Citation: Kurose T, Sugano E, Sugai A, Shiraiwa R, Kato M,
Mitsuguchi Y, Takai Y, Tabata K, Honma Y, Tomita H. Neuroprotective effect of a
dietary supplement against glutamate-induced excitotoxicity in retina. Int J
Ophthalmol 2019;12(8):1231-1237. DOI:10.18240/ijo.2019.08.01
·Basic Research·
Neuroprotective effect of a dietary supplement
against glutamate-induced excitotoxicity in retina
Takahiro Kurose1, Eriko Sugano2,
Akihisa Sugai2, Raki Shiraiwa2, Mariyo Kato1,
Yoko Mitsuguchi1, Yoshihiro Takai1, Kitako Tabata2,
Yoichi Honma1, Hiroshi Tomita2,3
1Rohto Pharmaceutical Co., Ltd.,
6-5-4 Kunimidai, Kizugawa, Kyoto 619-0216, Japan
2Laboratory of Visual Neuroscience,
Graduate Course in Biological Sciences, Iwate University Division of Science
and Engineering, 4-3-5 Ueda, Morioka, Iwate 020-8551, Japan
3Clinical Research, Innovation and
Education Center, Tohoku University Hospital, 1-1 Seiryo, Aoba, Sendai, Miyagi
980-8574, Japan
Correspondence to: Hiroshi Tomita and Eriko Sugano. Iwate University Division of Science and
Engineering, 4-3-5 Ueda, Morioka, Iwate 020-8551, Japan. htomita@iwate-u.ac.jp;
sseriko@iwate-u.ac.jp
Received: 2019-02-04
Accepted: 2019-06-04
Abstract
AIM: To
evaluate the neuroprotective effect of a dietary
supplement (ClearVision EX®; CV) against glutamate-induced
excitotoxicity in retina.
METHODS: We
evaluated the protective effects CV on glutamate-induced cell toxicity of an
immortalized mouse hippocampal cell line (HT-22) in vitro and
N-methyl-D-aspartate (NMDA) induced retinal injury in vivo. Once-daily
oral administration of CV or vehicle (5% Arabic gum) was started the day before
the NMDA injection and continued until the end of the study. Electroretinograms
(ERGs) were recorded to evaluate the retinal function at 2d after NMDA
injection. Furthermore, a histological evaluation, Western blot analysis, and immunohistochemistry
were performed for assessing the signal transduction pathway.
RESULTS: HT-22
cell death was induced by the addition of glutamate and co-incubation with CV
protected against it. Oral administration of CV inhibited the decrease in
scotopic threshold response amplitudes induced by the intravitreal injection of
NMDA and those of the thickness of the inner retinal layer in the histological
evaluation. The increased phosphorylated levels of extracellular signal-regulated
kinase (ERK) but not cAMP response element binding protein (CREB) or Akt were
observed 1h after NMDA injection in both the vehicle- and CV-treated rats;
however, pERK activation was no more upregulated at 3h after NMDA injection.
pERK upregulation was observed in Müller cells.
CONCLUSION: CV
shows a protective effect against both glutamate-induced HT-22 cell death and
NMDA-induced retinal damage. pERK upregulation in the Müller cells plays a key
role in the protective effect of CV against glutamate-induced retinal toxicity.
KEYWORDS: glutamate-induced toxicity; retinal
ganglion cells; HT-22 cells; pERK
DOI:10.18240/ijo.2019.08.01
Citation: Kurose T, Sugano E, Sugai A, Shiraiwa R, Kato M,
Mitsuguchi Y, Takai Y, Tabata K, Honma Y, Tomita H. Neuroprotective effect of a
dietary supplement against glutamate-induced excitotoxicity in retina. Int J
Ophthalmol 2019;12(8):1231-1237
INTRODUCTION
Retinal ganglion cells (RGCs), which
are the third-order neuron in the visual pathway in the retina, play a role in
transmitting visual signals from the retina to the brain. RGCs are injured in
various types of retinal diseases such as glaucoma[1],
ischaemic neuropathy[2], and hereditary optic
neuropathy[3]. The progression of these diseases
has a possibility of loss of vision. However, critical treatments have not been
developed yet.
There are various types of glaucoma
such as primary open angle glaucoma, primary angle-closure glaucoma, and
normal-tension glaucoma (NTG). Glaucoma is a neurodegenerative condition
characterized by a progressive loss of RGCs and associated with visual field
defects and thin optic nerve fiber layer[4]. One
of the causes of glaucomatous retinal injury is
thought to be related to ocular hypertension[5],
which is the major risk factor for glaucoma and causes the death of RGCs[6]. The molecular mechanisms underlying of the death of
RGCs, such as changes in axonal flow including neurotrophic factors[7-9], oxidative stress, ischaemia[10], and excitotoxicity[11], have been reported as potential contributors to
glaucomatous injury. Recently, inflammatory reactions related to the
progression of RGC death have been reported[12-13]. Thus, the mechanisms involved are very complicated
and effective neuroprotective agents have not been explored.
Neuroprotective effects of various
types of antioxidants on neuronal cell death have been reported. We have
reported that thioredoxin-2, a small redox-active protein[14],
protected retinal epithelial cells that played a critical role in retinal
degeneration from oxidant-induced apoptosis[15-16]. However, it takes a long time to develop gene
therapies for application in the clinical setting.
Dietary supplementation is a candidate approach for
protecting neurons from neurodegenerative conditions. One of the benefits of
using a dietary supplement is to apply it for human easily because the safety
data of all of contents in the dietary supplement have been established.
ClearVision EX® (CV) is a commercialized Japanese dietary vitamin
supplement based on the Age-Related Eye Disease Study (AREDS) formulation,
containing β-carotene, vitamin C, vitamin E, lutein, zeaxanthin, niacin, zinc,
and copper. From detailed description of contents in CV, it is expected to have
some anti-oxidant effects. Thus, in this study, we evaluated the efficacy of CV
for managing N-methyl-D-aspartate (NMDA)-induced retinal degeneration.
MATERIALS AND METHODS
Ethical Approval
The
use of animals in these experiments was in accordance with the ARVO Statement
for the Use of Animals in Ophthalmic and Vision Research and conducted with the
consent of the Animal Research Committee of Rohto Pharmaceutical Co., Ltd., Japan.
Cell Culture The immortalized mouse hippocampal
cell line HT-22 was cultured in Dulbecco’s modified Minimum Essential Medium
(Thermo Fisher Scientific, Tokyo, Japan) supplemented with 10% foetal bovine serum under a 5% CO2 atmosphere at
37℃. The culture medium was changed
every 3d and cells were passaged using a 0.02% ethylaminediaminetetraacetic acid/phosphate-buffered saline
(PBS) solution. We investigated appropriate glutamate concentrations and cell
densities to induce glutamate toxicities because cell toxicities depend on cell
densities. We determined the glutamate concentration to 1.5 mmol/L at the cell density of 5×103 cells/well
that showing about IC50 (data not shown). The HT-22 cells were
plated at a density of 5×103 cells/well in a 96-well plate and
incubated with glutamate (1.5 mmol/L) and CV (1.6-200
μg/mL). After incubation for 24h, cell viability was evaluated using an MTS
cell proliferation colorimetric assay kit (Funakoshi, Tokyo, Japan). Cell toxicity was evaluated using a lactate
dehydrogenase (LDH) cytotoxicity assay kit (Funakoshi, Tokyo, Japan).
Animals Sprague-Dawley rats (SLC Japan Inc.,
Shizuoka, Japan) aged 8wk were maintained under a 12-hour/12-hour light-dark cycle (light intensity: 300 lx) with unlimited access
to laboratory chow and water. Once-daily oral administration of CV (30, 100, or
300 mg/kg) or vehicle (5% arabic gum) was started on the day before NMDA
injection and continued until the end of the study. Thirty or fourteen rats
were used for electroretinogram (ERG) recordings and histological analysis
or Western blot analysis, respectively.
Reagents CV was supplied by Rohto
Pharmaceutical Co., Ltd (Osaka,
Japan).
Antibodies against extracellular signal-regulated kinase (ERK) (Cat No. 4695S),
phospho-ERK (pERK) (Cat No. 9101S), Akt (Cat No. 9272), pAkt (Cat No. 9271L),
cAMP response element binding protein (CREB) (Cat No. 9197S), pCREB (Cat No.
9198S), and β-actin (Cat No. sc-69879) were obtained from Cell Signaling
Technology (Tokyo, Japan). Glutamine synthetase (GS) antibody (Cat No. MAB302)
was obtained from Millipore (Tokyo, Japan). Secondary antibodies against rabbit
(Cat No. S3731) and mouse immunoglobulin G (IgG)
(Cat No. S3721) were obtained from Promega (Tokyo, Japan).
N-methyl-D-aspartate-induced Retinal
Damage Retinal damage was induced using
NMDA (Sigma, St. Louis, MO, USA). Under anesthesia by continuous inhalation of
isoflurane (Pfizer Japan Inc., Tokyo, Japan), 5 μL of NMDA solution (1 mmol/L) was injected into the vitreous by entering the eye
at the ora serrata by using a 32-gauge needle on a
Hamilton syringe.
Electroretinogram Recordings ERGs were recorded to evaluate the
retinal function at 2d after NMDA injection as
described previously[17]. In brief, the rats were
dark-adapted for a minimum of 12h. A small contact lens with an electrode was
then mounted on the cornea and a reference electrode was placed under the
tongue. The amplitude of the scotopic threshold response (STR, -5 log cds/m2) of ERG was determined.
Histological Analysis Ten days after the NMDA injection, the rats were euthanized by CO2 inhalation. The
eyes were enucleated and immersed in Karnovsky fixative solution and then
embedded in paraffin. Serial sections (4 μm) of whole eyes were cut
horizontally through the cornea and parallel to the optic nerve and stained
with haematoxylin and eosin. Light microscopy images
of the histological sections were photographed (Olympus IX71, Osaka, Japan), and the ratio of the thickness of the
whole retinal layer (WRL) and the inner plexiform layer (IPL) was calculated (IPL/WRL).
Western Blot Analysis Western blot analysis was performed
using the methods described previously[18].
Briefly, the retinas were isolated from the eyes 1 and 3h after NMDA injection.
Protein was extracted with a lysis buffer containing 10 mmol/L Tris-HCl (pH 7.5), 1 % Triton X-100, 0.5 % NP-40, 1 mmol/L EDTA, 150 mmol/L NaCl, 1× protease inhibitor
cocktail (Thermo Scientific, Tokyo,
Japan),
and 1× Halt™ phosphatase inhibitor cocktail (Thermo Scientific, Tokyo, Japan). The lysate was quantified using a bicinchoninic
acid assay (BCA) protein assay kit (Pierce, Rockford, IL). Thirty micrograms of
protein was electrophoresed on 4%-15%
Mini-PROTEIN TGX gels (Bio-Rad, Tokyo,
Japan)
and transferred onto a polyvinylidene difluoride (PVDF) membrane. After
blocking with a membrane, the membrane was incubated with the first antibody.
Following washing, the membrane was incubated with the secondary antibody, an
alkaline phosphatase-conjugated antibody. Chemiluminescence detection (CDP-Star
Detection Reagent; GE Healthcare, Tokyo, Japan) was performed according to the standard procedure.
Band densities were measured using ImageQuant software (GE Healthcare, Tokyo, Japan) and the density of each band was normalized to that
of β-actin.
Immunohistochemistry Immunohistochemistry was performed
on the retina obtained 3h after NMDA injection. Paraffin-embedded sections were
deparaffinized according to the standard procedure. The sections were treated
in citrate buffer (pH 6.0) in a microwave oven as an antigen retrieval method.
After blocking, sections were incubated in a solution of pERK antibody
(dilution 1:200) and GS antibody or in a control solution of anti-rabbit and
anti-mouse IgG overnight at 4℃.
The sections were then washed three times with PBS containing 0.01% Tween20 and
incubated with Alexa488- or Alexa594-conjugated anti-rabbit and anti-mouse IgG
solution at room temperature for 1h. Following washing, the sections were
covered with mounting media including 4’,6-diamidino-2-phenylindole (DAPI;
VECTASHIELD: Funakoshi, Tokyo).
Statistical Analysis Statistical analyses for in vitro
experiments were performed using GraphPad Prism 4 (GraphPad Software, San
Diego, CA, USA). Data are expressed as mean±SD
values. The statistical method used was Tukey’s multiple comparison test. Statistical analyses in vivo experiments were
performed using StatLight (Yukms Corp., Tokyo, Japan). Data are expressed as
mean±SD values. Statistical comparisons were made using Student’s t-test
or Dunnett’s test after checking for homoscedasticity.
RESULTS
Effect of CV on Glutamate-induced
Toxicities in HT-22 Cells Cell viability decreased upon the
addition of glutamate; however, after the addition of CV, the viability increased
and was higher than that of the control cells (Figure 1A). The increased cell
viabilities might indicate that CV affected cell proliferation. We tested the
effects of CV on cell proliferation (Figure 1B). CV treatment at a concentration
of 1.6 μg/mL positively affected cell proliferation. To evaluate cell toxicity
excluding the proliferative activities, we measured LDH release from cells
(Figure 1C). Increased LDH levels were observed in the glutamate-treated cells.
On the other hand, the increase in LDH levels was inhibited by the addition of
8 μg/mL of CV.
Figure 1 Effects of CV on
glutamate-induced toxicity in cultured HT-22 cells A: Cell viability was determined 24h after treatment with
1.5 mmol/L glutamate; B: The effect of CV on cell proliferation; C: Cell toxicity induced by glutamate was evaluated based
on the LDH release. Data shown as mean±SD (n=10, Dunnet’s multiple
comparison test, aP<0.01, bP<0.001
vs control).
Effects of CV on NMDA-induced
Retinal Toxicity RGC function was evaluated by
measurements of STRs by ERG recordings 2d after NMDA injection (Figure 2A). The
amplitudes of the STRs recorded in rat retinas treated with an intravitreal
injection of NMDA markedly decreased compared to those recorded in non-injected
rats. On the other hand, the decrease in the STRs tended to increase in a CV
dose-dependent manner, and a significant increase was observed in rats
administered a CV dose of 300 mg/kg compared to that in the
vehicle-administered rats. The thickness of the IPL of the histological
sections was evaluated (Figure 2C). The thickness of the inner retinal layer
clearly decreased in the NMDA-injected rats. CV administration had a protective
effect against inner retinal damage. The percentage of IPL thickness to that of
the WRL in the NMDA-injected retinas markedly decreased compared to that in the
non-injected retina. The decreased value of IPL/WRL in the vehicle-administered
rats was significantly higher than that in the rats administered CV at 300
mg/kg (Figure 2B).
Figure 2 Effect of CV on
NMDA-induced retinal damage in rats A: Intravitreal injection of NMDA decreased the STR
amplitude (n=12, aP<0.05 vs vehicle,
Dunnett’s test); B: The thickness of the IPL was determined and the data
are shown as the percentage to the WRL thickness (n=6, aP<0.05 vs vehicle, Dunnett’s test); C: The representative histological sections obtained from
rats with intravitreal injection of NMDA and treated with CV (30, 100, and 300
mg/kg) or vehicle.
Western Blot Analysis To investigate the effect of CV on
the early response related to the cell death pathway, we performed the Western
blot analysis (Figure 3). Phosphorylated or non-phosphorylated ERK, CREB and
Akt, normalized by β-actin and phosphorylated levels were investigated. pERK
levels increased at 1h after the treatment with vehicle and CV. The significant
difference was also detected in pERK of 3h after the treatment with CV and the
level of ERK in the CV-administered rats increased at 3h after the NMDA
injection. Significant differences in the levels of pERK/ERK, as a result, were
only detected at 1h after the NMDA injection in the vehicle group (Figure 3A).
On the other hand, the levels of phosphorylated and non-phosphorylated CREB and
Akt were unchanged both in the vehicle- and CV-administered rats (Figure 3B, 3C).
Figure 3 Western blot analysis
showing the effects of NMDA injection on ERK, CREB, and pAkt Western blots probed with antibodies
against pERK, ERK (A), pCREB, CREB (B), pAkt, Akt (C), and β-actin. Quantitative
analyses were performed (n=7, Tukey’s
multiple comparison test, aP<0.05, bP<0.01).
Immunohistochemistry To determine the area of pERK
activation, immunohistochemical studies were performed. In the normal retinas
(untreated retinas and retinas without NMDA injection), only weak pERK
immunoreactivity was observed in the whole retina (Figure 4A: green). Increased
pERK immunoreactivity was observed 3h after the NMDA injection (Figure 4B, 4C).
In the vehicle-treated rat retinas, pERK immunoreactivity was mainly observed
in the ganglion cell layer (GCL) and IPL (Figure 4B). In the CV-treated rat
retinas, the inner nuclear layer (INL) also showed immunoreactivity (Figure
4C).
Figure 4 Immunohistochemical
analysis of retinas 3h after NMDA injection pERK, GS, and nuclear staining are
seen in the retinas of the control group (no NMDA injection) (A), NMDA-injected
retinas treated with vehicle (B) or NMDA-injected retinas treated with CV (C)
as green (pERK), red (GS), and blue (DAPI), respectively. Arrow head and arrows
indicate pERK staining in the INL and double-positive immunoreactivity,
respectively.
DISCUSSION
The aim of this study was to
evaluate the effects of CV, a vitamin supplement based on AREDS formulation
that has been commercialized in Japan, on glutamate-induced excitotoxicity and
to investigate the mechanisms underlying its protective effects. Various kinds
of factors such as glutamate excitotoxicity, nitric oxide, generation of
reactive oxygen species (ROS), and reduction of blood flow are related to the
induction of glaucomatous injury, although elevated intraocular pressure (IOP)
is the main risk factor for glaucoma[19]. Among
these causes, we focused on retinal injury
induced by ROS. The two main results obtained were as follows: pre-administration
of CV protected the retina from NMDA-induced toxicity. The major effect
underlying the protective mechanism of CV is radical scavenging by vitamins.
NMDA-induced neuronal toxicities are
implicated in the pathological mechanisms of Alzheimer’s disease[20], Huntington disease[21],
and Parkinson’s disease[22]. One possible
mechanism of NMDA-induced neurotoxicity has been well identified as follows:
prolonged NMDA receptor activation causes the opening of a cationic channel and
increases Ca2+ concentration in depolarized cells, which results in
destabilization of mitochondrial membrane potentials. The release of ROS and
cytochrome c from the mitochondria lead to cell death[23]. To elucidate the possible
mechanism underlying the protective effect of CV, we used HT-22 cells. HT-22
cells, which lack an ionotropic glutamate receptor such as an NMDA receptor
under undifferentiated conditions[24], are a good
model for evaluating neuronal oxidative stress. It has been reported that
glutamate-induced cell death is induced by inhibition of the cystine/glutamate
exchanger. The inhibition leads to decreased glutathione production and subsequently causes
intracellular ROS production[25-26].
CV clearly showed a protective effect against glutamate-induced toxicity in
HT-22 cells, indicating that CV acted as a radical scavenger during oxidative stress.
In in vivo experiments, we
observed decreased STR amplitudes indicating dysfunction of RGCs 2d after the
intravitreal injection of NMDA. It is known that STRs are a representative
marker of RGC function[27-28].
Histological examinations indicated that the physiological dysfunctions
associated with the STRs were morphological damages but not transient. In the
histological examinations, we calculated the percentage of the inner retinal
layer to evaluate the damage to the inner retinal neurons. The protective
effects of CV were observed in both evaluations, ERGs and histological
examinations. We also analysed the signal transduction pathways
related to NMDA-induced retinal toxicity. There was a significant increase in pERK/ERK level in the vehicle-administered rats.
However, pERK level in CV-administered rat retinas continuously increased and a
significant difference was observed at 3h after NMDA injection. The
upregulation of pERK is well identified as a marker of NMDA-induced retinal
toxicity[29-30]. We previously
reported that pERK has a protective role in ischaemia-induced retinal damage[31].
Some studies have reported that co-injection of U0126, an ERK inhibitor, exacerbated
NMDA toxicity[30,32].
Therefore, we hypothesized that continuous pERK activation in Muller cells, but
not the level of pERK/ERK, was important for protecting the retina from
NMDA-induced toxicity.
In summary, CV had protective
effects against NMDA-induced retinal damage and cell death induced by oxidative
stress. We showed that Müller cells had a key role in the protective effect.
Our results indicate that oral vitamin supplementation may prevent retinal cell
death caused by oxidative stress. However, the data showing here is the
protective role of CV on ROS-related toxicities. The ganglion cell death caused
by glaucoma is not simple, as indicated above. Further studies using other
models such as increased IOP or optic nerve crush model is needed to confirm
the efficacy of CV to use as a daily supplementation for patients with
glaucoma.
ACKNOWLEDGEMENTS
We express our heartfelt appreciation to Ms. Misao Enomoto
of Laboratory of Visual Neuroscience for maintaining the cell culture used in
this study.
Foundations:
Supported by the Rohto Pharmaceutical
Co., Ltd. Furthermore, it was partly supported by Grants-in-Aid for Scientific
Research from the Ministry of Education, Culture, Sports, Science and
Technology of Japan (No.16H05485; No.16K15729;
No.16K11314; No.17H06330).
Conflicts of Interest: Kurose T, Kato M, Mitsuguchi Y, Takai Y,
Honma Y, are
employed by the Rohto Pharmaceutical Co., Ltd. Sugano E, None; Sugai
A, None; Shiraiwa R, None; Tabata K, None; Tomita H,
None.
REFERENCES