A mouse model for human deafness DFNB22 reveals that hearing impairment is due to a loss of inner hair cell stimulation Andrei N. Lukashkina,1, P. Kevin Legana, Thomas D. Weddella,1, Victoria A. Lukashkinaa,1, Richard J. Goodyeara, Lindsey J. Welsteada, Christine Petitb,c,d, Ian J. Russella,1,2, and Guy P. Richardsona,2 aSchool of Life Sciences, University of Sussex, Falmer, Brighton BN1 9QG, United Kingdom; bInstitut Pasteur, Unité de Génétique et Physiologie de l’Audition, F75015 Paris, France; cInstitut National de la Santé et de la Recherche Médicale, Unité Mixte de Recherche en Santé 587, Université Pierre et Marie Curie, 75724 Paris, France; and dCollège de France, 75005 Paris, France Edited by Thomas B. Friedman, National Institute on Deafness and Other Communication Disorders, National Institutes of Health, Rockville, MD, and accepted by the Editorial Board September 28, 2012 (received for review June 14, 2012) The gene causative for the human nonsyndromic recessive form of reticular lamina and the lower surface of the TM in an ex vivo deafness DFNB22 encodes otoancorin, a 120-kDa inner ear-specific preparation of the guinea pig cochlea provide evidence that, at protein that is expressed on the surface of the spiral limbus in the frequencies below 3 kHz, counterphase transverse movements of cochlea. Gene targeting in ES cells was used to create an EGFP the two surfaces generate pulsatile fluid movements in the sub- EGFP/EGFP EGFP/EGFP knock-in, otoancorin KO (Otoa ) mouse. In the Otoa tectorial space that could drive the hair bundles of the IHCs (10). mouse, the tectorial membrane (TM), a ribbon-like strip of ECM At higher frequencies, the two surfaces move in phase, and radial that is normally anchored by one edge to the spiral limbus and shear alone is thought to dominate. Theoretical studies (11) re- lies over the organ of Corti, retains its general form, and remains veal that the boundary layers will be vanishingly thin at high in close proximity to the organ of Corti, but is detached from the frequencies, that the fluid in the gap between the TM and the limbal surface. Measurements of cochlear microphonic potentials, reticular lamina will be inviscid, and that the hydrodynamic forces distortion product otoacoustic emissions, and basilar membrane on the hair bundle will be inertial. Although an overlying TM that motion indicate that the TM remains functionally attached to the is not directly attached to a hair bundle does not apply torque to electromotile, sensorimotor outer hair cells of the organ of Corti, the hair bundle (11), the inertial force of the fluid driving the hair and that the amplification and frequency tuning of the basilar bundle depends on its mass and therefore the size of the gap membrane responses to sounds are almost normal. The compound between the reticular lamina and the TM (11, 12). action potential masker tuning curves, a measure of the tuning of The TM is composed of radially arrayed collagen fibrils that the sensory inner hair cells, are also sharply tuned, but the thresh- are imbedded in a noncollagenous matrix composed of a number olds of the compound action potentials, a measure of inner hair of different glycoproteins, including Tecta, Tectb, otogelin, oto- cell sensitivity, are significantly elevated. These results indicate lin, and Ceacam16 (13–16). Mutations in Tecta cause recessive that the hearing loss in patients with Otoa mutations is caused (DFNB21) and dominant (DFNA8/12) forms of human heredi- – by a defect in inner hair cell stimulation, and reveal the limbal tary deafness (17 19), and a dominant missense mutation in fi attachment of the TM plays a critical role in this process. Ceacam16 (DFNA4) has been identi ed recently as a cause of late-onset progressive hearing loss in an American family (15). Mutations in Tecta are one of the most common causes of au- he sensory epithelium of the cochlea, the organ of Corti (Fig. tosomal-dominant, nonsyndromic hereditary hearing loss (20), T1), contains two types of hair cell, the purely sensory inner hair cells (IHCs) and the electromotile, sensorimotor outer hair and mouse models for the recessive (21) and dominant (22) cells (OHCs). These cells are critically positioned between two forms of deafness arising from mutations in Tecta have been GENETICS strips of ECM, the basilar membrane (BM) and the tectorial created. Together with data from a Tectb-null mutant mouse membrane (TM). Signal processing in the cochlea is initiated (23), these studies have provided evidence that the TM plays when sound-induced changes in fluid pressure displace the BM multiple roles in hearing (24). Although much is known about in the transverse direction, causing radial shearing displacements the structure of the TM, an ECM that is unique to the cochlea, between the surface of the organ of Corti (the reticular lamina) relatively little is known about how it attaches to the apical and the overlying TM (1). The radial shear is detected by the hair surface of the cochlear epithelium. Otoancorin, a product of the bundles of the IHCs and the OHCs (2), with the stereocilia of DFNB22 locus, is expressed on the apical surface of the spiral the OHC hair bundles forming an elastic link between the organ limbus and has been suggested to mediate TM attachment to this of Corti and the overlying TM (3). Deflection of the stereocilia region of the cochlear epithelium (25). In this study, we use gene gates the hair cell’s mechanoelectrical transducer (MET) chan- targeting to inactivate otoancorin. This provides a mouse model nels, thereby initiating a MET current (4) that promotes active for DFNB22, reveals a loss of IHC sensitivity as the primary mechanical force production by the OHCs, which, in turn, in- fluences mechanical interactions between the TM and the BM (5, 6). This nonlinear frequency-dependent enhancement pro- Author contributions: C.P., I.J.R., and G.P.R. designed research; A.N.L., P.K.L., T.D.W., V.A.L., cess, which boosts the sensitivity of cochlear responses to low- R.J.G., L.J.W., I.J.R., and G.P.R. performed research; P.K.L. contributed new reagents/ana- level sounds and compresses them at high levels, is known as the lytic tools; A.N.L., T.D.W., R.J.G., I.J.R., and G.P.R. analyzed data; and C.P., I.J.R., and G.P.R. cochlear amplifier (7). wrote the paper. Whereas the hair bundles of the OHCs are imbedded into the The authors declare no conflict of interest. TM and therefore directly excited by relative displacement of This article is a PNAS Direct Submission. T.B.F. is a guest editor invited by the the undersurface of the TM and the reticular lamina, those of the Editorial Board. IHCs are not in direct contact with the TM, and the way in which 1Present address: School of Pharmacy and Biomolecular Sciences, University of Brighton, they are driven by motion of the BM remains unclear. Intra- Brighton BN2 4GJ, United Kingdom. cellular recordings of the receptor potentials in IHCs indicate 2To whom correspondence should be addressed. E-mail: [email protected] or that the bundles are velocity-coupled (to fluid flow) at low fre- [email protected]. quencies and displacement-coupled at higher frequencies of This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. stimulation (2, 8, 9). Direct measurements of the motion of the 1073/pnas.1210159109/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1210159109 PNAS | November 20, 2012 | vol. 109 | no. 47 | 19351–19356 Downloaded by guest on September 27, 2021 fibril bundles of the TM are, as in heterozygotes and WT, dis- tributed radially within the core of the matrix in the homozygous mutants (Fig. 4 A and B). The covernet, a network formed by fibrils that are aligned mainly along the length of the upper surface of the TM, is also of normal appearance (Fig. 4 A and B). Hensen’s stripe, a ridge that runs longitudinally along the lower surface of the TM and is thought to engage the hair bundle of the IHCs, is, however, not visible in the homozygous mutants (Fig. 4 C and D).These observations show that otoancorin is required for adhesion of the TM to the spiral limbus and that the TM, despite the structural abnormalities described here, retains its Fig. 1. Schematic cross-section of WT cochlea. Spiral lamina (SLAM), spiral gross overall form, remaining in close proximity with the organ ligament (SLIG), inner pillar cells (IPC), outer pillar cells (OPC), Deiters’ cells of Corti. (DC), phalangeal process of DC (PhP), Claudius cells (CC), OHC, IHC, reticular EGFP/EGFP laminar (RL), spiral limbus (SL), and major noncellular elements (BM and TM). Cochlear Microphonic Potentials Are Symmetrical in Otoa Mice. Cochlear microphonic (CM) potentials are extracellular potentials derived from the transducer currents of the OHCs (1, 26– cause of deafness, and isolates a specific role for the limbal at- 28), and their symmetrical nature is known to result from the tachment of the TM in driving the hair bundles of the IHCs. presence of the TM biasing the operating point of the hair bundle (21, 29). Although the TM is detached from the spiral limbus in the Results OtoaEGFP/EGFP mice, the CM potentials recorded from the round GenetargetinginEScellswasusedtoreplacethefirst coding window of OtoaEGFP/EGFP mice are similar to those recorded exon of the murine Otoa gene with a cassette encoding EGFP from WT mice (Fig. 5A), being symmetric at low to moderate (Fig. 2A). Southern blotting with probes located external to the sound levels, becoming negatively and then positively asymmet- targeting vector was used to detect and confirm successful rical with increasing sound levels greater than 80 dB sound targeting (Fig.