Flexible Real-Time Polymerase Chain Reaction-Based Platforms for Detecting Deafness Mutations in Koreans: A Proposed Guideline for the Etiologic Diagnosis of Auditory Neuropathy Spectrum Disorder

Routine application of next-generation sequencing in clinical settings is often limited by time- and cost-prohibitive complex filtering steps. Despite the previously introduced genotyping kit that allows screening of the 11 major recurring variants of sensorineural hearing loss (SNHL) genes in the Korean population, the demand for phenotype- and variant-specific screening kits still remains. Herein, we developed a new real-time PCR-based kit (U-TOP™ HL Genotyping Kit Ver2), comprising six variants from two auditory neuropathy spectrum disorder (ANSD) genes (OTOF and ATP1A3) and five variants from three SNHL genes (MPZL2, COCH, and TMC1), with a distinct auditory phenotype, making this the first genotyping kit dedicated to ANSD. The concordance rate with Sanger sequencing, sensitivity, and specificity of this genotyping kit were all 100%, suggesting reliability. The kit not only allows timely and cost-effective identification of recurring OTOF variants, but it also allows timely detection of cochlear nerve deficiency for those without OTOF variants. Herein, we provide a clinical guideline for an efficient, rapid, and cost-effective etiologic diagnosis of prelingual ANSD. Our study provides a good example of continuing to update new key genetic variants, which will continuously be revealed through NGS, as targets for the newly developed genotyping kit.


Introduction
The realization of the importance of genomics in biological sciences has led to the use of genomics information for the development of precision medicine [1]. Sensorineural hearing loss (SNHL) is one of the most common neurosensory disorders in humans, affecting approximately 1 in every 1000 newborns and 1 in every 300 children by 4 years of age [2]. The early detection of SNHL and timely intervention are necessary to avoid devastating disabilities that are associated with auditory deprivation during the most sensitive period of brain development and plasticity [3,4]. The identification of the genetic etiology of SNHL could allow for the prediction of hearing prognosis and facilitate the establishment of an appropriate auditory rehabilitation strategy on the basis of underlying pathophysiological mechanisms [5,6]. Several studies have shown that congenital deafness is largely associated with genetic etiology, with a frequency of 50% or more [7]. Moreover, existing evidence indicates that there is a meaningful genetic contribution of Mendelian inheritance to post-lingual sensorineural deafness [8]. Recently, next-generation sequencing (NGS), based on molecular genetic testing, has been widely employed to identify candidate variants responsible for SNHL [9]. When time and cost is not an issue (i.e., for research purposes), and especially when blind to a specific auditory phenotype or genotype-phenotype correlation, NGS would be an extremely powerful tool that can scan all deafness genes. However, routine application of NGS in the everyday clinical setting is often limited by time and cost of the highly complex filtering steps [10][11][12]. As such, the development of molecular genetic testing that is more time-and cost-effective, with rapid screening capability, may be necessary in the context of precision medicine for SNHL.
We previously developed a genetic diagnostic kit that enabled screening of the prevalent variants of five major prelingual deafness genes in the Korean population [13]. The kit was highly flexible, time efficient, and cost effective, ideal for the everyday clinical setting [13]. Nevertheless, given an extreme etiologic heterogeneity, there is still demand for the development of additional convenient screening kits that ensure coverage of other important deafness genes. This is specifically the case when there is either a characteristic auditory phenotype or highly prevalent variants in a gene. For example, we recently found that four prevalent OTOF variants account for about 70% of Korean prelingual auditory neuropathy spectrum disorder (ANSD) with the anatomically intact cochlear nerve [14], making it unnecessary to go through other deafness genes irrelevant to such a phenotype. Furthermore, the predominant variant, p.Arg1939Gln, of OTOF, was frequently missed during NGS-based targeted exome sequencing due to capture failures and the poor coverage of the last exon of the cochlear isoforms of OTOF [14,15]. Considering this, a genotyping kit that facilitates screening exclusively for such prevalent OTOF variants would minimize the time and cost of NGS in Korean prelingual ANSD. This issue implicates some major variants of other deafness genes as well. Indeed, several prevalent variants that exert a potential hotspot/founder effect or manifest specific auditory phenotypes have been recently identified in the Korean population with varying degrees of SNHL [16][17][18][19][20].
Encouraged by the rising demand for phenotype-specific and variant-specific screening kits for deafness, we developed a genetic diagnostic kit (U-TOP™ HL Genotyping Kit Ver2), characterized by real-time PCR-based melting array techniques and peptide nucleic acid (PNA) probes, for screening of 11 variants from five deafness genes, which comprised two ANSD genes (OTOF and ATP1A3) and three SNHL genes (MPZL2, COCH, and TMC1), with a distinct auditory phenotype among Koreans. Any laboratory equipped with a real-time PCR machine can easily use this kit, without the limitation that at least one of the sequencers for Sanger sequencing or NGS must exist. In this study, we present the efficacy of this kit for detection of the corresponding variants from our previously established cohort in comparison with Sanger sequencing. We also prospectively tested the diagnostic yield of this kit among a prospectively recruited prelingual ANSD cohort. Based on this, we herein provide a clinical guideline on how otologists/pediatricians should deal with prelingual ANSD in children under this era of precise medicine and customized auditory rehabilitation.

Ethics Statement
This study was approved by the institutional review boards (IRBs) of Seoul National University Bundang Hospital (IRB-E-1905/540-001, 17 July 2020). We obtained written informed consent from all participants in this study. For child participants, written informed consent was obtained from their parents or guardians. We confirmed that all methods were carried out in accordance with relevant guidelines and regulations.

Participants
We collected patient samples with SNHL and family members who visited the clinics of SNUBH from May 2015 through May 2018, as well as those who executed genetic testing for causative variants. Among the 121 participants, we obtained 72 positive samples that had at least one of the 11 variants from the five genes (OTOF, ATP1A3, MPZL2, COCH, and TMC1) in homozygous, heterozygous, or compound heterozygous, as confirmed by Sanger sequencing. Forty-nine normal control subjects without any of the 11 variants were also collected.

Validation of the Real-Time PCR-Based Melting Array Genetic Diagnostic Kit Ver2
To validate the performance of the molecular diagnostic kit, which is referred to as the U-TOP™ HL Genotyping Kit Ver2 (SeaSun Biomaterials) [13], we tested the DNA samples and compared the results obtained using the U-TOP™ HL Genotyping Kit Ver2 with those obtained by Sanger sequencing. All experiments with randomly anonymous samples were blinded in terms of previously identified variants. Genomic DNA (gDNA) was extracted from the whole blood samples by using standard protocols (Gentra Puregene Blood Kit, Catalog No. 158389; Qiagen, Venlo, The Netherlands). The gDNA samples had 260 nm/280 nm absorbance ratios of over 1.5.

Real-Time PCR
Real-time PCR was performed using the U-TOP™ HL Genotyping Kit Ver2 with a CFX96 Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA). Here, 11 variants from five deafness genes, comprised of two ANSD genes (OTOF and ATP1A3) and three SNHL genes (MPZL2, COCH, and TMC1), were examined by using this kit as per the manufacturer's manual [13]. The data were analyzed with Bio-Rad CFX manager v1.6 software (Bio-Rad). These variants were characterized by the fluorescence signal of the detection probes and corresponding to T m , according to the standard protocol outlined by the manufacturer s manual [13].

Sanger Sequencing
All gDNA samples were examined individually with each of the 8 sets of primers in both directions and were confirmed by Sanger sequencing (Macrogen Inc., Seoul, Korea).

Statistical Analysis
To evaluate the performance of the U-TOP™ HL Genotyping Kit Ver2, the sensitivity, specificity, and 95% confidence intervals (CI) were calculated as described previously [13]. To assess the accuracy between the U-TOP™ HL Genotyping Kit Ver2 and Sanger sequencing, we calculated the kappa (κ) statistic, which reflects perfect agreement when the κ value is 0.81 and 1.00 [13]. Cohen s Kappa value between the two tests was determined as a reference to the following equation.
where p = the average positive predictive value (PPV) and negative predictive value (NPV) values from the reference articles; p 0 = the expected PPV and NPV for this study (equivalence); Z α/2 = 1.96; Z β = 0.842.

Design and Establishment of the U-TOP™ HL Genotyping Kit Ver2
Based on previous Korean and East Asian reports from several leading institutes, including ours, we selected 11 variants from five genes (ATP1A3, COCH, OTOF, MPZL2, and TMC1), which were frequently detected and were previously reported to cause distinctive audiological phenotypes. In detail, OTOF variants (p.Glu841Lys, p.Arg1856Trp, p.Leu1011Pro, p.Tyr1064Ter, and p.Arg1939Gln) were highly prevalent among Korean and East Asian prelingual ANSD (Supplementary Table S1). OTOF variants, including a founder variant (p.Arg1939Gln) among Koreans, account for approximately 90% of Korean prelingual ANSD cases with anatomically intact cochlear nerves [14]. Consistent with this, OTOF p.Arg1939Gln was found in 20 of the 26 alleles (76.9%) among Japanese congenital or early-onset ANSD cases and the founder effect was determined for this variant [21]. Additionally, the ATP1A3 Glu818Lys variant can cause postlingual-onset auditory synaptopathy, which is frequently accompanied by a distinct CAPOS (cerebellar ataxia, areflexia, pes cavus, optic atrophy, and sensorineural hearing loss) syndrome [16]. Although CAPOS syndrome is a rare disease that has been reported in less than 30 subjects thus far, the ATP1A3 variant, p.Glu818Lys, was detected in all cases [22]. In collaboration with our study [16], a recent study also demonstrated evidence for ANSD in most subjects with CAPOS syndrome caused by ATP1A3 p.Glu818Lys [23]. Furthermore, this kit would help to screen potential hotspot/founder variants that manifest specific auditory phenotypes. Very recently, we confirmed that a high proportion of Mendelian genetic contribution was due particularly to the p.Gln74* MPZL2 variant amongst patients with pediatric-onset mild-to-moderate SNHL [18]. The COCH variants are suggested to be a frequent cause of progressive cochleovestibular dysfunction in Koreans eventually requiring cochlear implantation [17]. Specifically, we reported two distinct COCH variants (p.Gly38Asp and p.Cys162Tyr), demonstrating a potential genotype-cochleovestibular phenotype correlation [17]. According to our recent work, the genetic etiology has been reported in 21 (52.5%) of 40 postlingually deafened cochlear implantees. Among them, the most frequent causative etiology was the TMC1 missense dominant variant, p.Asp572Asn, which was detected from three cases. Similarly, the TMC1 variant (p.Asp572Asn) accounts for about 4.4% (3/68) of progressive, postlingual autosomal dominant nonsyndromic hearing loss (ADNSHL) in the Chinese population [24]. In addition, the TMC1 variant (p.Arg34*) should be tested for in the genetic evaluation of subjects with autosomal recessive nonsyndromic hearing loss (ARNSHL) from North African and Southwest Asia [25].
A total of 11 variants were divided into three sets and were tested for validation of the efficacy of the U-TOP™ HL Genotyping Kit Ver2. Set 1 comprised three OTOF variants (p.Glu841Lys, p.Arg1856Trp, and p.Leu1011Pro) and one TMC1 variant (p.Asp572Asn). Set 2 comprised two COCH variants (p.Gly38Asp and p.Cys162Tyr), one OTOF variant (p.Tyr1064Ter), and one TMC1 variant (p.Arg34*). Set 3 comprised one OTOF variant (p.Arg1939Gln), one MPZL2 variant (p.Gln74*), and one ATP1A3 variant (p.Glu818Lys). Individually, all variants showed distinct melting curves that corresponded to each allele with a specific melting temperature (T m ) value for each diagnostic melting peak. The representative melting peaks that corresponded to both the heterozygous state and, whenever available, the homozygous state of each variant, as well as to those of the wild-type sequence, are depicted in Figure 1. While the heterozygous variants usually exhibited dual melting peaks at T m , except for two variants as a single heterozygous state (p.Arg1939Gln of OTOF and p.Arg34* of TMC1), the homozygous variant and wild-type sequence consistently exhibited a single melting peak at T m . Figure 1. Melting peaks of 11 variants from five deafness genes comprising two ANSD genes (OTOF and ATP1A3) and three SNHL genes (MPZL2, COCH, and TMC1) with a distinct auditory phenotype among Koreans. The blue, green, and red lines indicate wild type, heterozygous, and homozygous mutants, respectively. Each hearing loss set contains 3 or 4 peptide nucleic acid (PNA) probes labeled with FAM, HEX, Texas Red, and Cy5 channels.

Diagnostic Sensitivity, Specificity, and Concordance Rate
As compared with Sanger sequencing, the clinical sensitivity and specificity of the newly developed genotyping kit for the detection of target variants were 100% (95% confidence interval (CI): 94.9-100.0%) and 100% (95% CI: 92.7-100.0%), respectively (Table 1). A comparison study that included 127 clinical samples showed a concordance rate of 100% (95% CI: 96.9-100.0%) between the two methods. Accordingly, these data strongly indicate that the results obtained using U-TOP™ HL Genotyping Kit Ver2 were in perfect agreement with those obtained using Sanger sequencing.

Clinical Implications of the U-TOP™ HL Genotyping Kit Ver2 for ANSD
We also determined the diagnostic yield of the U-TOP™ HL Genotyping Kit Ver2 in the molecular diagnosis of prelingual ANSD by investigating six, prospectively recruited, consecutive cases with prelingual ANSD ( Table 2). All six subjects manifested bilateral ANSD, as defined by the electrophysiological tests that exhibited "absent ABR," but with the preservation of otoacoustic emissions or cochlear microphonics (Supplementary Figure S1). Importantly, the U-TOP™ HL Genotyping Kit Ver2 detected the homozygous missense variant p.Arg1939Gln of OTOF from two of the seven probands (33.3%), making it unnecessary to get the internal auditory canal (IAC) magnetic resonance imaging (MRI) from these two subjects. Among the other four probands who do not carry OTOF variants, two (50.0%) turned out to have a bilateral cochlear nerve deficiency (CND) based on IAC-MRI. In one proband (SB515-981), a meticulous review of the medical histories revealed predisposing factors associated with ANSD, such as prematurity and hypoxic damage. Interestingly, the hearing of one proband (SB554-1029) in whom neither the genotyping kit nor IAC-MRI revealed any abnormality, showed improvement over time. These results suggest that the newly developed genotyping kit not only allows the timely and cost-effective identification of OTOF variants prevalent in Korean prelingual ANSD subjects, but it also significantly contributes to timely imaging tests that allow the detection of cochlear nerve deficiency for those lacking these target OTOF variants. Indeed, it is worth noting that the correct etiologic diagnosis of prelingual ANSD of these six subjects were possible within one month (median 20 days, range 11-52 days), which would not have been possible without the application of this new genotyping kit.

Discussion
This newly developed genotyping kit (U-TOP™ HL Genotyping Kit Ver2), which is a real-time PCR-based method using the melting array techniques and PNA probes, merits special attention, because this-to the best of our knowledge-is the first genotyping kit highly specialized to diagnose ANSD, of which the molecular etiology has remained enigmatic even in this NGS era. Five prevalent variants from OTOF, and one recurring variant from ATP1A3, which are causative for prelingual ANSD and peri/postlingual ANSD, respectively, were included in this kit. Additionally, five recurring variants from three genes (MPZL2, COCH, and TMC1) that are known to cause a distinct auditory phenotype among Koreans were also included to complement the previously introduced U-TOP™ HL Genotyping Kit Ver1 in terms of coverage of recurring variants.
The analytical performance of the newly developed genotyping kit was evaluated by comparing it with Sanger sequencing in 121 clinical samples with varying degrees of SNHL. The newly developed genotyping kit showed a concordance rate of 100% with Sanger sequencing. The sensitivity and specificity of the genotyping kit were 100% with a Kappa value of 1.00, strongly indicating that the newly developed kit is reliable for screening SNHL in the Korean population.
Recent studies have shown a higher genetic load of OTOF variants explaining 91% of prelingual ANSD in Koreans with an anatomically intact cochlear nerve [14]. Furthermore, understanding the distribution pattern of the Korean OTOF alleles, which are predominantly concentrated on certain exons among Korean prelingual ANSD, could provide a justification for proposing a set of the most frequent OTOF variants to be included in the first-line screening [14,15]. As a result, the new genotyping kit developed in this study exclusively comprises the five most prevalent missense variants of OTOF causing prelingual ANSD in the Korean population, showing perfect agreement with Sanger sequencing analysis. Recently, the authors disclosed that the NGS-based targeted resequencing of the known 134 deafness genes does not necessarily guarantee the accurate detection of causative variants, particularly, the major allele (p.Arg1939Gln) of Korean ANSD with prelingual onset, due to capture failures and poor coverage of the last exon of the cochlear isoforms of OTOF [14,15]. Furthermore, Sanger sequencing of OTOF is often hindered by the large size of this gene, which comprises 46 exons, and additionally, not all clinics are equipped with a sequencer for Sanger sequencing [26]. Another de novo, recurring variant p.Glu818Lys in ATP1A3, which causes enigmatic perilingual/postlingual ANSD and is sometimes accompanied by CAPOS syndrome [16], was also included in this kit, making this kit an exceptionally efficient and accurate screening tool for Korean ANSD children. The beauty of this kit also lies in the fact that all the reactions can be achieved within a couple of hours.
Given this, early molecular etiologic diagnosis of ANSD children offers two advantages. First, early acquisition of genetic information using the genotyping kit allows the avoidance of unnecessary imaging studies for those with target OTOF variants, as evidenced by our two probands (SB543-1015 and SB) in the present study. Second, and conversely, the use of this kit can provide a guide for timely imaging to reveal ANSD-related anatomical abnormalities for those who lack these target OTOF variants. Previous studies consistently reported a relatively high incidence of abnormal brain MRI findings among children with ANSD [27,28]. For example, Roche et al. demonstrated one or more abnormal MRI findings in nearly 65% of children with ANSD, suggesting that MRI is the initial imaging technique of choice for children with ANSD [27]. Specifically, the presence of CND is largely associated with prelingual ANSD [29,30], particularly with the characteristics of unilateral auditory neuropathy [30]. Our prospective data showed that two (50.0%) of the four probands who do not carry OTOF variants based on the genotyping kit had bilateral CND based on IAC-MRI. Previous studies have also shown that CND is significantly associated with poor speech intelligibility or perception after cochlear implantation, posing potentially limited effectiveness and uncertain cost-benefits [31,32]. In addition, previous studies suggested that the sensitive period for successful CI outcomes for children with OTOF-related DFNB9 may be narrower than that for those with SLC26A4and GJB2-related deafness [14,33]. Supporting this, abnormal cortical auditory evoked potentials and delayed postsynaptic neurotransmission in children with ANSD may reflect abnormal maturation of the central auditory system, necessitating early intervention [34,35]. Collectively, OTOF-related DFNB9 with early intervention, i.e., before the age of 24 months, is considered a good candidate for CI with a good prognosis for outcome. Indeed, the application of this genotyping kit prompted the completion of etiologic diagnosis of most prelingual ANSD within approximately one month in our present study, which suggests that U-TOP™ HL Genotyping Kit Ver2 would significantly contribute to timely intervention and precise auditory rehabilitation of children with ANSD in Korea. Based on what we observed, as depicted in Figure 3, we propose a diagnostic pipeline for a more efficient, rapid, and cost-effective etiologic diagnosis of prelingual ANSD and provide a clinical guideline on how otologists/pediatricians should deal with prelingual ANSD children in this era of precise medicine and customized auditory rehabilitation. However, the relatively small sample size inherent from a pilot study precludes us from labeling the diagnostic yield of this kit for genetically diagnosing prelingual ANSD as 33%. A longitudinal follow-up study with a large cohort should be warranted to validate our preliminary results.
The implementation of the U-TOP™ HL Genotyping Kit Ver2 is not necessarily limited to ANSD. Indeed, the development of this new kit was based on the recent identification of several prevalent variants exerting a potential hotspot/founder effect or manifest specific auditory phenotypes in the Korean hearing-impaired population [15][16][17][18][19][20]. This new kit was designed to detect as many recurring and important deafness-causing variants as possible in conjunction with the previous version, U-TOP™ HL Genotyping Kit Ver1. By implementing this new kit, we expect that an additional 10.8% of the Korean pediatric-onset mild-to-moderate SNHL can be explained by the p.Gln74* MPZL2 variant, when compared with using only the previous kit. In addition, we expect that this new kit would help to predict the clinical course of DFNA9 subjects. Previously, our group reported distinct vestibular phenotypes depending on the location of COCH variants in DFNA9 subjects, indicating a potential genotype-phenotype correlation [17]. In detail, the p.Gly38Asp variant is closely linked to bilateral vestibulopathy, while the p.Cys162Try variant is likely to manifest a Meniere's disease-like phenotype [17]. Although the carrier frequency of the variants remains low, we believe that this new kit, based on the genotype-auditory correlation, could be used widely for specific auditory phenotypes, as a part of precision medicine. Likewise, we will continue to update new key genetic variants for deafness, which will continuously be revealed through NGS, as targets for the newly developed genotyping kit. The clinical applicability of this phenotype-specific and variant-specific genotyping kit surely outpowers that of NGS in a routine everyday clinical setting, especially when we are aware of the phenotype and can assume candidate genes. Figure 3. A diagnostic pipeline for more efficient, rapid, and cost-effective etiologic diagnosis of prelingual auditory neuropathy spectrum disorder (ANSD). This diagnostic strategy involves screening of the five most prevalent OTOF variants using a newly developed genotyping kit, imaging study, and exome sequencing for correct etiologic diagnosis of prelingual ANSD.

Conclusions
Taken together, our results suggest that U-TOP™ HL Genotyping Kit Ver2 for the screening of 11 variants from five deafness genes, comprised of two ANSD genes (OTOF and ATP1A3) and three SNHL genes (MPZL2, COCH, and TMC1), can be used as a reliable screening tool in the Korean population with a distinctive auditory phenotype. Particularly, this newly developed kit is highly specialized for diagnosing prelingual ANSD, allowing more efficient, rapid, and cost-effective etiologic diagnosis. Therefore, we have provided a clinical guideline on the management of prelingual ANSD children in this era of precise medicine and customized auditory rehabilitation.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
Data Availability: The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.