Genetics of Non-syndromic deafness

Deafness is a relatively common condition; it is estimated that 1 in 1000 children are born with a serious hearing impairment. Half of these cases can be considered to have a genetic origin. In addition a great number of individuals will suffer with age-related hearing loss with 50% of the population having a significant hearing impairment by age 80. Whilst isolated, or non-syndromic, deafness accounts for the majority of the inherited forms, hearing loss can occur with other features as part of a recognisable syndrome; over 450 distinct entities have been listed in Online Mendelian Inheritance in Man (OMIM).

Enormous progress has been made in the last few years in the localisation and identification of genes for deafness. As of 1995, no genes for non-syndromic deafness had been cloned, and only a handful of loci mapped, due to obstacles such as genetic heterogeneity, and the difficulties in distinguishing between genetic and environmental causes of deafness. Mapping techniques such as homozygosity mapping and the study of large families living in geographically isolated regions for several generations has allowed us to circumvent these problems. To date 76 loci for autosomal and X-linked forms of non-syndromic deafness have been reported (see Hereditary hearing loss homepage) and many causative genes identified. Interestingly, in the mapping of these loci, it has been observed that some autosomal recessive (denoted DFNB) and autosomal dominant forms (denoted DFNA) both map to the same chromosomal region, as do some syndromic and non-syndromic forms. It has subsequently been shown at the molecular level that different mutations in the same gene can result in these different deafness phenotypes.

A great number of genes have now been implicated in non-syndromic forms of deafness. As illustrated in the table, a diverse range of molecules have been implicated in deafness including transcription factors, structural components and molecules essential for potassium homeostasis and hair cell transduction. The successful identification of these genes has employed a range of strategies, a selection of which will be discussed here.

Candidate genes and Positional candidates
The positional candidate approach is a strategy which combines the knowledge of the map position of the disease locus with the availability of candidate genes mapping to the same chromosomal region. This strategy was successful in the identification of Connexin 26 (GJB2) as the first gene for an autosomal non-syndromic form of deafness (DFNB1 and DFNA3). Connexin 26 is part of a large gene family encoding gap junctions. These are plasma membrane channels formed by the association of six connexins to form a connexon. The formation of intercellular channels is possible through the interaction of connexons between adjacent cells. Gap junctions are found in many cell types and facilitate the exchange of small molecules between cells. A role for gap junctions in the recycling of potassium ions back to the endolymph, after stimulation of the sensory hair cells, has been suggested, and a loss of Cx26 would be expected to disrupt this potassium flow, thus leading to hearing loss. The identification of GJB2 as a “deafness” gene has been one of the most important breakthroughs in the genetics of this condition. After the initial identification of the DFNB1 locus which harbours this gene, studies on collections of families had suggested that this locus may be a major contributor to prelingual deafness. With the identification of the gene, it has been possible to assess the contribution of mutations in GJB2 in autosomal prelingual deafness. Over 50 different mutations in this gene have been identified, accounting for up to half of the genetic non-syndromic childhood deafness. The prevalence of one particular mutation, denoted 35delG, is extremely common in many populations with a carrier frequency of up to 1:25-1:30. The implication of this protein family in hearing has led to the characterisation of other connexin molecules, and their screening for mutations in deaf individuals.

Due to a limited knowledge of the genes involved in normal auditory function and development, one approach to select potential candidates is to identify genes that are either exclusively or preferentially expressed in the inner ear. The isolation and subsequent chromosomal assignment of these genes will render them excellent candidates for any deafness disorder mapping to the same chromosomal interval. This approach was successful in the identification of the DFNA9 gene. A human fetal cochlea cDNA library was constructed and a novel cochlear gene, COCH, was isolated, and demonstrated to be highly expressed in the cochlear and vestibular systems. The subsequent chromosomal assignment and identification of COCH mutations in deaf individuals has endorsed the applicability of this technique.

Mouse models
A great number of mouse mutants with defects in the auditory system have been described, providing a rich source of models for human deafness. The mouse has proved to be an invaluable model due to similarities in inner ear structure, range of inner ear defects and range of associated abnormalities. Conserved linkage groups between man and mouse allow the identification of orthologous genes. In addition, studies of the mouse also allow access to the inner ear at various developmental stages to study disease pathology. Deaf mouse mutants in conjunction with linkage analysis of families with deafness have been instrumental in the identification of human genes, including Myosin7A and Myosin15. Yet, despite this resource, a great number of human deafness loci do not have a corresponding “mouse model”; conversely there are a large number of deaf mouse mutants with no human homologue. This deficit is hoping to be addressed in a large ENU mouse mutagenesis programme, which will aim to provide a larger resource of characterised mouse models for human disease.

In summary, the deafness genes identified to date have given us a sneak preview into cochlear function at the molecular level. A number of genes remain to be found, but with the human genome sequence project completed and the generation and characterisation of novel mouse mutants underway, these invaluable advances will undoubtedly lead to a more thorough understanding of the molecular and cellular basis of normal ear function and development.

Non-syndromic deafness genes identified to date, illustrating a complex repertoire of proteins in the inner ear.
Type of protein		Encoded molecule
Extracellular matrix components	TECTA COCH COL11A2	alpha-tectorin coch collagen11alpha2	DFNA8/12 and DFNB21 DFNA9 DFNA13
Transcription factors	POU3F4 POU4F3	POU3F4 POU4F3	DFN3 DFNA15
Channel components	GJB2 GJB3 GJB6 KCNQ4	Connexin 26 Connexin 31 Connexin 30 KCNQ4	DFNB1/DFNA3 DFNA2 DFNA3 DFNA2
Cytoskeletal components	DIAPH1	Diaphanous	DFNA1
Ion transporters	PDS	Pendrin	DFNB4
Motor molecules	MYO7A MYO15 MYH9	Myosin 7A Myosin 15 MYH9	DFNB2/DFNA11 DFNB3 DFNA17
Synapse component	OTOF	Otoferlin	DFNB9
Intercellular adhesion molecules	CDH23	Cadherin	DFNB12
Junction protein	CLDN14	Claudin 14	DFNB29
Serine protease	TMPRSS3	TMPRSS3	DFNB10
“novel”	DFNA5		DFNA5

Selection of web sites relating to deafness
Hereditary hearing loss homepage. A web site dedicated to the genetics of deafness, regularly updated as loci and genes are reported. Good links to additional audiology-related sites Van Camp, G. and Smith, RJH.
http://www.uia.ac.be/dnalab/hhh/
Mouse mutants with hearing or balance problems. Steel, KP.
http://www.ihr.mrc.ac.uk/hereditary/MutantsTable.shtml
Database of genes expressed in the inner ear. Holme, RH., Bussoli, TJ., Steel, KP.
http://www.ihr.mrc.ac.uk/hereditary/genetable/index.shtml
Cochlea cDNA library, an invaluable resource for identifying candidate genes for deafness.
http://hearing.bwh.harvard.edu/
ENU mutagenesis programme.
http://www.mgu.har.mrc.ac.uk/mutabase/