Analysis of SLC26A4, FOXI1, and KCNJ10 Gene Variants in Patients with Incomplete Partition of the Cochlea and Enlarged Vestibular Aqueduct (EVA) Anomalies

Pathogenic variants in the SLC26A4, FOXI1, and KCNJ10 genes are associated with hearing loss (HL) and specific inner ear abnormalities (DFNB4). In the present study, phenotype analyses, including clinical data collection, computed tomography (CT), and audiometric examination, were performed on deaf individuals from the Sakha Republic of Russia (Eastern Siberia). In cases with cochleovestibular malformations, molecular genetic analysis of the coding regions of the SLC26A4, FOXI1, and KCNJ10 genes associated with DFNB4 was completed. In six of the 165 patients (3.6%), CT scans revealed an incomplete partition of the cochlea (IP-1 and IP-2), in isolation or combined with an enlarged vestibular aqueduct (EVA) anomaly. Sequencing of the SLC26A4, FOXI1, and KCNJ10 genes was performed in these six patients. In the SLC26A4 gene, we identified four variants, namely c.85G>C p.(Glu29Gln), c.757A>G p.(Ile253Val), c.2027T>A p.(Leu676Gln), and c.2089+1G>A (IVS18+1G>A), which are known as pathogenic, as well as c.441G>A p.(Met147Ile), reported previously as a variant with uncertain significance. Using the AlphaFold algorithm, we found in silico evidence of the pathogenicity of this variant. We did not find any causative variants in the FOXI1 and KCNJ10 genes, nor did we find any evidence of digenic inheritance associated with double heterozygosity for these genes with monoallelic SLC26A4 variants. The contribution of biallelic SLC26A4 variants in patients with IP-1, IP-2, IP-2+EVA, and isolated EVA was 66.7% (DFNB4 in three patients, Pendred syndrome in one patient). Seventy-five percent of SLC26A4-biallelic patients had severe or profound HL. The morphology of the inner ear anomalies demonstrated that, among SLC26A4-biallelic patients, all types of incomplete partition of the cochlea are possible, from IP-1 and IP-2, to a normal cochlea. However, the dominant type of anomaly was IP-2+EVA (50.0%). This finding is very important for cochlear implantation, since the IP-2 anomaly does not have an increased risk of “gushers” and recurrent meningitis.

In 2007, Yang et al. [38] described two DFNB4 families with heterozygous variants in the FOXI1 gene and proposed that this gene showed significantly decreased luciferase activation in promoter-reporter assays, suggesting that this variant compromised the ability of FOXI1 to transactivate SLC26A4 and was causally related to disease [38][39][40][41]. The FOXI1 gene is located on chromosome 5, contains two coding exons, and belongs to the forkhead family of transcription factors, characterized by a distinct forkhead domain. This gene may play an important role in the development of the cochlea and vestibulum, as well as in embryogenesis [42][43][44]. In addition, the encoded protein has been found to be required for the transcription of four subunits of a proton pump found in the inner ear, the kidney, and the epididymis [45][46][47]. Moreover, Yang et al. proposed the hypothesis of digenic inheritance in patients with monoallelic SLC26A4 variants associated with the heterozygous variants of the FOXI1 and KCNJ10 genes [38,[47][48][49]. The KCNJ10 gene is located on chromosome 1, contains one coding exon, and encodes a member of the inward rectifier-type potassium channel family, characterized by having a greater tendency to allow potassium to flow into, rather than out of, a cell [42][43][44]. The encoded protein may form a heterodimer with another potassium channel protein and may be responsible for the potassium buffering action of glial cells in the brain [42][43][44]. The OMIM database describes three family cases (https://omim.org/entry/600791, accessed on 2 December 2022) with digenic inheritance in patients with monoallelic SLC26A4 variants associated with heterozygous variants of the FOXI1 and KCNJ10 genes.
In Russia, the FOXI1 and KCNJ10 genes have not previously been analyzed among patients with HL. However, pathogenic variant analysis of the SLC26A4 gene was previously performed in 246 families with HL cases from the Bashkortostan Republic of Russia (Volga-Ural Region) [50] and in 20 patients with Pendred syndrome, EVA, and/or Mondini anomalies in a geographically dispersed sample [51]. Analyses of pathogenic variants of the SLC26A4 gene in 313 patients with HL in the Tyva and Altai Republics of Russia (Southern Siberia) revealed the different contributions of biallelic SLC26A4 pathogenic variants to the etiology of HL (28.2% in Tuvinian and 4.3% in Altaian patients) [52]. In total, 14 recessive pathogenic variants of the SLC26A4 gene were identified: c.  [50][51][52].
In the present study, consecutive genotype-phenotype analyses were performed in 165 deaf individuals from the Sakha Republic of Russia (Eastern Siberia), including clinical data collection, computed tomography, audiometric examination, and Sanger sequencing of the coding regions of the SLC26A4, FOXI1, and KCNJ10 genes associated with DFNB4.

Allelic Frequency and Clinical Significance of the Identified Variants in the SLC26A4, FOXI1, and KCNJ10 Genes
In six of the 165 patients (3.6%), CT scans revealed an incomplete partition of the cochlea (IP-1 and IP-2) in isolation or combined with an enlarged vestibular aqueduct (EVA) anomaly. In patients with incomplete partition of the cochlea (IP-1 and IP-2) and/or an enlarged vestibular aqueduct (EVA) anomaly, we sequenced the coding region of the SLC26A4, FOXI1, and KCNJ10 genes. In the SLC26A4 gene, five previously known variants were found, four of which were missense variants, namely c.85G>C p.(Glu29Gln) (10%), c.441G>A p.(Met147Ile) (20%), c.757A>G p.(Ile253Val) (20%), and c.2027T>A p.(Leu676Gln) (20%), with one variant that affected the donor splicing site c.2089+1G>A (IVS18+1G>A) (10%). Two synonymous variants were found in the FOXI1 gene: c.279G>A p.(Arg93=) (30%) and c.1044T>C p.(Tyr348=) (90%), and one previously known missense variant, c.811C>T p.(Arg271Cys) (10%), was found in the KCNJ10 gene. The allelic frequency of the SLC26A4, FOXI1, and KCNJ10 gene variants identified in this study is presented in Table 1. The allelic frequency of the identified variants in some populations according to gnomAD data is shown in Table S1. Among the variants present in the ClinVar database, the three SLC26A4 variants c.85G>C p.(Glu29Gln), c.2027T>A p.(Leu676Gln), and c.2089+1G>A (IVS18+1G>A) are considered to be pathogenic. One variant, c.757A>G p.(Ile253Val), was absent in ClinVar; however, it was described as likely pathogenic in the Deafness Variation Database (https://deafnessvariationdatabase.org/ accessed on 28 September 2022) ( Table 1). The two FOXI1 variants c.279G>A p.(Arg93=) and c.1044T>C p.(Tyr348=), and one KCNJ10 variant c.811C>T p.(Arg271Cys), were classified as benign or likely benign ( Table 1). The c.441G>A p.(Met147Ile) variant of the SLC26A4 gene has conflicting interpretations of pathogenicity: in ClinVar, it is listed as likely benign (1) and of uncertain significance (6) (https://www.ncbi.nlm.nih.gov/clinvar/variation/229258/ accessed on 28 September 2022) ( Table 1). Note. The variant with uncertain significance (VUS) is highlighted in bold, * Data were taken from the Deafness Variation Database (DVD). ** One sibling was excluded from calculation of allelic frequency. In total, from six patients (12 chromosomes), the allelic frequency was calculated on five unrelated patients (10 chromosomes).  [53]. However, the authors identified only heterozygous variants in patients presenting unilateral HL and ipsilateral EVA [53], which is not the cause of disease with the autosomal recessive type of inheritance. According to the ACMG guidelines, a variant of uncertain significance (VUS) should not be used in clinical decision-making [54,55]. Regardless of the results of in silico pathogenicity prediction programs (SIFT, Polyphen-2, PROVEAN, and Pathogenic variant Taster), which all rated it as damaging (Table S2), we applied an additional in silico pathogenicity assessment of this variant; since, in the most programs, computational prediction methods are based on a similar algorithm of evolutionary conservatism. However, the effect of a missense change depends, not only on the evolutionary conservation of an amino acid or nucleotide, but also on other criteria, such as the location and context within the protein sequence, and the biochemical consequence of the amino acid substitution [54]. However, the crystal structure of human pendrin protein is currently unknown, making a predictive analysis of the pathogenic effect of SLC26A4-variants on the function and/or structure of the pendrin protein impossible [16].

In Silico
In this connection, using the AlphaFold 2.0 AI system [56] and the PyMOL molecular graphics system [57], we modeled and aligned the mutant and native protein structures, to evaluate the likely pathogenic effect on the function and/or structure of pendrin (SLC26A4) ( Figures S1 and S2). A detailed in silico analysis and a discussion of the results are presented in the Supplementary Materials. As a result, the calculated similarity index of the two structures' RMSD (root mean square deviation of the atomic positions) was less than 2 Å, which indicates that the missense p.(Met147Ile) substitution theoretically does not violate the structural integrity of the SLC26A4 protein ( Figure S2). Probably, the pathogenic effect of this variant occurs at the functional level, since the analyzed p.(Met147Ile) substitution is located in a critical region of the core domain ( Figures S3 and S4), the violation of which can lead to incorrect substrate transport or the appearance of toxic conformations. Thus, the obtained data indicate that the c.441G>A p.(Met147Ile) variant of the SLC26A4 gene, which is located in an evolutionarily conservative region (the α3-helix region of the TMD) ( Figure S5 (Table S3). This finding provides additional evidence of the pathogenicity of the c.441G>A p.(Met147Ile) variant of the SLC26A4 gene. In accordance with the criteria recommended by ACMG [54], adapted for hereditary forms of HL [55], the c.441G>A p.(Met147Ile) variant was classified by us as "likely pathogenic" (Table S4).

Phenotypes of Families with Biallelic Variants in the SLC26A4 Gene
Considering the pathological effect of the identified variants, we present the clinical features of three families with biallelic, and probably causative, variants identified in the SLC26A4 gene.  (Table S3). This finding provides additional evidence of the pathogenicity of the c.441G>A p.(Met147Ile) variant of the SLC26A4 gene. In accordance with the criteria recommended by ACMG [54], adapted for hereditary forms of HL [55], the c.441G>A p.(Met147Ile) variant was classified by us as "likely pathogenic" (Table S4).

Phenotypes of Families with Biallelic Variants in the SLC26A4 Gene
Considering the pathological effect of the identified variants, we present the clinical features of three families with biallelic, and probably causative, variants identified in the SLC26A4 gene.
Family 1 Patient II:3 had congenital HL and was five years old, female, and Russian. In this patient, the homozygous c.441G>A p.(Met147Ile) variant was detected in the SLC26A4 gene ( Figure 1). This is the first report on the identification of the c.441G>A p.(Met147Ile) variant in the SLC26A4 gene in the homozygous state. Bilateral IP-1 on both sides, bilateral vestibule dilatation, and unilateral EVA were found in her CT scans ( Figure 1). The audiological examination revealed severe bilateral HL.   Figure 2). In addition, the observed audiological examination revealed profound bilateral HL in both sisters. No thyroid dysfunction was found in either sister. However, one of the sisters (Patient II:3) had some vestibular impairment and complained of dizziness.
in the compound heterozygous state (Figure 2). CT scans of both sisters revealed an IP-2 anomaly of the inner ear, represented by a cystic cochlea, a dilatated vestibule, and an enlarged vestibular aqueduct (EVA) on both sides ( Figure 2). In addition, the observed audiological examination revealed profound bilateral HL in both sisters. No thyroid dysfunction was found in either sister. However, one of the sisters (Patient II:3) had some vestibular impairment and complained of dizziness.

Clinical and Molecular Genetic Characteristics of Patients with the Studied Inner Ear Malformations
In total, the majority of patients with biallelic pathogenic variants in the SLC26A4 gene (patient codes 16, 17, and 192) had congenital severe or profound HL (75%), except for a patient with bilateral isolated EVA anomalies (patient code 1091), in whom we registered moderate HL (25%) ( Table 2). On the basis of the physical examination and analysis of the available clinical data of the patients with biallelic pathogenic variants in the SLC26A4 gene, we can conclude that the majority of them (75%) did not complain of episodic vertigo, clumsiness, and vomiting, which could be regarded as vestibular dysfunctions (Table 2) [58,59]. However, one patient with the c.2089+1G>A (IVS18+1G>A) and c.85G>C p.(Gly29Gln) variants in the compound heterozygous state (patient code 16) complained of vertigo (25%) ( Table 2). In this patient, we found the symptom triad of Mondini malformation, including a dilated vestibule merging into semicircular canals. A dilated vestibule merging into semicircular canals can be associated with vertigo. However, a similar Mondini malformation was found in her sister, who did not have vertigo. In this case, more in-depth observations were required. An analysis of thyroid dysfunction was carried out, by examining the patients via palpation, without involving other methods of examination (thyroid hormone levels, ultrasound, and perchlorate tests) [59,60]. Most patients with biallelic pathogenic variants in the SLC26A4 gene (75%) showed no signs of goiter. However, one patient (patient code 1091) with a homozygous c.2027T>A p.(Leu676Gln) pathogenic variant in the SLC26A4 gene (Table 2) underwent surgery on the thyroid gland for nodular goiter (25%). Thus, taking the identified phenotypic features and data of the molecular genetic analysis together, DFNB4 (OMIM #600791) was confirmed in three patients (patient codes 16, 17, and 192) ( Table 2). Pendred's syndrome (PDS, OMIM #274600) was confirmed in one patient (autosomal recessive HL with bilateral EVA combined with nodular goiter; patient code 1091) with a homozygous variant c.2027T>A p.(Leu676Gln) in the SLC26A4 gene (Table 2).

Clinical and Molecular Genetic Characteristics of Patients with the Studied Inner Ear Malformations
In total, the majority of patients with biallelic pathogenic variants in the SLC26A4 gene (patient codes 16, 17, and 192) had congenital severe or profound HL (75%), except for a patient with bilateral isolated EVA anomalies (patient code 1091), in whom we registered moderate HL (25%) ( Table 2). On the basis of the physical examination and analysis of the available clinical data of the patients with biallelic pathogenic variants in the SLC26A4 gene, we can conclude that the majority of them (75%) did not complain of episodic vertigo, clumsiness, and vomiting, which could be regarded as vestibular dysfunctions (Table 2) [58,59]. However, one patient with the c.2089+1G>A (IVS18+1G>A) and c.85G>C p.(Gly29Gln) variants in the compound heterozygous state (patient code 16) complained of vertigo (25%) ( Table 2). In this patient, we found the symptom triad of Mondini malformation, including a dilated vestibule merging into semicircular canals. A dilated vestibule merging into semicircular canals can be associated with vertigo. However, a similar Mondini malformation was found in her sister, who did not have vertigo. In this case, more in-depth observations were required. An analysis of thyroid dysfunction was carried out, by examining the patients via palpation, without involving other methods of examination (thyroid hormone levels, ultrasound, and perchlorate tests) [59,60]. Most patients with biallelic pathogenic variants in the SLC26A4 gene (75%) showed no signs of goiter. However, one patient (patient code 1091) with a homozygous c.2027T>A p.(Leu676Gln) pathogenic variant in the SLC26A4 gene (Table 2) underwent surgery on the thyroid gland for nodular goiter (25%). Thus, taking the identified phenotypic features and data of the molecular genetic analysis together, DFNB4 (OMIM #600791) was confirmed in three patients (patient codes 16, 17, and 192) ( Table 2). Pendred's syndrome (PDS, OMIM #274600) was confirmed in one patient (autosomal recessive HL with bilateral EVA combined with nodular goiter; patient code 1091) with a homozygous variant c.2027T>A p.(Leu676Gln) in the SLC26A4 gene (Table 2).  In two patients (patient codes 1095, 18), in the SLC26A4 gene, we found the c.757A>G p.(Ile253Val) variant in the single-heterozygous state, which was previously reported as likely pathogenic (Table 2). Since we did not find a second mutant allele in this gene and did not find any causative variants in the FOXI1 and KCNJ10 genes, we cannot relate this monoallelic SLC26A4-variant to the DFNB4 phenotype.

Discussion
In this study computed tomography of the temporal bones revealed IP-1, IP-2, IP-2+EVA, and isolated EVA anomalies in six of the 165 studied patients (3.6%). Among these six patients, Sanger sequencing revealed pathogenic biallelic SLC26A4-variants in four patients (4/6), monoallelic SLC26A4-variants in two patients (2/6), but no causative variants in FOXI1 (0/6) and KCNJ10 (0/6). The total contribution of biallelic (homozygous or compoundheterozygous) pathogenic variants in the SLC26A4 gene among patients with temporal bone anomalies was 66.7%. On the basis of the obtained clinical and molecular genetic data, DFNB4 was confirmed in three patients, and Pendred's syndrome was confirmed in one patient. The rate of the biallelic SLC26A4-variants in our study was slightly higher than in certain European studies (25-50%) [2,6,18,[29][30][31][32][33][34]. Probably this was due to the relatively low rate of monoallelic variants of the SLC26A4 gene in our sample. The monoallelic SLC26A4 cases (33.3%) were detected in patients of Siberian origin (two single-heterozygotes for c.757A>G p.(Ile253Val) variant) ( Table 2). Since, in this cohort of patients, we did not find a second mutant allele in this gene and did not find any causative variants in the FOXI1 and KCNJ10 genes, we supposed that in these SLC26A4-monoallelic patients, there are other pathogenic variants in the trans-position in the regulator region of this gene, similar to the hypothesis of the CEVA-haplotype [37].
Interestingly, in line with the radiographic classification proposed by Sennaroglu (2002Sennaroglu ( , 2017 [9,10], we found some differences in morphology of the inner ear anomalies between SLC26A4-monoallelic and SLC26A4-biallelic patients. Among the monoallelic SLC26A4 patients, the most common type of anomaly was IP-1 without EVA (50%). On the contrary, among the biallelic SLC26A4 patients, the dominant type of anomaly was IP-2 with EVA (50.0%) (Figure 4). In two patients (patient codes 1095, 18), in the SLC26A4 gene, we found the c.757A>G p.(Ile253Val) variant in the single-heterozygous state, which was previously reported as likely pathogenic (Table 2). Since we did not find a second mutant allele in this gene and did not find any causative variants in the FOXI1 and KCNJ10 genes, we cannot relate this monoallelic SLC26A4-variant to the DFNB4 phenotype.

Discussion
In this study computed tomography of the temporal bones revealed IP-1, IP-2, IP-2+EVA, and isolated EVA anomalies in six of the 165 studied patients (3.6%). Among these six patients, Sanger sequencing revealed pathogenic biallelic SLC26A4-variants in four patients (4/6), monoallelic SLC26A4-variants in two patients (2/6), but no causative variants in FOXI1 (0/6) and KCNJ10 (0/6). The total contribution of biallelic (homozygous or compound-heterozygous) pathogenic variants in the SLC26A4 gene among patients with temporal bone anomalies was 66.7%. On the basis of the obtained clinical and molecular genetic data, DFNB4 was confirmed in three patients, and Pendred's syndrome was confirmed in one patient. The rate of the biallelic SLC26A4-variants in our study was slightly higher than in certain European studies (25-50%) [2,6,18,[29][30][31][32][33][34]. Probably this was due to the relatively low rate of monoallelic variants of the SLC26A4 gene in our sample. The monoallelic SLC26A4 cases (33.3%) were detected in patients of Siberian origin (two single-heterozygotes for c.757A>G p.(Ile253Val) variant) ( Table 2). Since, in this cohort of patients, we did not find a second mutant allele in this gene and did not find any causative variants in the FOXI1 and KCNJ10 genes, we supposed that in these SLC26A4-monoallelic patients, there are other pathogenic variants in the trans-position in the regulator region of this gene, similar to the hypothesis of the CEVA-haplotype [37].
Interestingly, in line with the radiographic classification proposed by Sennaroglu (2002Sennaroglu ( , 2017 [9,10], we found some differences in morphology of the inner ear anomalies between SLC26A4-monoallelic and SLC26A4-biallelic patients. Among the monoallelic SLC26A4 patients, the most common type of anomaly was IP-1 without EVA (50%). On the contrary, among the biallelic SLC26A4 patients, the dominant type of anomaly was IP-2 with EVA (50.0%) (Figure 4). Our results are consistent with the published data [61]. Previously, Mey et al., in genotype-phenotype analyses of a large cohort of patients with nonsyndromic EVA, described the morphology of inner ear anomalies and found that, among patients with biallelic pathogenic variants in the SLC26A4 gene (87%), the predominant type of anomaly was IP-2+EVA (including cochlear anomalies, not only isolated EVA) [61]. A study by Our results are consistent with the published data [61]. Previously, Mey et al., in genotype-phenotype analyses of a large cohort of patients with nonsyndromic EVA, described the morphology of inner ear anomalies and found that, among patients with biallelic pathogenic variants in the SLC26A4 gene (87%), the predominant type of anomaly was IP-2+EVA (including cochlear anomalies, not only isolated EVA) [61]. A study by Mey et al. [61] indicated that the morphology of the inner ear of patients with biallelic SLC26A4 variants supported the pressure theory proposed by Sennaroglu (2002Sennaroglu ( , 2017 [9,10], allowing for a graded range of severity, ranging from the detectable displacement of the topmost interscalar partitions to a nearly normal cochlea [10]. Our study also supports this hypothesis, because among the biallelic SLC26A4 patients, we found all types of incomplete partition of the cochlea, from IP-1 and IP-2 to a normal cochlea ( Figure 5). Interestingly, in one biallelic SLC26A4 patient, we found IP-1 on one side and IP-1+EVA on the other side (Figure 2). In a related study by Mey et al. (2019) [61], the IP-1 malformation was not detected in a large cohort of biallelic SLC26A4 patients with retrievable images of their inner ear morphology. Our study is the first report about these types of anomalies in biallelic SLC26A4 patients.
Mey et al. [61] indicated that the morphology of the inner ear of patients with biallelic SLC26A4 variants supported the pressure theory proposed by Sennaroglu (2002Sennaroglu ( , 2017 [9,10], allowing for a graded range of severity, ranging from the detectable displacement of the topmost interscalar partitions to a nearly normal cochlea [10]. Our study also supports this hypothesis, because among the biallelic SLC26A4 patients, we found all types of incomplete partition of the cochlea, from IP-1 and IP-2 to a normal cochlea (Figure 5). Interestingly, in one biallelic SLC26A4 patient, we found IP-1 on one side and IP-1+EVA on the other side (Figure 2). In a related study by Mey et al. (2019) [61], the IP-1 malformation was not detected in a large cohort of biallelic SLC26A4 patients with retrievable images of their inner ear morphology. Our study is the first report about these types of anomalies in biallelic SLC26A4 patients. Considering that both types of cochlear anomalies are possible, radiographic differences between IP-1 and IP-2 cochlear anomalies in patients with biallelic SLC26A4 pathogenic variants are significant for the surgery stage of cochlear implantation. The clinical importance of the differences between incomplete partition anomalies is that the IP-2 malformation does not have an increased risk of a "gusher" from the cochleostoma during cochlear implantation, which earlier embryological (IP-1) malformations may Considering that both types of cochlear anomalies are possible, radiographic differences between IP-1 and IP-2 cochlear anomalies in patients with biallelic SLC26A4 pathogenic variants are significant for the surgery stage of cochlear implantation. The clinical importance of the differences between incomplete partition anomalies is that the IP-2 malformation does not have an increased risk of a "gusher" from the cochleostoma during cochlear implantation, which earlier embryological (IP-1) malformations may have [61]. In addition, patients with IP-2 have a lower risk of recurrent meningitis than patients with IP-1. This is due to the defective stapes footplate and cerebrospinal fluid filling the cochlea.
There is a cystic structure in the stapes footplate, which is easily infected during an attack of otitis media. This is very typical for IP-1 but not for IP-2 [10].

Patients
Data on 165 patients with HL were obtained from Republican Hospital No. 1-National Center of Medicine (Yakutsk, Russia), and Republican Hospital No. 2-Center for Emergency Medical Care (Yakutsk, Russia). All patients were examined by an audiologist (collection of complaints, anamnesis of life and illness, ENT examinations (otoscopic, rhinoscopic, and pharyngoscopic studies), acumetry, and sound reactivity testing) and a geneticist (hereditary load, type of inheritance, different syndromic and combined forms). Pediatric patients were additionally examined by a teacher of the deaf (assessment of speech status), a psychoneurologist (neurological status), an endocrinologist (skin, mucous membrane, palpation of the thyroid gland), and a cardiologist (auscultation of the heart). All the participants gave written informed consent to participate in the study. This study was approved by the local Biomedical Ethics Committee of Republican Hospital #2 (Protocol 2, Decision 2 of 24 December 2015).

Computed Tomography of the Temporal Bone
The pyramid of the temporal bone was examined on a 4 slice Somatom Sensation 4 CT scanner (Siemens, Germany) and a 64 slice Somatom Definition AS CT scanner (Siemens, Germany) in axial projection with a tomographic layer thickness of 1 mm, a table advance step of 1 mm, a reconstruction increment of 1 mm (InnerEarSpi program Version A40, Siemens, Germany), a voltage of 120 kV, and a current of 70 mA. During visualization of the structures of the pyramid of the temporal bone, 2D images were used, both in the native axial planes and in the MPR reformation mode using a "bone" filter with a window width of 4000 µN and a window level of +700 µN. In addition, radiographic classifications of temporal bone anomalies were used [7][8][9][10].

Pathogenic Variant Analysis
Blood samples were collected from all 165 patients, and DNA was extracted from the blood leukocyte fraction using the phenol-chloroform method. In six of the 165 patients with incomplete partition of the cochlea (IP-1 and IP-2) in isolation or combined with an enlarged vestibular aqueduct (EVA) anomaly, we sequenced the coding region of the SLC26A4, FOXI1, and KCNJ10 genes. To identify SLC26A4 pathogenic variants, DNA fragments of the 21 exons of SLC26A4, including the flanking intronic regions, were sequenced using the oligonucleotide primers [2]. The entire coding regions of FOXI1 and KCNJ10 were also amplified and sequenced according to standard procedures, using the original primer sequences, as follows: FOXI1:

Database Searches
Population and phenotypic databases were used to search for previously described variants, namely dbSNP, a database for single nucleotide polymorphisms and other classes of minor genetic variation (https://www.ncbi.nlm.nih.gov/snp/) accessed on 4 September 2021; The Genome Aggregation Database (gnomAD), a resource developed by an international coalition of investigators, with the goal of aggregating and harmonizing both exome and genome sequencing data from a wide variety of large-scale sequencing projects (https://gnomad.broadinstitute. org/ accessed on 4 September 2021); The International Genome Sample (IGSR), a resource that maintains and shares the human genetic variation resources built by the 1000 Genomes Project (https://www.internationalgenome.org/ accessed on 4 September 2021); Online Mendelian Inheritance in Man (OMIM), an online catalog of human genes and genetic disorders containing a representative sample of disease-associated variants (https://www.omim.org/ accessed on 4 September 2021); The Human Gene Pathogenic variant Database (HGMD), a database of human gene pathogenic variants and gene damage responsible for human hereditary diseases (http://www.hgmd.cf.ac.uk/ac/index.php accessed on 4 September 2021); Clin-Var, a freely accessible public archive of reports of the relationships among human variations and phenotypes, with supporting evidence (https://www.ncbi.nlm.nih.gov/clinvar/ accessed on 4 September 2022); and the Deafness Variation Database, which provides a comprehensive guide to the genetic variations in genes known to be associated with deafness (https://deafnessvariationdatabase.org/). These databases were accessed on 4 September 2022.

In Silico Analysis
The functional effects of the c.441G>A p.(Met147Ile) variant in the SLC26A4 gene were predicted using SIFT (https://sift.bii.a-star.edu.sg/index.html) accessed on 2 February 2022 [62], PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2) accessed on 2 February 2022 [63], Pathogenic variantTaster (https://www.pathogenic varianttaster.org/ accessed on 2 February 2022) [64], and PROVEAN (http://provean.jcvi.org/index.php) [65] (accessed on 23 May 2022). To model the spatial structure of the protein pendrin (SLC26A4), AlphaFold was used, which is a computational algorithm that can regularly predict protein structures with atomic accuracy, even in cases in which no similar structure is known [56]. AlphaFold produces a confidence metric for amino acid residues, as a predicted local distance difference test (pLDDT), on a scale of 0 to 100 [56]. An expected value of pLDDT > 90 is taken as the high accuracy cut-off (blue), pLDDT > 70 indicates low confidence and corresponds to a generally correct backbone prediction (turquoise color, good backbone prediction), pLDDT ≤ 70 indicates that we should also add substantial coverage for sequences without a good template in PDB (yellow color, should be considered with caution), and pLDDT < 50 indicates very low confidence (orange, should not be interpreted) [62]. Full details are available at: https://alphafold.ebi.ac.uk/about. accessed on 2 February 2022. Alignment of the mutant and normal structures of pendrin (SLC26A4) was carried out in the PyMol (Molecular Graphics System) graphics program, which provides 3D visualizations of proteins, small molecules, molecular surfaces, and trajectories [57]. Quantitative measurement of the similarity of the two spatial structures of proteins was carried out by calculating the root mean square deviation (RMSD) of the coordinates of the corresponding atomic positions, which is an indicator of the average distance between the atoms of super-imposed proteins [66]. The criterion for complete similarity of the two structures was an RMSD value less than 2 Å [67,68]. Full details about the PyMol program are available at https://pymol.org/2/#products. Accessed on 15 February 2022.

Conclusions
1. The total contribution of biallelic SLC26A4 cases among the patients with IP-1, IP-2, IP2+EVA, and isolated EVA anomalies was 66.7% (4 out 6 patients). In these cases, on the basis of the obtained clinical and molecular genetic data, DFNB4 was confirmed in three patients, and Pendred's syndrome was confirmed in one patient. All monoallelic SLC26A4 cases (2 out of 6 patients) were detected in patients of Siberian origin with the single-heterozygous variant c.757A>G p.(Ile253Val) (33.3%). Since, in this cohort of patients, we did not find a second mutant allele in this gene and did not find any causative variants in the KCNJ10 and FOXI1 genes, we supposed that in these patients, there are other pathogenic variants in the trans-position in the regulator region of this gene. In these cases, other extensive studies are required.
2. The morphology of the inner ear anomalies among patients with monoallelic SLC26A4 cases demonstrated that IP-1 without EVA (50%) was the most common type of anomaly. Among patients with biallelic pathogenic variants in the SLC26A4 gene, the predominant type of anomaly was IP-2+EVA (50.0%). However, for the first time, our study demonstrated that in biallelic SLC26A4 patients all incomplete partitions of the cochlea are possible, from IP-1 and IP-2 to a normal cochlea. Considering that both types of cochlear anomalies are possible, radiographic differences between IP-1 and IP-2 cochlear anomalies in patients with biallelic SLC26A4 pathogenic variants are significant for the surgical stage of cochlear implantation, since the IP-2 anomaly does not have an increased risk of "gushers" and recurrent meningitis, which is typical for IP-1 malformation.
Limitations of the Study. Although the strength of this study is the investigation of the three genes SLC26A4, FOXI1, and KCNJ10 associated with DFNB4, it also has some limitations. The design of our study was based on current data of phenotype series and hypothesis of digenic inheritance of SLC26A4 and FOXI1 genes or SLC26A4 and KCNJ10 genes, annotated in the OMIM database (OMIM https://omim.org/entry/600791 accessed on 2 December 2022). We did not screen other autosomal genes, such as GJB2 (three out of six patients in our sample had no changes in GJB2 gene) (Table 2), POU3F4, TMC1, and EPHA2 mentioned as associated with these phenotypes. In addition, since the CEVA-haplotype is associated with Caucasian genetic background, we could not test this hypothesis in SLC26A4-monoallelic patients of Siberian origin. However, we hope that further extensive studies of the SLC26A4 monoallelic cases, among patients from similar isolated populations, will contribute to assessing the role of "silent" variants of this gene.