Introduction

Hearing impairment is the most frequently inherited human sensory defect. Approximately one in 1000 children is affected with prelingual deafness and at least 50% of these cases are hereditary.1,2 In 80% of these cases, the hearing impairment is described as nonsyndromic, because it is not associated with other disorders.3 Isolated deafness forms transmitted in the recessive mode are the most frequent (85%) and the most severe.1 Thirty-one different DFNB loci have been mapped and 13 of the corresponding genes have been identified (http://www.uia.ac.be/dnalab/hhh/). We have investigated a large consanguineous Tunisian family segregating an autosomal recessive nonsyndromic hearing loss and mapped the gene responsible for this defect to chromosome 1p13.3-22.1.

Materials and methods

Subjects

We analysed a large consanguineous Tunisian family in which nonsyndromic hearing impairment is transmitted as an autosomal recessive trait (Figure 1). Individuals were evaluated for hearing loss by clinical examination and audiological tests including pure tone audiometry and evoked response audiometry. Air-conduction pure-tone average (ACPTA) thresholds in the conversational frequencies (0.5, 1 and 2 kHz) were calculated for each ear, and were used to define the severity of the hearing loss according to the better hearing ear: mild (25 dB⩽ACPTA⩽39 dB), moderate (40 dB⩽ACPTA⩽69 dB), severe (70 dB⩽ACPTA⩽89 dB) and profound (ACPTA⩾90 dB). All affected individuals are otherwise healthy. Syndromic forms of hearing loss had been excluded. Appropriate informed consent was obtained from all subjects.

Figure 1
figure 1

Haplotype analysis of DFNB32 family with nonsyndromic hearing impairment. Blackened circles and squares indicate affected members. Haplotypes for 20 polymorphic markers from the 1p13.3–22.1 chromosomal region and obligate recombination events are indicated. The haplotype assumed to carry the disease allele is indicated by the black bar.

Full size image

Genotyping and linkage analysis

Genomic DNA was extracted using a standard phenol-chloroform technique from 10 ml blood samples. The genome-wide screening was performed using 400 microsatellite markers at intervals of 10 cM (ABI PRISM 1 Linkage Mapping Set 2, Applied Biosystems). Multiplex PCR were performed with four markers. PCR products were analysed following electrophoresis on a 6% polyacrylamide gel. In order to confirm linkage and for fine mapping the genetic interval, 19 additional polymorphic microsatellite markers were typed, comprising 18 published (http://cedar.genetics.soton.ac.uk/pub/chrom1/map.html) and one new marker (D1S21401), which was determined by searching for possible (CA) repeats in the sequence data of BAC clones between markers D1S1166 and afmb014zb9. The PCR primers for the new marker were D1S21401F (AGAATTTGGAGGTTTTTGGG)/ D1S21401R (TTGCATCCAGGACTTTTGTT). PCR products were analysed on 6% denaturing polyacrylamide gels, transferred onto N+-Hybond membrane (Amersham) and hybridized with a polyAC probe labelled with α32P dCTP.

Linkage analysis was performed using the linkage 5.1 program.4 Two-point lod scores were calculated with the Mlink program. Deafness was assumed to be inherited in a recessive manner with complete penetrance. Recombination frequencies were assumed to be equal for males and females. An estimated frequency of 10–4 was used for the disease gene. The allele frequencies of the polymorphic markers were assumed to be equal.

Mutation analysis

Each of the 68 coding exons of COL11A1 and their flanking splicing sites were analysed by conformation sensitive gel electrophoresis (CSGE) according to Annunen et al.5 and at least 30 exons were sequenced using an ABI 377 automated sequencer. Sequences were compared with those of the different transcript variants of human COL11A1.

Results and discussion

Twenty-two individuals (five affected and 17 unaffected) from a large consanguineous Tunisian family affected with congenital autosomal recessive deafness, were investigated (Figure 1). The audiometric tests showed, for all affected individuals, a loss of hearing greater than 70 dB. All affected individuals are otherwise healthy, with normal life spans and without dysmorphic or other abnormal findings.

In order to map the gene responsible for deafness, we performed linkage analyses with microsatellite markers corresponding to the previously identified loci (http://www.uia.ac.be./dnalab/hhh/). For each of the candidate location we genotyped three markers within each locus selected based on published linkage study. No linkage was found indicating the involvement of a novel locus in this family. Therefore, we undertook a screening of the whole genome using microsatellite markers at intervals of 10 cM. A linkage was detected only with locus D1S206 on chromosome 1, thereby defining a novel deafness locus, DFNB32. In order to confirm linkage and for fine mapping the genetic interval, 18 additional polymorphic microsatellite markers from the Généthon panel (http://cedar.genetics.soton.ac.uk/pub/chrom1/map.html), were typed (Figure 1). Significant two-point lodscores were obtained with five microsatellite markers with a maximum of 4.36 for D1S2896 at a recombination fraction θ=0.00 (Table 1). Haplotype analysis disclosed key recombination events between D1S1166 and afmb014zb9 in affected individual IV:13, defining the centromeric boundary of the disease interval (Figure 1). To refine the centromeric limit, we generated a new microsatellite marker D1S21401. This marker was named according to the name of the BAC clone. No recombination was detected and a maximum lodscore of 4.96 was obtained with this marker. The telomeric limit of the interval is determined by a normal hearing individual IV:6, with recombination events excluding markers telomeric to D1S2868 (Figure 1). Therefore, the DFNB32 locus lies within the 16 Mb region delimited by D1S2868 and afmb014zb9 markers (http://www.genome.ucsc.edu/cgi-bin/hgGateway?db=hg11).

Table 1 Markers with significant two-point lodscores and their respective map position according to (http://www.genome.ucsc.edu)
Full size table

The DFNB32 critical region overlap with DFNA37 locus (http://www.uia.ac.be./dnalab/hhh/). There are two possible explanations for this finding. Firstly, the two deafness disorders could be caused by mutations in the same gene which would map within the overlapping interval of the two loci. On the other hand, the possibility of two deafness-related genes lying close to each other on 1p13.3-22.1 cannot be ruled out. Instances of both phenomena have been reported in previous literature. Mutations of Cx26 gene can cause nonsyndromic recessive (DFNB1) and dominant (DFNA3) deafness.6,7 On the other hand, two distinct genes, Cx26 and Cx30, account for deafness segregating with both DFNB1 and DFNA3 locus.6,7,8,910 So, only the isolation of DFNB32 and DFNA37 genes is expected to elucidate the relationship between the two disorders.

Moreover, several DFNB genes have been shown to be responsible of isolated and syndromic forms of deafness (http://www.uia.ac.be./dnalab/hhh/). This has been the case for the MYO7A gene. Mutations of MYO7A can cause both nonsyndromic hearing impairment (DFNB2 and DFNA11) and syndromic deafness (USH1B).11,12,13 The DFNB32 region contains the COL11A1 gene responsible for 2 syndromic deafness, Marshall and Stickler syndromes.14,15 A search of the Human Cochlear cDNA Library and EST Database (http://hearing.bwd.harvard.edu/cochlearcdnalibrary.htm) revealed five ESTs expressed mapping within the 1p13.3–22.1 area (Table 2). The region syntenic to the DFNB32 interval is located on the murine chromosome 3. Among the numerous deaf mouse mutants which have been reported (http://www.ihr.mrc.ac.uk/hereditary/MutantsTable.shtml), four (Chondrodysplasia, Osteopetrosis colony stimulating factor 1 KO, Droopy ear, Varttint-waddler) have been mapped to the orthologous region of chromosome 1p13.3–22.1. Therefore, it is reasonable to propose that DFNB32 and these mutants may represent defects of the same gene. The gene responsible for the chondrodysplasia mutant is the Col11a1. Therefore, the COL11A1 gene was considered as an excellent candidate because: (i) it is responsible for two syndromic deafness, (ii) it is expressed in the cochlea, and (iii) mutations in the homologue mouse mutant produce hearing impairment. In order to investigate COL11A1 as a candidate gene, we performed CSGE and sequence analysis. Analysis of the COL11A1 coding region in an affected family member disclosed 21 sequence variants. These variants have previously been detected in several unaffected individuals (unpublished data), thus ruling out their involvement in the disease pathogenesis. In addition to the COL11A1, the genes corresponding to the ESTs found are being screened for deafness-causing mutations.

Table 2 ESTs mapped in the DFNB32 interval
Full size table

In conclusion, the hearing disorder in the pedigree investigated maps to a new locus, DFNB32. Localisation of the locus DFNB32 is a first step towards the identification of a new gene that controls development and/or maintenance of normal hearing.