Учебники / Genetic Hearing Loss Willems 2004

158.Lanford PJ, Shailam R, Norton CR, Gridley T, Kelley MW. Expression of Math1 and HES5 in the cochlea of wildtype and Jag2 mutant mice. JARO 2000; 1(2):161–171.

159.Muller U, Littlewood E. Mechanisms that regulate mechanosensory hair cell di erentiation. Trends Cell Biol 2001; 11(8):334–342

160.Leon Y, Vazquez E, Sanz C, Vega JA, Mato JM, Giraldez E, et al. Insulin-like growth factor-I regulates cell proliferation in the developing inner ear, activating glycosylphosphatidylinositol hydrolysis and Fos expression. Endocrinology 1995; 136(8):3494–3503.

posing cochlear implants or genetic testing (1,2). On the other hand, some 90–95% of deaf children are born to hearing parents who are likely to ask many questions such as ‘‘Why is my child deaf?’’ ‘‘Will his or her hearing get worse?’’ or ‘‘Will my other children be similarly a ected?’’ (Table 1). Not surprisingly, then, most of these parents are in favor of genetic testing (3).

Only upon establishing the precise cause of a child’s deafness can we begin to answer questions such as those above with any degree of certainty. At least half of all prelingual cases of deafness have a genetic etiology, with at least 80% being inherited in an autosomal recessive manner (4–7). Thus upon presentation of a case of isolated deafness within a family wanting recurrence risks for deafness, only genetic testing can confirm a genetic etiology. Without a confirmed cause one can only provide the family with an empirical or average recurrence risk, which is usually given as somewhere between 1 in 5 to 1 in 18 in such cases (8,9). With our rapidly increasing understanding of the genetic causes comes the prospect of providing many families with a precise genetic cause and hence accurate recurrence risks and perhaps more information on issues such as the severity or progression of deafness, carrier status, and perhaps even prenatal diagnosis.

Many genetic disorders are caused by mutations in a single gene, e.g., cystic fibrosis, Huntington’s disease, and as such genetic counseling is largely straightforward. With perhaps as many as a hundred genes involved in deafness, however, genetic screening and counseling is a big challenge. Yet already research findings are having a positive impact, most notably the finding that mutations in a single gene, connexin 26 (GJB2), can account for up to 60% of prelingual nonsyndromic hearing impairment in some populations (10,11). Consequently the precise cause of a significant proportion of such cases can be determined with a relatively quick and simple diagnostic test. Unfortunately, the genetic heterogeneity of hearing impairment means that a single, simple genetic test for all a ected individuals is not yet available. Despite the rapid pace of genetic studies caution needs to be taken to ensure expectations of the general public are not raised too high and that the ethical implications of this work are fully realized, not least those of the

Table 1 Reasons for Genetic Testing

Define diagnosis/cause

Provide recurrence risks

Alter medical management

Determine carrier status

Prenatal diagnosis

Prepare for future, e.g., language and schooling needs

Terminate pregnancy

Deaf community themselves many of whom are fearful of the impact of this work. Note that ‘‘Deaf’’ refers to individuals who are culturally deaf, viewing deafness from the cultural or sociological perspective where deafness is a condition to be understood and preserved.

II.TESTING FOR DEAFNESS

Prelingual hearing impairment has a dramatic e ect on speech acquisition and consequently cognitive social development and education. The birth of a hearing-impaired child to hearing parents can often cause great concerns. With early identification and appropriate intervention before a child reaches 6 months language development can often be normal, while late diagnosis can have a devastating e ect on language acquisition, communications development, confidence, and social skills (12–14).

Depending upon where a child is born his or her chances of being detected at such an early age can vary. In England it is estimated than around 50% of the 840 children born each year with a permanent hearing impairment will not be identified until they are 18 months old, with 25% still left undiagnosed at 3 years old (15). Steps to introduce universal neonatal hearing screening, a noninvasive test measuring otoacoustic emissions, will hopefully lead to early detection in a much higher proportion of newborns (16).

In addition to the 1 in 1000 children with a prelingual hearing impairment a similar number will develop a hearing impairment before their teens (17). Although language will have been acquired by this age a prompt diagnosis is still essential to ensure the child’s continued educational and social development.

The most common form of hearing impairment is that a ecting adults. Approximately half of all adults over the age of 65 have a significant hearing impairment, which we tend to accept as part of getting old (18). The causes of this age-related hearing loss are less well understood and are likely to be much more complicated than childhood hearing loss. Nevertheless agerelated hearing impairment has a significant social impact and it seems likely that much of what we learn from studying childhood-onset deafness will help us to develop therapies to slow down or prevent age-related hearing loss. Since certain genetic backgrounds may make persons more sensitive to environmental factors such as noise, the issue of predictive testing for adult-onset deafness also exists. In fact it is already known that individuals with a particular mutation of the mitochondrial DNA (A1555G) will su er a permanent hearing loss shortly after taking aminoglycoside antibiotics (19).

Although genetic testing has the potential to detect children or adults who will become or are at risk of becoming deaf it seems highly unlikely, particularly given the costs, that genetic testing will be used to screen all

newborns for deafness genes. At present, genetic testing will be of most value in establishing the precise cause of deafness in an individual who is deaf, which will help him and his family answer many questions they might have.

III.DEAFNESS GENES

Just a cursory glance of the contents of this book shows the remarkable progress that has been made in the last 5 years in establishing the genetic causes of hearing impairment. More than 90 loci for nonsyndromic hearing impairment have been identified; 51 autosomal dominant, 39 autosomal recessive, and six X-linked. To date more than 20 of these genes have been identified. In addition, more than 30 genes causing syndromic forms of deafness have also been characterized (7,20). As we isolate more genes and study their e ects more closely the complex nature of these deafness genes becomes apparent. Mutations in the same gene can cause both syndromic and nonsyndromic deafness (e.g., MYO7A can cause Usher syndrome type 1b or NSSHI), others can cause both dominant and recessive forms (e.g., both DFNA8/12 and DFNB21 are caused by mutations in the TECTA gene), while the e ects of others may be modified by other genes (e.g., DFNM1 protects against DFNB26). Such findings are fascinating for the scientist researching these genes, posing many more questions, but also highlight the di culties sometimes facing the genetic counselor when dealing with a family with deafness.

IV. ESTABLISHING THE GENETIC CAUSE

With perhaps as many as 100 di erent genes involved in hearing impairment where does one start? Clearly screening all of these genes is not practical so steps must be taken to narrow down the number of potential genes as much as possible. Three key steps in this process are: clinical diagnosis, pattern of inheritance, and appropriate genetic testing. These are not mutually exclusive as the identification of the genetic cause may help establish the precise nature of a syndromic form of deafness for example.

An accurate clinical diagnosis is important not only for the individual’s health but also in directing the appropriate genetic testing. Approximately 70% of genetic hearing impairment is nonsyndromic whilst the remaining 30% is syndromic of which many hundreds of syndromes have been described (9). Identifying a syndromic deafness significantly reduces the number of genes to screen, to perhaps half a dozen or fewer.

In the absence of any family history establishing the precise clinical phenotype can be di cult, particularly where some features may not mani-

fest themselves until later in life. For example in Usher syndrome type 1, retinitis pigmentosa does not normally develop until the end of the first decade or shortly thereafter. Appropriate genetic testing, however, could lead to an early diagnosis of a syndromic form before many clinical features manifest themselves, allowing extremely early treatment of such conditions. Conversely, genetic testing may be able to exclude the possibility of such conditions developing in a deaf child, thus allaying any fears the parents may have.

The presence of other a ected family members will of course help establish a diagnosis and can help determine the pattern of inheritance of the genetic defect, directing genetic testing. In large enough families linkage analysis can be used, at the very least to exclude certain loci. Deafness in more than one generation may imply an autosomal dominant gene while the absence of any family history suggests the trait is most likely autosomal recessive, especially if consanguinity is present, but one cannot rule out the possibility of a new, spontaneous mutation.

In determining the pattern of inheritance one needs to beware of possible confounding factors such as reduced penetrance or the presence of modifier genes. Quite often other family members may not be available for examination and thus a verbal account of a distant relative being deaf needs to be taken with caution, especially in cases of late, adult-onset hearing impairment. Likewise environmental causes, such as meningitis, ototoxic drugs, or infection mother had in pregnancy, must also be excluded as the presence of a phenocopy may give the impression of an incorrect pattern of inheritance.

Having established whether the deafness is syndromic or nonsyndromic, its pattern of inheritance, and any other information, one now has a much better idea of which gene(s) is likely the cause. For example Cx26 is likely in a child with prelingual, severe, recessive deafness whereas a child with postlingual, low-frequency hearing loss inherited in a dominant manner is much more likely to have a mutation in the WFS1 gene. In fact, autosomal dominant deafness appears to be more varied than autosomal recessive deafness; i.e., some forms are prelingual though most are postlingual, some a ect high frequencies while others a ect low frequencies. Consequently in dominant families it is possible to narrow down, somewhat, the number of candidate genes based on the clinical phenotype. Unfortunately this does not appear to be the case for the recessive genes, and as such, determining a genetic cause is probably most challenging for nonsyndromic deafness.

At present screening of only a handful of deafness genes is carried out by DNA diagnostic laboratories on a routine service basis. Most mutation detection is only done on a research basis but it is likely that as more genes are identified and mutations within these genes are characterized more of this work will be moved to a service-based setting. Which genes are screened and which are not will depend primarily on how common a cause each is of

deafness. Thus routine screening for the Cx26 gene is now common and is o ered by DNA laboratories all over the world.

Strategies for screening genes will also need to be developed with the primary aim being to identify the genetic mutation in as many cases as possible. Thus routine screening of Cx26 usually takes account of the fact that a deletion of a single guanine in a stretch of six G’s (35delG) accounts for 70– 80% of DFNB1 cases in many Caucasian populations (10,11,21). However, the 35delG mutation appears to be almost absent in some parts of the world, e.g., Asia and Africa (22,23). In Japan the 235delC mutation is much more common, being present in 1–2% of cases (22,23), while the 167delT is present in 4% of Ashkenazi Jews (24). Specific screening strategies will therefore have to be developed for each country or population.

Cx26 has a single coding exon and as such is simple to screen. However, many of the other genes causing deafness are considerably larger, several having 50 or more exons. In these cases the best way of screening is to follow the example of the cystic fibrosis (CF) gene. The CF gene has 27 exons but from numerous studies it has become apparent that certain mutations are much more common than others. Consequently screening the CF gene now involves using a kit to detect the 10 or so most common mutations within that population. Such a strategy will be most appropriate for many of the deafness genes, though at present such data are available on very few of these gene to allow us to build up such a mutation profile of each gene. One example is the PDS gene where, from several surveys, it appears that mutations in certain exons are more common than in others; e.g., in Caucasians about 60% of PDS mutations lie in exons 6, 8, and 10 (25). Thus a strategy for screening the PDS gene would begin with these exons, though again, population di erences are likely to exist.

Ultimately any diagnostic service o ered will of course be largely limited by costs. Screening a very large gene that perhaps accounts for less than 1% of cases is unlikely to be o ered on a routine basis. As technology advances so will the capability to screen for increasing numbers of mutations. The costs of such a service will also have to be balanced against the benefits to the individual. At present, apart from providing accurate genetic counseling, the identification of the defective gene in nonsyndromic deafness o ers little direct health benefits to that individual. However, this may change as therapies are developed and genetic testing is already beneficial in determining whether or not the deafness is syndromic.

V.POTENTIAL PITFALLS AND PROBLEMS

Given that we have not yet identified all of the genes involved in deafness, it is clear that the genetic cause in many cases cannot yet be established. Even

where the gene has been identified there are still some instances in which it may be di cult to interpret with any certainty the e ect of a particular mutation that is found (Table 2). In some cases these may even raise more problems than they solve, making genetic counseling di cult.

The first question asked upon finding a new mutation within a gene is what e ect does that mutation have on the gene product? In the case of nonsense or frameshift mutations it is apparent that a truncated gene product is unlikely to function normally. Further studies to assess the mutation frequency in the general population and deaf persons will help determine whether a mutation is indeed pathogenic or merely a polymorphism. Missense mutations or those that might perturb a splice site are more di cult to interpret. Proving that a particular mutation causes deafness can sometimes be very di cult.

Learning more about the gene product helps us understand which nucleotides and amino acids are most important and likely to be pathogenic if altered. Animal models provide an excellent way of studying the role of specific genes in hearing, though species di erences can prove problematic. The most notable example is that of Cx26 where the knockout is embryonic lethal in mice owing to di erences between the human and mouse placenta (26). Similarly, functional and mRNA splicing studies, which are usually beyond the scope of diagnostic laboratories, do not always extrapolate to humans. One excellent example of this is also provided by mutations in the Cx26 gene. Using in vitro expression systems such as Xenopus oocytes, the M34T mutation has been shown to cause impaired intercellular coupling and abnormalities of tra cking and targeting of Cx26 (27,28). Although this work suggests the M34T mutation is dominant, clinical data in humans suggest it is recessive or even just a benign polymorphism (29–31). Thus the status of this mutation remains uncertain and providing counseling for families with such mutations is di cult as the genetic cause has not been determined for certain.

Although screening Cx26 can establish the genetic cause of perhaps 50% of autosomal recessive cases, the high carrier frequency of Cx26 muta-

Table 2 Potential Di culties in Establishing a Genetic Cause

No family history

Genetic heterogeneity—which gene to look at? Is the mutation pathogenic? (e.g., missense) Only 1 recessive mutation found

Animal/functional studies may not extrapolate to humans Does the mutation cause syndromic or nonsyndromic deafness? Reduced penetrance of some mutations

Mutation does not predict severity/age of onset Ethical considerations

tions [1 in 31 in Spain and Italy (20)] can also cause problems. Many deaf persons will be incidental carriers of Cx26 mutations, a separate gene being the cause of their deafness. Yet it would be unwise to dismiss the presence of a single recessive mutation in the Cx26 gene in a deaf child. A recent study has shown that a large deletion just upstream of the Cx26 gene, acting in trans with a mutation in the other copy of the Cx26 gene, or in the homozygous state, leads to deafness, possibly through disrupting the regulatory region of the Cx26 gene (32).

Such a scenario highlights how the absence of a mutation in a gene must be treated with caution and not taken as absolute and demonstrates how genetic testing needs to be continually updated as further findings come to light. In fact, the vast majority of deafness-causing mutations reported to date have been in the gene transcript. Yet the possibility remains that mutations in regulatory regions of these genes also exist that contribute to deafness. For instance, mutations in the PDS gene are thought to account for as much as 10% of childhood deafness (9) and as such the carrier frequency of mutations within the PDS gene may be relatively high. There are several reports of deaf individuals with inner ear defects and mapping to DFNB4 in whom only a single mutation in the coding region of the PDS gene has been found (25). As the regulatory regions of the PDS gene have not been characterized it is not possible to determine for certain whether the PDS gene is the cause of these individuals’ deafness.

Although identification of a genetic mutation can confirm a genetic cause there is often no clear genotype-phenotype correlation. For example, even siblings homozygous for Cx26 35delG may display variation in age of onset and severity of hearing impairment. This is even more extreme where mutations in a single gene can lead to both syndromic and nonsyndromic forms of deafness, e.g., COL11A2-Stickler syndrome or DFNA13; MYO7AUsh1b or DFNB2; PDSPendred syndrome or DFNB4; WFS1Wolfram syndrome or DFNA6/14 (7). In some instances the identification of a mutation in the relevant gene can confirm a diagnosis of a syndromic form. On the other hand, the identification of a mutation in, for example, the MYO7A or CDH23 genes in a child with prelingual hearing loss raises the possibility that that child may develop ocular problems later in life as part of Usher syndrome. Since we do not yet understand how mutations in these genes lead to Usher syndrome as opposed to nonsyndromic deafness, advising the family can be extremely di cult. While one does not unnecessarily wish to cause parents concern they are likely to want know as early as possible if their child has Usher syndrome so that they can prepare for the future of that child. Of course, the identification of mutations in another gene, such as Cx26, can exclude the possibility of syndromic deafness and help reassure the parents.