In this study, we investigated the clinical phenotype and molecular etiology of 2 persons in a family with inherited deafness across 3 generations. According to the progressive hearing loss and visual acuity loss in the proband and in some patients in the family, accompanied by the proband’s ophthalmic examination, USH3 was initially diagnosed. Two patients carried a homozygous variation of CLRN1 c.474T > A (P.Cys158Ter, NM_001256819.2) or c.302T > A (p.Val101Asp, NM_174878.3). Sanger sequencing was subsequently used to verify the likely pathogenic variants sites in the above two branches, and the verification structure revealed that the homozygous variation of CLRN1 c.474T > A (P.Cys158Ter, NM_001256819.2) or c.302T > A (p.Val101Asp, NM_174878.3) was coisolated with the genotype‒phenotype of the family.

USH3 mainly manifests as postlinguistic progressive bilateral symmetrical SHL, and vestibular function may decline, with or without RP occurring before 20 years of age [7]. However, patients with USH3 showed significant genetic heterogeneity and phenotypic specificity. At present, a CLRN1 gene mutation is believed to cause USH3A, and an HARS gene mutation causes USH 3B [15]. USH3 is relatively rare compared to USH1 and USH2, although Hope et al. [16] reported that USH3 accounted for approximately 20% of USH in Birmingham. Ness et al. [5]reported that this subtype accounted for approximately 40% of all USH patients in Finnish and Ashkenazi Jews. However, in the Spanish population, a study of 89 USH patients with no instances of USH3 was followed by analysis of a sample of 131 families with only 8 USH3 cases, an incidence of only 6% [8]. Familial USH3 has not been reported in the Chinese population; only sporadic cases have been reported due to eye disease.

Due to the high phenotypic heterogeneity of USH3 patients not only among different families but also within the same family, Ness et al. [5] believed that the phenotype of USH3 patients with SHL was relatively consistent, and this clinical phenotype was used as an important distinguishing feature to distinguish USH3 from UHS2 and USH1. Another clinical manifestation commonly used to distinguish USH3 from UHS2 and USH1 is the onset time of deafness.

Postlingual SHL is characteristic of USH3, whereas prelingual hearing loss is a feature of USH. In this study, the proband had typical USH3 clinical symptoms—from hearing, vision and typical RP—the onset age and development were similar to those found in the literature, but her older brother (II-1) developed binaural severe SHL after birth, night blindness since childhood and, poor binocular vision; moreover, his binocular vision is almost completely gone. Therefore, the clinical manifestations of USH3 patients in this family were not identical. In 1999, Adato et al. [17] found 2 patients with different clinical phenotypes in a nonclose family of Jewish Yemenis, including 1 patient with the USH1 clinical phenotype and the other with the USH3 clinical phenotype. In subsequent studies, the authors reported that the different clinical phenotypes in the same family occurred were because patients with the USH1 phenotype carried the complex heterozygous variation in MYO7A, while patients with the homozygous variation of the USH3 haplotype presented the USH3 phenotype. Ness et al. [5] also found that among 40 Ashkenazi Jews with USH, 16 patients were clinically classified as USH3. Although all patients had CLRN1 mutations, those with homozygous p.N48K showed more severe clinical phenotypes, with particularly prominent audiological manifestations, and the onset age ranged from infancy to over 35 years old. Aller et al. [8] reported that three Spanish families with USH also presented with postlingual bilateral symmetric severe SHL (at age 6: average hearing threshold 70 dB; age 17: average hearing threshold > 90 dB).In a study of CLRN1p.Y176X knockout mice, the clinical phenotype of the knockout mice was very similar to that of patients with the CLRN1p.Y176X mutation; however, hearing loss occurred early and progressed rapidly, and vestibular function loss occurred much later than hearing loss. These results indicate that clinical confusion may exist between USH3 and USH1 or between USH3 and USH2. However, due to delayed and progressive hearing loss, some patients whose hearing is clinically consistent with that of USH3 have been found to carry variants of USH2A, USH1B and USH1D. Therefore, according to our results and previously reported findings, progressive SHL and postlingual deafness are not important criteria for the diagnosis and differentiation of USH3. With the gradual increase in case data, USH3 clinical phenotypes have appeared in more forms than clinical phenotypes for USH diagnosis, and the genetic diagnosis has become much clearer and more reliable.

The main significance of this study is embodied in the following aspects. The NCBI website (CLRN1 clarin 1 [[Homo sapiens(human)] -Gene-NCBI (nih.gov)) recorded four transcripts (Fig. 1), which transcribed different amino acid sequences due to different exons contained. This partly explains why CLRN1 mutations cause heterogeneity in the clinical phenotype of patients. NM_174878.3 and NP_777367(isoform a is encoded by transcript variant 1) contain exons 0-1-2, encoding a clarin-1 protein containing 232 amino acids of approximately 25.8 kDa [18]. All the known CLRN1 mutations are located in the 3 exons of isoform a, except an intron variant in the intron region between exon 0 and exon 0b of isoform e. With the exception of the p.Y176X and P.P.120 K mutations found in Finnish populations [19] and the p.N48K variant found in Ashkenazi Jews [20], most of the other variants have been found in single families. In addition to these three common mutations, 21 missense and nonsense mutations, 2 occurrences of abnormal splicing, 7 deletion mutations, and 3 insertion mutations have been reported to date (Human Gene Mutation Database, Deadline 2023.03).

In 2017, Khan et al. [21] reported a CLRN1 intron mutation c.254–649T > G(NM_001256819.1, isoform e)in an inbred family from the Arabian Peninsula diagnosed with USH1, located in the intron region between exon 0 and exon 0b. Through minigene splicing experiments, it was proven that the mutation caused abnormal splicing of exons, resulting in frameshift and early appearance of stop codons. The authors then tested for the mutation site in seven untested Saudi USH1 patients with a related genetic mutation and found that two of them carried the mutation. Pertinently, c.254–649T > G, which is also the human pathogenic mutation reported to date for non-NM_174878.3 transcripts, represents a founder allele that may significantly contribute to deaf-blindness in people on the Arabian Peninsula.

Using the prediction software Alphafold2, we found that the homozygous variants [CLRN1 c.474T > A (P.Cys158Ter, NM_001256819.2) or c.302T > A (p.Val101Asp, NM_174878.3)] carried by the familial patients had different effects on different transcripts of CLRN1. Although we did not obtain direct evidence from patient-derived ILCLs, we confirmed the presence of specific transcripts in HEK293T cells, and through plasmid expression experiments, we observed protein expression differences caused by c.474T > A (P.Cys158Ter, NM_001256819.2). The mechanism may be that the mutation of noncanonical NM_001256819.2 leads to alternative splicing of the canonical transcript, which eventually leads to protein degradation.

The CLRN1c.474T > A (p.Cys158Ter) variant has not been reported before. To the best of the authors’ knowledge, there is no report about this variant in any databases, including The Single Nucleotide Polymorphism Database, The Human Gene Mutation Database, 1000 Genomes Project and ClinVar and Exome Sequencing Project v. 6500. As suggested by the ACMG/AMP guidelines, a pathogenicity analysis was performed: (a) According to the recommendations of ACMG/AMP guidelines, pathogenicity analysis was performed: (a) Through in vitro experiment verification of lymphocyte line construction and plasmid transfection experiment, it was found that CLRN1c.474T > A (p.Cys158Ter) variant led to abnormal protein expression and impaired gene function (strong pathogenic evidence PS3); (b) the novel variant CLRN1c.474T > A (p.Cys158Ter) was not identified in control groups, the frequency in the normal population database is “-” when compared that of GnomAD (moderate pathogenic evidence PM2); (c) Hearing loss and RP expression in proband and older siblings are highly characteristic of the CLRN1 gene (supporting evidence for pathogenesis, PP4) and (e) USH3 is a recessive genetic disorder. The affected siblings were found to carry CLRN1c.474T > A (p.Cys158Ter) homozygous mutations, while the remaining family members had either heterozygous or wild type mutations, confirming that the phenotype of hearing loss and retinitis pigmentosa co-segregated with genotype in this family (supportive pathogenic evidence, PP1). Taken together, the evidence for the c.474T > A mutation is “PS3 + PM2 + PP4 + PP1” and is judged to be a pathogenic mutation (very strong pathogenic evidence). The CLRN1c.474T > A (p.Cys158Ter, NM_001256819.2) mutation underlies the first report of USH3 in the Chinese population.