A novel missense variant in ESRRB gene causing autosomal recessive non-syndromic hearing loss: in silico analysis of a case

Hereditary hearing loss (HHL) is a common heterogeneous disorder affecting all ages, ethnicities, and genders. The most common form of HHL is autosomal recessive non-syndromic hearing loss (ARNSHL), in which there is no genotype–phenotype correlation in the majority of cases. This study aimed to identify the genetic causes of hearing loss (HL) in a family with Iranian Azeri Turkish ethnicity negative for gap junction beta-2 (GJB2), gap junction beta-6 (GJB6), and mitochondrially encoded 12S rRNA (MT-RNR1) deleterious mutations. Targeted genome sequencing method was applied to detect genetic causes of HL in the family. Sanger sequencing was employed to verify the segregation of the variant. Finally, we used bioinformatics tools and American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) guidelines to determine whether the detected variant might affect the corresponding protein or not. A novel homozygous missense mutation, c.499G>A (p.G167R), was identified in exon 5 of the ESRRB (estrogen-related receptor beta) gene. Healthy and affected family members confirmed the co-segregation of the variant with ARNSHL. Eventually, the variant's pathogenicity was confirmed by the in silico analysis and the ACMG/AMP guidelines. The study suggests that the detected variant, c.499G>A, plays a crucial role in the development of ARNSHL, emphasizing the clinical significance of the ESRRB gene in ARNSHL patients. Additionally, it would be helpful for genetic counseling and clinical management of ARNSHL patients and providing preventive opportunities.

The limited phenotypic variability and high heterogeneity of NSHL make it particularly difficult to diagnose by routine detection methods [5]. Recent advances in diagnostic methods such as next-generation sequencing (NGS) have opened a new window for heterogenic disorders; however, the production of thousands of variants per individual is the biggest challenge of NGS technology. Especially when directing clinical care, there is a need for correct interpretation and ethnic-specific filtering of test results to achieve precise outcomes [17][18][19][20]. In this light, this study aimed to employ the targeted sequencing of 127 known HL-causing genes panel in the proband of a family of Iranian Azeri Turkish ethnicity with consanguine marriage negative for GJB2, GJB6, and mitochondrially encoded 12S rRNA (MT-RNR1) mutations.

Participants
An affected Iranian individual of Azeri Turkish ethnicity, whose parents had a consanguineous marriage (Fig. 1A), was recruited from Ebnsina Medical Genetic Laboratory of Tabriz. The proband originated from Tabriz, a city in the northwest of Iran with majorly Azeri Turkish ethnicity [21]. Physical examination, audiological evaluation, clinical questionnaires on age, exposure to environmental factors, and history of other diseases were administered when the patient was being tested for GJB2, GJB6, and MT-RNR1 (A1555G) mutations. To evaluate the detected variant in the control population, 200 blood samples were collected in the same region (Tabriz). The majority (64%) of this population were male, with a mean age of 36 years and no history of HLL. Full informed consent was obtained from the participants in the study. The Ethics Committee of Tabriz University of Medical Sciences approved the study and methodology for investigating humans with ethical code No. 94/2-7/13.

Targeted genomic capturing and next-generation sequencing
Genomic DNA was extracted from 3 ml peripheral blood of the patient using a DNA Extraction Mini Kit (Favor-Prep, Taiwan, Cat. No.: FATGK001). Then, the quality and quantity of the DNA were determined with a spectrophotometer and a 1.2% agarose gel. We sent the DNA sample to the Beijing Genomics Institute (BGI-Clinical Laboratories, Shenzhen, China) to perform the targeted sequencing. The sequencing test was conducted using a custom-designed Nimblegen chip capturing the 127 targeted genes, followed by next-generation sequencing. The test platform examined > 95% of the target gene with a sensitivity of > 99%. Point mutation, micro-insertion, deletion, and duplication (< 20 bp) could be simultaneously detected. Illumina Base Calling software (bcl2fastq) was used to analyze the raw data. Human genome assembly hg19 (GRCh37) was used as the reference sequence. A single nucleotide variant detection toolkit (GATK v3.3.0) was utilized to detect single nucleotide variants (SNV/INDEL), and then the variants were annotated using ANNOVAR software. In addition, all variants were then filtered according to the mutations' type, frequency, and pathogenicity.

Variant confirmation by sanger sequencing
Primers surrounding the damaging variant were designed using Primer 3 software (forward: GCC ACC AGG CTA AAT CCT CTG and reverse: TTT CCA AGG TCA CCA GGC AGT) to evaluate the variant associated with the phenotype. After amplification by an Eppendorf thermocycler and using a SinaClon PCR Master Mix kit (Cat. No.: MM2011), Sanger sequencing was performed utilizing a Big Dye kit (Cat. No.: 4337455) and an Applied Biosystems instrument (3130xl Genetic Analyzer). Finally, Chromas software version 2.6.4 was used to observe the obtained sequences. In order to assess the detected variant in the control population, we extracted genomic DNA from the 200 blood samples using salting-out procedure. We then performed PCR amplification and Sanger sequencing using primers surrounding the candidate gene to obtain the sequences.
The effect of the mutation on the molecular level was analyzed by modeling software Yet Another Scientific Artificial Reality Application (YASARA) (http:// www. yasara. org/). The ESRRB protein model used for this analysis was according to the PDB file with an ID of 1lo1.

Clinical evaluations
In this family, both parents had normal hearing levels. A three-year-old boy was the proband (IV.3). In addition to him, both affected members of the family (V.5 and V.6) displayed bilateral severe to profound HL (Fig. 1a-c). As stated by parents, HL was congenital, and no other abnormality indicated syndromic HL. As much as remembered by the mother, there was no exposure to environmental factors, and no special disease would have made use of a particular medicine or antibiotic such as aminoglycosides.

Functional prediction and molecular modeling
The targeted genomic capturing in this family indicated a novel homozygous variant c.499G>A (p. Gly167Arg) in the ESRRB gene (estrogen-related receptor beta-NM_004452.1), submitted by our group with Ref SNP ID: rs155534214. The variant was co-segregated with the ARNSHL phenotype (Additional File 1) and was absent in 200 control individuals (data not shown). Based on the results of pathogenicity prediction, the c.499G>A variant was predicted as a damaging variant by all databases (Table 1), and more importantly, it had no frequency in population databases ( Table 1). The conservation evaluation of the 167th amino acid of the ESRRB protein indicated that glycine was a conserved amino acid ( Fig. 2A), where substituted arginine caused a large side chain in the protein and could destabilize it (Fig. 2B). Interactome analysis using the STRING database showed that the ESRRB protein interacted with the following proteins (

Discussion
The GJB2 mutations are the most common cause of deafness worldwide, including in Iran [7,22]. This gene is responsible for about 16-18% of ARNSHL cases in Iran; however, there is a significant gap in recognition of other responsible genes [23]. Considering the heterogeneity of ARNSHL and the high prevalence of HL combined with the high frequency of consanguineous marriages in Iran, it is crucial to identify any involvement of other responsible genes associated with this phenotype [23,24]. In this study, we used the targeted genomic capturing method and found a variant in the ESRRB gene causing ARNSHL.
The ESRRB gene, consisting of 11 exons, codes the ERRRB protein. It is well established that the defective ESRRB protein is strongly associated with HL in the mice model [25]. The ESRRB protein belongs to a family of orphan nuclear receptors with broad expression profiles, playing a generic role in regulating energy metabolism in mammals. In addition, this protein enhances self-renewal and reprogramming ability in cells, and a lack of ESRRB function has been shown to impair germ cell development [26]. This protein consists of two critical domains, including ligand-binding domain (LBD) and DNA-binding domain (DBD), which are recruited as the transcriptional regulators of estrogen target genes [25]. The LBD is located at the C-terminal of the ESRRB protein and is organized in a three-layer sandwich structure, containing 12 α-helices (H1-H12). This complex structure contributes to the structural stability and binding ability of the ESRRB protein and plays an essential role in the activation of transcription. The DBD is located in the N-terminal portion of the ESRRB protein, which contains a zinc finger domain and is responsible for DNA binding [27].
The ESRRB gene is the only known gene that acts as a transcription factor and is associated with ARNSHL, while other transcription factors are related to autosomal dominant non-syndromic hearing loss (ADNSHL) [28]. This gene is responsible for the recessive DFNB35 deafness forms. The DFNB35 locus was identified in 2003 when NSHL was being studied in a large Pakistani family [29]. There are currently more than 13 variants have been found in DFNB35, nearly half of which relates to Pakistani ethnicity (p.A110V, p.L320P, p.E340del, p.V342L, p.L347P, p.D245H) [25,30,31] and the rest belongs to the Czech Republic (p.R291L) [32], Tunisia (p.Y305H) [28], Turkey (p.V342GfsX44) [25], China (p.R382C, p.R6G) [33,34], the UAE (p.D353GfsX6) [35], and our discovered variant (p.G167R) in the Iranian population. Except for c.1156C>T [25] and c.1237G>A [32], which are reported as polymorphism, other variants are considered functional mutations that interfere with the function of the ESRRB protein ( Table 2). As shown in Fig. 4, most of the variants are accumulated in exon 8 of the ESRRB gene at the mRNA level, causing them to appear in the ligand-binding domain of the protein, as shown in Fig. 5. In this report, the sequence analysis of the ESRRB gene showed a missense variant c.499G>A, which occurred in exon 5 of the ESRRB gene and caused the substitution of glycine by arginine at position 167 in the DBD of the ESRRB-coded protein. This is the first mutation of the ESRRB gene identified in an Iranian ethnic group and can be considered a rare cause of ARNSHL in this population. The American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/ AMP) has published guidelines for the clinical interpretation of variants providing 28 criteria for classifying a variant as pathogenic/likely pathogenic (P/LP) or benign/ likely benign (B/LB) [36]. According to the ACMG/AMP guidelines, our discovered variant met the pathogenicity criteria and was considered "likely pathogenic" because it is a novel missense variant (PM5) within a functional domain (PM1). Additionally, this variant was absent in the controls (PM2), co-segregated in the pedigree (PP1), and computational analysis supports the damaging effects of the variant (PP3).
Multiple sequence alignment of the human ESRRB protein across other species showed that the 167th amino acid was located in the highly conserved position at DBD, and it seems that an amino acid change in this position interfered with the function of the ESRRB protein. Molecular modeling prediction revealed that substituting aspartic acid with a glycine residue would lead to a large side chain, possibly resulting in inappropriate interactions between residues. This abnormality is likely to destroy the 3D structure of the DBD and subsequently Table 1 Characteristics and the pathogenicity prediction of the identified variant The Polymorphism phenotyping v2 (PolyPhen-2) scores range from 0.0 (tolerated) to 1.0 (deleterious). The Sorting Intolerant from Tolerant (SIFT) values range from 0 to 1. The variant is predicted damaging if the score is ≤ 0.05, and tolerated if the score is > 0.05. The Functional Analysis through Hidden Markov Models (FATHMM) scores range from − 16.13 to 10.64. The smaller the score the more likely the variant has a damaging effect. If a FATHMM score is ≤ − 1.5 the corresponding variant is predicted as "D(AMAGING)"; on the contrary, it is predicted as "T(OLERATED)". MutationAssessor scores range from 5.135 to 6.49. MutationAssessor's functional impact of a variant: predicted functional, i.e. high ("H") or medium ("M"), or predicted non-functional, i.e. low ("L") or neutral ("N"). MutationTaster score ranges from 0 to 1 and a larger score means more likely to be deleterious. MutationTaster predictions are "A" ("disease_causing_automatic"), "D" ("disease_causing"), "N" ("polymorphism") or "P" ("polymorphism_automatic"). The Loss-of-function (LOF) tool scores < 0.7 are considered benign, scores < 0.2 are considered probably damaging and a score of 0.2 to 0.7 are possibly damaging. The Meta Support Vector Machine (MetaSVM) score ranges from − 2 to 3 and the larger scores indicate the variant is more likely to be damaging. The likelihood ratio test (LRT) ranges from 0 to 1 and the LRT predictions are D(eleterious), N(eutral), or U(nknown). Genomic Evolutionary Rate Profiling (tmGERP)++ scores range from − 12.   [25]. In another study, Chen et al. found that conditional ESRRB −/− mice were significantly deficient in vestibular function, and their endolymph production decreased. In addition, they showed a reduction in the expression of ion channels in the inner ear, highlighting the importance of ESRRB in inner ear homeostasis [33]. Furthermore, the ESRRB is also essential for functioning and developing inner ear cells through its interaction with estrogen, glucocorticoids, and thyroid hormones. These hormones are essential for inner ear cells so that ESRRB defects can influence cellular processes within the inner ear [25]. Even though ESRRB plays an important role in developing inner ear cells, the exact molecular mechanisms by which it affects other target proteins remains unclear. Nevertheless, Colin et al. observed specific patterns in the distribution of ESRRB transcripts in the developing mouse inner ear during embryonic development [25]. Our interactome analysis (Fig. 3) confirmed their results, which revealed ESRRB interacts with most embryogenesis proteins (such as Nanog, SOX2, KLF4, etc.), and their activities might be influenced by each other. Unraveling these regulatory mechanisms is critical to a thorough understanding of HL caused by ESRRB. Altogether, our study confirmed the pathogenicity of the identified ESRRB mutation by computational analysis. However, as a significant limitation, the functional consequences of the mutations were not assessed in this study, which would reveal the specific mechanism by which the c.499G>A mutation affects the ESRRB structure.

Conclusion
Our study evaluated the genetic cause of bilateral severe to profound HL in an Iranian family with Azeri Turkish ethnicity. Using a targeted gene panel, we identified a novel missense variant in the ESRRB gene, which has not already been reported in any literature or databases. Our study confirmed the diagnosis of c.499G>A in the ESRRB gene at the molecular level and verified its pathogenicity by in silico analysis and the ACMG/ AMP guidelines. It is worth mentioning that this report emphasizes the significance of consanguinity marriage, a phenomenon common in Iran that causes homozygous mutations in these types of disorders. Moreover, reporting novel variations can enhance the clinical management process and genetic counseling, thus offering opportunities to prevent the diseases.