<TITLE: Genetics of Hearing
ACADEMIC DOMAIN: medicine
DISCIPLINE: genetics
EVENT TYPE: lecture
FILE ID: ULEC130
NOTES: includes a short discussion at the end

RECORDING DURATION: 54 min 50 sec

RECORDING DATE: 21.2.2007

NUMBER OF PARTICIPANTS: circa 25

NUMBER OF SPEAKERS: 4

S1: NATIVE-SPEAKER STATUS: Finnish; ACADEMIC ROLE: senior staff; GENDER: male; AGE: 51-over

S2: NATIVE-SPEAKER STATUS: Dutch (Belgium); ACADEMIC ROLE: senior staff; GENDER: male; AGE: 31-50

S3: NATIVE-SPEAKER STATUS: Chinese; ACADEMIC ROLE: unknown; GENDER: male; AGE: unknown

S4: NATIVE-SPEAKER STATUS: Finnish; ACADEMIC ROLE: unknown; GENDER: female; AGE: unknown

SS: several simultaneous speakers>


<S1> so dear friends we are about ready to start i expect that there will be people dropping in still during the presentation and , for for me personally it's a great pleasure to have <NAME S2> here today because <NAME S2> is the key person establishing the modern genetics in the hearing area he started about 15 years ago by putting the home pages of hearing genes and er whenever do you look what's happening in the genetic fields of hearing was it's just easy way to go in the pages and look all these notations actually these pages make me dizzy because they are quite difficult to @read@ there are abbreviations and synonyms and tables but it it's a good source of knowledge and they're linked to original papers er in a way it's a bible er for modern genetics the previous bible was the syndrome book of different gene- er of different syn- syndromes and er that was about in 1950s 60s so r- really <NAME S2> has been a person that has promoted the hearing genes and he is in advancing in several important topics he is leading one of the genetic hearing groups called age-related hearing impairment and the coordinator for that and found several interesting genes in this very complex genetics that are determining how well we will hear when we get older and er in addition for that er he is also participant of the eurohear project which is huge project which are cloning some other genes for the hearing and er for me it's a great pleasure that we have been working for several years jointly , and it has really been kick for for also this university work because we have collected from the hearing centre a great number of samples for this project please <NAME S2> <S2> yeah [er] </S2> [start your] @lecture@ </S1>
<S2> <REFERS TO POWERPOINT SLIDES THROUGHOUT THE LECTURE> thank you very much er <NAME S1> it's er really my pleasure er to be here in er in tampere er it's an er okay thank you it's an experience er not only to visit here but also to er experience these very low temperatures er coming from belgium this is very unusual it's actually the first time er that i know what minus 25 really er means er when you're out in the street <SS> [@@] </SS> [okay erm] i would like to start this er presentation by the first ten minutes or so give you an overview of er er hearing impairments and what has been done er genetically so hearing impairments you can er certainly find it in congenital forms and in adult forms congenital forms are quite frequent one in about 650 children as we know now from neonatal er hearing screening and erm adult forms er of course are mu- even much more frequent at ten per cent by the age of 60 50 per cent by the age of 80 as er hearing impairments er that interferes with communication in some way er and there are some er common themes in the aetiology between these two but there are also er very important differences er they're both caused both by genes and by the environment and er genes are involved er say about 50 per cent in both forms but in the congenital forms it's either the gene that is causing it and i mean causing monogenic cell mutation in one gene er in most in most children is causing the hearing impairment so the purely genetic er hearing impairment we call you also call it mendelian because if it's er inheritance it will be either inherited er so dominantly or recessively er er or it can be an environmental cause and then it's a completely different cause and then genes have very little to do with it with the adult forms it's ma- mainly the complex disease and here it's not either the genes or either the environment but it's an interplay between the two so although genes are involved er for about 50 per cent it's erm the environment that causes the the hearing impairment but genes that make people susceptible so it's an interplay between the environment and the genes er that cause it so that's the very important difference between the two er so for the last er say ten years or more that ten years er genes for hearing impairment have been found but it's always in the first category where it's er the monogenic category so where the disease is er the disease is caused by by a single gene and the first gene for hearing impairment was cloned in 95 and it has been going up in a more or less linear way since then and we are well over 40 now so different genes causing non-syndromic hearing impairment of course there are also syndromic forms and many other genes are responsible for that i will not go into that er but for the non-syndromic forms so the pure hearing impairment er we know many different genes and these er genes lead to a very similar phenotype so the hearing impairment in in many cases is very similar or even indistinguishable the different genes can er can cause it and hearing impairment is one of the erm , i think most extreme examples of what we call genetic heterogeneity so that a single disease can be caused by many different genes and although we are at about 45 many more remain to be identified so er estimates are that there are er that there are at least 100 different genes that can cause non-syndromic hearing impairment er these genes are very different in nature and this is just a slide of a section to the ear where some of these genes the expression has been in- er of these genes has been indicated you have extracellular matrix proteins like COCH er expressed here in this spiral ligament or in in and in the limbus er you have er intracellular genes like myosin-7A which is an intracellular motor protein which is expressed in the hair cells you have alphatectorin which is an extracellular matrix protein er which is the most important structural component of the tectorial membrane you have gap erm gap junctions you have er er potassium channels in the stria vascularis for instance you have many different genes but in the end if there are if if there is a defect in one of these genes the result is the same and the hearing in- er the hearing apparatus will not function er properly and you will get a hearing impairment erm i will now give you a few slides from er from a a review where there is a very nice correlation between the genetic research people looking for genes and identifying genes responsible for hearing impairment and on the other hand erm the er inner ear biology and people looking into the function of the ear and into the molecular components responsible for that and these are stereocilia so the the hair cells er er on the er er er surface have these er these hairs the stereocilia and they are linked with different links like er tip-links and side-links and ankle-links and this has been known because of electromicroscopy that has been performed er but the the molecular nature of these links er was never really er solved and it was by genetic means that all these components really have been identified er so here you have these different links and you have all these symbols and here is the legend so you can see that the different molecular components or some of them have been identified there are adhesion proteins like cadherin-23 or protocadherin-15 erm they are linked through scaffolding er proteins like (xx) or harmonin or SANS er through er these er er actin filaments that are present inside the stereocilia there are also motor proteins these myosins are are present here so that movement i- is er is possible so you really have a link of the adhesion proteins to the er intracellular s- er cytoskeleton and none of these er genes would be known if it were not for genetics that has identified these through family studies and now later on through f- functional studies they found out what the function of these proteins are so i think it's really a very nice example of how genetics can help to er understand the working of the inner ear and these are these different er genes as you can see like cadherin-23 and protocadherin-15 that i mentioned the myosins and these are responsible for different types of deafness and DFN-A means a dominant type of non-syndromic deafness DFN-B means a recessive type of non-syndromic deafness er USH means usher syndrome combination of deafness and blindness so you can see there's these genes either cause dominant de- non-syndromic deafness or recessive non-syndromic deafness or usher syndrome or sometimes two so you can see here cadherin-23 causes a recessive type of non-syndromic deafness but other mutations cause usher usher syndrome so first these loci have been found genes have been cloned studied functionally and now the function er was understood and then the interplay between these er different molecules has been studied er to find out to find out that they are involved in the workings of the of the stereocilia so i think that's a a nice result also resulting from er genetic studies but er so another result from genetic studies is really that we can do genetic diagnostics and genetic diagnostics are important for several reasons for instance a child is born er it gets a hearing test er it's not hearing er so then we want to know why why is it not hearing er will there be be other problems is this syndromic er well is this usher syndrome maybe will this er child develop blindness later in life these are all very important questions so and er er sometimes i think clinical testing can solve some of these questions sometimes not and if we have a genetic diagnosis then we know which gene is responsible and then we can predict what the natural course of er of this child will be er but there are many problems for genetic diagnostics for hearing impairment i told you that it's very genetically heterogeneous so very many genes to analyse and some of them not yet known so that's not very good for genetic diagnostics erm there are many different mutations er if you take cystic fibrosis for instance there's one mutation that makes up 85 per cent of all mutations and there is only one cystic fibrosis gene so if you have a child with cystic fibrosis you know which gene to screen there's only one and you know which mutation to screen because there's one major mutation and if you screen for this one mutation which is very easy then in most of the cases you already have done what you should do and if you don't find it in this one mutation you screen the gene for mutations only one gene so you're done with deafness you have to screen 54 or 45 genes and even if you don't find it there well maybe you missed it because the gene has not been identified yet so it's er very difficult and also these are often large genes er cadherin-23 is extremely er large some of the other genes are very large so this is fairly difficult and er it's even getting er more and more difficult as more and more genes for deafness are being found more large genes the task er is becoming more and more difficult so there's really an increasing gap between basic rese- research and diagnostic application because diagnostic application also depends on money depends on social security or what the state is really er putting into this it's very expensive so the amount of money available for genetic testing is limited of course here so you cannot spend a fortune on every patient that's er that's obvious erm so er er some of you might be interested in er DNA diagnostics or it sounds very pessimistic from the previous er slide but there are some er some things that are f- possible so if you have probably autosomal recessive inheritance so normal hearing parents and a non-syndromic deafness in your child then there are some things you can do and the first thing is connexin-26 that you probably know er it's a very small gene so that's nice and it's responsible for quite a lot of cases so this is interesting and it has a recurrent mutation so one mutation makes up a ma- er majority so this is an exception really and it's very good that this is er possible so a lot of testing for connexin-26 is done then there are a few other genes that you might screen for instance if erm the tests indicate that this is auditory neuropathy then you should check for otoferlin mutations because that that's up to now the only gene er that's really known to er to cause this auditory neuropathy if er er CT-scanning or imaging of the inner ear shows an enlarged vestibular aquaduct then the pendrin er gene is something you should screen for quite often it's mutated er so that's another hint that er er you can have and it seems that TMC1 is another deafness gene that might be mutated more frequently so er some studies are being done to see how frequently and in the future it might s- also er be a candidate for for DNA diagnostics so this is if er if it's recessive and er most congenital deafness is recessive er you also have some dominant forms of monogenic hearing impairment so if it's non-syndromic and dominant so you seek in at least two generations the parent and the child and you have good indication that it's dominant and there are a few genes you can check but that's very limited if it's the low frequencies that are affected very rare but it does happen then you have to screen the WFS1 gene because erm chances of finding mutation if it's dominant and low frequency are very high er up to 80 per cent in all er if we get a patient with a low frequency hearing impairment and in two generations then we find a mutation in this gene doesn't oc- er very often happen but if it happens chances are high then there is a typical er mutation in the COCH gene which is very frequent in belgium and the netherlands and there's one with vestibular involvement late-onset so they lose their hearing in their 30s or their 40s quite rapidly er nearly profound hearing impairment by the time they are 60 er and together with their hearing er they have vestibular problems instability in the dark so the vestibular system er really goes down until until there is complete (xx) and er this is a very recognisable phenotype er hundreds of patients in belgium where i live er we have ma- had many er full DNA diagnostics and it's one of the very few instances where you can really predict if you look carefully in the phenotype the clinical characteristics of the patient you already know beforehand whether you will find a mutation or not because it's really so typical and er but this is not frequent in the rest of the world it's a (xx) it's autosomal dominant er and also TMC1 seems to be quite frequently involved in dominant er families we found quite a few dominant families with TMC1 in their genes er and then just to be complete there is mitochondrial inheritance and a limited number of mutations causes syndromic and non-syndromic hearing impairment there's also probably you know er aminoglycoside susceptibility so if you have a patient with hearing impairment and the patient seems to be exceptionally sensitive to aminoglycosides then there is one mutation you have to check the 15-55 mutation er so that's very important er this is not very impressive but still some things are possible for DNA diagnostics now i would <COUGH> excuse me like to tell you a little bit about connexin-26 erm you probably know about it so i will go over it very quickly but i would like to move in the direction of complex diseases and i think this is a good example connexin-26 is very frequent in the mediterranean population responsible for more than 25 per cent of all cases more than 50 per cent of all genetic cases there is one mutation called 35-del-G so deletion of the single gene in position 35 er responsible for about 85 per cent of all the mutation and the carrier frequencies at least around the mediterranean is about one in 35 so it's very frequent there in if you go up to the north in in europe usually it's less frequent er in belgium it's somewhere in between one in 50 or one in 100 as the carrier frequency i think finland is about one in 50 also carrier for connexin-26 so you also have er conne- connexin-26 deafness in finland er no doubt erm er connexin-26 traditionally in the literature you find that the function is really the transfer of potassium ions back to the endolymph erm you know that er hearing in in the hair cells really stop with the entering of potassium ions from the endolymph which with a very high potassium concentration so they enter the hair cells they leave the hair cells at the erm at the base er through KCNQ4 er channels they are taken up by the supporting cells here and then er transferred from one cell to another because these are gap junctions er the potassium is er recycled back to the endolymph this is what is traditionally thought to be the function of connexin-26 but some recent evidence also suggest that not only potassium but also other molecules are transferred from one cell to another like er second messengers like IP3 er fabio mammano in italy has showed very nicely er that connexin-26 gap junctions transfer IP3 so there might be much more than just the potassium that is being transferred er between these cells where connexin-26 is er is expressed <COUGH> er we recently performed a genotype-phenotype correlation study worldwide with 26 laboratories and we collected er audiometric data on more than 1500 patients we found many different mutations some truncating non-truncating er so- some that really destroy the protein and some that er might change only one aminoacids and we category these as non-truncating er 35-del-G has a major mutation of course but then besides 35-del-G only a handful of mutations for in the one to two per cent range and all the others were very er infrequent , erm the first thing we found was that truncating mutations were much worse than non-truncating so among the people with two truncating mutations because it's recessive you have two mutations er you see that the large majority have profound deafness mild was s- er deafness was very infrequent moderate also er some severe and the majority profound but if you look at the group with two non-truncating mutations you see that most were mild er some were moderate and severe and profound was really a mino- minority so there was a really striking difference and we could say that truncating mutations on average were more severe than non-truncating mutations so that already gives an indication and because we had so many patients in this study 1,500 we could look at the individual individual genotypes itself so this was a reference group and this is 35-del-G homozygotes so both chromosomes er 35-del-G and erm this is a puretone average so this is profound deafness here you see 110 120 er decibels and this is here mild and moderate hearing loss so you can see er one patient here with a er particularly mild er phenotype er mild hearing loss about 20 dB puretone average but some er in this region but not er very many and most of them are profound erm we found that s- so these are other truncating mutations most of them are in the same range but one mutation was more severe than 35-del-G and that's a big deletion er upstream of the gene and it deletes another connexin located er in that region connexin-30 so possibly for that region er the hearing impairment on average in these patients is more severe there is a (xx) mutation also and (xx) mutations are known to be very unpredictable and er this was statistically different from our reference group so the hearing loss in the (xx) mutation was r- also ranged from from mild to er to profound er maybe not so interesting but the interesting part was when we came to some of the non-truncating mutations and a few of them are really frequent like V37I maybe not so frequent in the western world but in certain part of asia er er like er china V37I is very frequent sometimes with carrier frequencies up to ten per cent of the population and there is really very much debate whether it's a real mutation or whether it's a polymorphism and does not cause hearing impairment the same is M34T it's very frequent also in the western world er but it there were some doubts whether it was really pathogenic or whether it was just a B9 polymorphism and our data really suggest that this mutation in combi- er in combination with 35-del-G er cause er a mild or moderate hearing impairment as you can see all these these dots are patients and they all really cluster here in the mo- more mild er region so i think our results really indicate that these two very frequent variants that they are pathogenic that they cause a hearing impairment but that the hearing impairment in a large majority is er mild er or moderate , erm what we really want to do now is to go further and erm look at the phenotypic variability and these are 35-del-G homozygotes so erm let me maybe go back to the previous slide here is a reference group as you can see most are in the profound group here some are mild and moderate if er we look at this in this histogram then these are numbers of patients and these are categories in in five dB going from mild here very few or from moderate severe profound as you can see the profound category is a lot most largest category but these are just 35-del-G homozygotes so the same genotype the same mutation so even among patients with the same mutation you get a different hearing loss and you can ask the question why is this patient so mildly affected he has two very clear connexin-26 mutations and these patients are completely deaf and these er can even do without the hearing aid er and function relatively er normal so er what's the difference between the two well as a geneticist we'd we believe that modifier genes might play a role and that these people might have other variants in other genes that will protect them against the deafness caused by connexin-26 and it will not not be very frequent but in some cases this might be the case another re- explanation might be an environmental factor but as this deafness is present from birth it has to be something maybe during pregnancy or so we don't know and we are interested to s- to know why there is this difference because if you know why maybe there could be some way of influencing this and maybe this could be a lead to a new and much better form of of therapy of course they will have cochlear implants and many of these children er i think most of them will receive a cochlear implant and do very well but if we know why maybe there are leads to to er something else so that's that's might be very interesting so we were thinking about genetics studies to to look for modifier genes whether we could find them and and we made a model and it's not so er not so interesting this so we we think okay if connexin-26 causes 150 dB of hearing impairment and there is er modifier genes influencing it with 15 decibels can can we find it er with a typical er snip which is a genetic marker with a typical er allele frequency of a a minor allele frequency of point two so 20 per cent of all chromosomes in the population would have this so this is something that's quite typical and then we found that under these circumstances we would need about 750 of these patients to have a reasonable power er to do such a study and we have these patients from our multi-centre study that i just talked to you about but these were just audiometric data so we need DNA from them and that's what we're really doing now is collecting as many patients as possible er and preferably those that have mild hearing impairment mild or moderate because these are the most rare and if we have enough of those we will start er our erm aural genetics studies we will do a whole genome association study which is the latest er form of genetic studies using these DNA-chip with er a large amount of markers up to er 500 or a 1,000 markers on on single chip and use these to try to identify these modifier genes so i think now we made er a transfer er from er purely genetic er diseases like er er GJB2 er er which is connexin-26 to complex inheritance because diseases you can really classify as being caused by the environment therefore hearing impairment like noise or infections or purely genetics like connexin-26 or KCNQ4 to some intermediate part where there is an interplay between the environment and er and genetics and one thing that is is known in that is is that age-related hearing impairment is a well-known example of something that's er has to do with er with both although i have to say about this slide that connexin-26 also is really not only purely genetics a- as we saw that among patients there's really really a difference and it might be that some environmental factors also play a role or other genes making it more complex so making it more towards er the centre also noise for instance is of course purely environment but some people are more susceptible than others so and genes er play a role in that so noise-induced hearing impairment also is a little bit more complex so this is really a little bit sim- simplified , erm if we take the example of age-related hearing impairment then er i have to tell you that erm people will lose their hearing in the high frequencies and this is er the B-50 so the medi- median of the male population er at 20 er people are hearing very well but at 70 you see that this is er the median er thresholds of er of the population and it really goes down and i think 60 is difficult to see probably but it's in it's in between here so you lose some decibels in the high frequencies once you get over 20 or 30 then er you start losing these and er males are more more affected than females but also females er have similar although less pronounced but er if you look really and this is er very vague er er very rough that this is an area that that's maybe most important for speech understanding then you can see that er er from an average male from age 70 might er start to experience some difficulties in normal er spoken conversation er of course this is a median and er half of half of the population would not have these problems and the other half will have much more problems er than this actually there is a big difference between individuals as y- you can see this is the fifth percentile and this is the 95th percentile it's the same graph so 70 years at the of age male and you can see that one in 20 of these 70-years-old will have a very good hearing and er will have hardly any hearing loss in the high frequencies while one in 20 will have a really bad hearing and will really need a hearing a hearing aid because he will have lost most of er this area that's important for for er erm normal speech so what is really causing this difference between these good-hearing individuals and these bad-hearing individuals well there have been several studies er including one recently now here in in finland in collaboration with your department and jyvskyl i have understood er looking at what we call the heritability and heritability is really a measure of to what extent genes are involved and erm there are differences between studies but i think on average it's about 50 per cent of this the variance of er of this of these thresholds are caused by genes i think in finland it goes higher than 50 per cent so that might have to do with the special properties of the finnish population which might be a little bit more homogeneous than the rest of europe but er the other studies were er approximately 50 per cent so genes are very important erm in this so we think that age-related hearing impairment is a complex disease with environmental factors and we know some of the environmental factors but also with genetic factors and these genetic factors they are really unknown at this point so we started some years ago a project the european project to look for genetic factors of age-related hearing impairment in collaboration er with <NAME S1>'s department and then groups across er across europe and er this is the ARHI consortium and these are all the cities that were involved er in this study so there is er here the tampere group there was also a group from oulu so two finnish groups and then er from italy er and denmark and er the netherlands and belgium and germany er and wales and erm this is really a phylogenetic tree that we made on the basis of some genetic results (xx) genetic result and the topology of this phylogenetic tree really varied with with some parameters especially in this group er so this was maybe not so reliable but always er no matter how we made this tree the two finnish groups er were clustering together very closely and were very different from the rest of europe so this is well-known among geneticists that finnish population is genetically different from the rest of the europe and the rest of europe actually is is quite similar er in terms of of our genes so what we are doing now is treating these as a single group as a european group and treating the finns separately as a finnish group to look for er for genes erm but before looking for genes we looked into environmental factors we had a questionnaire many patients were er collected and er just to see whether we could find some obvious things like noise and solvents er we could do th- these are all the the different er cities like antwerp er cardiff copenhagen kent (xx) oulu er padova tampere and tbingen i think and you can see red is er it's very significant and and green is er er less significant and and this er yellow is is in between so noise had a very important impact on hearing but of course this is trivial we already knew that but we could look for some other things like smoking and smoking was controversial in the literature but in our study it was quite clear that smoking was was bad for hearing at at an older age so especially if we looked at packyrs which is a er i think the traditional way of looking er of quantifying smoking so we found a very significant result and even in individual populations which are of course a lot er smaller we could we could see in effect that overall it was er it was quite quite clear that smoking had a er a detrimental influence on on hearing er one other er interesting er result was alcohol consumption and this is not really bad for your hearing on the other hand it's protective so if you consume alcohol in moderate amount on a regular basis er then we could show er that er this will protect your hearing erm and we found these results in some of the populations not in all of them just with a simple question er do you regularly er consume alcohol and people with a with a high consumption were excluded so people with moderate consumption and on a regular basis er they had on average a better hearing than people who did not consume any alcohol the strange thing in antwerp is that this was an interaction er with er with gender and that we found the effect in males but we did not find it in females er and er but er over- overall we found that the reason why we did not find it in females is we believe that the threshold might be lower in females because females er metabolise alcohol er less efficiently than men so it might also be protective but to a lesser extent maybe one drink a day might protect you and two drinks a day might not protect you and might even be er worse for for your hearing while men maybe can have two two or three dinks drinks a day er while still bre- being protected so that might be an explanation er here so this is this is quite new and erm er well i think it's an interesting erm side-effect of our of our study what we are really doing in ARHI er in the project now is looking for genes the first thing we have been doing already is the candidate gene analysis er and the first candidates are the ones that are responsible for monogenic deafness so these 45 genes there we have tested them er for association and erm the results are erm , not really that er that exciting in a way that the genes that are responsible for the monogenic hearing apparently do not play a very clear and a major role in age-related hearing impairment so i think we need other strategies and we need genome-wide strategies er a human being has about 30,000 different genes so when testing 45 maybe if we think about it it's not so surprising that we do not find really strong hints so we should look at the other genes and that's what we are doing er we have also collected families we have collected 1,200 er approximately family samples across europe big families about five people in the age-range erm b- being er roughly 50 and 70 er and erm sometimes we even have their their parents present so we can do linkage analysis look at the leads segregating in the family and see whether we can link it to to thr- hearing thresholds erm in the project we also collected sami samples this is work for the the oulu group 315 er people from different villages er among the sami have been collected the sami are genetically also very interesting they're a very old population er their er the size over er hundreds of generations has remained about the same and genetically this means that there is a lot what we called linkage disequilibrium er which really makes it easy to identify genes so these are are being studied we are also using er the latest technology which is whole-genome association studies so instead of using a candidate gene you will use a DNA-chip with 500,000 markers to look at all the genes at the f- at the same time er we're also interesting in interested in er mitochondria and mitochondrial DNA because there are s- there is some evidence that it might be involved in age-related hearing impairment so resequencing mitochondrial DNA in er in some of these i will not go into details of results here erm just tell you that one candidate that we've been looking at is KCNQ4 i alr- already mentioned it it's in the hair cells and it's responsible for potassium leaving the hair cells and this are thresholds of family members with the mutation a monogenic mutation of an antwerp family with a frame-shift (xx) activating mutation as you can see they lose their hearing in the high frequencies low frequencies are spared and it's er you see it starts quite rapidly at the age of er before ten years of age and when they are 50 and 60 you see that most of the high frequencies here is gone erm this is er the er age-related hearing impairment and we were struck by the similarity of the two so that's why it was our first candidate gene we thought okay this is very similar so maybe KCNQ4 is the gene responsible for age-related hearing impairment and indeed in the antwerp er cohort which we studied first we found some nice evidence that KCNQ4 might be er involved if we look at the other european population erm there was also some evidence but it's not so clear so we need more studies on KCNQ4 and er there might very well be be an effect but it's very difficult to prove at this point so i don't think it's really a major gene but it might have a have a contribution i would now like to switch to a few other complex diseases that we're working on and er i would like to say something about noise-induced hearing loss and then i'd with er otosclerosis as an example where we have found some genes that erm really er have very conclusive evidence of er of being involved for noise-induced hearing loss it's maybe not that strong with otosclerosis so as we go from one disease to the other the the evidence will get stronger so this is noise-induced hearing loss so a male age er 30 erm er working for ten years in a factory with a lot of noise you can see that some people are relatively insensitive to the noise and that other people really develop er er substantial hearing impairment and also this er er the variance between so the difference between the individuals we believe is to a large extent genetic er we do not have the direct evidence from age-related hearing impairment where there are lots of studies er this is really deduced from animal models where we see that some (xx) of mice or or some er mutant mice are much more sensitive than than others but we believe that genes also play a role here erm we are collaborating with a swedish group er from rebro and they have collected 114 noise susceptible and er 114 noise resistant er workers so the the ten per cent extremes of the distribution and we are comparing those er while we erm er categorise them also for age and for noise exposure level into different categories there's also a polish worker er population er that we study in collaboration er with people from lodz in in poland er this is er a large er collection of suscet- susceptible and resistant workers and erm we're looking er at the moment at candidate genes and what we mean by susceptible and resistant really is the ten per cent extremes of the distribution so this is to increase er power with a limited number of of samples er we have been looking at a handful of genes involved in oxidative stress because there is a lot of evidence that oxidative stress in the ear er plays a role er in a noise and this is an antioxidant enzyme er er catalysing really hydrogen er peroxide erm into oxygen and and water and erm it's er a gene of 13 axels and we started an association study based on snips that is er very simple genetic markers so this is the gene about 40,000 er base-pairs so an average gene with 13 axels that are er erm separated by introns and these are the genetic markers across the gene that we use and the orange markers er they change the protein so they're changing aminoacids er which is polymorphic in the population and the green markers are just markers we've chosen er from databases there are er very good databases from the hapmap project which is a spin-off of the er erm human genome project so we know the complete sequence of the human genome and we know how to choose the best markers to get all the variability in the gene that's really what what we did so with 12 markers here we can get most of the variability in this genome we did that in the polish and swedish population and we got some interesting P-values these are P-values but we only got er significant P-values if we calculate it statistically with an interaction of er the noise exposure erm if we just do a general er analysis not any significant but if we categorise for different exposure levels then several of these snips give er s- give er erm significant results both in sweden and in poland so these are people in factories with relatively low noise or in workplaces with low noise this is intermediate and this is high noise er more than 92 er decibels and er so we get these significant results and what does it really mean why do we only see it if we take into account this noise er level well erm as you can see from this histogram and this was for all snips exactly the same so that's why we think this is real er if we look at one particular snip for instance we see people with two alleles from one allele and two alleles from the other and these are the heterozygotes so they have one chromosome in one form the other chromosome in the other you see here for instance below 85 the resistant group is much larger among the G-G individuals while if we look in the highest category if there is more noise then the susceptible group is much higher what does this really mean well er it's difficult to tell but one can hypothesise about it for instance you could say that erm a gene here is a variant that will make you hear better and let's let's er imagine that say that people with a G on average hear better so that would explain why the resistant people here are more er there are more resistant people here than susceptible but r- because resistant is really defined as better thresholds so and they work in relatively low noise so if they do not have a lot of damage from the noise then you will have better-hearing people here but if this G does not only make you hear better but this also makes but also makes your ear more vur- vulnerable to loud noise then if you work in really loud noise then your hearing will decline more rapidly than the other ones' who do not hear as well to start with but their hearing also doesn't decline as as much in er in noise so we believe that an explanation like this might explain these different results and af- we as we have found these different results with all these things consistently across this gene we believe er that this gene is really involved in er making people er resistant or susceptible er towards towards noise , erm i would like to end the presentation by saying something on otosclerosis er it's er you all know otosclerosis of course it's an abnormal bone homeostasis er with a fixation of the stapes footplate so you can operate it frequent cause in in like adults i suppose also in finland it's er quite frequent like in the rest of europe and there are some homogenic fa- families er linked with (xx) but no genes have been found that tell er most of them are complex so erm we think environment and genes are involved so we started off with 600 belgian-dutch patients and 300 french patients and we er looked for matched controls we did the study and er we started with er a candidate gene er TGF-beta-1 er probably some people have er heard about it which is er really a a gene in- involved in the induction of chondrogenesis of the otic capsule and it's expressed b- in both adult otosclerotic bone but also during embryonic mhm er development so it really it suggests a role in bo- both on turnover and remodelling of the otic capsule so we thought it was er a good candidate gene so we looked for genetic markers and in the belgian-dutch population we found er one one marker here with an interesting P-value changing one aminoacid in er in this protein and er let me go over this er quite quickly then we looked in the french population and when we looked at the same er variant here again we got a positive P-value er or or a significant P-value and these are 300 french er patients so wanted to take it further and er our french collaborator collected another 300 otosclerosis cases again we er er calculate association and this time the association was even more pronounced and when we now combine the 600 belgian patients and the 600 fren- er french patients and do a simple analysis we get a P-value of ten to the minus six so this is really replicating from time to time so this is very consistent so TGF-beta-1 is really a gene involved in otosclerosis i think this is really very clear and this is proven and er because it changes an aminoacid and because TGF-beta has a growth factor we could do some er experiments er so what we did we we took wild-type er TGF-beta er or we took the variant er with with an ion-position er 263 and we transfected it into er the cell and er er it er it's a complex growth factor to SMAD signalling signalling and it goes to the nucleus (xx) genes (xx) then er transcribed and er what we really did was a number of transfection experiments with luciferase activity and then er we calculated the difference statistically and it was very er statistically significant we found that the protective variant was more active in in our er assay so the protective variant is really more active er a more active growth factor er than er the wild-type er TGF-beta so not only we found a very statistical er statistically significant difference between cases and control for this er variant but also we could show an effect and could show that er this variant really has a higher activity and that the higher activity might protect against otosclerosis so this is really interesting because although otosclerosis er is treated very well in most cases by surgery er sometimes the surgery is not enough you also have you can have a sensorineural hearing impairment which y- you cannot treat or operate er you can also er sometimes after operations er things progress again the sensorineural component can progress so TGF-beta might be er a pathway for a different type of treatment so this is possible so we can really start think about it and a more active form of TGF-beta is apparently protecting against otosclerosis so this might be something er to think about er in the future er okay i would like to end with er this slide it's that er although we are are doing and have been doing some candidate er gene studies i think with varying er levels of success er i think for otosclerosis it's a very clear success for the other diseases it's more erm i- it's maybe not 100 per cent proven now we would like to do genome-wide association studies using er er DNA-chips and we can we think we can do that by er pooling genomic DNA at a reasonable cost individual er genotyping is at this moment much too expensive to do in the future it will be it will be possible and i think this type of studies for the complex forms of hearing impairment in the future i i hope that these will er really give us some nice results and give and will identify some strong genes involved in these diseases er that might on a very long term er give rise to potential new er new forms of treatments for these er very frequent forms of hearing impairment okay thank you very much </S2>
<APPLAUSE>
<S1> so i think <NAME S2> is now open for questions if you have anything in your mind to ask from <NAME S2> , <NAME S3> yes </S1>
<S3> er thank you for your <COUGH> coming er and thank you and the beautiful lecture er you found that er KCNQ4 mutation mhm patients they have high frequent hearing loss <S2> mhm-hm yes </S2> i was thinking about maybe erm er when their cochlea were or the hair cells were si- er stimulated in this high frequency the hair cells they're mhm activated er more than (xx) activity erm than in low hair low frequency so is there there maybe more potassium actually emitted in the hair cells when they are muta- er a- activated er in a high frequency </S3>
<S2> so you you believe that on a physi- physiological basis that more potassium is accumulating in the high <S3> [correct] </S3> [frequencies] and that that might be a reason for the high frequency hearing loss <S3>  yes </S3> yeah yeah yeah yeah i think i think that's er that's possible i'm a geneticist so i'm not so much into electrophysiology of er of the cochlea but it it might very well explain er what we see audiometrically is that in in this family which is actually a family living around antwerp we see a pure high tone hearing loss er in this monogenic family and er so the low frequencies are spared so er i think if er i think you're right when you say that at the high frequencies maybe er more potassium er is is circulating and there are bigger fluxes in er <S3> [yes] </S3> [near] the base of of the cochlea in in the high frequencies so this might be very well er an explanation for the high frequency er (xx) although erm it's more difficult to if you look at other KCNQ4 families because this is with the (xx) mutation it really is strong in protein some other (xx) mutations in other families all over the world they they have another type of hearing impairment which is more pronounced in the high frequencies but also affecting the lower frequencies later in life so it's not really like the ski-slope audiometric configuration but it is more like er erm a line er er so a sloping one but not really er keeping the the low frequencies in- intact so er i i don't know how you would er explain something like that er </S2>
<S3> yeah it was also i think that this is a (xx) </S3>
<S2> that's more [(xx) yeah] </S2>
<S3> [(xx) yes] yeah more (predictable) <S2> yeah </S2> yeah i suppose it's (xx) </S3>
<S2> yeah , thank you </S2>
<S1> more questions . yeah <NAME S4> please </S1>
<S4> i go back to this common connexin er was it 35 </S4>
<S2> del-G [yeah] </S2>
<S4> [del-G] there's where you had this variable in the hearing impairment in people <S2> yeah [yeah] </S2> [and] you are looking for the modifier genes do you have any good candidates is it just picked from the air or are you going find it </S4>
<S2> yeah er well what we really want to do is whole-genome association but of course there are er candidates and the first candidates would be other genes involved in er potassium recycling er we have at the moment erm a number of patients and DNA from a number of patients and we have done a preliminary study with a handful of other er genes that might be involved in the potassium er recycling like connexin-30 also expressed in the cochlea KCN er Q4 KCNQ1 er and and a few others er these do not really give strong signals so these genes are not strong modifiers we already know that from the first study so but this is just a handful of genes among 30,000 as and you have to guess the right gene in the candidate gene approach so i don't think that's a good idea because things might be much more complex than than you can imagine and we know not so much about the physiology of the inner ear so that's why i really want to take it to a whole-gene- genome association study the problem with a whole-genome association study is that replication is key you will get positive and very significant results if you do that just by chance because you do so many tests that some will be very significant so what we have now is assessed that we can do the experiment but we need a second replication set so if we have some really significant results we can do them again in the second set and the one that replicates we think these would will be really true results but therefore we need more patients and especially the mild and moderate ones are rare so if you have any mild or moderate er 35-del-G homozygotes er please get in touch and we would love er to include you in the study so we look we were looking worldwide and er er we still need about 50 or 60 er mild or moderate patients and then we can do the replication and then we will get started with these experiments </S2>
<S1> the problem with 35-del-G is that in in oulu study they really had two per cent of the population with hard of hearing carrier of er 35-del-G so it's in finland it's not so common </S1>
<S2> yeah one in 50 about but it's it's about the same er compared to er to belgium <S1> yeah </S1> yeah what you were referring to is that er we include the age-related hearing impairment also and 35-del-G does not play a role in that er so we thought that maybe 35-del-G carriers would be more prone to age-related hearing impairment but that's not the case </S2>
<S1> okay thank you very much thank you </S1>
<S2> thank you </S2>
