*3.3. TRIOBP-4 in Hearing Loss*

In 2000, details were reported of a Palestinian family with nonsyndromic hereditary deafness, linked to a locus on chromosome 22, which was labeled as DFNB28 [5]. Homozygosity mapping implicated the *TRIOBP* locus, but no mutations were found in *TRIOBP-1*, the only open reading frame of *TRIOBP* known at that time. This directly led to the cloning of the long splice form *TRIOBP-5* and the discovery of a homozygous nonsense mutation within it [5] as well as separate mutations in other families with nonsyndomic deafness [5]. Simultaneously, studies of deafness linked-loci in families from India and Pakistan led to the discovery of a range of other *TRIOBP* mutations as well as cloning of *TRIOBP-4* and *6* [4]. Subsequently, a large number of studies have sequenced the *TRIOBP* gene in families or individuals with severe or profound prelingual hearing loss, revealing a wide range of seemingly pathogenic recessive mutations (Table 1, Figure 4b). These pathogenic mutations tend to be homozygous in patients with deafness, in many instances as a result of consanguinity. Patients have also been found with compound heterozygous expression of two different *TRIOBP* mutations.

The majority of mutations detected to date in patients are either nonsense or frameshift mutations in *TRIOBP-4*, which would lead to the expression of truncated TRIOBP-4 and longer splice variants such as TRIOBP-5 and 6, but with no predicted effect on TRIOBP-1. While many of the mutations lie with the large exon 7 (as in *TRIOBP-6*, Figure 1a), their location in the TRIOBP-4 protein varies considerably (Figure 4b). Many, but not all, of the predicted truncated proteins would still contain the R1 repeat region. Most either lack the R2 region, or would only partially express it. It is likely that these putative truncated proteins would be non-functional and degraded by the proteasome. Deafness in patients with these mutations therefore likely arises through lack of functional TRIOBP-4, which is consistent with the finding that mice lacking TRIOBP-4 (and longer isoforms) are profoundly deaf [50]. An alternative hypothesis would be that loss of the R2 region and/or C-terminal region of TRIOBP-4 would lead to expression of truncated proteins, which could interfere with normal stereocilia function. While the roles of these regions are not well characterized, it is notable that they are among the most highly conserved regions of *TRIOBP-4* in mammals (Figure 1c). In both instances, based on studies in mice [18], it is likely that stereocilia rootlets fail to form in the patients, leading to degeneration of stereocilia and thus hearing loss. Consistent with this, some patients with TRIOBP-4 mutations and hearing loss have had been successfully treated using cochlear implants, which bypass the need for stereocilia [50,51].

While most *TRIOBP* mutations implicated in deafness were found in patients with severe or profound hearing loss detected before speaking, compound heterozygous mutations have been reported in patients with moderate hearing loss or later onset severe hearing loss (Table 1). Notably, many of these patients either possess a mutation that lies 3′ of the *TRIOBP-4* reading frame, affecting longer splice variants only, or else have a mutation near the C-terminus of TRIOBP-4, meaning that the R2 domain would still be intact (Figure 4b). Potentially, the presence of some TRIOBP-4 functionality could therefore explain the milder phenotype, although there are instances of similar C-terminal mutations in patients with severe hearing loss.

Genome wide association studies have also shown that intronic SNP rs58389158 is associated with age related hearing impairment in non-Hispanic white individuals from California [52]. This SNP lies in an intron common to *TRIOBP-4*, *5*, and *6*. It is close to, and correlates strongly with, the coding SNP rs5756795, which leads to an F1187I protein variant [52]. The rs58389158 finding was replicated in the UK Biobank [52], in which rs5756795 was also found to be associated with both hearing difficulty and hearing aid use at the genome-wide level [53]. Common sequence variants in *TRIOBP* therefore appear to have an impact on hearing, in addition to rare nonsense and frameshift mutations.


**Table 1.** Published mutations in *TRIOBP* from individuals and families with hearing loss.


**Table 1.** *Cont*.

<sup>1</sup> Amino acid number of human TRIOBP-6, NM\_001039141.2 (for amino acid locations in TRIOBP-4 or 5, subtract 172 amino acids). <sup>2</sup> FS: Frameshift, NON: Nonsense, MIS: Missense. <sup>3</sup> HM: Homozygous, HT: Heterozygous, CHT: Compound heterozygous with the mutation indicated. <sup>4</sup> Numbers refer to isoforms, e.g., "5,6" indicates the mutation lies within TRIOBP-5 and TRIOBP-6. Key to locations: R1, R2: first and second repeats of TRIOBP-4, Centr.: Central coiled-coil domain of TRIOBP-1. <sup>5</sup> Country name, unless a more specific ethnicity was stated in the original paper. <sup>6</sup> Or prelingual, but severity not stated. <sup>7</sup> Personal communication of additional details by Prof. Shin-ichiro Kitajiri.
