New photic stimulating system with white light-emitting diodes to elicit electroretinograms from zebrafish larvae

The embryos of zebrafish (Danio rerio) of the RIKEN wild-type strain (RIKEN WT) were obtained from the National BioResource Project of the RIKEN Brain Science Institute (RIKEN, Saitama, Japan). The embryos and larvae were maintained under a 14-hour light (approximately 500 lx) and 10-hour dark cycle, and the temperature of the aquarium E3 medium (5 mM NaCl, 0.17 mM KCl, 0.33 mM CaCl₂, and 0.33 mM MgSO₄) was 28 °C. The age of the larvae was based on the days post-fertilization (dpf), and they were tested at an age of 5–7 dpf, which is the age that has been commonly used in studies on the visual system of zebrafish larvae. All experiments were performed at room temperature (28–30 °C) and in the afternoon. After the completion of the experiments, the larvae were euthanized by a lethal dose of 3-aminobenzoic acid methyl ester (MESAB; Sigma, St. Louis, MO, USA).

The research was conducted in full compliance and strict accordance with the Association for Research in Vision and Ophthalmology (ARVO) Resolution on the Use of Animals in Ophthalmic and Vision Research. The protocol was approved by Mie University Graduate School of Medicine Institutional Animal Care and Use Committee.

Our custom-made LED light stimulator (MAYO Co., Inazawa, Japan) was constructed in a quadrangular shape (Fig. 1a). The dimensions of the quadrangle were 8.0 cm wide, 8.0 cm deep, and 8.0 cm high. The bottom and top surfaces were open. A part of the right wall was cut so that the glass microelectrode could be mounted on the microelectrode holder located in the quadrangular light stimulator (Fig. 1a, arrow). The inside of the quadrangle was partitioned off with a plate which had a circular hole of 2.0 cm in diameter (Fig. 1b, arrow). After painting the inside of the quadrangle and partition matte white, twelve white LEDs (NS6W083BT, NICHIA Corporation, Tokushima, Japan) were installed on the inside of the lower part of the quadrangle (Fig. 1c, arrows). These LEDs served as the source for the stimulus and the background illumination.

All LEDs were covered with copper nets to reduce the electrical artifacts (Fig. 1c). The LEDs were driven by an electronic control unit (LS-100, MAYO Co., Inazawa, Japan), which controlled the light intensity (current) and duration of the stimulus. Light from the LEDs was reflected from the inner surface of the enclosure. This setup was designed so that it could be attached to the lower part of a stereo or dissecting microscope without blocking the view of the larvae (Fig. 1d).

All of the electrodes, stereomicroscope, and the LED light stimulator were placed inside a grounded shielded cage of 80 × 70 × 140 cm (Fig. 2). In our system, an anti-vibration table was not used to improve the signal-to-noise ratio [19].

After filling a glass micropipette with an opening of about 20 μm at the tip with E3 medium, a chloride silver wire electrode was inserted the micropipette and then fixed to a microelectrode holder (E45SW-F10PH, Warner Instruments, Hamden, CT, USA). The glass microelectrode holder was fixed and moved by a micromanipulator (UM-1PF: Narishige Group, Tokyo, Japan, Fig. 1d, arrow) so that the tip of the microelectrode was correctly placed on the center of the cornea using a stereomicroscope. The reference electrode was a chlorided silver pellet that was placed under a moist paper towel that was resting on a sponge in a 35 mm Petri dish containing E3 medium.

The chlorided silver wire in the microelectrode and the reference electrode were connected to a bioamplifier (AVB-10, Nihon Kohden, Tokyo, Japan). The electrical signals from the larva were differentially amplified 2000 times with bandpass cut-off frequencies of 0.8 and 300 Hz for all of the recordings. The amplified signals were fed to a PowerLab 2/25 instrument (ADInstruments Pty. Ltd., South Wales, Australia) using the Scope version 4.1 software (AD Instruments Pty. Ltd.), and data acquisition, storage, and analyses were performed with a personal computer (iMac^®; Apple Computer, Inc., Cupertino, CA, USA).

ERGs were recorded under scotopic (dark-adapted) and photopic (light-adapted) conditions. All specimens were dark-adapted for 30–40 min prior to the recordings, even though a shorter period of dark adaptation is reported to be sufficient in 5–7 dpf larvae [19]. All of the preparations and setups were done under dim red illumination to minimize light adaptation. After the dark adaptation, the larvae were anesthetized by submersion in a solution of 0.02% MESAB in E3 medium until the swimming motions stopped. They were then paralyzed by submersion in a 0.8 mg/ml Esmeron (Organon Teknika, Eppelheim, Germany) solution in E3 medium. They were then positioned on their side on a piece of moistened paper towel that was placed on a reference electrode in a Petri dish (Fig. 2.). Larvae’s entire body except the head was covered with a small strip of paper towel moistened with a solution of 0.02% MESAB in E3 medium. The damp paper towel kept the larvae moist during the experiments. Then, the Petri dish was placed on the stage of the microscope, and the glass microelectrode was positioned at approximately the center of the cornea. Although we did not supply the O₂ during the ERG recording, the experimental period was less than 30 min.

The electroretinograms (ERGs) were recorded under scotopic (dark-adapted) and photopic (light-adapted) conditions. After the larvae were positioned under the stereomicroscope under dime red illumination, they were dark-adapted additionally for more than three minutes in complete darkness prior to the scotopic ERG recording. For the photopic ERG recordings, the larvae were light-adapted to a background illumination (31.6 cd/m²) for over 5 min before the recordings. The stimulus intensity was changed in 0.5 log unit steps from −3.0 log units to 0 log units with log 0 corresponding to 3160 cd/m² (photopic unit). The background luminance was 31.6 cd/m² for photopic condition which was similar to that used in earlier studies [18]. The stimulus duration was 1000 ms which allowed the recordings of the ON and OFF responses separately. For each stimulus intensity, 3–10 responses were averaged with an inter-stimulus interval of 10–60 s for scotopic conditions or 3 s for photopic conditions.

The amplitude of the b-wave was measured from the trough of the a-wave to the peak of the positive b-wave. The implicit times of the b-wave were measured from the stimulus onset to the peak of the b-wave.

To evaluate the reliability of our recording conditions, we determined the variations in the b-wave amplitude among the larvae under the same conditions. These trials consisted of placing the glass electrode on the cornea and recording the ERGs three times under the same conditions from 12 larvae. The stimulus intensity was −1.0 log unit on a steady background of 31.6 cd/m².

Two-way layout ANOVA was used to evaluate the variations of the inter-recording setup trials and the inter-individual differences. To evaluate the reliability of the inter-recording trials, intra-class correlation coefficients (ICC) were also calculated. The ICCs were classified as: ‘excellent’ (≧.81), ‘good’ (.61–.80), ‘moderate’ (.41–.60), and ‘poor’ (≦.40) according to past biometrical studies [20, 21]. All statistical analyses were performed with the SPSS Statistics version 23. A P value of <0.05 was considered significant.

We were able to adjust the position of the larvae and place the glass electrode at approximately the center of the cornea easily and quickly. We were also able to confirm and adjust the position of the microelectrode at any time during the experiment. It took approximately one to two minutes for the set up and not more than 30 min for recording all scotopic and photopic ERGs.

The a-, b-, and d-waves were recorded under both scotopic and photopic conditions, and their shapes were similar to those reported earlier for zebrafish larvae [6, 13‐19]. Under scotopic conditions (Fig. 3a), positive b- and d-waves were present at the lowest stimulus intensity of −3.0 log units tested. The a-waves were very small and were first observed at −1.5 to −1.0 log units. As the intensity increased, the b-wave gradually increased while the positive d-waves were not clearly observed at the middle intensities of −1.5 to −1.0 log units. At the maximum 0 log unit intensity, the a-wave and a large negative-going wave after the b-wave were present while the positive-going d-wave appeared again.

Under the photopic conditions (Fig. 3b), the b- and d-waves were not observed at the lowest intensity (−3 log units) but were present at −2.5 log units. The amplitudes of both increased with the increase in stimulus intensities. The a-wave was observed at −1.5 to −1.0 log unit and increased with the increases in the stimulus intensities.

The changes in the amplitudes and implicit times of the scotopic ERG b-waves are plotted as a function of stimulus intensity in Fig. 4a, b, respectively (n = 12). The amplitudes of the scotopic ERG b-wave increased rapidly as the stimulus intensity increased at the lower stimulus intensities of −3.0 to −1.5 log unit; then, it increased more slowly until it plateaued at −0.5 to 0 log unit. In contrast, the scotopic ERG b-wave implicit times decreased monotonically with the increase in stimulus intensity.

The amplitudes of the b-wave amplitudes of the photopic ERG in which placing the glass recording electrode on the cornea and recording the ERGs were repeated three times under the same conditions from 12 larvae are plotted in Fig. 5. The differences in the b-wave amplitudes for the inter-recording trials were not significantly different (167.1 ± 85.3 μV, 169.9 ± 79.0 μV, 173.2 ± 85.5 μV; F = 0.71; P = 0.50; two-way layout ANOVA). However, the inter-individual differences were statistically significant (n = 12 larvae, F = 134.9; P < 0.001; two-way layout ANOVA). The inter-trial reliability (intra-class correlation coefficient, ICC) was 0.98 for the three trials. These results indicated that the inter-trial reliability of the ERG recorded by our system was ‘excellent,’ and the variability in the b-wave amplitudes was mainly caused by the differences between the subjects.