Effect of an Inducer of BiP, a Molecular Chaperone, on Endoplasmic Reticulum (ER) Stress-Induced Retinal Cell Death

Abstract

purpose. The effect of a preferential inducer of 78 kDa glucose-regulated protein (GRP78)/immunoglobulin heavy-chain binding protein (BiP; BiP inducer X, BIX) against tunicamycin-induced cell death in RGC-5 (a rat ganglion cell line), and also against tunicamycin- or N-methyl-d-aspartate (NMDA)-induced retinal damage in mice was evaluated.

methods. In vitro, BiP mRNA was measured after BIX treatment using semi-quantitative RT-PCR or real-time PCR. The effect of BIX on tunicamycin (at 2 μg/mL)-induced damage was evaluated by measuring the cell-death rate and CHOP protein expression. In vivo, BiP protein induction was examined by immunostaining. The retinal cell damage induced by tunicamycin (1 μg) or NMDA (40 nmol) was assessed by examining ganglion cell layer (GCL) cell loss, terminal deoxyribonucleotidyl transferase (TdT)-mediated dUTP nick-end labeling (TUNEL) staining, and CHOP protein expression.

results. In vitro, BIX preferentially induced BiP mRNA expression both time- and concentration-dependently in RGC-5 cells. BIX (1 and 5 μM) significantly reduced tunicamycin-induced cell death, and BIX (5 μM) significantly reduced tunicamycin-induced CHOP protein expression. In vivo, intravitreal injection of BIX (5 nmol) significantly induced BiP protein expression in the mouse retina. Co-administration of BIX (5 nmol) significantly reduced both the retinal cell death and the CHOP protein expression in GCL induced by intravitreal injection of tunicamycin or NMDA.

conclusions. These findings suggest that this BiP inducer may have the potential to be a therapeutic agent for endoplasmic reticulum (ER) stress–induced retinal diseases.

The endoplasmic reticulum (ER) is the cellular organelle in which proteins (destined for secretion or for diverse subcellular localizations) are not only synthesized, but acquire their correct conformation. Perturbations of the environment normally required for protein folding in the ER, or the production of large amounts of misfolded proteins exceeding the functional capacity of the organelle, trigger a pattern of physiological response in the cell, collectively known as the unfolded protein response (UPR). ^{1 2 3} The UPR serves to cope with ER stress by transcriptionally regulating ER chaperones and other ER-resident proteins, attenuating the overall translation rate, and increasing the degradation of misfolded ER proteins. ER stress is caused by the accumulation of unfolded proteins in the ER lumen, and it is associated with various neurodegenerative diseases such as Alzheimer’s, Huntington’s, and Parkinson’s diseases, and with type-1 diabetes. ^{4 5 6} Recent reports have shown that ER stress is also involved in a variety of experimental retinal neurodegenerative models, such as those of diabetic retinopathy, ⁷ retinitis pigmentosa, ^{8 9} and glaucoma. ^{10 11}  

Recently, we reported that BiP expression is upregulated in the retina after intravitreal injection of either tunicamycin or NMDA (a glutamate-receptor agonist). ^{12 13} Tunicamycin, a glucosamine-containing nucleoside antibiotic, produced by genus Streptomyces, is an inhibitor of N-linked glycosylation and the formation of N-glycosidic protein-carbohydrate linkages. ¹⁴ Tunicamycin, which reduces the N-glycosylation of proteins, causes an accumulation of unfolded proteins in the ER and thus induces ER stress. Awai et al. ¹⁵ previously had demonstrated that NMDA induces CHOP protein (a member of the CCAAT/enhancer-binding protein family induced by ER stress) in GCL and the inner plexiform layer (IPL), and that CHOP^-/- mice are more resistant to NMDA-induced retinal cell death than wild-type mice. These findings indicate that ER stress may be involved in these models of retinal injury. 

BiP, a highly conserved member of the 70 kDa heat shock protein family, is one of the chaperones localized to the ER membrane, ^{16 17} and it is a major ER-luminal Ca²⁺-storage protein. ^{18 19} BiP works to restore folding in misfolded or incompletely assembled proteins, ^{20 21 22} the interaction between BiP and misfolded proteins being dependent on its hydrophobic motifs. ^{23 24 25} Proteins stably bound to BiP are subsequently translocated from the ER into the cytosol, where they are degraded by proteasomes. ^{26 27} Previous reports have show that induction of BiP prevents the neuronal death induced by ER stress. ^{28 29 30 31} Hence, a selective inducer of BiP might attenuate ER stress and be a new, useful therapeutic agent for the treatment of ER stress-associated diseases. 

This seemed an interesting idea, and we recently identified BiP inducer X (BIX) while screening for low molecular mass compounds that might induce BiP using high-throughput screening (HTS) with a BiP reporter assay system (Dual-Luciferase Reporter Assay; Promega Corporation, Madison, WI). ³² We found that BIX preferentially induced BiP mRNA and protein in SK-N-SH cells and reduced tunicamycin-induced cell death. Intracerebroventricular pretreatment with BIX reduced the infarction size after focal cerebral ischemia in mice. In view of the retinal research described above, we wondered whether BIX might reduce the retinal ganglion cell loss and CHOP expression induced by tunicamycin or NMDA treatment. 

In the present study, we examined primarily whether induction of BiP might inhibit the retinal cell death induced by tunicamycin in RGC-5 cells in vitro, and/or that induced by tunicamycin or NMDA in mice in vivo. 

Materials and Methods

All experiments were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and they were approved and monitored by the Institutional Animal Care and Use Committee of Gifu Pharmaceutical University, Gifu, Japan. 

Materials

Dulbecco modified Eagle medium (DMEM) and NMDA were purchased from Sigma-Aldrich (St. Louis, MO). The other drugs used and their sources were as follows: BIX, 1-(3,4-dihydroxyphenyl)-2-thiocyanato-ethanone, was synthesized in the Department of Medicinal Chemistry, Gifu Pharmaceutical University, while tunicamycin was purchased from Wako (Osaka, Japan). Isoflurane was acquired from Nissan Kagaku (Tokyo, Japan) and fetal bovine serum (FBS) was from Valeant (Costa Mesa, CA). 

RGC-5 Culture

RGC-5 ³² were gifted by Neeraj Agarwal (Department of Pathology and Anatomy, UNT Health Science Center, Fort Worth, TX). Cultures of RGC-5 were maintained in DMEM supplemented with 10% FBS, 100 U/mL penicillin (Meiji Seika Kaisha Ltd., Tokyo, Japan), and 100 μg/mL streptomycin (Meiji Seika Kaisha. Ltd.) in a humidified atmosphere of 95% air and 5% CO₂ at 37°C. The RGC-5 cells were passaged by trypsinization every 3 days, as in our previous reports. ^{12 13 33} We used RGC-5 without any differentiation. 

RNA Isolation and Semi-Quantitative RT-PCR Analysis

To examine the effect of BIX on BiP mRNA expression, RGC-5 cells were seeded in six-well plates at a density of 1.4 × 10⁵ cells per well. After the cells had been incubating for 24 h, they were exposed to 50 μM BIX in 1% FBS DMEM for 0.5, 1, 2, 4, 6, 8, or 12 h, or to 2, 10, 50, or 150 μM BIX in 1% FBS DMEM for 6 h. Total RNA was extracted (RNeasy Mini Kit; QIAGEN KK, Tokyo, Japan) according to the manufacture’s protocol. The total RNA was divided into microtubes, and frozen to −80°C. RNA concentrations were determined spectrophotometrically at 260 nm. First-strand cDNA was synthesized in a 20-μl reaction volume using a random primer (Takara, Shiga, Japan) and Moloney murine leukemia virus reverse transcriptase (Invitrogen, Carlsbad, CA). PCR was performed in a total volume of 30 μL containing 0.8 μM of each primer, 0.2 mM dNTPs, 3 U Taq DNA polymerase (Promega), 2.5 mM MgCl₂, and 1× PCR buffer. The amplification conditions for the semi-quantitative RT-PCR analysis were as follows: an initial denaturation step (95°C for 5 minutes), 20 cycles of 95°C for 1 minute, 55°C for 1 minute, and 72°C for 1 minute, and a final extension step (72°C for 7 minutes). The numbers of amplification cycles for the detection of BiP and β-actin were 18 and 15, respectively. The primers used for amplification were as follows: BiP: 5′-GTTTGCTGAGGAAGACAAAAAGCTC-3′ and 5′-CACTTCCATAGAGTTTGCTGATAATTG-3′; β-actin: 5′-TCCTCCCTGGAGAAGAGCTAC-3′ and 5′-TCCTGCTTGCTGATCCACAT-3′. 

PCR products were resolved by electrophoresis through 6% (w/v) polyacrylamide gels. The density of each band was quantified using an imaging program (Scion Image Program; Scion Corporation, Frederick, MD). 

Real-Time PCR

Real-time PCR (TaqMan; Applied Biosystems, Foster City, CA) was performed as described previously. ³⁴ Single-stranded cDNA was synthesized from total RNA using a high-capacity cDNA archive kit (Applied Biosystems). Quantitative real-time PCR was performed using a sequence detection system (ABI PRISM 7900HT; Applied Biosystems) with a PCR master mix (TaqMan Universal PCR Master Mix; Applied Biosystems), according to the manufacturer’s protocol. mRNA expression was measured by real-time PCR using a gene expression product (Assays-on-Demand Gene Expression Product; Applied Biosystems) and a BiP probe (Assay ID Details: Mm00517691). The thermal cycler conditions were as follows: 2 minutes at 50°C and then 10 minutes at 95°C, followed by two-step PCR for 50 cycles consisting of 95°C for 15 seconds followed by 60°C for 1 minute. For each PCR, we obtained the slope value, R ² value, and linear range of a standard curve of serial dilutions. All reactions were performed in duplicate. The results are expressed relative to the β-actin (Assay ID Details: Mm00661904) internal control. 

Cell Viability

To examine the effects of BIX on the cell death induced by tunicamycin (2 μg/mL) or staurosporine (an ER stress-independent apoptosis inducer, 30 nm) RGC-5 cells were seeded at a low density of 700 cells per well into 96-well plates. After pretreatment with BIX for 12 h, tunicamycin or staurosporine was added to the cultures for 48 h or 24 h, respectively. Cell death was assessed on the basis of combination staining with the fluorescent dyes Hoechst 33342 and propidium iodide (PI; Molecular Probes, Eugene, OR) or the change in fluorescence intensity after the cellular reduction of WST-8 to formazan. Hoechst 33342 (λex 350 nm, λem 461 nm) and PI (λex 535 nm, λem 617 nm) were added to the culture medium at final concentrations of 8 and 1.5 μM, respectively, for 30 minutes. Images were collected using a CCD camera (DP30VW; Olympus America, Center Valley, PA) via an epifluorescence microscope (IX70; Olympus, Tokyo, Japan) fitted with fluorescence filters for Hoechst 33342 (U-MWU; Olympus), and PI (U-MWIG; Olympus). In WST-8 assay, cell viability was assessed by culturing cells in a culture medium containing 10% WST-8 (Cell Counting Kit-8; Dojin Kagaku, Kumamoto, Japan) for 3 h at 37°C, with quantification being achieved by scanning with a microplate reader at 492 nm. ³⁵ This absorbance is expressed as a percentage of that in control cells (which were in 1% FBS DMEM) after subtraction of background absorbance. 

Animals

Mice used were male adult ddY mice (Japan SLC, Hamamatsu, Japan), male adult Thy-1-cyan fluorescent protein (CFP) transgenic mice (The Jackson Laboratory, Bar Harbor, Maine), ³⁶ and ER stress–activated indicator (ERAI)-transgenic mice carrying the F-XBP1ΔDBD-venus, a variant of green fluorescent protein (GFP) fusion gene, which allows effective identification of cells under ER stress in vivo, as previously described by Iwawaki et al. ³⁷ Briefly, when ER stress in ERAI transgenic mice was induced, splicing of mRNA encoding the XBP-1 fusion gene occurs and the spliced form of F-XBP1ΔDBD-venus fusion gene could be translated into fluorescent protein. Thus, it is visualized by the fluorescence intensity arising from the XBP-Δ-venus fusion protein during ER stress, and we measured it by fluorescence microscopy and immunoblotting in the present study. 

All mice were kept under controlled lighting conditions (12 h:12 h light/dark). The mouse genotype was determined by applying standard PCR methodology to tail DNA. 

NMDA- or Tunicamycin-Induced Retinal Damage

NMDA- or tunicamycin-induced retinal damage was produced as previously reported by Siliprandi et al. (1992). ³⁸ Briefly, mice were anesthetized with 3.0% isoflurane and maintained with 1.5% isoflurane in 70% N₂O and 30% O₂ via an animal general anesthesia machine (Soft Lander; Sin-ei Industry Co. Ltd., Saitama, Japan). The body temperature was maintained between 37.0 and 37.5°C with the aid of a heating pad. Retinal damage was induced by the injection (2 μL/eye) of NMDA (Sigma-Aldrich) at 20 mM or tunicamycin at 1 μg/mL dissolved in 0.01 M PBS with 5% dimethyl sulfoxide (DMSO). Each solution was injected into the vitreous body of the left eye under the above anesthesia. One drop of 0.01% levofloxacin ophthalmic solution (Santen Pharmaceuticals Co. Ltd., Osaka, Japan) was applied topically to the treated eye immediately after the intravitreal injection. Seven days after the injection, eyeballs were enucleated for histologic analysis. For comparative purposes, nontreated retinas from each mouse strain were also investigated. BIX (0.5 or 5 nmol) or vehicle (5% DMSO in PBS) was co-administered with the NMDA or tunicamycin in each mouse. 

Histologic Analysis

In mice under anesthesia produced by an intraperitoneal injection of sodium pentobarbital (80 mg/kg), each eye was enucleated and kept immersed for at least 24 h at 4°C in a fixative solution containing 4% paraformaldehyde. Six paraffin-embedded sections (thickness, 4 μm) cut through the optic disc of each eye were prepared in a standard manner and stained with hematoxylin and eosin. The damage induced by NMDA or tunicamycin was then evaluated, with three sections from each eye being used for the morphometric analysis, as described below. Light-microscope images were photographed, and the cells in the GCL at a distance between 375 and 625 μm from the optic disc were counted on the photographs in a masked fashion by a single observer (Y.I.). Data from three sections (selected randomly from the six sections) were averaged for each eye and used to evaluate the cell count in the GCL. 

Retinal Flatmounts and Analysis in Transgenic Mice

Transgenic mice were given an overdose of sodium pentobarbital, and retinas were dissected out and fixed for 30 minutes in 4% paraformaldehyde diluted in 0.1 M phosphate buffer (PB) at pH 7.4. Retinas were subsequently washed with PBS at room temperature, flatmounted on clean glass slides using fluorescent mounting medium (Dako Corp., Carpinteria, CA), and stored in the dark at 4°C for 1 week. The damage induced by tunicamycin was then evaluated, with four sections (dorsal, ventral, temporal, and nasal) from each eye being used for the morphometric analysis, as described below. At various times after intravitreal injections (24 h in ERAI mice and 7 days in Thy-1-CFP transgenic mice), fluorescent images were photographed (×200, 0.144 mm²) using an epifluorescence microscope (BX50; Olympus) fitted with a CCD camera (DP30VW; Olympus). In the case of Thy-1-CFP transgenic mice, Thy-1-CFP-positive cells at a distance of 1 mm from the optic disc were counted on the photographs in a masked fashion by a single observer (Y.I.). Data from the four parts of each eye were used to evaluate the RGC count. ³⁹  

Immunostaining

Eyes were enucleated as described under Histologic Analysis, fixed in 4% paraformaldehyde overnight at 4°C, immersed in 25% sucrose for 48 h at 4°C, and embedded in optimum cutting temperature (OCT) compound (Sakura Finetechnical Co. Ltd, Tokyo, Japan). Transverse 10-μm thick cryostat sections were cut and placed onto slides (MAS COAT; Matsunami Glass Ind., Ltd., Osaka, Japan). Immunohistochemical staining was performed according to the following protocol: Briefly, tissue sections were washed in 0.01 M PBS for 10 minutes, and then endogenous peroxidase was quenched by treating the sections with 3% hydrogen peroxide in absolute methanol for 10 minutes, followed by a pre-incubation with 10% normal goat serum. They were then incubated overnight at 4°C with the following primary antibodies: against CHOP (1:1000 dilution in PBS; Santa Cruz, CA), and against BiP/GRP78 (1:1000 dilution in PBS; BD Transduction Laboratories, Lexington, KY). Sections were washed and then incubated with biotinylated anti-rabbit IgG or anti-mouse IgG. They were subsequently incubated with the avidin-biotin-peroxidase complex for 30 minutes, and then developed using diaminobenzidine (DAB) peroxidase substrate. Images were obtained using a digital camera (COOLPIX 4500; Nikon, Tokyo, Japan). 

Quantitation of Density

In the DAB-labeled areas of anti-BiP/GRP78 (BD Transduction Laboratories) and anti-CHOP (Santa Cruz) in the GCL and IPL at a distance between 475 and 525 μm (50 × 50 μm) from the optic disc, retinal DAB-labeled cell density was evaluated by means of appropriately calibrated computerized image analysis, using median density in the range of 0 to 255 as an analysis tool (Image Processing and Analysis in Java, Image J; National Institute of Mental Health, Bethesda, MD) and averaged for two areas. ⁴⁰ The data lie within the dynamic range of this assays 

Briefly, light-microscope images of the above-mentioned areas were photographed, inverted in a gradation sequence using image editing software (Adobe Photoshop 5.5; Adobe Systems Inc., San Jose, CA), and then optical intensity was evaluated using Image J. The score for the negative-control (nontreated with first antibody), as the background value, was subtracted from the scores. 

TUNEL Staining

TUNEL staining was performed according to the manufacturer’s protocol (In Situ Cell Death Detection Kit; Roche Biochemicals, Mannheim, Germany) to detect the retinal cell death induced by NMDA. Mice were anesthetized with pentobarbital sodium at 80 mg/kg, IP, 24 h after intravitreal injection (either of NMDA 40 nmol/eye or of tunicamycin 1 μg/eye). The eyes were enucleated, fixed overnight in 4% paraformaldehyde, and immersed for 2 days in 25% sucrose with PBS. The eyes were then embedded in a supporting medium (OCT compound) for frozen-tissue specimens. Retinal sections at 10-μm thick were cut on a cryostat at −25°C, and stored at −80°C until staining. After twice washing with PBS, sections were incubated with terminal TdT enzyme at 37°C for 1 h, then washed 3 times in PBS for 1 minute at room temperature. Sections were subsequently incubated with an anti-fluorescein antibody-peroxidase conjugate at room temperature in a humidified chamber for 30 minutes, then developed using DAB tetrahydrochloride peroxidase substrate. Light-microscope images were photographed, and the labeled cells in the GCL at a distance between 375 and 625 μm from the optic disc were counted in two areas of the retina in a masked fashion by a single observer (Y.I.). The number of TUNEL-positive cells was averaged for these two areas, and plotted as the number of TUNEL-positive cells. 

Western Blot Analysis

RGC-5 cells were lysed using a cell-lysis buffer (RIPA buffer R0278; Sigma-Aldrich) with protease (P8340; Sigma-Aldrich) and phosphatase inhibitor cocktails (P2850 and P5726; Sigma-Aldrich), and 1 mM EDTA. In vivo, mice were euthanatized using sodium pentobarbital at 80 mg/kg, IP, and their eyeballs were quickly removed. The retinas were carefully separated from the eyeballs and quickly frozen in dry ice. For protein extraction, the tissue was homogenized in the cell-lysis buffer using a homogenizer (Physcotron; Microtec Co. Ltd., Chiba, Japan). The lysate was centrifuged at 12,000g for 20 minutes, and the supernatant was used for this study. Assays to determine the protein concentration were performed by comparison with a known concentration of bovine serum albumin using a BCA protein assay kit (Pierce Biotechnology, Rockford, IL). A mixture of equal parts of an aliquot of protein and sample buffer with 10% 2-mercaptoethanol was subjected to 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The separated protein was then transferred onto a polyvinylidene difluoride membrane (Immobilon-P; Millipore Corporation, Bedford, MA). For immunoblotting, the following primary antibodies were used: rabbit anti-CHOP polyclonal antibody (1:1000; Santa Cruz), mouse anti-β-actin monoclonal antibody (1:4000; Sigma-Aldrich), and rabbit anti-green fluorescent protein (GFP) polyclonal antibody (1:1000; Medical & Biological Laboratories Co. Ltd., Nagoya, Japan). The secondary antibody used was either goat anti-rabbit HRP-conjugated (1:2000) or goat anti-mouse HRP-conjugated (1:2000). The immunoreactive bands were visualized using a chemiluminescent substrate (SuperSignal West Femto Maximum Sensitivity Substrate; Pierce Biotechnology). The band intensity was measured using an imaging analyzer (Lumino Imaging Analyzer; Toyobo, Osaka, Japan) and a gel analysis electrophoresis analysis software (Gel Pro Analyzer; Media Cybernetics, Atlanta, GA). 

Statistical Analysis

Data are presented as the means ± SE. Statistical comparisons were made by way of Dunnett’s test or Student’s t-test using statistical analysis software (STAT VIEW version 5.0; SAS Institute, Cary, NC). P < 0.05 was considered to indicate statistical significance. 

Results

BiP mRNA in RGC-5 Preferentially Induced by BIX

To clarify whether BIX (Fig. 1)induces BiP in RGC-5, we used semi-quantitative RT-PCR and real-time PCR, using a specific primer and a TaqMan probe recognizing BiP mRNA, respectively. Real-time PCR revealed that the level of BiP mRNA was significantly elevated at 0.5 to 12 h (peak at approximately 6 h) after treatment with 50 μM BIX (Fig. 2A) . At 6 h after treatment with BIX (2 to 150 μM), BiP mRNA was increased concentration-dependently (Fig. 2B) . Next, we used real-time PCR to investigate whether BIX might affect the expressions of any other genes related to the ER stress response, such as GRP94, calreticulin, protein kinase inhibitor of 58 kDa (p58^IPK), or asparagine synthetase (ASNS; Fig. 2C ). Real-time PCR revealed significant inductions of ASNS and calreticulin mRNAs at 6 h after treatment with BIX at 2 and 10 μM, respectively. At 50 μM, BIX induced the mRNAs for GRP94 at 12 h, calreticulin at 6 and 12 h, p58^IPK at 6 h, and ASNS at 12 h. In contrast, GRP94 mRNA was significantly reduced at 4 h after treatment with 50 μM BIX. 

Protective Effect of BIX against ER Stress–Induced Cell Death in RGC-5 Cells

To investigate whether BIX can prevent the cell death induced by ER stress, RGC-5 cells were pretreated for 12 h with or without BIX, then treated with 2 μg/mL tunicamycin, and finally incubated for a further 48 h. Fluorescence micrographs of Hoechst 33342 and PI staining revealed 38.4 ± 4.5% cell death (n = 8) at 48 h after tunicamycin treatment, (control: 0.9 ± 0.2%, n = 8), and pretreatment with BIX at 1 and 5 μM significantly reduced this cell death (Figs. 3A 3B) . Next, we evaluated the expression of CHOP protein after tunicamycin treatment. There was no CHOP protein expression in either nontreated or BIX-treated cells (Figs. 3C 3D) . On the other hand, tunicamycin markedly induced CHOP protein, while pretreatment with BIX at 5 μM reduced this expression to almost half the value seen after tunicamycin treatment alone (Figs. 3C 3D) . 

Effects of BIX on Cell Damage Induced by Staurosporine in RGC-5 Culture

To investigate whether BIX protects non-ER stress-induced cell death, we examined staurosporine-induced cell death. Staurosporine at 30 nM for 24 h reduced cell viability to approximately 60% of control (Fig. 4) . There was no statistical difference between BIX (1 and 5 μM)-treated and vehicle-treated group. 

BiP Protein in the Mouse Retina Induced by Intravitreous Injection of BIX

Compared with that in the nontreated retina, BiP protein expression in GCL and IPL was significantly increased at 6 and 12 h after intravitreal injection of BIX (5 nmol; Figs. 5A 5B ). Optical density analysis confirmed that administration of BIX induced BiP protein in vivo (Fig. 5D) . 

Protective Effect of BIX against Tunicamycin-Induced Retinal Damage in Mice

Tunicamycin decreased the cell number in GCL at 7 days after its intravitreal injection (vs. nontreated retinas; Figs. 6A 6B ). There was significantly less cell loss in GCL when BIX (5 nmol) was co-administered with the tunicamycin (Figs. 6B 6C 6D) . In addition, intravitreal injection of tunicamycin increased the number of TUNEL-positive cells in GCL at 24 h (vs. nontreated retinas; Figs. 6E 6F ). BIX (5 nmol), when co-administered with the tunicamycin, significantly reduced the number of TUNEL-positive cells (vs. tunicamycin alone; Figs. 6F 6G 6H ). 

Protective Effect of BIX against Tunicamycin-Induced Retinal Damage in Thy-1-CFP Transgenic Mice

In this experiment on Thy-1-CFP transgenic mice, we confirmed the effect of BIX in a larger retinal area than that evaluated in Figure 6D . We counted the number of Thy-1-CFP–positive cells (in flatmounts) in the four white areas shown 1 mm from the center of the optic disc in Figure 7E , and then totaled these values. In the Thy-1-CFP–transgenic mouse retina, axonal fibers were evenly and densely distributed. There were congested CFP-positive cells in the vehicle-treated retina (Fig. 7A) , and no change was observed in BIX-treated retinas without tunicamycin treatment (Fig. 7B) . Intravitreal injection of tunicamycin decreased the Thy-1-CFP–positive cell count at 7 days (vs. vehicle-treated retina; Figs. 7A 7C ). BIX at 5 nmol, when co-administered with the tunicamycin, significantly inhibited this cell loss (Figs. 7C 7D 7F) . 

Effect of BIX on Tunicamycin-Induced CHOP Expression in Mice

Representative photograph of a nontreated retina is shown in Figure 8A . No change was observed in the BIX-treated retina (Fig. 8B) . Optical density analysis of CHOP protein immunoreactivity in GCL and IPL showed that intravitreal injection of tunicamycin (1 μg) significantly increased the level of CHOP protein at 72 h after the injection (Fig. 8C) . BIX (5 nmol), when co-administered with the tunicamycin, significantly inhibited this effect (Figs. 8D 8E) . 

Effect of BIX on Tunicamycin-Induced XBP-1 Expression in ERAI Mice

In ERAI mice, the fluorescence intensity arising from the XBP-1–venus fusion protein (indicating ER stress activation) can be easily visualized, allowing evaluation of the effect of ER stress on the retina. In the representative photographs of flatmount retinas from ERAI mice shown in Figure 9 , no difference was observed between vehicle-treated and BIX-treated retinas (Fig. 9B) . Intravitreal injection of tunicamycin (1 μg) induced XBP-1–venus expression (vs. the vehicle-treated retina; Fig. 9C ). Immunoblot analysis of XBP-1–venus protein expression in the retina (using an anti-GFP antibody) showed that intravitreal injection of tunicamycin (1 μg) significantly raised the level of XBP-1–venus protein, and that BIX (5 nmol), when co-administered with the tunicamycin (Fig. 9D) , significantly inhibited this effect (Figs. 9E 9F) . 

Protective Effect of BIX against NMDA-Induced Retinal Damage in Mice

A representative photograph of a nontreated retina is shown in Figure 10A . Intravitreal injection of NMDA (a) decreased the cell number in GCL at 7 days (Figs. 10B 10E)and (b) increased the number of TUNEL-positive cells in GCL at 24 h (vs. nontreated normal retina; Fig. 10F ). BIX (5 nmol), when co-administered with the NMDA, significantly reduced (vs. NMDA alone) both the cell loss in GCL (Figs. 10D 10E)and the number of TUNEL-positive cells (Fig. 10F) . On the other hand, there was no statistical difference between BIX (0.5 nmol)- and vehicle-treated group in NMDA-induced cell death in GCL (Figs. 10C 10E) . 

Effect of BIX on NMDA-Induced CHOP Expression in Mice

A representative photograph of a nontreated retina is shown in Figure 11Aand no change was detected between nontreated retina and the BIX-treated retina without NMDA treatment (Figs. 11A 11B) . Optical density analysis of CHOP protein immunoreactivity in GCL and IPL showed that intravitreal injection of NMDA (40 nmol) significantly increased the level of CHOP protein at 72 h after the injection (Fig. 11C) . When co-administered with the NMDA, BIX (5 nmol) significantly inhibited this effect (Figs. 11D 11E) . 

Discussion

In the present study, we confirmed that BIX preferentially induces BiP mRNA in RGC-5. Although it also induced GRP94, calreticulin, p58^IPK, and ASNS, these inductions were lower than that of BiP. This is consistent with our previous study that BIX preferentially induced BiP with slight inductions of GRP94, calreticulin, and CHOP mediated by the activating transcription factor 6 (ATF6) pathway accompanied by activation of ERSEs, and that BIX does not affect the pathway downstream of IRE1 or the translational control branch downstream of PERK in SK-N-SH cells. ³² Therefore, BIX is not just an ER stressor such as tunicamycin or thapsigargin, and we consider that the induction of BiP by BIX is mediated by the ATF6 pathway in RGC-5 similar to that in SK-N-SH cells. Next, we evaluated the effects of BIX, as a preferential inducer of BiP, on ER stress–induced in vitro cell death in RGC-5 (a rat ganglion cell-line) and in vivo retinal damage in mice. We found that BIX reduced tunicamycin-induced cell death in RGC-5 and also reduced both tunicamycin-induced and NMDA-induced retinal damage in mice. Our previous study revealed that BIX (a) reduced tunicamycin-induced cell death in SK-N-SH cells, (b) contributed to the induction of BiP expression via the ATF-6 pathway (but not via the PERK or IRE1 pathways), and (c) on intracerebroventricular injection, prevented the neuronal damage induced by focal ischemia in mice. ³² Furthermore, immunostaining revealed that intravitreal injection of BIX significantly induced BiP protein in mouse retina. Particularly, it expressed in GCL and IPL (versus both the normal and the sham retina). On the other hand, there was little protective effect of BIX against RGC-5 damages after staurosporine treatment. Staurosporine is well known as a nonspecific inhibitor of protein kinases and initiates caspase-dependent apoptosis in many cell types. ^{41 42} Our previous studies revealed that staurosporine induced cell death without any changes in the expression of BiP or CHOP protein. ^{12 43} Furthermore, preliminary study showed that treatment with BIX (1 and 5 μM) did not inhibit RGC-5 cell death 48 h after serum deprivation, which does not induce any UPR-responses such as BiP or CHOP (unpublished data). These results strongly support that BIX selectively protects cell damage induced by ER stress. 

Recently, we reported that in mice, increased expressions of XBP-1 splicing, BiP, and CHOP could be detected after the induction of retinal damage by tunicamycin, NMDA, or an elevation of intraocular pressure. ¹³ That report was the first to demonstrate an involvement of ER stress and BiP in retinal cell death in mice. Hence, in the present study we asked whether BIX can prevent such retinal damage. By histologic analysis and TUNEL staining, we estimated that BIX reduced tunicamycin-induced retinal damage in GCL. However, the cell counts in partial cross sections provide a comparatively small sample on which quantitative morphometry can be used to judge such an effect. Therefore, we used Thy-1-CFP transgenic mice ³⁶ to examine the effect of BIX in a large retinal area. This transgene contains a CFP gene under the direction of regulatory elements derived from the mouse Thy-1 gene, and the transgenic mice express CFP protein in RGC and in the inner part of the IPL of the retina. ³⁶ Our results show that BIX exerted a protective effect against tunicamycin-induced retinal damage in the Thy-1-CFP transgenic mouse. However, it is possible that microglial cells become co-labeled with CFP by phagocytosis of the dying RGCs. In this study, we evaluated CFP-positive cells in 7 days after tunicamycin injection. In our previous and preliminary studies, activated microglia cells in GCL were increased at 3 days after NMDA injection ¹³ and their increases were almost ceased within the 7 days (unpublished data). Furthermore, microglial cells can be distinguished with neuronal cells by their morphologic features. ⁴⁴ In fact, microglial cells were scarcely observed at seven days after tunicamycin injection, similar to that at seven days after NMDA injection. When we investigated the effect of BIX on NMDA-induced retinal damage in ddY mice, we found that it significantly attenuated such damage. NMDA is well known to induce RGC death and optic-nerve loss (effects mediated by excitatory glutamate receptor), and such neuronal death is believed to play a role in many neurologic and neurodegenerative diseases. ^{45 46} Recently, Uehara et al. ⁴⁷ noted that mild exposure to NMDA induced apoptotic cell death in primary cortical culture, and they demonstrated this effect to be caused by an accumulation of polyubiquitinated proteins and increases in XBP-1 mRNA splicing and CHOP mRNA (reflecting activation of the UPR signaling pathway). They also found that protein-disulphide isomerase, which assists in the maturation and transport of unfolded secretory proteins, prevented the neurotoxicity associated with ER stress. These findings suggested that activation of ER stress may participate in the retinal cell death occurring after NMDA-receptor activation and/or an ischemic insult. ⁴⁷  

In our investigation of the mechanisms underlying the above-mentioned effects, we focused on CHOP. Since CHOP is a member of the CCAAT/enhancer-binding protein family that is induced by ER stress and participates in ER-mediated apoptosis, CHOP may be a key molecule in retinal cell death. ⁴⁸ We found that treatment with tunicamycin induced apoptotic cell death in RGC-5 and also induced a production of ER stress–related proteins (BiP, the phosphorylated form of eIF2α, and CHOP protein). BIX reduced both the cell death and the CHOP protein expression induced by tunicamycin in RGC-5 in vitro. BIX also attenuated the CHOP protein expression induced by either tunicamycin or NMDA in the mouse retina in vivo. As mentioned above, BIX may affect CHOP protein expression through ATF6 pathway, but no change was observed in BIX-treated RGC-5. In our previous data in SK-N-SH cells, BIX slightly increased CHOP mRNA only at 2 h after the treatment. Expression of CHOP is mainly regulated by three transcription factors—ATF4, cleaved ATF6, and x-box binding protein-1 (XBP-1)—which are downstream effectors during ER stress in similar to other ER chaperones. These differences between BiP and CHOP expression by BIX may be due to the difference of their promoters. CHOP promoter contains at least two ERSE motifs (CHOP ERSE-1 and CHOP ERSE-2) located in opposite directions with a 9 bp overlap, and one of ERSEs is inactive. ⁴⁹ On the other hand, BiP promoter has three functional ERSE motifs of the rat GRP78 promoter (ERSE-163, ERSE-131, and ERSE-98). ⁵⁰ These variations in each promoter may contribute to the differences among the expressions of ER chaperons induced by BIX and the lack of CHOP expression. 

Subsequently, we monitored XBP-1 activation in the mouse retina in vivo, using ERAI transgenic mice. ³⁷ Effective identification of cells under ER stress conditions is possible in the retina in these mice, as described in our previous report. ¹³ Here, ERAI mice carrying the F-XBP1ΔDBD–venus expression gene were used to monitor ER stress. The fluorescence intensity arising from the X-box binding protein (XBP1)-venus fusion protein, indicating ER stress activation, was increased in cells within GCL and IPL at 24 h after injection of tunicamycin into the vitreous. BIX significantly reduced this expression, indicating that BIX may attenuate the retinal damage induced by ER stress–associated factors. In our previous study, ³² we found that BIX induced BiP protein expression via the ATF-6 pathway (not via other ER stress–associated factors such as the PERK and IRE1 pathways) in SK-N-SH cells. Possibly, the protective mechanism underlying the effect of BIX on the mouse retina may be the same as that revealed by our previous study, but further experiments will be needed to clarify this issue. 

In conclusion, we have demonstrated that BIX, a preferential inducer of BiP, inhibits both the neuronal cell death induced by ER stress in vitro in RGC-5 cells and in vivo in the mouse retina. Hence, an increase in BiP might be one of the targets of mechanisms bestowing neuroprotection in retinal diseases. 

 Supported in part by a Grant-in-Aid (No.18209053) for scientific research from the Ministry of Education, Science, Sports, Culture of the Japanese Government; by a research grant (No.18210101) from the Ministry of Health, Labor, and Welfare of the Japanese Government, and by Grant-in-Aid for Japan Society for the Promotion of Science Fellows.

 Submitted for publication April 4, 2008; revised July 25, 2008; accepted October 27, 2008.

 Disclosure: Y. Inokuchi, None; Y. Nakajima, None; M. Shimazawa, None; T. Kurita, None; M. Kubo, None; A. Saito, None; H. Sajiki, None; T. Kudo, None; M. Aihara, None; K. Imaizumi, None; M. Araie, None; H. Hara, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

The structure of BIX (1-(3,4-dihydroxyphenyl)-2-thiocyanate-ethanone).

Figure 2.

 

Effect of BIX on BiP mRNA expression BIX in RGC-5 cells. (A) Time-dependent induction of BiP mRNA after treatment with 50 μM BIX and (B) concentration-dependence of BIX-induced BiP mRNA expression are each shown by semi-quantitative RT-PCR (upper panel) and real-time PCR (lower panel). β-Actin mRNA is shown as an internal control. (C) Induction of GRP94, calreticulin, p58^IPK, and ASNS mRNAs at 6 h after treatment with 2 or 10 μM BIX and at 2 to 12 h after treatment with 50 μM BIX. Data are shown as mean ± SE (n = 3 or 4). *P < 0.05, **P < 0.01 versus 0 μM (A and B) or 0 μM/0 h (C).

Figure 3.

 

Effects of BIX on tunicamycin-induced cell death and CHOP protein expression in RGC-5 cells. (A) RGC-5 cells were pretreated with vehicle or with 1 μM BIX for 12 h, and then immersed in fresh medium (control) or in medium supplemented with 2 μg/mL tunicamycin (Tm; labeled Tm or Tm + BIX). Upper photomicrographs show Hoechst 33342 and lower ones propidium iodide (PI) staining at 48 h after tunicamycin stimulation. Scale bar represents 25 μm. (B) Numbers of PI-positive cells after tunicamycin treatment. Pretreatment of cells with BIX (1 and 5 μM) significantly reduced the amount of cell death (vs. cells treated with tunicamycin alone). (C) Immunoblot of CHOP protein shows that tunicamycin induced significant CHOP expression, and that pretreatment of cells with BIX at 5 μM reduced this expression with no change in the level of β-actin. Upper panel shows CHOP and lower panel shows β-actin. (D) Quantitative representations of β-actin-based tunicamycin-induced CHOP protein expression (in arbitrary units). Data are shown as mean ± SE (n = 6 or 8). *P < 0.05, **P < 0.01 versus tunicamycin alone.

Figure 4.

 

Effect of BIX on the cell death induced by staurosporine in RGC-5. RGC-5 cells were pretreated with vehicle or with BIX (1 or 5 μM) for 12 h, and then immersed in fresh medium (control) or in medium supplemented with staurosporine at 30 nM. At the end of this culture period, cell death was assessed by WST-8 assay (Cell Counting Kit-8; Dojin Kagaku). Data are shown as mean ± SE (n = 6).

Figure 5.

 

BiP protein expression in the mouse retina induced by intravitreous injection of BIX. Immunostaining probed with an antibody against BiP/GRP78. (A) Nontreated, (B) 6 h, and (C) 12 h after intravitreous injection of BIX (5 nmol). (D) Expression ratio for BiP induction intravitreous injection of BIX is represented as the ratio of intensity values. Data are shown as mean ± SE (n = 6). *P < 0.05, **P < 0.01 versus nontreated normal retina. Each scale bar represents 25 μm.

Figure 6.

 

Effects of BIX on retinal damage induced by intravitreal injection of tunicamycin in mice. Hematoxylin and eosin staining of cross-sections of (A) nontreated, (B) Tm-treated, and (C) Tm plus BIX-treated mouse retinas at seven days after intravitreal injection of tunicamycin (1 μg) either alone or with BIX (5 nmol). (D) Damage was evaluated by counting cell numbers in GCL at seven days after the above injections. TUNEL staining of cross-sections of (E) nontreated, (F) Tm-treated, and (G) Tm plus BIX-treated mouse retinas at seven days after the above injections. (H) Effect of BIX on Tm-induced expression of TUNEL-positive cells at 24 h after the above injections. Data are shown as mean ± SE (n = 9 or 10). *P < 0.05 versus tunicamycin alone. Scale bars each represent 25 μm.

Figure 7.

 

Effects of BIX on retinal damage induced by intravitreal injection of tunicamycin in Thy-1-CFP transgenic mice. Mouse retinas (flatmounts) at seven days after intravitreal injection of (A) vehicle, (B) BIX (5 nmol), (C) Tm (1 μg), or (D) Tm (1 μg) plus BIX (5 nmol). Damage was evaluated by counting Thy-1-CFP–positive cell numbers in the four white areas shown in (E) (each area 0.144 mm² × 4 areas; total 0.576 mm²) at seven days after the above intravitreal injections. (F) Effect of BIX against Tm-induced damage (indicated by decreased number of Thy-1-CFP–positive cells) at seven days after intravitreal injection. Data are shown as mean ± SE (n = 9 or 10). *P < 0.05 versus tunicamycin alone. Scale bar represents 25 μm.

Figure 8.

 

Effect of BIX on tunicamycin-induced CHOP expression in the mouse retina. Mouse retinas (cross-sections) either (A) nontreated or at three days after intravitreal injection of (B) BIX (5 nmol), (C) Tm (1 μg) or (D) Tm (1 μg) plus BIX (5 nmol). (E) Relative density of CHOP protein expression in GCL and IPL at three days after the above intravitreal injections. Data are shown as mean ± SE (n = 6). **P < 0.01 versus tunicamycin alone. Scale bar represents 25 μm.

Figure 9.

 

Effect of BIX on tunicamycin-induced XBP-1 venus expression in ERAI mice. Mouse retinas (flatmounts) at 24 h after intravitreal injection of (A) vehicle, (B) BIX (5 nmol), (C) Tm (1 μg) or (D) Tm (1 μg) plus BIX (5 nmol). (E) Upper panel shows XBP-1 venus protein expression, while lower panel shows β-actin protein expression at 24 h after the above injections. (F) Western blot analysis showing effect of BIX on Tm-induced expression of β-actin-based XBP-1 venus protein expression at 24 h after the above injections. Data are shown as mean ± SE (n = 8). *P < 0.01 versus tunicamycin alone. Scale bars in main photomicrographs and in insets represent 25 and 5 μm, respectively.

Figure 10.

 

Effects of BIX on retinal damage induced by intravitreal injection of NMDA in mice. Mouse retinas (cross-sections) either (A) nontreated or at seven days after intravitreal injection of (B) NMDA (40 nmol) alone or (C, D) NMDA (40 nmol) plus BIX (0.5 or 5 nmol). (E) Damage was evaluated by counting cell numbers in GCL at seven days after the above intravitreal injections. (F) Effect of BIX on NMDA-induced expression of TUNEL-positive cells at 24 h after intravitreal injection of NMDA (40 nmol) either alone or with BIX (5 nmol). Data are shown as mean ± SE (n = 9 or 10). **P < 0.01 versus NMDA-treated control group. Scale bar represents 25 μm.

Figure 11.

 

Effect of BIX on NMDA-induced CHOP expression in mice. Mouse retinas (cross-sections) either (A) nontreated or at three days after intravitreal injection of (B) NMDA (40 nmol) alone or (C) NMDA (40 nmol), or (D) NMDA (40 nmol) plus BIX (5 nmol). (E) Relative density of CHOP protein expression in GCL and IPL at three days after the above intravitreal injections. Data are shown as mean ± SE (n = 6). **P < 0.01 versus NMDA alone. Scale bar represents 25 μm.

The authors thank Masayuki Miura (Department of Genetics, Graduate School of Pharmaceutical Sciences, University of Tokyo, Tokyo, Japan) for the kind gift of ERAI mice, and Rumi Uchibayashi and Shunsake Imai for technical assistance.