(C) PLOS One
This story was originally published by PLOS One and is unaltered.
 . . . . . . . . . .



From hidden hearing loss to supranormal auditory processing by neurotrophin 3-mediated modulation of inner hair cell synapse density [1]

['Lingchao Ji', 'Kresge Hearing Research Institute', 'Department Of Otolaryngology Head', 'Neck Surgery', 'University Of Michigan', 'Ann Arbor', 'Michigan', 'United States Of America', 'Beatriz C. Borges', 'David T. Martel']

Date: 2024-07 

Plots of peak latency recorded at 16 kHz against sound stimulus level show that latencies of ABR peaks I–V are not altered by Ntf3-KD ( A ) and Ntf3-OE ( B ). n = 14–20, ns = p > 0.05 by two-way ANOVA. The data underlying this figure can be found in S1 Data . Error bars represent SEM. ABR, auditory-brainstem response.

Mean amplitude vs. level functions for ABR peaks I–IV in Ntf3-KD ( A ) and Ntf3-OE ( B ) mice and their respective controls at 16 kHz. Whereas peak I amplitudes are reduced in Ntf3-KD mice, the amplitudes of the other peaks remain normal, indicative of central compensation ( A ). In contrast, Ntf3 overexpression increases amplitudes of ABR peaks I to IV ( B ). N = 14–20, ns = p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, ***p < 0.0001 by two-way ANOVA. The data underlying this figure can be found in S1 Data . Error bars represent SEM. ABR, auditory-brainstem response.

Whereas peak I amplitude was reduced in mice with reduced Ntf3 cochlear expression, later ABR peak amplitudes were normal in these mice ( Fig 5A ), suggesting that decreased sound-evoked activity of the auditory nerve leads to central gain in several higher auditory centers, as seen after noise-induced or age-related synaptic loss [ 5 , 11 ]. In contrast, Ntf3 overexpression resulted in increased amplitudes for ABR peaks I–IV ( Fig 5B ), indicating that increased IHC synapse density enhances sound-evoked signaling along the ascending auditory pathway. All peak latencies were normal in both mutants ( Fig 6A and 6B ), suggesting that auditory nerve myelination and conduction velocity was not affected by the altered Ntf3 levels and synapse densities [ 38 ].

DPOAE ( A, D ) and ABR ( B, E ) thresholds in Ntf3-KD and Ntf3-OE mice are not different than their controls. In contrast, Ntf3 knockdown reduces ABR P1 amplitudes ( C ), whereas overexpression leads to increased peak I amplitudes ( F ). Representative traces of DPOAEs ( G ) and ABRs ( H ). n = 15–24. ABR P1 amplitudes were assessed at 80 dB SPL. ns = p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, ***p < 0.0001 by two-way ANOVA. The data underlying this figure can be found in S1 Data . Error bars represent SEM. ABR, auditory-brainstem response; DPOAE, distortion product otoacoustic emission.

At 8 weeks of age, we assessed cochlear function by measuring distortion product otoacoustic emissions (DPOAEs), which reflect outer hair cell function, and ABRs, which reflect the summed responses of the auditory nerve and several higher auditory centers [ 33 ]. While peak I reflects synchronous sound-evoked activity of auditory-nerve fibers, the second peak (peak II) is dominated by contributions from cochlear-nucleus bushy cells, and later waves represent bushy-cell targets in the superior olivary complex and inferior colliculus [ 33 – 37 ]. These recordings confirmed that, as we reported earlier [ 19 ], reduced Ntf3 expression by IHC supporting cells reduces ABR peak I amplitudes without changing ABR and DPOAE thresholds ( Fig 4A–4C ), while increasing Ntf3 leads to normal thresholds with increased peak I amplitudes ( Fig 4D–4F ). An example raw recording of DPOAE and ABR waveform is shown in Fig 4G and 4H . Like the changes in synapse density, the effects of Ntf3 levels on peak I amplitudes are stronger in the middle and high frequencies, which reflect responses arising from the middle/basal cochlear regions.

Representative confocal images of IHC synapses at the 16 kHz cochlear region from Ntf3-KD ( A ) and Ntf3-OE ( E ) mice and their respective controls immunolabeled for presynaptic ribbons (CtBP2—red), postsynaptic receptor patches (GluA2—green), and hair cells (Myo7a - blue). Mean counts (± SEM) of ribbons ( B, F ), GluA2 patches ( C, G ), and colocalized markers ( D, H ) in Ntf3 KDs and OEs. n = 5, ns = p > 0.05, * p < 0.05, ** p < 0.01, *** p < 0.001, *** p < 0.0001. Synaptic markers were compared by two-way ANOVA. The data underlying this figure can be found in S1 Data , the raw images were deposited in the Dryad repository ( https://doi.org/10.5061/dryad.k6djh9w8v ). Error bars represent SEM. IHC, inner hair cell.

We also immunostained cochleas to assess the numbers of synapses between IHCs and auditory-nerve fibers. Whereas almost all auditory-nerve fibers contact a single IHC, each IHC is contacted by numerous auditory-nerve fibers. Each of these glutamatergic synaptic contacts can be identified as a closely apposed pair of puncta in cochleas immunostained for CtBP2, a major component of the presynaptic ribbon, and GluA2, a subunit of the glutamate receptors localized at the postsynaptic terminals ( Fig 3A and 3E ). In a normal mouse, the mean number of synapses per IHC follows an inverted U-shaped function, peaking in mid-cochlear regions at a value of roughly 20 synapses per IHC [ 2 , 19 ]. As we previously reported [ 19 ], reduced supporting-cell Ntf3 expression levels decreased IHC synapse density in the basal half of the cochlea by as much as 20%, while increased supporting-cell Ntf3 expression levels increased IHC synapse density in the same cochlear regions by as much as 30% ( Fig 3B–3D and 3F–3H ). The stronger impact of supporting-cell derived Ntf3 on synapse density in the basal half of the cochlea likely reflects the fact that endogenous Ntf3 levels are lower in this region [ 30 ], making it more sensitive to changes in expression.

mRNA level of Ntf3 and VGF, a gene downstream of TrkC signaling, are reduced in Ntf3-KD cochleas ( A ) and increased in Ntf3-OE cochleas ( B ). Furthermore, cochlear of Ntf3 and VGF mRNA levels are correlated ( C ). In contrast, cortical Ntf3 mRNA level is slightly decreased in Ntf3-KD mice ( D ) and unchanged in Ntf3-OE mice ( E ). No changes in VGF mRNA levels are observed in the brains of either Ntf3-KD or Ntf3-OE mice ( D, E ). n = 6–8, ns = p > 0.05, * p < 0.05, ** p < 0.01, mRNA levels were compared by two-tailed unpaired t test. The data underlying this figure can be found in S1 Data . Error bars represent SEM. CNS, central nervous system.

Quantitative RT-PCR ( Fig 2A and 2B ) confirmed that cochlear Ntf3 levels were reduced in the Ntf3 flox/flox ::Plp1-CreER T , i.e., Ntf3 Knockdown (Ntf3-KD) mice and increased in the Ntf3 STOP ::Plp1-CreER T , i.e., Ntf3 Overexpressor (Ntf3-OE), mice. Similarly, expression of VGF mRNA, a gene downstream of neurotrophin receptor signaling [ 31 ], was decreased in the cochleas of Ntf3-KD mice ( Fig 2A ) and increased in Ntf3-OE mice ( Fig 2B ). Furthermore, there was a clear correlation between the mRNA levels of Ntf3 and VGF ( Fig 2C ), indicating that the changes in Ntf3 expression impact TrkC signaling in the inner ear. Since the Plp1 gene is also expressed in oligodendrocytes in the brain, and auditory-driven behaviors such as GPIAS are modulated by cortical circuits [ 32 ], we also measured the levels of Ntf3 and VGF in the cerebral cortex. We found a small decrease in cortical Ntf3 mRNA levels in Ntf3-KD mice ( Fig 2D ), no change in Ntf3-OE mice ( Fig 2E ), and most importantly, no changes in VGF mRNA levels in either of the mutants ( Fig 2D and 2E ), indicating that the manipulation of Ntf3 expression in Plp1-expressing cells is unlikely to have a direct effect on the central nervous system (CNS).

Between the ages of 8 and 15 weeks, mutant and control mice underwent a variety of behavioral and physiological tests ( Fig 1 ). At 16 weeks, cochlear tissues were harvested to measure Ntf3 expression levels and the number of synapses per IHC (synapse density). We present the latter analyses first, which show that the Ntf3 manipulations had the expected effects on the cochlea.

To modify the levels of Ntf3 expression in IHC supporting cells, we used cell-specific inducible gene recombination as in prior studies [ 19 ]. Briefly, the Plp1-CreERT transgenic line [ 27 ] was used to drive gene recombination in IHC supporting cells via tamoxifen treatment during the neonatal period [ 28 ]. This CreERT transgene was combined with either conditional Ntf3 KO alleles (Ntf3 flox/flox ) [ 29 ] to reduce Ntf3 expression in these cells, or an inducible Ntf3 overexpression transgene (Ntf3 STOP ) to increase it [ 19 ]. Since Ntf3 is expressed by both IHCs and their surrounding supporting cells in the cochlea [ 30 ], knockout of the Ntf3 gene from IHC supporting cells in Ntf3 flox/flox ::Plp1-CreER T reduces cochlear Ntf3 level but does not eliminate it [ 19 ]. As controls, we used mice with the conditional Ntf3 alleles without the CreERT transgene.

IHC synapse density influences auditory processing but not the startle reflex or sensory gating

To examine the impact of IHC synapse density on auditory processing, we tested 3 auditory-driven behaviors; the acoustic startle response (ASR) [39], and 2 behaviors that involve modification of the ASR, prepulse inhibition of the ASR (PPI) [40], and gap-prepulse inhibition of the ASR (GPIAS) [41]. These tests have been a mainstay of studies on temporal processing and hearing-in-noise deficits in animal models [22,42–44].

The ASR is a reflexive and rapid burst of muscular activity in response to a sudden, brief, and intense sound. This is a robust and consistent behavior, easily quantified by measuring the whole-body startle response [45]. To investigate the impact of alterations in IHC synapse density on the ASR, we measured responses to moderate-intensity narrowband stimuli (50 ms 65 dB SPL) centered at different frequencies (8, 12, 16, 24, and 40 kHz) (Fig 7A1) and response to a high-intensity startle stimulus (20 ms 120 dB SPL broadband noise burst) (Fig 7B1), to detect hyper-responsiveness to innocuous sound and to determine if responses to the prepulse stimuli are altered by changes in Ntf3 levels. As expected, the moderate intensity prepulse stimulus alone did not induce startle responses in either mutant or control mice (Fig 7A2 and 7A3). The high-intensity startle stimulus elicited strong responses with amplitudes that were not different between control and mutants (Fig 7B2 and 7B3). These results indicated that changes in Ntf3 expression, synapse density, and ABR peak I amplitudes do not affect reflexive motor responses to sound.

PPI is commonly used to assess sensorimotor gating, i.e., the ability of a sensory stimulus to suppress a motor response [26]. The PPI assay is quantified as the decrease in the ASR when a prepulse stimulus is presented a few milliseconds before the startle stimulus [46]. We used the 50 ms prepulse stimulus described above ending 50 ms prior to the startle stimulus (Fig 7C1) and quantified the magnitude of PPI as the fractional ASR reduction, i.e., 1 minus the ratio of startle magnitude with and without prepulse [46]. On average, the prepulse inhibited the startle response by 40%. There was no significant difference between control and mutant mice in PPI (Fig 7C2 and 7C3), indicating that changes in synapse density and associated changes in auditory-nerve activity do not affect sensorimotor gating.

PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 7. IHC synapse density does not influence the acoustic startle response or prepulse inhibition. Schematics of the protocols for prepulse stimulus (A1), startle stimulus (B1), and PPI stimuli (C1). (A1) The noise prepulse stimulus is a narrowband noise (4 kHz width around variable center frequencies, 65 dB SPL, 50 ms duration). (B1) The startle stimulus is a broadband noise (120 dB SPL, 20 ms duration). (C1) PPI consists of a noise prepulse and a startle stimulus that starts 50 ms after the prepulse. (A2, A3) Reactivity to prepulse is not significantly different between mutant and control mice. (B2, B3) Loud sound (120 dB SPL) elicits startle responses with amplitudes that were similar in mutant mice and their control littermates. (C2, C3) The degree of prepulse inhibition of the startle response by a prepulse was determined using the formula . On average, the prepulse inhibit the startle response by 40%. There is no significant difference between control and mutant mice. N = 20–24 mice/group for response to prepulse (A2, A3); n = 18–20 mice/group for startle response (B2, B3); n = 11–17 mice/group for PPI (C2, C3). ns = p > 0.05 by two-tailed unpaired t tests (B2, B3) or two-way ANOVA (C2, C3). The data underlying this figure can be found in S1 Data. Mean ± SEM are shown. IHC, inner hair cell; PPI, prepulse inhibition. https://doi.org/10.1371/journal.pbio.3002665.g007

Finally, we measured GPIAS, which is used to examine auditory temporal processing [20,22,25,47] and correlates with speech recognition in humans [48,49]. In GPIAS, the suppression of the ASR is induced by the presentation of a brief silent gap in a continuous background noise instead of a mild sound stimulus in silence. Gap detection was measured by presenting animals in broadband background noise (BBN) with gaps of various durations (3 to 50 ms) ending 50 ms before the startle stimulus (Fig 8A). Importantly, the startle amplitudes in background noise were normal in mice with altered Ntf3 expression (Fig 8B and 8C), indicating that synapse number does not alter the salience of the startle stimulus in the presence of background noise.

As done by others [14,22], we quantified GPIAS as the fractional reduction of startle, i.e., 1 minus the ratio of the startle magnitude with and without a gap ( ). Analysis of gap inhibition as a function of gap duration (Fig 8D and 8E) showed that, consistent with previous reports [22], GPIAS is stronger with longer gaps. More importantly, Ntf3-KD mice showed a significant decrease in the gap inhibition, whereas Ntf3-OE mice showed a significant increase compared to their respective controls, indicating that IHC synapse density influences this modification of the ASR. Importantly, gap inhibition for both genotypes was stable across different sessions (S1 Fig), indicating that the results were not influenced by the age of the mice withing the 8 weeks of testing. Furthermore, there was a strong correlation between gap inhibition and ABR P1 amplitudes (S2 Fig), providing evidence that the magnitude of the sound-evoked auditory potentials is critical for the GPIAS.

The conclusion that IHC synapse density influences GPIAS was also supported by 2 additional methods of quantitative analysis of the GPIAS, gap detection threshold and Rd’. Gap detection threshold, which is used to measure the temporal acuity for acoustic transients, is defined as the gap duration that elicits 50% of the maximal inhibition level [22,42,50–54]. As for gap inhibition, gap-detection threshold was higher in Ntf3-KD mice (18.65 ± 3.473 ms) compared to their controls (8.38 ± 1.413 ms), and lower in Ntf3-OE mice (6.712 ± 0.8304 ms) compared to their controls (10.92 ± 1.775 ms) (Fig 8F and 8G). Similar conclusions could be reached by analysis of Rd’, ( ), a parameter that reflects the salience of each gap condition for each mouse [55]. As shown in panels 8H and 8I, analysis of Rd’ as a function of gap duration showed similar trends as for gap inhibition, i.e., Ntf3-KD mice showed a significant decrease in the Rd’ curve, whereas Ntf3-OE mice showed a significant increase compared to their respective controls (Fig 8H and 8I).

PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 8. Ntf3 expression levels influence gap detection thresholds in broadband background noise. (A) Schematic depiction of NO-GAP trials (left) and GAP trials (right). NO GAP trials consisted of a startle sound (120 dB SPL, 20 ms duration) presented in continuous noise background (broadband noise, BBN, 65 dB SPL). In contrast, in the GAP trials, a silent gap in the background noise of variable length (0–50 ms) was presented ending 50 ms before the startle stimulus (S). (B, C) ASR amplitudes for the NO-GAP trials were similar in Ntf3 mutant mice and their control littermates. (D, E) Show the level of gap inhibition vs. gap length and for Ntf3 KD and OE mice, respectively. The inhibition of the startle reflex increases as the gap duration increases. (F, G) Show gap detection thresholds. Gap detection threshold is increased in Ntf3-KD mice (H) and decreased in Ntf3-OE mice (I) compared to their littermate controls. (H, I) Show level of Rd’ vs. gap length for Ntf3 KD and OE mice, respectively. n = 7–20 mice/group, *p < 0.05, **p < 0.01, ***p < 0.001, ***p < 0.0001 by two-tailed unpaired t test (B, C, H, and I) or two-way ANOVA (D, E, F, and G). The data underlying this figure can be found in S1 Data. Mean ± SEM are shown. ASR, acoustic startle response; BBN, broadband background noise. https://doi.org/10.1371/journal.pbio.3002665.g008

Since the spectral components and bandwidth of background noise also affects the behavioral gap detection [55], we presented gaps of 50-ms duration when mice were subjected to narrowband background noise (NBN) centered around different frequencies (6 to 10 kHz, 10 to 14 kHz, 14 to 18 kHz, 22 to 26 kHz, and 38 to 42 kHz; spectral width: 4 kHz) (Fig 9A). Whereas Ntf3 expression levels did not influence the startle responses at any frequency (Fig 9B and 9C), the mutant mice showed consistent differences with the littermate controls with NBN at all frequencies (Fig 9D and 9E). These results indicate that the difference in suppression level caused by altered Ntf3 expression cannot be attributed to impact on startle amplitudes, and that the influence of Ntf3 expression levels on gap inhibition level is not restricted to any specific frequency.

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3002665

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/