Hereditary hearing impairment shows extreme genetic heterogeneity and more than 40 different loci have been reported. One of these, DFNA2, was localized to chromosome 1p34 on the basis of linkage analysis in two large families with autosomal dominant nonsyndromic hearing loss originating from Indonesia and the United States1. Subsequently, hearing loss in three additional large families from Belgium and the Netherlands was found to map to the same region2.

Two different genes are now reported to be the 'DFNA2' gene. In December's issue of Nature Genetics, Jia-Hui Xia and colleagues described two small Chinese families with nonsyndromic autosomal dominant hearing loss in which they found a missense and a nonsense mutation in GJB3, which encodes connexin 31 (ref. 3). As GJB3 is located on chromosome 1p34, it was proposed, in an accompanying News & Views that GJB3 is a good candidate for involvement in DFNA2 families4. Extensive sequence analysis of the coding region and the 5´ UTR of GJB3 in all 5 DFNA2 families revealed no mutations, however (data not shown). A recent report in Cell reports the identification of a novel member (KCNQ4) of the potassium channel family KCNQ, also located on chromosome 1p34. Its authors found its mutation to segregate with hearing loss in a small French family, and claimed the identification of the DFNA2 gene.

Which one of these claims is correct? An obvious point to consider is that DFNA2 (Fig. 1) is the name of a chromosomal locus identified by linkage analysis, rather than a gene. Our analysis indicates that the candidate region delineated by the Indonesian family does not overlap with the candidate region of one of the Dutch families. The candidate regions of the remaining families overlap with both candidate regions (Fig. 1). KCNQ4 is located in the candidate region of four families ( Fig. 1); it is therefore possible that KCNQ4 is responsible for hearing loss in these 'DFNA2' families, but conclusive data awaits mutation analysis. The centromeric recombinant of the Indonesian family excludes this gene as a possible candidate gene for hearing loss in this family, suggesting the presence of a third deafness gene in this region.

Figure 1: Schematic representation of the candidate regions indicated by the DFNA2 families, relative to the genetic map of chromosome 1 (not drawn to scale).
figure 1

Black boxes indicate the candidate regions; the positions of GJB3 and KCNQ4 are indicated by arrow.

As both GJB3 and KCNQ4 are responsible for nonsyndromic autosomal dominant hearing loss and are present in the DFNA2 region, each of them can be considered to be a 'DFNA2' gene (a hearing loss gene from the DFNA2 region), rather than the 'DFNA2' gene. As nonsyndromic deafness is a paradigm of genetic heterogeneity6, the presence of different genes responsible for nonsyndromic deafness in a small chromosomal region is not all that surprising. This situation may not be limited to hearing loss, as two different genes have been suggested at one locus for retinitis pigmentosa7,8, another prime example of genetic heterogeneity. The question posed in our title is therefore legitimate, and DFNA2 may not be the only locus to which it applies.