Mutations in genes encoding components of the mitochondrial DNA (mtDNA) replication machinery cause mtDNA depletion syndromes (MDSs), which associate ocular features with severe neurological syndromes. Here, we identified heterozygous missense mutations in single-strand binding protein 1 (SSBP1) in 5 unrelated families, leading to the R38Q and R107Q amino acid changes in the mitochondrial single-stranded DNA-binding protein, a crucial protein involved in mtDNA replication. All affected individuals presented optic atrophy, associated with foveopathy in half of the cases. To uncover the structural features underlying SSBP1 mutations, we determined a revised SSBP1 crystal structure. Structural analysis suggested that both mutations affect dimer interactions and presumably distort the DNA-binding region. Using patient fibroblasts, we validated that the R38Q variant destabilizes SSBP1 dimer/tetramer formation, affects mtDNA replication, and induces mtDNA depletion. Our study showing that mutations in SSBP1 cause a form of dominant optic atrophy frequently accompanied with foveopathy brings insights into mtDNA maintenance disorders.

Inherited retinal dystrophies (IRDs) are a clinically and genetically heterogeneous group of diseases with more than 250 causative genes. The most common form is retinitis pigmentosa. IRDs lead to vision impairment for which there is no universal cure. Encouragingly, a first gene supplementation therapy has been approved for an autosomal recessive IRD. However, for autosomal dominant IRDs, gene supplementation therapy is not always pertinent because haploinsufficiency is not the only cause. Disease-causing mechanisms are often gain-of-function or dominant-negative, which usually require alternative therapeutic approaches. In such cases, genome-editing technology has raised hopes for treatment. Genome editing could be used to (i) invalidate both alleles, followed by supplementation of the wild type gene, (ii) specifically invalidate the mutant allele, with or without gene supplementation, or (iii) to correct the mutant allele. We review here the most prevalent genes causing autosomal dominant retinitis pigmentosa and the most appropriate genome-editing strategy that could be used to target their different causative mutations.

Autosomal dominant congenital stationary night blindness (adCSNB) is caused by mutations in three genes of the rod phototransduction cascade, rhodopsin (RHO), transducin α-subunit (GNAT1), and cGMP phosphodiesterase type 6 β-subunit (PDE6B). In most cases, the constitutive activation of the phototransduction cascade is a prerequisite to cause adCSNB. The unique adCSNB-associated PDE6B mutation found in the Rambusch pedigree, the substitution p.His258Asn, leads to rod photoreceptors desensitization. Here, we report a three-generation French family with adCSNB harboring a novel PDE6B mutation, the duplication, c.928-9_940dup resulting in a tyrosine to cysteine substitution at codon 314, a frameshift, and a premature termination (p.Tyr314Cysfs*50). To understand the mechanism of the PDE6β1-314fs*50 mutant, we examined the properties of its PDE6-specific portion, PDE6β1-313. We found that PDE6β1-313 maintains the ability to bind noncatalytic cGMP and the inhibitory γ-subunit (Pγ), and interferes with the inhibition of normal PDE6αβ catalytic subunits by Pγ. Moreover, both truncated forms of the PDE6β protein, PDE6β1-313 and PDE6β1-314fs*50 expressed in rods of transgenic X. laevis are targeted to the phototransduction compartment. We hypothesize that in affected family members the p.Tyr314Cysfs*50 change results in the production of the truncated protein, which binds Pγ and causes constitutive activation of the phototransduction thus leading to the absence of rod adaptation.

Inherited retinal diseases (IRDs) display an enormous genetic heterogeneity. Whole exome sequencing (WES) recently identified genes that were mutated in a small proportion of IRD cases. Consequently, finding a second case or family carrying pathogenic variants in the same candidate gene often is challenging. In this study, we searched for novel candidate IRD gene-associated variants in isolated IRD families, assessed their causality, and searched for novel genotype-phenotype correlations. Whole exome sequencing was performed in 11 probands affected with IRDs. Homozygosity mapping data was available for five cases. Variants with minor allele frequencies ≤ 0.5% in public databases were selected as candidate disease-causing variants. These variants were ranked based on their: (a) presence in a gene that was previously implicated in IRD; (b) minor allele frequency in the Exome Aggregation Consortium database (ExAC); (c) in silico pathogenicity assessment using the combined annotation dependent depletion (CADD) score; and (d) interaction of the corresponding protein with known IRD-associated proteins. Twelve unique variants were found in 11 different genes in 11 IRD probands. Novel autosomal recessive and dominant inheritance patterns were found for variants in Small Nuclear Ribonucleoprotein U5 Subunit 200 (SNRNP200) and Zinc Finger Protein 513 (ZNF513), respectively. Using our pathogenicity assessment, a variant in DEAH-Box Helicase 32 (DHX32) was the top ranked novel candidate gene to be associated with IRDs, followed by eight medium and lower ranked candidate genes. The identification of candidate disease-associated sequence variants in 11 single families underscores the notion that the previously identified IRD-associated genes collectively carry > 90% of the defects implicated in IRDs. To identify multiple patients or families with variants in the same gene and thereby provide extra proof for pathogenicity, worldwide data sharing is needed.