Abstract
Background: Low-frequency non-syndromic hearing loss (LFNSHL) is a rare form of hearing loss (HL). It is defined as HL at low frequencies (≤2,000 Hz) resulting in a characteristic ascending audiogram. LFNSHL is usually diagnosed postlingually and is progressive, leading to HL affecting other frequencies as well. Sometimes it occurs with tinnitus. Around half of the diagnosed prelingual HL cases have a genetic cause and it is usually inherited in an autosomal recessive mode. Postlingual HL caused by genetic changes generally has an autosomal dominant pattern of inheritance and its incidence remains unknown. Summary: To date, only a handful of genes have been found as causing LFNSHL: well-established WFS1 and, reported in some cases, DIAPH1, MYO7A, TNC, and CCDC50 (respectively, responsible for DFNA6/14/38, DFNA1, DFNA11, DFNA56, and DFNA44). In this review, we set out audiological phenotypes, causative genetic changes, and molecular mechanisms leading to the development of LFNSHL. Key Messages: LFNSHL is most commonly caused by pathogenic variants in the WFS1 gene, but it is also important to consider changes in other HL genes, which may result in similar audiological phenotype.
Introduction
Hearing loss (HL) is the most prevalent disability of the human senses with a prevalence of approximately 2 in 1,000 live births [Watkin and Baldwin, 2012]. While there can be many preventable environmental factors negatively impacting hearing, from infections and ototoxic drugs to noise exposure, in almost half the confirmed cases with prelingual HL the cause is genetic [Morton et al., 2006]. HL can be either syndromic, as one of many symptoms of a given disorder, or non-syndromic, when HL is the only clinical manifestation. To date, 124 genes [van Camp and Smith, 2022, https://hereditaryhearingloss.org/, accession date: 10/11/22] have been identified as being involved in non-syndromic HL. Most cases of non-syndromic prelingual HL are inherited in an autosomal recessive mode (∼75–80%), while 20% are autosomal dominant. Very rarely HL is caused by changes in genes located on the X chromosome or in mitochondrial DNA (∼2%) [Sheffield and Smith, 2019]. HL with a genetic cause is given an additional designation, i.e., DFN (for deafness) with extra letters signifying the mode of inheritance (A, dominant; B, recessive; X, X-linked) and numbers indicating the order of discovery, e.g., DFNB1A refers to HL caused by pathogenic changes in the GJB2 gene, which was the first recessive HL gene discovered.
Audiological phenotypes in HL are diverse, and pathogenic changes in certain genes can cause a characteristic type of HL, e.g., affecting certain frequencies. Low-frequency non-syndromic hearing loss (LFNSHL) is defined as HL at frequencies <2,000 Hz but with normal hearing at higher frequencies [Bespalova et al., 2001]. It is usually developed postlingually (after speech acquisition) and progresses over time, usually involving other frequencies (a flat audiogram), and is sometimes accompanied by tinnitus. Generally, it is inherited in an autosomal dominant mode. Since initially it does not impair speech or speech recognition, it is less debilitating than other types of NSHL (and hence it is diagnosed later in life, when it progresses to other frequencies). LFNSHL has been tied to only a few genes – commonly WFS1, DIAPH1, MYO7A, and TNC reported in a few families, and in the case of CCD50, it is reported in only one family worldwide.
WFS1
WFS1 is a gene mapped to chromosome 4p16.1, spanning 33.4 kb of genomic DNA, containing 8 exons, transcribed to 3.6 kb cDNA, and translated into an 890 amino acid long glycoprotein called wolframin. Exon 1 is noncoding, while exon 8 encompasses 68% of the whole transcript [Inoue et al., 1998]. Wolframin is a transmembrane protein present in endoplasmic reticulum (ER) and plays a role in Ca2+ ion homeostasis [Takei et al., 2006] and ER stress response [Yamaguchi et al., 2004; Gharanei et al., 2013]. WFS1 is expressed ubiquitously in human tissues [de Falco et al., 1986, Philbrook et al., 2005] but especially in the nervous system and β-islets of the pancreas [Strom et al., 1998; Hofmann et al., 2003].
Mutations in WFS1 cause Wolfram syndrome (OMIM: 222300), a rare autosomal recessive neurodegenerative disorder characterized by diabetes insipidus, juvenile-onset diabetes mellitus, optic atrophy, and deafness (DIDMOAD, another name for the disease). Other symptoms include urinary tract and neuropsychiatric problems. It is a terminal disease, with progressive neurodegeneration leading to blindness and deafness, brainstem atrophy, and thence death by respiratory failure [Pallotta et al., 2019]. WFS1 mutations can also cause DFNA6/14/38, autosomal dominant LFNSHL (OMIM: 600965) [Bespalova et al., 2001], and Wolfram-like syndrome (OMIM: 614296), an autosomal dominant disease with symptoms similar to DIDMOAD – but without diabetes insipidus and with progressive, congenital deafness [Valéro et al., 2008]. Aditionally one pathogenic WFS1 variant segregated with an autosomal dominant congenital nuclear cataract in an Irish family [Berry et al., 2013]. WFS1 mutations have been correlated with type 1 diabetes [Li et al., 2020b] and psychiatric illnesses such as depression, bipolar disorder, schizophrenia, suicidal tendencies, and mood disorders [Munshani et al., 2021]. Interestingly, it is hypothesized that WFS1 may play a role in another neurodegenerative disorder, Alzheimer disease [Li et al., 2020a; Chen et al., 2022].
Such variety of diseases caused by changes in one gene poses the question about the localization and type of those mutations and their influence on the function of the translated protein. Indeed, mutations causing DIMOAD are in a homozygous or complex heterozygous state. They are usually truncating – producing a shorter, nonfunctioning transcript or leading to a nonsense-mediated decay of the transcript, while those causing DFNA6/14/38 are heterozygous and missense (Table 1). The proposed mechanism of pathogenicity of this transmembrane protein, assuming it is a multimer, is that when mutations on both alleles cannot produce a correct, translatable transcript, lack of functioning wolframin leads to ER stress, Ca2+ dyshomeostasis, and in turn neurodegeneration. In the case of DFNA6/14/38, one functioning allele is enough to keep full functionality in most of the tissues, but it is not enough to maintain normal hearing (i.e., mechanotransductive processes sensitive to ion metabolism). In other words, there is a dominant-negative effect based on the assumption that wolframin dimerizes or multimerizes [Cryns et al., 2003].
Nucleotide change . | Protein change . | Exon . | Domain . | Progressive HL . | Vertigo . | Tinnitus . | Inner ear malformations . | Familial segregation . | Family origin . | Reference . |
---|---|---|---|---|---|---|---|---|---|---|
c.511G>A | p.Asp171Asn | 5 | Extracellular 1 | Yes | No | Yes | ? | Not performed | Portugese | Gonçalves et al., 2014 |
c.577A>C | p.Lys193Gln | 5 | Extracellular 1 | ? | ? | ? | ? | Sporadic | German | Cryns et al., 2002 |
c.908T>C | p.Leu303Pro | 8 | N-terminal | +/− | No | No | ? | Yes | Japanese | Kobayashi et al., 2018 |
c.923C>G | p.Ser308Cys | 8 | N-terminal | No | Yes | No | ? | Yes | Japanese | Kobayashi et al., 2018 |
c.1235T>C | p.Val412Ala | 8 | N-terminal | ? | ? | ? | No | Yes | Korean | Choi et al., 2017 |
c.1480G>A | p.Gly494Ser | 8 | N-terminal | Yes | No | No | No | Not performed | Japanese | Kasakura-Kimura et al., 2017 |
c.1982A>G | p.Asn661Ser | 8 | N-terminal | No | No | No | ? | Yes | Japanese | Kobayashi et al., 2018 |
c.2005T>C | p.Tyr669His | 8 | N-terminal | No | No | No | ? | Yes | Taiwanese | Tsai et al., 2007 |
c.2020G>T | p.Gly674Trp | 8 | N-terminal | Yes | No | +/− | No | Yes | Chinese | Li et al., 2021 |
c.2021G>A | p.Gly674Glu | 8 | N-terminal | ? | ? | ? | ? | Yes | Dutch | Cryns et al., 2002 |
c.2021G>T | p.Gly674Val | 8 | N-terminal | ? | ? | ? | ? | Yes | Dutch | Cryns et al., 2002 |
c.2045A>G | p.Asn682Ser | 8 | N-terminal | No | No | Yes | ? | Sporadic | Japanese | Kobayashi et al., 2018 |
c.2054G>C | p.Arg685Pro | 8 | N-terminal | Yes | No | No | No | Yes | American | Bramhall et al., 2008 |
c.2086C>T | p.His696Tyr | 8 | N-terminal | Yes | +/− | +/− | No | Yes | Japanese | Sun et al., 2011 |
c.2096C>T | p.Thr699Met | 8 | N-terminal | ? | ? | ? | ? | Yes | German | Cryns et al., 2002 |
c.2108G>A | p.Arg703His | 8 | N-terminal | ? | ? | ? | ? | Sporadic | Japanese | Sun et al., 2011 |
c.2115G>C | p.Lys705Asn | 8 | N-terminal | Yes | ? | ? | No | Yes | German | Kunz et al., 2003 |
c.2146G>A | p.Ala716Thr | 8 | N-terminal | Yes | No | No | ? | Yes | Canadian | Young et al., 2001 |
+/− | No | +/− | ? | Yes | Japanese | Kobayashi et al., 2018 | ||||
? | ? | ? | ? | Yes | Japanese | Fukuoka et al., 2007 | ||||
c.2266C>T | p.Thr699Met | 8 | N-terminal | ? | ? | ? | ? | Yes | Dutch | Bespalova et al., 2001; Lesperance et al., 2003 |
c.2300–2302delTCA | p.delIle767 | 8 | N-terminal | ? | ? | ? | ? | Yes | Dutch | Cryns et al., 2002 |
c.2311C>T | p.Asp771His | 8 | N-terminal | +/− | ? | ? | ? | Yes | Swiss | Gürtler et al., 2005 |
c.2316G>A | p.Ala716Thr | 8 | N-terminal | Yes | ? | Yes | ? | Yes | Dutch, Irish | Bespalova et al., 2001; Lesperance et al., 2003 |
c.2335G>A | p.Val779Met | 8 | N-terminal | ? | ? | Yes | ? | Sporadic | American | Bespalova et al., 2001; Lesperance et al., 2003 |
c.2385G>C | p.Glu795Asp | 8 | N-terminal | Yes | No | Yes | ? | Sporadic | Japanese | Kobayashi et al., 2018 |
c.2419A>C | p.Ser807Arg | 8 | N-terminal | ? | ? | ? | ? | Yes | English | Cryns et al., 2002 |
c.2492G>A | p.G831Asp | 8 | N-terminal | ? | ? | ? | ? | Yes | American | Cryns et al., 2002 |
c.2507A>C | p.Lys836Thr | 8 | N-terminal | Yes | No | +/− | No | Yes | Japanese | Fujikawa et al., 2010 |
Yes | No | No | No | Yes | Japanese | Kasakura-Kimura et al., 2017 | ||||
+/− | +/− | +/− | ? | Yes | Japanese | Kobayashi et al., 2018 | ||||
c.2576G>A | p.Arg859Gln | 8 | N-terminal | Yes | +/− | +/− | No | Yes | American | Hildebrand et al., 2008 |
c.2576G>C | p.Arg859Pro | 8 | N-terminal | ? | ? | ? | ? | Sporadic | Japanese | Kasakura-Kimura et al., 2017 |
c.2576G>A | p.Arg859Pro | 8 | N-terminal | No | ? | ? | ? | Yes | American | Gürtler et al., 2005 |
c.2590G>A | p.Glu864Lys | 8 | N-terminal | ? | ? | ? | ? | Yes | Japanese | Fukuoka et al., 2007 |
Yes | ? | ? | ? | Yes | Japanese | Kobayashi et al., 2018 | ||||
No | No | No | No | Sporadic | Japanese | Kasakura-Kimura et al., 2017 | ||||
? | No | No | No | Sporadic | Chinese | Guan et al., 2020 | ||||
c.2591A>G | p.Glu864Gly | 8 | N-terminal | Yes | No | No | No | Yes | Chinese | Niu et al., 2017 |
c.2656T>C | p.Leu829Pro | 8 | N-terminal | Yes | ? | Yes | ? | Yes | American | Bespalova et al., 2001; Lesperance et al., 2003 |
c.2662G>A | p.Gly831Asp | 8 | N-terminal | ? | ? | Yes | ? | Yes | American | Bespalova et al., 2001; Lesperance et al., 2003 |
c.2700G>A | p.Ala844Thr | 8 | N-terminal | No | No | +/− | No | Yes | Japanese | Noguchi et al., 2005 |
Nucleotide change . | Protein change . | Exon . | Domain . | Progressive HL . | Vertigo . | Tinnitus . | Inner ear malformations . | Familial segregation . | Family origin . | Reference . |
---|---|---|---|---|---|---|---|---|---|---|
c.511G>A | p.Asp171Asn | 5 | Extracellular 1 | Yes | No | Yes | ? | Not performed | Portugese | Gonçalves et al., 2014 |
c.577A>C | p.Lys193Gln | 5 | Extracellular 1 | ? | ? | ? | ? | Sporadic | German | Cryns et al., 2002 |
c.908T>C | p.Leu303Pro | 8 | N-terminal | +/− | No | No | ? | Yes | Japanese | Kobayashi et al., 2018 |
c.923C>G | p.Ser308Cys | 8 | N-terminal | No | Yes | No | ? | Yes | Japanese | Kobayashi et al., 2018 |
c.1235T>C | p.Val412Ala | 8 | N-terminal | ? | ? | ? | No | Yes | Korean | Choi et al., 2017 |
c.1480G>A | p.Gly494Ser | 8 | N-terminal | Yes | No | No | No | Not performed | Japanese | Kasakura-Kimura et al., 2017 |
c.1982A>G | p.Asn661Ser | 8 | N-terminal | No | No | No | ? | Yes | Japanese | Kobayashi et al., 2018 |
c.2005T>C | p.Tyr669His | 8 | N-terminal | No | No | No | ? | Yes | Taiwanese | Tsai et al., 2007 |
c.2020G>T | p.Gly674Trp | 8 | N-terminal | Yes | No | +/− | No | Yes | Chinese | Li et al., 2021 |
c.2021G>A | p.Gly674Glu | 8 | N-terminal | ? | ? | ? | ? | Yes | Dutch | Cryns et al., 2002 |
c.2021G>T | p.Gly674Val | 8 | N-terminal | ? | ? | ? | ? | Yes | Dutch | Cryns et al., 2002 |
c.2045A>G | p.Asn682Ser | 8 | N-terminal | No | No | Yes | ? | Sporadic | Japanese | Kobayashi et al., 2018 |
c.2054G>C | p.Arg685Pro | 8 | N-terminal | Yes | No | No | No | Yes | American | Bramhall et al., 2008 |
c.2086C>T | p.His696Tyr | 8 | N-terminal | Yes | +/− | +/− | No | Yes | Japanese | Sun et al., 2011 |
c.2096C>T | p.Thr699Met | 8 | N-terminal | ? | ? | ? | ? | Yes | German | Cryns et al., 2002 |
c.2108G>A | p.Arg703His | 8 | N-terminal | ? | ? | ? | ? | Sporadic | Japanese | Sun et al., 2011 |
c.2115G>C | p.Lys705Asn | 8 | N-terminal | Yes | ? | ? | No | Yes | German | Kunz et al., 2003 |
c.2146G>A | p.Ala716Thr | 8 | N-terminal | Yes | No | No | ? | Yes | Canadian | Young et al., 2001 |
+/− | No | +/− | ? | Yes | Japanese | Kobayashi et al., 2018 | ||||
? | ? | ? | ? | Yes | Japanese | Fukuoka et al., 2007 | ||||
c.2266C>T | p.Thr699Met | 8 | N-terminal | ? | ? | ? | ? | Yes | Dutch | Bespalova et al., 2001; Lesperance et al., 2003 |
c.2300–2302delTCA | p.delIle767 | 8 | N-terminal | ? | ? | ? | ? | Yes | Dutch | Cryns et al., 2002 |
c.2311C>T | p.Asp771His | 8 | N-terminal | +/− | ? | ? | ? | Yes | Swiss | Gürtler et al., 2005 |
c.2316G>A | p.Ala716Thr | 8 | N-terminal | Yes | ? | Yes | ? | Yes | Dutch, Irish | Bespalova et al., 2001; Lesperance et al., 2003 |
c.2335G>A | p.Val779Met | 8 | N-terminal | ? | ? | Yes | ? | Sporadic | American | Bespalova et al., 2001; Lesperance et al., 2003 |
c.2385G>C | p.Glu795Asp | 8 | N-terminal | Yes | No | Yes | ? | Sporadic | Japanese | Kobayashi et al., 2018 |
c.2419A>C | p.Ser807Arg | 8 | N-terminal | ? | ? | ? | ? | Yes | English | Cryns et al., 2002 |
c.2492G>A | p.G831Asp | 8 | N-terminal | ? | ? | ? | ? | Yes | American | Cryns et al., 2002 |
c.2507A>C | p.Lys836Thr | 8 | N-terminal | Yes | No | +/− | No | Yes | Japanese | Fujikawa et al., 2010 |
Yes | No | No | No | Yes | Japanese | Kasakura-Kimura et al., 2017 | ||||
+/− | +/− | +/− | ? | Yes | Japanese | Kobayashi et al., 2018 | ||||
c.2576G>A | p.Arg859Gln | 8 | N-terminal | Yes | +/− | +/− | No | Yes | American | Hildebrand et al., 2008 |
c.2576G>C | p.Arg859Pro | 8 | N-terminal | ? | ? | ? | ? | Sporadic | Japanese | Kasakura-Kimura et al., 2017 |
c.2576G>A | p.Arg859Pro | 8 | N-terminal | No | ? | ? | ? | Yes | American | Gürtler et al., 2005 |
c.2590G>A | p.Glu864Lys | 8 | N-terminal | ? | ? | ? | ? | Yes | Japanese | Fukuoka et al., 2007 |
Yes | ? | ? | ? | Yes | Japanese | Kobayashi et al., 2018 | ||||
No | No | No | No | Sporadic | Japanese | Kasakura-Kimura et al., 2017 | ||||
? | No | No | No | Sporadic | Chinese | Guan et al., 2020 | ||||
c.2591A>G | p.Glu864Gly | 8 | N-terminal | Yes | No | No | No | Yes | Chinese | Niu et al., 2017 |
c.2656T>C | p.Leu829Pro | 8 | N-terminal | Yes | ? | Yes | ? | Yes | American | Bespalova et al., 2001; Lesperance et al., 2003 |
c.2662G>A | p.Gly831Asp | 8 | N-terminal | ? | ? | Yes | ? | Yes | American | Bespalova et al., 2001; Lesperance et al., 2003 |
c.2700G>A | p.Ala844Thr | 8 | N-terminal | No | No | +/− | No | Yes | Japanese | Noguchi et al., 2005 |
Nucleotide numeration is based on transcript NM_006005.3, codon numeration on NP_005996.2. HL, hearing loss; ?, not assessed in the study; +/−, varied among family members.
Table 1 presents WFS1 variants classified as causative for LFNSHL. Here, variants are excluded which have been identified in patients who did not have the characteristic ascending audiogram. In many cases, vestibular symptoms or inner ear morphology were not reported, which makes confirmation of pathogenicity of the studied variants difficult. Vertigo, tinnitus, and endolymphatic hydrops are typical features of Meniere disease, which is characterized by an ascending audiogram [Perez-Carpena and Lopez-Escamez, 2020].
All variants except for two (c.511G>A, c.577A>C) are positioned in exon 8 of WFS1. It is characteristic for LFNSHL-causing variants since variants causing DIDMOAD are spread across the whole gene [Khanim et al., 2001; Cryns et al., 2003]. Also, all except one (c.2300–2302delTCA) are single nucleotide substitution variants, resulting in an amino acid change. Interestingly, two variants have been independently identified a few times in West Asian populations (c.2507A>C, c.2590G>A). Some variants accumulate in the same nucleotides or codons, suggesting the existence of mutational hot spots (e.g., nucleotide c.2576, codons p.Gly674, p.Arg859, p.Glu864). Variants in WFS1 might be one of the most common causes of LFNSHL in many populations [Lesperance et al., 2003; Fukuoka et al., 2007; Moteki et al., 2016], especially since the presence of the same variant in families in different parts of the world, but with the same phenotype, suggest that they are not private variants.
DIAPH1
The DIAPH1 gene is located on chromosome 5q31.3 and produces a 5.7-kb transcript encoding 1,272 amino acid protein diaphanous homolog 1 or Diaphanous-related formin-1 (DRF1). DRF1 belongs to the formin family, subfamily DIAPH. As with every formin, it takes part in F-actin polymerization. The main features of DRF1 are its domains – FH1 and FH2, which are responsible for binding (FH1) actin monomers or their joining (FH2). Their activity is regulated by the N-terminal GTP-ase binding domain (GBD), Diaphanous inhibitory domain (DID), and C-terminal Diaphanous autoregulatory domain (DAD). The protein is stabilized by a coiled coil domain found between the DID and FH1 domains [Kang et al., 2017]. In their resting state, the DID and DAD domains interact, preventing exposure of the FH1 and FH2 domains. When GTPase Rho is present, it binds to the GBD domain, separating the DID and DAD domains and in consequence activating the FH1 and FH2 domains, leading to F-actin polymerization [Miyoshi et al., 2021].
DIAPH1is expressed ubiquitously in human tissues and DRF1 is found in cell cytoplasm [Uhlén et al., 2015, The Human Protein Atlas https://www.proteinatlas.org/ENSG00000131504-DIAPH1/tissue, accession date: 10/11/22]. It has also been found in the cochlea and spiral ganglion of a mouse model [Ninoyu et al., 2020]. DRF1 is responsible for cytoskeletal organization and as a consequence for cell shape and migration [Wen et al., 2004], cell polarization [Yamana et al., 2006], mitosis and mechanotransduction [Jégou et al., 2013], microtubule stabilization [Wen et al., 2004], and signal transduction [Palander and Trimble, 2020]. It is involved in brain development [Ercan-Sencicek et al., 2015] and platelet production [Zuidscherwoude et al., 2019].
DIAPH1 is the first gene described as an autosomal dominant deafness gene (DFNA1, OMIM: 124900). HL-causing variants are characterized in Table 2. Mutation in DIAPH1was reported in a large (78 affected individuals) Costa Rican family as responsible for progressive HL, starting at low frequencies in childhood and later developing at all frequencies [Lynch et al., 1997]. The same audiological profile has not been repeated in other families with deafness and variants in DIAPH1 – usually, this type of HL starts at high frequencies and progresses rapidly. It is usually accompanied by mild macrothrombocytopenia (MTP) and mild neutropenia, suggesting a syndromic character of the DIAPH1-caused HL [Stritt et al., 2016; Ganaha et al., 2017; Neuhaus et al., 2017; Bastida et al., 2018; Westbury et al., 2018; Karki et al., 2021; Rabbolini et al., 2022]. Later, it was found that the low-frequency HL phenotype in the Costa Rican family might have been associated with endolymphatic hydrops, found in one of the family members [Lalwani et al., 1998]. In some studies, mainly focused on the molecular mechanism of DIAPH1 mutations or on isolated HL, blood tests were not done. Since the presence of MTP or neutropenia could not be assessed, those studies neither confirm nor deny the existence of DIAPH1-related syndrome [Lynch et al., 1997; Lalwani et al., 1998; Ueyama et al., 2016; Kang et al., 2017]. To date, only one family has been described with a variant of DIAPH1 causing non-syndromic, progressive HL starting at high frequencies but without MTP [Kim et al., 2019]. In one large Chinese family (15 affected individuals), a variant of DIAPH1 segregated with HL similar to the originally described DFNA1 and without MTP but with features of auditory neuropathy [Wu et al., 2020]. DIAPH1 is also responsible for autosomal recessive seizures, cortical blindness, and microcephaly syndrome (SCBMS, OMIM: 616632) [Ercan-Sencicek et al., 2015; Al-Maawali et al., 2016] and is implicated in cancer development [Miao et al., 2021].
Nucleotide change . | Protein change . | Exon . | Domain . | Progressive HL . | Type of HL at diagnosis . | MTP . | NP . | AU . | Family origin . | Reference . |
---|---|---|---|---|---|---|---|---|---|---|
c.793G>T | p.Ala265Ser | 8 | DID | No | HF | No | No | No | Korean | Kim et al., 2019 |
c.1589T>G | p.Ile530Ser | 15 | CC | ? | HF | ? | ? | ? | Korean | Kang et al., 2017 |
c.3551_3552del | p.Glu1184AlafsTer11 | 26 | DAD | Yes | LF | No | No | Yes | Chinese | Wu et al., 2020 |
c.3575_3661del | p.Glu1192_Gln1220del | 27 | DAD | Yes | HF | Yes | Yes | No | English | Westbury et al., 2018 |
c.3610C>T | p.Arg1204X | 27 | DAD | Yes | HF | ? | ? | No | Japanese | Ueyama et al., 2016 |
c.3637C>T | p.Arg1213X | 27 | DAD | Yes | HF | Yes | Yes | No | American | Stritt et al., 2016 |
Japanese | Ganaha et al., 2017 | |||||||||
Japanese | Karki et al., 2021 | |||||||||
Australian | Bastida et al., 2018;Rabboloni et al., 2022 | |||||||||
c.3624_3625del | p.Arg1210SerfsTer31 | 27 | DAD | Yes | HF | Yes | Yes | No | German | Neuhaus et al., 2017 |
c.3629_3630del | p.Arg1210GlyfsTer31 | 27 | DAD | Yes | HF | Yes | Yes | No | English, Spanish | Westbury et al., 2018 |
c.3661+1G>T | p.Arg1212ValfsX22 | - | DAD | Yes | LF | ? | ? | No | Costa Rican | Lynch et al., 1997;Lalwani et al., 1998 |
Nucleotide change . | Protein change . | Exon . | Domain . | Progressive HL . | Type of HL at diagnosis . | MTP . | NP . | AU . | Family origin . | Reference . |
---|---|---|---|---|---|---|---|---|---|---|
c.793G>T | p.Ala265Ser | 8 | DID | No | HF | No | No | No | Korean | Kim et al., 2019 |
c.1589T>G | p.Ile530Ser | 15 | CC | ? | HF | ? | ? | ? | Korean | Kang et al., 2017 |
c.3551_3552del | p.Glu1184AlafsTer11 | 26 | DAD | Yes | LF | No | No | Yes | Chinese | Wu et al., 2020 |
c.3575_3661del | p.Glu1192_Gln1220del | 27 | DAD | Yes | HF | Yes | Yes | No | English | Westbury et al., 2018 |
c.3610C>T | p.Arg1204X | 27 | DAD | Yes | HF | ? | ? | No | Japanese | Ueyama et al., 2016 |
c.3637C>T | p.Arg1213X | 27 | DAD | Yes | HF | Yes | Yes | No | American | Stritt et al., 2016 |
Japanese | Ganaha et al., 2017 | |||||||||
Japanese | Karki et al., 2021 | |||||||||
Australian | Bastida et al., 2018;Rabboloni et al., 2022 | |||||||||
c.3624_3625del | p.Arg1210SerfsTer31 | 27 | DAD | Yes | HF | Yes | Yes | No | German | Neuhaus et al., 2017 |
c.3629_3630del | p.Arg1210GlyfsTer31 | 27 | DAD | Yes | HF | Yes | Yes | No | English, Spanish | Westbury et al., 2018 |
c.3661+1G>T | p.Arg1212ValfsX22 | - | DAD | Yes | LF | ? | ? | No | Costa Rican | Lynch et al., 1997;Lalwani et al., 1998 |
Nucleotide numeration according to transcript NM_005219.5, codon numeration according to NP_005210.3. All variants segregated with HL within patients’ families.
AU, auditory neuropathy; CC, coiled coil; DAD, C-terminal Diaphanous autoregulatory domain; DID, Diaphanous inhibitory domain; HF, high frequency; HL, hearing loss; LF, low frequency; MTP, macrothrombocytopenia; NP, neutropenia.
As with WFS1, the biomolecular mechanism underlying the comparatively mild autosomal dominant disease (and the severe autosomal recessive syndrome caused not only by pathogenic variants in the same gene but in the same protein domain) has been the subject of research. In the case of DIAPH1, the proposed mechanism is gain-of-function: HL-causing variants in the DAD domain (leading to its truncation) result in compromised autoinhibition and constitutive activity. Variants in other domains are also predicted to be disruptive for autoinhibition [Kim et al., 2019; Kang et al., 2017]. Consequent uncontrolled F-actin polymerization may lead to an abnormal cytoskeleton in the hair cells and macrothrombocytes [Lakha et al., 2021; Rabbolini et al., 2022], disturbing their proper function. Interestingly, some studied variants retain autoinhibitory activity (c.793G>T, c.1589T>G, c.3661+1G>T), while others do not (c.3610C>T, c.3637C>T, c.3624_3625del) [Ueyama et al., 2016; Lakha et al., 2021]. SCBMS, on the other hand, is caused by loss-of-function variants which truncate DIAPH1 protein in the FH2 domain, rendering it completely dysfunctional. Interestingly, patients with SCBMS do not experience HL [Ercan-Sencicek et al., 2015; Al-Maawali et al., 2016].
MYO7A
The MYO7A gene is located on chromosome 11q13.5 and encodes an unconventional myosin VIIA protein (MYO7A). Myosin VIIA belongs to the myosin family, a group of actin-based, ATP-dependent motor proteins, which take part in cellular processes engaging the cytoskeleton, such as extra- and intracellular transport [Li et al., 2016], organelle trafficking [Soni et al., 2005], cell division, shape, and motility [Heissler and Sellers, 2016]. They are characterized by an evolutionary-conserved N-terminal motor domain and neck domain, consisting of various number of calmodulin-binding IQ motifs and various C-terminal domains. In the case of myosin VIIA, there are 5 IQ motifs after the motor domain, followed by the coiled coil domain and two repeats of domains MyTH4 (myosin tail homology 4) and FERM (band 4.1-ezrin-radixin-moesin), interspaced with the SH3 (SRC homology 3) domain [Sato et al., 2017]. Myosin VIIA is expressed in the retina and inner ear [Hasson et al., 1995] – especially in the retinal pigment epithelium and tip-links of hair cell stereocilia [Grati and Kachar, 2011], making it a protein required for normal sight and hearing.
Pathogenic variants in MYO7Amay lead to Usher syndrome type 1B (USH1B, OMIM: 276900), DFNA11 (OMIM: 601317), and DFNB2 (OMIM: 600060). Usher syndrome type 1B is an autosomal recessive disease characterized by profound congenital deafness, retinitis pigmentosa, and vestibular dysfunction [Galbis-Martínez et al., 2021]. Only a few families with DFNA11 and LFNSHL have been reported (Table 3). All of them experienced postlingual, progressive HL which started at low frequencies and later encompassed all frequencies, resulting in a flat or sloping audiogram.
Nucleotide change . | Protein change . | Exon . | Domain . | Tinnitus . | Vertigo . | Family origin . | Reference . |
---|---|---|---|---|---|---|---|
c.1847G>A | p.Arg616Gln | 16 | Motor | No | No | Korean* | Joo et al., 2022 |
c.2003G>A | p.Arg668His | 17 | Motor | ? | ? | Chinese | Sang et al., 2013 |
c.2011G>A | p.Gly671Ser | 17 | Motor | Yes** | No | Chinese | Sun et al., 2011 |
Chinese | Li et al., 2018 | ||||||
c.2164G>C | p.Gly722Arg | 18 | Motor | No | No | American | Street et al., 2004 |
c.2557C>T | p.Arg853Cys | 21 | 5IQ | No | +/− | German | Boltz et al., 2004 |
Nucleotide change . | Protein change . | Exon . | Domain . | Tinnitus . | Vertigo . | Family origin . | Reference . |
---|---|---|---|---|---|---|---|
c.1847G>A | p.Arg616Gln | 16 | Motor | No | No | Korean* | Joo et al., 2022 |
c.2003G>A | p.Arg668His | 17 | Motor | ? | ? | Chinese | Sang et al., 2013 |
c.2011G>A | p.Gly671Ser | 17 | Motor | Yes** | No | Chinese | Sun et al., 2011 |
Chinese | Li et al., 2018 | ||||||
c.2164G>C | p.Gly722Arg | 18 | Motor | No | No | American | Street et al., 2004 |
c.2557C>T | p.Arg853Cys | 21 | 5IQ | No | +/− | German | Boltz et al., 2004 |
All variants segregated with HL within patients’ families. Nucleotide numeration based on transcript NM_000260.4, codon numeration based on NP_000251.3.
5-IQ, neck domain with 5 IQ motifs; ?, not assessed in the study; +/−, varied among family members.
*Variant identified in patient with LFNSHL and his son without HL; assumed familial segregation since son was younger than the age of HL onset in the proband.
**Both families reported tinnitus at the time of HL onset.
Interestingly, most MYO7A variants co-segregating with HL (either DFNA11 or DFNB2) are located in the motor domain, while those responsible for USH1B are found all over the gene length [Azaiez et al., 2018, https://deafnessvariationdatabase.org/gene/MYO7A, accession date: 10/11/22]. Similarly to Wolfram syndrome, USH1B-causative variants are truncating in nature, while those leading to HL are missense, resulting in single amino acid change and presumably changed protein properties.
DFNA11 is phenotypically very variable, and presented LFNSHL cases are rare. Many MYO7A variants are identified in patients with flat or sloping audiograms. HL severity and frequency range may vary even within the one family [Street et al., 2004] – suggesting the existence of a genetic modifier involved in DFNA11 development [Street et al., 2011].
TNC
TNC, a gene placed on chromosome 9q33.1, produces a 8.5-kb transcript translated into an extracellular matrix glycoprotein tenascin-C. Tenascin-C is composed of four types of domains: tenascin assembly domain, epidermal growth factor-like (EGF-L) repeats, fibronectin-type III-like repeats (FNIII), and a fibrinogen-like globe domain [Spring et al., 1989; Jones et al., 1990]. The tenascin assembly domain is responsible for oligomerization (as tenascin occurs physiologically as a hexamer), EGF-L is a ligand for the EGF receptor (EGF-R), and the FNIII domains allow for interactions with proteins and growth factors present in the extracellular matrix [Jones and Jones, 2000].
The TNC transcript undergoes alternative splicing, resulting in many isoforms, usually differentiated by an exact number of FNIII domains. A canonical transcript includes 14.5 EGF-L repeats and maximally 17 FNIII repeats − of which 8 are constitutive [Giblin and Midwood, 2015].
Tenascin-C is highly expressed during neural development, later restricting itself to smooth muscles [Tucker et al., 1993], the central nervous system [Tucić et al., 2021], stem cell niches [Chiquet-Ehrismann et al., 2014], and is transiently present in wound regeneration [Matsuda et al., 1999]. It is also expressed in the basilar membrane of the organ of Corti and the basal part of hair cells in mouse [Kwiatkowsta et al., 2016].
This glycoprotein serves many functions: cell migration, spreading, and survival; embryogenesis; and tissue remodeling [Jones and Jones, 2000; Giblin and Midwood, 2015]. Tenascin-C works as a promotor and inhibitor of growth in different cell populations, supporting its role as a boundary molecule [Treloar et al., 2009; Kwiatkowska et al., 2016]. It has also been involved in cancer as its disordered function promotes tumorogenesis [Midwood and Orend, 2009], angiogenesis [Kobayashi et al., 2016], metastasis [Lowy and Oskarsson, 2015], and chronic inflammation [Marzeda and Midwood, 2018].
In the case of HL, it has recently been found to be causative for DFNA56 (OMIM: 615629) [Zhao et al., 2013]. To date, only three variants (Table 4) have been described as causing DFNA56 [Zhao et al., 2013; Jin et al., 2022], and phenotypically, only one of them could be described as LFNSHL – c.5317G>A [Zhao et al., 2013]. The underlying molecular pathomechanism of DFNA56 is still unknown.
Nucleotide change . | Protein change . | Exon . | Domain . | Progressive HL . | Type of HL . | Familial segregation . | Family origin . | Reference . |
---|---|---|---|---|---|---|---|---|
c.1641C>A | p.Cys547X | 3 | EGF-L | ? | ? | Sporadic | Chinese | Jin et al., 2022 |
c.5317G>A | p.Val1773Met | 19 | FNIII (constitutive) | Yes | LF | Yes | Chinese | Zhao et al., 2013 |
c.5368A>T | p.Thr1796Ser | 19 | FNIII (constitutive) | ? | HF | Yes | Chinese | Zhao et al., 2013 |
Nucleotide change . | Protein change . | Exon . | Domain . | Progressive HL . | Type of HL . | Familial segregation . | Family origin . | Reference . |
---|---|---|---|---|---|---|---|---|
c.1641C>A | p.Cys547X | 3 | EGF-L | ? | ? | Sporadic | Chinese | Jin et al., 2022 |
c.5317G>A | p.Val1773Met | 19 | FNIII (constitutive) | Yes | LF | Yes | Chinese | Zhao et al., 2013 |
c.5368A>T | p.Thr1796Ser | 19 | FNIII (constitutive) | ? | HF | Yes | Chinese | Zhao et al., 2013 |
Nucleotide numeration based on transcript NM_002160.4, codon numeration based on NP_002151.2.
EGF-L, epidermal growth factor-like repeats; FNIII, fibronectin-type III-like repeats; HL, hearing loss; HF, high frequency; LF, low frequency.
CCDC50
The CCDC50 gene is located at chromosome 3q28 and codes for two alternative transcripts (8.6 kb and 1.9 kb) differentiated by the exclusion of exon 6 in the shorter transcript. Both transcripts are translated into isoforms (482 and 306 amino acid long, respectively) of the coiled-coil domain-containing protein 50 − CCDC50. CCDC50 is a soluble cytoplasmic protein and is ubiquitously expressed in both isoforms. It plays a crucial role in neural development [Min et al., 2020] and is also present in the mouse inner ear at different developmental stages, suggesting the role of CCDC50 in its development [Modamio-Høybjør et al., 2007].
CCDC50 has been identified as causative for autosomal dominant NSHL (DFNA44, OMIM: 607453) in one Spanish family [Modamio-Høybjør et al., 2003]. The HL was postlingual, progressive, starting at low frequencies, and over time extending to the higher frequencies, developing into a flat audiological profile. No tinnitus or vertigo was reported in the affected family, and inner ear malformations were excluded by computer tomography imaging. The pathogenic variant CCDC50 c.1394_1401dupCACGGCAT (p.Phe468HisfsX37) segregated with HL in the family in an autosomal dominant mode. To date, it is the only family with HL caused by variants in CCDC50. Since CCDC50 colocalizes with microtubules in the cytoskeleton of cells in the inner ear, it has been hypothesized that derangements of the cytoskeleton in the adult cochlea over time may lead to progressive HL [Modamio-Høybjør et al., 2007]. Interestingly, CCDC50 has been found to be upregulated in serum of mine workers suffering from noise-induced HL [Tumane et al., 2021].
Other Genes
Recently, CENPP was reported as a candidate gene for LFNSHL [Robles-Bolivar et al., 2022]. Variant in CENPP (c.849T>A, p. Cys283X) was found in three generational Swiss family suffering from a mild low-frequency HL which progressed over time, resulting in a sloping audiogram. The presence of endolymphatic hydrops was excluded by MRI in all family members except for one who was diagnosed with Meniere disease. The CENPP variant was inherited in an autosomal dominant mode.
SLC26A4 is known as causative gene for two autosomal recessive diseases – DFNB4 (OMIM: 600791) and Pendred syndrome (OMIM: 274600), both characterized by temporal bone abnormalities, which may result in an air-bone gap in the low frequencies [Honda and Griffith, 2022]. Variants in MYO6 gene are a known cause of DFNA22 (OMIM: 606346). The audiometric curve in DFNA22 may vary even among family members, and while it mostly causes progressive HL at middle and high frequencies, sometimes it may affect low frequencies [Oka et al., 2020].
Despite its typical audiological phenotype encompassing all tested frequencies and resulting in a flat audiogram, a variant in EYA4 was also found to be the cause of quickly progressive HL in low and middle frequencies in one Japanese patient [Abe et al., 2018]. While segregation analysis was not performed, mild mitral valve regurgitation present in the patient was consistent with a cardiac phenotype found in some cases of EYA4-caused SNHL. Pedigree analysis suggested an autosomal dominant or maternal inheritance. The same variant was described in conjunction with severe HL in a different patient, with dip in high frequencies [Kim et al., 2015].
Conclusion
Low-frequency non-syndromic HL is a rare, audiologically characteristic form of HL. As described here, to date, it has been reported as caused by only 4 of 124 NSHL genes, and clinically, LFNSHL is classified as an essential manifestation of only one disease entity − DFNA6/14/38. In the case of DFNA1, DFNA11, and DFNA56, LFNSHL was reported in individual families and did not represent a typical audiological phenotype, characterized by an ascending audiogram. WFS1 remains the most common genetic cause of LFNSHL, although heterozygotic changes in this gene do not always result in LFNSHL [Kytövuori et al., 2017; Kobayashi et al., 2018].
While the reviewed data indicate its clinical rarity, LFNSHL might go undiagnosed due to the late age of onset and additional, albeit rare, factors like environmental exposure to xenobiotics [Long and Tang, 2022]. Since LFNSHL does not impact speech recognition at first, this makes it less bothersome for the patient than high-frequency HL – but less likely to be diagnosed. In the case of older patients, the true nature of LFNSHL may also be obfuscated by presbycusis. All of these aspects might make LFNSHL more prevalent than initially assumed. Further studies are needed to assess the prevalence of LFNSHL and determine its full range of etiological factors.
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
Funding Sources
This research was funded by the National Science Centre, Poland (Grant No. 2016/22/E/NZ5/00470) and the Institute of Physiology and Pathology of Hearing.
Author Contributions
Nina Sara Gan has been responsible for data collection, analysis, interpretation and drafting the article. Dominika Oziębło, Henryk Skarżyński, and Monika Ołdak critically reviewed the article. All authors gave their final approval of version to be published.
Additional Information
ORCID numbers of the authors: Nina Sara Gan, https://orcid.org/0000-0001-8998-746X; Dominika Oziębło, https://orcid.org/0000-0002-3454-8002; Henryk Skarżyński, https://orcid.org/0000-0001-7141-985; Monika Ołdak, https://orcid.org/0000-0002-4216-9141.