Whole Exome Sequencing in Hereditary Ocular Disease Leads to Recognition of Novel and Recurrent Disease-Causing Mutations: Pros and Cons

Hereditary ocular diseases show an extreme genetic and phenotypic heterogeneity, ranging from mild retinal dysfunctions to severe congenital forms of blindness. Most of the many genes associated with these diseases encode signaling and structural proteins are involved in the developing eye. Another category of serious visual impairments concerns the degeneration of neurons as photoreptors, related to the various forms of retinitis pigmentosa. All modes of Mendelian inheritance occur and many are sporadic cases. Identification of the underlying genetic basis for families and affected individuals, understanding genotype-phenotype correlations and developing therapeutic approaches are therefore highly challenging. In this study whole exome sequencing (WES) was performed on Hiseq2500 platform (Illumina) as a powerful strategy for assessment of multiple candidate disease genes (> 450) at the same time in 11 unrelated families (16 patients) with different eye defects, including microcornea/ coloboma / microphthalmia / anophthalmia, Axenfeld-Rieger syndrome, Retinitis Pigmentosa, Leber congenital amaurosis and Stargardt macular degeneration. Afterward, Sanger sequencing was performed to confirm and determine whether any of the candidate variants co-segregated with the disease phenotype in the families. We evaluated the diagnostic yield, the spectrum of clinical referrals, the challenge of variants’ interpretation and the genetic heterogeneity of such diseases.Our data indicate that this approach enables us to genetically diagnose approximately 80 % of the patients (n = 13) with variant(s) in known disease-associated genes. We revealed four pathogenic variants in RAB3GAP1 (p.(Tyr958*), CHD7 (p.Ala1347Glnfs*25), KCNK9 (p.Gly266Arg), CDH23 (p.Leu1343Phe) and TULP1 (p.Gly266Val, four of them are novel and not reported in the literature or dbSNP: RAB3GAP1, CHD7, CDH23, and TULP1 . In addition, we identified 6 known disease-associated variants, previously reported, in USH2A (p.Asp347Gly & p.Leu3606Pro), FOXC1 (p.Asp261Argfs*45), STRA6 (p.Arg655His), GUCY2D (p.Pro130Leufs*36), and ABCA4 (p.Ser1696Asn). The identified mutation spectrum involved novel variants and previously described recurrent mutations Further, we pointed out two variants probably explaining the abnormal ocular phenotype in: TULP1 (c.823-17G>C) and CDH23 (p.Cys1045Phe). The vast majority of mutations have not been reported in the Italian population (only USH2A; p.Asp347Gly) .We also identified a novel phenotype for mutations in KCNK9. In one family, identified phenotypes were different from the previously reported clinical findings associated with the causative gene in STRA6.WES can rapidly identify variants in various families affected by different forms of hereditary ocular diseases in contrast to Sanger sequencing of candidate genes. In fact the genetic and phenotypic heterogeneity of eye diseases impair straightforward genotype-phenotype relationships . Without WES, we would not have been able to arrive at a molecular diagnosis in 13 cases, even by using commercial panels of genes involved in eye defects that, although continuously upadated, do not included all the possible causative genes.We discussed the challenges experienced in data analysis and showed examples of cases where the detection of the causative variant required further investigation due to the low coverage of some genes, or would be valued unlikely applying the standard WES analysis. For instance, in two families, we were not able to detect mutations in genes currently associated with inherited eye diseases in humans or in animal models.