Deafness, Autosomal Recessive 4, With Enlarged Vestibular Aqueduct
A number sign (#) is used with this entry because of evidence that autosomal recessive deafness-4 (DFNB4) with enlarged vestibular aqueduct (EVA) is caused by homozygous or compound heterozygous mutation in the SLC26A4 gene (605646) on chromosome 7q22.
Mutation in the FOXI1 gene (601093) has been found to be a rare cause of EVA. EVA may also be rarely caused by digenic inheritance of heterozygous mutations in the SLC26A4 and FOXI1 genes, or in the SLC26A4 and KCNJ10 (602208) genes.
Mutations in the SLC26A4 gene also cause Pendred syndrome (PDS; 274600), a disorder comprising congenital sensorineural hearing loss, cochlear abnormalities (EVA or Mondini dysplasia), and thyroid enlargement (goiter).
DescriptionDFNB4 with enlarged vestibular aqueduct is characterized by pre- or perilingual onset of sensorineural or mixed hearing loss, which may be fluctuating or progressive. The hearing loss is associated with temporal bone abnormalities, most commonly enlargement of the vestibular aqueduct, but it can also include the more severe Mondini dysplasia, a complex malformation in which the normal cochlear spiral of 2.5 turns is replaced by a hypoplastic coil of 1.5 turns (summary by Campbell et al., 2001 and Pryor et al., 2005). Enlarged vestibular aqueduct is the most common form of inner ear abnormality and can be associated with disequilibrium symptoms in a minority of patients (Valvassori, 1983; Jackler and de la Cruz, 1989; Levenson et al., 1989; Arcand et al., 1991; Belenky et al., 1993; Okumura et al., 1995).
Clinical FeaturesGriffith et al. (1996) reported a family in which 2 brothers had sensorineural hearing loss and enlarged vestibular aqueduct with no other abnormalities. Their parents were unaffected. The authors suggested autosomal recessive or X-linked inheritance with variable expressivity of the disorder in this family.
Abe et al. (1997) reported 3 families in which 2 sibs in each had congenital, high-frequency, fluctuating sensorineural hearing loss associated with enlargement of the vestibular aqueduct. Both parents in all 3 families were unaffected, suggesting autosomal recessive inheritance of the disorder.
Abe et al. (1999) studied 13 patients from 9 Japanese families and 2 patients from a Caucasian family who had congenital high frequency-dominant fluctuating sensorineural hearing loss and EVA on CT scan. Gadolidium-enhanced MRI confirmed the enlarged endolymphatic duct and sac. Hearing loss in some patient was progressive but with fluctuations, and about one-third had a history of vertigo. The perchlorate discharge test was performed in 8 patients from 6 of the families; all results were normal. Three of these families had been described previously by Abe et al. (1997).
Li et al. (1998) studied a large consanguineous family from southwest India in which 10 individuals ranging in age from 5 to 38 years were affected with congenital, profound, nonsyndromic autosomal recessive deafness. No goiter was palpable in any of the affected individuals and, although the perchlorate discharge test was not available, several other tests of thyroid function were normal. Axial and coronal computerized tomography of the temporal bone showed bilateral large vestibular aqueducts in all 3 affected individuals who were studied, with no Mondini-type cochlear malformation.
MappingBaldwin et al. (1995) described a large Middle-Eastern Druze family with recessive nonsyndromic deafness and demonstrated linkage between deafness in this family and 7q31 with a lod score exceeding 5.5. Baldwin et al. (1995) designated the locus DFNB4. In addition, they found that deafness in 3 other Druze pedigrees, including 1 related to the linked family, was not linked to 7q31. Thus, there appear to be multiple nonallelic mutations for deafness in this genetic isolate. On the basis of a personal communication from Baldwin (1998), Li et al. (1998) purported that the Israeli-Druze family indeed had Pendred syndrome. Affected members of this family were later found to have goiters.
Everett et al. (1997) identified SLC26A4 (605646) as the gene mutant in Pendred syndrome (PDS; 274600) in 3 families. The gene maps to 7q31. They pointed out that DFNB4 also maps to 7q31 and considered it likely that the DFNB4 individuals reported actually have PDS, rather than mutations in another gene.
By linkage analysis in 9 Japanese families and 1 Caucasian family with sensorineural hearing loss associated with EVA, Abe et al. (1999) localized the gene responsible to 7q31, with a maximum multipoint lod score of 3.647. The EVA candidate gene region was found to lie in a 1.7-cM interval between flanking markers D7S501 and D7S2425. Although this region overlaps the region containing the gene responsible for Pendred syndrome, these patients did not fulfill the criteria for PDS.
Molecular GeneticsMutations in the SLC26A4 Gene
In affected members of a large consanguineous family from southwest India with DFNB4 with EVA, Li et al. (1998) found linkage to chromosome 7q31 and demonstrated that affected individuals were compound homozygotes for 2 mutations in exon 13 of the PDS gene (605646.0004).
Usami et al. (1999) screened the SLC26A4 gene for mutations in 6 families with congenital nonsyndromic high frequency, fluctuating, sometimes progressive sensorineural hearing loss, and enlarged vestibular aqueduct diagnosed by CT. One patient had a history of vertigo; none had Mondini malformation. Affected individuals in 4 of the 6 families were homozygous or compound heterozygous for SLC26A4 mutations (605646.0009-605646.0015).
Campbell et al. (2001) found mutations in the SLC26A4 gene in 5 of 6 multiplex families with EVA (83%) and in 4 of 5 multiplex families with Mondini dysplasia (80%), implying that mutations in the SLC26A4 gene are the major genetic cause of these temporal abnormalities. In their analyses of Pendred syndrome and DFNB4, they found that the 2 most common mutations, T416P (605646.0006) and IVS8+1G-A (605646.0007), were present in 22% and 30% of families, respectively.
Recessive mutations in the anion transporter gene SLC26A4 are known to be responsible for Pendred syndrome and for nonsyndromic hearing loss associated with EVA. However, a large percentage of patients with these phenotypes lack mutations in the SLC26A4 coding region in one or both alleles. Yang et al. (2007) identified and characterized a key transcriptional regulatory element in the SLC26A4 promoter that binds FOXI1 (601093), which is a transcriptional activator of SLC26A4. They found 9 patients with Pendred syndrome or nonsyndromic EVA who were heterozygous for a novel -103T-C mutation (605646.0027) in this regulatory element of the SLC26A4 gene that interfered with FOXI1 binding and completely abolished FOXI1-mediated transcriptional activation.
Mutation in the FOXI1 Gene
In 2 families given a diagnosis of enlarged vestibular aqueduct, Yang et al. (2007) found heterozygosity for a mutation in the FOXI1 gene (601093.0002). Although both of these families were classified by the authors as 'nonsyndromic EVA,' in one of them goiter reminiscent of Pendred syndrome was noted. Both alleles of the SLC26A4 gene were wildtype. The FOXI1 mutation showed significantly decreased luciferase activation in promoter-reporter assays, suggesting that this variant compromised the ability of FOXI1 to transactivate SLC26A4 and was causally related to disease.
Digenic Inheritance
Yang et al. (2007) reported a patient with DFNB4 and EVA who was compound heterozygous for a mutation in 2 different genes. The patient had a heterozygous mutation in the SLC26A4 gene (605646.0028) and a heterozygous mutation in the FOXI1 gene (601093.0001). This finding was consistent with their observation that EVA occurs in the mouse mutant doubly heterozygous for mutations in these 2 genes, and the results supported a dosage-dependent model for the molecular pathogenesis of nonsyndromic EVA that involves SLC26A4 and its transcriptional regulatory machinery. Yang et al. (2007) stated the this was the first example of digenic inheritance to be verified as a cause of human deafness.
Yang et al. (2009) sequenced the KCNJ10 gene (602208) in 89 patients who had a clinical diagnosis of EVA/Pendred syndrome and were known to carry only 1 SLC26A4 coding sequence mutation; promoter mutations and deletions of SLC26A4 were excluded in this patient cohort. In 2 patients, Yang et al. (2009) identified missense mutations in KCNJ10 (P194H, 602208.0008 and R348C, 602208.0009, respectively). The former patient carried a F335L mutation in SLC26A4 (605646.0031), and the latter a splice site mutation (605646.0029). Both KCNJ10 mutations reduce potassium conductance activity, which is critical for generating and maintaining the endocochlear potential.
Genotype/Phenotype CorrelationsScott et al. (2000) compared 3 common Pendred syndrome allele variants with 3 PDS mutations reported only in individuals with nonsyndromic hearing loss. The mutations associated with Pendred syndrome exhibited complete loss of pendrin (SLC26A4)-induced chloride and iodide transport, while alleles unique to patients with DFNB4 were able to transport both iodide and chloride, albeit at a much lower level than wildtype pendrin. The authors hypothesized that the residual level of anion transport was sufficient to eliminate or postpone the onset of goiter in individuals with DFNB4. They proposed a model for pendrin function in the thyroid in which pendrin transports iodide across the apical membrane of the thyrocyte into the colloid space.
Tsukamoto et al. (2003) screened 10 Japanese families with Pendred syndrome, 32 Japanese families with bilateral sensorineural hearing loss associated with EVA, and 96 unrelated Japanese controls for mutations in the SLC26A4 gene. They identified causative mutations in 90% of the typical Pendred syndrome families and in 78.1% of those with sensorineural hearing loss with EVA. None of their patients had the Mondini malformation. Tsukamoto et al. (2003) noted that the same combination of mutations resulted in variable phenotypic expression (see, e.g., 605646.0011 and 605646.0012), suggesting that these 2 conditions are part of a continuous spectrum of disease.
Pryor et al. (2005) evaluated the clinical phenotype and SLC26A4 genotype of 39 patients with EVA from 31 families, definitively classifying 29 individuals. All 11 PDS patients had 2 mutant SLC26A4 alleles, whereas all 18 nonsyndromic EVA patients had either 1 or no SLC26A4 mutant alleles. Pryor et al. (2005) concluded that PDS and nonsyndromic EVA are distinct clinical and genetic entities, with PDS being a genetically homogeneous disorder caused by biallelic SLC26A4 mutations, and at least some cases of nonsyndromic EVA being associated with a single SLC26A4 mutation. They noted that the detection of a single mutant SLC26A4 allele is incompletely diagnostic without additional clinical evaluation to differentiate PDS from nonsyndromic EVA.
Albert et al. (2006) analyzed the SLC26A4 gene in 109 patients from 100 unrelated French Caucasian families with nonsyndromic deafness and enlarged vestibular aqueduct and no mutation in the GJB2 gene (121011). They identified 91 allelic variants in 40 unrelated families (prevalence of SLC26A4 mutations, 40%). There were 18 compound heterozygous and 6 homozygous families; Albert et al. (2006) noted that patients with biallelic mutations had more severe deafness, an earlier age of diagnosis, and a more fluctuating course than patients in whom no mutation was identified. Albert et al. (2006) estimated that up to 4% of nonsyndromic hearing impairment could be caused by SLC26A4 mutations.
In 71 families with EVA, Choi et al. (2009) used sequence analysis of SLC26A4 coding and conserved noncoding regions and CGH microarray analysis, and compared segregation of EVA among families with 2, 1, or no detectable mutant alleles of SLC26A4. EVA segregation ratios were similar in families with 1 or 2 mutant alleles, but the segregation ratio for families with 1 mutation was significantly higher than that of families with no SLC26A4 mutations. Haplotype analyses revealed discordant segregation of EVA with SLC26A4-linked STR markers in 8 of 24 families with no mutation in SLC26A4. Choi et al. (2009) concluded that families with EVA and 1 detectable mutation in SLC26A4 were likely to be segregating EVA as a trait caused by that mutation in combination with a second occult mutant allele of SLC26A4 or of another autosomal gene. In contrast, EVA appeared to be a nongenetic or complex trait with a significantly lower recurrence rate in families with no detectable SLC26A4 mutation.
Chattaraj et al. (2017) performed genotype-haplotype analysis and massively parallel sequencing of the SLC26A4 gene in patients with EVA and only 1 detected mutant allele in the SLC26A4 gene. The authors identified a shared novel haplotype, termed CEVA (Caucasian EVA), composed of 12 uncommon variants upstream of SLC26A4. The presence of the CEVA haplotype on 7 of 10 mutation-negative chromosomes in a National Institutes of Health discovery cohort and 6 of 6 mutation-negative chromosomes in a Danish replication cohort was higher than the observed prevalence of 28 of 1,006 Caucasian control chromosomes (p less than 0.0001 for each EVA cohort). The corresponding heterozygous carrier rate was 28 of 503 (5.6%). The prevalence of CEVA (11 of 126) was also increased among EVA chromosomes with no mutations detected (p = 0.0042). Chattaraj et al. (2017) concluded that the CEVA haplotype causally contributes to most cases of Caucasian EVA, being present in cases where only 1 mutation is detected by traditional exonic sequencing, and possibly in some cases where no mutation has been detected.
Population GeneticsWang et al. (2007) identified a total of 40 SLC26A4 mutations, including 25 novel mutations, among 107 Chinese patients with EVA from 101 families. Overall, SLC26A4 mutations were identified in 97.9% of patients. The most common mutation was a splice site transition (IVS7-2A-G; 605646.0029), which accounted for 57.6% of mutant alleles. Park et al. (2005) identified the same splice site mutation in 9 (20%) of 45 mutant alleles in a study of Korean EVA patients. In 15 patients from 13 unrelated Chinese families with deafness and EVA, Hu et al. (2007) identified the IVS7-2A-G mutation in 5 (22.3%) of 22 mutant alleles. Reviewing previously published studies involving Chinese patients, the authors stated that IVS7-2A-G accounted for 69.1% (76 of 110) of all mutant alleles in the Chinese, suggesting a founder effect.
Pourova et al. (2010) screened the SLC26A4 gene in 303 Czech patients with early-onset hearing loss. The patients were divided into 3 groups: 22 with EVA and/or Mondini malformation on imaging, 220 patients without imaging available, and 61 patients with EVA/Mondini-negative imaging studies. Biallelic SLC26A4 mutations were found in 6 (27.3%) patients in the first group, 2 (0.9%) patients in the second group, and none (0%) in the third group; 4 of the 8 patients with biallelic mutations had goiter, consistent with Pendred syndrome. Monoallelic SLC26A4 mutations were found in 3 (13.6%) patients in the first group, 12 (5.5%) patients in the second group, and 3 (4.9%) patients in the third group. The most frequent mutations were V138F (605646.0024) and L445W (605646.0018), in 18% and 8.9% alleles, respectively. Among 13 patients with bilateral EVA, 6 (46%) carried biallelic mutations. No biallelic mutations were found in EVA-negative patients, but 4.9% had monoallelic mutations. Overall, biallelic mutations were found in only 2.7% of all patients, but were more common in familial cases. The findings also suggested that a single SLC25A4 mutation may contribute to the phenotype, perhaps in concert with mutations in other genes.
NomenclatureIn the title of their paper, Baldwin et al. (1995) referred to the form of deafness that maps to 7q31 as DFNB4. The same symbol was used by Fukushima et al. (1995) for a locus on chromosome 14 (600792). The chromosome 14 locus is, in fact, symbolized DFNB5.