Association of cochlear outer hair cell - type II spiral ganglion afferents with protection from noise-induced hearing loss

The medial olivocochlear (MOC) efferent feedback circuit projecting to the cochlear outer hair cells (OHCs) confers protection from noise-induced hearing loss and is generally thought to be driven by inner hair cell (IHC) - type I spiral ganglion afferent (SGN) input. Knockout of the Prph gene (PrphKO) encoding the peripherin type III intermediate filament disrupted the OHC - type II SGN innervation and virtually eliminated MOC – mediated contralateral suppression from noise delivered to the opposite ear, measured as a reduction in cubic distortion product otoacoustic emissions. Electrical stimulation of the MOC pathway elicited contralateral suppression indistinguishable between wildtype (WT) and PrphKO mice, indicating that the loss of contralateral suppression was not due to disruption of the efferent arm of the circuit; IHC – type I SGN input was also normal, based on auditory brainstem responses. High-intensity, broadband noise (108 dB SPL, 1 hour) produced permanent hearing loss in PrphKO mice, but not in WT littermates. These findings associate OHC-type II input with MOC efferent - based otoprotection at loud sound levels.


185
Electrically-evoked contralateral suppression 186 To directly drive the contralateral MOC efferent innervation of the OHCs, these fibres were electrically 187 stimulated at the point where they cross the midline on the floor of the fourth ventricle of the brainstem.

188
Mice were initially anaesthetized with a k/x/a cocktail, maintained on a heat pad and eye ointment applied 189 as described in the section Animals. This was followed by shaving the head to provide a clean surgical 190 field for the dorsal-occipital approach to the fourth ventricle. A tracheostomy was performed, and the 191 animals were ventilated throughout the experiment with a respiration rate of 100 breaths per minute and 192 10 cm peak inspiratory pressure (Kent Scientific TOPO™ Dual Mode Ventilator, Torrington, CT, USA).

193
The animal was then positioned onto a teeth bar to stabilize the head position. The dorsal brainstem was 194 accessed by removing muscles and connective tissue between the C1 (Atlas) vertebra and the occiput. This was followed by an incision in the dura covering the foramen, and the stimulating electrodes (two 196 platinum iridium wires with 500 µm exposed tips, 400 µm separation) were positioned with one electrode 197 on the midline of the fourth ventricle floor and the second electrode ipsilateral, using a micromanipulator.

198
Twitching of the ears in response to a test electrical stimulation (monophasic 150 µs pulses, 200 Hz) ~ 199 9 5s, indicated the correct position. α-D-tubocurarine (1.25 mg/kg, i.p.) was then administered to achieve 200 muscle paralysis, which included paralysis of the stapedius reflex. The right (ipsilateral) external auditory 201 meatus of the mouse was coupled to the DPOAE probe. A baseline measurement was obtained as the 202 cubic DPOAE around 16 kHz at 50 or 55 dB SPL primaries (f1 = f2 amplitude, producing DPOAEs ~ 10 203 dB above the noise floor); 20 averaged measurements (10 samples each) over 1 minute. The electrically-204 evoked contralateral suppression was then recorded during 25 seconds of stimulation (8 data point 205 averages), followed by 2.5 min of recorded recovery. The DPOAE amplitudes were analysed relative to 206 the noise floor. These experiments proved particularly challenging, with 3 / 16 WT mice and 3 / 7 KO 207 mice providing useable data. However, electrically-evoked DPOAE suppression for each of the 208 successful WT and PrphKO mice (n = 3 mice per genotype) were obtained from 3 -5 repeats per mouse. 211 To assess the otoprotection conferred by OHC -type II SGN -mediated MOC efferent cochlear amplifier 212 suppression, noise-induced hearing loss was assessed in PrphKO versus WT mice using auditory 213 brainstem responses (ABRs) and cubic DPOAE measurements with acute noise presentation. For ABR, 214 following k/x/a anaesthesia induction, subdermal platinum electrodes were inserted subcutaneously at 215 the vertex (+), over the mastoid process (-), and in the hind flank (ground) (after Cederholm et al., 2012).

216
Click (100 µs alternating polarity) or pure tone pips (4, 8, 16, 24 and 32 kHz) stimuli (5 ms, 0.5 ms 217 rise/fall time, 10/s) were delivered using an EC1 electrostatic speaker, with the generated ABR potentials 218 amplified, filtered and averaged 512 times using the TDT System 3 workstation. The ABR threshold for 219 each frequency was determined as the lowest intensity (5 dB steps from 70 dB SPL) at which the P2 220 ABR wave could still be visually observed above the noise floor (~ 100 nV). Cochleae were then exposed 221 to 108 dB SPL 'open field' broadband white noise (4 -32 kHz; 2 nd order Bessel filter) for 1 hour. The 222 white noise stimulus was generated using custom software with a National Instruments A/D driving an 223 amplifier (BIEMA model Q250, Altronic, Northbridge, WA, Australia) and delivered via an MF1 224 10 speaker (TDT) positioned at midline, 15 cm in front of the mouse (n = 9 for each genotype). The noise 225 was calibrated at ear level. ABR threshold shifts were determined immediately post-noise by 226 remeasurement before the mice recovered from the anaesthesia. The mice were then rested for 14 days 227 before being re-tested to determine irreversible hearing loss (permanent threshold shift (PTS)).

239
Disruption of the outer spiral bundle (type II SGN) fibre tracts in Prph knockout cochleae 240 The normal representation of type I and type II SGN neurite projections within the adult mouse 241 cochlea was established in wildtype (WT) tissue using β-III tubulin and NF200 immunofluorescence 242 ( Figure 1; Video 1). The type II SGN sub-population was specifically identified using peripherin 243 immunofluorescence in the WT cochleae ( Figure 1A). The type II SGN somata were predominantly  (Figures 1 -4). The type II fibres then become basally projecting outer spiral fibres within OSB       validated in the current study using the cubic (2f1-f2) DPOAE (Fig. 10A,B), which is sensitive to changes 517 in the gain of the cochlear amplifier (Mills andRubel, 1996, Frank andKossl, 1996), as mediated by noise, with 60 dB SPL pure tone drivers around 20 kHz in the ipsilateral ear, the peak noise-induced 524 reduction in DPOAE in WT mice was -11.5 ± 3.0 dB (n = 7), measured 3 s after noise onset ( Figure 7C).

525
In contrast, only a residual contralateral suppression effect was detected in the KO mice at this level 526 (peak = -1.6 ± 1.6 dB; p = 0.003, one sample t-test, n = 7). The difference in peak contralateral 527 suppression between WT and KO mice with the 96 dB SPL noise presentation was therefore 9.9 dB (p = 528 0.013; unpaired t-test). The peak contralateral suppression in the WT mice with 82 dB SPL noise (65 dB 529 SPL drivers around 28 kHz) was -3.5 ± 1.3 dB, measured at 6 s after noise onset ( Figure 7D). There was     and respond to sound stimulation, this single labelled type II SGN was 'silent' (Robertson, 1984 which may support GABAergic transmission as part of a reciprocal microcircuit in the early post-natal 770 period in rats (Weisz, 2020). Changes in these cochlear connections could potentially be a factor.

771
Any of these potential mechanisms demonstrate a critical role for peripherin in afferent feed to 772 contralateral suppression. However, in the absence of evidence supporting alternatives, we feel that 773 parsimony favors loss of OHC -type II SGN input as the root cause of loss of contralateral suppression.

774
Parenthetically, a mouse study that selectively reduced IHC -type I afferent input by treating the cochlear In summary, these findings suggest that from moderately loud sound levels, the cochlear OHC-type II 798 SGN afferent pathway contributes to MOC efferent regulation of the cochlear amplifier that confers 799 protection from noise-induced hearing loss.