Next-Generation Sequencing Identifies Pathogenic Variants in HGF, POU3F4, TECTA, and MYO7A in Consanguineous Pakistani Deaf Families

Background Approximately 70% of congenital deafness is attributable to genetic causes. Incidence of congenital deafness is known to be higher in families with consanguineous marriage. In this study, we investigated the genetic causes in three consanguineous Pakistani families segregating with prelingual, severe-to-profound deafness. Results Through targeted next-generation sequencing of 414 genes known to be associated with deafness, homozygous variants c.536del (p. Leu180Serfs∗20) in TECTA, c.3719 G>A (p. Arg1240Gln) in MYO7A, and c.482+1986_1988del in HGF were identified as the pathogenic causes of enrolled families. Interestingly, in one large consanguineous family, an additional c.706G>A (p. Glu236Lys) variant in the X-linked POU3F4 gene was also identified in multiple affected family members causing deafness. Genotype-phenotype cosegregation was confirmed in all participating family members by Sanger sequencing. Conclusions Our results showed that the genetic causes of deafness are highly heterogeneous. Even within a single family, the affected members with apparently indistinguishable clinical phenotypes may have different pathogenic variants.


Background
Congenital hearing loss affects 1‰-2‰ of newborns worldwide. Among them, approximately 70% of deafness is attributable to genetic causes [1]. The genetics of hearing loss is extremely heterogeneous as, to date, more than 120 genes are reported to cause nonsyndromic hearing loss (Hereditary Hearing Loss Homepage; https://hereditaryhearingloss.org).
Similarly, numerous genes are known to cause syndromic hearing loss [2][3][4][5][6][7][8][9][10]. In recent years, next-generation sequencing (NGS) has been increasingly implemented in the genetic diagnosis of heterogeneous diseases including hearing loss, providing a high-throughput, efficient approach to target hundreds of causative genes or even the whole exome for the identification of pathogenic variants [11,12]. Due to the extremely high heterogeneity of genetic hearing loss, however, determining the pathogenicity of the candidate variants can be challenging in many cases. For example, rare, benign variants are not always distinguishable from the true pathogenic variants following general guidelines for sequencing data interpretation [13,14]. Reporting of rare variants with phenotypic cosegregation in large families, therefore, will provide valuable references for genetic diagnosis of deafness in isolated cases [14].
The deafness-associated genes play diverse roles in the development, function, and maintenance of the inner ear. Variants in these genes correspondently lead to variable auditory phenotypes in regard to onset, severity, progression, audiogram profile, and accompanying syndromic features [1]. In some cases, different variants in the same gene, such as MYO7A, may result in distinct phenotypes in nonsyndromic deafness DFNA11 and DFNB2 [15,16] or syndromic deafness USH1B [17]. Documentation and analysis of the genotype-phenotype correlation covering a broad range of novel or previously less characterized variants are therefore necessary to facilitate genetic diagnosis of hearing loss.
Approximately 80% of genetic deafness are inherited in an autosomal recessive manner [18]. In many cases, they are closely related to consanguineous marriage, which is common in regions of the Middle East, South Asia, South America, and Africa [19,20]. Studies in Iranian and Palestinian populations, for example, showed that, respectively, 65% and 58% of deaf children were born to parents having consanguineous marriage [21,22]. A number of deafnessassociated genes are discovered by linkage analyses based on homozygosity mapping of large consanguineous deaf families [23,24].
In this study, we performed targeted NGS in three Pakistani consanguineous families and identified novel variants in TECTA and POU3F4 and previously reported variants in MYO7A and HGF as the pathogenic causes for prelingual, severe-to-profound deafness. Interestingly, in a large, multigenerational consanguineous family, we identified two separated variants in HGF and POU3F4, illustrating the complex genetic heterogeneity of deafness.

Subjects and Clinical
Evaluations. Three consanguineously married families (PK-DD-KA-01 (Figure 1), PK-DD-RP-01 (Figure 2), and PK-DB-OKA-01 ( Figure 3)) were enrolled from three districts (Muzafargarh, Rajanpur, and Okara, respectively) in Punjab (Pakistan) with multiple individuals suffering from deafness. Written informed consents were obtained from all participants and/or their parents before their enrollment in this study. Family members affected with deafness were examined in the ear, nose, and throat (ENT) wards of their respective District Head Quarter Hospital by medical specialists. Clinical evaluations included complete medical history interview, comprehensive physical examination, and pure tone audiometry (PTA). Otoscopic examination was performed to exclude hearing loss due to infections, trauma, or other environmental factors. The vestibular function was evaluated by medical history inquiry and behavioral testing. This study was approved by the Research Ethics Committee of the Peking University Shenzhen Hospital.

Genetic Analysis.
Genomic DNA was extracted from peripheral blood samples of the enrolled subjects by using the QIAamp DNA Blood Mini Kit (QIAGEN, Shanghai). Targeted NGS was performed in the proband from each family (arrows in Figures 1(a), 2(a), and 3(a)) plus individual IV-5 from Family PK-DB-OKA-01. The customized capture panel (MyGenostics, Beijing) targeted the exonic region of 414 known deafnessassociated genes; in addition to the exonic region, the panel also included all intronic, intergenic, and noncoding regions containing variants that reported in HGMD (147 genes, 751 variants) (Supplementary Table S1). Candidate pathogenic variants were defined as nonsynonymous (including nonsense, missense, splice-site, and indels) variants with minor allele frequencies (MAFs) less than 0.005 in public databases including 1000 Genomes, dbSNP, and GnomAD. Potential pathogenic effect of the candidate variants was evaluated by in silico tools MutationTaster, PROVEAN, SIFT, and PolyPhen-2. Cosegregation of the deafness phenotype and the pathogenic variants was confirmed in participating family members by PCR amplification and Sanger sequencing; the results were shown in each pedigree map. Pathogenicity of the variants was classified following the guidelines of ACMG 2015 [13].  Figure 3(a)) that were suffering from bilateral, prelingual, severe-to-profound sensorineural hearing loss (Figures 1(c), 2(c), and 3(c)). All affected subjects in Families PK-DD-KA-01 and PK-DD-RP-01 and half (11/22) of the affected subjects in Family PK-DB-OKA-01 were born to parents with consanguineous marriage. No ear malformation, vestibular dysfunction, developmental abnormality, or syndromic symptom were identified in enrolled subjects.  Figure 3(a)), respectively. All candidate variants have minor allele frequencies lower than 0.0001 in the public database gnomAD and are categorized as likely pathogenic based on ACMG guidelines (Table 1).     b)). This novel variant is not seen in the gnomAD databases and has not been previously reported in association with hearing loss. It substitutes an evolutionarily conserved, acidic residue glutamic acid to an alkaline residue lysine at position 236 (Figure 3(e)) and is predicted to be deleterious or probably damaging by in silico tools Mutation-Taster, PROVEAN, SIFT, and PolyPhen-2 (Table 1).

Discussion
Hearing loss is one of the major disabilities worldwide, which is often induced by loss of sensory hair cells [25][26][27][28][29] and spiral ganglion neurons [30][31][32][33][34] in the inner ear cochlea. Hearing loss could be caused by genetic factors, aging, chronic cochlear infections, infectious diseases, ototoxic drugs, and noise exposure [35][36][37][38][39][40][41][42], and genetic factors account for around 70% of the hearing loss. In the present study, we are reporting the genetic causes of the prelingual, severe-to-profound deafness in three consanguineous Pakistani families through targeted NGS approach. Variants c.536del (p. Leu180Serfs * 20) in TECTA and c.706G>A (p. Glu236Lys) in POU3F4 are novel, while c.3719 G>A (p. Arg1240Gln) in MYO7A and c.482 +1986_1988del in HGF were previously reported in only very limited cases associated with deafness [43,44]. In support of their pathogenicity, all four variants segregated with multiple affected and unaffected family members consistent with the autosomal recessive or X-linked inheritance modes. Considering that homozygosity of rare variants is quite rare in the general population, the data generated during the present study will provide valuable genetic evidence regarding the genetic basis of deafness.
The phenotypes of the three families in this study are all characterized as prelingual, severe-to-profound sensorineural hearing loss. Consistently, a similar type of hearing loss has also been associated with many variants in TECTA, MYO7A, HGF, and POU3F4 in previous reports. TECTA encodes α-tectorin, which is one of the main components to comprise the tectorial membrane in the cochlea [45]. Unlike dominant TECTA variants, which are associated with milder hearing loss DFNA8/12 (MIM 601543), the recessive TECTA variants often result in prelingual, severe-toprofound hearing loss (DFNB21, MIM 603629) [46][47][48][49][50]. Like several other recessive, truncating variants in TECTA, the c.536del (p. Leu180Serfs * 20) variant identified in this study causes a frameshift; this will result either in an abortive protein truncated in exon 4 or in no protein at all due to nonsense-mediated mRNA decay [51] and likely results in loss of function (Figure 4(a)). MYO7A is extensively expressed in the hair cells and plays an important role in maintaining stereocilia differentiation and morphology [52,53]. The c.3719 G>A (p. Arg1240Gln) variant in MYO7A locates in the highly conserved first MyTH4 domain  Neural Plasticity (Figure 4(b)). It has been previously reported to be associated with Usher syndrome 1B (MIM 276900), characterized by congenital, severe-to-profound hearing loss, and late-onset retinitis pigmentosa before or after puberty [44,54]. Since the two affected children with the c.3719 G>A (p. Arg1240Gln) homozygous variant in our study were 7 and 9 years old without apparent visual abnormality, the potential visual dysfunction remains to be determined at older age. HGF encodes a hepatocyte growth factor and is expressed in the stria vascularis of the mouse inner ear. The c.482+1986_1988del variant identified in this study has been previously reported to cause prelingual, profound deafness (DFNB39, MIM 608265) [43]. A Hgf 10 bp deletion in homozygous mutant mice, which fully encompasses the 3 bp deletion in humans, also displayed profound hearing loss at 4 weeks age (Figure 3(d)) [55]; Hgf 10 bp deletion in homozygous mice causes low expression of Hgf in the cochlea, which leads to developmental defect of the stria vascularis and reduced endocochlear potential in the cochlea [55]. The 3 bp deletion in the intronic region possibly has a similar mechanism to cause hearing loss. POU3F4 encodes a POU domain transcription factor expressed in a spiral ligament and spiral limbus [56]. Most DFNX2 (MIM 304400) patients with POU3F4 variants have profound hearing loss with or without developmental abnormality of the conductive components [57,58]. Although the temporal bone abnormities could not be confirmed in this family, the PTA results of affected males in the family only presented sensorineural hearing loss, which is less common than mixed hearing loss in patients with POU3F4 variants. The novel c.706G>A (p. Glu236Lys) variant identified in this study is located in the highly conserved POU-specific domain (Figure 4(c)), which is similar to a previously reported c.707A>C; p. (Glu236Ala) variant in a Turkey family [59]. Overall, our results are consistent with known genotype-phenotype correlation of the corresponding genes.
−/+ −/+ X + /X + X + /X + X + /X + X + /Y +/+ +/+ +/+   Neural Plasticity As summarized above, in this study, rare pathogenic variants were identified in four separate deafness-associated genes, TECTA, MYO7A, HGF, and POU3F4, which have distinct expression profiles and functions in the inner ear. Nevertheless, all four variants resulted in an almost uniformed type of prelingual, severe-to-profound deafness, representing yet another example of extremely high genetic heterozygosity for hearing loss. Interestingly, within the large pedigree of Family PK-DB-OKA-01 with multiple consanguineous marriages, five of the fifteen affected individuals actually have a hemizygous, X-linked variant as a separate cause of hearing loss irrelevant to the consanguineous marriage pattern (Figure 3(a)). Since most novel deafness-causative genes were originally identified through genetic analysis of such large pedigrees based on assumptions that all affected family members share a single pathogenic variant, our results suggested that caution should remain against such complexed inheritance patterns involving two or multiple genetic causes.

Conclusions
In summary, our study of three consanguineous families with prelingual, severe-to-profound deafness revealed a rather heterogeneous variant spectrum in the corresponding Pakistani deaf communities. Next-generation sequencing illustrates its advantages in resolving such complexed cases.      Neural Plasticity

Data Availability
The original data is available upon request by contacting the corresponding author.

Conflicts of Interest
The authors declare no conflict of interest.