Age-related changes in auditory nerve-inner hair cell connections, hair cell numbers, auditory brain stem response and gap detection in UM-HET4 mice

Abstract

This study compared the timing of appearance of three components of age-related hearing loss that determine the pattern and severity of presbycusis: the functional and structural pathologies of sensory cells and neurons and changes in gap detection (GD), the latter as an indicator of auditory temporal processing. Using UM-HET4 mice, genetically heterogeneous mice derived from four inbred strains, we studied the integrity of inner and outer hair cells by position along the cochlear spiral, inner hair cell-auditory nerve connections, spiral ganglion neurons (SGN), and determined auditory thresholds, as well as pre-pulse and gap inhibition of the acoustic startle reflex (ASR). Comparisons were made between mice of 5-7, 22-24 and 27-29 months of age. There was individual variability among mice in the onset and extent of age-related auditory pathology. At 22-24 months of age a moderate to large loss of outer hair cells was restricted to the apical third of the cochlea and threshold shifts in the auditory brain stem response were minimal. There was also a large and significant loss of inner hair cell-auditory nerve connections and a significant reduction in GD. The expression of Ntf3 in the cochlea was significantly reduced. At 27-29 months of age there was no further change in the mean number of synaptic connections per inner hair cell or in GD, but a moderate to large loss of outer hair cells was found across all cochlear turns as well as significantly increased ABR threshold shifts at 4, 12, 24 and 48 kHz. A statistical analysis of correlations on an individual animal basis revealed that neither the hair cell loss nor the ABR threshold shifts correlated with loss of GD or with the loss of connections, consistent with independent pathological mechanisms.

Keywords: NT3; age-related hearing loss; aging; auditory; cochlea; gap detection.

Figure 1

Representative images of surface preparations from mid-cochlea (3 mm from the apex) of mice at 5-7 months of age (A), 22-24 months of age (B) and 27-29 months of age (C), co-labeled with phalloidin (red label) to image hair cells and scars in regions of missing hair cells and with CTBP2 immunolabeling (green puncta) to mark the ribbons in inner hair cell – auditory nerve synapses. Outer hair cell loss is seen at 22-24 months (B) and at 27-29 months (C) and not at 5-7 months (A). CTBP2 immunolabeling (green puncta) is seen at inner hair cell bases at all ages. Quantitative assessment of hair cell loss is seen in shown in Figure 2. Quantitative assessment of CTBP2 puncta (marker for ribbons and synapses) per inner hair cell is shown in Figure 4. Bar = 10 μm.

Figure 2

The mean percentage of inner and outer hair cells that are missing at different locations along the cochlear spiral from apex (left) to base (right) in three groups of UM-HET4 mice terminated at 5-7 months of age (15 mice) (3A); at 22-24 months of age (26 mice) (3B) or at 27-29 months of age (22 mice) (3C). Error bars show standard error of the mean.

Figure 3

Comparison of Auditory Brain Stem Response (ABR) measures in separate groups of UM-HET4 mice, assessed at 5-7 months (15 mice), 22-24 months (26 mice) and 27-29 months (22 mice) of age respectively. A small but significant increase (p<0.01) in the mean of auditory ABR thresholds at 4 kHz is seen at 22-24 months compared to 5-7 months, no significant changes were seen at other frequencies. At 27-29 months of age there were significant elevations (p<0.01) in the means of ABR thresholds at all four frequencies tested either when compared to 5-7 months mice or 22-24 months. These mice were also used for the gap detection and morphological assays described in this manuscript. Asterisks indicate significance (p<0.01). Significance was tested using the Student t test (Sigmaplot)

Figure 4

A) Box plot showing the number of CTBP2 puncta per inner hair cell in four regions of interest (1, 2, 3 and 4 mm from the apex) in three groups of mice, terminated at 5-7 months (15 mice), 22-24 months (26 mice) and 27-29 months (22 mice) of age respectively. The means at 22-24 months are significantly lower than those at 5-7 months in all four regions of interest (p <.01). The means at 27-29 months are not significantly different from those at 22-24 months. Boxes indicate 25th-75th percentile of data; bar indicates median; and whiskers indicate 5th-95th percentile of data. B) When the four regions of interest are combined there is a significant decrease in the overall mean number of connections at 22-24 and 27-29 months of age compared to 5-7 months of age (p <.01) Asterisks indicate significance (p<0.01). Significance was determined using ANOVA with a Bonferroni/Dunn Post-hoc test (Statview).

Figure 5

Box plot comparing mean spiral ganglion neuron counts from groups of UM-HET4 mice at 5-7 months of age (5 mice) and 22-24 months of age (23 mice). . The mean number at 22-24 months of age is significantly reduced in comparison with 5-7 months (p <.01). Asterisks indicate significance (p<0.01). Significance was determined using ANOVA with a Bonferroni/Dunn Post-hoc test (Statview).

Figure 6

Comparison of pre-pulse (A) and Gap (B) inhibition of the acoustic startle reflex (ASR) in mice at 5-7 months (15 mice), 22-24 months (based on 18 of the 26 mice, 8 that did not jump are not included) or 27-29 months (based on 13 mice – 9 mice that did not jump are not included) of age. At 5-7 months of age both the pre-pulse and the gap produce a mean inhibition of the ASR of 30-35%. At 22-24 months of age there is no longer a significant gap-inhibition of the ASR while the pre-pulse inhibition remains in the 30-35% range. The difference in Gap Detection between 5-7 months, 22-24, and 27-29 months is significant (p< 0.01), while pre-pulse inhibition has not changed, indicating an age-related decrease in gap detection. At 27-29 months of age there is no longer a significant pre-pulse or gap inhibition of the ASR, indicative of the large hearing loss in many animals. Asterisks indicate significance (p<0.01). Significance was tested using the Student t test (Sigmaplot)

Figure 7

qRT-PCR results comparing gene expression of Ntf3, Bdnf and Gdnf at 5-7 months versus 22-24 months (A) and at 5-7 months versus 27-29 months (B), in whole cochleae from pooled samples (from 2-3 right cochleae) of UM-HET4 mice (5 pools from 5-7 month old mice, 14 pools from 22-24 month old mice; 5 pools from 27-29 month old mice). Ntf3 gene expression in 22-24 month old animals was significantly decreased compared to 5-7 month old mice (p < 0.02). At 27-29 months of age, Ntf3 expression remained significantly decreased from 5-7 month old animals.. Compared to 5-7 months, the decrease in Bdnf expression was not significant at 22-24 months, but became significant in the 27-29 months old animals (p< 0.01). There were no significant changes in Gdnf expression. Significance was determined using an ANOVA with a Bonferroni/Dunn Post-hoc test (Statview) comparing the Δ₁ values of the young animals to the the Δ₁ values of the older animals. The error was calculated using the standard deviation of the Δ₁ as 2^{−(Δ₁Δ₂Ct-stdev)} and2^{−(Δ₁Δ₂Ct+stdev)}.

Figure 8

Scatter plot comparing the number of CTBP2 immunolabeled ribbons per IHC base (as a marker for inner hair cell – auditory nerve connections) to normalized Gap Inhibition. Mice had best Gap Inhibition at 12 kHz, so the region of interest 2 mm from the apex, corresponding to the 12kHz frequency regions (Viberg and Canlon, 2004) was tested for correlation with Normalized Gap Inhibition at 12 kHz, across individual mice at 22-24 months of age. Data are based on 16 of the 26 mice; 10 mice without a robust jump response at 12 kHz are not included. Correlation was weak based on a modeled linear regression/best fit analysis (Sigmaplot).

Figure 9

Scatter plot comparing ABR thresholds changes at 4 kHz (where there was greatest change) and Normalized Gap Inhibition across individual mice at 22-24 months of age. No correlation is observed based on a modeled linear regression/best fit analysis (Sigmaplot).

Figure 10

Scatter plot comparing loss of outer hair cells in apical third of the cochlea (where there was greatest loss) versus Normalized Gap Inhibition across individual mice at 22-24 months of age. No correlation is observed based on a modeled linear regression/best fit analysis (Sigmaplot).