rss
J Med Genet 38:e13 doi:10.1136/jmg.38.5.e13
  • Electronic letter

A cysteine substitution in the zona pellucida domain of α-tectorin results in autosomal dominant, postlingual, progressive, mid frequency hearing loss in a Spanish family

So far, 37 different loci for autosomal dominant, sensorineural, non-syndromic hearing loss (ADSNSHL), have been mapped and 11 genes have been cloned.1 Among them, theTECTA gene, mapped to chromosome 11q22-24 (locus DFNA8/A12), encodes α-tectorin,2 a non-collagenous component of the cochlear tectorial membrane. This membrane is an extracellular matrix that covers the apical surface of the sensory epithelium in the cochlea and plays an important role in transmitting the mechanical energy of sound to the mechanosensitive stereociliary bundles of the hair cells, where the sound is transduced into neural potentials. In previous studies, five different missense mutations resulting in ADSNSHL have been described in the TECTA gene in fourDFNA8/A12 families, affecting different domains of the protein and showing different phenotypes (table 1): a prelingual, non-progressive hearing loss affecting mid frequencies or a postlingual, progressive, high frequency hearing impairment.

Table 1

TECTA mutations described in the DFNA8/A12 families

In this study, we report a Spanish family with a novel phenotype of postlingual, progressive, mid frequency hearing loss resulting from a new mutation in the TECTA gene.

Materials and methods

The family consists of 47 members including nine affected subjects with ADSNSHL (fig 1A). Appropriate informed consent was obtained from all those studied. Clinical examination was performed and blood samples were obtained from 33 family members. Environmental factors were eliminated as the cause of deafness in all affected family members. Features suggestive of syndromic anomalies were not present. Otoscopic examination and use of the tuning fork test ruled out conductive hearing loss. Pure tone audiometry was performed to test for air conduction (frequencies of 125-8000 Hz) and bone conduction (frequencies of 250-8000 Hz). Affected subjects showed bilateral sensorineural hearing impairment. In the beginning, the hearing loss in this family is mild, mainly affecting mid frequencies (500, 1000, and 2000 Hz) (fig 1B) and later it progresses to moderate-severe hearing loss involving all the frequencies. Linear regression analysis, based on all available audiograms from the nine patients, showed a 0.4 dB/year age linked progression of the 0.5-2 KHz average hearing loss. The onset of the hearing loss ranged from childhood (<9 years) in affected members III.2, III.3, IV.1, IV.7, and V.2 to the second decade in the remaining affected subjects. No evidence of vestibular dysfunction was observed, except in patient IV.10 who reported occasional dizziness. In addition, there were symptoms of occasional tinnitus in four of nine affected members (III.5, IV.5, IV.7, and IV.10).

Figure 1

(A) Pedigree and haplotype analysis of the Spanish family S063. Only representative members have been included. The order of the markers was set integrating genetic and physical data from previous studies.3 5 Black symbols represent affected subjects. Haplotypes are represented by bars, with the haplotype associated with hearing loss in black. (B) Audiograms from four different affected members of the family showing decreased hearing particularly at mid frequencies (500, 1000, and 2000 Hz). Only results for the right ear are presented.

The subjects studied were genotyped for microsatellite markers close to all the described DFNA loci. Linkage analysis was performed using the LINKAGE 5.1 software package,6 setting the frequency of the deafness gene to 0.0001 and considering marker allele frequencies to be equal to each other.

Results

In all cases negative results for linkage were obtained except for six microsatellite markers close to theDFNA8/A12 locus. A maximum lod score of 3.67 was shown for marker D11S4089 at theta=0.0. Extensive alterations of the disease gene frequency, or of the allele frequencies of microsatellite markers, did not significantly change the lod scores. Detection of mutations in the TECTA gene was carried out by heteroduplex analysis followed by DNA sequencing of exons 10, 14, 17, 18, and the 5′ end of intron 9 where all the previously reported mutations were located.2-4 7 A novel mutation was detected in exon 17 of theTECTA gene. At nucleotide position 5509, a T to G transversion was found, which would produce a C1837G amino acid substitution (fig 2). This change results in the loss of the unique restriction site for enzyme Alw44I at exon 17. On this basis, we developed an easy screening test for the C1837G mutation. A 306 bp DNA fragment including exon 17 from theTECTA gene was amplified using specific primers designed at flanking intronic positions (forward primer: 5′-GAT TTG CCT TTC GTA ATA ACT GT-3′, reverse primer: 5′-AGG ACA ATA AAT GTG CAA ACA CT-3′). After cleavage with the restriction enzymeAlw44I, two bands of 150 bp and 156 bp were seen in controls. In the affected members, an additional band corresponding to the 306 bp undigested product was also present. This missense mutation appeared in heterozygosity and was shown to segregate completely with the affected status in this family. In addition, it was not present in 100 unrelated Spanish controls.

Figure 2

DNA sequences showing the TECTA missense mutation in the Spanish family S063. Electropherograms for the regions immediately surrounding the 5509T>G mutation at exon 17 are shown. An affected subject and a control are depicted. Codon 1837 is boxed.

Discussion

The α-tectorin precursor is proteolytically processed into three polypeptides: a module containing a region with similarity to the G1 domain of entactin,8 a central module containing von Willebrand factor (vWf) type D repeats similar to zonadhesin (ZA),9 and a module consisting of a zona pellucida domain (ZP).10 It is assumed that these three polypeptides are cross linked to each other by disulphide bridges and interact with β-tectorin11 to form the non-collagenous matrix of the tectorial membrane. The mutations in the Belgian and Austrian families change conserved amino acids from the ZP domain, whereas in the Swedish and French families cysteine residues from the ZA-like domain are substituted (table 1). The mutation in the French family changes one of the vicinal cysteine residues of the vWf-type D4 repeat that play an important role as a catalytic site for disulphide bonded multimer assembly of vWf.12 In the Spanish family, the novel mutation is located in the ZP domain (fig 3A) and results in the replacement of a cysteine by a glycine. The cysteine residue at this position is fully conserved in all known ZP domains from other proteins (fig 3B), so the C1837G substitution might disrupt the proper interaction between the different tectorin polypeptides altering the mechanotransductional properties of the tectorial membrane leading to an inefficient transmission of sound. This hypothesis has also been proposed for the previously reported mutations in theDFNA8/A12families.2-4

Figure 3

(A) Domain structure of the human α-tectorin protein. D0-D4 units represent the von Willebrand factor type D repeats (Zonadhesin-like). Asterisks indicate the position of the different missense mutations. The new C1837G mutation is boxed. (B) Multiple amino acid alignment of proteins homologous to the α-tectorin zona pellucida domain. Only a short region encoded by exon 17 and containing the C1837G mutation is shown. The alignment includes α-tectorin from human (Hs α-tect2) and mouse (Mm α-tect13), uromodulin from human (Hs uro14), mouse (Mm uro15), rat, and bovine (Rn uro and Bt uro16), glycoprotein 2 from dog (Cf GP217), human (Hs GP218), and rat (Rn GP219 20), and frog thyroid regulated glycoprotein 18 (Xl 1821).  

It has been postulated that the phenotypic differences among the previously reported DFNA8/A12 families (table 1) could be the result of the altered domain in α-tectorin in each family. Thus, mutations in the ZP domain appeared to lead to prelingual stable hearing loss mainly involving mid frequencies, while mutations in the ZA-like domain resulted in progressive hearing loss starting in the high frequencies.2-4 The Spanish family reported here shows a novel phenotype not observed in the previously described DFNA8/A12 families. The alteration of the ZP domain also affects mainly the mid frequencies in the beginning but in contrast it results in progressive hearing loss. This suggests that the alteration of one determined domain from α-tectorin determines the range of affected frequencies (ZA-like, high frequencies; ZP, mid frequencies) independently of the type of mutation. On the other hand, the progression or stability of the hearing loss seems to be related to the type of substituted residue in each domain. In particular, the substitution of the same type of residue (cysteine) in two different domains (ZA-like and ZP) appears to lead to a progressive and generally postlingual hearing loss that may be caused by a progressive deterioration of the tectorial membrane owing to the improper cross linking of the α-tectorin polypeptides. Identification of new mutations in additional families is needed to determine the validity of this putative genotype-phenotype correlation.

Acknowledgments

We are grateful to the Spanish family who made this research possible and Dr C Somoza-Castillo for patient referral. This work was supported by grants from the Fondo de Investigaciones Sanitarias FIS-96/1556, the Comisión de Investigación de Ciencia y Tecnología CICYT-SAF 99-0025, and the European Community QLRT 1999-00988.

References