Age-related hearing loss (AHL), or presbycusis, is the most common type of human hearing impairment, affecting about half the population by age 80. Little is known about the genetic causes of human presbycusis because of its late onset and confounding factors such as noise trauma, disease or ototoxic drugs. Our approach for investigating the genetic basis of human presbycusis is to use inbred mouse strains that exhibit progressive hearing loss as models (Noben-Trauth and Johnson, 2009). We have mapped multiple AHL loci in mice by analyses of linkage crosses, recombinant inbred strains and chromosome substitution strains.
ahl: We mapped the first mouse AHL locus (ahl) to mid Chromosome (Chr) 10 in C57BL/6J mice (Johnson et al., 1997) and later showed that the locus affects hearing loss in many other inbred mouse strains (Johnson et al., 2000). In collaboration with Dr. Konrad Noben-Trauth, we showed that ahl corresponds with an inbred strain variant of the cadherin 23 (Cdh23) gene (Noben-Trauth et al., 2003), which encodes a tip link protein of the hair cell mechanotransduction apparatus. The ahl allele is common to many inbred strains and confers a predisposition to hearing loss with the degree of hearing loss dependent on strain background (Kane et al., 2012). To formally test the hypothesis that a specific Cdh23 splice site variant underlies ahl-related hearing loss, we have produced reciprocal knockin (KI) mouse strains, one in which the AHL resistance allele of the 129S1 strain (Cdh23753G) is replaced with the AHL susceptibility allele of the C57BL/6J strain (Cdh23753A), and the other in which the Cdh23753A allele of the C57BL/6J strain is replaced with the Cdh23753G allele of the 129S1 strain. We are now aging mice from each of these KI strains for periodic ABR analysis to compare their hearing loss phenotypes with those of mice from control strains.
ahl2: We mapped a second AHL locus (ahl2) to Chr 5, which contributes to hearing loss in NOD/LtJ mice (Johnson and Zheng, 2002), but have been unable to identify a probable candidate gene for this locus.
mt-Tr and ahl4: The mitochondrial DNA of the A/J strain harbors a deleterious variant of the gene encoding arginine tRNA (mt-Tr). We showed that the A/J variant of mt-Tr worsens hearing loss in mice that also have the predisposing ahl allele of Cdh23 (Johnson et al., 2001). We later identified a nuclear gene locus on distal Chr 10 (ahl4) that also contributes to the hearing loss of A/J strain mice (Zheng et al., 2009) and showed that a missense mutation (H55N) of the citrate synthase gene (Cs) is the likely underlying cause (Johnson et al., 2012b). Citrate synthase is located in the mitochondrial matrix and is the rate-limiting enzyme of the citric acid cycle. A nucleotide variant in exon 3 of Cs is the only known DNA difference within the 5.5 Mb ahl4 candidate gene interval that is unique to the A/J strain and that causes a nonsynonymous codon change. The Cs mutation likely contributes to the hearing loss of A/J mice by reducing ATP production and increasing free radical production in mitochondria of cochlear hair cells. The A/J mouse with mutations in Cs and mt-Tr thus provides a new model system for in vivo studies of mitochondrial dysfunction and hearing loss.
ahl8: We mapped a locus on distal Chr 11 (ahl8) that contributes to hearing loss in DBA/2J mice (Johnson et al., 2008). We later identified a missense mutation of the fascin-2 gene (Fscn2) as the underlying cause of the ahl8 phenotype (Shin et al., 2010). Fscn2 encodes an actin bundling protein that is expressed in the inner ear hair cell bundle. SNP analysis identified a unique nucleotide change in the protein coding sequence of Fscn2 in DBA/2J mice that would change a highly conserved arginine residue to histidine. The closely related substrains DBA/1J, DBA/2HaSmnJ, DBA/2DaJ, and DBA/2NCr do not have this Fscn2 variant and mice of these strains exhibit a later onset and slower progression of hearing loss than do DBA/2J mice. We produced a D2.B6-Fscn2 congenic strain and developed transgenic mice in which a C57BL/6J-derived BAC containing the wildtype Fscn2 gene is integrated into the genome of DBA/2J mice. We found that the C57BL/6J-derived Fscn2 congenic region and the C57BL/6J-derived Fscn2 transgene improved the hearing of mice with an otherwise DBA/2J genome. These rescue experiments confirmed the causative nature of the R109H Fscn2 variant in conferring an increased susceptibility to age-related hearing loss in DBA/2J mice. Later collaborative experiments using double-mutant mice showed that fascin-2 cooperates with beta-actin to maintain stereocilia length and auditory function (Perrin et al., 2013).
Mouse deafness mutations provide valuable models of human hereditary hearing disorders and entry points into molecular pathways important to the hearing process. We assess hearing in mice with naturally occurring mutations and in mice from mutagenesis programs. If the mutant mice have an inherited hearing impairment, genetic crosses are set up to map the mutation. Each new mutation is characterized for inner ear pathologies and, when the underlying gene is identified, its temporal and spatial expression patterns are assessed along with those of other genes that may act in the same pathways of inner ear development or maintenance. Below we describe in more detail some of the hearing-related mutations we have discovered.
Slc25a13hspn: Mice homozygous for the spontaneous Slc25a13 deletion mutation (hspn) are deaf with balance defects, whereas targeted Slc25a13 knockout mice display no phenotypic abnormalities. We show that inner ear malformations of Slc25a13hspn/hspn mice are similar to those of mice with knockout mutations of Dlx5, a developmentally important transcription factor gene located 660 kb from Slc25a13. We also show that Dlx5 expression is severely reduced in the otocyst but not the branchial arches of Slc25a13hspn/hspn embryos, indicating that the Slc25a13hspn deletion eliminates otic-specific enhancers of Dlx5. In addition, transheterozygous Slc25a13+/hspn Dlx5+/– mice exhibit non-complementation with inner ear dysmorphologies similar to those of Slc25a13hspn/hspn and Dlx5–/–embryos, verifying a cis-acting effect of the Slc25a13hspn deletion on Dlx5 expression. Slc25a13hspn/hspn mice thus provide a new animal model for studying long-range enhancer effects on Dlx5 expression in the developing inner ear.
Atp6v1b1vtx: Mutations of the human ATP6V1B1 gene cause distal renal tubular acidosis (dRTA) associated with sensorineural hearing impairment; however, mice with a knockout mutation of Atp6v1b1 were reported to exhibit a compensated acidosis and normal hearing. We discovered a new spontaneous mutation (vtx) of Atp6v1b1 in an MRL/MpJ (MRL) colony of mice. In contrast to the reported phenotype of the knockout mouse, which was developed on a primarily C57BL/6 (B6) strain background, MRL-Atp6v1b1vtx/vtx mutant mice exhibit profound hearing impairment, which is associated with enlarged endolymphatic compartments of the inner ear. The abnormal inner ear phenotype of MRL- Atp6v1b1vtx/vtx mice was lost when the mutation was transferred onto the B6 background, indicating the influence of strain-specific genetic modifiers. Because MRL-Atp6v1b1vxt/vtx mice do not recapitulate the metabolic acidosis of dRTA patients, they provide a new genetic model for nonsyndromic deafness with enlarged vestibular aqueduct (EVA).
Ush1cdfcr: We mapped two new recessive mutations causing circling behavior and deafness to the same region on Chromosome 7 and showed they are allelic by complementation analysis. One was named “deaf circler” (allele symbol dfcr) and the other “deaf circler 2 Jackson” (allele symbol dfcr‑2J). Both were shown to be mutations of the Ush1c gene (Johnson et al., 2003), the mouse ortholog of the gene responsible for human Usher syndrome type IC and for the nonsyndromic deafness disorder DFNB18. The qualitatively different mutations provided a means to evaluate the functional domains of the USH1C protein in cochlear mechanotransduction.
Tmhshscy: We discovered that a mutation in a previously uncharacterized gene was responsible for the deafness and circling phenotype of a new mouse mutation we called hurry-scurry (hscy). We named the gene “tetraspan membrane protein of hair cell stereocilia” (Tmhs) because of its predicted protein structure and inner ear localization (Longo-Guess et al., 2005). Subsequently, mutations in the homologous human gene (designated LHFPL5) were shown by others to be responsible for DFNB67, an autosomal recessive nonsyndromic deafness locus. We later generated a mouse with a targeted null mutation and lacZ reporter gene that demonstrated the same phenotype as the hscy mutation and showed high Tmhs directed expression in the cochlear and vestibular hair cells of the inner ear, supporting involvement of Tmhs in stereocilia development and function (Longo-Guess et al., 2007). The TMHS protein subsequently was shown to be an integral component of the mechanotransduction machinery of cochlear hair cells (Xiong et al., 2012).
Clic5jbg: We showed that the spontaneous jitterbug (jbg) mutation, which causes deafness and vestibular dysfunction when homozygous, is a 97-base pair intragenic deletion of the chloride intracellular channel 5 gene, Clic5 (Gagnon et al., 2006). Using an antibody provided by our collaborator Mark Berryman, Ohio University, we detected specific CLIC5 immunofluorescence in stereocilia of both cochlear and vestibular hair cells and in microvilli on the apical surface of Kolliker's organ during inner ear development, consistent with the known association of this protein with actin-based cytoskeletal structures of other polarized epithelial cells. Further supporting these results, our collaborator Peter Gillespie, Oregon Hearing Research Center, showed by mass spectrometry that CLIC5 is expressed at high levels in hair cell bundles isolated from the chicken utricle. In a recent collaboration, CLIC5 was shown to stabilize membrane-actin filament linkages at the base of hair cell stereocilia in a complex with radixin, taperin, and myosin VI (Salles et al., 2014). A mutation in the homologous human CLIC5 gene was recently shown by others to underlie the DFNB109 nonsyndromic deafness disorder.
Elmod1: We used a positional cloning approach to identify ELMO domain containing 1 (Elmod1) as the gene underlying two naturally occurring, allelic mutations (rda and rda-2J) that cause hearing and balance deficits in mice (Johnson et al., 2012a). Elmod1 is a poorly characterized gene with no previously reported mutant phenotypes. We determined that the deafness associated with the Elmod1rda and Elmod1rda-2J mutations is caused by cochlear hair cell dysfunction, as indicated by conspicuous elongations and fusions of inner hair cell stereocilia and progressive degeneration of outer hair cell stereocilia. ELMOD1 has been shown by others to function as a GTPase-activating protein (GAP) for the Arf family of small G proteins. Our finding connecting ELMOD1 deficiencies with stereocilia dysmorphologies thus establishes a link between small regulatory GTPases and the actin cytoskeleton of hair cell stereocilia.
Duox2thyd and Tpotee: Thyroid hormone (TH) is essential for proper cochlear development and function, and TH deficiencies cause variable hearing impairment in humans and mice. We have discovered spontaneous mouse mutations in two genes – dual oxidase 2 (Duox2) and thyroid peroxidase (Tpo) – that cause congenital hypothyroidism with associated dwarfism and profound hearing impairment. DUOX2 is an NADPH oxidase that generates hydrogen peroxide in thyroid follicular cells, which is needed by TPO for the oxidation of iodide and the production of thyroid hormone. TPO and DUOX2 mutations have been found to underlie many cases of congenital hypothyroidism in human patients. We identified the first mouse model for DUOX2-related hypothyroidism, the thyroid dyshormonogenesis (thyd) mutation of Duox2 (Johnson et al., 2007). We also characterized the first mouse models for TPO-related hearing loss, the teeny (Tpotee) and teeny 2J (Tpotee-2J) mutations (Johnson et al., 2014). The maturation of cochlear structures is delayed and tectorial membranes are abnormally thick in the Duox2 and Tpo mutant mice.
In addition to the mutations described above, we have characterized the auditory phenotypes of a retroviral insertion mutation of the eyes absent 1 gene (Eya1bor; Johnson et al., 1999), intragenic deletions of the phosphate regulating endopeptidase gene (PhexHyp, Hyp-2J, Hyp-Duk; Lorenz-Depiereux et al., 2004), and a missense mutation of the otoferlin gene (Otofdeaf5Jcs; Longo-Guess et al., 2007). We currently are at various stages in the analyses of nine other new mouse mutations that cause hearing loss and vestibular dysfunction, including ones that disrupt inner ear development, stria vascularis maturation, endolymph homeostasis, cochlear duct extension, and hair cell degeneration.
Mouse mutations often show variable phenotypes on different strain backgrounds. In collaborations with other investigators, we mapped genetic modifiers of hearing in mice with mutations in several different deafness-related genes, including Atp2b2dfw (Noben-Trauth et al., 1997; Zheng and Johnson, 2001), Tubtub (Ikeda et al., 2002), and Gpr98frings (Johnson et al., 2005). In another collaboration, we discovered a genetic interaction between mutations of the cadherin genes Cdh23 and Pcdh15 that causes a progressive hearing loss in digenic heterozygous mice and in human patients with compound mutations (Zheng et al., 2005). We wrote a collaborative review paper describing these and other hearing modifiers (Johnson et al., 2006) and since then have collaborated on additional projects to evaluate genetic modifiers of hearing in other mouse mutations, including Sobpjc (Calderon et al., 2006), Sod1KO (Johnson et al., 2010), and Pou1f1dw (Fang et al., 2011). We currently are studying strain background effects on hearing loss manifestation in mice with a mutation of the Atp6v1b1 gene.