Japanese Journal of Ophthalmology

, Volume 57, Issue 1, pp 120–125

Possible implications of acid-sensing ion channels in ischemia-induced retinal injury in rats


  • Takatomo Miyake
    • Department of Ophthalmology and Visual ScienceNagoya City University Graduate School of Medical Sciences
    • Department of Ophthalmology and Visual ScienceNagoya City University Graduate School of Medical Sciences
  • Tsutomu Yasukawa
    • Department of Ophthalmology and Visual ScienceNagoya City University Graduate School of Medical Sciences
  • Shinya Ugawa
    • Department of Anatomy and NeurobiologyNagoya City University Graduate School of Medical Sciences
  • Shoichi Shimada
    • Department of Neuronal Cell BiologyOsaka University Graduate School of Medicine
  • Yuichiro Ogura
    • Department of Ophthalmology and Visual ScienceNagoya City University Graduate School of Medical Sciences
Laboratory Investigation

DOI: 10.1007/s10384-012-0213-9

Cite this article as:
Miyake, T., Nishiwaki, A., Yasukawa, T. et al. Jpn J Ophthalmol (2013) 57: 120. doi:10.1007/s10384-012-0213-9



Retinal ischemia in eyes with diabetic retinopathy and retinal vein occlusion leads to local tissue acidosis. Acid-sensing ion channels (ASICs) are expressed in photoreceptors and other neurons in the retina, and may play a role in acid-induced cell injury. The purpose of this study was to investigate the neuroprotective effects of amiloride, an ASIC blocker, on induced retinal ischemia in rats.


Transient retinal ischemia was induced in male Long–Evans rats by the temporary ligation of the optic nerve. Just before the induction of ischemia, the experimental eyes underwent intravitreal injection of amiloride. On day 7, the retinal damage in eyes that underwent amiloride treatment (and in those that did not undergo the treatment) was evaluated by histology and electroretinogram (ERG).


Transient retinal ischemia caused retinal degeneration with thinning of the inner layer of the retina. The blockage of ASICs with amiloride significantly prevented retinal degeneration. ERG demonstrated that the reduction in a- and b-wave amplitudes induced by the transient retinal ischemia was significantly prevented by the application of amiloride.


The present study suggests that ASICs might, at least in part, play a pathophysiological role in ischemia-induced neurodegeneration. Blockage of ASICs may have a potential neuroprotective effect in ocular ischemic diseases.


AmilorideAcid-sensing ion channel (ASIC)NeuroprotectionRetinal ischemia


Retinal ischemia is a common cause of visual impairment and blindness in a variety of retinal diseases such as diabetic retinopathy and retinal vein occlusion. While severe capillary nonperfusion leads to retinal edema in or around the ischemic lesions and neovascularization, transient or chronic retinal ischemia is associated with local retinal degeneration, reflecting neuronal cell death [1]. Several self-reinforcing destructive cascades, such as neuronal depolarization, calcium influx, oxidative stress initiated by energy failure, increased glutamatergic stimulation, and leukocyte-associated reperfusion injury accompanying the generation of reactive oxygen species and inflammatory cytokines, are implicated in the cell deaths [1, 2].

In brain ischemia, neurons and glial cells become depolarized with the loss of ATP, leading to excessive synaptic release of glutamate and reduced glutamate uptake [3]. Accumulation of glutamate in the extracellular space overactivates the postsynaptic N-methyl-d-aspartate (NMDA) receptors, resulting in intracellular Ca2+ overload and subsequent neuronal cell death [48]. However, recent clinical trials using NMDA receptor antagonists have not produced effective results in cases of brain ischemic injury [9, 10]. Similarly, an intravitreal injection of NMDA induced retinal neuronal death, whereas blockage of the NMDA receptor failed to completely protect retinal neurons against transient retinal ischemia [11]. These observations indicate that there must be some other mechanisms underlying ischemic tissue damage. Generally, in ischemic lesions, anaerobic glucose metabolism predominantly accompanies the accumulation of lactic acid and the release of H+ from ATP hydrolysis, resulting in marked reduction of the tissue pH (local tissue acidosis) [12], and the protons generated may be a possible link between ischemia and the injury.

Acid-sensing ion channels (ASICs) are neuronal proton-gated cation channels related to the degenerin/epithelial sodium channel (DEG/ENaC) superfamily [1316]. Because local tissue acidosis can activate the channels, ASICs are likely to be key molecules in ischemia-induced retinal injury [15]. ASICs comprise four proteins that form functional subunits (ASIC1a, ASIC1b, ASIC2a, and ASIC3) and two modulatory subunits (ASIC2b and ASIC4) which, by themselves, are inactive. ASICs have strong Na+ selectivity, and the excess Na+ influx through the channels into cells can cause swelling and rupture of the cells [17]. Importantly, ASIC1a is permeable to not only Na+ but also Ca2+, and the activation of ASIC1a could cause Ca2+ toxicity in ASIC1a-expressing cells.

Accumulating research on brain ischemia suggests that ischemia-derived acidosis provokes neuronal degeneration via aberrant activation of ASICs [1820]. In acidic conditions, extracellular protons continuously activate local ASICs, thereby causing cytotoxicity through the excess influx of Na+ and Ca2+ into the cells. It is reported that amiloride, an ASIC blocker, potently attenuated ischemia-induced brain damage in experimental animal models [19, 21]. In particular, knockout of the ASIC1a gene protected brains suffering from ischemic insults [19, 20]. Thus, ASICs are increasingly implicated in ischemic tissue damage, and ASIC1a is a key molecule involved in Ca2+ toxicity in brain ischemia.

ASICs are also expressed in photoreceptors and other cells in the retina, where they regulate phototransduction [2225]. In the present study, we evaluated the neuroprotective effects of amiloride to elucidate possible implications of ASICs in ischemic retina in vivo.

Materials and methods

Animal model

Transient retinal ischemia was induced by a previously described method with minor modifications [26]. Male-pigmented Long–Evans rats (200–250 g) were used. Only the right eye of each rat was subjected to ischemia. The rats were anesthetized with an intraperitoneal injection of ketamine (86 mg/kg) and xylazine (13 mg/kg). The pupils were dilated with 0.5 % tropicamide and 2.5 % phenylephrine hydrochloride. After lateral conjunctival peritomy and disinsertion of the lateral rectus muscle, the optic nerve head of the right eye was exposed using blunt dissection. A 6-0 nylon suture was passed around the optic nerve and tightened until the blood flow ceased in all the retinal vessels. After 60 min of ischemia, complete nonperfusion was confirmed through an operating microscope, and the suture was removed. Reperfusion of the vessels was also observed microscopically. Eyes that failed to reperfuse within 5 min were excluded from experiments. Sham-operated rats underwent the same surgery but without tightening of the suture.

Amiloride treatment

Amiloride (Sigma–Aldrich, St. Louis, MO, USA) stock solution (20 mM) dissolved in dimethyl sulfoxide was prepared. It was subsequently diluted to a concentration of 2.4 mM with normal saline. Intraocular injection was done just before the induction of ischemia. The 2.4 mM solution was injected intravitreally with a 30 G needle (Becton–Dickinson, Franklin Lakes, NJ, USA) attached to a 10 μl Hamilton syringe under an operating microscope. A concentration of 100 μM amiloride was chosen in our experiments based on the previous observation that it was the most consistently effective concentration for blocking inward currents in Müller cells [22]. The amount of amiloride solution injected (2.5 μl) was determined by assuming an average vitreous chamber volume of 60 μl [27]. Vehicle-treated rats were given the same volume of saline. Sham-operated rats were used as controls. All experiments were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.

Retinal thickness

Four eyes from four rats each in the amiloride-treated, vehicle-treated, and sham-operated control groups were obtained seven days after reperfusion. When we extracted the eyeballs, we checked the top of the dorsal point (using the optic axis as the landmark) and stained it with hematoxylin and eosin for the purposes of orientation. The rats were killed with an overdose of anesthetic and the operated eyes were immediately enucleated. The posterior portion of each eye was immediately frozen on dry ice in Tissue-Tek (Sakura Finetek, Tokyo, Japan) OCT compound at optimal temperature. Cryostat sections were cut through the optic nerve every 10 μm and fixed in 4 % paraformaldehyde in phosphate-buffered saline for staining by hematoxylin and eosin. Sections were cut perpendicular to the retinal surface. To quantify the retinal damage induced by transient retinal ischemia, we measured changes in retinal thickness using the method described by Hughes [28]. The thickness of the overall retina from outer segment to inner limiting membrane was measured.

Two sections from each eye were measured. The thickness of each section was measured in the retina at a distance of 1.5 mm from the center of the optic nerve head (nasal and temporal portions) and the values were averaged.

Electroretinogram (ERG) examination

At seven days after reperfusion, full-field ERG was recorded by a previously described method with minor modifications [29]. The animals were dark-adapted overnight and anesthetized with an intraperitoneal injection of ketamine (86 mg/kg) and xylazine (13 mg/kg). The rats were placed on a heating pad to maintain their body temperatures at close to 37 °C. The pupils were dilated with topical 0.5 % tropicamide and 0.5 % phenylephrine HCl, and ERGs were recorded with a gold wire loop electrode (Corneal Electrode, MAYO Corporation, Inazawa, Aichi, Japan) placed on the cornea, which had been anesthetized with 1 % tetracaine; a drop of 1 % methylcellulose was also applied to assure good contact. A differential electrode was placed under the skin on the forehead, and a neutral electrode was inserted subcutaneously near the tail. Electrical signals were amplified with a gain of 4,000 and a bandwidth of 0.3–500 Hz. The signals were digitized at a rate of 10 kHz rate using a data acquisition system (PowerLab ML825, AD Instruments, Bella Vista, Australia). Scotopic ERGs were recorded with an interstimulus interval of 3–60 s, and the interval increased with increasing stimulus intensity. The mice were placed in a Ganzfeld bowl (SG-2002 stand-alone Ganzfeld, LKC Technologies Inc. Gaithersburg, MD, USA), and 20–30 stroboscopic stimuli were averaged using a repetition rate of 1 s to record the photopic ERG. The maximum luminance was 2.2 log cd s/m2, and neutral density filters were used to reduce the stimulus intensity. Steps ranging from −5.0 to 2.2 log cd s/m2 were used to elicit the scotopic ERG. Sham-operated rats were evaluated as controls. Five different rats from each of the treated and control groups were measured seven days after the treatment.

Statistical analysis

All values are presented as the mean ± SEM. Statistical comparisons between two groups were performed using the unpaired t test. ANOVA was used to compare three or more conditions, with post hoc comparisons tested using the Bonferroni procedure. Probabilities of P < 0.05 were considered to be statistically significant.


Retinal thickness

Light microscopic examinations of retinal sections were performed in eyes that underwent the sham operation and in eyes that were treated with vehicle or amiloride just before the induction of transient ischemia (Fig. 1a). After the ischemia, retinal thickness was significantly decreased in vehicle-treated rats (159 ± 5.5 μm, n = 4) compared with sham-operated rats (173 ± 3.6 μm, n = 4) (P < 0.05), verifying ischemia-induced retinal injury. Retinal damage was significantly reduced in the amiloride-treated group (172 ± 8.3 μm, n = 4) (P < 0.05) (Fig. 1b). There was no significant difference between the sham-operated and amiloride-treated groups.
Fig. 1

The effect of intravitreal amiloride on retinal thickness in eyes with transient retinal ischemia. a Histological sections of the retina on day 7 in sham-operated eyes or in eyes treated with vehicle or amiloride just before the induction of ischemia. Scale bar 50 μm. b Full retinal thickness was measured. After ischemia, the retinal thickness was significantly decreased in vehicle-treated rats as compared with sham-operated rats. Retinal damage was significantly reduced in the amiloride-treated group (*P < 0.05)

Scotopic ERG

Representative scotopic ERGs elicited by increasing the stimulus intensities from sham-operated rats and vehicle-treated or amiloride-treated ischemic rats are shown in Fig. 2a. Plots of a- and b-wave amplitude at high intensity (2.2 log cd s/m2) for all groups of rats are shown in Fig. 2b, c. In the ERG evaluation, retinal ischemia caused reduced a- and b-wave amplitudes (a-wave 172.3 ± 76.2 μV, b-wave 435.0 ± 199.5 μV). The amiloride treatment significantly prevented the reductions in both waves (a-wave 324.1 ± 123.2 μV, b-wave 851.4 ± 331.8 μV) (P < 0.05), and there was no significant difference between the sham-operated and amiloride-treated groups.
Fig. 2

In vivo effects of retinal ischemia and amiloride treatment on ERGs. At seven days after ischemia–reperfusion injury, full-field ERGs were recorded. a Representative scotopic ERGs elicited by different stimulus intensities for sham-operated, vehicle-treated, and amiloride-treated ischemic rats. Mean a-wave (b) and b-wave (c) amplitudes obtained at high intensity (2.2 log cd s/m2). Significantly reduced a- and b-wave amplitudes (a-wave 172.3 ± 76.2 μV, b-wave 435.0 ± 199.5 μV) were induced with ischemia–reperfusion injury (*P < 0.05). The amiloride treatment prevented the reductions in the amplitudes of both waves (a-wave 324.1 ± 123.2 μV, b-wave 851.4 ± 331.8 μV) (*P < 0.05)


Amiloride sensitivity is the hallmark of ASICs. ASIC1a is expressed in cone photoreceptors and in horizontal as well as some amacrine, bipolar, and retinal ganglion cells (RGCs) [24]. ASIC2a and ASIC2b are expressed in photoreceptors and some cells in the inner nuclear layer [25]. ASIC3 is present in rod photoreceptors and in horizontal as well as some amacrine and retinal ganglion cells [30]. These ASICs are considered to be negative modulators of phototransduction. In pathological conditions, ASICs play a role in RGC death induced by hypoxia [31]. In the present study, we investigated the protective effect of amiloride against ischemic damage using the rat transient retinal ischemia model. Our morphological results clearly demonstrate the preservation of retinal thickness in amiloride-treated eyes, and suggest that the amiloride treatment ameliorated retinal ischemic injury through ASIC blockade.

In our experiments, scotopic ERG was used to evaluate retinal function. The a-wave provides information about photoreceptor function (mainly rod photoreceptors), while the b-wave provides information about the functional activity of Müller cells, bipolar cells, or both. In the ischemic rats, the amiloride treatment led to significant suppression of decrease in scotopic electroretinogram a- and b-wave amplitudes, implying that 100 μM amiloride preserves not only retinal morphology but also functions in eyes suffering from a transient ischemic insult, which is in agreement with previous observations that amiloride increased the amplitudes of a- and b-waves on scotopic ERG in the normal retina [32]. Apart from this, the knockdown of ASIC1a by antisense oligonucleotides and the in vivo blocking of ASIC1a activity by PcTx1, a specific inhibitor of ASIC1a, were reported to cause substantial decreases in the photopic a- and b-waves of the ERG and in oscillatory potentials in normal adult brown Norway rats [24]. Those findings suggest that among ASIC subtypes, ASIC1a, at least in part, plays a significant role in normal phototransduction. Hence, photopic electrophysiological recordings will be needed to clarify the mechanism by which ASIC1a affects the retinal ischemic injury.

The amiloride treatment did not extinguish all of the injury induced by the retinal ischemia–reperfusion, indicating that there must be some other underlying mechanisms for the ischemic tissue damage. In the brain, the NMDAR-Ca2+/calmodulin-dependent protein kinase II (CaMKII) cascade was found to functionally couple to ASICs and to contribute to acidotoxicity during ischemia, raising the possibility that specific blockade of the NMDAR/CaMKII-ASIC coupling could reduce neuronal death after ischemia and other pathological conditions involving excessive glutamate release and acidosis [33]. Therefore, specific inhibition of the NMDAR/ASIC coupling may be effective in retinal ischemic injury too. The susceptibility of RGCs to NMDA injury varies depending on the RGC type [34]. Similar differences in susceptibility to amiloride treatment may exist among cell types, partly because not all RGCs express a high density of ASIC1a channels [35]. It is worth noting that ASIC1a is Ca2+ permeable, and that the activation of the channel can directly induce Ca2+ toxicity in the cells. The detailed distributions of transcripts of ASIC1a and its proteins need to be elucidated in order to achieve a better understanding of the efficacy of the amiloride treatment.

In the retina, edaravone, a free-radical scavenger that is used in stroke patients and has some protective effects without affecting the circulation in various animal models, significantly reduced retinal ischemic damage [36]. In addition, intraocular irrigation with d-allose reduces extracellular glutamate and attenuates oxidative stress during vitrectomy, and this behavior may protect the retina against ischemia-induced damage [37]. Those observations indicate that oxidative stress plays a pivotal role in the mechanism of the retinal damage, and the uncoupling of oxidative stress and ASICs may potently reduce the ischemic injury.

Prior to the present experiments, we examined the protective effect of amiloride at estimated concentrations of 10, 100, and 500 μM, and found that 10 or 500 μM amiloride conversely worsened the retinal damage (data not shown). Needless to say, amiloride and its derivatives are also known to block other cation channels and exchangers, such as T-type Ca2+ channels [38] and Na+/H+ antiporters [39], as well as other member proteins belonging to the DEG/ENaC superfamily. In particular, there is the possibility that the concentration of the drug used may modulate the activities of T-type Ca2+ channels [10] and epithelial sodium channels [40], both of which are also expressed in the retina. Therefore, further studies using selective inhibitors for each channel, knockout mice, and gene silencing are needed to determine the specific pathophysiological roles of ASICs.

In conclusion, approximately 100 μM amiloride prevented ischemia-induced retinal injury in rats, suggesting that ASICs are implicated in ischemia-induced retinal degeneration and that ASIC blockers such as amiloride and benzamil might be potent neuroprotective agents in eyes with retinal ischemic diseases. Further experiments will be needed to determine which ASIC subtypes are implicated in ischemia-induced retinal injury, and whether the blockade of ASICs by amiloride or other agents provides neuroprotective benefits rather than late-onset drug neurotoxicity.


We wish to thank K. Miyata and M. Kondo for their technical assistance in measuring the ERG.

Conflict of interest

The authors declare that they have no competing financial interests.

Copyright information

© Japanese Ophthalmological Society 2012