Transcorneal Electrical Stimulation Promotes Survival of Photoreceptors and Improves Retinal Function in Rhodopsin P347L Transgenic Rabbits

Abstract

Purpose.: To determine whether transcorneal electrical stimulation (TES) has neuroprotective effects on the photoreceptors, and whether it slows the rate of decrease of the electroretinogram (ERG) in rhodopsin P347L transgenic (Tg) rabbits.

Methods.: Six-week-old Tg rabbits received TES through a contact lens electrode on the left eye weekly for 6 weeks. The right eyes received sham stimulation on the same days. Electroretinograms (ERGs) were recorded before and at 12 weeks after the TES. After the last ERG recordings, the animals were euthanized for morphologic analysis of the retinas. Immunohistochemical (IHC) analysis was performed to detect the immunostaining by peanut agglutinin (PNA) and rhodopsin antibodies in the retinas.

Results.: The a- and b-wave amplitudes of the photopic ERGs and the b-wave amplitudes of the scotopic ERGs at higher stimulus intensities were significantly larger in the TES eyes than in the sham stimulated eyes (P < 0.05, respectively). Morphologic analyses showed that the mean thickness of the outer nuclear layer (ONL) in the visual streak at 12 weeks was significantly thicker in TES eyes than in sham-stimulated eyes (P < 0.05). IHC showed that the immunostaining by PNA and rhodopsin antibody in the TES-treated retinas was stronger than that in the sham-stimulated retinas.

Conclusions.: TES promotes the survival of photoreceptors and preserves the ERGs in Tg rabbits. Although further investigations are necessary before using TES on patients, these findings indicate that TES should be considered for therapeutic treatment for RP patients with a P347L mutation of rhodopsin.

Introduction

Patients with RP have a progressive loss of rod and cone photoreceptors that leads to a severe decrease in the visual acuity and a severe constriction of the visual field. ^1,2 The worldwide prevalence of RP is approximately 1 in 4000, meaning that more than 1 million individuals are affected worldwide. ³ As such, RP is one of the leading causes of blindness in the world. 

Many promising treatments to save or restore vision in RP patients are being investigated clinically and experimentally. ^4–9 Electrical stimulation (ES) of the retina is one of the methods that is being tried because it is less invasive than other treatments and has been shown to have neuroprotective properties on the visual system. ^10–18 ES of the transected optic nerve stump in rats promoted the survival of axotomized retinal ganglion cells (RGCs) in vivo. ¹⁰ Transcorneal electrical stimulation (TES) in rats was reported to rescue axotomized RGCs ^11,12 and promote axonal regeneration of injured RGCs. ^13,14 TES was also shown to improve the visual function of patients with traumatic optic neuropathy and nonarteritic ischemic optic neuropathy. ¹⁵  

We have demonstrated that TES promoted the survival of photoreceptors and preserved the retinal function of Royal College of Surgeons (RCS) rats, an animal model of RP. ¹⁶ Ni et al. ¹⁷ also reported that TES had neuroprotective effects on the photoreceptors after phototoxicity in rats. In a preliminary clinical trial, Schatz et al. ¹⁸ demonstrated that TES improved the visual function in RP patients. 

However, RP is a genetically heterogeneous disease, and mutations in several photoreceptor-specific and some nonspecific genes are known to cause RP. ¹⁹ Therefore, it is necessary to examine the neuroprotective effect of TES on the photoreceptors in the retinas of various RP animal models to determine which genetic type of RP is responsive to TES. 

Rhodopsin Pro 347 Leu (P347L) transgenic (Tg) rabbits have been generated by Kondo et al. ²⁰ This sequence of alterations is similar to those in human patients with autosomal dominant RP (adRP) with the rhodopsin P347L mutation. ^21,22 This animal model has a rod-dominated, progressive photoreceptor degeneration with regional variations in the pattern of photoreceptor loss. ^20,23  

The purpose of this study was to determine whether TES has a neuroprotective effect on the photoreceptors and improves the amplitudes of the electroretinogram (ERG) in Tg rabbits. Our morphologic and electrophysiological analyses showed that TES had a neuroprotective effect on the photoreceptors and improved the amplitudes of the ERG of Tg rabbits. 

Materials and Methods

Animals

All experimental procedures were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and the procedures were approved by the Animal Research Committee, Osaka University Graduate School of Medicine. Five Tg rabbits were purchased from the Kitayama Labes Co. (Ina, Japan). They were raised on a 12-hour dark 12-hour light cycle with an ambient light intensity of 100 lux. 

Transcorneal Electrical Stimulation

The rabbits were anesthetized intramuscularly with a mixture of medetomidine (0.3 mg/kg, Domitor; Orion Corporation, Espoo, Finland), midazolam (4 mg/kg, Dormicum, Astellas Pharma Inc., Tokyo, Japan), and butorphanol (5 mg/kg, Betorphal; Meiji Seika Pharma, Co., Ltd., Tokyo, Japan). For the electrical stimulation, the corneas were also anesthetized with a drop of 0.4% oxybuprocaine HCl, and a contact lens electrode with inner and outer concentric electrodes (Mayo Corporation, Nagoya, Japan) was placed on the cornea with a drop of 2.5% methylcellulose to maintain good electrical contact and prevent corneal drying. Biphasic rectangular current pulses (700 μA, 10 ms/phase duration) were delivered at a frequency of 20 Hz from an electrical stimulation system (Stimulator: SEN-7320, Nihon Kohden, Tokyo, Japan; Isolator: WPI, Sarasota, FL) through the contact lens electrode. 

TES was given to 6-week-old rabbits for 1 hour once a week until the animals were 12 weeks old. Only the left eye was electrically stimulated. The same type of contact lens electrode was placed on the right eyes but no electrical current was delivered (sham stimulation). 

Electroretinograms

ERGs were recorded from the animals at 6 weeks of age just before the beginning of the TES and after the end of the TES treatments at 12 weeks of age. For the TES, animals were anesthetized intramuscularly with a mixture of medetomidine (0.3 mg/kg), midazolam (1 mg/kg), and butorphanol (1 mg/kg). The pupils were dilated with 2.5% phenylephrine hydrochloride and 0.5 % tropicamide. 

After 1 hour of dark adaptation, the animals were restrained in a box and were prepared for the recordings under dim red light. ERGs were recorded from both eyes simultaneously with a corneal electrode carrying LEDs creating a mini-Ganzfeld stimulator (WLS-20, Mayo Corporation). A 2.5% hydroxypropyl methylcellulose ophthalmic solution was used with the corneal contact lens electrode. The reference electrode and a ground electrode were inserted subcutaneously into the left ear and the nose, respectively. 

The luminance of the scotopic ERG stimuli was increased from −5.0 to 1.48 log cd·s/m² in 0.5 or 1.0 log unit steps. After the scotopic ERG recordings, animals were light-adapted for 30 minutes, and the photopic ERGs were recorded. The luminance of photopic ERG stimuli was increased from −1.0 to 1.95 log cd·s/m², and the stimuli were presented on a white background of 25 cd/m². 

The responses were amplified, band pass filtered from 0.3 to 1000 Hz, and digitized at 3.3 kHz. A computational ERG recording system (Neuropack μ; Nihon Kohden, Tokyo, Japan) was used to average the ERG responses. Five to 20 responses were averaged with interstimulus intervals from 1 to 10 seconds depending on the intensity of the stimulus. 

ERG Analysis

The scotopic (dark-adapted) and photopic (light-adapted) a-wave amplitudes were measured from the prestimulus baseline to the peak of the a-wave, and the b-wave amplitude was measured from the trough of the a-wave to the peak of b-wave. 

To determine the significance of differences in the ERG amplitudes between TES electrically stimulated eyes and sham-stimulated eyes for the full intensity range, we plotted the average ratio of the TES-treated to the sham-stimulated eyes at all intensities and performed statistical analyses. ^22–24  

Histological Analysis

Immediately after the final ERG recordings, the rabbits were euthanized with an overdose of pentobarbital sodium. The eyes were removed and placed in a mixture of 10% neutral buffered formalin and 2.5% glutaraldehyde in 0.1 M phosphate buffer (PB) for 30 minutes at room temperature. Then eyes were trimmed, and part of the eye cups, including the optic nerve, were postfixed in 4% glutaraldehyde in 0.1 M PB at 4°C. The tissues were trimmed, embedded in paraffin, sectioned vertically, and stained with hematoxylin and eosin for light microscopy. All sections were cut along the vertical meridian of the eye passing through the optic nerve. Five serial sections of each eye were analyzed for each experimental animal. 

The degree of retinal degeneration was assessed by measuring the thickness of the outer nuclear layer (ONL), inner nuclear layer (INL), and ganglion cell layer (GCL). Photographs were taken of the superior and inferior hemispheres at 10 defined points with a camera attached to a light microscope (E800; Nikon, Tokyo, Japan). The first photograph was taken at approximately 2 mm from the center of the optic nerve head, and subsequent photographs were taken every 2 mm more peripherally. The thickness of ONL, INL, and GCL were measured on the photographs (Scion Image analyzer; Scion Corp., Frederick, MD). Each eye was coded so that the investigator making the measurements was masked to treatment of the eye. 

Immunohistochemistry

The paraffin-embedded sections (5 μm) were processed for immunofluorescence staining with antirhodopsin antibody (1:100; RET-P1; Santa Cruz Biotechnology, Santa Cruz, CA), followed by Cy3-conjugated anti-mouse IgG (1:200), and FITC-conjugated peanut agglutinin (1:100) (PNA; Invitrogen, Carlsbad, CA), a lectin that binds specifically to rabbit cone photoreceptors. The TES-treated and sham-stimulated sections were observed with a fluorescence microscope (E800; Nikon). 

Statistical Analysis

Data were analyzed with a commercial software (JMP8; SAS Institute Japan, Tokyo, Japan). The data were expressed as the means ± SDs or SEMs. Comparisons between two groups were made by Student's t-tests when the data were normally distributed or by the Mann-Whitney rank-sum test when the data were not normally distributed. Statistical significance was set at P < 0.05. 

Results

Effect of TES on Survival of Photoreceptors in Tg Rabbits

Representative retinal sections in the area of the visual streak from 12-week-old Tg rabbits that had TES (left eye) or sham stimulation (right eye) are shown in Figures 1A and 1B. The number of rows of nuclei in the ONL at the visual streak was two to three and the nuclei were closely packed in the retina receiving TES (Fig. 1A). In the sham-stimulated retina, only one row of nuclei was found in the ONL at the visual streak and they were loosely packed (Fig. 1B). In contrast, there was no difference in the structure and thickness of the ONL in other areas of the retina away from the visual streak between the TES-treated and sham-stimulated retinas (Figs. 1C, 1D). The architecture and thickness of the middle and inner retinal layers were well preserved in both TES-treated and sham-stimulated retinas (Figs. 1A–D). 

Figure 1. 

 

Photomicrographs of TES-treated and sham-stimulated retinas from 12-week-old Tg rabbits. Retinal sections of the visual streak from TES-treated retina (A) and sham-stimulated retina (B). Peripheral retinas at 6 mm superior to the optic nerve head from TES-treated retina (C) and sham-stimulated retina (D). Scale bar = 50 μm.

Quantitative analyses showed that the thickness of the ONL in the visual streak in the TES-treated eyes was 13.9 ± 3.3 μm (mean ± SD, n = 5) which was significantly thicker than that in the sham-stimulated eyes (8.8 ± 2.8 μm, n = 5, P < 0.05) (inferior hemisphere 1). In contrast, there was no significant difference in the mean ONL thickness outside the area of the visual streak (Fig. 2A). Thus, TES promoted the survival of photoreceptors in the area of the visual streak at 12 weeks of age. 

Figure 2. 

 

Thickness of the ONL (A), the INL (B), and the GCL (C) along the vertical meridian measured at 10 retinal locations at 2-mm intervals. Mean ± SD of five Tg rabbits are plotted. There was a significant difference of the mean ONL thickness between TES-treated retinas (○) and sham-stimulated retina (▪) at the visual streak (Student's t-test for two groups; *P < 0.05).

To determine whether TES affected other layers of the retina, we measured the thickness of the INL and GCL. There were no significant differences of the mean thickness of INL and GCL between the TES retinas and in the sham retinas (n = 5 each; Figs. 2B, 2C). 

Effect of TES on Electroretinograms of Tg Rabbits

To evaluate the electrical properties of the rod and cone systems of rabbits, we recorded full-field scotopic and photopic ERGs. The scotopic ERGs elicited by different stimulus intensities from 6- and 12-week-old Tg rabbits are shown in Figure 3A. The amplitudes of the scotopic ERGs recorded from the eyes of 12-week-old Tg rabbits were not reduced compared with those from the eyes of 6-week-old Tg rabbits. The intensity–response curves for the a- and b-waves are plotted in Figure 3B. Scotopic ERG a-wave amplitudes of TES-treated eyes were not significantly different from those of sham-stimulated eyes. However, the b-wave amplitudes of the TES-treated eyes were slightly but significantly larger than those of the sham-stimulated eyes at the higher stimulus intensities. 

Figure 3. 

 

Scotopic ERGs recorded from 6- and 12-week-old rhodopsin P347L Tg rabbits. (A) Scotopic ERGs elicited by eight different stimulus intensities. (B) Scotopic ERG mean amplitude versus flash intensity for the a- and b-waves in the TES-treated (○) and sham-stimulated eyes (▪) (n = 5, each, mean ± SEM). Average ratio (TES/sham) of the a- (C) and b-wave (D) amplitudes at 12 weeks of age (n = 5, each, mean ± SEM). Pointwise comparison indicated a significant difference in b-wave amplitudes at 1.48 and 0.95 log cd·s/m² (Student's t-tests for two groups; *P < 0.05).

We plotted the ratio (TES/sham-stimulated eye) of the amplitudes of the a- and b-waves for all intensities and performed statistical analyses on the differences (Figs. 3C, 3D). The differences in the ratios of the a-waves were not significant for all intensities. On the other hand, the ratios of the b-wave amplitudes were significantly larger at stimulus intensities higher than 0.95 log cd·s/m² (P < 0.05) in the TES-treated eyes. 

The photopic ERGs obtained from Tg rabbits at 6 and 12 weeks of age are also shown in Figure 4A. The amplitudes of the TES-treated and sham-stimulated eyes at 12 weeks of age were slightly reduced compared with the ERGs recorded from 6-week-old Tg rabbits but the differences were not significant. However, the responses in the eye treated with TES were larger than those treated with sham stimulation (Fig. 4A). 

Figure 4. 

 

Photopic ERGs recorded from 6- and 12-week-old rhodopsin P347L Tg rabbits. (A) Photopic ERGs elicited by five different stimulus intensities. (B) Photopic ERG mean amplitude versus flash intensity for the a- and b-waves in the TES-treated (○) and sham-stimulated (▪) retinas. Average ratio (TES/sham) of the a- (C) and b-wave (D) amplitudes at 12 weeks of age (n = 5, each, mean ± SEM). Pointwise comparison indicated a significant difference in a-wave amplitudes at 0.95 to 1.95 log cd·s/m² (Student's t-tests for two groups; *P < 0.05, **P < 0.01), and in b-wave amplitudes at 1.48 and 1.95 log cd·s/m² (Student's t-tests for two groups; *P < 0.05, **P < 0.01).

The intensity–response curve for the a- and b-waves are plotted in Figure 4B. We also plotted the average ratio of TES-treated to sham-stimulated eyes at all intensities (Figs. 4C, 4D). For a-waves, there were significant differences between TES-treated and sham-stimulated eyes at 0.95 to 1.95 log cd·s/m² (P < 0.05, respectively). For b-waves, there were significant differences between them at 1.48 and 1.95 log cds/m² (P < 0.05). 

Immunohistochemistry

Immunostaining with an antirhodopsin antibody and PNA lectin showed that the intensities of the immunostaining for both antirhodopsin antibody and PNA were stronger in the TES-treated retina (Figs. 5A–C) than the sham-stimulated retina (Figs. 5D–F). 

Figure 5. 

 

Immunohistochemical analysis of rod and cone photoreceptors triple labeled with rhodopsin (green), PNA (red), and DAPI (blue) in TES-treated (A–C) and sham-stimulated retinas (D–E) at 12 weeks of age (approximately 4 mm inferior to the optic nerve head). Intensities of rhodopsin and PNA immunostaining are stronger in the TES-treated ratio than in the sham-stimulated retina. Scale bar = 50 μm.

Discussion

Our electrophysiological and histological analyses showed that TES led to the survival of photoreceptors in the visual streak, and it also led to the preservation of ERG responses at higher stimulus intensities in rhodopsin P347L Tg rabbits. Although the cause of the photoreceptor degeneration in Tg rabbits is different from that in RCS rats and the phototoxic-induced degeneration in rats, ^{20,23,25–27} TES also had a neuroprotective effect on the photoreceptors in Tg rabbits. These findings indicate that TES might have a similar neuroprotective effect on photoreceptors whose degeneration has different causes. 

In the histological analysis, only the photoreceptors in the visual streak were rescued by TES, and in the areas outside the visual streak, the number of photoreceptors in the TES-treated retina was not significantly different from that in sham-stimulated retina. In Tg rabbits, the loss of photoreceptors was maximum in the visual streak where the photoreceptor density is highest, and the loss of photoreceptors was not significantly different at other regions outside visual streak at 12 weeks of age. ²⁰ Therefore, at 12 weeks of age, the loss of photoreceptors was striking only in the visual streak, indicating that the neuroprotection of photoreceptors was limited to the visual streak. 

Immunohistochemical analysis showed that the intensity of both PNA and rhodopsin immunostainings was stronger in the TES-treated retinas than in the sham-stimulated retinas in the visual streak. 

However, the results of ERGs indicated that TES preserved the cone components better than rod components, although in Tg rabbits the rod components are more affected than the cones. ^20,23 Although it was not determined why the cone components were better preserved than the rod components, one possibility is that TES promoted the survival of both rod and cone photoreceptors, and the rescued rods secreted a cone viability factor to rescue the cone photoreceptors. ²⁸ Otherwise, at 12 weeks of age, photoreceptors near the visual streak were much more affected than those outside the visual streak, ²⁰ therefore the differences of ERG amplitudes of full field ERGs between TES-treated and sham-stimulated retinas might be detected only at higher stimulus intensities. 

There are some possible mechanisms for the neuroprotection of photoreceptors. First, TES increased the expression of the mRNA and protein levels of neurotrophic factors (e.g., insulin-like growth factor-1 (IGF-1), brain-derived neurotrophic factor (BDNF), ciliary neurotrophic factor, or B-cell lymphoma-2 in the retinas after TES. ^11,17 A second possibility is that TES reduced the expression of the TNF super families and Bax, which are related to apoptosis signaling in retinal cells. ²⁹ Cultured rat Müller cells exposed to electrical currents have been shown to express IGF-1, BDNF, and fibroblast growth factor-2 (FGF-2). ^30–32 Other types of electrical stimulation to the retinas, such as subretinal electrical stimulation, increases the expression of FGF-2 in the retinas. ³³ Unfortunately, we did not determine whether the expression of any of these neurotrophic factors was increased after TES in the Tg rabbit retinas. 

Another possible mechanism for the TES-induced neuroprotection was an increase of chorioretinal blood circulation by TES. ^34,35 In clinical studies, TES has been shown to improve the visual function of patients with retinal artery occulsion. ^36,37 Thinning of the vascular plexus and the development of aberrant vessels have been reported in RP patients and animal models of RP. ^38–41 This indicates that retinal blood circulation might be reduced in Tg rabbits. TES might have some neuroprotective effects on photoreceptors by increasing chorioretinal blood circulation. 

We did not examine whether TES was neuroprotective for the photoreceptors in the peripheral retina. In Tg rabbits at the age of 48 weeks, almost all of the photoreceptors were lost ^20,27 ; however, it takes a long time to investigate the neuroprotective effects of TES on the entire retina until the age of 48 weeks from 6 weeks, so it is difficult to continue the treatment until 48 weeks because weekly anesthesia and treatment put a heavy load on animals and is adverse to the animal welfare for long-term experiments. The results that TES did have neuroprotective effects on photoreceptors in the visual streak at 12 weeks of age were enough to lead us to determine the neuroprotection of TES on the photoreceptors in Tg rabbits. 

Rhodopsin P347L Tg rabbits are an adRP model of human RP. Our results indicate that TES might have a neuroprotective effect on the photoreceptors in RP patients with the same mutation. Schatz et al. ¹⁸ performed a prospective, randomized sham-controlled clinical study, and reported that TES improved the visual function in RP patients. From these neuroprotective effects of TES already published and our results, TES might exert a neuroprotective effect on photoreceptors of different animals with RP. Additional investigations on different animal models are necessary to determine which type of RP was the indication of TES treatment. 

In conclusion, TES had a neuroprotective effect on the photoreceptors in the visual streak of rhodopsin P347L Tg rabbits, which is a model of human adRP. These results support and encourage clinical trials of TES for RP patients. 

Acknowledgments

The authors thank Yuko Furukawa and Emi Higasa for technical assistance, and Duco I. Hamasaki for help with the manuscript. 

Footnotes

 Disclosure: T. Morimoto, None; H. Kanda, None; M. Kondo, None; H. Terasaki, None; K. Nishida, None; T. Fujikado, None