Missense mutations in the COCH gene, which is expressed preferentially at high levels in the inner ear, cause the autosomal dominant sensorineural deafness and vestibular disorder, DFNA9 (OMIM 601369). By in situ hybridization of mouse and human inner ear sections, we find high-level expression of COCH mRNA in the fibrocytes of the spiral limbus and of the spiral ligament in the cochlea, and in the fibrocytes of the connective tissue stroma underlying the sensory epithelium of the crista ampullaris of the semicircular canals. A polyclonal antibody against the human COCH protein product, cochlin, was raised against the N-terminal 135 amino acid residues of cochlin, corresponding to the Limulus factor C-homology (cochFCH) domain; this domain harbors all five known point mutations in DFNA9. On western blots of human fetal cochlear extracts, anti-cochlin reacts with a cochlin band of the predicted full-length size as well as a smaller isoform. Immunohistochemistry performed with anti-cochlin shows staining predominantly in the regions of the fibrocytes of the spiral limbus and of the spiral ligament in mouse and in human fetal and adult tissue sections. These sites correspond to those areas that express COCH mRNA as determined by in situ hybridization, and to the regions of the inner ear which show histological abnormalities in DFNA9. The fibrocytes expressing mRNA and protein products of COCH are the very cell types which are either absent or markedly reduced and replaced by eosinophilic acellular material in temporal bone sections of individuals affected with DFNA9.

We have cloned a novel human gene, designated PFET1 (predominantly fetal expressed T1 domain) (HUGO-approved symbol KCTD12 or C13orf2), by subtractive hybridization and differential screening of human fetal cochlear cDNA clones. Also, we have identified the mouse homolog, designated Pfet1. PFET1/Pfet1 encode a single transcript of approximately 6 kb in human, and three transcripts of approximately 4, 4.5, and 6 kb in mouse with a 70% GCrich open reading frame (ORF) consisting of 978 bp in human and 984 bp in mouse. Both genes have unusually long 3¢ untranslated (3¢ UTR) regions (4996 bp in human PFET1, 3700 bp in mouse Pfet1) containing 12 and 5 putative polyadenylation consensus sequences, respectively. Pfetin, the protein encoded by PFET1/Pfet1, is predicted to have 325 amino acids in human and 327 amino acids in mouse and to contain a voltage-gated potassium (K + ) channel tetramerization (T1) domain. Otherwise, to date these genes have no significant homology to any known gene. PFET1 maps to the long arm of human chromosome 13, in band q21 as shown by FISH analysis and STS mapping. Pfet1 maps to mouse chromosome 14 near the markers D14Mit8, D14Mit93, and D14Mit145.1. The human 6 kb transcript is present in a variety of fetal organs, with highest expression levels in the cochlea and brain and, in stark contrast, is detected only at extremely low levels in adult organs, such as brain and lung. Immunohistochemistry with a polyclonal antibody raised against a synthetic peptide to PFET1 sequence (pfetin) reveals immunostaining in a variety of cell types in human, monkey, mouse, and guinea pig cochleas and the vestibular system, including type I vestibular hair cells.

As auditory genes and deafness-associated mutations are discovered at a rapid rate, exciting opportunities have arisen to uncover the molecular mechanisms underlying hearing and hearing impairment. Single genes have been identified to be pathogenic for dominant or recessive forms of nonsyndromic hearing loss, syndromic hearing loss, and, in some cases, even multiple forms of hearing loss. Modifier loci and genes have been found, and investigations into their role in the hearing process will yield valuable insight into the fundamental processes of the auditory system.