Two Different Connexin 26 Mutations in an Inbred Kindred Segregating Non-Syndromic Recessive Deafness: Implications for Genetic Studies in Isolated Populations

Minerva M. Carrasquillo 0 2 4 Joel Zlotogora 0 2 3 4 Saleh Barges 0 1 2 4 Aravinda Chakravarti 0 2 4 0 Medical Center , Jerusalem, Israel 1 Kupat Holim Klalit, Department of Family Physicians , Afula, Israel 2 and University Hospitals of Cleveland , Cleveland OH 44106, USA 3 Department of Human Genetics , Hadassah 4 Department of Genetics and Center for Human Genetics, Case Western Reserve University School of Medicine - Non-syndromic recessive deafness (NSRD) is the most common form of prelingual hereditary hearing loss. To date, 10 autosomal NSRD loci (DFNBs) have been identified by genetic mapping; at least three times as many additional loci are expected to be identified. We have performed linkage analyses in two inter-related inbred kindreds, comprised of >50 affecteds, from a single Israeli-Arab village segregating NSRD. Genetic mapping by two-point and multi-point linkage analysis in 10 candidate regions identified the segregating gene to be on human chromosome 13q11 (DFNB1). Haplotype analysis, using eight microsatellite markers spanning 15 cM in 13q11, suggested the segregation of two different mutations in this kindred: affected individuals were homozygotes for either haplotype or compound heterozygotes. The gene for the connexin 26 gap junction protein, recently shown to be mutant in both dominant and recessive deafness, maps to this locus. We identified two distinct mutations, W77R and Gdel35, both of which likely inactivate connexin 26. The Gdel35 change likely occurs at a mutational hotspot within the connexin 26 gene. The recombination of marker alleles at the polymorphisms studied in 13q11, at known map distances from the mutations, allowed us to estimate the age of the mutations to be 35 generations (75125 years). This study independently confirms the identity of connexin 26 as an NSRD gene. Importantly, we demonstrate that in small populations with high rates of consanguinity, as compared with large outbred populations, recessive mutations may have very recent origin and show allelic diversity. The phenotype of deafness is easily recognized in humans, and since it does not compromise fertility or longevity in patients, pedigrees segregating the disorder can be identified with ease (1,2). The classification of deafness as conductive or perceptive (sensorineural/neural) requires, however, detailed audiologic examination. The clearest phenotype recognized is in individuals who have deafness at birth or in whom hearing loss occurs before 3 years of age, termed prelingual deafness. Although congenital, this type of hearing loss may be inherited or acquired by prenatal infection (rubella), middle-ear disease and maternal drug therapy during pregnancy (thalidomide) and the like (3). Several genetic surveys and family studies have established the incidence of prelingual deafness, not associated with recognized syndromes, as ~ 1/1000 births with >60% of cases being hereditary (1). Of familial cases, >70% are estimated to arise from the segregation of recessive mutations at a minimum of 36 loci (4). The high incidence of hereditary deafness has led to an arduous search for the genes involved in both syndromic and non-syndromic forms (see ref. 2 for review). The large number of recessive mutations underlying non-syndromic prelingual hearing loss (NSRDs) is not in doubt and is well-supported by observations in the offspring of consanguineous unions. However, the exact number of such genes remains unknown. To date, 10 autosomal NSRD loci (also called DFNBs) have been mapped to human chromosomes: DFNB1 (13q11) (5); DFNB2 (11q13.5) (6); DFNB3 (17p11.2-q12) (7); DFNB4 (7q31) (8); DFNB5 (14q12) (9); DFNB6 (3p14-p21) (10); DFNB7 (9q13-q21) (11,12); DFNB8/10 (21q22) (13,14); DFNB9 (2p22-p23) (15); and DFNB12 (10q21-q22) (16). However, gene identification has been hampered since most autosomal NSRD loci segregate in only a few pedigrees. Nevertheless, while this study has been in progress, mutations in connexin 26 (17) and myosin 7A (18), contributing to both recessive and dominant nonsyndromic deafness, have recently been demonstrated in multiple families. Since individual NSRD mutations are expected to be rare recessives, studies in multiple inbred populations are necessary to identify the majority of loci. Inbred populations with a founder effect for such mutations are ideal for eventual gene identification. During the course of genetic surveys in different villages in Israel, we identified one population in which the prevalence of profound, isolated, non-syndromic, congenital deafness, as well as less severe forms, is ~ 2%. This isolate is a Muslim Israeli-Arab village of ~ 8000 inhabitants located in the Two point lod scores were calculated between the segregating NSRD gene in a subset of nuclear families and genetic markers at 10 candidate regions corresponding to the ten known autosomal NSRD loci. In each case, two genetic markers within 12 cM of the NSRD locus were studied. Lod scores were calculated at recombination values of 0, 5 and 10% assuming recessive inheritance. From refs a(11), b(12), c(14) and d(13). lower Galilee and fully integrated into the life of the state of Israel. The extant members of this community are divided into large inter-related kindreds, called Hamulas, and can trace their ancestry to a few founders about eight generations ago. Two of these kindreds, which we studied and designated as Hamula A and H, each represent 3040% of all household units in the same village. Hearing loss is present in both kindreds, and there are many cases of intermarriage between the kindreds. We performed two-point linkage analyses on the 10 candidate NSRD regions listed earlier for a small subset of nuclear families and detected significant linkage to DFNB1 (3) on chromosome 13q11. Specifically, we observed a maximum two-point lod score of 4.9 at marker D13S175; multi-point analyses using eight markers in 13q11 showed a maximum lod score of 21.2 at D13S175. Interestingly, haplotype analysis using eight microsatellite markers spanning 15 cM in 13q11 suggested the segregation of two different mutations with affected individuals being homozygous for either mutation or compound heterozygotes. Subsequently, we investigated the gene for the gap junction protein connexin 26 (Cx26). Since it is a positional candidate gene which maps to the same region on 13q11 as DFNB1 (17), it is a biological candidate by virtue of the ability of hexameric assemblies of these proteins to compose electrical synapses which couple some neurons (19), and, it has recently been shown to be mutant in individuals whose deafness phenotype segregates with DFNB1 (17). By DNA sequencing and single stranded conformation polymorphism analyses, we identified two distinct mutations which co-segregate with deafness in our families confirming the role of Cx26 in hereditary deafness. The finding of two mutations in an inbred population, which we show have likely arisen in the last three to five (...truncated)