STRC Gene Mutations, Mainly Large Deletions, Are a Very Important Cause of Early-Onset Hereditary Hearing Loss in the Czech Population

GENETIC TESTING AND MOLECULAR BIOMARKERS Volume 22, Number 2, 2018 ª Mary Ann Liebert, Inc. Pp. 127–134 DOI: 10.1089/gtmb.2017.0155

STRC Gene Mutations, Mainly Large Deletions, are a Very Important Cause of Early-Onset Hereditary Hearing Loss in the Czech Population

Simona Poisson Markova´,1 Dana Sˇ afka Brozˇkova´,1 Petra Lasˇsˇuthova´,1 Anna Me´sza´rosova´,1 Marcela Kru˚tova´,1 Jana Neupauerova´,1 Dagmar Rasˇkova´,2 Marie Trkova´,2 David Staneˇk,1 and Pavel Seeman1

Introduction: Hearing loss (HL) is the most common sensory deficit in humans. HL is an extremely heterogeneous condition presenting most frequently as a nonsyndromic (NS) condition inherited in an autosomal recessive (AR) pattern, termed DFNB. Mutations affecting the STRC gene cause DFNB type 16. Various types of mutations within the STRC gene have been reported from the U.S. and German populations, but no infor- mation about the relative contribution of STRC mutations to NSHL-AR among Czech patients is available. Methods and Patients: Two hundred and eighty-eight patients with prelingual NSHL, either sporadic (n = 207) or AR (n = 81), who had been previously tested negative for the mutations affecting the GJB2 gene, were included in the study. These patients were tested for STRC mutations by a quantitative comparative fluorescent polymerase chain reaction (QF-PCR) assay. In addition, 31 of the 81 NSHL-AR patients were analyzed by massively parallel sequencing using one of two different gene panels: 23 patients were analyzed by multiplex- ligation probe amplification (MLPA); and 9 patients by SNP microarrays. Results: Causal mutations affecting the STRC gene (including copy number variations [CNVs] and point mutations) were found in 5.5% of all patients and 13.6% of the 81 patients in the subgroup with NSHL-AR. Conclusion: Our results provide strong evidence that STRC gene mutations are an important cause of NSHL- AR in Czech HL patients and are probably the second most common cause of DFNB. Large CNVs were more frequent than point mutations and it is reasonable to test them first by a QF-PCR method—a simple, accessible, and efficient tool for STRC CNV detection, which can be combined by MLPA.

Keywords: NSHL, DFNB, STRC, CNV, hearing loss, deafness

Introduction 2001, 2011). Mutations of the STRC gene have been described mainly in association with mild-to-moderate HL (Francey

Downloaded by Faculty of Medicine at Charles University from online.liebertpub.com 02/21/18. For personal use only. earing loss (HL) is the most common sensory deﬁcit; et al., 2012; Moteki et al.,2016). Hits prevalence among newborns ranges from 1:500 to The STRC gene is located on chromosome 15q15.3. This 1:1000 (Fortnum and Davis, 1997; Fortnum et al., 2001; chromosomal region is highly complex due to a large seg- Morton and Nance, 2006). In more than 50% of cases, HL mental duplication involving three other genes, HISPPD2A is genetically determined, being an extremely genetically (MIM: 610979), CATSPER2 (MIM: 607249), and CKMT1A heterogeneous condition. Nonsyndromic autosomal-recessive (MIM: 613415), whose pseudogenes are (beside CKMT1A) hearing loss (NSHL-AR) is the most frequent one, and is also inactive as well (Knijnenburg et al.,2009).Furthermore,the known as DFNB (Marazita et al., 1993). homology between STRC gene and pSTRC pseudogene is al- The STRC gene (MIM: 606440) encodes a highly conserved most 100% (99.6% of the coding areas and 98.9% with intronic protein called Stereocilin, which is necessary for proper hair areas included) (Francey et al., 2012). pSTRC is inactive due to cell function and is defective in DFNB type 16 (Verpy et al., a nonsense mutation in exon 20 (Verpy et al., 2001).

1DNA Laboratory, Department of Paediatric Neurology, 2nd Faculty of Medicine, Motol University Hospital, Charles University in Prague, Prague, Czech Republic. 2Gennet, Prague, Czech Republic.

127 128 POISSON MARKOVA´ ET AL.

Therefore a comprehensive diagnostics for STRC muta- homozygous, but also in heterozygous state. However, fur- tions has been very challenging. Copy number variations ther verification is needed to confirm its validity and differ- (CNVs), namely deletions, usually affect the entire or a large entiate whether the identified variants originate from the gene part of STRC gene, but also point mutations, small deletions, or from the pseudogene. In regions with such high homology insertions, and gene conversions were reported (Deafness and complexity we cannot rely on position of markers nor on Variation Database, HGMD Professional). allele frequencies available in public databases. The STRC deletions frequently involve also the CATSPER2 For SNVs, further differentiation, for example, Long gene, which is associated with male infertility. This syndrome Range polymerase chain reaction (LR-PCR) to obtain the is known as deafness infertility syndrome (DIS–MIM: 611102). gene specific product, without contamination by product Recent reports from United States have shown that muta- from the pseudogene. LR-PCR product is then used as tem- tions affecting the STRC gene contribute to 16.1% of all di- plate for a nested PCR (as described e.g., in Vona et al., agnosed cases of patients presenting genetic HL, counting for 2015). This allows specific amplification and sequencing of nearly 20% in the cohort of patients with the AR inheritance only specific gene regions, with excluded pseudogene re- pattern in these countries. In the U.S. population it seems to gions. CNVs can be tested or verified by, for example, SNP be the second most frequent cause of hereditary HL, after microarray (Microarray SNP chip), multiplex-ligation probe mutations in GJB2 (DFNB1), especially in Caucasians and amplification (MLPA), or—as the method of first choice used Hispanics (Sloan-Heggen et al., 2016). In Germany, ob- in this article—quantitative comparative fluorescent poly- served rates of causal mutations of STRC accounted for about merase chain reaction (QF–PCR). 6% of patients in GJB2 negative patients (Vona et al., 2015). In other European countries, mutations of STRC were as well observed (e.g., Sommen et al., 2016). Patients and Methods Bioinformatic analysis of data from massively parallel Two hundred and eighty-eight patients with nonsyndromic sequencing (MPS) can be used for detection of single nu- early-onset, presumed autosomal recessive (dominant in- cleotide variants (SNVs) and due to advances in bioinfor- heritance pattern excluded) HL who were previously tested matic analysis also for the presence of larger CNVs, mainly in negative for pathogenic mutations in the coding part of the Downloaded by Faculty of Medicine at Charles University from online.liebertpub.com 02/21/18. For personal use only.

FIG. 1. Diagram of STRC mutation identification process. CNV, copy number variation; HL, hearing loss; MLPA, multiplex-ligation probe amplification; QF-PCR, quantitative comparative fluorescent polymerase chain reaction; SNV, single nucleotide variant. STRC IS IMPORTANT CAUSE OF HEARING LOSS IN CZECHIA 129

GJB2 gene and also for the noncoding mutation c.-23 + 1G>A a were included in the study. Patients with recognized syndromic HL were excluded. In 81 of these patients, the occurrence of deafness was apparently autosomal recessive— familial (with at least one sibling also affected by a hearing impairment or the family pedigree otherwise strongly suggested autosomal recessive inheritance). The occurrence of early HL in the remaining group of 207 patients was sporadic or not speciﬁed. All patients or responsible adult signed in- formed consent for genetic testing of early-onset hereditary HL, and study of ethical standards complies with the current laws of the country of origin. DNA was isolated from blood

by classical isolation methods. Ready to use PCR master GAAATTCGAGACCACCCTGA mixes from Top-Bio (Top-Bio, Prague, Czech Republic) were used for PCRs. in Selected Methods Five methods (QF-PCR, SNP microarray/Microarray SNP a

chip/, MLPA, panel based MPS, Sanger sequencing) were Other details used for detection and conﬁrmation of various types of mu- 19) tations affecting the STRC gene. The tested patients were hg divided into two subgroups—the mutation identiﬁcation process is visualized in Figure 1. Regions used for CNV detection in STRC gene by QF-PCR,

MLPA, and SNP microarray/Microarray SNP chip/ are sum- NM153700, marized in Table 1 and schematically visualized in Figure 2. dbSNP Nucleotide change Forward primer Reverse primer

QF-PCR for CNV detection Gene ( Regions with different lengths in intron 18 and intron 26

between STRC and pSTRC sequences were selected using STRC ClustalX (Clustal: Multiple Sequence Alignment) and Meld (GNOME) software to compare gene and pseudogene sequences. Furthermore, PCR primers for each region with different lengths were designed to sequences with 100% STRC and pSTRC homology; therefore, they amplify both regions of the gene STRC and of the pseudogene pSTRC simultaneously and under identical conditions for subsequent relative quantification (see primers in Table 1, see Fig. 3). Forward primers were marked by fluorescein dye (fluorescein amidite [FAM]). The PCR resulted in two products with different lengths: intron 18 (gene: 267 bp, pseudogene: 264 bp) and intron 26 (gene: 104 bp, pseudogene: 106 bp). PCR conditions in detail are mentioned in Supplementary file 1 (Supplementary Data are available online at www.liebertpub .com/gtmb). PCR products were further processed by fragment analysis on an automated Genetic Analyzer ABI 3130 (Applied Bio- systems) and analyzed using GeneMapper (Applied Biosys- Downloaded by Faculty of Medicine at Charles University from online.liebertpub.com 02/21/18. For personal use only. tems) software and RStudio (RStudio, Inc.). Intron 26Exon 23Intron 18Intron chr15:43892580-43892683 25Intron 19 chr15:43895374-43895479 ATGGGTGGACTGGATGGAAG chr15:43897648-43897714 chr15:43892964-43893072 chr15:43897314-43897383 Intron 23 GCCCTATCATAGACCTTCCCC Intron 22 chr15:43894493 chr15:43895643 rs12050645 rs2260160 T/C T/C DNA samples from all 288 patients were analyzed by this Exon/intron Areas included method.

Panel-based MPS Regions Used for Copy Number Variation Detection in

Libraries were prepared using targeted enrichment meth- 1. ods (ﬁrst design: Agilent HaloPlex: 46 NSHL-AR genes—21 patients; second design: Agilent HaloPlex HS: 71 NSHL-AR Table genes—10 patients) and sequenced on Illumina MiSeq NGS

sequencer. (2015). The data were analyzed, including alignment and variant

calling, using combination of NextGENe (SoftGenetics) et al. Galaxy and SureCall (Agilent). Annotation of variants was

done using Alamut Batch for assembly hg19 with access to Vona MLPA, multiplex-ligation probe ampliﬁcation; QF-PCR, quantitative comparative ﬂuorescent polymerase chain reaction. a QF-PCR Intron 18 chr15:43898589-43898855 CCTCTGATTTCGGGTAAAAGG Human Gene Mutation Database (HGMD Professional) and MLPA Kit P461 DIS A1-0616 Exon 24HumanOmniExpress-24v1-1 chr15:43893565-43893640 Intron 23 chr15:43893818 rs2447196 A/G 130 POISSON MARKOVA´ ET AL.

FIG. 2. Schematic visualization of regions tested by QF-PCR, MLPA, and microarray.

public databases (e.g., Exome Aggregation Consortium/ One sample was analyzed using the first design and is in- ExAC/and 1000Genomes). cluded as well in the cohort of 21 patients mentioned above. CNV analysis of MPS sequencing data was done using NextGENe software, which includes a model for CNV Microarray and MLPA analysis of MPS data (CNV batch and Dispersion and Hidden Markov Model—Normalized counts). To validate our results obtained by the QF-PCR by an DNA samples from patients with identified suspicious independent method, nine samples with identified CNVs variants in STRC considered as pathogenic or likely patho- were further analyzed using a microarray SNP (Illumina In- genic were further verified by Sanger sequencing method, finium OmniExpress-24v.1.1). using LR-PCR and Nested PCR (as described in Vona et al., Further 23 samples were analyzed by MLPA (MRC- 2015—with slight variation of chemistry/Top-Bio/and PCR Holland P461 DIS A1-0616). conditions used for LR—consult Supplementary file 1) to The selection of these samples for analyses was partly exclude pSTRC contamination. Final PCR products were influenced by DNA quantity limitations. sequenced with the BigDye Terminator v3.1 Kit (Applied Biosystems) and analyzed on the ABI 3130 Genetic Analyzer Results (Applied Biosystems). The cause of HL/deafness was attributed to the STRC gene in The MPS data analysis was divided in two groups as 16 (5.5%) samples of 288, by use of different methods and their follows: combination. In the subgroup of patients with familial occur- (1) Data from 21 patients with familial occurrence of HL rence of deafness, the attribution of deafness to STRC gene was Samples were analyzed using Agilent HaloPlex target substantially higher—in 11 (13.58%) samples from 81 tested. enrichment method of 46 genes. These patients were selected For detailed summary of results from all three methods without previous knowledge of CNV STRC. combined, see Supplementary file 2. (2) Data from 11 patients with heterozygous STRC gene deletion, as detected by QF-PCR Patients tested by QF-PCR method, whose results were QF-PCR for CNV detection interpreted as a heterozygous deletion of the STRC gene, Classification criteria. The ratios between the peaks were further examined for the presence of point mutations— height from the gene and from the pseudogene guided the SNVs on the other/nondeleted allele of using a gene panel result classification (Table 2, Fig. 3). target enrichment and MPS (as described above). Ten sam- Six classification categories were created as follows: ples were prepared using second design of the gene panel. ‘‘Homozygous gene deletion,’’ ‘‘Heterozygous gene deletion,’’ Downloaded by Faculty of Medicine at Charles University from online.liebertpub.com 02/21/18. For personal use only.

FIG. 3. Examples of ‘‘normal’’ homozygous gene deletion and heterozygous gene deletion visualized from QF-PCR results. STRC IS IMPORTANT CAUSE OF HEARING LOSS IN CZECHIA 131

Table 2. Criteria Used for QF-PCR Result Classiﬁcation Region Intron 18 Intron 26

Classiﬁcation categories Peak2 (size 267 bp)/peak1 (size 264 bp) Peak1 (size 104 bp)/peak2 (size 106 bp)

Peak ratio interval classiﬁcation criteria ‘‘Normal’’ 0.82–1.2 0.82–1.2 Homozygous gene deletion 0 0 Heterozygous gene deletion 0.4–0.6 0.4–0.6 Homozygous pseudogene deletion Null Null Heterozygous pseudogene deletion 1.78–2.2 1.78–2.2 Not classiﬁed Any other result Any other result

‘‘Heterozygous pseudogene deletion,’’ ‘‘Homozygous pseu- ratios in a total of 19 (23.5%) samples. From those, four dogene deletion,’’ ‘‘Normal,’’ and ‘‘Not classifiable.’’ In- (4.9%) samples were classified into the homozygous and tervals for peak ratio criteria classification were established eight (9.9%) samples into the heterozygous gene deletion during validation (Table 2). category. One sample (1.2%) was classified into the category To validate this classification approach by an independent homozygous pseudogene deletion. Six (7.4%) samples could method, MLPA was performed on 23 selected samples. The not be classified. results (Supplementary file 2) were not contradictory to the In the remaining 62 (76.5%) samples, the ratios of the classification by QF-PCR. But due to very different lengths peaks were classified as ‘‘Normal.’’ The results are summa- and positions of the deletion affecting the STRC gene and also rized in Table 3. due to different regions of STRC which are used for testing by each of the methods used, the results must not be necessarily Panel-based MPS identical and both can be true. To classify the samples in the categories ‘‘Homozygous deletion’’ and ‘‘Heterozygous de- (1) Data of 21 patients with familial occurrence of HL, letion,’’ even one region (intron 18 or intron 26) was suffi- randomly selected cient to meet the category classification criteria. Sequence alterations (SNVs) affecting the STRC gene were detected in three (14.3%) of 21 patients (Table 4). In Result summary. In 46 of 288 samples analyzed (16%), the sample 38F, two probably pathogenic SNVs were de- the peak ratio was not classified as ‘‘Normal’’ (we further tected. In two samples, gene deletions were detected using address them as ‘‘Abnormal’’). CNV bioinformatic analysis. In the sample 6F one probably Of these, eight (2.8%) samples were classified as ‘‘Homo- pathogenic SNV and the heterozygous gene deletion were zygous gene deletion’’—due to the absence of at least one of detected (and confirmed by MLPA). In the sample 4F, two the gene specific peaks/products, and 16 (5.6%) were clas- overlapping large deletions of different sizes were detected sified as ‘‘Heterozygous gene deletion.’’ One sample (0.35%) (and confirmed on the Microarray and by MLPA). was classified into the category ‘‘Probably heterozygous pseu- In addition, in the sample from the affected sibling of the dogene deletion’’ and one sample (0.35%) into the category patient 38F, we confirmed the presence of both mutations in ‘‘Probably homozygous pseudogene deletion.’’ The peak ratios the heterozygous state by Sanger sequencing of the STRC of 20 (6.9%) samples were not classified. gene-specific LR-PCR product. In the remaining 242 (84%) samples, the peak ratios were (2) Data from 11 patients with altered STRC and pSTRC classified as ‘‘Normal.’’ regions detected by QF-PCR In the subgroup of 81 patients with HL with familial In a subgroup of 11 patients classified as ‘‘Heterozygous deafness occurrence, we have detected ‘‘Abnormal’’ peak gene deletion,’’ the second likely pathogenic mutation

Downloaded by Faculty of Medicine at Charles University from online.liebertpub.com 02/21/18. For personal use only. Table 3. Summary of QF-PCR Results Patients with Patients with ‘‘familiar’’ ‘‘sporadic’’ deafness deafness QF-PCR results summarized Percentage occurrence Percentage occurrence Percentage Total number of tested patients 288 81 207 Category Homozygous gene deletions 8 2.78 4 4.94 4 1.93 Heterozygous gene deletions 16 5.56 8 9.88 8 3.86 Homozygous pseudogene deletions 1 0.35 1 1.23 0 0.00 Heterozygous pseudogene deletions 1 0.35 0 0.00 1 0.48 Not classiﬁed 20 6.94 6 7.41 14 6.76 Totally ‘‘Abnormal’’ 46 15.97 19 23.46 27 13.04 ‘‘Normal’’ 242 84.03 62 76.54 180 86.96 132 POISSON MARKOVA´ ET AL.

Table 4. Summary of Detected Mutations in Three Clariﬁed Samples Using Panel-Based Massively Parallel Sequencing

Sample ID First allele Second allele 38F c.3217C>T: p.R1073* (class 5) c.4402C>T: p.R1468* 6F c.3217C>T: p.R1073* (class 5) Deletion: Chr15:43891845-43903207 4F Deletion: Chr15:43891845-43897622 Deletion: Chr15:43891845-44010407

affecting the opposite (undeleted) allele was detected in se- with previously excluded mutations in the GJB2 gene. In the ven patients (64%). In four patients, no further mutation has subgroup of patients with familial occurrence of HL the yet been identified in the coding sequence of the STRC gene percentage was even significantly higher—23.46%. and these patients or at least some of them may be only Evaluation of the QF-PCR results was based on compar- heterozygous carriers of the STRC deletion by chance. The ative quantitation, namely on ratios between peaks from PCR detection percentage for pathogenic or causal mutations was products from STRC gene and STRC pseudogene, while the higher in the group with familial deafness occurrence (as group intervals for the diagnostics were established during summarized in Table 5). validation of this method. To verify its applicability, MLPA Detected STRC variants are summarized in Table 6. analysis was performed on 23 selected samples (where In 10 from 11 tested patients, data of CNV bioinformatic DNA quantity allowed it). As the results corresponded to analysis using NextGENe were suggestive for STRC het- our classification, we believe that this classification is ap- erozygous gene deletion. plicable in laboratory practice. In 20 samples, the ratio of QF-PCR peaks’ height did not Microarray and MLPA allow a straight forward classification (being in the category ‘‘Not classified’’). As the region is a part of a large segmental In the overlapping tested regions, the results of microarray duplication involving three other genes and also frequent and MLPA corresponded or were not contradictory to the recombinations, these ratios may be indicative for STRC/ classification of QF-PCR (Supplementary file 2). pSTRC duplication, double duplication, or even gene conversions and recombinations. These possibilities are probably Discussion indistinguishable by QF-PCR, and ‘‘Not classified’’ results Increasing attention has been paid to the STRC gene as a need further analysis by other method or combination of contributor to hereditary HL during the last few years. The other methods and also by testing of other relatives of these complexity of the chromosomal region and the STRC gene patients. locus and the presence of the almost identical pseudogene When the peak height ratios were classified as ‘‘Hetero- hindered wider routine diagnostic testing of this gene in the zygous gene deletion,’’ panel-based MPS was performed past and are challenging for every molecular genetic method. when possible. In 7 of the 11 samples analyzed (64%), a Several approaches to its comprehensive testing have second probably pathogenic mutation has been identified on been recently suggested for routine diagnostic practice (e.g., the opposite (undeleted) allele. The second detected muta- Mandelker et al., 2014). tion was mostly truncating inactivating mutations predicted Our results provide solid evidence that mutations of to be loss of function alleles, in two cases, stop mutations, STRC are a very important cause of HL in the Czech pop- and another (c.875 + 1G>A: p.?) caused by the loss of the ulation and may be the second most frequent cause of early- second copy of the gene due to splicing alterations. This is onset hereditary HL after the GJB2 gene mutations. A supporting the causality of these variants in combination of similar situation may exist in the surrounding Central Eu- the STRC deletion on the other allele. For the remaining four ropean countries. mutations, the possible mechanism causing the HL is un- Using the QF-PCR method, we have managed to identify known. ‘‘Abnormal’’ alterations in peaks with PCR products from In total, from all detected STRC sequence alterations Downloaded by Faculty of Medicine at Charles University from online.liebertpub.com 02/21/18. For personal use only. STRC or pSTRC regions in 46 (15.94%) of tested patients (point mutations), p.Arg1073* was the most frequent and was

Table 5. Detection Percentage of Second Pathogenic or Causal Mutations in Patients With Heterozygous STRC Gene Deletion Patients with Patients Total number ‘‘familiar’’ with ‘‘sporadic’’ MPS of patients with gene of tested deafness deafness heterozygous deletion patients: Percentage occurrence Percentage occurrence Percentage Patients with heterozygous 11 7 4 gene deletion Identiﬁed second mutation 7 63.64 6 85.71 1 25.00 Nonidentiﬁed cause of HL 2 18.18 0 0.00 2 50.00 Other cause of HL 2 18.18 1 14.29 1 25.00

HL, hearing loss; MPS, massively parallel sequencing. STRC IS IMPORTANT CAUSE OF HEARING LOSS IN CZECHIA 133

Table 6. Summary of Detected Second Pathogenic/Probably Pathogenic Mutations in Patients with Heterozygous Gene Deletion (Class 5: Pathogenic Mutation, Class 4: Probably Pathogenic Variant) Sample ID First allele Second allele Commentary 5F Heterozygous gene deletion c.2171_2174del p.V724G*6 (class 5) rs786200883, described as pathogenic 6F, 10F Heterozygous gene deletion c.3217C>T: p.R1073* (class 5) Described as pathogenic in ClinVar database, no rs so far attributed. 7F Heterozygous gene deletion c.2784G>T: p.R928S (class 4) Novel 9F Heterozygous gene deletion c.3275G>A: p.C1092Y (class 4) rs727503447 (0.07% in NFE) 11F Heterozygous gene deletion c.4517 T>C: p.L1506P (class 4) Novel 10S Heterozygous gene deletion c.875 + 1G>A: p.? (class 4) Novel predicted alteration of splicing and exon skip

Variant classiﬁcation is according to ACMG criteria (Richards et al., 2015).

detected in three (Sample ID: 6F, 10F, 38 F) out of eight Conclusion patients. This point mutation might be prevalent in the Czech/ Mutations in STRC, mostly CNVs, are a very important Central European population and is not yet reported in the cause of HL in Czech patients, and our results support the HGMD database; however, it is already submitted as patho- evidence of its importance as observed recently in some other genic in ClinVar. countries. The validation of the QF-PCR method results by inde- QF-PCR is a simple, very fast, and cost-effective method pendent methods (MLPA, microarray) supports its validity; for screening of larger STRC deletions affecting the gene however, its limitations have been as well revealed. Due to from intron 18 to intron 26. In the case where the peak ratio is the very different extent of the deletions, other CNVs, and classified as a ‘‘Homozygous gene deletion,’’ the cause of HL also due to the difference of the sites used for testing by each is attributable to defective STRC. of these methods, the direct comparison of these methods is In the case of classification into the ‘‘Heterozygous gene not possible and there is no gold standard method for CNV deletion’’ category, the search for a second causal mutation STRC detection affecting the gene. However, the use of dif- (SNV) should be performed, either by MPS or classical ferent methods detecting CNV in combination is very useful, Sanger sequencing methods. For better reliability and for and we recommend it for better reliability. Due to the size of better estimation of the size of the CNVs within the altered the region, it is better to use also the MLPA method, which region, a combination of more quantitative methods is crucial CATSPER2, also tests the areas in the surrounding genes ( and recommended. CKMT-1B). Moreover, the CNVs, which concern gene duplication, seem to be easier to evaluate and to give more robust results. Acknowledgments From all tested patients, we attribute the cause of he- This study was supported by GAUK 1216214 and MH CR reditary HL to STRC gene in 5.5% of all and to 13.58% of AZV 16-31173A. familial cases. These results closely correspond to the frequencies observed in Germany (Vona et al., 2015). QF-PCR Author Disclosure Statement however does not detect SNV or small deletions and insertions and not all of the 288 patients were tested by MPS No competing financial interests exist. for STRC point mutations, so the final frequency may even be higher. References Downloaded by Faculty of Medicine at Charles University from online.liebertpub.com 02/21/18. For personal use only. As the bioinformatic prediction of deletions nearly always corresponded to the verified result, the use of MPS is an Fortnum H, Davis A (1997) Epidemiology of permanent effective tool not only for the detection of sequence alter- childhood hearing impairment in Trent Region, 1985–1993. ations but also for the detection of STRC deletions—it can Br J Audiol 31:409–446. detect both SNVs and CNVs. However, it still remains rela- Fortnum HM, Summerfield AQ, Marshall DH, et al. (2001) Prevalence of permanent childhood hearing impairment in the tively expensive and requires advanced bioinformatic and United Kingdom and implications for universal neonatal technical equipment and support. Hopefully MPS will be hearing screening: questionnaire based ascertainment study. accessible to every patient with suspected hereditary HL after BMJ 323:536–540. the GJB2 gene testing in the future. Francey LJ, Conlin LK, Kadesch HE, et al. (2012) Genome- Until this time, especially in the Caucasian and Hispanic wide SNP genotyping identifies the Stereocilin (STRC) gene populations, QF-PCR can be used as a method of pre- as a major contributor to pediatric bilateral sensorineural screening, in patients with early HL and family history hearing impairment. Am J Med Genet A 158A:298–308. compatible with autosomal recessive inheritance. This can Knijnenburg J, Oberstein SA, Frei K, et al. (2009) A homo- help to identify the cause of early-onset hereditary HL in as zygous deletion of a normal variation locus in a patient with much as 6% of patients with already excluded mutations of hearing loss from non-consanguineous parents. J Med Genet GJB2 gene. 46:412–417. 134 POISSON MARKOVA´ ET AL.

Mandelker D, Amr SS, Pugh T, et al. (2014) Comprehensive cing approach combined with a mutation classiﬁcation sys- diagnostic testing for stereocilin: an approach for analyzing tem. Hum Mutat 37:812–819. medically important genes with high homology. JMD 16: Verpy E, Leibovici M, Michalski N, et al. (2011) Stereocilin 639–647. connects outer hair cell stereocilia to one another and to the Marazita ML, Ploughman LM, Rawlings B, et al. (1993) tectorial membrane. J Comp Neurol 519:194–210. Genetic epidemiological studies of early-onset deafness in Verpy E, Masmoudi S, Zwaenepoel I, et al. (2001) Mutations in a the U.S. school-age population. Am J Med Genet 46:486– new gene encoding a protein of the hair bundle cause non- 491. syndromic deafness at the DFNB16 locus. Nat genet 29:345–349. Morton CC, Nance WE (2006) Newborn hearing screening—a Vona B, Hofrichter MAH, Neuner C, et al. (2015) DFNB16 is a silent revolution. N Eng J Med 354:2151–2164. frequent cause of congenital hearing impairment: implementa- Moteki H, Azaiez H, Sloan-Heggen CM, et al. (2016) Detection tion of STRC mutation analysis in routine diagnostics. Clin and conﬁrmation of deafness-causing copy number variations Genet 87:49–55. in the STRC gene by massively parallel sequencing and comparative genomic hybridization. Ann Otol Rhinol Lar- Address correspondence to: yngol 125:918–923. Simona Poisson Markova´,MD Richards S, Aziz N, Bale S, et al. (2015) Standards and DNA Laboratory guidelines for the interpretation of sequence variants: a joint Department of Paediatric Neurology consensus recommendation of the American College of 2nd Faculty of Medicine Medical Genetics and Genomics and the Association for Motol University Hospital Molecular Pathology. Genet Med 17:405–424. Charles University in Prague Sloan-Heggen CM, Bierer AO, Shearer AE, et al. (2016) Prague 150 06 Comprehensive genetic testing in the clinical evaluation of Czech Republic 1119 patients with hearing loss. Hum Genet 135:441–450. Sommen M, Schrawen I, Vandeweyer G, et al. (2016) DNA E-mail: [email protected]; diagnostics of hereditary hearing loss: a targeted resequen- [email protected] Downloaded by Faculty of Medicine at Charles University from online.liebertpub.com 02/21/18. For personal use only.