Methods

Animals, Housing and Ethics Statement

12 male Mongolian gerbils, purchased from Janvier (Le Genest-Saint-Isle, France) were housed in standard animal racks (Bio A.S. Vent Light, Zoonlab, Emmendingen, Germany) in groups of 3 to 4 animals with free access to water and food at a room temperature of 20 to 25C under a 12h/12h dark/light circle. At the beginning of the experiments, the animals were 12 weeks old and had been at least two weeks in the animal facility for habituation. The use and care of the animals was approved by the state of Bavaria (Regierungspräsidium Mittelfranken, Ansbach, Germany, No. 54-2532.1-02/13).

Experimental protocol

The animals were investigated twice with the behavioral paradigm of the gap prepulse inhibition of the acoustic startle response (GPIAS) to evaluate possible behavioral indications of a tinnitus percept and with auditory brainstem response (ABR) audiometry under anesthesia to obtain the hearing thresholds. The respective first measurements were performed five days before a monaural acoustic noise trauma (2 kHz, 115 dB SPL, 75 min; anesthetized) or sham trauma (2 kHz, 65 dB SPL, 75 min; anesthetized). The second ABR measurements were obtained 4 days after that trauma, and GPIAS was obtained immediately before sacrificing the animals (CO2 euthanasia) 13 days after the trauma. The brains of the sacrificed animals were extracted, cryo-sliced, immunostained and immunohistologically evaluated.

Behavioral measurements

For the GIPAS measurements for tinnitus assessment, a custom-made open-source setup was used as detailed in (Gerum et al. , 2019). In short, the animals were placed in an acrylic glass restrainer tube, closed with a wire mesh at the front side and a cap at the back end, and placed on a sensor platform fixed to a vibration-damped table. Movements of the sensor platform were registered using a 3D acceleration sensor. Two loudspeakers were placed at a distance of 10 cm in front of the animal. One presenting the 115 dB SPL startle stimulus (Neo 25 S, SinusLive, noise burst 20ms, flattened with 5 ms sin2ramps) and the other the 60 dB SPL spectral noise background (CantonPlus XS.2). Spectral noise was centered at 1 to 16 kHz in octave steps with one octave bandwidth, with and without a gap of silence of 50 ms (flanked by 20 ms sin2 ramps, 10ms complete silence) starting 100 ms before the startle stimulus.
During the two measurements, animals were given 15 min of habituation in darkness in the tube. Prior to the actual measurement, five habituation stimuli were presented to “level” the startle responses. Each stimulus was repeated 30 times (15 with and 15 without gap), summing up to 120 stimuli, which took roughly 30 min.
The complete evaluation of the GPIAS measurements were performed using custom-made Python programs. The GPIAS effect was quantified by calculating the median from the full combinatorial startle amplitudes as a response to gap and no gap pre-stimulus (for details see Schillinget al. , 2017). Statistics on the mean GPIAS results were performed with Statistica 14 (TIBCO Software GmbH, Munich, Germany; cf. below).

Brainstem audiometry

For the frequency-specific ABR measurements the animals were anesthetized with a ketamine-xylocaine solution (ketamine 500 mg/kg, xylazine 25 mg/kg). The animals were placed on a remote-controlled heating pad set to 37°C. Individual audiograms of both ears were obtained for stimulation frequencies between 1 and 16 kHz in octave steps for stimulation intensities ranging from 0 to 90 dB SPL in 5 dB steps (6 ms duration with 2 ms sine-square ramps). For each ear, stimulus, and intensity, 300 repetitions were presented. The complete measurement of one ear took around 30 min. Three silver wires were used as electrodes and were placed subcutaneously retroaural above the bulla of the tested ear (recording electrode), central between both ears (reference electrode) and at the basis of the tail (ground electrode). The signal was recorded differentially between recording and reference electrode and filtered (bandpass filter 400 to 2000 Hz) via a Neuroamp 401 amplifier (JHM, Mainaschaff, Germany).

Monaural acoustic noise and sham trauma

Animals were put under deep ketamine-xylocaine anesthesia (cf. above). One ear (pseudorandomly selected) was plugged with foam (3M™ earplugs 1110, 3M, Neuss, Germany) adding at least 20 dB attenuation in the given frequency range (Stuermer & Scheich, 2000). The animals were placed on a remote-controlled heating pad set to 37°C with the non-plugged ear towards a speaker (CantonPlus XS 2). The acoustic noise trauma (2 kHz, 115 dB SPL, 75 min) or sham trauma (2 kHz, 65 dB SPL, 75 min) was only performed on that ear.

Brain extraction, cryo-slicing and immunofluorescence staining

The sacrificed animals were decapitated and the brains were extracted carefully. They were placed in a 2% paraformaldehyde / 2% sucrose solution on a rocking shaker for 8 hours at 4°C. The tissue was then washed with phosphate-buffered saline (PBS) three times over one minute each, and then transferred into a 10% sucrose solution on a rocking shaker for 8 hours at 4°C. The incubation was finalized with 15% sucrose solution and finally 30% sucrose solution on a rocking shaker for 8 hours each at 4°C, washing was performed with PBS two times for one minute each. After drying, the tissue was shock-frosted in liquid nitrogen at ‑196°C and cur on a cryotome in 18 µm thick slices, which were transferred onto object slides and air-dried for 30 min at ambient temperature. Until staining, the slides were stored at -20°C in a fridge,
After defrosting and drying, every 10th object slide was used and fixated with -20°C cold methanol for three minutes and dried for 10 minutes. After that the slices were permeabilized with 2 % horse-serum in 0.1 % Triton X-100 in PBS solution over 2.5 hours. Then the incubation with Wisteria Floribunda Lectin-Fluorescein-5-isothiocyanate (WFA-FITC) started, with WFA-FITC 1:200 in 0.5 % horse-serum/PBS over 5 hours at room temperature. It was stopped with washing with PBS three times over five minutes each. Finally, the slides were mounted with IS Mounting-Medium PI.

Microscopy and densitometry

Immunostained tissue slices were investigated with a BZ 9000 Keyence fluorescence microscope (4x magnification) and single pictures were combined to a full slice picture with the BZ-II software. We used a 1360x1024 pixel resolution, green fluorescence, exposure adjustments: shadow 11, highlight 165, gamma 0.7, exposure time 1/1.1 sec.
The combined slice-pictures were evaluated with ImageJ Fiji (version 1.53) and only the slices containing the primary auditory cortex were evaluated. For marker-intensity reference, always the brainstem of the same slide was used. The appropriate slices and cortical / subcortical areas were identified using anatomical markers based on a Mongolian gerbil brain atlas (plates 28-33 in Radtke-Schuller et al. , 2016). The areas of interest ipsi and contralateral to the trauma ear were marked and their luminance was calculated using the inbuilt program function. The luminance values of cortical areas were divided by the reference brainstem area values (luminance ratio). Examples of the brain slices of animals with trauma and behavioral signs of tinnitus, with trauma but without behavioral signs of tinnitus and sham trauma without tinnitus are depicted in Figure 1 .