A genetic approach to understanding Commentary inner ear function See related article, pages 1447–1455. James F. Battey, Jr.

National Institute on Deafness and Other Communication Disorders, NIH Building 31, Room 3C02, 31 Convent Drive MSC 2320, 9000 Rockville Pike, Bethesda, Maryland 20892, USA. Phone: (301) 402-0900; Fax: (301) 402-1590; E-mail: [email protected].

Within the temporal bone lies the inner tures are observed. About 70% have non- endolymph, resulting in profound hear- ear, home to both the snail-shaped syndromic hearing impairment, where ing loss (8). Completing the postulated cochlea, critical for auditory function, abnormal auditory function is the only recycling loop, channels at the secreting and the vestibule, providing critical sen- obvious clinical outcome resulting from surface of the marginal cells are made sory input for balance. Both organs mutation. Hundreds of different up of subunits that are encoded by the house sensory hair cells, cellular trans- syndromes have hearing impairment as KvLQT1 and ISK . In humans, ducers of either sound waves (auditory) a cardinal feature, and over sixty genetic mutations in KvLQT1 or ISK result in or linear or angular acceleration loci are known to harbor a gene whose Jervell and Lange-Nielsen syndrome, a (vestibular) into electrical impulses. mutation results in nonsyndromic hear- disorder characterized by autosomal These unique cells have at their apices a ing impairment that is inherited in a recessive congenital hearing impair- bundle of microvilli containing parallel mendelian fashion. These numbers con- ment and a long QT interval on electro- actin filaments referred to as stereocilia, tinue to grow at a rapid rate, indicating cardiograms (9). Mice engineered to lack whose deflection by as little as a that there may be as many as a five hun- a functional isk gene cannot produce nanometer results in the opening of dred different gene products whose normal endolymph and show profound mechanicallygated ion channels. function is essential for hearing. hearing impairment (10). In the cochlea, stereocilia are bathed In this issue of the JCI, Lee et al. (11) by endolymph, an extracellular fluid Genetic dissection of potassium report that inactivation of the KvLQT1 with unusually high potassium concen- recycling in the cochlea gene results in profound hearing tration. When the stereocilia are deflect- Studies of hereditary hearing impair- impairment and circling behavior char- ed, potassium from the endolymph ment have allowed the identification of acteristic of vestibular abnormality. This flows through transducer channels in many critical, and previously unknown, mouse model for Jervell and Lange- the hair cells, resulting in depolariza- molecular components of the auditory Nielsen syndrome recapitulates hearing tion. Following depolarization, inner system, including some that appear to impairment and manifests a more hair cells release neurotransmitter at define the unique biological pathway of severe vestibular abnormality, but fails synapses with afferent neurons, while potassium recycling. Mutations in the to demonstrate the electrocardiograph- outer hair cells undergo an oscillating gene KCNQ4 result ic abnormalities observed in humans. change in length that both increases in a dominant form of progressive hear- Histopathologic examination of the sensitivity and fine tunes responses to ing loss (4); this channel is postulated to inner ear of these mice reveals severe sound frequencies. This electrochemical be of importance in transporting potas- anatomic disruption of the cochlear and gradient between endolymph and hair sium ions out of hair cells. Upon leaving vestibular end organs, providing addi- cells is thought to require the active, the hair cells, potassium ions are postu- tional insight into the auditory and ATP-dependent recycling of potassium lated to pass through a network of gap vestibular abnormalities found in the from the depolarized hair cells back into junctions between supporting cells and mouse model and in patients with the endolymph. The discovery that hear- fibrocytes in the cochlea, ending up in Jervell and Lange-Nielsen syndrome. ing impairment can result from muta- the epithelial marginal cells of the stria tions in genes whose products are essen- vascularis, the structure that secretes Mouse models for hereditary tial to this putative recycling pathway potassium-rich endolymph. hearing impairment are essential suggests that potassium recycling is 26 and connexin 31 are compo- These and many other elegant studies required for normal auditory function. nents of these gap junctions. Mutations underscore the importance of mouse Given the remarkable sensitivity and in the GJB2 gene, encoding connexin 26, models in furthering understanding of specificity of auditory function, it is not are the most common cause of heredi- human genetic disease and gene func- surprising that mutations in hundreds tary hearing impairment in Americans, tion. When a mutated human gene is of different genes cause profound and accounting for about one fourth of mapped to a chromosomal locus, a progressive hearing impairment in chil- autosomal recessive hearing impair- mouse model with a mutated gene in dren (for reviews, see refs. 1–3). Roughly ment (5) and a much higher percentage the corresponding region of the mouse one child in two thousand is born with in other populations (6, 7). A sodium- genome can be very helpful in identify- hearing impairment that may compro- potassium-chloride cotransporter pro- ing the mutated gene. The power of this mise development of normal language tein encoded by Slc2a2 helps to pump approach is amply demonstrated by the skills. About 30% of these individuals potassium into the marginal cells that effort to clone a deafness gene at the have syndromic hereditary hearing secrete endolymph. Mice with muta- DFNB3 locus, where a mouse model, the impairment, where other clinical fea- tions in this gene fail to produce -2 mouse, was available. The

The Journal of Clinical Investigation | December 2000 | Volume 106 | Number 12 1431 mutated gene was determined by identi- genes. Using this approach, the genes basing intervention strategies on genet- fying the gene that would rescue normal underlying one type of autosomal domi- ic information. We are entering a new hearing in the shaker-2 mouse (12). nant deafness (DFNA9) (15), as well as golden age in our approach to hearing Once the gene (myo15) was found in the one type of autosomal recessive deafness loss, both in terms of our fundamental mouse, human MYO15 was rapidly (DFNB9) (16), were identified. More understanding of auditory function and cloned and mutations were identified in recently, Zheng et al. identified a collec- our ability to provide precise diagnosis, families with hereditary deafness map- tion of genes selectively expressed in accurate prognosis, and optimal inter- ping to the DFNB3 locus (13). The shak- outer, but not inner, hair cells (17). Outer vention for individuals with hereditary er-2 mouse was used for morphology hair cells function both to amplify and to hearing impairment. studies at the ultrastructural level, which fine-tune the auditory signal, and they are showed a striking loss of normal hair cell essential for important hearing out- Acknowledgments stereocilia, the presumptive cause of comes, such as speech recognition. They The author thanks Thomas B. Fried- hearing loss. Without an animal model perform these functions by changing man, Andrew Griffith, Donald Luecke, for hereditary hearing impairment, it is their length in response to depolarization, and Robert Dobie for critical com- difficult if not impossible to determine activating a voltage-dependent motor in ments. The author also thanks John the anatomical, biochemical, and cellu- the lateral that changes Ashkenas for editing and providing an lar basis for the phenotype, or to use the membrane surface area. Inner hair opportunity for this Commentary. gene rescue to prove unequivocally that cells do not share this function, but James F. Battey, Jr. is the Director, the disease-causing gene has been iden- instead release neurotransmitter at National Institute on Deafness and tified. Indeed, when a mouse model does synapses with afferent auditory neurons. Other Communication Disorders, NIH. not already exist, it is often necessary to The motor molecule was identified as a create one as the first step in under- protein called prestin that is abundantly 1. Friedman, T., et al. 2000. Modifier genes of heredi- tary hearing loss. Curr. Opin. Neurobiol. 10:487–493. standing how the gene mutation results expressed in outer, but not inner, hair 2. Morell, R.J. 1999. Recent progress in hereditary in hearing impairment. cells. Creation of a mouse lacking a func- hearing loss. Current Opinion in Otolaryngology & Hundreds of additional mouse mod- tional prestin is in progress and will pro- Head and Neck Surgery. 7:259–265. 3. Steel, K.P. 1999. The benefits of recycling. Science. els for hearing and vestibular defects vide a powerful tool for understanding 285:1363–1364. will emerge in the next few years as a this aspect of auditory function. 4. Kubisch, C., et al. 1999. KCQN4, a novel potassium result of systematic efforts to introduce channel expressed in sensory outer hair cells, is mutated in dominant deafness. Cell. 96:437–446. mutations at random into sperm at Potential clinical implications 5. Cohn, E.S., et al. 1999. Clinical studies of families high rates. After mutagenesis, the prog- Within the last four years, the research with hearing loss due to mutations in the connex- eny will be screened for auditory and community has identified nearly twenty in 26 gene (GJB2/DFNB1). Pediatrics. 103:548–550. vestibular abnormalities, as well as a genes whose mutation results in non- 6. Estivill, X., et al. 1998. Connexin-26 mutations in sporadic and inherited sensorineural deafness. host of other phenotypes of biomedical syndromic hereditary hearing impair- Lancet. 351:394–398. interest. Successful efforts of this type ment. One of these genes, GJB2, is 7. Morell, R., et al. 1998. Mutations in the connexin are already under way in the United responsible for about 25% of hereditary 26 gene (GJB2) among Ashkenazi Jews with non- syndromic recessive deafness. N. Engl. J. Med. Kingdom and Germany, and have just hearing impairment in Americans, and 339:1500–1505. begun in the United States in several an even higher percent in certain defined 8. Dixon, M.J., et al. 1999. Mutation of the Na-K-Cl centers supported by collaborations subpopulations (6, 7). Knowledge of co-transporter gene Slc12a2 results in deafness in mice. Hum. Mol. Genet. 8:1579–1584. between institutes within the NIH. It is these important genes will allow devel- 9. Neyroud, N., et al. 1997. A novel mutation in the not unreasonable to speculate that opment of genetic tests, leading to rapid potassium channel gene KvLQT1 causes the Jervell within the next five years there may be a and precise diagnosis of hereditary hear- and Lange-Nielsen cardioauditory syndrome. Nat. Genet. 15:186–189. mouse model available for most if not ing impairment. Increased diagnostic 10. Vetter, D.E., et al. 1996. Inner ear defects induced by all genes, or combination of genes, precision is particularly important when null mutation of the isk gene. Neuron. whose mutations result in nonsyn- a child with hearing impairment is born 17:1251–1264. 11. Lee, M.P., et al. 2000. Targeted disruption of the dromic hearing impairment. This col- in a family where both parents and all Kvlqt1 gene causes deafness and gastric hyperplasia lection of mouse models would be a close relatives have normal hearing. In in mice. J. Clin. Invest. 106:1447–1455. remarkable resource for identifying the this situation, it is difficult to determine 12. Probst, F.J., et al. 1998. Correction of deafness in shaker-2 mice by an unconventional myosin in a corresponding genes in humans, under- the etiology of hearing impairment in BAC transgene. Science. 280:1444–1447. standing gene function, and dissecting the absence of genetic tests. Knowledge 13. Wang, A., et al. 1998. Association of unconvention- the molecular pathways essential for of the etiology is essential to enable al myosin MYO15 mutations with human non- syndromic deafness DFNB3. Science. auditory function in unprecedented informed counseling for the affected 280:1447–1451. detail (for a review of mouse models for individual, as well as his or her parents 14. Probst, F.J., and Camper, S.A. 1999. The role of hearing loss, see ref. 14). and other relatives. Greater diagnostic mouse mutants in the identification of human hereditary hearing loss genes. Hear. Res. 130:1–6. Other approaches to identify precision may arm clinicians with 15. Robertson, N.G. 1997. Mapping and characteriza- that are critical for auditory function in important clues about other medical tion of a novel cochlear gene in human and in particular are also bearing fruit. Success- problems that accompany syndromic mouse: a positional candidate gene for a deafness disorder, DFNA9. Genomics. 46:345–354. ful efforts are underway to identify genes hearing impairment. Future clinical 16. Yasunaga, S., et al. 1999. A mutation in OTOF, selectively expressed in the cochlea, and to research may uncover new and different encoding otoferlin, a FER-1-like protein, causes determine whether these genes map to ways to improve the development of lan- DFNB9, a nonsyndromic form of deafness. Nat. Genet. 21:363–369. chromosomal loci known to harbor one guage and communication skills among 17. Zheng, J., et al. 2000. Prestin is the motor protein of or more hereditary hearing impairment individuals with hearing impairment, cochlear outer hair cells. Nature. 405:149–155.

1432 The Journal of Clinical Investigation | December 2000 | Volume 106 | Number 12