OTOF mutation analysis with massively parallel DNA sequencing in 2,265 Japanese sensorineural hearing loss patients

The OTOF gene (Locus: DFNB9), encoding otoferlin, is reported to be one of the major causes of non-syndromic recessive sensorineural hearing loss, and is also reported to be the most common cause of non-syndromic recessive auditory neuropathy spectrum disorder (ANSD). In the present study, we performed OTOF mutation analysis using massively parallel DNA sequencing (MPS). The purpose of this study was to reveal the frequency and precise genetic and clinical background of OTOF-related hearing loss in a large hearing loss population. A total of 2,265 Japanese sensorineural hearing loss (SNHL) patients compatible with autosomal recessive inheritance (including sporadic cases) from 53 otorhinolaryngology departments nationwide participated in this study. The mutation analysis of 68 genes, including the OTOF gene, reported to cause non-syndromic hearing loss was performed using MPS. Thirty-nine out of the 2,265 patients (1.72%) carried homozygous or compound heterozygous mutations in the OTOF gene. It is assumed that the frequency of hearing loss associated with OTOF mutations is about 1.72% of autosomal recessive or sporadic SNHL cases. Hearing level information was available for 32 of 39 patients with biallelic OTOF mutations; 24 of them (75.0%) showed profound hearing loss, 7 (21.9%) showed severe hearing loss and 1 (3.1%) showed mild hearing loss. The hearing level of patients with biallelic OTOF mutations in this study was mostly severe to profound, which is consistent with the results of past reports. Eleven of the 39 patients with biallelic OTOF mutations had been diagnosed with ANSD. The genetic diagnosis of OTOF mutations has significant benefits in terms of clinical decision-making. Patients with OTOF mutations would be good candidates for cochlear implantation; therefore, the detection of OTOF mutations is quite beneficial for patients, especially for those with ANSD.


Introduction
Hearing loss is one of the most frequent congenital sensory disorders, with one out of every 500 newborns having bilateral hearing loss [1]. It is reported that 50-60% of these cases show a genetic etiology, with 80% of them demonstrating autosomal recessive hearing loss [2]. The OTOF gene (Locus: DFNB9), encoding otoferlin, is reported to be one of the frequent causes of non-syndromic recessive sensorineural hearing loss. To date, more than 160 mutations in OTOF have been reported, and most of the patients with OTOF mutations have stable, prelingual and severe to profound hearing loss. OTOF is also known to be the most common cause of non-syndromic recessive auditory neuropathy spectrum disorder (ANSD) [3][4][5]. ANSD is a unique form of hearing loss characterized by the absence of or severe abnormalities in auditory brainstem response (ABR) and the presence of otoacoustic emissions (OAE). OTOF is mainly expressed in cochlear inner hair cells, and is necessary for synaptic exocytosis at the ribbon synapse [6]. While the function of the inner hair cells is impaired, that of the outer hair cells is preserved for the first one or two years; therefore, hearing loss due to OTOF gene mutation can also present as ANSD.
Recently, targeted exon sequencing of selected genes using massively parallel DNA sequencing (MPS) technology has been developed, enabling us to analyze massive amounts of data both relatively quickly and inexpensively improve the molecular diagnostic rate of hearing loss patients [7][8][9]. Although we previously reported the prevalence of hearing loss with OTOF mutations on the basis of Sanger sequencing [5], it is both time-consuming and costly to analyze a large number of patients by this method as the OTOF gene has a large number of exons. In this study, we conducted a genetic analysis of the OTOF gene in 2,265 Japanese hearing loss patients by MPS. The purpose of this study was to reveal the frequency and precise genetic and clinical background of OTOF-related hearing loss in a large hearing loss population.

Subjects
A total of 2,265 Japanese sensorineural hearing loss (SNHL) patients compatible with autosomal recessive inheritance (including sporadic cases) from 53 otorhinolaryngology departments nationwide participated in this study. Hearing loss was evaluated using pure-tone audiometry (PTA) classified by a pure-tone average over 500, 1000, 2000 and 4000 Hz in the better hearing ear. For children who could not undergo PTA, we used an average over 500, 1000, 2000 Hz for either auditory steady stem response (ASSR) or conditioned oriented reflex audiometry (COR), or the response threshold (dBnHL) from ABR. The severity of hearing loss was classified as follows: normal hearing, <25dB; mild hearing loss, 25-39dB; moderate hearing loss,

Variant analysis
Amplicon libraries were prepared using an Ion AmpliSeq Custom Panel (Applied Biosystems, Life Technologies), according to the manufacturer's instructions, for 68 genes including all the exons of the OTOF gene (NM_194248, NM_194323) reported to cause non-syndromic hearing loss (S1 Table). The detailed protocol was described elsewhere [10]. After preparation, the amplicon libraries were diluted to 20pM and equal amounts of 6 libraries for 6 patients were pooled for one sequence reaction.
Emulsion PCR and sequencing were performed according to the manufacturer's instructions. The detailed protocol was described elsewhere [10]. MPS was performed with an Ion Torrent Personal Genome Machine (PGM) system using an Ion PGM 200 Sequencing Kit and an Ion 318 Chip (Life Technologies).
The sequence data were mapped against the human genome sequence (build GRCh37/ hg19) with a Torrent Mapping Alignment Program. After sequence mapping, the DNA variant regions were piled up with Torrent Variant Caller plug-in software. After variant detection, their effects were analyzed using ANNOVAR software [11,12]. The missense, nonsense, insertion/deletion and splicing variants were selected from among the identified variants. Variants were further selected as less than 1% of 1) the 1,000 genome database [13], 2) the 6,500 exome variants (http://evs.gs.washington.edu/EVS/), 3) the Human Genetic Variation Database (dataset for 1,208 Japanese exome variants) [14], 4) the 333 in-house Japanese normal hearing loss controls, and 5) 1,000 control data in the deafness variation database [15]. All the mutations found in this study were confirmed by Sanger sequencing using exon-specific custom primers.

Results
Hearing level of the participating 2,265 patients was diagnosed as follows: mild hearing loss, 215 patients; moderate hearing loss, 679 patients; severe hearing loss, 524 patients; profound hearing loss, 599 patients; and unknown, 248 patients. The mutations found in this study were categorized into pathogenic, likely pathogenic, benign, likely benign and variant of uncertain significance according to the ACMG (American College of Medical Genetics) standards and guidelines [16]. The mutations judged to be pathogenic variants and likely pathogenic variants are presented in Table 1. Ten mutations including 6 previously reported variants and 4 novel variants (p.R425X, p.Y474X, p.W717X, p.L1003fs, p.Y1064X, p.Q1072X, p.I1449fs, p.R1856Q, p.R1172Q, c.4960+2T>C) were categorized as pathogenic variants. Five mutations (p.P489S, p.H513R, p.R1583H, p.R1792C, p.R1792H) were categorized as likely pathogenic variants. The 5 likely pathogenic variants were thought to be likely pathogenic because 1) they were found with previously reported pathogenic variants in trans (in different alleles of the gene): PM3 (p. R1856Q or p.R1172Q), 2) they were not found in the control: PM2, 3) the prediction programs   scores support their pathogenicity: PP3 and 4) co-segregation with family members with disease: PP1. The mutations judged to be likely benign and variants of uncertain significance are presented in S2

Discussion
In this study, 39 (1.72%) of 2,265 SNHL patients compatible with autosomal recessive (including sporadic cases) inheritance carried homozygous or compound heterozygous mutations in the OTOF gene. Two patients with heterozygous mutations (p.Y474X and p.R1172Q) showed an ANSD phenotype, and it is strongly suspected that they had OTOF related deafness. Possible explanations for these heterozygous cases are 1) the co-existence of copy number variations, 2) the existence of a second mutation in the exonic region that could not be covered in this study or in regulatory region of OTOF, which was not explored, 3) the contribution to hearing loss of an additional modulatory gene, and 4) the existence of a mutation in another gene (DIAPH3, AIFM1, ATP1A3 and mitochondrial 12SrRNA) which causes non-syndromic ANSD not examined in this study [18], so that the patients were just coincidental carriers of the OTOF mutations.
DFNB59 gene (also called as PJVK gene), reported to be a cause of non-syndromic ANSD [19], was also included in this study and no mutation was found in these two patients. We also performed copy number variation analysis for 68 genes (including OTOF and DFNB59) but did not identified any copy number variations in these two patients. Therefore, it is assumed that the frequency of hearing loss patients with OTOF mutations is at least 1.72% among autosomal recessive or sporadic SNHL cases. It was previously reported that OTOF mutations accounted for 1.  [24] populations. In this study, we analyzed 2,265 SNHL patients by MPS, which is the largest population analyzed to date. We had analyzed 160 SNHL patients by Sanger sequencing and reported that OTOF mutations accounted for 3.2-7.3% of recessive severe to profound SNHL [5]. This frequency is higher than that observed in this study. The main reason for this difference is thought to be that the subjects in this study included mild to The hearing levels in patients with biallelic OTOF mutations in this study were mostly severe to profound: 75.0% (24/32) had profound hearing loss, and 21.9% (7/32) had severe hearing loss. The commonly observed phenotype in patients with OTOF mutations is nonprogressive, congenital and severe to profound hearing loss. This is consistent with the results of this study. Genotype-phenotype correlations of OTOF have been discussed in past reports [3,18,23]. Patients with truncating mutations (nonsense and frameshift) or splice-site mutations basically show severe to profound hearing loss. Concerning non-truncating mutations (missense mutation and in-frame deletion), hearing level varies depending on each mutation or co-existing mutation; therefore some of them could show mild to moderate hearing loss [18,25]. In this study, only 1 patient (3.1%) had mild hearing loss; however, both mutations carried by the patient were variants of uncertain significance (p.D217G and p.A1802V), and it is unclear whether the true etiology of the hearing loss in this patient is due to mutations in the OTOF gene. Rare cases of temperature-sensitive ANSD, a particular form of ANSD, have been reported in some populations [3,24]; however, no temperature-sensitive ANSD was observed in this study.
As shown in Table 2, p.R1172Q was frequently identified in the patients participating in this study. Sixteen of 39 patients (41.0%) with biallelic OTOF mutations had homozygous p. R1172Q mutations. Eighteen of 39 patients (46.2%) had p.R1172Q with another mutation in compound heterozygosity. In summary, 34 of 39 patients (87.2%) had at least one p.R1172Q mutation. p.R1172Q has been proven to be a founder mutation [3], and as it was quite frequently detected in this study we believe it to be an important mutation in Japanese SNHL patients. p.Y474X (12.8%: 5/39), p.R1856Q (7.7%: 3/39), p.Q1072X (5.1%: 2/39) and p. Y1064X (5.1%: 2/39) were also detected in two or more patients. Populations among the various races have different mutation spectra and recurrent mutations. p.Q829X is quite frequently detected in Spanish [26], c.2905_2923delinsCTCCGAGCGGCA in Argentinean [21], p.V1778F in Ashkenazi Jewish [27], p.E57X and p.R1792H in Saudi Arabian [28] and p. E1700Q in Taiwanese [25] populations. Each recurrent mutation among these populations means that OTOF-related hearing loss is a major etiology of hearing loss in the respective country, not only in Japan.
The significance of the genetic diagnosis of OTOF mutations lies in its benefits for clinical decision-making. OTOF mutations represent one of the etiologies of ANSD. ANSD is heterogeneous disorder, and the outcomes of cochlear implantation for patients with ANSD vary [29,30]. Cochlear implantation has been reported to be effective for the patients with OTOF mutations as the OTOF gene mutations result in disruption of the synaptic exocytosis of inner hair cells, and the auditory nerves and spiral ganglions are preserved in patients with these mutations [31][32][33]. It is reported that outcome of cochlear implantation is predictable to some extent for various gene mutations [34]. Patients with OTOF mutations are predicted to show good outcomes; therefore, the detection of OTOF mutations is quite beneficial for the patients, especially for those with ANSD.
In this study, we also investigated whether the patients with biallelic OTOF mutations experience episodes of vertigo, and found that 31 of 32 patients (96.9%) with biallelic OTOF mutations had no such episodes. Although the OTOF gene is also expressed in vestibular hair cells, otoferlin-deficient mice show no evidence of apparent vestibular dysfunction based on gross evaluation by vestibular testing [6]. To date there have been no reports of a relationship between OTOF mutations and episodes of vertigo, and our results also showed that the clinical symptoms of vertigo are rarely observed in OTOF-related hearing loss patients.
In Japan, genetic testing for patients with SNHL using the Invader assay to screen for 46 mutations in 13 deafness genes was approved by the Ministry of Health, Labour and Welfare for inclusion in social health insurance coverage in 2012. Furthermore, the genetic testing was expanded in 2015 to allow screening for 154 mutations in 19 deafness genes using targeted genomic enrichment with MPS combined with the Invader assay [35]. We previously analyzed 717 hearing loss patients and achieved a 30% (212/717) diagnostic rate. OTOF mutations were also included among those 154 mutations and were identified in some cases. The pathogenic variants identified in this study will be added to this social health insurance-based genetic testing and further improvement in the diagnostic rate is expected.
Supporting information S1 Table.