Molecular composition and function of the spiral ganglion neuron peripheral synapse in mice Reijntjes, Daniël Onne Jilt
DOI:
10.33612/diss.93524048
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Reijntjes, D. O. J. (2019). Molecular composition and function of the spiral ganglion neuron peripheral synapse in mice. University of Groningen. https://doi.org/10.33612/diss.93524048
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Download date: 28-06-2021
Chapter 4
Acquired hearing loss in sodium-activated potassium
channel knockout mice
Reijntjes, D.O.J., Schubert, N.M.A., van Tuinen, M., Pyott, S.J.
81
Abstract
The spiral ganglion neurons relay all acoustic information to the brain with high temporal fidelity. The specific ion channel composition of the spiral ganglion neu- rons allows the spiral ganglion neurons to activate with high temporal precision. In particular, potassium channels are important for shaping both the passive and ac- tive membrane properties that allow high temporal fidelity. Recently, we described two novel potassium channels (KN a1.1and KN a1.2) in the spiral ganglion neurons whose activity is dependent on voltage changes and intracellular sodium concentra- tions. Genetic deletion of the genes encoding for these ion channels leads to a reduc- tion in ABR wave I amplitude slopes in vivo in KN a1 DKO mice and shorter spike latency and smaller amplitude in vitro without loss of the spiral ganglion neurons.
As a reduction in wave I amplitude is often correlated to synaptopathy, we investi- gated whether the reduction in ABR wave I amplitude in young adult mice would lead to exacerbated auditory defects with increasing age and after noise exposure.
In this study we show that KN a1DKO mice have increased vulnerability to auditory threshold shifts and synapse loss with ageing and after noise exposure. Specifically, KN a1DKO mice show increased wave I threshold shifts at 12 weeks of age, wave I latency slopes and synapse loss at 24 weeks, and OHC loss at 36 weeks, compared to WT mice. Furthermore, KN a1DKO mice show larger wave I threshold shifts and more synapse loss after noise exposure than WT mice. These results suggest that ion channel dysfunction in the spiral ganglion neurons can cause an uncommon type of auditory pathology that has not been well studied and thus warrants further inves- tigation.
4.1. Introduction
The auditory nerve is made up of spiral ganglion neurons (SGNs) that form synap- tic connections to the inner hair cells (Spoendlin, 1985). The SGNs initiate the fir- ing of action potentials when the inner hair cells release glutamate in response to auditory stimulation. The released glutamate then activates α-amino-3-hydroxy-5- methyl-4-isoxazolepropionic acid (AMPA)-type glutamate receptors on the postsynap- tic SGN dendrite (Glowatzki and Fuchs, 2002). To accurately convey auditory stimuli, the SGNs have to fire both rapidly and faithfully in response to the released glutamate (Meyer and Moser, 2010). The spiral ganglion neurons are able to fulfil these demand- ing tasks, likely because of specific ion channels present in their synaptic terminals.
Especially voltage gated K+channels are thought to play a major role in establishing the required active and passive membrane properties, such as the membrane poten- tial level, as well as action potential thresholds and firing properties (Oak and Yi,
4.2. Materials & Methods 83
2014; Reijntjes and Pyott, 2016; Rusznák and Szucs, 2009). Specifically, a subset of sodium dependent voltage gated K+ channels (KNa channels), is increasingly con- sidered to contribute in shaping these responses.
Two KN achannels have been identified (Kaczmarek, 2013), KN a1.1(SLO2.2/Slack), and KN a1.2(SLO2.1/Slick). These ion channels are co-activated by an increase in in- tracellular Na+concentration and membrane depolarization. Examination of KN a1.1 and KN a1.2double knockout mice at 6 weeks of age (KN a1DKO) indicate that KN a1.1 and KN a1.2are expressed in SGNs and, furthermore, that genetic deletion of these channels leads to a reduction in auditory brainstem response (ABR) wave I ampli- tude slopes in vivo without a loss of hair cells. In addition, genetic deletion of these channels leads to a reduced stimulation threshold and accelerated spiking of cultured SGNs in vitro (Reijntjes et al., 2019). Therefore, it is likely that KN a1.1, and KN a1.2 contribute toward a reduction in excitability of the SGNs both in vivo and in vitro.
It is currently unclear whether loss of KN a1.1, and KN a1.2, and therefore a pos- sible increase in SGN excitability, contributes to an increased vulnerability of the SGNs to pathology. A reduction in ABR wave I amplitude is correlated to a reduction in the number of synaptic contacts between SGNs, and inner hair cells, as seen in a specific type of pathology called synaptopathy (Liberman and Kujawa, 2017). Synap- topathy develops with either increasing age or after noise exposure. The reduction in ABR wave I amplitude at six weeks of age suggests that in KN a1DKO mice might exhibit synaptopathy at younger ages, without noise exposure. However, this previ- ous study did not find a reduction in the number of synaptic contacts in KO mice for KN a1.1, and KN a1.2at six weeks of age (Reijntjes et al., 2019). Therefore, the aim of this study was to follow up on these findings and investigate the effects of ageing and noise exposure on auditory responses and auditory nerve fiber survival in KN a1 DKO mice.
In order to examine a possible increase in vulnerability of the synaptic contacts between SGNs and inner hair cells in KN a1 DKO mice, we performed two sets of experiments. In the first set of experiments, we examined auditory function, and synapse survival in aged cohorts of mice (6 weeks, 12 weeks, 24 weeks, 36 weeks). In the second set of experiments, we examined auditory function and synapse survival 24 h and also 168 h after either a 100 dB SPL or a 110 dB SPL noise exposure. In this study we found that KN a1DKO mice have accelerated loss of synaptic terminals both with age and after noise exposure. However, these effects did not indicate that the reduction in ABR wave I amplitude results in a rapid progressive loss of synaptic terminals. These findings further suggest that cochlear synaptopathies can involve both functional and structural loss of the synapses between the IHCs and SGNs.
4.2. Materials & Methods
4.2.1. Animals
KN a1DKO mice (Reijntjes et al., 2019) were acquired from the breeding facilities at the University Medical Centre Groningen, the Netherlands. C57BL/6 mice from the same institute were used as control mice for all experiments. Hereafter, these C57BL/6 mice will be referred to as WT mice. Mice were a mix of both genders. In our ageing cohort, mice were tested for auditory function at the age of either 6, 12, 24, or 36 weeks, followed by termination for tissue assessment. In our noise exposure cohort, mice were tested for a baseline auditory function at 5 weeks of age. They were subsequently exposed to either moderate or severe noise at 6 weeks, and follow- up auditory function measurements were performed 24 h and 168 h (7 days) after noise exposure. Each group in the ageing and noise exposure cohorts consisted of 6-9 mice. All animal experiments were approved and conducted in accordance with Dutch animal welfare laws.
4.2.2. Auditory brainstem responses
Auditory brainstem responses (ABRs) were recorded as described in (Reijntjes et al., 2019). Briefly, mice were anesthetized with an intraperitoneal injection of 75 mg/kg ketamine and 1 mg/kg dexmedetomidine. ABRs were recorded following free field Click (25 µs), and pure tone (8, 16, and 32 kHz, 5000 µs) stimulation. Wave I thresholds, latencies, and amplitudes were manually assessed blinded to genotype and experimental condition. Wave I amplitude and latency input-output regression slopes were calculated for stimuli intensities from 40-90 dB (Reijntjes et al., 2019).
4.2.3. Noise exposure
Two levels of noise exposure (100/110 dB SPL) were used to compare noising regimes that cause either transient wave I threshold shifts or permanent ABR wave I threshold shifts. For noise exposure with a 100 dB SPL, mice (one WT and one KN a1 DKO) were placed in a two-compartment mesh cage inside an acoustic chamber and exposed to a 8-16 kHz noise band at a 100 dB SPL for 120 min using a free field speaker positioned 10 cm from the mesh cage. For noise exposure at 110 dB SPL for 120 min, mice were anesthetized using the same procedure described for ABR recordings, with the injection of additional sedative every 30 min. Anesthetized mice were positioned 10 cm in front of the free field speaker equidistant to the mesh cages used for the unsedated mice. Previous experiments indicated no difference between unsedated and sedated mice in terms of auditory pathology for 100 dB SPL noise exposures (Reijntjes et al., 2018).
4.2. Materials & Methods 85
4.2.4. Inner ear dissection and immunohistochemistry
To dissect the organ of Corti, mice were either anesthetized with 4% isofluorane gas, or with intraperitoneal injection of 75 mg/kg ketamine and 1 mg/kg dexmedeto- midine, and subsequently decapitated and the skull bisected. Inner ears were mi- crodissected in PBS. A small hole was made over the apical turn of the cochlea and inner ears were subsequently placed in a 4% PFA solution for 1 h. Organs of Corti were then further dissected in PBS and placed in blocking buffer for at least 1 h at room temperature. To visualize pre- and postsynaptic structures of the spiral gan- glion neuron - inner hair cell synapses, organs of Corti were subsequently fluores- cently labeled for C-terminal-binding protein 2 (CTBP2) and AMPA-type glutamate receptor 2/3 (GluA2/3) as in (Braude et al., 2015) (CTBP2; mouse monoclonal IgG1, BD Biosciences; cat. no. 612044, at a concentration of 0.83µg/ml (1:300), GluA2/3;
rabbit polyclonal, Milipore AB1506, at a concentration of 0.33 µg/ml (1:300)).
4.2.5. Image acquisition and analysis
To determine specific cochlear frequency regions of the isolated organs of Corti, low magnification images of the isolated organs of Corti were obtained using a fluores- cent microscope (Leica DM4000, with a 5× dry, NA 0.15 objective, and Leica Applica- tion Suite 4.3 software). The obtained images were used to create cochlear frequency maps using a freely available ImageJ plugin (http://www.masseyeandear.org/research /ent/eaton-peabody/epl-histology-resources) and a published place-frequency map (Müller et al., 2005). Subsequently, high magnification z-stacks of three cochlear frequency regions (8, 16, and 32 kHz) were obtained using a Leica SP8 confocal sys- tem (Leica Microsystems CMS GmbH with a 63× oil, NA 1.4, objective).
To quantify the number of synapses per inner hair cell, CTBP2 and GluA2/3 puncta were detected using the “spots” function in Imaris 6.4 software (Bitplane).
Only paired combinations of CTBP2, and GluA2/3 were considered to form a func- tional synapse. The total number of synapses counted in each z-stack was then di- vided by the total number of inner hair cells.
4.2.6. Statistics
All values are presented as the mean ± standard error of the mean. Data were analyzed using (M)ANOVA analysis with subsequent post-hoc tests whenever possi- ble. For some data, the assumptions of normality were violated substantially and could not be corrected for by transforming the data. In these cases, non-parametric Kruskal-Wallis with subsequent Mann-Whitney-U tests, were applied. Bonferroni corrections were applied for multiple comparisons for all comparisons for a given fre- qeuncy (clicks, 8kHz, 16 kHz, 32kHz) in the ageing experiments. Bonferroni correc-
tions were applied for multiple comparisons for all comparisons for a given exposure paradigm (Baseline, 100dB 24 h, 100 dB 168 h, 110dB 24 h, 110 dB 168 h) in the noising experiments. All analyses were performed in R (R Core Team, 2018).
4.3. Results
4.3.1. ABR wave I threshold responses in ageing KN a1DKO mice
To compare the deterioration of peripheral auditory function with age between KN a1DKO and WT mice, ABR wave I thresholds were assessed for both genotypes for all age groups (6, 12, 24, and 36 weeks of age). For both genotypes, ABR wave I thresholds increased with age across all frequencies. ABR wave I thresholds were not different between KN a1DKO and WT mice for click stimuli over all ages (Figure 4.1A).
Wave I thresholds were significantly higher in KN a1DKO mice than in WT mice for 8 kHz pure tones at 24 weeks (Mann-Whitney U-test, p < 0.05) but no differences were observed for the other age groups (Figure 4.1B). Wave I thresholds were not different between KN a1DKO and WT mice for 16 kHz stimuli over all ages (Figure 4.1C). Wave I thresholds were significantly higher in KN a1DKO mice than in WT mice for 32 kHz pure tones at 12 weeks (Mann-Whitney U-test, p < 0.01) but not different for the other age groups (Figure 4.1D). Thus, KN a1DKO mice exhibit an accelerated increase in ABR wave I thresholds compared to WT mice for specific frequencies at specific ages.
4.3.2. Outer hair cell survival in ageing mice
To investigate whether the accelerated increase in absolute thresholds in KN a1 DKO mice was correlated to an accelerated increase in the loss of OHCs, we com- pared OHC counts between KN a1DKO and WT mice. We quantified the number of OHCs in a 100 µm section at specific cochlear frequencies and compared these num- bers between genotypes over all age groups. Example images are shown at the 16 khz region for 6 and 36 week old animals (Figure 4.2A-D). For both genotypes, OHC numbers remained relatively stable up to 36 weeks of age for all frequencies (Figure 4.2E-G). At 36 weeks of age, severe loss of OHCs was observed in the 32 kHz region for both KN a1DKO and WT mice. In all but one KN a1DKO mouse, no further OHCs could be observed at and beyond 32kHz (Figure 4.2C). There were no significant dif- ferences between KN a1DKO and WT mice for the 6 week old, 12 week old, and 24 week old groups over all frequencies. In the 36 week old group, we found that KN a1 DKO OHC numbers were significantly higher than in WT mice but this difference was due to the one mouse with some remaining OHCs (three way-ANOVA with post- hoc test, p < 0.001). Thus, OHC numbers were not different between KN a1DKO, and WT animals until at least 36 weeks, which is at a later age than the observed thresh- old differences and thus, OHC loss seems to appear after increases in ABR wave I threshold in KN a1DKO mice with age.
4.3. Results 87
Figure 4.1. Auditory brainstem response wave I thresholds for KN a1DKO and WT mice at different ages. Peak I auditory brainstem response thresholds were recorded for KN a1DKO and WT mice in response to either click stimuli, or to 8 kHz, 16 kHz, and 32 kHz tones. Stimulus intensity was increased stepwise from 20 dB SPL to 90 dB SPL in 5 dB increments. Thresholds were marked at the first intensity where a clear peak was seen above the noise floor, and the peak could be continuously traced at higher intensities. Thresholds over age are presented in response to Click stimuli (A), in response to 8 kHz tones (B), in response to 16 kHz tones (C), and in response to 32 kHz tones (D). Data was compared between KN a1DKO and WT mice using three-way ANOVA with follow-up post hoc analysis. Normality assumptions were violated and therefore the comparisons were redone using nonparametric statistics.
Results were no different between both methods of analysis, non-parametric results are indicated in the Figure panels. Significance is indicated by: * = p < 0.05, ** = p < 0.01.
4.3.3. ABR wave I amplitude slopes in ageing KN a1DKO mice
A previous study (Reijntjes et al., 2019) found that wave I amplitude slopes were reduced in KN a1DKO compared to WT mice at six weeks of age. To assess whether the wave I amplitude slopes deteriorate even further with age in KN a1DKO mice, we assessed wave I amplitude slopes between KN a1DKO and WT mice over the dif- ferent age groups. Similar to our previous findings, ABR wave I amplitude slopes were smaller in KN a1 DKO mice than in WT mice. For both genotypes, ABR wave I amplitude slopes decreased with age. For click stimuli, the amplitude slopes were smaller (three-way ANOVA, p at least < 0.05) for the KN a1DKO compared to the WT mice in all age groups except at 36 weeks of age (Figure 4.3A).
Figure 4.2. Outer hair cell counts per cochlear frequency region for both KN a1DKO and WT mice.Organs of Corti were harvested from mice at four different time points, 6 weeks, 12 weeks, 24 weeks, and 36 weeks of age. Subsequently, the outer hair cells were labelled with antibodies against either Myosin 7A or Prestin and mounted. Cochlear frequency was determined by creating a frequency map using low magnification images. Confocal images were then created for three specific frequency regions (8 kHz, 16 kHz, and 32 kHz). Outer hair cells were counted in Imaris using either Myosin 7A or Prestin labeling.
Outer hair cell numbers were normalized to 100 µm sections for comparison. Example prestin labelling at 16 kHz is shown for 6 week old WT mice (A), 6 week old KN a1DKO mice (B), 24 week old WT mice (C), and 24 week old KN a1DKO mice (D). Outer hair cell numbers were calculated for 100 µm sections for KN a1DKO and WT mice at 8 kHz (E), 16 kHz (F), and 32 kHz (G). Data were compared using three-way ANOVA with follow-up post hoc analysis. Significance is indicated by: *** = p < 0.001.
For 8kHz pure tone stimuli, the amplitude slopes were smaller (three-way ANOVA, p at least < 0.001) for the KN a1DKO compared to the WT mice at 6 weeks of age but not for the other age groups (Figure 4.3B).
4.3. Results 89
Figure 4.3. Auditory brainstem response wave I amplitude slopes for KN a1DKO and WT mice at different ages.Auditory brainstem responses were recorded at four ages, 6 weeks, 12 weeks, 24 weeks, and 36 weeks of age. Amplitude-intensity graphs were created for peak I amplitudes obtained for stimuli between 40 dB SPL and 90 dB SPL. Slopes were then calculated from fitted regression lines and compared between KN a1DKO and WT mice over four stimuli, and four age groups. Regression lines were only fitted if corresponding thresholds were lower than 85 dB SPL. Auditory brainstem response peak I amplitude slopes over four ages are presented for Click stimuli (A), 8 kHz tones (B), 16 kHz tones (C), and 32 kHz tones (D). Inset in (D) shows example amplitude-intensity ABR peak I curves at 16 kHz for different ages.
Data were compared using three-way ANOVA with follow-up post hoc analysis. Significance is indicated by: * = p < 0.05, *** = p < 0.001.
For 16kHz pure tone stimuli, the amplitude slopes were smaller (three-way ANOVA, p at least < 0.001) for the KN a1DKO compared to the WT mice at 6, and 12 weeks of age but not for the other age groups (Figure 4.3C). For 32kHz pure tone stimuli, the amplitude slopes were smaller (three-way ANOVA, p at least < 0.05) for the KN a1 DKO compared to the WT mice at 6 weeks of age but not for the other age groups (Figure 4.3D). Thus, with increasing age, the reduction in wave I amplitude slope is larger for the WT mice than for KN a1DKO mice, resulting in similar wave I amplitude slopes at 36 weeks of age. These results suggest that the deterioration in wave I amplitude slope for the KN a1DKO mice does not persist but instead levels off with increasing age.
4.3.4. ABR wave I latency slopes in ageing KN a1DKO mice
Our previous study (Reijntjes et al., 2019) found that in vitro action potential latency was reduced in spiral ganglion neurons cultured from KN a1DKO mice com- pared to spiral ganglion neurons cultured from WT mice. However, ABR wave I la- tency slopes were found to be similar in KN a1DKO compared to WT mice at 6 weeks in vivo(Reijntjes et al., 2019). To examine whether the wave I latency slopes remain similar between KN a1 DKO and WT mice with increasing age, we compared wave I latency between the two genotypes over all age groups. For both genotypes, ABR wave I latency slopes became larger with age. For click stimuli, the latency slopes were larger (three-way ANOVA, p at least < 0.001) for the KN a1 DKO compared to the WT mice at 24 and 36 weeks of age (Figure 4.4A) but not different at the other age groups. For 8kHz pure tone stimuli, the latency slopes were larger (three-way ANOVA, p at least < 0.01) for the KN a1DKO compared to the WT mice at 24 weeks of age but not for the other age groups (Figure 4.4B). For 16kHz pure tone stimuli, the latency slopes were larger (three-way ANOVA, p at least < 0.01) for the KN a1DKO compared to the WT mice at 24, and 36 weeks of age but not for the other age groups (Figure 4.4C). For 32kHz pure tone stimuli, latency slopes were not different between KN a1DKO and WT mice (three-way ANOVA) for all age groups (Figure 4.4D). Thus, the latency slopes for the KN a1DKO mice increased more with age compared to WT mice for clicks, 8 kHz and 16 kHz stimuli, implying that wave I latencies became shorter (faster) in the KN a1DKO mice with age.
4.3.5. Inner hair cell-Spiral ganglion neuron synapse survival in KN a1DKO mice
A reduction in wave I amplitude slopes is correlated to a reduction of the total number of IHC-SGN synapses in ageing mice (Sergeyenko et al., 2013). In order to investigate whether the reduced amplitude slopes in KN a1DKO mice resulted from synapse loss, we quantified the number of synapses in all age groups across three different cochlear regions (8, 16, and 32 kHz). Examples of fluorescently labelled synaptic markers at 32 kHz over the different age groups indicated a loss of synap- tic markers with increasing age in both the WT (Figure 4.5A-D) and the KN a1DKO (Figure 4.5E-H) mice. At 8 kHz, there was no difference in the number of synapses (three-way ANOVA) at any age group between KN a1 DKO mice and WT mice (Fig- ure 4.5I). At 16 kHz, there was no difference in the number of synapses (three-way ANOVA) at any age group between KN a1 DKO mice and WT mice (Figure 4.5J). At 32 kHz, the number of synapses was smaller for the KN a1DKO mice compared to the WT mice at 24 weeks of age (three-way ANOVA, p at least < 0.01), but similar at the other age groups (Figure 4.5K). Thus, the KN a1DKO mice had an accelerated loss of synapses with age at 32 kHz compared to the WT mice.
4.3. Results 91
Figure 4.4. Auditory brainstem response wave I latency slopes for KN a1DKO and WT mice at different ages. Auditory brainstem responses were recorded at four ages, 6 weeks, 12 weeks, 24 weeks, and 36 weeks of age. Amplitude-intensity graphs were created for peak I amplitudes obtained for stimuli between 40 dB SPL and 90 dB SPL. Slopes were then calculated from fitted regression lines and compared between KN a1DKO and WT mice over four stimuli, and four age groups. Regression lines were only fitted if corresponding thresholds were lower than 85 dB SPL. Auditory brainstem response wave I latency slopes over four ages are presented for Click stimuli (A), 8 kHz tones (B), 16 kHz tones (C), and 32 kHz tones (D).
Data were compared using three-way ANOVA with follow-up post hoc analysis. Significance is indicated by: ** = p < 0.01, *** = p < 0.001.
4.3.6. ABR wave I threshold responses in KN a1 DKO mice after noise exposure
To further assess the vulnerability of the peripheral auditory system in KN a1DKO mice, we performed several noise exposure experiments. We investigated whether KN a1DKO mice were more vulnerable to threshold shifts following exposure to ei- ther a 100 dB SPL or a 110 dB SPL noise for 2 h. ABR wave I thresholds were examined at three time points, one week before noise exposure (baseline), 24 h after noise exposure, and 168 h (7 days) after exposure. At baseline, wave I thresholds were elevated in KN a1DKO mice compared to the WT mice for 32 kHz pure tone stimuli (MANOVA, p < 0.05) but were not different for the other stimuli (Figure 4.6A).
Figure 4.5. Inner hair cell – auditory nerve fibre synapse counts in KN a1DKO and WT mice at different ages. Organs of Corti were harvested at the end of the experiment. Subsequently, presy- naptic ribbons, and postsynaptic GluA2 glutamate receptor subunits were fluorescently labelled. Cochlear frequency was determined by creating a frequency map using low magnification images. Confocal images were then created for a region of about ten inner hair cells of three specific frequency regions (8 kHz, 16 kHz, and 32 kHz). Example images of synapse counts are shown at 32 kHz for WT mice for the different ages (A-D) with CTBP2 labelling in green and GluA2 labelling in red. Example images of synapse counts are shown at 32 kHz for KN a1DKO mice for the different ages (E-H). Functional synapses were classi- fied as synapses that contained labelling for both CTBP2 and GluA2 subunit. Functional synapses were calculated per IHC and compared between KN a1DKO and WT mice for the different frequency regions, 8 kHz (I), 16 kHz (J), and 32 kHz (J). Data were compared using three-way ANOVA with follow-up post hoc analysis. Significance is indicated by: ** = p < 0.01.
In our ageing cohort we did not observe a statistical difference for the 32 kHz ABR threshold between KN a1DKO and WT mice. However, in our ageing cohort we ap- plied non-parametric statistics instead of parametric statistics which likely reduced the power to detect these differences. ABR wave I thresholds for the KN a1DKO mice were higher (MANOVA, p < 0.001) than in the WT mice for 16 kHz stimuli but not different for the other frequencies at either 24 h and 168 h after a 100 dB SPL noise exposure (Figure 4.6B-C). ABR wave I thresholds for the KN a1DKO mice were higher (MANOVA, p at least < 0.05) than in WT mice for click stimuli, 24h after a 110 dB SPL exposure (Figure 4.6D) and for click and 8 kHz stimuli at 168 h after a 110 dB SPL exposure (Figure 4.6E).
4.3. Results 93
Figure 4.6. Auditory brainstem response wave I thresholds for KN a1DKO and WT mice after noise exposure. Auditory brainstem responses were recorded for both KN a1DKO and WT mice on three occasions. The first was one week before noise exposure (baseline), 24 hours after noise exposure, and 168 hours after noise exposure. Mice were allocated to two groups that received a different intensity noise exposure. The first group was exposed to a 100 dB SPL narrowband noise (8-16kHz) for two hours.
The second group was exposed to the same noise for the same duration at an intensity of 110 dB SPL.
Wave I auditory brainstem response thresholds were recorded for KN a1DKO and WT mice in response to either click stimuli, or to 8 kHz, 16 Khz, and 32 khz tones. Stimuli intensity was increased stepwise from 20 dB SPL to 90 dB SPL in 5 dB increments. Thresholds were marked at the first intensity where a clear peak was seen above the noise floor and the peak could be continuously traced at higher intensities.
Thresholds over frequency are presented for baseline (A), 24 hours after a 100 dB SPL noise exposure (B), at 168 hours after a 100 dB SPL noise exposure (C), 24 hours after a 110 dB SPL noise exposure (D), and at 168 hours after a 110 dB SPL noise exposure (E). Data was compared between KN a1DKO and WT mice by MANOVA using the responses at the four stimuli as dependent variables, with subsequent follow-up post hoc analysis. Significance is indicated by: * = p < 0.05, ** = p < 0.01, *** = p < 0.001.
Thus, a higher susceptibility of KN a1DKO mice to threshold elevation was observed for both noise exposures but was shifted toward lower frequency stimuli at the higher intensity exposure.
4.3.7. Wave I amplitude slopes in KN a1 DKO mice after noise exposure
To assess whether noise exposure further reduced wave I amplitude slopes in KN a1 DKO mice, we compared wave I amplitude slopes between KN a1DKO and WT mice after noise exposure. At baseline, wave I amplitude slopes were smaller (three-way ANOVA, p at least < 0.05) in KN a1 DKO than in WT mice (Figure 4.7A). Wave I amplitude slopes were smaller (three-way ANOVA, p at least < 0.01) 24h after a 100 dB SPL noise exposure in KN a1DKO than in WT mice (Figure 4.7B).
Figure 4.7. Auditory brainstem response wave I amplitude slopes for KN a1DKO and WT mice after noise exposure. Auditory brainstem responses were recorded for both KN a1DKO and WT mice on three occasions. The first was one week before noise exposure (baseline), 24 hours after noise exposure, and 168 hours after noise exposure. Mice were allocated to two groups that received a different intensity noise exposure. The first group was exposed to a 100 dB SPL narrowband noise (8-16kHz) for two hours.
The second group was exposed to the same noise for the same duration at an intensity of 110 dB SPL.
Wave I auditory brainstem response amplitudes were recorded for KN a1DKO and WT mice in response to either click stimuli, or to 8 kHz, 16 Khz, and 32 khz tones. Amplitude-intensity graphs were created for peak I amplitudes obtained for stimuli between 40 dB SPL and 90 dB SPL. Slopes were then calculated from fitted regression lines and compared between KN a1DKO and WT mice at baseline, 24 hours after, and 168 hours after both noise exposures. Regression lines were only fitted if corresponding thresholds were lower than 85 dB SPL. Auditory brainstem response peak I amplitude slopes for the four stimuli are presented for baseline (A), 24 hours after a 100 dB SPL noise exposure (B), at 168 hours after a 100 dB SPL noise exposure (C), 24 hours after a 110 dB SPL noise exposure (D), and at 168 hours after a 110 dB SPL noise exposure (E). Data was compared between KN a1DKO and WT mice by three-way ANOVA, with subsequent follow-up post hoc analysis. Significance is indicated by: * = p < 0.05, ** = p < 0.01, *** = p < 0.001.
Wave I amplitude slopes were smaller for 16 kHz stimuli (three-way ANOVA, p <
0.001) in KN a1 DKO than in WT mice but not different for the other stimuli 168 h after a 100 dB SPL noise exposure (Figure 4.7C). Wave I amplitude slopes were smaller for 16 kHz stimuli (three-way ANOVA, p < 0.001) in KN a1DKO than in WT mice but not different for the other stimuli 24 h after a 110 dB SPL noise exposure (Figure 4.7D). Wave I amplitude slopes were not different between KN a1 DKO and WT mice 168 h after a 110 dB SPL noise exposure (Figure 4.7E). Thus, ABR wave I amplitudes were reduced in both KN a1 DKO and WT mice after a 100 dB noise exposure across most frequencies to an extent where the wave I amplitudes were still lower in KN a1 DKO mice. After a 110 dB noise exposure, ABR wave I amplitude levels were shifted to equivalent thresholds for both strains across most frequencies.
4.3. Results 95
Figure 4.8. Auditory brainstem response wave I latency slopes for KN a1DKO and WT mice after noise exposure. Auditory brainstem responses were recorded for both KN a1DKO and WT mice on three occasions. The first was one week before noise exposure (baseline), 24 hours after noise, and 168 hours after noise exposure. Mice were allocated to two groups that received a different intensity noise exposure. The first group was exposed to a 100 dB SPL narrowband noise (8-16kHz) for two hours.
The second group was exposed to the same noise for the same duration, at an intensity of 110 dB SPL.
Wave I auditory brainstem response latencies were recorded for KN a1DKO and WT mice in response to either click stimuli, or to 8 kHz, 16 Khz, and 32 khz tones. Amplitude-intensity graphs were created for peak I amplitudes obtained for stimuli between 40 dB SPL and 90 dB SPL. Slopes were then calculated from fitted regression lines and compared between KN a1DKO and WT mice at baseline, 24 hours after, and 168 hours after both noise exposures. Regression lines were only fitted if corresponding thresholds were higher than 85 dB SPL. Auditory brainstem response peak I latency slopes for the four stimuli are presented for baseline (A), 24 hours after a 100 dB SPL noise exposure (B), at 168 hours after a 100 dB SPL noise exposure (C), 24 hours after a 110 dB SPL noise exposure (D), and at 168 hours after a 110 dB SPL noise exposure (E). Data was compared between KN a1DKO and WT mice by three-way ANOVA, with subsequent follow-up post hoc analysis. Significance is indicated by: * = p < 0.05, *** = p < 0.001.
4.3.8. Wave I latency slopes in KN a1 DKO and WT mice after noise exposure
In this study, KN a1 DKO mice developed larger wave I latency slopes with age than WT mice. To test whether this same effect occurred after noise exposure, we compared wave I latency slopes between KN a1DKO and WT mice. At baseline, wave I latency slopes for click stimuli were larger (Kruskal wallis, p < 0.001) in KN a1DKO mice than in WT mice but not different for the other stimuli (Figure 4.8A). We did not observe a similar difference in click ABR wave I latency slope in our ageing cohort and it is unclear why these results would be different. However, this difference in click ABR wave I latency slope between KN a1DKO and WT mice did not persist over other frequencies or after noise exposure.
Figure 4.9. Inner hair cell – auditory nerve fibre synapse counts in KN a1DKO and WT mice after noise exposure. Mice were allocated to two groups that received a different intensity noise expo- sure. The first group was exposed to a 100 dB SPL narrowband noise (8-16kHz) for two hours. The second group was exposed to the same noise for the same duration, at an intensity of 110 dB SPL. Organs of Corti were harvested 168 hours after noise exposure. Subsequently, presynaptic CTBP2, and postsynaptic GluA2 glutamate receptor subunits were fluorescently labelled. Cochlear frequency was determined by creating a frequency map using low magnification images. Confocal images were then created for a region of about ten inner hair cells of three specific frequency regions (8 kHz, 16 kHz, and 32 kHz). Functional synapses were classified as synapses that contained labelling for both CTBP2 and GluA2 subunit. Exam- ple images of synapse labelling with CTBP2 labelling in green and GluA2 labelling in red are shown for control WT mice (A), control KN a1DKO mice (B), noise exposed WT mice (C), and noise exposed KN a1 DKO mice (D) at 16 kHz regions. Functional synapses were calculated per IHC and compared between KN a1DKO and WT mice for the different frequency regions. Functional synapse counts are presented for baseline (E), a 168 hours after a 100 dB SPL noise exposure (F), and a 168 hours after a 110 dB SPL noise exposure (G). Data were compared using three-way ANOVA with follow-up post hoc analysis. Significance is indicated by: ** = p < 0.01.
Wave I latency slopes were not different between KN a1DKO and WT mice 24 h af- ter a 100 dB SPL noise exposure (Figure 4.8B). Wave 1 latency slopes were also not different between KN a1DKO and WT mice 168 h after a 100 dB SPL noise exposure (Figure 4.8C). Wave I latency slopes for 16 kHz stimuli were larger (Kruskal wallis, p
< 0.05) in WT mice than in KN a1DKO mice but not different for the other frequencies (Figure 4.8D) 24 h after a 110 dB SPL noise exposure. Wave I latency slopes were not different between KN a1DKO and WT mice 168 h after a 110 dB SPL noise expo- sure (Figure 4.8E). Therefore, across most frequencies, wave I latency slopes did not appear to be altered in KN a1DKO mice compared to WT mice after noise exposure.
4.3.9. Synapse survival in KN a1DKO mice after noise exposure
To examine whether KN a1DKO mice are more vulnerable than WT mice to synapse loss after noise exposure, we compared synapse numbers over three different cochlear frequency regions from tissues collected 168 h after either a 100 dB SPL or a 110 dB SPL noise exposure. The fluorescent labeling for the synaptic markers showed simi-
4.3. Results 97
lar patterns of immunoreactivity at 16 kHz for both WT (Figure 4.9A) and KN a1DKO (Figure 4.9B). After noise exposure, synaptic damage could be clearly detected in both WT and KN a1DKO mice but was more pronounced in the KN a1DKO than in the WT as exemplified in Figure 4.9C-D. At baseline, the number of synapses was compara- ble between KN a1DKO mice and WT mice (Figure 4.9E). After a 100 dB SPL noise exposure the number of synapses was comparable between KN a1DKO mice and WT mice (Figure 4.9F). After a 110 dB SPL noise exposure the number of synapses was smaller (three-way ANOVA, p < 0.01) in KN a1DKO mice at 16 kHz than in WT mice but comparable for the other frequency regions (Figure 4.9E). Thus, KN a1DKO mice were more susceptible to synapse loss at the 16 kHz region than WT mice following a 110 dB SPL noise exposure.
4.4. Discussion
Hidden hearing loss is a recently identified pathology of the inner ear suggested to be an early stage in the development of age-related (ARHL) and noise-induced (NIHL) hearing loss. Hidden hearing loss is hallmarked by loss of the spiral ganglion neuron (SGN) – inner hair cell synapses (IHC) and a reduction in ABR wave I amplitude despite preservation of hearing thresholds. KN a1 DKO mice exhibit a loss of ABR wave I amplitude without the loss of the SGN-IHC synapses at six weeks of age, which led to the prediction that KN a1 DKO mice may be more vulnerable to ARHL and NIHL than wildtype (WT) mice. KN a1DKO mice might then be an attractive model to study both the development of ARHL and NIHL and the mechanisms that underlie hidden hearing loss. The aim of this study was to test the prediction that KN a1DKO mice are more vulnerable to ARHL and NIHL than WT mice. We found that indeed, KN a1DKO mice exhibit increased auditory deficits and SGN-IHC synapse loss with ageing and after noise exposure.
4.4.1. Development of age-related and noise-induced hearing loss in KN a1DKO mice
Age-related hearing loss
In this study, we aged WT and KN a1DKO mice on a C57BL/6 background up to 36 weeks of age to investigate the pathogenesis of ARHL by examining both functional and morphological changes with age. Both WT and KN a1DKO mice showed ARHL as evidenced by increasing ABR wave I absolute thresholds at about 24 weeks (6 months) of age. This pattern of ARHL is expected for the WT strain, which is known to develop early onset high frequency hearing loss (Liu et al., 2012; Schettino and Lauer, 2013;
Q. Y. Zheng et al., 1999). Evidence of accelerated ARHL could be observed in KN a1 DKO mice even on this background strain. Specifically, KN a1DKO mice showed in-
creased ABR wave I absolute thresholds compared to WT mice at 32 kHz at 12 weeks and 8 kHz at 24 weeks. We further observed a reduction in ABR wave I amplitude slopes with age in both WT mice, similar to a previous report (Takeda et al., 2017), and in KN a1 DKO mice. Other studies are in agreement with a reduction of ABR wave I amplitudes with increasing age ( gerbil, Boettcher et al., 2003; CBA/CaJ mice, Halonen et al., 2016; rhesus monkey, Ng et al., 2015; CBA/CaJ mice, Sergeyenko et al., 2013; rhesus monkey, Torre and Fowler, 2000). As expected from our previous observations, KN a1 DKO mice showed significantly reduced ABR wave I amplitude I/O slopes compared to WT mice at all frequencies at 6 weeks of age (Reijntjes et al., 2019). At this age, no differences in absolute threshold were observed between the KN a1DKO and WT mice nor in the number of IHC-SGN synapses. With increasing age, ABR wave I amplitude I/O slopes were reduced in KN a1DKO compared to WT mice for click stimuli at 12 and 24 weeks and at 16 kHz at 24 weeks. Examination of the ABR wave I latency I/O slopes indicated little age-related changes in WT mice similar to results found in a previous study (Takeda et al., 2017). In contrast, al- though ABR wave I latency I/O slopes were not found to be different between KN a1 DKO and WT mice at 6 weeks of age, KN a1DKO mice showed age-related increases in ABR wave I latency I/O slopes at about 24 weeks (6 months) of age. ABR wave I latency I/O slopes were significantly larger in KN a1DKO compared to WT mice for click stimuli and at 16 kHz at 24 and 36 weeks of age and at 8 kHz at 24 weeks of age, indicative of shorter latencies.
Morphological analysis of the organs of Corti indicated that neither KN a1 DKO nor WT mice showed gradual age-related loss of either IHCs or OHCs. However, at 36 weeks of age, a steep decline in the number of OHCs was observed for both geno- types at the 32 kHz region. Other studies have reported similar steep declines in OHC numbers between 6 and 9 months of age in WT mice (Hequembourg and Liber- man, 2001; Johnson et al., 2017; Takeda et al., 2017). Another study suggests that at older ages, OHC loss starts to affect lower frequency regions as well with loss up to 80% of OHCs at regions > 8 kHz in WT mice (Johnson et al., 2010). Other strains of mice such as the CBA/CaJ strain are more resistant to OHC loss, showing hardly any loss at ages up to 18 months (Kane et al., 2012). We further observed that in WT mice, the IHC-SGN synapse numbers were reduced across all frequencies at 36 weeks of age. Another study indicated a similar extent of synapse loss in WT mice at this age (Takeda et al., 2017). For the KN a1DKO mice we observed a similar loss of synapses with age that was accelerated compared to WT mice specifically at 32 kHz at 24 weeks of age. Thus, KN a1DKO mice display a similar pattern of ARHL as WT mice but with accelerated threshold shifts, increased ABR wave I I/O latency slopes, and accelerated synapse loss.
4.4. Discussion 99
Noise-induced hearing loss
Both WT and KN a1DKO mice showed NIHL as evidenced by increased ABR wave I absolute thresholds for the 16 and 32 kHz tones after 2 h noise exposure to either a 100 or 110 dB SPL. These results are similar to previous reports (Takeda et al., 2017). In addition, we observed larger threshold shifts in the KN a1DKO mice, both for the 16 and 32 kHz tones after the 100 dB SPL exposure and for the 8 kHz tone and click stimuli after the 110 dB SPL exposure. ABR wave I amplitude I/O slopes were larger in KN a1DKO than in WT mice at baseline as seen before in our ageing cohort and previous study (Reijntjes et al., 2019). ABR wave I amplitude slopes appeared to be decreased after noise exposure, particularly for the 110 dB SPL exposure in both WT and KN a1DKO mice. However, we did not verify if differences were statistically significant. For the 110 dB SPL exposure, KN a1DKO mice no longer showed smaller ABR wave I amplitude slopes than WT mice. This observation suggests that noise exposure to high sound intensities in WT mice results in a similar loss of ABR wave I amplitude slope as lower sound intensity noise exposure on the KN a1DKO mice. We observed no differences in ABR wave I latency I/O slopes for the WT mice which fits with previous reports (Takeda et al., 2017). There were no differences in ABR wave I latency I/O slopes for the KN a1DKO mice observed either.
We did observe the loss of IHC-SGN synapses up to 40% at the 32 kHz region in both WT and KN a1DKO mice independent of exposure intensity. These findings are similar to previous reports of synapse loss after noise exposure (Kujawa and Liber- man, 2015). In addition, KN a1DKO mice lost up to 40% of their IHC-SGN synapses after the 110 dB SPL noise exposure at the 16 kHz region which was not observed in WT mice.
Investigation of the progression of acquired hearing loss in KN a1 DKO mice is made difficult because of the C57BL/6 background phenotype. C57BL/6 mice are known to develop early onset high frequency hearing loss (Liu et al., 2012; Schettino and Lauer, 2013; Q. Y. Zheng et al., 1999). This phenotype likely reduces the ability to detect differences in the progression of acquired hearing loss between KN a1DKO and WT mice on this strain. The application of different noise exposure intensities between 94 and 116 dB SPL has previously shown differences in the pathology of the cochlea in mice (Wang et al., 2002). We predicted that the application of different noise intensity regimes would increase our ability to detect differences between KN a1 DKO and WT mice in the progression of acquired hearing loss. Indeed, the two noise regimes used revealed specific differences in vulnerability to NIHL between KN a1 DKO and WT mice. The lower 100 dB SPL noising regime indicated an increased vulnerability of the KN a1DKO mice to ABR wave I threshold shifts. The higher 110 dB SPL noising regime indicated an increased vulnerability of the KN a1DKO mice to IHC-SGN synapse loss. Additional manipulations in both intensity and the noising frequency band might reveal further differences particularly at the high frequency
regions. Nevertheless, in this study we found that with both noise exposure regimes, as with ARHL, the progression of hearing loss is similar for KN a1DKO and WT mice but accelerated for the KN a1DKO mice.
Together, the findings from our ageing experiment and our noise exposure exper- iment indicate that KN a1DKO mice show similar patterns of progression of ARHL and NIHL compared to WT mice. In addition, KN a1DKO mice exhibit an acceleration of this pathology. However, it remains unclear whether this acceleration of pathology is simply an acceleration of the mechanisms underlying ARHL and NIHL or whether different mechanisms underlie this acceleration of pathology in the KN a1DKO mice.
In either scenario, further characterization of KN a1DKO should indeed be useful in examining the mechanisms underlying the pathogenesis of ARHL, NIHL, and hidden hearing loss.
4.4.2. Mechanisms underlying increased vulnerability of KN a1 DKO mice to development of age-related hearing loss and noise- induced hearing loss
To understand how the loss of KN a1channels increases the vulnerability of mice to ARHL and NIHL it becomes important to study the underlying mechanisms of KN a1function. In this study, we did not directly investigate the molecular function of this channel, but we did observe an increase in ABR wave I latency I/O slopes at about 24 weeks (6 months) of age in KN a1DKO mice. In addition, in a previous study we reported that KN a1channels shape the excitability of the SGNs, and this altered excitability might influence the vulnerability to ARHL and NIHL (Reijntjes et al., 2019). Together with the observed reduction in ABR wave I amplitude in KN a1DKO mice, variation in SGN excitability may underlie the increased vulnerability of these mice to ARHL and NIHL.
Studying other ion channels that regulate SGN excitability and ABR wave I am- plitude and their effects on ARHL and NIHL may indicate shared mechanisms that lead to hearing loss. Unfortunately, the identity of many proteins that shape SGN excitability are yet to be discovered. Two further ion channels that regulate SGN ex- citability besides KN a1have been studied in vitro and in vivo for their contribution to normal hearing. Voltage gated potassium channels (KV) of the KV1 family together with hyperpolarization-activated cyclic nucleotide–gated (HCN) channels have been shown to control SGN excitability in vitro (Q. Liu et al., 2014). Specifically, these channels have opposing effects on SGN soma excitability: KV1 channels serve to reduce SGN excitability and HCN1 channels increase SGN excitability. Knockout of these ion channels shows a complementary opposing effect in vivo. Knockout of KV1.1 results in shorter ABR wave I latency (Allen and Ison, 2012) and knockout of the HCN1 subunit results in longer ABR wave I latency (Kim and Holt, 2013). Thus, KN a1, KV1.1, and HCN1 channels all regulate SGN excitability and influence ABR
4.4. Discussion 101
wave I latency. However, only KN a1channels were shown to affect ABR wave I am- plitudes. Since the contribution of KV1.1 and HCN1 channels to ARHL and NIHL were not studied, it remains speculation how loss of these ion channels would affect ARHL and NIHL hearing loss. However, as these ion channels seem to affect SGN excitability and ABR wave I latency but not ABR wave I amplitude, they might pro- vide further models to complement the KN a1DKO mice to investigate whether loss of ABR wave I amplitude or altered SGN excitability is important in accelerating ARHL and NIHL. Therefore, further studies should make use of these models to examine the effect of altered SGN excitability and these ion channels in general on ARHL and NIHL.
In this study, both KN a1 DKO and WT mice show threshold shifts and reduced ABR wave I I/O amplitude slopes that precedes loss of OHCs, IHCs, and IHC-SGN afferent synapses. This pattern of hearing loss is different from the pattern predicted by the recently described model of hidden hearing loss mostly in CBA/CaJ mice. In this model, hidden hearing loss is caused by auditory synaptopathy, where loss of the IHC-SGN synapses is the primary event leading to hearing loss. The C57BL/6 mouse strain is known to be homozygous for the Cdh23ahlgene that drives early onset hearing loss, and thus these mutations may give rise to a different pattern of hearing loss. On top of this defect, KN a1DKO display a rapid reduction of ABR wave I I/O amplitude slope indicative of another pattern of hearing loss. In other species such as gerbil, age related hearing loss is dependent on a combination of strial and neu- ral degeneration (Heeringa and Koppl 2019 in print). Therefore, this study and other studies would suggest that hidden hearing loss caused by auditory synaptopathy may be less generalizable to mice and mammals than has been proposed and just one of various mechanisms and patterns that contribute to hearing loss.
