Novel genes in familial recurrent corneal erosion dystrophy: identification with NGS and characterisation in a zebrafish model

Abstract

Purpose: Corneal dystrophies are a genetically heterogeneous group of disorders. A New Zealand (NZ) family with a unique autosomal dominant recurrent corneal erosion dystrophy is previously described (Vincent et al, Mol. Vis., 2009). We aimed to identify the underlying genetic cause in this family and to characterise the disease mechanism utilising zebrafish.

Methods: 20 members of a 3-generation NZ family were phenotyped and DNA samples collected. Candidate corneal dystrophy genes were excluded as published previously. Genome-wide scanning using a SNP microarray and subsequent exome sequencing of two affected individuals was undertaken. Bioinformatic analysis, protein prediction, and splice site software were used to identify putative pathogenic variants within the linkage region. Known genetic variants were excluded using public databases of human variation, and changes validated with Sanger sequencing, as well as 2 additional families with a similar phenotype (Tasmania and UK). Expression was analysed in human cornea with immunohistochemistry (IHC) and RT-PCR, and in zebrafish with whole-mount in situ hybridisation. Morpholinos were used to transiently knockdown gene expression during zebrafish development, and CRISPR-Cas9 is being utilised to both create knockout lines and to introduce identified genetic variants.

Results: A genome-wide scan of the NZ family identified only one significant peak, on chromosome 10. Exome sequencing and bioinformatics identified candidate variants in three genes (COL17A1, DNAJC9 and FRMPD2) in this region. Segregation was confirmed for COL17A1 (G>A, new splice donor site leading to deletion of 18 amino acids) and DNAJC9 (C>T, missense) in the NZ family. The COL17A1 variant is present in the Tasmanian and UK families. IHC on human cornea indicates COL17A1 in the epithelium and DNAJC9 in the Bowman layer. Both proteins are similarly expressed in zebrafish cornea. Zebrafish lacking Col17a1a and Dnajc9 during development demonstrate phenotypic abnormalities.

Conclusions: The COL17A1 splice variant causing deletion of 18 amino acids is likely to be the causative mutation in our recurrent corneal erosion families; however, DNAJC9 cannot be excluded as a modifier gene. A zebrafish model of this dystrophy is being developed.