1. Missense mutations of the LCCL domain of COCH cause the dominant auditory/vestibular disorder DFNA9. A National Cancer Institute (NCI) group led by Dr. Colin Stewart has created a Coch partial deletion mouse line in which the entire protein downstream of the LCCL domain, including two VWF domains, has been deleted by homologous recombination. Although they created this line to test a hypothesis about the requirement of Coch for fetal implantation, we evaluated the auditory function of this mutant as a potential model for DFNA9 and the function of Coch in the auditory system. We observed normal hearing in heterozygous and homozygous mice, indicating that the VWF domains are not required for hearing. Our results provide evidence that is consistent with published hypotheses that DFNA9 mutations act via a gain-of-function or dominant negative effect upon other genes, and raise the possibility that Coch itself is not required for normal hearing. 2. In a study of a Korean cohort of deaf probands, we found that the most common mutant GJB2 (encoding connexin 26) allele, 235delC, was probably derived from a common founder. The 235delC-linked haplotypes on these chromosomes were compared to those from 235delC chromosomes from other East Asian populations, and reveal the approximate time of origin of this mutation on an ancient Asian founder. 3. We previously identified and reported a gene, TMC1, underlying hereditary, nonsyndromic recessive deafness at the DFNB7/B11 locus and nonsyndromic dominant progressive hearing loss at the DFNA36 locus on chromosome 9q31. We also identified dominant and recessive mutations in the mouse Tmc1 ortholog in the hearing loss mutants Beethoven (Bth) and deafness (dn), respectively. We have now cloned and sequenced 6 other TMC genes, TMC2-8, which all are predicted to encode proteins with 6-10 membrane-spanning domains. We also identified Tmc homologs in other mammalian and vertebrate species, including Caenorhabditis elegans and Drosophila melanogaster. We determined that TMC6 and TMC8 correspond to the recently reported EVER1 and EVER2 genes, in which recessive truncating mutations cause epidermodysplasia verruciformis (EV). EV is characterized by recurrent and chronic cutaneous papillomavirus infections and an increased susceptibility to non-melanoma skin cancers. In order to understand the structure and function of the Tmc genes and proteins, we have begun the generation of mice segregating knockout (null) alleles of Tmc1, Tmc2, and Tmc3. We have already observed germline transmission of our Tmc1 knockout allele and plan to begin characterizing homozygous mutants in the near future. We have also generated and characterized rabbit antibodies raised against synthetic peptides corresponding to Tmc1, Tmc2, ..., and Tmc6, as well as antisera raised against Tmc1- and Tmc2-GST fusion proteins. We are nearing completion of fusion protein constructs for the remaining Tmc genes for immunization of rabbits. We have completed initial analyses of the immunoreactivity of our antisera on mouse organs of Corti and plan to utilize our knockout mouse tissues as negative control substrates to confirm the specificities of the observed patterns. Finally, we have established heterologous mammalian tissue culture systems for the expression of Tmc1 and Tmc2. We are using these systems to examine physical interactions, document specificities of antibodies, and to search for transmembrane ion movements associated with expression of these proteins. We have generated parallel expression constructs containing each of the known missense mutations in DFNA36, DFNB7, and Bth (mouse) as negative control proteins, as well as for studies of the pathogenetic mechanisms associated with the mutations. We have used a yeast two-hybrid screen to isolate genes encoding proteins that interact with Tmc1. We are using several approaches to test the in situ relevance of these candidate interacting proteins, and are concurrently using the same two-hybrid library to identify proteins that interact with Tmc2. 4. We have made progress in our effort to positionally clone the gene mutated in the mouse Twirler strain. Heterozygous Twirler mice have inner ear malformations and obesity, whereas homozygous mice are born with cleft palate and die at birth. The critical interval containing the Twirler gene is approximately 1 Mb and contains several known genes along with several predicted genes. We have completed nucleotide sequence analysis all of the exons in these genes and not found a likely pathogenic mutation. We are initiating additional analyses to further search for the Twirler mutation and gene. 5. We have identified the genetic basis for a distinctive form of hereditary nonsyndromic hearing impairment in which there is prelingual profound deafness at high frequencies, with variably penetrant postlingual progressive loss of hearing at lower frequencies. The affected individuals are homozygous for a missense mutation, F1888S, of CDH23 (encoding cadherin 23). We used a candidate gene approach to identify a second genetic locus at which a dominant or recessive allele may modify the expressivity of the low frequency hearing loss phenotype in F1888S homozygotes. We are currently testing the functional consequences associated with a missense substitution on a candidate modifier allele. 6. We have shown that one allele (N48K) of USH3 accounts for most, if not all, cases of type 3 Usher syndrome in Ashkenazi Jews (AJs). USH3 was previously thought to be a rare form of Usher syndrome, but we found that USH3 accounts for a high proportion (40%) of all types of Usher syndrome in AJs. Interpreted in the context of studies from Finland, our results indicate that USH3 is common only in populations whose genetic history has permitted the expansion of a founder allele. Taken together with the results of a collaborative study with the Section on Human Genetics (NIDCD) on type 1 Usher syndrome in the same cohort, N48K and a single prevalent founder allele of PCDH15 account for approximately 80% of all Usher syndrome in AJs, providing the basis for efficient and effective molecular genetic screening for Usher syndrome in this population. 7. Enlargement of the vestibular aqueduct (EVA) is thought to be the most commonly detected radiologic malformation in temporal bones of individuals with hearing loss. A significant proportion of EVA cases have been reported to be associated with mutations of the SLC26A4 gene, in which mutations cause Pendred syndrome. PS is an autosomal recessive disorder comprised of bilateral sensorineural hearing loss and thyroid goiter. Goiter is an incompletely penetrant feature, but the perchlorate discharge test is a comparatively sensitive and specific method to detect the underlying iodide organification defect in PS thyroid glands. EVA is a universal finding in the ears of affected PS individuals. We have now ascertained nearly 50 EVA probands at the NIH Clinical Center. Analysis of our first 35 families reveals an unexpected finding: PS is correlated with two mutant SLC26A4 alleles, and nonsyndromic EVA is associated with one or zero mutant SLC26A4 alleles. Furthermore, the range of auditory phenotypes differs between the nonsyndromic EVA and PS groups. The observed correlation of thyroid and auditory phenotypes with SLC26A4 genotype suggests that most, if not all, nonsyndromic EVA is not associated with bi-allelic SLC26A4 mutations. Based upon our data, we hypothesize that one or more other genetic or environmental factors may act alone or in combination with a single SLC26A4 mutation to cause EVA. Our results have immediate significant implications for clinical practice: The detection of a single SLC26A4 mutation in an individual with EVA is not diagnostic for either PS or nonsyndromic EVA.
Showing the most recent 10 out of 76 publications
