Genetic Factors Contribute to the Phenotypic Variability in GJB2-Related Hearing Impairment

Recessive variants in GJB2 are the most important genetic cause of sensorineural hearing impairment (SNHI) worldwide. Phenotypes vary significantly in GJB2-related SNHI, even in patients with identical variants. For instance, patients homozygous for the GJB2 p.V37I variant, which is highly prevalent in the Asian populations, usually present with mild-to-moderate SNHI; yet severe-to-profound SNHI is occasionally observed in approximately 10% of p.V37I homozygotes. To investigate the genomic underpinnings of the phenotypic variability, we performed next-generation sequencing of GJB2 and other deafness genes in 63 p.V37I homozygotes with extreme phenotypic severities. Additional pathogenic variants of other deafness genes were identified in five of the 35 patients with severe-to-profound SNHI. Furthermore, case-control association analyses were conducted for 30 unrelated p.V37I homozygotes with severe-to-profound SNHI against 28 p.V37I homozygotes with mild-to-moderate SNHI, and 120 population controls from the Taiwan Biobank. The severe-to-profound group exhibited a higher frequency of the crystallin lambda 1 (CRYL1) variant (rs14236), located upstream of GJB2, than the mild-to-moderate and Taiwan Biobank groups. Our results demonstrated that pathogenic variants in other deafness genes and a possible modifier, the CRYL1 rs14236 variant, may contribute to phenotypic variability in GJB2-realted SNHI, highlighting the importance of comprehensive genomic surveys to delineate the genotype-phenotype correlations. Recessive variants in GJB2 are the most important genetic cause of sensorineural hearing impairment (SNHI) worldwide. Phenotypes vary significantly in GJB2-related SNHI, even in patients with identical variants. For instance, patients homozygous for the GJB2 p.V37I variant, which is highly prevalent in the Asian populations, usually present with mild-to-moderate SNHI; yet severe-to-profound SNHI is occasionally observed in approximately 10% of p.V37I homozygotes. To investigate the genomic underpinnings of the phenotypic variability, we performed next-generation sequencing of GJB2 and other deafness genes in 63 p.V37I homozygotes with extreme phenotypic severities. Additional pathogenic variants of other deafness genes were identified in five of the 35 patients with severe-to-profound SNHI. Furthermore, case-control association analyses were conducted for 30 unrelated p.V37I homozygotes with severe-to-profound SNHI against 28 p.V37I homozygotes with mild-to-moderate SNHI, and 120 population controls from the Taiwan Biobank. The severe-to-profound group exhibited a higher frequency of the crystallin lambda 1 (CRYL1) variant (rs14236), located upstream of GJB2, than the mild-to-moderate and Taiwan Biobank groups. Our results demonstrated that pathogenic variants in other deafness genes and a possible modifier, the CRYL1 rs14236 variant, may contribute to phenotypic variability in GJB2-realted SNHI, highlighting the importance of comprehensive genomic surveys to delineate the genotype-phenotype correlations. Hearing impairment is the most common inherited sensory disorder in the world. Approximately 2% of children worldwide experience bilateral sensorineural hearing impairment (SNHI),1Morton C.C. Nance W.E. Newborn hearing screening--a silent revolution.N Engl J Med. 2006; 354: 2151-2164Crossref PubMed Scopus (1202) Google Scholar two-thirds of which are attributed to genetic causes.2Hilgert N. Smith R.J.H. Van Camp G. Forty-six genes causing nonsyndromic hearing impairment: which ones should be analyzed in DNA diagnostics?.Mutat Res. 2009; 681: 189-196Crossref PubMed Scopus (391) Google Scholar Of the >200 genes that have been linked to hereditary hearing impairment, recessive variants in the gap junction protein β 2 (GJB2) gene [Online Mendelian Inheritance in Man (OMIM) number 220290] are the most important,3Kenna M.A. Feldman H.A. Neault M.W. Frangulov A. Wu B.L. Fligor B. Rehm H.L. Audiologic phenotype and progression in GJB2 (connexin 26) hearing loss.Arch Otolaryngol Head Neck Surg. 2010; 136: 81-87Crossref PubMed Scopus (77) Google Scholar possibly accounting for approximately 20% of children with SNHI globally.4Chan D.K. Chang K.W. GJB2-associated hearing loss: systematic review of worldwide prevalence, genotype, and auditory phenotype.Laryngoscope. 2014; 124: E34-E53Crossref PubMed Scopus (205) Google Scholar GJB2 encodes connexin 26,5Kelsell D.P. Dunlop J. Stevens H.P. Lench N.J. Liang J.N. Parry G. Mueller R.F. Leigh I.M. Connexin 26 mutations in hereditary non-syndromic sensorineural deafness.Nature. 1997; 387: 80-83Crossref PubMed Scopus (1258) Google Scholar which is mainly expressed in the supporting cells of the inner ear.6Chen Y. Hu L. Wang X. Sun C. Lin X. Li L. Mei L. Huang Z. Yang T. Wu H. Characterization of a knock-in mouse model of the homozygous p.V37I variant in Gjb2.Sci Rep. 2016; 633279Google Scholar Connexin 26 is a transmembrane protein that forms cell-cell gap junctions that maintain potassium recycling and intercellular signaling in the inner ear.7Jagger D.J. Forge A. Connexins and gap junctions in the inner ear--it's not just about K⁺ recycling.Cell Tissue Res. 2015; 360: 633-644Crossref PubMed Scopus (75) Google Scholar,8Wingard J.C. Zhao H.B. Cellular and deafness mechanisms underlying connexin mutation-induced hearing loss - a common hereditary deafness.Front Cell Neurosci. 2015; 9: 202Crossref PubMed Scopus (108) Google Scholar More than 400 GJB2 variants have been reported as pathogenic or likely pathogenic in the Deafness Variation Database (University of Iowa, Iowa City, IA, http://deafnessvariationdatabase.org, last accessed January 13, 2023). Several pathogenic GJB2 variants are particularly prevalent in certain ethnic groups, such as NM_004004.6:c.101T>C (p.Met34Thr) and NM_004004.6:c.35delG (p.Gly12ValfsTer2) in White individuals,9Gürtler N. Kim Y. Mhatre A. Müller R. Probst R. Lalwani A.K. GJB2 mutations in the Swiss hearing impaired.Ear Hear. 2003; 24: 440-447Crossref PubMed Scopus (26) Google Scholar,10Gasparini P. Rabionet R. Barbujani G. Melçhionda S. Petersen M. Brøndum-Nielsen K. Metspalu A. Oitmaa E. Pisano M. Fortina P. Zelante L. Estivill X. High carrier frequency of the 35delG deafness mutation in European populations: Genetic Analysis Consortium of GJB2 35delG.Eur J Hum Genet. 2000; 8: 19-23Crossref PubMed Scopus (352) Google Scholar NM_004004.6:c.167delT (p.Leu56ArgfsTer26) in Ashkenazi Jews,11Morell R.J. Kim H.J. Hood L.J. Goforth L. Friderici K. Fisher R. Van Camp G, Berlin CI, Oddoux C, Ostrer H, Keats B, Friedman TB: Mutations in the connexin 26 gene (GJB2) among Ashkenazi Jews with nonsyndromic recessive deafness.N Engl J Med. 1998; 339: 1500-1505Crossref PubMed Scopus (482) Google Scholar and NM_004004.6:c.109G>A (p.Val37Ile), usually abbreviated as p.V37I, and NM_004004.6:c.235delC (p.Leu79CysfsTer3) in East Asian populations.12Hwa H.L. Ko T.M. Hsu C.J. Huang C.H. Chiang Y.L. Oong J.L. Chen C.C. Hsu C.K. Mutation spectrum of the connexin 26 (GJB2) gene in Taiwanese patients with prelingual deafness.Genet Med. 2003; 5: 161-165Abstract Full Text Full Text PDF PubMed Scopus (94) Google Scholar, 13Oguchi T. Ohtsuka A. Hashimoto S. Oshima A. Abe S. Kobayashi Y. Nagai K. Matsunaga T. Iwasaki S. Nakagawa T. Usami S.I. Clinical features of patients with GJB2 (connexin 26) mutations: severity of hearing loss is correlated with genotypes and protein expression patterns.J Hum Genet. 2005; 50: 76-83Crossref PubMed Scopus (79) Google Scholar, 14Dai P. Yu F. Han B. Yuan Y. Li Q. Wang G. Liu X. He J. Huang D. Kang D. Zhang X. Yuan H. Schmitt E. Han D. Wong L.J. The prevalence of the 235delC GJB2 mutation in a Chinese deaf population.Genet Med. 2007; 9: 283-289Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar, 15Kim S.Y. Kim A.R. Han K.H. Kim M.Y. Jeon E.H. Koo J.W. Oh S.H. Choi BY: Residual hearing in DFNB1 deafness and its clinical implication in a Korean population.PLoS One. 2015; 10e0125416Google Scholar The severity of GJB2-related SNHI is closely correlated with genotype. Patients carrying two alleles of truncating variants, such as c.35delG and c.235delC, tend to exhibit more severe hearing loss, but those with at least one allele of non-truncating variants, such as c.101T>C (p.Met34Thr) and c.109G>A (p.Val37Ile), usually demonstrate only mild-to-moderate hearing loss.13Oguchi T. Ohtsuka A. Hashimoto S. Oshima A. Abe S. Kobayashi Y. Nagai K. Matsunaga T. Iwasaki S. Nakagawa T. Usami S.I. Clinical features of patients with GJB2 (connexin 26) mutations: severity of hearing loss is correlated with genotypes and protein expression patterns.J Hum Genet. 2005; 50: 76-83Crossref PubMed Scopus (79) Google Scholar,16Chen P.Y. Lin Y.H. Liu T.C. Lin Y.H. Tseng L.H. Yang T.H. Chen P.L. Wu C.C. Hsu C.J. Prediction model for audiological outcomes in patients with GJB2 mutations.Ear Hear. 2020; 41: 143-149Crossref PubMed Scopus (15) Google Scholar,17Snoeckx R.L. Huygen P.L. Feldmann D. Marlin S. Denoyelle F. Waligora J. et al.GJB2 mutations and degree of hearing loss: a multicenter study.Am J Hum Genet. 2005; 77: 945-957Abstract Full Text Full Text PDF PubMed Scopus (435) Google Scholar However, significant variability in hearing severity exists, even in patients with the same genotype.16Chen P.Y. Lin Y.H. Liu T.C. Lin Y.H. Tseng L.H. Yang T.H. Chen P.L. Wu C.C. Hsu C.J. Prediction model for audiological outcomes in patients with GJB2 mutations.Ear Hear. 2020; 41: 143-149Crossref PubMed Scopus (15) Google Scholar,18Shen J. Oza A.M. Del Castillo I. Duzkale H. Matsunaga T. Pandya A. et al.Consensus interpretation of the p.Met34Thr and p.Val37Ile variants in GJB2 by the ClinGen hearing loss expert panel.Genet Med. 2019; 21: 2442-2452Abstract Full Text Full Text PDF PubMed Scopus (58) Google Scholar In a survey of 15 laboratories and clinics in multiple countries, 75% and 25% of p.V37I homozygous patients exhibited mild-to-moderate and severe-to-profound SNHI, respectively. Among patients carrying c.35delG homozygotes, 11% presented with mild-to-moderate SNHI.18Shen J. Oza A.M. Del Castillo I. Duzkale H. Matsunaga T. Pandya A. et al.Consensus interpretation of the p.Met34Thr and p.Val37Ile variants in GJB2 by the ClinGen hearing loss expert panel.Genet Med. 2019; 21: 2442-2452Abstract Full Text Full Text PDF PubMed Scopus (58) Google Scholar Several genetic factors, such as variants in other deafness genes (eg, gap junction genes) or cis-regulatory elements of GJB2, have been proposed to modulate the phenotypic variability of GJB2 variants.19Del Castillo F.J. Del Castillo I. DFNB1 non-syndromic hearing impairment: diversity of mutations and associated phenotypes.Front Mol Neurosci. 2017; 10: 428Crossref PubMed Scopus (60) Google Scholar,20Moisan S. Le Nabec A. Quillévéré A. Le Maréchal C. Férec C. Characterization of GJB2 cis-regulatory elements in the DFNB1 locus.Hum Genet. 2019; 138: 1275-1286Crossref PubMed Scopus (7) Google Scholar However, because of the high heterogeneity and diverse distribution of GJB2 variants in populations, the genetic factors contributing to its phenotypic variability are difficult to delineate and remain largely uncertain. Because of the extraordinarily high prevalence of the GJB2 p.V37I variant and the implementation of large-scale pilot newborn genetic screenings in the past decade,21Wu C.C. Tsai C.H. Hung C.C. Lin Y.H. Lin Y.H. Huang F.L. Tsao P.N. Su Y.N. Lee Y.L. Hsieh W.S. Hsu C.J. Newborn genetic screening for hearing impairment: a population-based longitudinal study.Genet Med. 2017; 19: 6-12Abstract Full Text Full Text PDF PubMed Scopus (53) Google Scholar, 22Lu C.Y. Tsao P.N. Ke Y.Y. Lin Y.H. Lin Y.H. Hung C.C. Su Y.N. Hsu W.C. Hsieh W.S. Huang L.M. Wu C.C. Hsu C.J. Concurrent hearing, genetic, and cytomegalovirus screening in newborns, Taiwan.J Pediatr. 2018; 199: 144-150.e1Abstract Full Text Full Text PDF PubMed Scopus (26) Google Scholar, 23Wu C.C. Hung C.C. Lin S.Y. Hsieh W.S. Tsao P.N. Lee C.N. Su Y.N. Hsu C.J. Newborn genetic screening for hearing impairment: a preliminary study at a tertiary center.PLoS One. 2011; 6e22314Google Scholar we were able to establish a cohort with a single GJB2 genotype homozygous for p.V37I. This cohort was used to investigate the genomic underpinnings responsible for the phenotypic differences in subjects with the same GJB2 genotype. Both next-generation sequencing (NGS) and case-control genetic association approaches were used for analysis. Our results demonstrate that pathogenic variants in other deafness genes and a possible modifier in the crystallin lambda 1 (CRYL1), namely the variant NM_015974.3 (CRYL1): c.261T>C (p.Gly87Gly), rs14236, may contribute to the phenotypic variability in GJB2 p.V37I homozygotes. From 2005 to 2021, a total of 505 patients with SNHI were confirmed to have the homozygous GJB2 p.V37I variant via Sanger sequencing at the National Taiwan University Hospital (Taipei, Taiwan). All patients in the present study cohort were interviewed and diagnosed with non-syndromic hereditary hearing impairment by practicing otologists at National Taiwan University Hospital. The patients with coexisting acquired risk factors for childhood SNHI, including cytomegalovirus infection, premature birth, meningitis, neonatal icterus, and previous exposure to noise or ototoxic medications, were excluded. All patients diagnosed with SNHI underwent audiological assessments using tone burst auditory brainstem response, behavioral audiometry, or pure tone audiometry pertinent to their age and cognitive status.16Chen P.Y. Lin Y.H. Liu T.C. Lin Y.H. Tseng L.H. Yang T.H. Chen P.L. Wu C.C. Hsu C.J. Prediction model for audiological outcomes in patients with GJB2 mutations.Ear Hear. 2020; 41: 143-149Crossref PubMed Scopus (15) Google Scholar,24Lin P.H. Hsu C.J. Lin Y.H. Lin Y.H. Lee H.Y. Wu C.C. Liu T.C. Etiologic and audiologic characteristics of patients with pediatric-onset unilateral and asymmetric sensorineural hearing loss.JAMA Otolaryngol Head Neck Surg. 2017; 143: 912-919Crossref PubMed Scopus (33) Google Scholar Hearing levels were determined by averaging the hearing thresholds of the better ear at 0.5, 1, 2, and 4 kHz and categorized as mild (25 to 40 dB), moderate (41 to 70 dB), severe (71 to 90 dB), or profound (>90 dB).25Clark J.G. Uses and abuses of hearing loss classification.ASHA. 1981; 23: 493-500PubMed Google Scholar Patients with severe or profound SNHI were designated as cases as the phenotype mismatched with that of p.V37I homozygosity, and those with mild-to-moderate SNHI typical for p.V37I homozygosity were designated as disease controls. The authors purposefully included 30 cases and 28 disease controls to study two extremes of the phenotypic spectrum of p.V37I homozygosity. The whole-genome sequencing data of 120 subjects retrieved from the Taiwan Biobank, which is representative of the Taiwanese population,26Wei C.Y. Yang J.H. Yeh E.C. Tsai M.F. Kao H.J. Lo C.Z. Chang L.P. Lin W.J. Hsieh F.J. Belsare S. Bhaskar A. Su M.W. Lee T.C. Lin Y.L. Liu F.T. Shen C.Y. Li L.H. Chen C.H. Wall J.D. Wu J.Y. Kwok P.Y. Genetic profiles of 103,106 individuals in the Taiwan Biobank provide insights into the health and history of Han Chinese.NPJ Genom Med. 2021; 6: 10Crossref PubMed Scopus (85) Google Scholar were ascertained as population controls. The Taiwan Biobank recruited volunteer participants aged >20 years without a cancer diagnosis at enrollment.27Feng Y.A. Chen C.Y. Chen T.T. Kuo P.H. Hsu Y.H. Yang H.I. Chen W.J. Su M.W. Chu H.W. Shen C.Y. Ge T. Huang H. Lin Y.F. Taiwan Biobank: a rich biomedical research database of the Taiwanese population.Cell Genom. 2022; 2100197Google Scholar Most participants of Taiwan Biobank did not have any critical illness at the time of enrollment. Although the Taiwan Biobank provided no audiometric data, the 120 subjects in the present study did not carry known pathogenic genetic variants causing hearing impairment in 724 genes by the same criteria as the present study cases. All patients included in this study were of the Han Taiwanese ethnicity. All eligible patients and/or their parents were provided informed consent before participation in this study. All subjects from the Taiwan Biobank also provided informed consent for data collection and analysis (TWBR11106-05). This study was conducted in accordance with the guidelines of the Declaration of Helsinki and approved by the Research Ethics Committees of the National Taiwan University Hospital (201912141RIND, 201803092RINB, and 202108074RINC). All cases and disease controls were subjected to NGS targeting 214 known deafness genes (Supplemental Table S1).28Wu C.C. Lin Y.H. Lu Y.C. Chen P.J. Yang W.S. Hsu C.J. Chen P.L. Application of massively parallel sequencing to genetic diagnosis in multiplex families with idiopathic sensorineural hearing impairment.PLoS One. 2013; 8e57369Google Scholar The NGS panel was designed to include the entire GJB2 gene (ie, exons, intron 1, and the 5-kb region upstream of exon 1). Exonic regions of five other gap junction genes (GJA1, GJB1, GJB3, GJB4, and GJB6) that had been linked to hereditary hearing impairment were also included in the NGS panel. Genomic DNA extracted from the peripheral blood mononuclear cells of the patients was fragmented into approximately 800 bp using the sonication method (Covaris, Woburn, MA), followed by measurement of DNA fragment length and concentration using an Agilent Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA) and Qubit (Thermo Fisher Scientific, Waltham, MA), respectively. The TruSeq Library Preparation Kit (Illumina Inc., San Diego, CA) was used for library preparation. Probe capture-based enrichment was performed using a SeqCap EZ Hybridization and Wash Kit (Roche NimbleGen, Madison, WI). Well-constructed libraries were pooled according to the concentration and expected read numbers and sequenced using Illumina MiSeq (Illumina Inc.) to generate paired-end reads of 300 nucleotides with an average read depth of 150 folds. Targeted panel sequencing data of cases and disease controls, as well as the whole-genome sequencing data of the Taiwan Biobank population controls,26Wei C.Y. Yang J.H. Yeh E.C. Tsai M.F. Kao H.J. Lo C.Z. Chang L.P. Lin W.J. Hsieh F.J. Belsare S. Bhaskar A. Su M.W. Lee T.C. Lin Y.L. Liu F.T. Shen C.Y. Li L.H. Chen C.H. Wall J.D. Wu J.Y. Kwok P.Y. Genetic profiles of 103,106 individuals in the Taiwan Biobank provide insights into the health and history of Han Chinese.NPJ Genom Med. 2021; 6: 10Crossref PubMed Scopus (85) Google Scholar adhered to the Broad Institute (Cambridge, MA) best practice workflow using Sentieon Genomics software version 201808 (San Jose, CA).29Kendig K.I. Baheti S. Bockol M.A. Drucker T.M. Hart S.N. Heldenbrand J.R. Hernaez M. Hudson M.E. Kalmbach M.T. Klee E.W. Mattson N.R. Ross C.A. Taschuk M. Wieben E.D. Wiepert M. Wildman D.E. Mainzer L.S. Sentieon DNASeq variant calling workflow demonstrates strong computational performance and accuracy.Front Genet. 2019; 10: 7Crossref PubMed Scopus (98) Google Scholar Paired-end reads were mapped to hg19 (ucsc.hg19.fasta, GATK bundle 2.8) using Sentieon BWA-MEM (Burrows-Wheeler Alignment-Maximal Exact Matches), followed by sorting and quality matrix calculations. Duplicate removal, insertion/deletion realignment, and base quality score recalibration were performed, followed by variant calling. All single-nucleotide variants and small insertions/deletions were called using Sentieon Haplotyper version 201808. Single-nucleotide variants and insertions/deletions were annotated using ANNOVAR version 2018041630Wang K. Li M. Hakonarson H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data.Nucleic Acids Res. 2010; 38: e164Crossref PubMed Scopus (9053) Google Scholar and filtered by minor allele frequency in population databases [Genome Aggregation Database (gnomAD) East Asian population31Karczewski K.J. Francioli L.C. Tiao G. Cummings B.B. Alföldi J. Wang Q. et al.The mutational constraint spectrum quantified from variation in 141,456 humans.Nature. 2020; 581: 434-443Crossref PubMed Scopus (4635) Google Scholar and Taiwan Biobank26Wei C.Y. Yang J.H. Yeh E.C. Tsai M.F. Kao H.J. Lo C.Z. Chang L.P. Lin W.J. Hsieh F.J. Belsare S. Bhaskar A. Su M.W. Lee T.C. Lin Y.L. Liu F.T. Shen C.Y. Li L.H. Chen C.H. Wall J.D. Wu J.Y. Kwok P.Y. Genetic profiles of 103,106 individuals in the Taiwan Biobank provide insights into the health and history of Han Chinese.NPJ Genom Med. 2021; 6: 10Crossref PubMed Scopus (85) Google Scholar] with ≤2% and CADD Phred version 1.6 ≥ 10.32Kircher M. Witten D.M. Jain P. O'Roak B.J. Cooper G.M. Shendure J. A general framework for estimating the relative pathogenicity of human genetic variants.Nat Genet. 2014; 46: 310-315Crossref PubMed Scopus (4113) Google Scholar The minor allele frequencies of the East Asian population were from the gnomAD version 2.1.1 whole-genome sequencing data (https://gnomad.broadinstitute.org, last accessed May 7, 2022). The frequencies in the Taiwanese population were from the GRCh37 version next-generation sequencing data in the Taiwan Biobank browser (Taiwan Biobank, Taipei, Taiwan; https://taiwanview.twbiobank.org.tw/browse38, last accessed May 7, 2022). Variants with a sequencing depth lower than 30 folds were excluded. Selected variants were classified as pathogenic/likely pathogenic by the American College of Medical Genetics and Genomics guideline-based VarSome website (https://varsome.com, last accessed June 22, 2023)33Kopanos C. Tsiolkas V. Kouris A. Chapple C.E. Albarca Aguilera M. Meyer R. Massouras A. VarSome: the human genomic variant search engine.Bioinformatics. 2019; 35: 1978-1980Crossref PubMed Scopus (1013) Google Scholar and the Deafness Variation Database (University of Iowa, Iowa City, IA; http://deafnessvariationdatabase.org, last accessed January 13, 2023).34Azaiez H. Booth K.T. Ephraim S.S. Crone B. Black-Ziegelbein E.A. Marini R.J. Shearer A.E. Sloan-Heggen C.M. Kolbe D. Casavant T. Schnieders M.J. Nishimura C. Braun T. Smith R.J.H. Genomic landscape and mutational signatures of deafness-associated genes.Am J Hum Genet. 2018; 103: 484-497Abstract Full Text Full Text PDF PubMed Scopus (173) Google Scholar All pathogenic variants were matched with the inheritance modes recorded in Online Mendelian Inheritance in Man (OMIM; Johns Hopkins University, Baltimore, MD, https://omim.org, last accessed January 13, 2023). The NM_accession numbers in the text and table are from RefGene (hg19_refGeneVersion.txt; https://annovar.openbioinformatics.org/en/latest/user-guide/download/#additional-databases, December 18, 2019). The rsID numbers are from dbSNP (hg19_avsnp150.txt; https://annovar.openbioinformatics.org/en/latest/user-guide/download/#additional-databases, December 6, 2019). Case-control analyses were based on unrelated cases without additional pathogenic variants as well as those of disease controls. Individual genomic variant calling format files of all cases, disease controls, and Taiwan Biobank population controls were joined together in a cohort variant call format file using the Sentieon joint calling function. Variant quality score recalibration was used for all variants. Genotype quality control was achieved using KGGSeq software version 2021101335Li M.X. Gui H.S. Kwan J.S. Bao S.Y. Sham P.C. A comprehensive framework for prioritizing variants in exome sequencing studies of Mendelian diseases.Nucleic Acids Res. 2012; 40: e53Crossref PubMed Scopus (201) Google Scholar with default settings (Supplemental Table S2). For variant-level quality control, variants were annotated as PASS in the variant quality score recalibration filtering score. The variant call format file from KGGSeq was converted to the plink format for Fisher exact test using the PLINK version 1.9 default allelic mode.36Purcell S. Neale B. Todd-Brown K. Thomas L. Ferreira M.A. Bender D. Maller J. Sklar P. de Bakker P.I. Daly M.J. Sham P.C. PLINK: a tool set for whole-genome association and population-based linkage analyses.Am J Hum Genet. 2007; 81: 559-575Abstract Full Text Full Text PDF PubMed Scopus (21869) Google Scholar An in-house Python version 3.7 script was used for downstream variant selection and intersection of the two associations (Figure 1A). Intersecting significant variants (P ≤ 0.05; odds ratio ≥ 2) were selected from the two association analyses. The linear model statistic was calculated using the lm ( ) function in R version 4.2.1 [https://cran.rstudio.com; through RStudio version 2022.07.0 (https://docs.posit.co/previous-versions/connect/#2022070)] to determine the relationship between heterozygosity and phenotypes. The authors used NGS and genetic association approaches to explore the causative variants in other deafness genes and potential genetic modifiers, respectively (Figure 1A). Four-hundred and forty-eight of the 505 p.V37I homozygotes showed typical mild-to-moderate SNHI (disease controls), and the remaining 57 GJB2 p.V37I homozygotes exhibited atypical phenotypes of severe-to-profound SNHI (cases), of whom the authors randomly selected 35 subjects and scrutinized their NGS data to search for additional pathogenic variants in other deafness genes (Supplemental Figure S1). After excluding the five cases with pathogenic variants in other deafness genes, NGS data of the other 30 cases were subsequently compared with those of 28 subjects selected from 448 disease controls as well as those of 120 Taiwan Biobank population controls for case-control association analyses separately. The 30 cases comprised 19 and 11 subjects with profound and severe SNHI, respectively, and the 28 controls comprised 11 and 17 subjects with moderate and mild SNHI, respectively. The sex, age at audiogram acquisition, and hearing threshold information of cases and disease controls are listed in Supplemental Table S3. The average age of the cases was 9.8 years, and that of the disease controls was 20.3 years, confirming that the cases (younger age with more severe phenotype) and disease controls (older age with milder phenotype) accurately represented the two extremes of the phenotypic spectrum of p.V37I homozygosity and addressed the GJB2 progressive nature. Genetic variants with Fisher exact P ≤ 0.05 and odds ratio ≥ 2 in both association analyses were ascertained for subsequent analyses. Causative variants in other deafness genes were identified in 5 of the 35 cases (Table 1), but none of the 28 disease controls. The five cases included two patients (DE6616 and DE7352) with homozygous SLC26A4 variants (NM_000441.2(SLC26A4): c.[919-2A>G]; [919-2A>G]) and one patient (DE6760) with compound heterozygous SLC26A4 variants NM_000441.2(SLC26A4): c.919-2A>G()2168A>G. All three patients were subsequently confirmed to have enlarged vestibular aqueduct in temporal bone imaging studies, which was consistent with the phenotype caused by pathogenic variants in SLC26A4 (OMIM number 600791).37Usami S. Abe S. Weston M.D. Shinkawa H. Van Camp G. Kimberling W.J. Non-syndromic hearing loss associated with enlarged vestibular aqueduct is caused by PDS mutations.Hum Genet. 1999; 104: 188-192Crossref PubMed Scopus (261) Google Scholar, 38Wu C.C. Chen P.J. Hsu C.J. Specificity of SLC26A4 mutations in the pathogenesis of inner ear malformations.Audiol Neurootol. 2005; 10: 234-242Crossref PubMed Scopus (16) Google Scholar, 39Li Q. Zhu Q.W. Yuan Y.Y. Huang S.S. Han D.Y. Huang D.L. Dai P. Identification of SLC26A4 c.919-2A>G compound heterozygosity in hearing-impaired patients to improve genetic counseling.J Transl Med. 2012; 10: 225Crossref PubMed Scopus (8) Google Scholar Patients DE2855 and DE4492 had heterozygous pathogenic variants NM_005982.4 (SIX1): c.533G>A (p.Arg178Lys) and NM_005422.2 (TECTA): c.6250+2T>G, respectively. Pathogenic variants of both SIX1 (OMIM number 605192; DFNA23)40Mosrati M.A. Hammami B. Rebeh I.B. Ayadi L. Dhouib L. Ben Mahfoudh K. Hakim B. Charfeddine I. Mnif J. Ghorbel A. Masmoudi S. A novel dominant mutation in SIX1, affecting a highly conserved residue, result in only auditory defects in humans.Eur J Med Genet. 2011; 54: e484-e488Crossref PubMed Scopus (26) Google Scholar and TECTA (OMIM number 601543; DFNA8/12)41Yasukawa R. Moteki H. Nishio S.Y. Ishikawa K. Abe S. Honkura Y. Hyogo M. Mihashi R. Ikezono T. Shintani T. Ogasawara N. Shirai K. Yoshihashi H. Ishino T. Otsuki K. Ito T. Sugahara K. Usami S.I. The prevalence and clinical characteristics of TECTA-associated autosomal dominant hearing loss.Genes (Basel). 2019; 10: 744Crossref PubMed Scopus (14) Google Scholar,42Hildebrand M.S. Morín M. Meyer N.C. Mayo F. Modamio-Hoybjor S. Mencía A. Olavarrieta L. Morales-Angulo C. Nishimura C.J. Workman H. DeLuca A.P. del Castillo I. Taylor K.R. Tompkins B. Goodman C.W. Schrauwen I. Wesemael M.V. Lachlan K. Shearer A.E. Braun T.A. Huygen P.L. Kremer H. Van Camp G. Moreno F. Casavant T.L. Smith R.J. Moreno-Pelayo M.A. DFNA8/12 caused by TECTA mutations is the most identified subtype of nonsyndromic autosomal dominant hearing loss.Hum Mutat. 2011; 32: 825-834Crossref PubMed Scopus (67) Google Scholar have been linked to autosomal dominant non-syndromic SNHI. Overall, the results demonstrated that in approximately 14% (5/35) GJB2 p.V37I homozygotes with severe-to-profound SNHI, the atypical phenotype may result from additional pathogenic variants in other deafness genes on top of homozygous GJB2 p.V37I variant.Table 1Pathogenic Variants in Other Deafness Genes Identified from GJB2 p.V37I Homozygous Cases with Severe-to-Profound SNHIPatientHLGeneVariant lociZygositygnomAD, %TWB, %ACMGDVDOMIMDE2855ProfoundSIX1NM_005982.4: c.533G>A (p.Arg178Lys)HeteroN/AN/ALikely pathogenic (PM5, PM1, PM2, PP3)N/AADDE4492SevereTECTANM_005422.2: c.6250+2T>GHetero0.06N/ALikely pathogenic (PVS1, PM2)PathogenicAD/ARDE6616∗Variant loci from https://www.ncbi.nlm.nih.gov/nuccore. ∗Patients diagnosed with de