (n = 6, *P < 0.05 versus the COH group, #P < 0.05 versus the COH + NaHS group, and P < 0.05 versus the COH + PD98059 group). Bars represent means ± SD.

Impact of ERK 1/2 Inhibition on TNF-α Level in COH Rats Treated With NaHS

TNF-α is known to play a key role within inflammation-related mechanisms and to act as one of the major initiators that regulate extrinsic apoptotic pathway in glaucoma. We applied the ELISA method to quantify the TNF-α level in the retina. As presented in Figure 11, the TNF-α level in the retina of COH rats was significantly reduced by NaHS treatment (n = 6, P < 0.001 versus the COH group), while similar reductions of retinal TNF-α level were observed in the PD98059-treated group and the NaHS + PD98059-treated group (n = 6, both P < 0.001 versus the COH group). The differences of retinal TNF-α levels between any two of the three treated groups were negligible (n = 6, P > 0.99, = 0.56, and = 0.07 for the COH + NaHS group versus the COH + PD98059 group; the COH + NaHS group versus the COH + NaHS + PD98059 group; and the COH + PD98059 group versus the COH + NaHS + PD98059 group, respectively). The result reflected that the suppressive effect brought about by H₂S supplement on TNF-α production in experimental glaucoma resembled the effect caused by blocking ERK 1/2 pathway, and it was likely that there was no extra interactive effect introduced when ERK 1/2 pathway was blocked in experimental glaucoma being treated with NaHS. 

Figure 11

 

TNF-α production was suppressed by both H₂S supplement and ERK 1/2 pathway inhibition in COH rats. Quantitative analysis of retinal TNF-α level of different groups (n = 6, *P < 0.05 versus the COH group). Bars represent means ± SD.

Discussion

In our previous work, we concluded that H₂S supplement with exogenous donor significantly attenuated the apoptosis of RGC, thus effectively improving RGC survival in experimental glaucoma, and the neuroprotection by H₂S probably correlated with the suppression of glial activation, inhibition of proinflammatory pathways, attenuation of oxidative stress, and downregulation of autophagy in retina.²⁸ However, the mechanisms underlying the phenomena still remained unclear. In this study, we additionally introduced an inhibitor of ERK 1/2 pathway, one of the most fundamental signaling pathways within mammalian cellular pathophysiological regulation, in combination with H₂S supplement to treat COH rats. We expected to uncover a modicum of signaling mechanism under the neuroprotection achieved by H₂S in experimental glaucoma. 

The role of ERK 1/2 pathway activation in the mechanism of neuronal apoptosis has been reported to be dual, meaning it acts to promote cellular apoptosis in certain circumstances and to restrain cellular apoptosis in some other circumstances.³³ It was reported that the activity of ERK 1/2 pathway was obviously downregulated in the pathological progressions of experimental glaucoma and optic nerve crush, and further enhancement of ERK 1/2 pathway effectively improved the survival of RGC.³⁴ However, other researchers reported that the phosphorylation of ERK 1/2 was remarkably increased under conditions of retinal ischemia–reperfusion injury and the COH model, and injured RGC were significantly restored after pharmacological intervention was introduced to block the ERK 1/2 pathway.^35,36 Another study concluded that Bcl-2 family expression and abnormal glial activation correlated closely with the apoptosis of RGC in experimental glaucoma, but not the ERK 1/2 signaling pathway.¹⁵ In our present study, we found that the inhibitory effect of H₂S supplement on retinal ERK 1/2 signaling activity was similar to that of PD98059 application in experimental glaucoma. However, simple blockade of the ERK 1/2 pathway did not alter the regulation of Bcl-2 and Bax expressions, nor did it affect the cleavage of caspase 3, a primary apoptotic effector. Thus it was probable that the ERK 1/2 pathway inhibition accompanied by H₂S supplement was insufficient to directly hold back the apoptotic progression of RGC in experimental glaucoma. 

Hu et al.²⁹ found that pharmacological intervention with NaHS remarkably downregulated the activity of ERK 1/2 signaling pathway in Parkinson's disease models in vivo and in vitro, resembling the blocking effect of PD98059 treatment alone, while gp91^phox expression was notably downregulated with both interventions; the authors concluded that the antioxidative stress effect of H₂S in neurodegenerative disease might be attributed to its impact of inhibition on ERK 1/2 pathway. The results in our present study coincided with those of Hu et al. showing that the oxidative stress–related indicator was significantly downregulated by both NaHS treatment and PD98059 in experimental glaucoma. Additionally, we observed a more powerful antioxidative stress effect generated by NaHS than by PD98059 alone, and there was a positive synergistic effect between the two drugs. Given that H₂S acts as a potent antioxidant by itself and additionally enhances other antioxidants by promoting uptake of glutamate and synthesis of reduced glutathione, our result reflected that the origin of the antioxidative stress effect of H₂S in experimental glaucoma might not be limited to but be more than inhibition of ERK 1/2 signaling activity.⁴ Furthermore, our study demonstrated that H₂S supplement and ERK 1/2 pathway blockade generated similar effects in terms of attenuating glial activation, NF-κB pathway activity, and TNF-α production. It is probable that the activation of these three pathological processes partly depends on the ERK 1/2 signaling in experimental glaucoma, and that the inhibitory effect of H₂S on ERK 1/2 pathway contributes to the suppression of these activities. Since it has been reported by other researchers that suppression of glial activation, NF-κB pathway activation, NADPH oxidase activity, and TNF-α production correlated positively with RGC survival improvement in glaucomatous neuropathy, we supposed that the inhibitory effect on ERK 1/2 pathway might partly contribute to the neuroprotection by H₂S in experimental glaucoma.^37–40 

It was reported in the field of central nervous system disease that the activity of cellular autophagy, which contributed to neuronal lesions, was enhanced along with activation of ERK 1/2 signaling pathway, and the blockade of ERK 1/2 pathway effectively downregulated the expression of autophagic indicator proteins such as Beclin-1 and LC3 II,⁴¹ Zhang et al.⁴² also found in the light-induced retinopathy model that the autophagy of photoreceptors correlated closely with the activation of ERK 1/2 signaling, and further inhibition of ERK 1/2 pathway with PD98059 significantly downregulated the autophagic activity, leading to remission of photoreceptor lesions. However, our present study showed that simple blockade of ERK 1/2 pathway failed to suppress the cellular autophagic activity in experimental glaucoma, which is inconsistent with the conclusions by others mentioned above. We considered that the inconsistency might be attributable to the difference in targeted cell types and the complexity of pathological mechanisms in glaucoma. 

In the present study, we aimed to investigate the role of ERK 1/2 pathway underlying the impact of H₂S in glaucoma treatment, trying to better understand the mechanism by which H₂S exerts neuroprotection in glaucoma. However, there were certain limitations and deficiencies. For example, the sample size of rats subjected to various experimental procedures was relatively small. Also, since the research work spanned a relatively long period, some grouped animals were inevitably operated on at different dates and handled by different staff members, and this might introduce additional systematic bias and result in potential impairment of reproducibility. To some extent, our research techniques lacked enough specificity and precise quantification, especially in evaluations of activities of various pathways; as ERK 1/2 pathway was found to be insufficient to dominate the neuroprotective effect of H₂S in glaucoma, there exists a large unknown area with respect to a panoramic view of underlying mechanisms, which calls for further exploration and more extensive research work in the near future. 

In conclusion, the present study indicates that inhibition of ERK 1/2 pathway might partly contribute to the neuroprotection by H₂S in experimental glaucoma, through suppression of glial activation, NF-κB pathway activity, oxidative stress, and TNF-α production. However, our study also revealed that the involvement of ERK 1/2 pathway was probably insufficient to initiate the neuroprotection by H₂S, which meant that comprehension of the underlying mechanisms calls for further investigations. Being limited to resources provided, this study is only a minor step into the field of H₂S therapy in glaucoma. Further detailed and in-depth research is to be explored. 

Acknowledgments

Supported by the National Natural Science Foundation of China (No. 81371014, No. 81470639, No. 81670861, No. 81870652); the Shanghai Science and Technology Committee Project Foundation (No. 16411961900, No. 17140903000); and the Shanghai Municipal Commission of Health and Family Planning Project Foundation (No. 201640190, No. 2018JP008). The authors alone are responsible for the content and writing of the paper. 

Disclosure: S. Huang, None; P. Huang, None; H. Yu, None; Z. Lin, None; X. Liu, None; X. Shen, None; L. Guo, None; Y. Zhong, None