Genetic testing for sporadic hearing loss using targeted massively parallel sequencing identifies 10 novel mutations Xiaodong Gu1*, Luo Guo1,2*, Haiting Ji1, Shan Sun1,2,Renjie Chai3, Lei Wang4, Huawei Li1,3,4,5§ 1
Department of Otolaryngology, Hearing Research Institute, Affiliated Eye and ENT Hospital, Fudan
University, Shanghai, China; 2Central laboratory, Eye and ENT Hospital of Shanghai Medical School, Fudan University. Shanghai, 200031, P.R.China; 3Key Laboratory for Developmental Genes and Human Disease, Ministry of Education, Institute of Life Sciences, Southeast University, Nanjing 210096, China; 4Institutes of Biomedical Sciences, Fudan University, Shanghai, 200032, P.R.China; 5
State Key Laboratory of Medical Neurobiology, Fudan University, Shanghai, China;
*These authors contributed equally to this work §
Corresponding author
Address: Room 611, Building 9, No. 83, Fenyang Road, Xuhui District, Shanghai, China, 200031. Telephone: 86-21-64377134-669
Fax: 86-21-64377151
E-mail:
[email protected] Conflict of interest statement: None
This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/cge.12431 This article is protected by copyright. All rights reserved
Acknowledgements This work was supported by grants from the Major State Basic Research Development Program of China (973 Program) (2011CB504506, 2010CB945503) and the National Natural Science Foundation of China (No. 81230019, 81070793, 81371094) and was sponsored by the Program for Changjiang Scholars and Innovative Research Team in University (IRT1010), the Specialized Research Fund for the Doctor Program of Higher Education (20120071110077), the Program of Leading Medical Personnel in Shanghai, and the Shanghai Rising-Star Program (12QA1400500), the Fundamental Research Funds for the Central Universities (2242014R30022, NO2013WSN085).
Abstract The genetic heterogeneity of nonsyndromic hearing loss (NSHL) has hampered the identification of its pathogenic mutations. Several recent studies applied targeted genome enrichment (TGE) and massively parallel sequencing (MPS) to simultaneously screen a large set of known hearing loss (HL) genes. However most of these studies were focused on familial cases. To evaluate the effectiveness of TGE and MPS on screening sporadic NSHL patients, we recruited 63 unrelated sporadic NSHL probands, who had various levels of HL and were excluded for mutations in GJB2, MT-RNR1, and SLC26A4 genes. TGE and MPS were performed on 131 known HL genes using the Human Deafness Panel oto-DA3. We identified 14 pathogenic variants in STRC, CATSPER2, USH2A, TRIOBP, MYO15A, GPR98, and TMPRSS3 genes in eight patients (diagnostic rate = 12.7%). Among these variants, 10 were novel compound heterozygous mutations. The identification of pathogenic mutations could predict the progression of HL, and guide diagnosis and treatment of the disease. Keywords: sporadic nonsyndromic hearing loss, mutation screening, targeted genome enrichment, massively parallel sequencing
This article is protected by copyright. All rights reserved
Introduction Hearing loss (HL) is a common sensory disorder in humans, and affects 360 million people worldwide (http://www.who.int/pbd/deafness/estimates/en/index.html). About 60% of the congenital sensorineural HL cases are caused by genetic factors. Nonsyndromic hearing loss (NSHL), in which hearing impairment is the only obvious clinical abnormality, accounts for 70% of genetic cases (1). To date, mutations in more than 60 genes are known to cause NSHL (http://hereditaryhearingloss.org). Due to the genetic heterogeneity of NSHL, most of the genetic tests for HL focus on several common mutations in frequently mutated genes. More recently, targeted genome enrichment (TGE) and massively parallel sequencing (MPS) were applied to screen hearing loss mutations (2-8). So far, most of the studies using TGE and MPS have focused on familial cases, with a diagnostic rate that varies from 28% to 62%(2-5, 7, 8); to our knowledge, only two previous studies have investigated sporadic cases, and they reported diagnostic rate of 20% and 32% respectively(3, 5). However, this is a critical issue to be addressed as the majority of HL patients seeking genetic counseling are sporadic. The identification of pathogenic mutations in the less commonly mutated genes could predict the progression of HL, thus guide diagnosis and treatment of the disease. In this study, we aimed to test the efficiency of TGE and MPS in genetic diagnosis for sporadic HL. We recruited 63 sporadic NSHL probands who were excluded for mutations in the GJB2, MT-RNR1, and SLC26A4 genes, and sequenced a total of 131 HL genes using the commercial target sequencing panel (Human Deafness Panel oto-DA3, Otogenetics Corporation., Norcross, GA USA).
This article is protected by copyright. All rights reserved
Materials and Methods Subjects Sixty-three NSHL probands with various levels of HL were recruited from HL patients seeking genetic testing and counseling through the otology clinic at the Eye and ENT Hospital of Fudan University (Table S1). Full history, clinical evaluations, and audiograms were available for each patient. In accordance with the ethics committee of Fudan University, all participants or their parents gave written, informed consent to participate in this study.
Library preparation and massively parallel sequencing Genomic DNA was obtained from whole blood of the probands. Sequencing library was prepared using DNA library preparation kit (New England Biolabs, Ipswich, MA USA, catalog# E6040) and Human Deafness Panel oto-DA3 (Otogenetics Corporation., Norcross, GA USA). Captured library DNA was sequenced on an Illumina Hiseq2000 machine using 100 bp paired-end reads.
Bioinformatics and validation of the variants We implemented a bioinformatics analysis pipeline with open source software. The pipeline consisted of three parts: variant calling, variant filtering and candidate variant validation (overview is shown in Figure 1). Sanger sequencing was performed to validate the candidate variants. A SYBR Green-based quantitative PCR (qPCR) analysis was performed to verify the CNVs detected by MPS.
Results Targeted massively parallel sequencing of sporadic HL probands
This article is protected by copyright. All rights reserved
A total of 131 known HL genes were targeted for sequencing (Table S2). The target region contains 1647 exons and their exon-intron boundaries covering 478kbp of the human genome. A detailed statistics for the MPS data of each sample is shown in Table S3. After MAF filtering 249 variants of high sequencing quality in 62 probands (an average of 4 variants per proband) were identified (Table S4). Since sporadic HL is consistent with a possible recessive inheritance, 25 biallelic variants in 14 patients were retained for further inspection (Table 1). With the presumption of recessive inheritance, compound heterozygous variants should be on different homologous chromosomes. Phase analysis of compound heterozygous variants were performed by sequencing the probands’ parents (if the parental DNA was available) or by analyzing the MPS data. Candidate variants in J468 and G557 patients were excluded (Figure S1).
Pathogenic variants Twenty-one candidate variants from 12 probands were retained after phase analysis (Table 1). We required a compatible phenotype with previous studies to support a confident diagnosis. Overall, fourteen pathogenic variants were identified. The pathogenicity of missense variants were evaluated by four in silico methods. We identified compound heterozygous TRIOBP (OMIM 609761) variants in proband G571 (Table 1). Audiograms showed bilateral congenital profound HL, which was consistent with the previously reported familial cases (Figure 2) (9). The two nonsense mutations (p.R861X and p.R920X; NM_001039141) in proband G571 were both located in exon 6 of TRIOBP. Compound heterozygous TMPRSS3 (OMIM 605511) mutations were identified in proband G562 who showed moderate HL at the age of seven (Figure 2). The novel nonsense mutation (p.S99X; NM_032404) led to the deletion of the enzyme’s protease domain.
This article is protected by copyright. All rights reserved
Four compound heterozygous MYO15A (OMIM 602666) mutations were identified. Proband J495 was identified with two variants in the motor domain, and proband J173 was identified with one variant in the motor domain and one variant in the FERM domain. Audiograms of the two probands showed congenital severe HL with a typical slope toward high frequencies (Figure 2). A molecular model of MYO15A motor domain was constructed based on the X-ray structure of the dictyostelium discoideum myosin motor domain (PDB ID 1d0xA) to show the position of the motor domain variants (Figure S2). Compound heterozygous USH2A (OMIM 608400) mutations were identified in proband J406, who showed bilateral mild-to-profound sensorineural HL with a lower threshold in high frequencies (Figure 2). The novel missense mutation p.M5099T was located in the conserved functional domain. Compound heterozygous GPR98 (OMIM 602851) mutations were identified in proband G567 (Table 1), who showed a typical phenotype: congenital HL with a slope toward high frequencies (Figure 2). Proband J457 was three years old when diagnosed with moderate HL at high frequencies (Figure 2). We identified a homozygous deletion of STRC (OMIM 606440) and CATSPER2 (OMIM 607249) in this proband (Figure S3).
Variants of uncertain significance Seven variants of uncertain significance were identified (Table 1). Functional studies should be applied to determine the pathogenicity of these variants. Two heterozygous missense TRIOBP (OMIM 609761) variants were identified in proband J449 in exon 6 and 8. Audiograms revealed severe to profound HL for the left ear, and mild to moderate HL for the right ear (Figure S4). To date, all of the fifteen mutations (including the 2 novel variants) causing DFNB28 deafness in the TRIOBP gene are located in exon 6, and fourteen of them are truncating nonsense and frameshift mutations.
This article is protected by copyright. All rights reserved
Mutations in TJP2 (OMIM 607709) have only previously been reported once in a family with late-onset autosomal dominant HL. We identified a homozygous TJP2 missense variant in proband J431, which suggested an autosomal recessive inheritance pattern, however no recessive TJP2 mutation has been reported to cause NSHL. In proband G564, we identified a missense PCDH15 (OMIM 605514) variant (p.S362R; NM_001142767) in the highly conserved third extracellular cadherin domain which may cause an adhesion defect, and a truncating frameshift insertion (p.N1626Rfs*17; NM_001142771.1) in the cytoplasmic domain of the CD3 isoform. The latter was the first mutation found in the third extracellular cadherin domain of PCDH15. The parental DNA of the proband were unavailable, so co-segregation analysis of the variants were not applied. The proband manifested moderate HL at age of 12 (Figure S4), which was not compatible with previous studies. Audiograms of J540 showed moderate HL at all frequencies (Figure S4). Neither the onset age (48 years old) nor the degree of HL was compatible with the previously reported GPR98 cases.
Discussion We recruited a total of 89 NSHL probands, and 26 (29.2%) of them was identified with mutations in the GJB2, MT-RNR1 or SLC26A4. The remaining 63 undiagnosed probands were studied by TGE and MPS. Fourteen pathogenic variants, including ten novel compound heterozygous variants, in seven genes were identified in 12.7% (8/63) of the patients. All the novel pathogenic missense mutations were predicted as deleterious by at least three in silico prediction methods. No individual HL gene recorded a mutation detection rate of more than 4%, which demonstrated the genetic heterogeneity of Chinese sporadic NSHL patients. To our knowledge, the homozygous deletion of STRC and CATSPER2 and the heterozygous mutations in TRIOBP and TMPRSS3 were identified for the first time in these three genes among Chinese
This article is protected by copyright. All rights reserved
HL patients. As shown in Table S1, severe to profound HL group showed a higher diagnostic rate compared to the moderate HL group, while no variant was identified in the mild HL group, likely due to its small sample size. Unlike familial cases, the candidate variants in sporadic cases lack support from co-segregation analysis, so we required compatible phenotypes with previous studies to reduce false positive diagnosis. To our knowledge, only two previous studies have investigated sporadic cases(3, 5). Yang et al. sequenced 79 HL genes in 93 sporadic probands who were excluded of mutations in 3 common mutated genes and obtained a mutation detection rate of 20%. Shearer et al. sequenced 54 to 66 HL genes in 34 sporadic NSHL patients and diagnosed 11 patients (32%) with biallelic mutations, but the pathogenicity of some of these mutations was not evaluated(3). The diagnostic rate of our study (12.7%, 8/63) is slightly lower than the sporadic study carried out by Yang et al. in the Chinese Hans population. The low diagnostic rate of our study is likely due to the following reasons: 1) more than half of the probands recruited were mild-to-moderate HL patients, whereas the probands in the study carried out by Yang et al. were all severe-to-profound patients. Additionally, the rate of genetic causes of mild-to-moderate HL is lower than severe-to-profound patients; 2) in order to avoid false diagnosis, we used a stricter standard to classify a variant as pathogenic, and this might have sacrificed some true pathogenic mutations; 3) we postulated an autosomal recessive inheritance pattern for sporadic HL and only focused on biallelic variants, which may have missed some variants inherited in autosomal dominant pattern, such as de novo mutations; 4) due to the imbalance of capturing efficiency, variants may have been missed or filtered out as a result of insufficient coverage or low quality; and 5) some pathogenic variants might be overlooked due to the drawbacks of current mutation analysis strategy and sequencing technology, for example, intronic variants and variants in the exonic splicing enhancers that might affect normal splicing, CNVs involving only one
This article is protected by copyright. All rights reserved
or two exons. Based on our results and previous studies, we speculate that using targeted MPS, the diagnostic rate of sporadic NSHL would be 10 to 20%. In order to improve the diagnostic rate, we need to carry out functional studies on variants of uncertain significance and improve enrichment and sequencing technologies. Additionally, synonymous variants and intronic variants need to be investigated for their impact on splicing. In summary, we identified 10 novel pathogenic mutations in sporadic NSHL patients using TGE and MPS. Our data showed that in 12.7% of the sporadic NSHL patients excluded for mutations in GJB2, MT-RNR1, and SLC26A4 genes, pathogenic mutations can be identified in the other known HL genes. Our results will help us to improve the genetic diagnosis of sporadic NSHL patients.
References 1.
Nance WE. The genetics of deafness. Mental retardation and developmental disabilities research reviews
2003: 9: 109-119. 2.
Brownstein Z, Friedman LM, Shahin H et al. Targeted genomic capture and massively parallel sequencing
to identify genes for hereditary hearing loss in Middle Eastern families. Genome biology 2011: 12: R89. 3.
Shearer AE, Black-Ziegelbein EA, Hildebrand MS et al. Advancing genetic testing for deafness with
genomic technology. Journal of medical genetics 2013: 50: 627-634. 4.
Shearer AE, DeLuca AP, Hildebrand MS et al. Comprehensive genetic testing for hereditary hearing loss
using massively parallel sequencing. Proceedings of the National Academy of Sciences of the United States of America 2010: 107: 21104-21109. 5.
Yang T, Wei X, Chai Y et al. Genetic etiology study of the non-syndromic deafness in Chinese Hans by
targeted next-generation sequencing. Orphanet journal of rare diseases 2013: 8: 85. 6.
Mutai H, Suzuki N, Shimizu A et al. Diverse spectrum of rare deafness genes underlies early-childhood
hearing loss in Japanese patients: a cross-sectional, multi-center next-generation sequencing study. Orphanet journal of rare diseases 2013: 8: 172. 7.
Choi BY, Park G, Gim J et al. Diagnostic application of targeted resequencing for familial nonsyndromic
hearing loss. PloS one 2013: 8: e68692. 8.
Baek JI, Oh SK, Kim DB et al. Targeted massive parallel sequencing: the effective detection of novel
causative mutations associated with hearing loss in small families. Orphanet journal of rare diseases 2012: 7: 60. 9.
Kitajiri S, Sakamoto T, Belyantseva IA et al. Actin-bundling protein TRIOBP forms resilient rootlets of hair
cell stereocilia essential for hearing. Cell 2010: 141: 786-798.
This article is protected by copyright. All rights reserved
Legends Figure 1. Overview of candidate variant filtering strategy. BWA, Burrows-Wheeler Aligner; GATK, Genome Analysis Toolkit; MAF, minor allele frequency.
This article is protected by copyright. All rights reserved
Figure 2. Audiogramss of diagnosed probands. Right R ear dennoted with “O O”, left ear den noted with “X X”.
This articlee is protected by b copyright. Alll rights reserve ed
Table 1. Biallelic variants identified by targeted massively parallel sequencing in 63 sporadic NSHL patients. aD: Deleterious; T: Tolerated. bD: Deleterious; N: Neutral. cD: Probably damaging; P: possibly damaging; B: benign. dD: disease causing; B: benign
This article is protected by copyright. All rights reserved