The zebrafish has become an established model organism for the study of hearing and balance systems in the past two decades. The classical approach to examine hair cells is to use dye to conduct selective staining, which shows the number and morphology of hair cells but does not reveal their function. Startle response is a behavior closely related to the auditory function of hair cells; therefore it can be used to measure the function of hair cells. In this study, we developed a device to measure the startle response of zebrafish larvae. By applying various levels of stimulus, it showed that the system can discern a 10 dB difference. The hair cell in zebrafish can regenerate after damage due to noise exposure or drug treatment. With this device, we measured the startle response of zebrafish larvae during and after drug treatment. The results show a similar trend to the classical hair cell staining method. The startle response was reduced with drug treatment and recovered after removal of the drug. Together it demonstrated the capability of this behavioral assay in evaluating the hair cell functions of fish larvae and its potential as a high-throughput screening tool for auditory-related gene and drug discovery.
Due to its miniature size, prolific reproduction, and the external development of the transparent embryo, the zebrafish is a leading model for developmental and genetic studies, as well as in toxicology and omics-based research [
The zebrafish carries numerous valuable features as a model in auditory research. For instance, several dozens of hearing-related genes have been discovered in zebrafish and many of them similarly influence the inner ear of humans and other vertebrates [
Loss of sensory hair cells is the leading cause resulting in deafness or hearing deficits, and the process is not reversible in mammalian vertebrates. There is no or very limited hair cell regeneration after hair cell damage or death. Postnatal hair cell death in humans is often induced by bacterial infections, damage from prolonged noise exposure, and treatments with certain ototoxic drugs such as aminoglycoside antibiotics or chemotherapeutic agents. In contrast to mammalian vertebrates, robust hair cell regeneration occurs in most nonmammalian vertebrates, including zebrafish [
The hair cell regeneration in zebrafish is usually assessed through staining the hair cells and microscopically counting the cell number. In brief, the drug dose-dependent hair cell death can be examined with a particular ototoxin, such as neomycin, and subsequent time-lapsed cell regeneration can be investigated with borderline-hair cell death that is achieved by appropriate drug dose [
Functional examination of zebrafish hair cell is difficult due to the lack of reliable quantification methods, compared to the electrophysiological measurement of auditory brainstem response, or otoacoustic emissions in mice. Zebrafish do harbor a rich repertoire of motor behaviors neurologically initiated by their sensory organs, either the lateral line system or the auditory system [
Deviant from previous systems for startle response measurement, significant improvement has been made to increase the accuracy. Using the system, we have successfully quantified the startle response in zebrafish (1) immediately after, (2) one day after, and (3) three days after drug exposure. The hair cells in the lateral line were also stained and counted at stages (1) and (3) to verify the damage and regeneration. The startle response results showed similar trends as what hair cell counting did but with much less effort. It demonstrated that this system can facilitate regenerative research in the zebrafish and improve and expedite our understandings in regenerative pathways and regulations in hair cell development and regeneration.
Wild-type TU fish line was raised and maintained in a recirculating aquaculture system according to standards described by Kimmel et al. [
Neomycin was used to induce damage in neuromast hair cells. It was applied to 7-dpf zebrafish larvae in the culture medium for duration of 24 hours. At the end of drug treatment, 8-dpf zebrafish were incubated in 8
An instrument system was developed to measure the startle responses of the fish larvae. The schematic of the system was shown in Figure
Recording the startle response in zebrafish. (a) Instrumentation for the measurement of startle response. (b) Moving traces identified from multiple picture frames after delivering a stimulus. (c) Characteristic C-bend motion identified in a single picture frame from a subset of zebrafish.
A digital camera system was mounted on a microscope frame to monitor the Petri dish and zebrafish from the top. With the transillumination, the fish larva body appears as dark region in each image frame and the fish larvae were segmented from the background with in-house software, developed within MATLAB (MathWorks, MA, USA). With the segmented fish body, the position of the fish within the Petri dish can be located. By connecting the position in each frame for a fish larva, its movement during each experiment can be extracted from the recorded video. As proposed in [
To verify the efficacy of the instrument system, an experiment was performed to measure the startle responses of zebrafish larvae to sound stimulus with different intensity in fish that were treated with or without ototoxic drug. To test the relationship between the startle response and stimulus level, 400 Hz tone bursts with 3 different sound levels were applied to the amplifier to drive the vibrator. The stimulus mid-level was chosen by visually observing that more than 5 larvae (without drug treatment) showed significant movement. The high level is about 10 dB above and the low level is about 10 dB below the mid-level. For each stimulus level, 10 repeats were performed to achieve the statistical significance. Between each stimulus, 100 sec of break was applied to avoid the adaptation, as suggested in [
With the instrument system introduced earlier, we performed the startle responses as well as the traditional hair cell counting technique to monitor the recovery of the auditory function of the zebrafish larvae. In each test, 10 larvae were placed in the Petri dish. Two testing systems were used in parallel so that in total 20 larvae were tested for each experiment. The stimulus waveform was a tone burst of 160 ms with 30 ms rise and fall time, as shown in Figure
Time course for the measurement of startle response in zebrafish.
One hundred zebrafish larvae were used in the present study. At 7 dpf, larvae were divided into three groups: control (i.e., 0) and 0.16
In parallel with the startle response test, the hair cell damage by neomycin treatment was confirmed by observing and counting the hair cell with staining. As in previous test, the larvae were divided into three groups, with neomycin concentration of 0, 0.16, or 1.6
We quantified the startle response by zebrafish larvae’s moving distance upon sound stimulation. Figure
Characterizing the startle response. (a) Mean moving distance of larvae with 400 Hz sound of intensity linearly grown from 5 to 60 with an arbitrary unit. This results in a sound level in about 20 dB. The values are
Previous studies have mostly demonstrated that neomycin exposure ablated hair cells in the lateral line in a dose-dependent manner [
Neomycin-induced neuromast hair cell damage and regeneration. (a) Confocal image of lateral line neuromasts under neomycin treatment in wild-type zebrafish. (b) Average number of neuromast hair cells in each group. Each group consists of 10 7 dpf zebrafish larvae treated with respective concentration for 24 h and then allowed to recover for 72 h to assess hair cell regeneration. All neomycin-treated larvae showed decreased number of hair cells to some extent; statistical analyses were performed using Student’s
Using the same experimental condition with neomycin treatment, we also evaluated the startle response with our in-house instrument system. The tone burst attributes were the same as previously described and the tone frequency was 400 Hz. The startle responses were checked 24 hrs and 72 hrs after the drug treatment. The mean moving distance of the control group is used as a reference at each checkpoint. The responses of the drug treatment groups were normalized by that of the control group to eliminate the possible variation in startle responses between different days. At 8 dpf (0 h in Figure
Startle response of fish larvae with neomycin treatment of different level of concentration and subsequent recovery period after drug removal. Mean moving distance after sound stimulation was used as the quantification parameter. To eliminate the variation of different days, the values are scaled by that of the control group in each day. The value in each column from left to right is
In this study, a mini shaker was used as the driver to deliver acoustic vibration to the Petri dish and produce the sound stimulus to fish larvae. Although this is not a direct sound generation, it is an effective way of delivering sound stimulus. Using a load speaker in air is not efficient because of the air-water interface, where 95% of sound energy is reflected back. An aquatic speaker can be used underwater but it is not practical in this setup because the water level is only a few millimeters inside the Petri dish. The mini shaker was previously used in [
In one of the previous systems [
With the experimental protocol in the present study, the test sensitivity was comparable between the morphological hair cell counting method in Figure
In this study, we developed a behavioral assay to evaluate the auditory function of hair cells by measuring the startle response of zebrafish larvae. By applying various level of stimulus, results showed that the system can discern a 10 dB sound level difference. Using the system, we investigated the hair cell damage and regeneration in the lateral line neuromasts of zebrafish larvae. The result from this system shows similar trend to the traditional hair cell counting methods. The startle response was reduced with neomycin treatment and recovered with hair cell regeneration. These results demonstrated the capability of this behavioral assay in evaluating the hair cell functions of zebrafish larvae and its potential as a high-throughput screening tool for auditory-related gene and drug discovery.
The authors declare that there are no competing interests regarding the publication of this paper.
This work was supported by the National Natural Science Foundation of China (Grant no. 81470701), the Shenzhen Overseas Talents Innovation Plan (Grant no. KQCX20140522150857838) awarded to FC, and the US Army Medical Research and Materiel Command (Grant no. W81XWH1410006) awarded to HL.