Genetics of progressive hearing loss: a link between hearing impairment and dysfunction of mechanosensory hair cells

Abstract

Hearing loss is the most common form of sensory impairment that affects the population in developed countries. Although it is not a widely appreciated, a large number of children (1–2 per 1000) are born deaf. Most of us, however, know from personal experience in our circle of friends and family that the disease frequently affects elderly people [1–2]. Hearing loss is a complex disease which can be subdivided into categories depending on several criteria: the part of the ear that is affected (external and middle ear in conductive hearing loss; the inner ear in sensorineural forms of the disease); the age of disease onset (congenital, pre/post lingual, age-related, presbycusis); its degree of severity (moderate, mild, severe, profound); the affected frequencies (Low 2000Hz); and the association with other symptoms such as vision loss (syndromic) or not (non syndromic). Hearing impairment can be triggered by environmental factors such as exposure to excessive noise, chemicals, and certain medications that as a side effect perturb auditory function [3]. Many forms of hearing loss are of genetic origin and genetic predisposition is also thought to be an important factor in the development of noise- and age-dependent forms of the disease. 15% of all genetic forms of deafness are inherited as dominant traits, while 80% are transmitted recessively; the remaining patients carry mutations in the sex chromosomes or even in the genome of mitochondria, the organelles within our cells that generate energy [1–2]. While recessive deafness generally leads to congenital and profound hearing loss, dominant deafness tends to be late-onset and progressive. To date, approximately 50 genes have been linked to hearing loss but many more genes are thought to be involved [101]. These genes encode proteins with various functions such as transcription factors, adhesion receptor, cytoskeletal component, connexins, ion channels and synaptic proteins. In order to define and treat the pathophysiological changes that cause hearing loss, it will be essential to define the function of the genes that are linked to auditory impairment. As age-related hearing loss is a fast growing problem due to the increase in life expectancy, it will also be important to determine how interactions between genetic and environmental factors come into play to determine disease onset and progression. Progress in genomic research has accelerated the pace of auditory research by providing novel tools for the identification of genetic changes that cause hearing loss. The identification of the causative mutations has been greatly aided by the elucidation of the human genome. However, it is still a difficult task to identify the mutations and requires access to clinical data and DNA samples from members of families that are afflicted with genetic forms of the disease. An alternative to establish a connection between genetic changes and hearing loss is to focus on vast numbers of affected and non-affected humans and scrutinize the entire genome for features shared between affected individuals. This is the aim of the recent genome-wide association studies (GWAS). One of these studies has already focused on the identification of genetic alterations that confer an increased risk for the development of age-related hearing loss. This study identified a highly significant single nucleotide polymorphism (SNP) in the gene coding a metabotropic glutamate receptor (GRM7) [4]. With the availability of affordable next-generation sequencing technologies, future studies will probably identify additional significant SNPs and susceptibility genes by using a better genomic coverage and a higher number of patients. There is no doubt that GWAS is an instrumental tool in the identification of polymorphisms that are most commonly associated with hearing loss in humans. However, some mutations are rare and need to be identified by alternative strategies. Genetically tractable animal models are particularly important in this regard. Studies that were carried out over the last few years have indeed demonstrated that the mouse is a particularly useful model to study the human disease. The inner ear of mice and humans share many structural and functional similarities and the genetic program that controls the development and function of the auditory sense organ appears to be quite conserved. Capitalizing on these premises, several laboratories have shown that mutations in the same genes frequently lead to hearing loss in both humans and mice. For examples, studies focusing on mice in large breeding colonies that are maintained for research purposes have led to the identification of spontaneous mutations with defects in hearing function. These mice can be detected in the mouse colonies with relative ease because the inner ear also contains sensors for head movement. Therefore, affected mice are frequently hyperactive and circle. The identification of the mutations that cause inner ear dysfunction in these mice has led to the identification of several genes that are also linked to hearing loss in humans [5–6] The identification and analysis of sporadic mutations in mouse colonies is time consuming because the mutations are rare and has let only to a limited numbers of animal models for auditory research. New strategies have therefore been developed to increase the repertoire of mouse strains with defined auditory phenotypes by treating mice with mutagens. The alkylating agent ENU has been chosen most commonly because it is highly effective and introduces point mutations randomly. Because many human genetic diseases are caused by point mutations, some of the mice that are obtained by ENU mutagenesis are expected to be valuable animal models for the diseases. The first systematic ENU mutagenesis to identify mouse lines that are afflicted with recessive forms of hearing loss was carried out in our laboratory [7]. We chemically induced mutations randomly in the genome of male gametes, bred the mice for several generations and identified among them those with heritable hearing loss. One of the more than 20 lines from our screen, named samba, contains a mutation that affects a previously uncharacterized gene named Loxhd1 (lipoxygenase homology domain 1) [8]. We resolved the complete gene structure of Loxhd1 and demonstrated that it is expressed in the inner ear of mice. Loxhd1 encodes a protein of 2068 amino acids with an unusual structure. It consists of a 15 so-called PLAT (Polycystin/Lipoxygenase/Alpha-toxin) domains. A single PLAT domain consists of 120 amino acids and is present in proteins of diverse functions such as lipoxygenases, pancreatic lipase, the rab9-interacting protein 1, and the α-toxin of clostridium perfringens [9–12]. While the function of PLAT domains is unknown, previous works suggested that PLAT domains are able to interact with the plasma membrane, and in some cases also with other proteins [13]. Crystallographic studies have shown that PLAT domains form a β-sandwich consisting of two sheets, each of four strands, creating a highly hydrophobic pocket. The amino acids in the pocket are highly conserved in PLAT domains and one of them is mutated in samba mice, changing it from a hydrophobic residue (isoleucine 1342) into a polar one (asparagine) [8]. This is predicting to be deleterious for the proper folding of the PLAT domain, and may affect the overall structure of the entire LOXHD1 protein. Based on our findings in mice, we searched for mutations in human families segregating autosomal recessive hearing loss. We linked a genomic region on human chromosome 18 that includes the LOXHD1 gene to hearing impairment in a consanguineous Iranian family. Subsequent analysis of the genomic DNA by sequencing revealed a missense mutation in LOXHD1 that is homozygous only in affected family members [8]. The mutation is predicted to truncate LOXHD1 after the 5th PLAT domain. Affected members of the family show preferential hearing loss in the mid to high frequency range that can be detected already in childhood and is aggravated during ageing, leading finally to complete deafness. Interestingly, only three genes have so far been linked to progressive forms of hearing loss that are inherited as recessive traits. We have linked mutations in a gene called pejvakin (PJVK) to this form of hearing loss and others established a link between myosin 3a (MYO3A) and the disease [7, 14]. As the LOXHD1 gene is a very large, it spans more than 160 kB, it will be important to determine whether other LOXHD1 mutations might cause other forms of disease such as congenital or age-related hearing impairment. Interestingly, LOXHD1, pejvakin and MYO3a are expressed in the inner ear in hair cells, the mechanosensory cells that convert sound waves and head movements into electrical signals to provide our sense of hearing and balance. These findings suggest that defects in hair cell function might be more generally the cause for progressive forms of hearing loss in humans. The mechanically sensitive organelle of hair cells is the so-called hair bundle, an array of actin rich protrusions that emanate from the apical hair cell surface. Mechanosensors with actin-rich protrusions are found in the highly developed Chordates but not in the evolutionarily divergent Ecdysozoa (insect, nematode). LOXHD1 is found in the former but not the latter, raising the possibility that LOXHD1 is an evolutionary invention that has enabled the emergence of actin-based mechanosensory structures. In summary, our findings demonstrate that the mouse is a powerful instrument for auditory research, enabling the discovery of genes that cause hearing loss in humans and providing model systems to study disease mechanisms. One of the next big challenges is to utilize these mouse lines for the development of therapeutic strategies towards the treatment of hearing loss.