Fig 1.
NAC/HPN-07 treatment reduced the blast-induced down-regulation of Arc in the central auditory system.
Examples of Arc immunohistochemical staining in the AC, the IC and the DCN of the naive control group (arrows in A, D and G), the vehicle-treated, blast- exposed group (arrows in B, E and H) and the blast exposed, therapeutic-treated group (arrows in C, F and I). The number of Arc-positive neurons in the AC, the IC and the DCN was counted and statistical analyzed (J, K L). Statistical significance was determined by ordinary one-way ANOVA (Tukey correction for multiple comparisons). The number of Arc-positive neurons was significantly decreased in the AC (J, * p < 0.05), the IC (K, * p < 0.05) and the DCN (*** p < 0.001) in vehicle-treated rats exposed to blast (B+P) compared to naive controls (NC). Significant recovery of Arc-positive neurons was observed in the AC and the IC in rats exposed to blast and then treated with NAC/HPN-07 (B+T) compared to vehicle-treated rats exposed to blast (*, p < 0.05). No significant recovery was observed in the DCN after NAC/HPN-07 treatment (L, p > 0.05), although these animals showed a trend of therapeutic recovery in this auditory center. Numbers in parentheses represent the number of animals evaluated in each group. The scale bar in I = 50 μm, and applies to A-I.
Table 1.
Summary of blast-induced biomarker changes in the auditory system.
Fig 2.
NAC/HPN-07 treatment reduced chronic accumulation of GluR2 in the DCN of blast-exposed rats.
Examples of GluR2 immunostaining in the DCN from the naive control group (arrows in A), the vehicle-treated, blast-exposed group (arrows in B), and the blast-exposed, therapeutic-treated group (arrows in C). GluR2 positive-neurons were largely restricted to the fusiform soma layer (FSL) in the naive control group (arrows in A) and in blast-exposed rats treated with NAC/HPN-07 (arrows in C). In contrast, GluR2-positive neurons were observed in all three layers (the FSL, the middle layer, ML, and the deep layer, DL) of the DCN in vehicle-treated, blast-exposed rats (arrows in B). The number of GluR2-positive neurons in the DCN was counted and statistically analyzed (D). Statistical significance was determined by ordinary one-way ANOVA (Tukey correction for multiple comparisons). A significantly increased number of GluR2- positive neurons was counted in the DCN of the vehicle-treated, blast-exposed group (B+P) compared to naive controls (NC, *** p < 0.001). Statistically fewer GluR2-positive neurons were counted in the DCN in the blast-exposed, NAC/HPN-07-treated group (B+T) compared to the vehicle-treated, blast-exposed group (B, * p < 0.05). There was no significant difference in the number of GluR2-positive neurons between the NC and the B+T groups (p > 0.05). Numbers in parentheses represent the number of animals evaluated in each group. The scale bar in C = 50 μm, and applies to A-C.
Fig 3.
NAC/HPN-07 treatment reduced chronic blast-induced up-regulation of GABAAR-α1 in the central auditory system.
Examples of GABAAR-α1 immunostaining in the DCN (A-C) and in the AC (E-G) from rats in the naive control group (NC, arrows in A and E), the vehicle-treated, blast-exposed group (B+P, arrows in B and F), and the blast-exposed, therapeutic-treated group (B+T, arrows in C and G). The number of GABAAR-α1-positive neurons in the DCN (D) and the AC (H) were counted and statistically analyzed. Statistical significance was determined by ordinary one-way ANOVA (Tukey correction for multiple comparisons). Significantly increased numbers of GABAAR-α1-positive neurons were counted in the DCN and the AC of rats in the vehicle-treated, blast-exposed group (B+P) compared to naive controls (NC, *** all p < 0.001). In NAC/HPN-07-treated animals (B+T), the number of GABAAR-α1-positive neurons in the DCN and AC was significantly reduced (* p < 0.05 and *** p < 0.001, respectively) compared to the vehicle-treated, blast-exposed group (B). There was no significant difference between the NC and B+T groups in the DCN (p > 0.05). Numbers in parentheses represent the number of animals evaluated in each group. The scale bar = 50 μm in G, and applies to A-C and E-G.
Fig 4.
Specification of GABAAR-α1 positive neurons in the DCN after blast exposure.
To identify the cells that persistently up-regulated GABAAR-α1 in the DCN at nine-weeks post-blast exposure, GABAAR-α1 and GAD67 (a marker for inhibitory neurons, or interneurons) co-labelling was performed in tissue sections from the naive control group (A, D, G, J), the vehicle-treated, blast-exposed group (B, E, H, K), and the blast-exposed, NAC/HPN-07-treated group (C, F, I, L). GABAAR-α1 and GAD67 dual-labeled neurons (arrows in J, K, L) are observed in the DCN of all conditions, while more dual-labeled neurons were observed in the DCN of the vehicle-treated, blast-exposed group (arrows in K). In these animals, all GABAAR-α1-positive neurons were co-labeled with GAD67, specifying them as inhibitory neurons. The scale bar = 10 μm in L, applies to A-L.
Fig 5.
NAC/HPN-07 treatment attenuated chronic blast-induced up-regulation of VR1 in spiral ganglion neurons.
Examples of VR1 immunostaining in spiral ganglia from the naive control group (A), the vehicle-treated, blast-exposed group (B) and the blast-exposed, therapeutic-treated group (C). The number of VR1-positive SGNs were counted and statistically analyzed (D). Statistical significance was determined by ordinary one-way ANOVA (Tukey correction for multiple comparisons). A significantly increased number of VR1-positive neurons were observed in the SG of the vehicle-treated, blast-exposed group (B+P) compared to naive controls (NC, *** p < 0.001). A significant reduction in the number of VR1-positive SGNs was observed in blast-exposed rats treated with NAC/HPN-07 (B+T) compared to the vehicle-treated, blast-exposed group (B+P, *** p < 0.001). There was no significant difference in the number of VR1-positive SGNs between the NC and the B+T groups (p > 0.05). Numbers in parentheses represent the number of animals evaluated in each group. The scale bar in C = 50 μm, applies to A-C.
Fig 6.
NAC/HPN-07 treatment significantly reduced the development of a chronic blast-induced tinnitus percept.
Tinnitus index scores of the acoustic startle were increased in blast-exposed rats treated with vehicle (saline) but not altered in those receiving therapeutic NAC/HPN-07 treatment when compared with naïve, unexposed controls. Mean (± SE) scores are shown for each group, where n = 17–20 rats/group. Data were obtained at 8 weeks after blast exposure. Statistical significance by a mixed ANOVA: * p < 0.05, the main effect of group.
Fig 7.
Incidence and frequency-dependent manifestations of tinnitus percept in blast-exposed rats.
(A.) summarizes the incidence of tinnitus in the therapeutic-treated group in comparison to the vehicle-treated group. Data were obtained at 8 weeks after blast exposure. Statistical significance was determined by Fisher’s exact test: ** p < 0.01. A 50% reduction in the incidence of a chronic tinnitus percept was observed in the therapeutic treatment group relative to the vehicle treatment group. (B.) depicts the number of times that each of the four test frequencies was deemed a tinnitus percept frequency in the vehicle treatment group. These values were then expressed as a proportion of the total number of vehicle-treated rats with tinnitus. Here, “tinnitus frequency” was defined as the test frequency with an index score that exceeded the 95% confidence interval of the companion naïve control group. The blast-induced tinnitus percepts were characterized by a wide frequency range with a peak of around 20 kHz.
Fig 8.
The shock tube blast exposure paradigm did not induce behavioral evidence of hyperacusis.
Both acoustic startle response (ASR) and acoustic prepulse inhibition (PPI) of startle remained unchanged in blast-exposed rats that underwent vehicle treatment compared to unexposed, naive control rats, suggesting that blast overpressures did not result in hyperacusis-like behavior. Mean (± SE) values are shown for ASR (A.) and PPI (B.), where n = 20 rats/group. Data were obtained at 8 weeks after blast exposure.
Table 2.
ABR threshold shifts after blast exposure.
Fig 9.
NAC/HPN-07 treatment normalized the blast-induced enhancement of the wave V-to-wave I amplitude ratio of ABRs.
In vehicle-treated, blast-exposed rats, significant reductions in ABR wave-I amplitudes were measured at 4, 16, and 24 kHz (A.). Naïve control-like ABR wave-V amplitudes were measured at 4, 8, and 16 kHz (B.), thus leading to enhancement in wave-V/I amplitude ratios (C.), where 8 kHz presented an exception: no changes in ABR measurements. In the therapeutic-treated cohort, the changes in ABR wave-I amplitudes across the test frequency range (D.) were approximately equivalent to those measured for the ABR wave-V amplitudes (E.), such that the ABR wave-V/I ratio was maintained within a normal range (F.). Symbol keys in each row apply to the three panels in the same row. ABR wave-I amplitudes (A., D.), ABR wave-V amplitudes (B., E.), and wave-V/I amplitude ratios (C., F.) were normalized to naïve control means. Means (± SE) in response to tone bursts at 80 dB SPL are shown, where n = 34–40 ears/group. Data were obtained at 8 weeks after blast exposure. Statistical significance was determined by t-test with Holm-Bonferroni correction: + p < 0.05, ++ p < 0.01 and +++ p < 0.001, represent the simple effect of group at different test frequencies.
Fig 10.
Time-dependent effects of treatment on ABR threshold sensitivity.
After initial threshold stabilization, the reduced amplitudes of ABR wave I progressively recovered with increasing post-blast exposure time in the NAC/HPN-07 therapeutic treatment group (B.) but not in the vehicle treatment group (A.). ABR wave-I amplitudes were normalized with respect to mean values in age-matched, unexposed naïve controls. Mean (± SE) amplitudes in response to tone bursts at 80 dB SPL are shown, where n = 34–40 ears/group. Data were obtained at 4 and 8 weeks after blast exposure, respectively. Statistical significance was determined by paired t-test with Holm-Bonferroni correction: + p < 0.05, the simple effect of time at 8 and 24 kHz.
Fig 11.
NAC/HPN-07 treatment rescued the blast-induced loss of IHC afferent synapses.
A.–C. represent maximum projections from confocal z stacks of the IHC area in the 16 kHz region immunolabeled with antibody against CtBP2 (red) to show synaptic ribbons. D.-E. summarize mean (± SE) counts of presynaptic ribbons per IHC at each of the tonotopic positions for animals in each experimental cohort, as indicated in the symbol keys, where n = 12–23 ears/group. Data were obtained at approximately nine weeks after blast exposure. Statistical significance was determined by t-test with Holm-Bonferroni correction: ++ p < 0.01 and +++ p < 0.001, representing the simple effect of group at 32 and 48 kHz, respectively; by a mixed ANOVA: ** p < 0.01, the main effect of group.
Fig 12.
Tinnitus-related enhancement of ABR wave V/I ratios correlated with IHC de-afferentation.
Loess regression suggested that the enhancement of the wave V-to-wave I ratio after blast injury is likely due to loss of cochlear synapses rather than damage to cochlear sensory cells (see Results). Loess curves were fitted to the empirical data in a scatterplot of wave-V/I ratio vs. either ribbon count (A.) or threshold shift (B.) with a smoothing span = 0.75 and a polynomial degree = 2 (quadratic). ABR wave-V/I ratios and ABR threshold shifts, as well as IHC presynaptic ribbon counts, were averaged in blast-exposed ears from each vehicle (red dots)- or therapeutic (green dots)-treated animal (n = 44) for frequencies from 4–32 kHz and then normalized to mean values from naïve, unexposed control ears (n = 12). Data were obtained at 8–9 weeks after blast exposure. Loess curves and linear trendlines (that the loess fit captures in the data) were depicted by solid lines and dashed lines, respectively.
Fig 13.
Predicted probabilities from logistic regression modeling of tinnitus relative to ABR wave V/I amplitude ratio.
Observed ABR data were extracted from Fig 9C and 9F, and corresponding behavioral binary data coded as 1 (presence of tinnitus) or 0 (absence of tinnitus) were extracted from Fig 7B. Abscissa depicts the maximum ratio V/I across frequencies and ears for each subject.
Table 3.
Measures of fit for logistic regression.