Test animals and husbandry
Wild type adult zebrafish (AB line) were initially obtained from China Zebrafish Resource Center (CZRC, China) and reared at the zebrafish facility of the University of Saint Joseph, Macao. Fish were maintained in 10 L tanks in a standalone housing system (model AAB-074-AA-A, Yakos 65, Taiwan) with filtered and aerated water (pH balanced 7–8; 400–550 μS conductivity) at 28 ± 1 °C and under a 12:12 light: dark cycle. Animals were fed twice daily with live artemia and dry powder food (Zeigler, PA, USA). The fish used in this study were 6–8 months old, both males and females (1:1), with a total length of 2.2–3.1 cm. The total number of specimens tested was 30 for the auditory sensitivity measurements and inner ear morphological analysis (6 fish per experimental group), and 78 for the Novel Tank Diving assay (15-18 fish per group).
All experimental procedures complied with the ethical guidelines regarding animal research and welfare enforced at the Institute of Science and Environment, University of Saint Joseph, and approved by the Division of Animal Control and Inspection of the Civic and Municipal Affairs Bureau of Macao (IACM), license AL017/DICV/SIS/2016. This study was conducted in compliance with the ARRIVE guidelines60.
Noise treatments
Prior to acoustic treatments, all subjects were transferred to 4 L isolation glass tanks that were placed in a quiet lab environment (Sound Pressure Level, SPL: ranging between 103 and 108 dB re 1 μPa) for a minimum of 7 days. These tanks had no filtering system but were subject to frequent water changes, and the light, temperature and water quality were kept similar to the stock conditions. This adaptation period was important to reduce potential effects of noise conditions from the zebrafish housing system.
After this period, groups of six zebrafish were transferred into separate acoustic treatment glass tanks (dimensions: 59 cm length × 29 cm width × 47 cm height; 70 L)—Fig. 1 Supplementary, where they remained 24 h in acclimation. Each tank was equipped with an underwater speaker (UW30, Electro-Voice, MN, USA) housed between two styrofoam boards (dimensions: 3 cm thick × 29 cm width × 47 cm height) with a hole in the centre, positioned vertically in one side of the tank. Another similar sized board was positioned in the opposite side of the tank and fine sand was placed in the bottom to minimize transmission of playback vibrations into the tank walls. Each treatment tank was mounted on top of styrofoam boards placed over two granite plates spaced by rubber pads to reduce non-controlled vibrations.
Four acoustic treatment tanks were prepared for this study to be used alternately between trials and cleaning procedures, but only two were used simultaneously. When two tanks were being used, one contained specimens under acclimation and the other fish under a specific acoustic treatment. The tanks were housed in a custom-made rack and placed at least 1 m apart to minimize acoustic interferences. The tanks were used randomly for the different treatments across the various trials.
The speakers were connected to audio amplifiers (ST-50, Ai Shang Ke, China) that were connected to laptops running Adobe Audition 3.0 for windows (Adobe Systems Inc., USA). After the acclimation period, specimens were exposed to white noise playbacks (bandwidth: 100–3000 Hz) at 150 dB re 1 µPa for 24 h, starting in the morning between 10 and 11 a.m. The bandwidth adopted covered the best hearing range of zebrafish27, as well as the frequency range of most anthropogenic noise sources, such as pile driving and vessels2.
Sound recordings and SPL measurements were made with a hydrophone (Brüel & Kjær type 8104, Naerum, Denmark; frequency range: 0.1 Hz–120 kHz, sensitivity of − 205 dB re 1 V/μPa) connected to a hand-held sound level meter (Brüel & Kjær type 2270). Noise level was adjusted with the speaker amplifier so that the intended amplitude (LZS, RMS sound level obtained with slow time and linear frequency weightings: 6.3 Hz–20 kHz) was achieved at the centre of the tanks before each treatment. A variation in SPL of ±10 dB was registered in the closest and farthest points (in relation to the speaker). The sound spectra of the noise treatments were relatively flat similar to the setup described in a prior study by Breitzler et al.27.
Moreover, the acoustic treatments were calibrated with a tri-axial accelerometer (M20-040, frequency range 1–3 kHz, GeoSpectrum Technologies, NS, Canada) with the acoustic centre placed in the middle of the tank. The sound playback generated was about 120 dB re 1 m/s2, with most energy in the horizontal axis perpendicular to the speaker, which was verified based on previously described methods using a MATLAB script paPAM16.
In this study four sound treatments were used with varying temporal patterns similar to Sabet et al.18—Fig. 1: continuous noise (CN); intermittent regular noise with a fast pulse rate—1 s pulses interspersed with 1 s silence (IN1,1); intermittent regular noise with a slow pulse rate—1 s pulses interspersed with 4 s silence (IN1,4) and intermittent random noise—1 s pulses interspersed with 1, 2, 3, 4, 5, 6 or 7 s silent intervals in randomized sequence (RN1,7) leading to a mean interval of 4 s. All intermittent patterns had 5 ms ramps to fade in and fade out pulses for smooth transitions. In the “control” treatment tank, the amplifier connected to the speaker was switched on but without playback.
After each treatment, two specimens were tested for audiometry, two were tested with the NTD assay and another two were euthanized and dissected for inner ear morphological analysis.
Auditory sensitivity measurements
Auditory Evoked Potential (AEP) recordings were conducted immediately after noise treatments. The AEP recording technique adopted followed previously described procedures27. The recordings were conducted in a rectangular plastic tank (50 cm length × 35 cm width × 23 cm height) equipped with an underwater speaker (UW30) positioned in the bottom and surrounded by fine sand. A custom-built sound stimulation system with enhanced performance at lower frequencies (< 200 Hz) was mounted in the centre of the tank wall facing the front and consisted of a vibrating plexiglass disc driven by a mini-shaker (Brüel & Kjær 4810)—see further details in Breitzler et al.27. The experimental tank was placed on top of an anti-vibration air table (Vibraplane, KS kinetic systems, MA, USA), which was housed in a walk-in soundproof chamber (2.13 × 2.13 × 2.0 m, 120a-3, IAC Acoustics, North Aurora, IL, USA) constructed as a Faraday cage.
Test adult zebrafish were slightly anaesthetized in 0.12 g/L tricaine methanesulfonate bath (MS-222, Arcos Organics, NJ, USA) buffered with equal concentration of sodium bicarbonate27,31. Specimens were positioned in a custom-designed sponge holder, where they remained immobilized with a fine net covering the upper body maintaining normal breathing.
The holder was positioned so that the fish head was just beneath the water surface. Two stainless steel electrodes (0.40 mm diameter, 13 mm length, Rochester Electro-Medical, Inc., FL, USA) were used. The recording electrode was positioned firmly against the skin over the brainstem region, and the reference electrode was placed on the side of the body.
Both sound stimuli presentation and AEP recordings were accomplished using the workstation from TDT (Tucker-Davis Technologies, FL, USA). The AEP recording was fed into a low impedance head stage (RA4LI, TDT) connected to a pre-amplifier (RA4PA, TDT, 20× amplification), band-pass filtered (0.1–1 kHz), and digitized (16 bit, ± 4 mV). The output was then sent to a Multi-I/O processor (RZ6, TDT). Sound stimuli and AEP recordings were controlled with SigGen and BioSig TDT software. Stimuli consisted of tone bursts of 100, 200, 400, 600, 800, 1000, 2000, 4000, and 6000 Hz, presented randomly, with 20 ms duration and 2 ms rise/fall time. Tone stimuli ranged from 2 (100 Hz) up to 80 (6000 Hz) complete cycles and were presented at least 1000 times, half at opposite polarities (180° phase shifted). Prior to each experiment, a hydrophone (Brüel & Kjær 8104) connected to a sound level meter (Brüel & Kjær 2270) was used for calibrating the stimuli at the position that would be occupied by the fish head in the recording tank. Although it would be ideal to calibrate the system also in particle motion, this was not possible due to the size of available accelerometer and setup constraints. Nevertheless, the information provided should be enough for a comparison between experimental groups.
For each frequency, tones were initially presented at 140 dB re 1 µPa and then in 2.5 dB decreasing steps. The auditory threshold was defined as the lowest SPL at which a visible and repeatable AEP response was identified in at least two averaged waveforms. The validation of an auditory response was based on at least two out of three possible criteria: (1) general waveform shape matching response from previous sound level (visual inspection); (2) presence of a delay or increased latency in the auditory response measured in a consistent AEP peak; and (3) presence of a spectral peak in the FFT analysis of the auditory response that is double of the stimulation frequency.
The peak latency of auditory responses was determined for each individual at 1000 Hz and 120 dB re 1 µPa stimulation. This measure was determined based on the time interval between the tone stimulus onset and the maximum peak of the AEP response. An overall mean TTS was also calculated for each noise treatment group based on the mean auditory thresholds obtained for the several frequencies subtracted by the mean thresholds obtained from control group at the same frequencies.
Inner ear morphological analysis
Saccular epithelia were obtained from selected specimens immediately following noise treatments. Fish were euthanized with overdose MS-222 (300 mg/L) buffered with sodium bicarbonate. We followed the methods previously described to extract zebrafish inner ear sensory epithelia61 and double stain saccular hair cells and pre-synaptic Ribeye b31.
Firstly, the fish heads were cut and the ventral part of the skull was cracked open, and then fixed in 10% neutral buffered formalin solution (Sigma, USA) at 4 °C overnight. After fixation, samples were thoroughly rinsed with PBS and then dissected under a dissecting stereomicroscope (Stemi 2000CS, Zeiss).
In order to stain the hair cells (HC) and Ribeye b, saccular epithelia were first permeabilized with 1% Triton X-100 RT for 2 h, followed by blocking with 10% goat serum in PBS and incubated with primary antibody against the Ribeye b protein (mouse anti-zebra Ribeye b monoclonal antibody, gift from Dr. T. Nicolson, Stanford University, CA, USA; 1:5000) at 4 °C overnight. In the following day, samples were firstly rinsed with 1% PBST for 2 h, then incubated with Alexa Fluor 647 IgG2a secondary antibody (Invitrogen, USA; 1:500) and Alexa Fluor 488 phalloidin (Invitrogen, USA; 1:500) for 2 h, and finally rinsed in PBS for 30 min. Saccular epithelia were further labelled with DAPI (Invitrogen, USA; 1:1000) for visualization of cell nuclei and whole mounted with Fluoromount-G (Southern Biotech, USB).
The tissue samples were imaged with a confocal imaging system (STELLARIS 5 LIA, Leica Microsystems CMS GmbH) using Leica Application Suite software X (LAS X, Leica Microsystems Leider Lane) and further analysed with image J (Version 1.53e, National Institutes of Health, USA). Both HC bundles and Ribeye b puncta were quantified manually in seven non-overlapping squared regions of 900 mm2 located across the length of the rostral-caudal axis of the saccule, as shown in Fig. 4A and according to previous studies31. All HC bundles and Ribeye b puncta within or overlapping the square outlines were included in the counts.
Novel Tank Diving assay
To evaluate the impact of noise treatments on swimming activity and anxiety-like behaviour, two fish from each acoustic treatment tank were tested with the Novel Tank Diving (NTD) assay62 immediately following the treatments. NTD was tested in a rectangular-shaped glass transparent tank (26 cm length × 13 cm width × 19.5 cm height) filled with 4 L of system water. To perform NTD test, a single fish was simply transported to the experimental tank and the behaviour recorded with a digital camera (SONY HDR-PJ675, Japan) for 6 min. The water from the experimental tanks was replaced between trials to avoid potential effects from chemical cues.
Videos were analysed using EthoVision XT 15 (Noldus Information Technology, Netherlands) to generate a tracking line for each individual fish. To measure the vertical exploratory activity, the tank was divided into two equally sized areas (top and bottom), as shown in Fig. 5D. Specific parameters were calculated based on the overall 6 min recording: mean velocity (cm/s)—overall swimming distance divided by recording time; latency to the top zone (s)—time interval between the beginning of the trial until the first entry in the top zone; percentage (%) of time spent in the bottom zone—amount of time in the bottom zone per each consecutive 60 s, and the number of top zone entries.
Data analysis
The effects of noise treatment on auditory thresholds were tested with Repeated Measures Analysis of Variance (RM ANOVA). The noise treatment was considered a between-subject factor, while the different frequencies were the repeated measures (within-subject factor). The impact of noise treatment on the peak latency was tested with one-way ANOVA. The effect of the acoustic treatment on hair cell number and Ribeye b punctua were tested with two-way ANOVA, with treatment and epithelial region as factors.
Regarding the behavioural tests, the effect of noise treatment on the time spent in the bottom zone was tested with RM ANOVA, considering noise treatment as a between subject factor, and different time points of the trial as repeated measures. Furthermore, time spent in the bottom (per min), the mean velocity, latency to the top zone, percentage (%) of time spent in the bottom zone (per 60 s), and the number of entries in the top zone were compared between treatments using one-way ANOVA.
ANOVAs were followed by Tukey’s pairwise comparison post hoc test. All assumptions for parametric analyses were confirmed through the inspection of normal probability plots and by performing the Levene’s test for homogeneity of variances. Graphs and statistical tests were performed using Prism 7 for Mac OS X (GraphPad, USA) and SPSS v26 (IBM Corp. Armonk, USA).
Source: Ecology - nature.com
