gms | German Medical Science

The vast majority of inherited hearing impairment are non syndromic. The exact number of genes involved in genetic hearing impairment (HI) is not known, but it is thought that mutations in at least 100 genes can be associated with hearing impairment.

Over the last decade, we have seen a tremendous growth in the localisation and identification of genes for non syndromic hearing impairment. It has become clear that this condition is extremely genetically heterogeneous. Chaib et al. [Ref. 6] reviewed the frequency and inheritance of congenital isolated deafness and causes for difficulties associated with mapping of deafness genes. They noted that in the U.S., deafness affects 1 in 1000 children at birth or during infancy. About 75% of the inherited forms of congenital isolated deafness have an autosomal recessive mode of transmission. Difficulties in the localization of deafness genes are due to several factors, including extreme genetic heterogeneity of the defect (there are an estimated 40 deafness genes segregating in the population); the absence of clinical criteria to allow differentiation between inner ear defects; and the high proportion of marriages between deaf persons in Western countries leading to coexistence of several defective genes responsible for clinically indistinguishable phenotypes in one family [Ref. 6]. Currently (mid 2005), over 100 different loci have been found to be associated with hereditary hearing impairment: 41 disorders are associated with autosomal dominant mode of inheritance, 45 with autosomal recessive, 6 with X-linked and 6 are maternally inherited on the mitochondrial DNA. Of these, 20 genes have been cloned for autosomal dominant (DFNA) (Table 1 [Tab. 1]), 20 genes for autosomal recessive (DFNB) (Table 2 [Tab. 2]), 1 for X-linked (DFN) (Table 3 [Tab. 3]) and 6 genes in the mitochondrial DNA (Table 4 [Tab. 4]). Furthermore, two modifier genes have been found (Table 5 [Tab. 5]): DFNM1 can dramatically reduce the penetrance of DFNB26 -linked haplotype in homozygous, therefore preventing hearing impairment to occur [Ref. 37], and DFNM2 which should account for hearing impairment in A1555G mitochondrial mutation carriers not exposed to aminoglycosides [Ref. 5]. A continuously updated overview of the field can be found in the Hereditary Hearing loss Homepage (http://webhost.ua.ac.be/hhh/).

For each gene various disease causing mutations have been described, exhibiting significant genotype and phenotype heterogeneity. Even when the inheritance of the trait in a family is evident, the substantial genotype-phenotype heterogeneity does not allow screening for a genetic condition that might be associated with a certain audiometric configuration, unless the subject pertains to a large familial group so that a linkage study might be feasible. In fact, there are several different mutated genes that can give rise to low frequency or high frequency HI or affect all frequencies. In the audiological clinics, diagnostic tests available have been found to have poor sensitivity and specificity [Ref. 29]. The exact carrier frequency of specific gene mutations in the population would also be an important information for the clinicians, as carrier frequencies of various mutations have been shown to differ between countries. For example, the mutation 35delG is most prevalent in southern Europe (apparently 4,0%), whereas apparently it only accounts for 1.7-2.5% in Northern Europe and North America. However, the samples tested are small and further studies based on larger populations are needed. Although it may be unrealistic, the audiological physician would like the possibility that a specific audiometric configuration points to a specific gene mutation- improving and facilitating the diagnostic process. In other words, the clinician wants to know how a specific gene mutation affects the hearing i.e. correlate the genetic disorder to its clinical presentation and possibly prospect future specific treatment for HI.

Indeed, there are only few cases in which a consistent genotype-phenotype correlation has been found and these are discussed below in this chapter. Although it cannot be expected that a certain classification of hearing impairment, configuration of the audiogram and involvement of specific frequencies will correlate 100% to a specific mutation in a gene, it has never been discussed and agreed upon what 'correlation' means [Ref. 9]. For example, we assume that there is correlation when 90% of the phenotype in subjects homozygous for the 35delG mutation in Connexin 26 will have profound HI. But this is not always the case. In some cases we have incomplete or variable penetrance due to modifier genes, hetereoplasmy etc., therefore most of the time it is not easy to point at a certain genotype based on the clinical presentation.

The insufficient descriptions of phenotypes in genetic journals and genotypes in audiological journals is also a barrier to correlating phenotypes to genotypes and thus guidelines to describe phenotypes/genotypes in reports on non syndromic HHI has been proposed [Ref. 30]. In addition, selection bias is often found in even extensive reports on hereditary HI, because most often the subjects with severe/profound HI are predominantly included in these studies being the patients mostly referred to the clinics and attracting major attention. On the other hand, subjects with mild to moderate hearing impairment or late onset hearing impairment are often misdiagnosed and the clinical traits are more easily confused with phenocopies of non genetic origin.

As we gain more knowledge on the function of the gene in the ear, i.e. the function of the specific protein coded by the gene, as well as the effect that the various kinds of mutations have on the protein function, we will probably understand more of the physiology of hearing and of mechanisms which lead to hearing impairment.

The rapidly increasing understanding of genetics of HI implies a potential for a wider use of genetic counselling and molecular genetic testing in the clinical practice. The complexity and the amount of available biologic information can present greater challenges for ENT doctors and audiologists. It is becoming more and more important to be able to provide an interdisciplinary approach to hearing impairment with teams including clinical audiology, medical genetics, pathology, molecular biology experts. Future challenges include the understanding of the interaction between genes and the environmental damaging factors such as noise, ototoxic molecules, and to which degree presbyacusis, otosclerosis, sudden hearing loss or other complex forms of hearing impairment are influenced by individual genetic predisposition and, based of these knowledge tailor new treatments for hearing disorders.

The clinical presentation of the different genetic hearing impairments is not specific and present several phenocopies, making it difficult to identify the underlying genetic disorder. Here some of the autosomal dominant non syndromic hearing impairment phenotypes are described.

Severity varies notably across disorders. In Figure 1 [Fig. 1] and 2 [Fig. 2] it can be seen that autosomal dominant disorders tend to be less severe in comparison to autosomal recessive disorders. It is also clear that no specific trait can be identified for a given genotype. In fact, several conditions can give rise to high frequency loss or involve all frequencies, etc. Although not in all cases, more often a recessive condition leads to a profound, early onset hearing impairment, while autosomal dominant condition can span from a mild to profound impairment involving various frequencies.

Even in autosomal dominant conditions, where the family history is suggestive of a inherited form of hearing impairment, a certain audiometric trait does not orient our diagnosis toward the underlying genotype.

DFNA2 (1p34): onset 15-25 yrs, involves initially the HF then progresses to involve all frequencies. Deteriorates to severe in about 10 yrs, incomplete penetrance [Ref. 8]. A Swedish family was found to carry both haplotype DNFA2 and DFNA12. hearing impairment was more severe in patients carrying both haplotypes simultaneously [Ref. 1].

Gene: GJB3 named also connexin 31, belongs to the gap junction family protein [Ref. 47], can also cause hyperkeratosis syndromes with or without hearing impairment [Ref. 38].

Gene: KCNQ4 encodes a potassium channel protein containing six transmembrane domains, a highly conserved P-loop lining the channel pore, and two cytoplasmic domains. Four KCNQ4 subunits aggregate to form a functional channel. In situ hybridization revealed that the KCNQ4 message was present only in outer hair cells of the organ of Corti, suggesting that KCNQ4 might play a role in the recycling of potassium ions to the endolymph after hair cell stimulation [Ref. 20].

DFNA3 (13q11-12): prelingual onset, can involve both HF or all frequencies, stable or progressive, hearing impairment can be mild to profound. Penetrance is incomplete.

Gene: GJB2, also named connexin 26, encodes for a gap junction. These proteins are plasma membrane channels formed by six subunit proteins called connexins. The formation of intercellular channels is possible through interaction with connexins in the plasma membrane of adjacent cells [Ref. 19].

Gene: GJB6 also named connexin 30 [Ref. 18].

DNFA5 (7p15): onset in early childhood, rapidly progresses to severe hearing impairment. Penetrance is complete [Ref. 44].

Gene: DFNA5 is expressed in the cochlea and in various other tissues. It shows no significant homology to any other known gene, and no clues to its function have been found despite extensive computational analysis [Ref. 45].

DFNA7 (1q21-23): onset at 11-30 yrs, HF, progressive, mild to severe hearing impairment, incomplete penetrance [Ref. 12].

DFNA9 (14q12-13): onset at 11-30 yrs, HF to all frequencies, progressive, severe to profound hearing impairment, vestibular anomalies, complete penetrance [Ref. 28].

Gene: COCH. Encodes a 550-amino acid involved in different functions such as hemostasis, complement system, immune system, and extracellular matrix assembly [Ref. 39].

DFNA11 (11q12.3-21): onset at 11-30 yrs, HF, progressive, mild to severe hearing impairment, complete penetrance [Ref. 42].

Gene: MYO7A is an unconventional myosin, that share structurally conserved heads which move along actin filaments using actin-activated ATPase activity, and have divergent tails presumably to move different macromolecular structures relative to actin filaments. In the ear, myosin VIIA is present in both inner and outer hair cells, although expression is greater in the former. In the eye, it is localized to microvilli projections in retinal pigmentary epithelial cells and photoreceptor cells. Mutations in myosin VIIA have been identified in Usher syndrome type 1b, a recessively inherited disease characterized by congenital deafness, vestibular dysfunction and retinitis pigmentosa. Mutations in MYO7A were also found in nonsyndromic autosomal recessive deafness (DFNB2) and autosomal dominant deafness (DFNA11) [Ref. 25].

DFNA16 (2q23-24.3): onset at 11-30 yrs, progressive, HF sometimes fluctuating that responds to steroids, severity not specified. Penetrance is not known [Ref. 14].

DFNA17 (22q12.2-13.3): onset 4 -10 yrs, HF or all frequencies are involved, progressive, moderate to profound hearing impairment, complete penetrance [Ref. 24].

Gene: MYH9, a nonmuscle-myosin heavy-chain gene is responsible of DFNA17 phenotype. Expression of MYH9 was immunolocalized in the organ of Corti, the subcentral region of the spiral ligament, and the Reissner membrane [Ref. 23].

DFNA20/26 (17q25): onset 11-30 yrs, HF, progressive, severity is variable, progressive, complete penetrance [Ref. 32], [Ref. 48].

DFNA25 (12q21-24): onset 11-30 yrs, HF, progressive, severity is variable, unknown penetrance [Ref. 15].

DFNA30 (15q25-26): onset 11-30 yrs, HF + MF, progressive to the IV decade of life, then stable, moderate to severe, little interfamilial variability, normal vestibular function, tinnitus variable, complete penetrance [Ref. 27].

DFNA43 (2p12): onset 11-30 yrs, HF to all freq., progressive, mild to profound, little interfamilial variability, normal vestibular function, tinnitus variable, complete penetrance [Ref. 13].

DFNA4 (19q13): onset 11-30 yrs, all freq., progressive, from moderate to profound hearing impairment, unknown penetrance [Ref. 7].

Gene: MYH14 encodes one of the heavy chains of the class II nonmuscle myosins and is highly expressed in mouse cochlea [Ref. 11]. The role of MYOH14 in hearing has not been clarified, yet.

DFNA8/A12 (11q22-24): onset congenital-10 yrs, all freq. (audiogram is typically a U shaped with maximum loss at 2kHz), from moderate to severe, stable, onset 11-30 aa, all freq., progressive, from moderate to profound hearing impairment, little intrafamilial variability, unknown penetrance [Ref. 34]. Mutations also cause a form of autosomal recessive hearing impairment (DFNB21) [Ref. 33].

Gene: TECTA. Conserved structural features include an amino-terminal hydrophobic signal sequence for translocation across the membrane and a carboxy-terminal hydrophobic region characteristic of precursors of glycosylphosphatidylinositol-linked membrane-bound proteins. Three polypeptide domains, a module containing a region homologous to the G1 domain of entactin, a module similar to zonadhesin, and a module consisting of a zona pellucida domain, are cross-linked to each other by disulfide bridges and interact with beta-tectorin to form the non-collagenous matrix of the tectorial membrane

DFNA10 (6q22-23): variable onset, all freq., progressive, from moderate to severe, complete penetrance [Ref. 35].

Gene: EYA4 (Eyes absent 4), a member of the vertebrate Eya family of transcriptional activators. In two unrelated different mutations in EYA4 were found, both of which create premature stop codons. Although EYA proteins interact with members of the SIX and DACH protein families in a conserved network that regulates early embryonic development, this finding shows that EYA4 is also important post-developmentally for continued function of the mature organ of Corti [Ref. 46].

DFNA15 (5q31): onset 11-30 yrs, all freq., progressive, from moderate to severe, vestibular dysfunction, complete penetrance [Ref. 43].

Gene: POU4F3 is a member of the family of POU domain transcription factors. This group of proteins shares a POU-specific domain and a POU homeodomain, both of which are required for high-affinity binding to DNA target sites. Targeted mutagenesis of both Pou4f3 alleles leads to profound deafness and vestibular dysfunction in "knock-out" mice.

DFNA20/26 (17q25): congenital onset, all freq., profound, unknown penetrance [Ref. 10].

Gene: Zhu et al. [Ref. 49] found missense mutations in highly conserved actin domains of the ACTG1 gene. This gene encodes for a gamma actin. It is the major component of the thin filaments of muscle cells and of the cytoskeletal system of nonmuscle cells. The exact mechanism by which this actin causes hearing loss has not been defined, yet.

DFNA36 (9q13-q21): onset 5 to 10 yrs, all, progresses rapidly to profound hearing impairment, normal vestibular function, unknown penetrance

Gene: TMC1 is member of a gene family predicted to encode transmembrane proteins. Tmc1 mRNA is expressed in hair cells of the postnatal mouse cochlea and vestibular end organs and is required for normal function of cochlear hair cells [Ref. 22].

DFNA41 (12q24-qter): onset 10-20 yrs, all freq., progressive, severity is from moderate to profound, unknown penetrance [Ref. 3].

DFNA6/14/38 (4p16.3): onset 11-30 yrs, initially it involves the low frequencies and high frequencies then progresses involving all frequencies, usually stable, incomplete penetrance [Ref. 2].

Gene: WFS1, is the gene responsible for Wolfram syndrome, an autosomal recessive disorder characterized by diabetes mellitus and optic atrophy, and often, deafness [Ref. 41]. WFS1 encodes for wolframin, an integral, endoglycosidase H-sensitive membrane glycoprotein expressed predominantly in the endoplasmic reticulum (ER) of different tissues including the CNS [Ref. 41]. The ER localization of WFS1 suggests roles in membrane trafficking, protein processing and/or regulation of ER calcium homeostasis.

Many patients carrying WFS1 mutations have also tinnitus accompanying LFSNHI, but there are otherwise no associated features such as vertigo. LFSNHI worsens over time without progressing to profound deafness; in contrast to low-frequency hearing loss linked to DFNA1, caused by mutations in the DIAPH1 gene that is associated with progression to profound deafness by the fourth decade of life.

DFNA1 (5q31): onset 11-30 yrs, LF, progressive to profound deafness bu age 40-50, mild to severe, complete penetrance.

Gene: DIAPH1 belongs to formin gene family which are involved in cytokinesis and establishment of cell polarity. All formins share Rho-binding domains in their N-terminal regions. Rho regulates actin polymerization, which may be particularly important for hair cells [Ref. 26].

DFNA12 (11q22-24): onset congenital -10 yrs, MF, stable in some patients, progressive in others. The hearing loss was moderate to severe, a pure tone audiogram showing a U-shaped form with maximum loss at 2, 000 Hz. There was great interfamilial variability, complete penetrance.

Gene: TECTA (see DFNA8)

DFNA13 (6p21.3): onset 11-30 yrs, MF, from mild to severe, stable, unknown penetrance [Ref. 4].

Gene: COL11A2 encodes for a collagen protein. Mutations in COL11A2 cause autosomal dominant and autosomal recessive OSMED syndrome, a combination of osteochondrodysplasia and hearing impairment. Autosomal dominant OSMED is also called non-ocular Stickler syndrome. Mutations in COL11A2 have been found in two families with non-syndromic hearing impairment. Electron microscopy of the tectorial membrane in COL11A2 knockout mice suggests that an abnormal tectorial membrane is responsible for the hearing impairment [Ref. 31].

DFNA21 (6p21-22): onset congenital -10 yrs, MF and HF, progressive, from moderate to severe, incomplete penetrance [Ref. 21].

DFNA23 (14q21-22): onset congenital -1 yr, MF and HF, variable severity, stable, complete penetrance [Ref. 40].

DFNA24 (4q35-qter): prelingual onset, MF and HF, stable, variable severity, unknown penetrance [Ref. 16].

DFNA28 (8q22): onset >7 yrs, MF and HF, variable severity, progressive, unknown penetrance [Ref. 36].

It can be noticed that although some differences exist between different conditions, no rigid clinical classification is possible due to progression which changes hearing configuration with age and or exposure to worsening environmental factors. Variable or incomplete penetrance and heterogeneity within the same familial group also contribute to no clear cut clinical pictures. Still a lot of information is lacking regarding some clinical characteristic in few conditions such as DFNA22 caused by mutations in the unconventional myosin VI gene, (Myo6). In DFNA22 hearing impairment is progressive, with onset during childhood (6 to 8 years of age at onset of first audiometric abnormalities). By the age of approximately 50 years, affected individuals invariably had profound sensorineural hearing impairment. No vestibular dysfunctions were detected. In this condition, no information was given regarding frequencies initially and subsequently involved, penetrance or tinnitus. For most of the conditions we have described, only the locus segregating with that phenotypical trait is known. A locus may contain more than one gene as in the case of DFNA2 and DFNA3, but even when genes and their products are known, it is not easy to predict the phenotypic expression, since the mechanisms which determine hearing impairment are complex and still to be understood. Furthermore, as shown in Figure 3 [Fig. 3], different mutations in the same gene can give rise to different modalities of inheritance. This can be partially explained by the type of variation of protein function that a certain mutation introduces.

Recessive mutations more often produce a "loss of function" and for most genes a normal functional product is guaranteed by the half normal protein produced by the normal hallele, therefore clinical symptoms will be present only in the mutated homozygous condition.

Dominant mutations in most cases result in "gain of function", therefore the mutated protein gains a new deleterious function. Proteins that group to form multimeric structures can also present a dominant - negative effect. For example: 6 connexin monomers must assemble to form a connexon, that forms a gap junction. If half of the monomers are mutated and if only one altered monomer is sufficient to enable the function of the whole connexon, then only 1/64 connexons (2⁶), will function properly. In such case, a condition of heterozygosity is sufficient to produce a clinical phenotype and the condition will be dominant.

Given the difficulties in correlating genotype and phenotype, it is important that as much information as possible is collected on clinical presentation. That constant update of genoptype-phenotype correlation is made as more knowledge is acquired on the genes, products and their function, and of the change in function induced by different mutations. Guidelines have been designed for a minimum set of clinical information to be used by researchers when describing genetic non syndromic hearing impairment. This can become a useful tool to monitor differences in clinical characteristics introduced by various mutations within the same condition or among different conditions [Ref. 30]. Guidelines were also proposed for testing and interpretation of results of connexin 26 [Ref. 29] to try to comply with problems concerning genotype-phenotype correlations and clinical interpretation of genetic results.