cVEMP correlated with imbalance in a mouse model of vestibular disorder

Male ICR mice purchased from Japan SLC (Hamamatsu, Japan) were randomly bred and used for experiments at 4–5 weeks of age. The mice were maintained under specific pathogen-free (SPF), constant temperature (23 ± 2 °C), and 12-h light/dark cycle conditions.

The method described in a previous report was used [18]. For the IDPN group, mice at 4 weeks of age were orally administered IDPN (Wako) in saline (28 mmol/kg body weight) once with a sonde (FTP-22-25, Prime Tech Ltd.). For the control group, mice at the same age were administered only saline in the same way. All behavioral tests and cVEMP testing were performed before and 3–5 days after IDPN administration.

The method reported previously was used for the rotarod test [19]. Mice (4 weeks of age) were placed on a stable rod of 5.7 cm in diameter (47600-Mouse Rota-Rod, UGO BASILE, Italy), and then the rod was rotated at a constant speed of 10 rpm/min for 10–20 s as training. The mice were then placed on the rod before and 3 days after administration of IDPN, and the rod was rotated at a constant speed of 25 rpm/min up to 300 s. Each trial was aborted if the mouse fell from the rod, and the latency until slipping on the rotarod was recorded. For the beam crossing test, mice were forced to walk on an elevated round wooden bar of 15 mm in diameter, and the time taken to traverse the bar of 30 cm in length was recorded twice up to 10 s. Before the test, mice walked on a round wooden bar of 22 mm in diameter once for training. Triplicated measurements were performed with 5-min intervals in each trial for the rotarod and beam crossing tests. For the air-righting reflex test, the method reported previously was used [20]. This test was performed after finishing the other behavior tests. A soft sponge was used to relieve the dropping impact. All animals were set in a supine position at 0.50 m in height and then dropped. The time it took for each animal to assume a prone position with all four legs in down in the air was determined. Each animal’s behavior was recorded with a high-speed camera (EXILIM EX-10, CASIO) three times.

We performed perfusion fixation with 4% paraformaldehyde (PFA) for mice in the control and exposure groups 1–2 weeks after IDPN administration. After collection of inner ears, we further immersed the inner ears in EDTA for 2 days at 4 °C for decalcification. Paraffin sections of inner ears of 3 μm in thickness were prepared. Immunostaining with a polyclonal antibody against Myosin VIIa (1:100, Proteus BioScience) was performed after antigen retrieval had been performed with a TE buffer (pH 9.0) at 90–92 °C for 10 min. Alexa Fluor 594-labeled donkey anti-rabbit IgG (1:1000, Invitrogen) was used as a secondary antibody followed by counterstaining with 4′,6-diamidino-2-phenylindole (DAPI). The stained specimens were observed under a fluorescent microscope with phase contrast (Leica DM6000B). A researcher who was blinded to the results of cVEMP testing counted the numbers of hair cells in vestibular sensory epithelia. Hair cells were defined by the expression of Myosin VIIa in the cytoplasm and nuclear exclusion of the staining [21, 22].

After treatment with IDPN, perfusion fixation with 4% paraformaldehyde (PFA) was performed. The utricles were isolated after dissection of inner ears. Hair bundles were labeled with fluorescein-phalloidin (Wako, 068–06261) diluted with PBS (1:200). The stained specimens were observed under a fluorescent microscope with phase contrast (Leica DM6000B).

cVEMP was recorded in mice by using methods established in previous studies [10, 12, 23]. Mice were first anesthetized under inhalation of isoflurane (Wako) to set platinum-needle electrodes in the neck extensors and a reference electrode in the occipital area. When we set up the measurement system for cVEMP, we pathologically confirmed the exact site of the neck extensor muscles after opening the targeting area in order to accurately place the needle electrodes in the neck extensor muscles. In addition, we always placed the electrodes at the same sites with reference to a position of the spinal cord at the level of C3. Electromyography (EMG) potentials were monitored to check whether mice were awake. After awaking, recording of cVEMP in EMG was started with an auditory brainstem response (ABR) system (Tucker-Davis Technologies, FL, USA) and PowerLab (AD Instruments Pty. Ltd) by stimulation with sound at 1000 Hz, SPL of 90 dB, rise-fall of 0.2 ms, interval of 200 ms, and 10-ms flat that was generated by a sound generator system (Sound Stimulator DPS-725, Dia Medical System CO., LTD, Japan). The sound stimulation was given by a speaker to the right ear of each mouse at a distance of 20 cm from the speaker. The onset was 0 ms After stimuli 200 times, integrated EMG showed similar potentials in the control group and IDPN group. During VEMP recording, the animal’s neck was stretched by lifting up the inhalation mask with polypropylene string to induce tonus of the sternocleidomastoid muscle. The inhalation mask was set at 2 cm in height from the ground (Additional file 1: Figure S1). All recordings were performed under wakefulness. We followed the definition of “normal” cVEMP in previous reports [12, 23, 24], which was the presence of a positive (p1)-negative (n1) waveform (i) at a 6- to 9-ms latency (ii) with a p1-n1 amplitude of 5 to 20 μV and (iii) the presence of a preceding negative wave just before cVEMP. In addition to this definition, we regarded the waveform as cVEMP when it appeared under the condition of (iv) stimulation of air-conducted sound at 1000 Hz and 90 dB (v) with the tension of the sternocleidomastoid muscle (SCM) by head tilt at 2 cm in height from the ground. We measured the background amplitudes under the condition of sound stimulation but not head tilt. We used the amplitudes “without head tilt” at the same latencies as those of cVEMPs obtained under the conditions of sound stimulation and head tilt. We normalized cVEMP amplitudes by subtracting the background amplitudes of the electromyography (EMG) potentials from the cVEMP amplitudes of cVEMP. We used the normalized cVEMP amplitudes for Fig. 3d and left graphs of Figs. 4 and 5. The results of comparisons of the control and IDPN groups (Fig. 3) and correlations (Figs. 4 and 5) were similar to the results without normalization.

Two-tailed Student’s t test was used to determine a significant difference between the two groups for parametric data [25]. Welch’s t test was also used to determine a significant difference between the two groups for parametric data with unequal variance. Spearman’s rank correlation coefficient and Spearman’s rank correlation coefficient were used to determine a significant correlation between nonparametric variables. A difference with p < 0.05 was considered to indicate statistical significance. All statistical analyses were performed using JMP Pro (version 13.0.0; SAS Institute Inc., Cary, NC, USA).

Rotarod, beam crossing, and air-righting reflex tests were performed before and 3–4 days after oral exposure one time to IDPN at 28 mmol/kg in order to determine balance functions in the control group and IDPN group. Before exposure to IDPN, the two groups showed comparable scores in the rotarod, beam crossing, and air-righting reflex tests (Fig. 1b–d). Latency until slipping on the rotarod in the IDPN group was significantly decreased compared to that in the control group (Fig. 1b). The time taken to traverse the beam in the IDPN group was significantly increased compared to that in the control group (Fig. 1c). Latency until air-righting reflex in the IDPN group was significantly increased compared to that in the control group (Fig. 1d). Next, immunohistostaining of vestibular hair cells from mice after oral exposure to IDPN for 1–2 weeks was performed to morphologically determine the influence of IDPN on vestibular hair cells. The numbers of hair cells in the saccule, utricle, and cupula from mice in the IDPN group were significantly decreased compared to those in the control group (Fig. 2). In phalloidin staining of vestibular hair cells from mice after oral exposure to IDPN, the numbers of hair bundles in the saccule and utricle from mice in the IDPN group were significantly decreased compared to those in the control group (Additional file 1: Figure S2).

At 4 days after the oral administration of IDPN, cVEMP measurements were performed for mice in the control and IDPN groups (Fig. 3). cVEMP waves in the control group (upper) and the IDPN group (lower) are presented at the same scale. The waveforms of cVEMPs were identified as cVEMPs with a positive peak (p1, indicated by black triangles) and those with a negative peak (n1, indicated by white triangles) by the appearance of preceding negative waves (indicated by gray arrows) just before p1 indicated by black triangles. cVEMPs in the two groups were comparable before exposure to IDPN (Fig. 3b). After administration of IDPN, the IDPN group showed decreased cVEMP amplitude and prolonged cVEMP latency compared to those in the control group (Fig. 3c). cVEMP amplitudes and latencies were compared in the control group and IDPN group (Fig. 3d, e). The amplitude in the IDPN group was significantly smaller than that in the control group, while the two groups showed comparable amplitudes before administration of IDPN (Fig. 3b–d). The p1 latencies in the IDPN group were significantly prolonged compared to those in the control group (Fig. 3e). We also determined the amplitudes of the preceding negative waves in the control and IDPN groups (Additional file 1: Figure S3). The amplitudes of the preceding negative waves were not significantly different in the two groups (Additional file 1: Figure S3A), while the amplitudes of the p1-n1 waveforms showed a significant difference in the two groups (Additional file 1: Figure S3B). ABR was also measured to confirm the cVEMP waveforms in mice. Typical waveforms of ABR at 20–80 dB of 4 kHz sound were detected in the control group but not in the IDPN group (Additional file 1: Figure S4) as previously reported [26, 27].

The correlations between scores of behavior tests and amplitude or latency of cVEMP were determined (Fig. 4). There were significant correlations of time to traverse on the beam with cVEMP amplitude (r = − 0.8011, p < 0.0001) and latency (r = 0.3808, p = 0.0379) (Fig. 4b), and there were significant correlations of air-righting reflex with cVEMP amplitude (r = − 0.42806, p = 0.0065) and latency (r = 0.3844, p = 0.0360) (Fig. 4c). On the other hand, only the cVEMP amplitude had a significant correlation with the rotarod test score (r = 0.6239, p = 0.0002) (Fig. 4a).

Finally, correlations of the numbers of hair cells in the saccule, utricle, and cupula with the amplitude or latency of cVEMP were examined (Fig. 5) because associations of the utricle and cupula with cVEMP were previously reported [28, 29]. The number of hair cells in the saccule was significantly correlated with cVEMP amplitude (r = 0.5153, p = 0.0343) but not with cVEMP latency (r = 0.4052, p = 0.1066) (Fig. 5a), and there were significant correlations of the number of hair cells in the utricle with both cVEMP amplitude (r = 0.6191, p = 0.0081) and latency (r = 0.6039, p = 0.0103) (Fig. 5b). The number of hair cells in the cupula was significantly correlated with cVEMP amplitude (r = 0.6242, p = 0.0301) but not with cVEMP latency (r = 0.5232, p = 0.0809) (Fig. 5c).