MET currents and otoacoustic emissions from mice with a detached tectorial membrane indicate the extracellular matrix regulates Ca2+ near stereocilia

The aim was to determine whether detachment of the tectorial membrane (TM) from the organ of Corti in Tecta/Tectb−/− mice affects the biophysical properties of cochlear outer hair cells (OHCs). Tecta/Tectb−/− mice have highly elevated hearing thresholds, but OHCs mature normally. Mechanoelectrical transducer (MET) channel resting open probability (Po) in mature OHC is ∼50% in endolymphatic [Ca2+], resulting in a large standing depolarizing MET current that would allow OHCs to act optimally as electromotile cochlear amplifiers. MET channel resting Po in vivo is also high in Tecta/Tectb−/− mice, indicating that the TM is unlikely to statically bias the hair bundles of OHCs. Distortion product otoacoustic emissions (DPOAEs), a readout of active, MET‐dependent, non‐linear cochlear amplification in OHCs, fail to exhibit long‐lasting adaptation to repetitive stimulation in Tecta/Tectb−/− mice. We conclude that during prolonged, sound‐induced stimulation of the cochlea the TM may determine the extracellular Ca2+ concentration near the OHC's MET channels.


Introduction
The tectorial membrane (TM) is a strip of extracellular matrix that lies atop of the organ of Corti (Fig. 1A). It is attached medially to the spiral limbus and laterally to the tips of the hair bundles of the outer hair cells (OHCs), electromotile cells that can amplify the motion of the basilar membrane at low sound pressure levels and sharpen its frequency selectivity (Dallos, 1992;Ashmore, 2018). The electromotility of OHCs, the rapid voltage-dependent contractions of the basolateral membrane uniquely due to the presence of the motor protein prestin, a modified anion exchanger (Zheng et al. 2000;Liberman et al. 2002), form the basis of the so-called cochlear amplifier. The TM is known to be essential for normal hearing and has been proposed to play a number of roles (Lukashkin et al. 2010). There is evidence that it acts as an inertial mass that can influence the timing and gain of the cochlear amplifier (Mammano & Nobili, 1993;Gummer et al. 1996;Legan et al. 2000). Furthermore, it ensures that the stereociliary bundles of the inner hair cells (IHCs), which are not coupled directly to the TM, are displaced by fluid flow in the sub-tectorial space (Legan et al. 2005;Nowotny & Gummer, 2006). More recently, the TM has been shown to stabilize the cochlear amplifier (Cheatham et al. 2018), and that it may act as a source of Ca 2+ required for mechanoelectrical transduction and thus able to modulate the ionic environment around the hair-cell stereocilia (Strimbu et al. 2019).
In Tecta ENT/ ENT mice with a functional-null mutation in TECTA, a major non-collagenous component of the TM, the TM fails to form correctly and is no longer associated with the apical surface of the organ of Corti ( Fig. 1B; see also Legan et al. 2000). The hearing thresholds in Tecta ENT/ ENT mutant mice are considerably elevated compared with those recorded in wild-type mice. Importantly, cochlear microphonics (CMs), a measure of the OHC receptor potential, were found to be phase-shifted and asymmetrical in form in Tecta ENT/ ENT mice. This suggested that the hair bundles of OHCs were responding to fluid flow rather than displacement in the absence of an attached TM, and no longer sitting with ∼50% of their mechanoelectrical transducer (MET) channels open at rest (Legan et al. 2000). These observations also led to the suggestion that the TM normally serves to bias the OHC hair bundles by shifting their position towards the excitatory direction, such that they operate around the mid-point of the relationship between bundle displacement and the amplitude of the receptor potential.
Thus far, however, the properties of the MET channels, the basolateral potassium conductance and prestin-based electromotility in the hair cells from mice lacking a functional TM have not been described and compared with those in wild-type mice. Therefore, the cellular origins of the observed hearing loss in these mice remain uncertain. In this study we used mice homozygous for null mutations in Tecta and Tectb (referred to as Tecta/Tectb −/− mice) that, like the Tecta ENT/ ENT mice, no longer have a TM in contact with the organ of Corti (Fig. 1B). OHC development is normal in the absence of a functional TM, and MET channels have a resting open probability of ∼50% in the presence of an extracellular concentration of Ca 2+ similar to that found in the cochlear endolymph in vivo (40 μM, referred to hereafter as endolymphatic-like Ca 2+ ). This large open probability ensures the optimal activation of the prestin. Analysis of auditory function in vivo revealed that long-lasting adaptation of distortion product otoacoustic emissions (DPOAEs) is completely absent in the Tecta/Tectb −/− mice, suggesting that the MET current in OHCs fails to adapt in the absence of TM. Therefore, TM may be responsible for regulating the Ca 2+ concentration at or near the MET channels, thereby allowing adaptation of hair-cell MET currents during prolonged low-level repetitive auditory stimulation.

Ethics statement
In the UK, experiments were performed in accordance with Home Office regulations under the Animals (Scientific Procedures Act) 1986 (PPL_PCC8E5E93) and following approval by the Ethical Review Committees of the Universities of Sheffield (180626_Mar) and Sussex. In Germany, care and use of the animals and the experimental protocol were reviewed and approved by the University of Tübingen, Veterinary Care Unit and the Animal Care and Ethics Committee of the regional board of the Federal State Government of Baden-Württemberg, Germany (permission number AZ 35/9185.82-2 §8a Abs.1 dated 21.07.16), and followed the guidelines of EU Directive 2010/63/EU for animal experiments.

Generation of Tecta/Tectb double knockout mouse (Tecta/Tectb −/− )
A targeting vector, which was designed and constructed by Vector Biolabs (Eagleville PA, USA) for deleting Tecta and expressing a Tectb-IRES-Egfp minigene under the control of the endogenous Tecta promotor, was generated using a combination of PCR and conventional cloning techniques. The vector consisted of a 2832 bp left arm with the ATG start codon of the Tecta open reading frame (ORF) at the 3' end fused via a PmeI linker to the Tectb ORF, followed by an IRES, the Egfp ORF, an SV40 polyadenylation signal sequence, a neomycin resistance cassette flanked with loxP sites, a 4903 bp right arm from the Tecta gene beginning 314 bp 3' of exon 2 and a thymidine kinase cassette. The Tecta arms were prepared from a 129SvEvBrd genomic DNA clone as described previously (Legan et al. 2000). For the generation of transgenic mouse line Tecta tm6Gpr , embryonic stem cells were transfected with I-CeuI linearized targeting vector and resistant colonies selected as described (Legan et al. 2000). Individual colonies were picked and screened by Southern blotting and correctly targeted clones identified. Transgenic mice were prepared by microinjection of mouse blastocysts and a chimeric male founder was crossed with a wild-type S129SvEv female. Offspring carrying the insertion were then crossed with a beta-actin Cre line to remove the floxed neomycin selection cassette and, once it had been established that the selection cassette had been deleted, the offspring were characterized.
Initial characterization of the Tecta tm6Gpr transgenic line showed that while TECTA protein cannot be detected in mice homozygous for this allele, Tectb is not expressed under control of the Tecta promoter as originally intended because the PmeI linker introduces a frame shift, placing the entire Tectb coding sequence out of frame with the Tecta start codon. EGFP is, however, still expressed from the IRES under the control of the Tecta promoter and the spatial-temporal pattern of EGFP expression in the developing cochlea is very similar to that previously described with in situ hybridization using antisense probes for Tecta (Rau et al. 1999). The Tecta tm6Gpr mouse, a Tecta null mutant mouse expressing EGFP at the Tecta locus, was subsequently crossed with the Tectb tm1Gpr null mutant mouse line to produce a Tecta tm6Gpr/tm6Gpr , Tectb tm1Gpr/tm1Gpr double-null mutant mouse line that is referred to in this paper as the Tecta/Tectb −/− double knockout mouse. Mice were bred onto a C57BL6/N background for at least five generations and wild-type C57BL/6N mice were used as controls.
Genotyping for the Tecta tm6Gpr allele was done by PCR with KAPA2G Hot Start DNA polymerase (Sigma-Aldrich, UK) using primers TectbMGGF1 (CTCCCTGATAACCTACACTTC) and MmTectaEX1R1 (GAGCATGCTGATCAAGAGCTGTAGG) to amplify a wild-type product of 351 bp, and primers TectbMGGF1 and TectbMGGR1 (AACACAAGGATGACATCTGC) to amplify a mutant product of 339 bp. When resolved on a 1.5% agarose gel in 1 × TBE buffer a single 351 bp band indicates a Tecta +/+ genotype, two bands of 351 bp and 339 bp indicates a Tecta +/tm6Gpr genotype and a single 339 bp band indicates a Tecta tm6Gpr/tm6Gpr genotype. Genotyping for the Tectb tm1Gpr allele was done by PCR with Fast Start Taq DNA polymerase (Roche, UK) in the presence of uracil-DNA glycosylase (Roche) and nucleotide mixes containing dUTP using primers MbKOF2 (GATTCAAGTGGTAACTGAGCTTCC) and MbKOR1 (GGCCAGGTCGCGATTGTTCTGTATC) to amplify a wild-type product of 376 bp, and primers MbKOF2 and PGKR9 (TGCACGAGACTAGTGAGACGTGCTA) to amplify a mutant product of 550 bp. When resolved on a 1.5% agarose gel in 1 × TBE buffer a single 376 bp band indicates a Tectb +/+ genotype, two bands of 550 bp and 376 bp indicates a Tectb +/tm1Gpr genotype and a single 550 bp band indicates a Tectb tm1Gpr/tm1Gpr genotype.

Single-hair cell electrophysiology
Tissue preparation. Apical-coil OHCs from wild-type and Tecta/Tectb −/− double knockout mice of either sex were studied in acutely dissected organs of Corti J Physiol 599.7 from postnatal day 7 (P7) to P31, where the day of birth is P0. After killing the mice using a Home Office-approved Schedule 1 method (cervical dislocation), cochleae were rapidly dissected and kept in the following extracellular solution (in mM): 135 NaCl, 5.8 KCl, 1.3 CaCl 2 , 0.9 MgCl 2 , 0.7 NaH 2 PO4, 5.6 D-glucose, 10 Hepes-NaOH, 2 sodium pyruvate. Eagle's minimum essential medium amino acid solution (X50, without L-Glutamine) and vitamin solution (X100) were added from concentrates (ThermoFisher Scientific, UK); pH was adjusted to 7.5, ∼308 mOsmol kg −1 . Dissected cochleae were transferred to a microscope chamber, immobilized using a nylon mesh fixed to a stainless steel ring (Marcotti et al. 2003) and continuously perfused with the above extracellular solution. The sensory epithelia were viewed using an upright microscope (Olympus, Japan; Leica, Germany) with Nomarski differential interference contrast optics (X60 or X63 water immersion objectives and X10 or X15 eyepieces). All recordings were performed at room temperature (∼22°C) unless otherwise stated.
Whole-cell patch clamp. Voltage and current recordings were performed using an Optopatch amplifier (Cairn Research Ltd, UK). Patch pipettes, with resistances of 2-3 MΩ, were pulled from soda glass capillaries and the shank of the electrode was coated with surf wax (Mr Zoggs Sex Wax, CA, USA). Basolateral currents were measured using the following intracellular solution (in mM): 131 KCl, 3 MgCl 2 , 1 EGTA-KOH, 5 Na 2 ATP, 5 Hepes-KOH, 10 sodium phosphocreatine (pH 7.3). For MET recordings, the intracellular solution contained (in mM): 106 L-glutamic acid, 20 CsCl, 10 Na 2 phosphocreatine, 3 MgCl 2 , 1 EGTA-CsOH, 5 Na 2 ATP, 5 HEPES and 0.3 GTP (adjusted to pH 7.28 with 1 M CsOH; 294 mOsmol kg −1 ). An L-glutamic acid-based intracellular solution was used as it preserves cellular ultrastructure and improves the stability of recordings (Kay, 1992). A similar solution has been used extensively for investigating the biophysical properties of mammalian cochlear hair cells (e.g. Corns et al. 2018;2020;Jeng et al. 2020a,b). Data acquisition was performed using pClamp software (Molecular Devices, USA) using a Digidata 1440A. Data were lowpass filtered at 5 kHz (8-pole Bessel). Offline data analysis was performed using Origin 2019 software (OriginLab, USA). Membrane potentials were corrected for the residual series resistance (R s ) after compensation, and liquid junction potential (K + -and Cs + -based intracellular solution: −4 mV and −11 mV measured between electrode and bath solution, respectively). For some experiments, a gravity-fed local perfusion system was used to apply solutions with different extracellular Ca 2+ concentrations (500 μM Ca 2+ and endolymphatic-like 40 μM Ca 2+ ) either alone or with 200 μM of the MET channel blocker dihydrostreptomycin (DHS, Sigma, UK) (Marcotti et al. 2005).
Hair bundle stimulation. MET currents were elicited using a fluid jet from a pipette driven by a 25 mm diameter piezoelectric disc (Kros et al. 1992;Corns et al. 2014;2018). The fluid-jet pipette tip had a diameter of 8-10 μm and was positioned at about 8 μm from the hair bundles to elicit a maximal MET current. Mechanical stimuli were applied as steps or 50 Hz sinusoids.
Electromotile response. Electromotility was estimated in OHCs at room temperature (∼22°C) by applying a depolarizing voltage step from the holding potential of −64 mV to +56 mV and recorded using a CCD camera (Thorlabs DCU224M). The camera was attached to a microscope (Olympus), equipped with a X60 water immersion objective (Olympus LUMPlanFL N). The acquired images were stack-sliced along a vertical axis of each OHC and the contraction was measured on the image stack as length change of the cell. All images were analysed in ImageJ and the measurements were calibrated using a stage graticule (10 μm = 130 pixels).

Non-linear membrane
capacitance. Nonlinear (voltage-dependent) capacitance was measured from P18 and P24 OHCs using whole-cell patch-clamp recordings. In order to block most of the ion channels in hair cells, the intracellular solution in the pipette contained (in mM): 125 CsCl, 3 MgCl 2 , 1 EGTA-CsOH, 5 Na 2 ATP, 5 Hepes-CsOH, 5 tetraethylammonium (TEA), 5 4-aminopyridine (4-AP) (pH was adjusted with CsOH to 7.28; 290 mOsmol kg −1 ). Real-time changes in nonlinear membrane capacitance (C N-L ) were investigated using the capacitance tracking-mode of the Optopatch amplifier (Cairn Research Ltd, UK) during the application of a 4 kHz sine wave of 13 mV RMS. From the holding potential of −84 mV, hair cells were subjected to a voltage ramp from −154 mV to +96 mV over 2 s. The capacitance signal from the Optopatch was lowpass filtered at 250 Hz and sampled at 5 kHz.

Immunofluorescence microscopy
Inner ears from wild-type and Tecta/Tectb −/− mice (n = 4 for each experiment) were removed by dissection and fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS, pH 7.4) for 20 min at room temperature. Cochleae were microdissected, rinsed three times for 10 min in PBS, and incubated for 1 h at room temperature in PBS supplemented with 5% horse serum (HS) and 0.5% Triton X-100. The samples were then incubated overnight at 37°C with the primary antibody in PBS supplemented with 1% HS. Primary antibodies were: mouse anti-myosin 7a (1:1000, Developmental Studies Hybridoma Bank, #138-1C), rabbit anti-myosin 7a (1:200, Proteus Biosciences, #25-6790), rabbit anti-prestin (1:5000, kindly provided by Robert Fettiplace), rabbit anti-SK2 (1:500, Sigma-Aldrich, P0483) and goat anti-choline acetyltransferase (ChAT, 1:500, Millipore, AB144P). All primary antibodies were labelled with species-appropriate Alexa Fluor secondary antibody for 1 h at 37°C. Samples were then mounted in VECTASHIELD. The z-stack images were captured with a Nikon A1 confocal microscope equipped with Nikon CFI Plan Apo 60X Oil objective in the Light Microscope Facility at the University of Sheffield. Image stacks were processed with Fiji Image Analysis software (ImageJ, NIH and LOCI Laboratory of Optical and Computational Instrumentation, USA).

Toluidine blue staining
After glutaraldehyde fixation, cochleae were washed three times in 0.1 M sodium cacodylate buffer pH 7.2 and post-fixed in 1% osmium tetroxide in 0.1 M sodium cacodylate buffer for 3 h at room temperature. Samples were then washed three times in sodium cacodylate buffer and decalcified in 0.5 M EDTA pH 8.0 containing 0.1% glutaraldehyde for 3 days at 4°C. Samples were then washed briefly in water, dehydrated through an ascending ethanol series, equilibrated in propylene oxide and embedded in epoxy resin (TAAB 812). Blocks were cured at 60°C for 24 h and trimmed with a glass knife after which semi-thin 1 micron sections were cut on a Reichert Ultracut E ultramicrotome using a histo-grade Diatome diamond knife. Sections were dried onto glass slides and stained briefly with Toluidine blue before viewing on a Zeiss Axioplan 2 wide-field microscope. Images were captured using a Jenoptik ProgRes C3 CCD camera.

Scanning electron microscopy (SEM)
Cochleae were fixed by perfusing the cochlea with 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.2), and immersed in the same fixative overnight. Cochleae were then washed three times in 0.1 M sodium cacodylate buffer pH 7.2 and decalcified in 0.5 M EDTA pH 8.0 for 2-3 days at 4°C. Pieces of organ of Corti were then dissected in 0.1 M sodium cacodylate following the method described previously for cochlear wholemounts (Legan et al. 2014). Samples were then post-fixed in 1% osmium tetroxide for 3 h at room temperature, washed in cacodylate buffer and dehydrated through a series of ascending concentrations of ethanol. Following critical-point drying, samples were mounted on SEM stubs and sputter-coated with platinum before viewing in a Jeol JSM-6700F SEM operating at 5 kV.

In vivo hearing tests
Hearing tests. All recordings were performed on anaesthetized mice in a soundproof chamber (IAC, Niederkrüchten, Germany) as previously described (Knipper et al. 2000). In short, a multi-function IO-Card (PCI-6052E, National Instruments, USA) housed in a PC was used for stimulus generation and recording. Stimuli were delivered open field to the ear by loudspeakers either placed 3 cm lateral to the animal's pinna or as a closed field for otoacoustic measurements. Sound pressures were controlled with attenuators and amplifiers (Wulf Elektronik, Germany) and calibrated online prior to each measurement.

Auditory brainstem responses (ABR).
ABRs evoked by short-duration sound stimuli represent the summed neuronal activity along the auditory pathway (e.g. Möhrle et al. 2016). Briefly, ABRs were evoked by gated clicks, noise bursts, and pure tone stimuli of gradually increasing sound pressure in 5 dB steps at a repetition rate of 60/s. Clicks (100 μs) and noise-burst stimuli (1 ms random phase frozen noise voltage signal) were rectangularly gated to produce sound signals emitted from the speaker dominated by a spectral peak at 7.9 kHz with a plateau of energy up to 35 kHz and a 30 dB roll-off from 35 to 65 kHz. Compared with the click stimulus, the noise-burst stimulus contained more energy at higher frequencies (>10 kHz). Pure tones were cosine square J Physiol 599.7 (cos 2 ) gated (3 ms duration, 1 ms rise/fall times) and presented for frequencies 2-45 kHz with two steps per octave. Stimuli were presented in open field by a loudspeaker (DT-911, Beyerdynamic, Heilbronn, Germany) and calibrated online with a microphone (B&K 4191,Brüel & Kjaer,Denmark) placed near the ear. The recorded signal was amplified (100 dB), filtered (0.2-5 kHz) and responses of alternating phase or polarity were summed to exclude the stimulus artefact and CMs. ABR thresholds were determined as the minimal sound pressure level able to induce a response that was identified by visual inspection of the averaged signal (Rüttiger et al. 2013;Möhrle et al. 2017).
Electrocochleographic recordings. We studied electrical potentials of cochlear OHCs and auditory nerve fibres by electrocochleography in living anaesthetized mice (wild-type: four males and five females; Tecta/Tectb −/− : two males and 10 females; aged 2-6 months, average age 3.7 and 3.4 months for wild-type and Tecta/Tectb −/− mice, respectively). Age and sex of the mice did not influence the electrical response. For these experiments, the surgery was done as previously described (Zampini et al. 2011). Briefly, mice were anaesthetized as described above (Anaesthesia) and 20-40 μl of 2% Xylocain (AstraZeneca, Wedel, Germany) were applied subcutaneously at sites of surgical cuts. The bony auditory bulla was exposed by cutting the skin behind the ear and carefully traversing muscles, nerves and connective tissues avoiding displacing. A small hole (0.6 mm diameter) was drilled into the bulla and the round window niche of the cochlea visualized using an ultra-low speed drill rotation (down to two turns/s) in order to avoid any vibration-induced damage to the cochlea. A silver wire electrode insulated by varnish and silicone ending in a small silver bead was placed within the niche. The skin above the ear was closed and the mouse placed in the sound-attenuating booth in front of a loudspeaker for recording.
Compound action potential (CAP) threshold responses from the auditory nerves were determined by stimulation with short-tone pips (3 ms duration, 1 ms rise/fall times cos 2 gated, 32-96 repetitions with stimulus interval 16 ms and alternating polarity) presented with 5 dB incremental steps from 0 to 110 (wild-type) or 40 to 120 dB SPL (Tecta/Tectb −/− ) between 2 and 64 kHz. Electrical potentials were amplified (80 dB) and filtered between 0.2 and 5 kHz before being sampled at 20 kHz A/D rate, averaged, and saved to file. Thresholds were determined from individual ears from averaged waveform responses as the lowest SPL resulting in a signal visually distinguishable from noise. CAP amplitude growth. Electrical responses were recorded for 20 ms long 18 kHz pure tone stimuli (0.2 ms cos 2 gated) of 0-100 dB SPL (for wild-type mice) or 20-120 dB SPL (for Tecta/Tectb −/− mice). Responses were amplified, filtered (DC, 50 kHz low pass), sampled at 100 kHz A/D rate, and averaged for 64 repetitions (ISI 50 ms). For CAP input-output analysis, the averaged waveform was digitally filtered (0.2-5 kHz, 6-pole, Hanning window, DC removed and phase corrected by twofold filtering with original and time-inverted time-amplitude series) and inspected for lowest and highest amplitude deflections (peaks) within a time window of 1 to 6 ms after stimulus onset. The peak-to-peak amplitude delineates the amount of synchronously active fibres within the auditory nerve as a CAP. The growth of the CAP is registered for each individual ear and the resulting growth functions from all ears averaged and presented as mean and standard deviation (SD).
Cochlear microphonic peak responses. CM signals, which are extracted from the electrocochleographic recordings, are mostly symmetric, fast potential changes from OHCs following the auditory stimulus pressure changes. Resulting average traces were analysed for their positive and negative peak of deflection within a 5 ms time window at 14-19 ms after stimulus onset (steady-state response). Resulting peak values were averaged for positive and negative peaks separately and displayed as a function of pressure (in Pascal).
Distortion product otoacoustic emissions. OHC function was assessed by the response strength and response threshold from the growth function and the distortion product audiogram of the cubic DPOAE in response to synchronous presentation of two stimulus tones (primaries 1 and 2). The cubic 2f1-f2 distortion product of the DPOAE for primary-tone frequency (f) f2 = 1.24 × f1 and primary-tone level (L) L2 = L1-10 dB were recorded in the soundproof chamber as previously described (Engel et al. 2006). Pairs of tones with frequencies between f2 = 4 kHz and 32 kHz were presented directly into the ear canal by means of a metal coupler connected to two loudspeakers (DT-911, Beyerdynamic). The emission signals were recorded by a microphone (MK 231, Microtech, Gefell, Germany;Preamplifier Brüel & Kjaer 2670, Naerum, Denmark) connected to the coupler. Emission signals were recorded during sound presentation of 262 ms and averaged four times for each level and primary-tone frequency. For the distortion product audiogram, the 2f1-f2 distortion product amplitude was measured at constant L2 of 50 dB SPL with pairs of primaries with frequencies between f2 = 4 kHz and 32 kHz in four steps per octave. The growth function of the 2f1-f2 distortion product amplitude was measured for L1 ranging from -10 to 65 dB SPL with pairs of primary tones with frequencies between f2 = 4 kHz and 32 kHz in half-octave steps. DPOAE thresholds were defined as the L1 level that could generate a 2f1-f2 signal reliably exceeding about 5 to 10 dB above noise level with noise level typically at −20 dB SPL. DPOAE adaptation. During repeated loud sound stimulation, DPOAE can exhibit both short-lasting (fast) adaptation, which occurs within the 100 ms of the stimulus presentation (Kujawa & Liberman, 2001), and long-lasting adaptation that lasts more than minutes following the stimulus in rats (Zhao et al. 2018) and humans (Narahari et al. 2017). Fast DPOAE adaptation is believed to be mediated by cholinergic medial olivocochlear (MOC) efferent neurons (Robertson & Gummer 1985;Maison et al. 2003) and middle-ear muscle reflexes (MEMR: Horner, 1986). In order to elicit DPOAE adaptation in the ipsilateral ear, we repetitively exposed the ear to the two stimulus tones (primaries) with the frequencies f2 = 11.3 kHz and f1 = 9.11 kHz and with L1 = 60 dB SPL and L2 = 50 dB SPL for primaries 1 and 2, respectively (Wolter et al. 2018). Primaries were switched on synchronously, maintained for 100 ms and presented 384 times with 350 ms recording intervals. Onset and offset were cos 2 gated with a 2 ms ramp. The aim of our work was to investigate the intrinsic activity of OHCs (by means of DPOAE), while avoiding the high stimulus levels that would activate the MOC and/or MEMR or induce adaptation effects by sound conditioning (for rabbits at ca 85 dB SPL: Luebke et al. 2015). Therefore, we first determined the DPOAE I/O growth function for f2 = 11.3 kHz and L1 = −10 to 65 dB SPL (non-adapted situation).
After inducing DPOAE adaptation, the DPOAE growth function for f2 = 11.3 kHz was measured again. A change in the form of the f2 = 11.3 kHz DPOAE growth function indicates whether DPOAE responses had adapted to the louder stimulus.
Data handling and presentation. Stimulus presentation, potential recording and filing was done by custom-made computer programs based on LabWindows-CVI (National Instruments) cards and libraries (CAP.exe, University of Tübingen). For single-sweep and averaged responses, data were filtered and processed by custom-made analysis computer programs (SingleSWEEP.exe and PEAK.exe, University of Tübingen). Results were exported to text files, arranged and visualized with Microsoft Excel.

Statistical analysis
For in vitro recordings, statistical comparisons of means were made by Student's two-tailed t test or, for multiple comparisons, using one-way ANOVA. Mean values are quoted ± SD, where P < 0.05 indicates statistical significance.
For in vivo recordings, data were compared for statistical difference by means of 2-way ANOVA (Graphpad Prism, San Diego, USA) and post hoc testing, with alpha-level corrected for multiple testing following the Bonferroni-Holms method or the Fisher's exact probability test for categorical frequency 2 × 2 contingency tables. For the frequency-dependent ABR and DPOAE data, and the waveform analysis, we used 2-way ANOVA and a post hoc Holm-Sidak's multiple comparison test. For click and noise burst stimulated ABRs, 2-sided Student's t test was used. In figures, statistical significance for pairwise comparisons are indicated by asterisks. n.s. denotes statistically non-significant results. Mean values are quoted ± SD, where P < 0.05 indicates statistical significance.

TM is detached from the organ of Corti in Tecta/Tectb −/− mice
Toluidin blue-stained semi-thin sections and SEM were used to compare the structure of the organ of Corti and hair cell stereociliary bundles in wild-type (Tecta/Tectb +/+ ) and Tecta/Tectb double knockout (Tecta/Tectb −/− ) mice (Fig. 1C). In wild-type mice the TM is attached medially to the spiral limbus and extends laterally, lying over the apical surface of the sensory hair cells (Fig. 1a). In the Tecta/Tectb −/− mice, a residual TM is present but it is no longer associated with the lumenal surface of the organ of Corti and is instead attached to Reissner's membrane (Fig. 1D). The structure, organisation and orientation of the hair bundles of the inner (not shown) and outer hair cells in the wild-type (Fig. 1E) and the Tecta/Tectb −/− (Fig. 1F) mice are very similar. Material assumed to be derived from the TM is, however, often seen associated with the tips of the tallest row of stereocilia in the hair bundles of the OHCs in the wild-type (Fig. 1E, arrows), but not in the Tecta/Tectb −/− mice (Fig. 1F). The structural phenotype of the cochlea in Tecta/Tectb −/− mice is therefore very similar to that described previously for mice homozygous for a functional-null mutation ( ENT) in Tecta (Tecta ENT/ ENT : Legan et al. 2000).

Tecta/Tectb −/− mice have a low-to mid-frequency hearing loss
The hearing in Tecta/Tectb −/− mice was investigated using ABRs, a measure of the activity of the auditory neurons downstream of IHCs. Compared with wild-type mice (Figs. 2A-2E), aged-matched Tecta/Tectb −/− mice had significantly higher ABR thresholds for click and J Physiol 599.7 noise-burst stimuli (Figs. 2A, 2B), and for pure-tone stimuli ranging from 2 to 22.6 kHz (Fig. 2C). ABR thresholds were not detected in Tecta/Tectb −/− mice for pure-tone stimulus frequencies below 4 kHz at the highest stimulus level tested (110 dB SPL). At higher pure-tone stimulus frequencies (32 and 45.3 kHz), the ABR thresholds of Tecta/Tectb −/− mice were not significantly higher than those measured in wild-type mice (Fig. 2C). ABR thresholds are therefore raised in the most sensitive region of the hearing range in Tecta/Tectb −/− mice, but are comparable to those of wild-type mice at high frequencies (Fig. 2C). Extracellular recordings made from the round window of the cochlea in response to pure-tone stimuli reveal that the thresholds of the CAP are increased across the entire frequency range in the Tecta/Tectb −/− mice (Fig. 2D). Thresholds are elevated by up to 60 dB in the mid-frequency range in the Tecta/Tectb −/− mice and, above threshold, the CAP waveform amplitude grows slowly, relative to that in wild-type mice (Fig. 2E). These findings are in accordance with those from the Tecta DENT/DENT mouse and confirm that the reduced ABR sensitivity of Tecta/Tectb −/− mice occurs primarily at the level of the organ of Corti.  (Legan et al. 2000). Abbreviations: TM: tectorial membrane, OHCs: outer hair cells, IHC: inner hair cell. Red and orange lines indicate the medial efferent and type II afferent fibres, respectively, making contacts with the OHCs. Blue lines indicate the type I afferent fibres making contacts with the IHCs. Violet lines indicate the lateral efferent fibres making contact with the type I afferent terminals. Cells depicted in grey represent the non-sensory cells within the organ of Corti. Note that the TM is in intimate contact with the tips of the tallest-row of stereocilia of the OHCs of wild-type (A) but not in Tecta −/− (B) mice. C and D, Toluidine blue-stained semi-thin plastic sections from the apical coil of the cochlea of P21 wild-type and double knockout (Tecta/Tectb −/− ) mice, respectively. The tectorial membrane (TM) is attached to the spiral limbus in the wild-type (C, arrow) but completely detached in the mutant (D, arrow). Scale bars: 50 μm. E and F, scanning electron microscopy images of the stereociliary bundle of mid-apical OHCs from P40 wild-type (E) and Tecta/Tectb −/− (F) mice. Additional material is often associated with the tips of the taller row of stereocilia in the wild-type (arrows, E) but not with the Tecta/Tectb −/− OHC bundles (F). Scale bars: 2 μm.

Cochlear microphonics indicate symmetric OHC receptor potentials
CM potentials were recorded from the round window in response to an 18 kHz tone ranging in intensity from 50 to 100 dB SPL in the wild-type mice, and from 70 to 120 dB SPL in the Tecta/Tectb −/− mice (Fig. 3A,B). Despite the variability in the phase of the response relative to the stimulus, the averaged waveform of CM in the Tecta/Tectb −/− mice is symmetrical and similar to that of the wild-type. However, we noted a progressive phase shift with increasing stimulus level in the averaged response from theTecta/Tectb −/− mice (Fig. 3A,B), which was evident towards the later cycles within the first 0.5 ms of the stimulus onset. This indicates that, in Tecta/Tectb −/− mice, OHC hair bundles are likely to be fluid coupled to the stimulus. When the peak amplitudes of the CM waveforms measured at steady state 14-19 ms after stimulus onset are averaged across all ears of the same genotype and plotted as a function of pressure, the CM waveform was found to be symmetric in both genotypes (Fig. 3C,D). Furthermore, these data reveal that, the OHCs in both the wild-type and the Tecta/Tectb −/− mice are operating around the steepest point on their input-output (transfer) functions. For examples of CM waveforms from individual animals, and the peak CM responses as function of pressure for all animals used to

Figure 2. Auditory brainstem responses (ABRs) and compound action potentials (CAPs) in Tecta/Tectb mice
A and B, ABR thresholds evoked with a broadband click stimulus (A) and a noise-burst stimulus (B) from wild-type (8 mice, 16 ears) and Tecta/Tectb −/− (8 mice, 15 ears) mice. Thresholds from single ear measurements are shown as dots superimposed on the average bar graph data. * P < 0.0001 in both A and B (2-sided Student's t test). C, ABR thresholds for a series of frequency-specific pure-tone stimuli from wild-type (8 mice, 12 ears) and Tecta/Tectb −/− (8 mice, 15 ears) mice. Note that ABR thresholds increase across most of the frequency hearing range (P < 0.0001, two-way ANOVA; for Sidak's post hoc test see Statistical Summary). For two of the ears tested in the Tecta/Tectb −/− mice, at 32-45 kHz ABR threshold could not be detected even with the highest stimulus levels. D, average CAP thresholds for frequency-specific pure-tone stimuli from wild-type and Tecta/Tectb −/− mice (P < 0.0001, 2.8-45.2 kHz range, two-way ANOVA; for Sidak's post hoc test see Statistical Summary). E, growth of CAP responses across stimulus levels in wild-type and Tecta/Tectb −/− mice. In Tecta/Tectb −/− mice, CAP I/O function for 18 kHz stimuli above threshold was significantly different compared to that recorded in wild-type mice (P < 0.0001, two-way ANOVA; for Sidak's post hoc test see Statistical Summary). For the single data points in panel E, see Supplementary Data Set. Data are means ± SD. J Physiol 599.7 Figure 3. Cochlear microphonic signals in wild type and Tecta/Tectb −/− mice A, cochlear microphonic (CM) responses to 18 kHz pure-tone stimuli (gated with 0.2 ms cos 2 ramp at onset) of increasing level averaged from 10 wild-type ears (top, black) and 18 Tecta/Tectb −/− ears (bottom, red). Stimulus levels ranged from 50-100 dB SPL for the wild-type mice, and from 70-120 dB SPL for the Tecta/Tectb −/− mice. Responses shown were recorded during the first 0.5 msec after onset of the stimulus. B, top three pairs of traces: averaged CM responses shown in panel A from wild-type (black) and Tecta/Tectb −/− (red) mice are compared at stimulus levels eliciting potentials of comparable amplitude, at 60 and 90, 75 and 100, 100 and 120 dB SPL, respectively. For a comparison of CM responses from two individual animals see Supplemental Data Fig. 3B. Bottom: the green sinewave depicts the voltage driving the speaker that is delivering the stimulus and is located 3.8 cm from the pinna; the orange sinusoid shows the acoustic sound wave reaching the inner ear. At stimulus levels >100 dB SPL, the driving voltage to the loudspeaker leads to significant electro-magnetic artifacts in the stimulus response that occur before the acoustic sound wave reaches the inner ear. construct the averages shown in Fig. 3C, see Supplemental Data set.

DPOAE signals indicate lowered sensitivity but regular response strength in Tecta/Tectb −/− mice
We measured DPOAEs to provide a better understanding of the functional consequences of the detached TM in Tecta/Tectb −/− mice. In Tecta/Tectb −/− mice, the amplitudes of the DPOAEs at 50 dB SPL stimulus level were greatly reduced (Fig. 4A) and the thresholds for evoking minimum DPOAE responses were increased to very high levels (Fig. 4B) compared with those in wild-type mice. DPOAE amplitude growth (I/O) function for stimuli with f2 = 11.3 kHz (the frequency with the best DPOAE thresholds) revealed significantly lower values for DPOAE signal strength for stimulation levels between 20 and 55 dB SPL in Tecta/Tectb −/− mice (Fig. 4C), although responses reached the levels of the wild-type mice above 55 dB SPL (Fig. 4C). As shown for the CM responses (Fig. 3), these results indicate that the stereociliary bundles of the OHCs are being stimulated by high-level sound simulation in the absence of an attached TM.
During repeated loud sound stimulation, DPOAE responses can exhibit both fast and long-lasting adaptation. Fast DPOAE adaptation, which can occur within 4 ms of the stimulus onset and can be elicited at levels of 50 dB SPL (Horner, 1986), is driven by neuronal feedback from the CNS via the MOC efferent system and the MEMRs. CNS feedback was not present in our recordings from either genotype (data not shown). In contrast to fast DPOAE adaptation, long-lasting adaptation can persist for minutes (Narahari et al. 2017;Zhao et al. 2018) and is not dependent on the feedback from the CNS. The characteristics of long-lasting DPOAE adaptation were investigated by exposing mice to 387 repetitions 100 ms long (inter-stimulus interval 350 ms) primary tones of 60 dB SPL over 2.25 min (135 s) (Figs. 4D-4F), a stimulus level that produced equally strong DPOAE responses in both genotypes (Fig. 4C). After presentation of the adapting stimulus for 2.25 min, the 2f1-f2 DPOAE response was found to be significantly reduced, compared with that obtained before exposure, in wild-type (Figs. 4D, 4F) but not in Tecta/Tectb −/− mice (Figs. 4E, 4F).

Tecta/Tectb deletion does not affect hair-cell maturation
In order to ascertain whether the observed hearing defects (Fig. 2) are only due to the absence of the TM (Fig. 1), and not due to defects in hair-cell development, we investigated whether IHCs (Fig. 5) and OHCs (Figs. 6, 7) in Tecta/Tectb −/− mice become mature, fully functional sensory receptors. Adult IHCs from wild-type mice express two characteristic K + currents, a rapidly activating, large-conductance, Ca 2+ -activated K + current I K,f , and a negatively activating delayed-rectifier K + current, I K,n (Figs. 5A-5C) (Kros et al. 1998;Jeng et al. 2020c), with the latter carried by KCNQ4 channels (Kubisch et al. 1999). Both I K,f and I K,n in IHCs from Tecta/Tectb −/− mice were not significantly different from those recorded in wild-type mice (Figs. 5D, 5E). Adult IHCs from Tecta/Tectb −/− mice also exhibited voltage responses (Fig. 5F) and resting membrane potentials (Fig. 5G) that were similar to those in wild-type cells. Similar to IHCs, mature OHCs express I K,n (Marcotti & Kros, 1999;Jeng et al. 2020a), which was present in Figure 4. DPOAE responses adapt in wild-type but not in Tecta/Tectb −/− mice A, maximal amplitude of DPOAE signals from DP-grams measured over a range of primary-tone frequencies f2 = 4-32 kHz at stimulation level L1 = 50 dB SPL in both wild-type (black: 8 mice, 16 ears) and Tecta/Tectb −/− (red: 8 mice, 15 ears) mice. Maximal DPOAE responses were present in both genotypes at frequencies f2 = 6.7-26.7 kHz, but were found to be significantly reduced in Tecta/Tectb −/− mice (t test, * P < 0.0001). B, DPOAE thresholds were significantly increased in Tecta/Tectb −/− compared with wild-type mice for most of the frequency range used for hearing in mice (P < 0.0001, two-way ANOVA; for Sidak's post hoc test 'asterisks' see Statistical Summary). C, growth function of DPOAE signals with increasing stimulus level indicates loss of active amplification in the closeto-threshold range in the Tecta/Tectb −/− mice, while high-level DPOAE signals reach amplitudes in a similar range as the wild-type mice (P < 0.0001, two-way ANOVA; for Sidak's post hoc test 'asterisks' see Statistical Summary). Lines: 95% confidence interval (CI 95 ) . D and E, DPOAE growth function for f2 = 11.3 kHz stimulated 2f1-f2 DPOAE signals using 60 dB SPL stimulation (384 × 100 ms long stimulus presentations, inter-stimulus interval 350 ms, within 2.25 min) in both wild-type (black: 6 mice, 11 ears) and Tecta/Tectb −/− (red: 4 mice, 8 ears) mice. Continuous black and red lines show the mean signals before stimulation and the CI 95 . Symbols show the signals directly after stimulation. Measured noise floor and critical threshold level of −15 dB SPL are illustrated by grey crosses and area, respectively. After stimulation, DPOAE signals were significantly reduced within 15−45 dB SPL stimulation in wild-type (P = 0.0060, two-way ANOVA, for Sidak's post hoc test see Statistical Summary) but not in Tecta/Tectb −/− (P = 0.7661) mice. Data are means ± CI 95 (before) and ± SD (after). F, comparison of 2f1-f2 adaptation in wild-type (black) and Tecta/Tectb −/− (red) mice. The difference in adaptation is significantly dependent on stimulus level (P both wild-type and Tecta/Tectb −/− mice (Figs. 6A-6C). Current-clamp experiments also revealed normal voltage responses in OHCs from both genotypes (Figs. 6D, 6E) with a resting membrane potential that was not significantly different in wild-type and Tecta/Tectb −/− mice (Fig. 6F). The membrane capacitance was also similar in OHCs from wild-type (11.8 ± 0.5 pF, n = 4) and Tecta/Tectb −/− mice (12.2 ± 0.9 pF, n = 5, P = 0.4706, t test). Mature OHCs are also the primary target of the cholinergic MOC neurons (Maison et al. 2003) which, by releasing the neurotransmitter acetylcholine (ACh) at their efferent terminals (Simmons et al. 1996), modulate OHC electromotility and therefore mechanical amplification in the adult cochlea (Guinan, 1996). Efferent inhibition of OHCs by ACh is caused by Ca 2+ influx through α9α10-nAChRs activating hyperpolarizing, small-conductance, Ca 2+ -activated K + channels (SK2 channels: Oliver et al. 2000;Katz et al. 2004;Marcotti Figure 5. The basolateral membrane properties of adult IHCs are indistinguishable between wild-type and Tecta/Tectb −/− mice A and B, currents recorded from IHCs of wild-type (black) and Tecta/Tectb −/− (red) post-hearing mice. Currents were elicited by using depolarizing voltage steps (10 mV increments) from the holding potential of −64 mV to the various test potentials shown by some of the traces. The arrow next to I K,n indicates the deactivating tail currents; while that next to I K,f indicates activation. C, steady-state current-voltage curves obtained from IHCs of wild-type (P22-P24) and Tecta/Tectb −/− (P22-P28) mice. For recordings from individual hair cells in panel C, see Supplementary Data Set. D and E, size of the isolated I K,f (D), which was measured at −25 mV and at 1 ms form the onset of the voltage step, and I K,n (E) that was measured as the difference between the peak and steady-state deactivating tail current at −124 mV; for a similar analysis see: Marcotti et al. 2003. I K,f size: wild-type 2.9 ± 0.8 nA, n = 5; Tecta/Tectb −/− 2.7 ± 1.1 nA, n = 4, P = 0.7650, t test. I K,n size: wild-type 153 ± 47 pA, n = 4; Tecta/Tectb −/− 208 ± 54 pA, n = 5, P = 0.1584. F, voltage responses recorded from adult IHCs of wild-type and Tecta/Tectb −/− mice elicited by applying hyperpolarizing and depolarizing current injections from their respective membrane potentials. G, average IHC resting membrane potential was not significantly different between wild-type (−74.5 ± 4.7 mV, n = 5) and Tecta/Tectb −/-(−71.1 ± 2.3 mV, n = 5, P = 0.1853,t test). Single cell value recordings (closed symbols) are plotted behind the average values. Number of IHCs investigated is shown above the average data points. Values are means ± SD. Recordings were performed at room temperature. et al. 2004). Immunolabelling experiments confirmed that efferent cholinergic terminals, as visualized by ChAT immunoreactivity, and SK2 channels were both present in the OHCs of both wild-type and Tecta/Tectb −/− mice (Fig. 6G).
As OHCs become functionally mature, from about P7 onwards, they acquire somatic motility (Marcotti & Kros, 1999;Abe et al. 2007), which is required for cochlear amplification and is driven by the motor protein prestin (SLC26A5: Zheng et al. 2000;Liberman et al. 2002). Voltage steps of 120 mV from the holding potential of −64 mV caused OHCs from both wild-type and Tecta/Tectb −/− mice to shorten (Figs. 7A, 7B) by an amount (Fig. 7C) similar to that measured previously in other strains (Marcotti & Kros, 1999;Abe et al. 2007). This was further investigated using non-linear (voltage-dependent) capacitance changes (C N-L ), an electrical signature of electromotility in OHCs. We found that the maximum size of C N-L in P18-P24 OHCs (Figs. 7D-7F) was comparable to that previously reported (e.g. Oliver & Fakler, 1999;Abe et al. 2007;Jeng et al. 2020a) and not significantly different in the two genotypes, consistent with the qualitative similar distribution of prestin observed in both genotypes (Fig. 7G).
These results show that OHCs and IHCs in Tecta/Tectb −/− mice mature normally and acquire the characteristic pattern of efferent innervation pattern, indicating that they do not require interactions with the TM to become fully functional receptors.

Mechanoelectrical transduction in mature OHCs from Tecta/Tectb −/− mice
Considering that the basolateral membrane properties of OHCs are indistinguishable in wild-type and Tecta/Tectb −/− mice, we sought to investigate whether the absence of DPOAE adaptation could be due to defects in the intrinsic characteristics of the MET apparatus. The structure of the hair bundles of P40 OHCs appeared comparable in wild-type and Tecta/Tectb −/− mice (Fig.  1E,F), and we were able to record large MET currents in OHCs from both genotypes by displacing their hair bundles in the excitatory and inhibitory directions using a 50 Hz sinusoidal force stimulus from a piezo-driven fluid jet (Corns et al. 2014;2018). At negative membrane potentials (−121 mV) in an extracellular solution containing 1.3 mM Ca 2+ , a large inward MET current could be elicited in P10 OHCs from both wild-type and Tecta/Tectb −/− mice by deflecting the bundles towards the taller stereocilia (i.e. in the excitatory direction) (Fig. 8A). The resting current flowing through open MET channels in the absence of mechanical stimulation Figure 6. Basolateral membrane currents in OHCs from Tecta/Tectb −/− mice A and B, currents recorded from a wild-type (P20) and a Tecta/Tectb −/− (P22) OHC, respectively. Cells were held at −84 mV, and currents were elicited by voltage steps (10 mV increments). The holding currents are plotted as zero. C, average peak current-voltage relation obtained from four wild-type (P20-24) and six Tecta/Tectb −/− (P22-23) OHCs. For recordings from individual hair cells in panel C, see Supplementary Data Set. D and E, voltage responses from mature OHCs of wild-type (D) and Tecta/Tectb −/− mice (E), elicited using a series of hyperpolarizing or depolarizing current injections (shown next to the traces). F, resting membrane potential in wild-type (P20-24: −74.3 ± 3.1 mV, n = 4) and Tecta/Tectb −/− (P22-27: −72.8 ± 4.5 mV, n = 5, P = 0.5790, t test) OHCs. G, maximum intensity projections of confocal z-stack images taken from mature apical-coil OHCs (P21) of wild-type and Tecta/Tectb −/− mice. Immunostaining for SK2 channels (green) and ChAT (red), which is used to visualize the efferent olivocochlear innervation of OHCs; Myo7a (blue) was used as the hair cell marker. Similar staining was seen in four additional mice for each genotype. Scale bars: 10 μm. J Physiol 599.7 was reduced when the bundles were moved towards the shorter stereocilia (i.e. in the inhibitory direction) in both wild-type and Tecta/Tectb −/− OHCs (Fig. 8A,  arrows). By stepping the membrane potential from −121 mV to +99 mV, the MET currents became outward during excitatory bundle stimulation, consistent with the non-selective permeability of the MET channels to cations. At positive potentials, which are near the Ca 2+ equilibrium potential and strongly reduce Ca 2+ entry via the MET channels, OHCs exhibited a larger resting MET current (Fig. 8A: arrowheads). This phenomenon is consistent with Ca 2+ entry driving adaptation as previously demonstrated in hair cells from lower vertebrates (Assad et al. 1989;Crawford et al. 1989;1991) and mammals (Corns et al. 2014;Marcotti et al. 2016).
In previous studies it has only been possible to record MET currents in vitro from wild-type mouse OHCs prior to the onset of hearing at stages <P12, possibly because the transduction apparatus is damaged during physical removal of the TM. However, when using preparations from Tecta/Tectb −/− mice in which the TM is no longer associated with the surface of the organ of Corti, we were able to record a large MET current in OHCs even after the onset of hearing (Fig. 8B: see also Jeng et al. 2020a). The maximal MET current at −121 mV in OHCs was plotted as a function of postnatal A and B, electromotile responses from a wild-type (P22) and a Tecta/Tectb −/− (P19) OHC, respectively. The images of the OHCs in the left panels and right panels are obtained before and after, respectively, the application of a voltage step of 120 mV from the holding potential of −64 mV. Stacked images from repeating steps (right panels) are shown along a vertical axis of the basal-pole of the OHCs. Red and green lines indicate the position of the cell before and after the application of the voltage, respectively. Scale bar: 5 μm. C, OHC contraction measured from wild-type (P22: 542 ± 81 nm, n = 5) and Tecta/Tectb −/− (P19-27: 532 ± 85 nm, n = 6, P = 0.8360, t test) OHCs. D, voltage-dependent non-linear capacitance (C N-L ) recorded from P18 OHCs and IHCs by applying a voltage ramp from −154 mV to +96 mV. Note that C N-L was present in the OHCs from both genotypes. E and F, average C N-L was similar between wild-type and Tecta/Tectb −/− apical-coil P18-P25 OHCs (E: wild-type 4.98 ± 0.84 pF, n = 7; Tecta/Tectb −/− 5.07 ± 1.94 pF, n = 7, P = 0.8979, t test), even after normalization to the average OHC membrane capacitance C m (F: wild-type 0.48 ± 0.08, n = 7; Tecta/Tectb −/− 0.42 ± 0.13, n = 7, P = 0.3465, t test). Using the above protocol, IHCs only showed very small capacitance changes (183 ± 129 fF, n = 4, D-F: blue lines and dots), which were due to exocytosis of synaptic vesicles. Number of recordings are shown above the data points. Recordings are at room temperature. G, maximum intensity projections of confocal z-stacks taken from the apical cochlear region of wild-type and Tecta/Tectb −/− mice at P21 stained with antibodies against prestin (green) and Myo7a (OHC marker: blue). Experiments were repeated from four mice of each genotype. Scale bars: 10 μm. . The number of OHCs tested from left to right: wild-type7 (P7), 5 (P8), 2 (P10); Tecta/Tectb −/− 4 (P10), 3 (P13), 5 (P14), 4 (P15), 5 (P18). The dashed grey lines are a Boltzmann fit depicting the predicted changes in P open with age, which was obtained by fitting the shown data extrapolated to 0 since, like most of the other features of the MET current, it develops postnatally (e.g. Lelli et al. 2009). E and F, MET currents recorded at −81 mV from a P18 Tecta/Tectb −/− OHC by displacing the hair bundle with force-step stimuli (top panels). The MET current was elicited by 50 ms positive (excitatory) and negative (inhibitory) driver voltages to displace the stereociliary bundles (V Piezo ). The excitatory recordings in panel E were obtained using progressively larger driver voltages to show MET current adaptation more clearly. Upon termination of the inhibitory stimulus, the MET current showed evidence of rebound adaptation (E, arrow). J Physiol 599.7 development in wild-type (P7-P10) and Tecta/Tectb −/− (P10-P18) mice (Fig. 8C). The size of the MET current in OHCs of Tecta/Tectb −/− mice after the onset of hearing (P13-P18: 1146 ± 257 pA, n = 16, at −121 mV) was not significantly different from that measured in wild-type mice just before the onset of hearing (P7-P10: 1051 ± 237 pA, n = 14, P = 0.300, t test). The MET current was also not significantly different between the two genotypes in pre-hearing OHCs (wild-type: P7-P10, 1051 ± 237 pA, n = 14; Tecta/Tectb −/− : P10: 870 ± 113 pA, n = 4, P = 0.165, t test).

Figure 7. Prestin-driven electromotility in OHCs is normal in Tecta/Tectb −/− mice
Despite the similar MET current size, the open probability of MET channels at rest in 1.3 mM Ca 2+ was significantly larger in OHCs from mature mice (Tecta/Tectb −/− P13-P18: 0.204 ± 0.074 at −121 mV; 0.535 ± 0.109 at + 99 mV, n = 17) than that measured in cells of pre-hearing mice (wild-type P7-P10: 0.101 ± 0.030 at −121 mV, n = 14, P < 0.0001; 0.407 ± 0.088 at +99 mV, P = 0.0014) (Fig. 8D). However, the open probability of MET channels in P7-P10 wild-type OHCs (see above) was not significantly different from that measured in pre-hearing OHCs from Tecta/Tectb −/− mice (P10: 0.133 ± 0.058 at −121 mV, n = 4, P = 0.523; 0.448 ± 0.176 at +99 mV, P = 0.165). These data show that while the maximum size of the MET current is already reaching mature-like values during the second postnatal week (pre-hearing), the Ca 2+ sensitivity of the MET channel may only reach stable, possibly mature, characteristics after the onset of hearing. When OHC stereociliary bundles were deflected by excitatory fluid-jet force steps, from a holding potential of −81 mV, the MET current in mature OHCs from Tecta/Tectb −/− mice declined or adapted over time (Fig. 8E). MET current adaptation to small bundle deflections was best fitted with a single exponential with a time constant of 0.73 ± 0.28 ms (n = 5) (Fig. 8F), a value not significantly different from the fast time constant previously measured in pre-hearing OHCs using a fluid jet (0.65 ± 0.31 ms, n = 14, P = 0.612: Corns et al. 2014). When inhibitory step deflections (negative driver voltages) were applied to the OHC hair bundle, the MET current shut off, revealing the small fraction of current flowing at rest. A transient rebound inward MET current was evident at the offset of the large inhibitory step (downward dip indicated by the arrow in Fig. 8E).
Calcium sensitivity of the MET channels in mature OHCs from Tecta/Tectb −/− mice A recent investigation has indicated that the MET channels located at the tip of the OHC stereocilia may, due to the TM acting as a source of Ca 2+ , be exposed to a Ca 2+ concentration that is much higher (≥ ∼500 μM: Strimbu et al. 2019) than that known to be present in end-olymph (20-40 μM: Bosher & Warren, 1978;Ikeda et al. 1988;Salt et al. 1989;Wood et al. 2004). The effect of these two concentrations of extracellular Ca 2+ on the OHC MET currents in mature Tecta/Tectb −/− mice (P15-P21) is shown in Fig. 9A (40 μM Ca 2+ ) and Fig. 9B (500 μM Ca 2+ ). The size of the MET current at the membrane potential of −81 mV and in 40 μM Ca 2+ (1.48 ± 0.25 nA, n = 9) was significantly larger than that measured in 500 μM Ca 2+ (0.92 ± 0.29 nA, n = 8, P = 0.0003, Tukey's post hoc test, one-way ANOVA, Fig. 9C). The MET current size in 1.3 mM Ca 2+ (1.09 ± 0.26 nA, n = 20) was significantly smaller than that measured in 40 μM Ca 2+ (P = 0.0021), but comparable to the size in 500 μM Ca 2+ (P = 0.2769, Tukey's post hoc test, one-way ANOVA). Extracellular Ca 2+ is known to be a permeant blocker of the hair-cell MET channel (Ricci & Fettiplace, 1998;Marcotti et al. 2005), so the increased current amplitude in 40 μM Ca 2+ is caused by the partial relief of this block. In agreement with previous observations from wild-type mice (Johnson et al. 2012;Corns et al. 2014), the presence of 40 μM Ca 2+ increased the resting open probability (P o ) of the MET channels in the absence of mechanical stimulation (0.50 ± 0.13, n = 9, Fig. 9D), which was significantly larger than that measured in the presence of 500 μM (0.10 ± 0.05, n = 8, P < 0.0001, Fig. 9D) and 1.3 mM Ca 2+ (0.19 ± 0.08, n = 20, P < 0.0001, Tukey's post hoc test, one-way ANOVA). This resting open probability (P o ) of the MET channel in 40 μM Ca 2+ (0.50 ± 0.13, n = 9) was not significantly different from that measured at +99 mV (from Fig. 8D: 0.51 ± 0.12, n = 20, P = 0.7635, t test), a value that is near the Ca 2+ equilibrium potential and strongly reduces Ca 2+ entry into the MET channels (Assad et al. 1989;Crawford et al. 1989).
In order to test the physiological effects caused by the high Ca 2+ concentration proposed to be present near the MET channel (∼500 μM: Strimbu et al. 2019), relative to an endolymphatic concentration of 20-40 μM (see above), on mature OHCs, we performed current-clamp experiments while perfusing the stereociliary bundle with different Ca 2+ concentrations (Figs. 9E-9H). In the presence of 1.3 mM Ca 2+ , which is the Ca 2+ concentration present in the perilymph and also that normally used to perform in vitro recordings from hair cells, OHCs had a resting membrane potential (V m ) of −68.7 ± 4.2 mV (n = 5). When the hair bundles of OHCs were superfused with 500 μM Ca 2+ , the resting V m did not change significantly (−66.5 ± 4.4 mV, n = 5, P = 0.8814, Tukey's post hoc test, one-way-ANOVA) compared with that measured in 1.3 mM Ca 2+ . The small depolarization in 500 μM Ca 2+ was prevented when it was perfused together with the MET channel blocker dihydrostreptomycin (DHS) (Marcotti et al. 2005). In the presence of DHS the V m of OHCs became even more hyperpolarized (−71.3 ± 3.6 mV, n = 5) than that recorded in 1.3 mM Ca 2+ , which was due to the block of the small resting MET current. However, OHCs showed a large depolarization in the presence of 40 μM Ca 2+ (−42.9 ± 6.6 mV, n = 5), which was significantly greater than that observed in either 1.3 mM or 500 μM Ca 2+ (P < 0.0001 for both comparisons, Tukey's post hoc test, one-way ANOVA, Figs. 9E, 9F) and is in agreement with previous observations made in mice , and in rat OHCs with fully developed hearing (∼−40 mV: Oliver & Fakler, 1999;Mahendrasingam et al. 2010). We also showed that 200 μM DHS was also able to block the large membrane depolarization caused by 40 μM Ca 2+ (Figs. 9G, 9H), indicating that the change V m in it is driven by the opening of the MET channels (P = 0.0272, one-way ANOVA). Figure 9I shows the membrane currents recorded from the same OHC shown in Fig. 9E at the end of the experiment, which highlights the presence of viable OHCs. Since the resting V m of OHCs is crucial to optimally activate prestin (Oliver & Fakler, 1999), we determined where the cells would be operating on the voltage-dependent non-linear capacitance curve (Fig. 7D) at the average V m obtained using the different Ca 2+ concentrations (Fig. 9F). We found that in the presence of 40 μM Ca 2+ surrounding the stereociliary bundle, but not in the proposed higher [Ca 2+ ] (Strimbu et al. 2019), the V m of OHCs is sitting at the best activation voltage for prestin modulation (Fig. 9J).

Discussion
The results of this study show that detachment of the TM from the surface of the organ of Corti (Fig. 1) causes a significant elevation of ABR thresholds in Tecta/Tectb −/− mice (up to ∼60 dB in the 8-15 kHz frequency range). The possible contribution of dysfunctional hair cells to the hearing phenotype in Tecta/Tectb −/− mice can be excluded as the TM is not required for the functional maturation of their biophysical properties (Figs. 5-7), including those of the MET apparatus (Fig. 8). The TM is, however, required to ensure that the primary stimulus (sound-induced basilar membrane motion) drives the hair bundles of the hair cells with high sensitivity especially at low-to-moderate sound pressure levels (Fig. 2). The results also demonstrate that, in vitro, the resting open probability of the MET channel in mature OHCs, a manifestation of the channel's Ca 2+ -dependent adaptation characteristics, reaches values of ∼50% in a concentration of Ca 2+ similar to that in endolymph (40 μM). The large resulting resting MET current depolarizes mature OHCs close to the best level for activation of the motor protein prestin, thus providing optimal cochlear amplification ( Fig. 9). In vivo, a large open probability of the MET channels was encountered irrespective of the presence or absence of a TM, indicating that the TM is unlikely to statically bias the resting position of the OHC stereociliary bundle (Fig. 3). Finally, long-lasting adaptation of DPOAEs following prolonged stimulation depends on the presence of the TM (Fig. 4), indicating that the TM may contribute to regulating Ca 2+ levels around the stereocilia and therefore MET channel adaptation in vivo.

Mechanoelectrical transduction in adult OHCs from
Tecta/Tectb −/− mice round window were elevated in the Tecta/Tectb −/− mouse, one can conclude that the TM is required to efficiently drive the deflection of the hair bundle. Amplification of the basilar membrane motion by the OHCs will therefore be reduced in the absence of a TM, as will the motion of the fluid in the sub-tectorial space that drives the displacement of the hair bundles of the IHCs. This, in turn, will reduce neurotransmitter release at the ribbon synapses of the IHCs and result in an increase in ABR and CAP thresholds.
A previous study performed on mice with a detached TM (Tecta ENT/ ENT mice: Legan et al. 2000) provided evidence for an asymmetry of the peak amplitude CM responses to positive and negative sound pressure levels, indicating only a small fraction of the MET channels were open at rest. This finding suggested that the TM statically biases the position of the hair bundles of the OHCs in wild-type mice in the excitatory direction, thereby allowing the cells to sit with 50% of their MET channels open and operate around the most sensitive (steepest) region of the input-output function (Legan et al. 2000). However, considering the narrow operating range of the OHC, estimated to be <200 nm for mouse OHCs (Géléoc et al. 1997;He et al. 2004;Corns et al. 2014;Fettiplace & Kim, 2014;Marcotti et al. 2016) it is hard to conceive how the TM can achieve such a precise spatial bias on the hair bundles. Furthermore, the symmetry of the CM waveforms in the wild-type and the Tecta/Tectb −/− mice are very similar. The data therefore fail to provide evidence that the TM statically biases the operating point of the hair bundles of the OHCs, and indicates that the OHCs have 50% of MET channels open at rest irrespective of the presence or absence of a TM. Considering the morphological similarity of the cochlea in the Tecta ENT/ ENT (Legan et al. 2000) and the Tecta/Tectb −/− mice, the discrepancy in the findings between the two mouse lines is unclear. Possible reasons could include differences in genetic background (mixed/variable C57BL/6J-S129SvEv: Legan et al. 2000;C57BL/6N: present study) or the averaging of data across a larger sample size. Furthermore, we found that DPOAE responses in Tecta/Tectb −/− mice at L1 levels >40 dB SPL lack long-lasting adaptation, a process that adjusts the OHC response into the physiological range. This finding provides evidence that the TM is required for adaptation of the hair-cell MET channel, a process that is known to require Ca 2+ (Assad et al. 1989;Crawford et al. 1991;Corns et al. 2014;Marcotti et al. 2016).

MET channel adaptation during repetitive sound stimulation may require the TM
Recently, it has been shown that the Ca 2+ concentration within the TM of the guinea pig is likely to be much higher (≥500 μM depending on the cochlear region: Strimbu et al. 2019; see also : Anniko & Wroblewski, 1980) than that found in the surrounding endolymph (20-40 μM: Bosher & Warren, 1978;Ikeda et al. 1988;Salt et al. 1989;Wood et al. 2004). Furthermore, very loud sounds, similar to those used to cause a temporary threshold shift, can deplete Ca 2+ from the TM (Strimbu et al. 2019). These data have led to the suggestion that the TM may release Ca 2+ and thus elevate the concentration in the vicinity of the MET channels to more than is required to sustain normal transduction and adaptation (Strimbu et al. 2019). However, the CM recordings made from the round window of Tecta/Tectb −/− mice in vivo showed that the OHC receptor potential is symmetrical in shape (Fig. 3), which is only likely to occur in the presence of endolymphatic-like extracellular Ca 2+ (Fig. 9: i.e. the open probability of the MET channel is ∼50% in 40 μM Ca 2+ ).
In the presence of 500 μM Ca 2+ surrounding the hair bundle, similar to that found in certain regions of the TM (Strimbu et al. 2019), the open probability of the MET channel drops to ∼10%, a value that is inconsistent with the symmetrical CM responses. Additionally, in the presence of 40 μM Ca 2+ , the large MET current flowing into the OHCs depolarizes them to near −40 mV, a potential that reduces the membrane time constant and allows optimal activation of the motor protein prestin, both crucial for normal cochlear amplification. In contrast, the relatively hyperpolarized membrane potential found in the presence of 500 μM Ca 2+ (∼−65 mV) is suboptimal and would compromise amplification (Fig. 9). It therefore seems likely that, either with or without the TM, the set point of the OHC receptor transfer function is determined principally and by the endolymphatic Ca 2+ concentration. As adaptation of the DPOAEs following repetitive lower-level stimulation depends on the presence of the TM, our findings raise the possibility, as yet to be proven, that MET current adaptation in vivo may rely on the TM regulating the Ca 2+ concentration near the MET channel. The TM may directly contribute to increase the Ca 2+ concentration near the MET channels (Strimbu et al. 2019) and/or act as a 'diffusion barrier' for Ca 2+ extruded by the stereocilia via the plasma membrane Ca 2+ -ATPase PMCA2, a protein that is highly expressed in OHCs but less so in IHCs (Chen et al. 2012;Fettiplace & Nam, 2019).