Usher syndrome (USH) is an autosomal recessively inherited disease characterized by sensorineural hearing loss (SNHL) and retinitis pigmentosa (RP) with or without vestibular dysfunction. It is highly heterogeneous both clinically and genetically. Recently, variants in the arylsulfatase G (ARSG) gene have been reported to underlie USH type IV. This distinct type of USH is characterized by late-onset RP with predominantly pericentral and macular changes, and late onset SNHL without vestibular dysfunction. In this study, we describe the USH type IV phenotype in three unrelated subjects. We identified three novel pathogenic variants, two novel likely pathogenic variants, and one previously described pathogenic variant in ARSG. Functional experiments indicated a loss of sulfatase activity of the mutant proteins. Our findings confirm that ARSG variants cause the newly defined USH type IV and support the proposed extension of the phenotypic USH classification.

CRISPR-Cas9-based genome-editing is a highly efficient and cost-effective method to generate zebrafish loss-of-function alleles. However, introducing patient-specific variants into the zebrafish genome with CRISPR-Cas9 remains challenging. Targeting options can be limited by the predetermined genetic context, and the efficiency of the homology-directed DNA repair pathway is relatively low. Here, we illustrate our efficient approach to develop knock-in zebrafish models using two previously variants associated with hereditary sensory deficits. We employ sgRNA-Cas9 ribonucleoprotein (RNP) complexes that are micro-injected into the first cell of fertilized zebrafish eggs together with an asymmetric, single-stranded DNA template containing the variant of interest. The introduction of knock-in events was confirmed by massive parallel sequencing of genomic DNA extracted from a pool of injected embryos. Simultaneous morpholino-induced blocking of a key component of the non-homologous end joining DNA repair pathway, Ku70, improved the knock-in efficiency for one of the targets. Our use of RNP complexes provides an improved knock-in efficiency as compared to previously published studies. Correct knock-in events were identified in 3–8% of alleles, and 30–45% of injected animals had the target variant in their germline. The detailed technical and procedural insights described here provide a valuable framework for the efficient development of knock-in zebrafish models.

The USH2A variant c.2276 G > T (p.(Cys759Phe)) has been described by many authors as a frequent cause of autosomal recessive retinitis pigmentosa (arRP). However, this is in contrast with the description of two asymptomatic individuals homozygous for this variant. We therefore assessed pathogenicity of the USH2A c.2276 G > T variant using extensive genetic and functional analyses. Whole genome sequencing and optical genome mapping were performed for three arRP cases homozygous for USH2A c.2276 G > T to exclude alternative genetic causes. A minigene splice assay was designed to investigate the effect of c.2276 G > T on pre-mRNA splicing, in presence or absence of the nearby c.2256 T > C variant. Moreover, an ush2ap.(Cys771Phe) zebrafish knock-in model mimicking human p.(Cys759Phe) was generated and characterized using functional and immunohistochemical analyses. Besides the homozygous c.2276 G > T USH2A variant, no alternative genetic causes were identified. Evaluation of the ush2ap.(Cys771Phe) zebrafish model revealed strongly reduced levels of usherin expression at the photoreceptor periciliary membrane, increased levels of rhodopsin localization in the photoreceptor cell body and decreased electroretinogram (ERG) b-wave amplitudes compared to wildtype controls. In conclusion, we confirmed pathogenicity of USH2A c.2276 G > T (p.(Cys759Phe)). Consequently, cases homozygous for c.2276 G > T can now receive a definite genetic diagnosis and can be considered eligible for receiving future QR-421a-mediated exon 13 skipping therapy.