Identifying the genetic causes of paediatric cataract in Australian families

Cataracts are a clouding of the normally transparent lens of the eye. They result in varying degrees of visual impairment due to light scattering that occurs as light passes across the cataract affected lens. Paediatric cataracts can be present at birth (congenital) or develop anytime through childhood up to the late teenage years. Paediatric cataract has an estimated prevalence of 2.2-13.6 per 10,000 births worldwide and approximately one in four cases has a genetic basis to their disease. Inherited paediatric cataracts can occur as an isolated phenotype, in combination with other ocular features or as part of a syndrome. To date, over 50 genes and loci have been linked to isolated paediatric cataracts and over 200 have been linked to syndromic forms. Studying the genetic causes of paediatric cataracts will enable a better understanding of the molecular mechanisms that underpin a healthy lens and in doing so, create opportunities to improve preventive care and treatment options for all cataract cases. The identification of new genes and mutations will improve molecular diagnostic rates, with current known genes accounting for only 60-70% of familial cases. The overarching aim of this project is to identify the genetic basis of disease in unsolved families from Australia’s largest repository of paediatric cataract DNA samples. This aim was achieved through three complementary studies: a gene screen of known isolated paediatric cataract genes; novel gene discovery using a combination of linkage analysis and massively parallel sequencing; and extended investigation of variant pathogenicity using in vitro and in vivo methodologies. The gene screen aimed to identify putative disease-causing variants in known genes, to provide a molecular diagnosis for numerous families, and investigate variants impacting non-coding regions of known genes. Thirty-three families were investigated, containing one to seven individuals with DNA available and a maximum of four affected cases. Whole genome sequencing (WGS) was performed on one affected individual in each family. Rare (MAF ≤0.01) and predicted pathogenic SNPs and indels were filtered from the coding and non-coding regions of 50 isolated congenital cataract genes. Sanger sequencing was used to assess variant segregation in the family. Positive control families CSA182 and CSA196 had their respective known coding BFSP1 c.1124delA and non-coding FLT c.-168G>T variants identified, confirming the effectiveness of filtering parameters used. Segregating coding variants were observed in 29% (9/31) of families, with 89% (8/9) of those classed as pathogenic or likely pathogenic. Five of these variants were identified in connexin genes GJA3 and GJA8, a recurrent variant was observed in CRYAA, and variants in COL4A1 and LEMD2 are likely to be disease causing and may cause additional pathologies. Incomplete penetrance was identified for coding variants in six families, with variants in the BFSP2 (CQLD130) and MIP (CRCH4) genes strong candidates. Additionally, variants in CRYBB1 (CTAS34) and CRYBA2 (CSA196) are hypothesised to be altering disease severity in addition to other causative variants identified in those families. Non-coding variants were identified in five families. A putative disease-causing c.-383G>T variant in PAX6 (CTAS34) resides in the gene’s P1 promoter and, in family CSA169, a deep intronic NHS variant was identified, with both prioritised for further investigation. The remaining unsolved families form a cohort available for screening novel genes and for novel gene discovery. The second study utilised genome-wide parametric linkage analysis and massively parallel sequencing for novel gene discovery. Three families were investigated, each containing DNA for 16 (CRCH24) or 7 (CRCH65 and CSA93) individuals, with a minimum of 5 affected cases. Regions with a LOD>1 were investigated for rare, predicted pathogenic variants in WGS of an affected individual from each family. In the largest family, CRCH24, linkage analysis eliminated 97.74% of the genome from investigation and structural variant calling identified a 2.8kb intronic NHS deletion, that overlaps with a comparable deletion in a published family with isolated congenital cataracts. This deletion is highly likely to be causing disease and is hypothesised to remove a deep intronic feature required for correct transcript splicing, a novel mechanism for cataracts that has been prioritised for further investigation. In families CRCH65 and CSA93 whole exome sequencing in additional cases was used to filter SNPs and indels. Linkage analysis reduced the investigated genomic areas (LOD>1) to 1.87% and 3.88% in families CRCH65 and CSA93 respectively. In family CRCH65, a PRR12 variant causing a p.Thr74Pro residue change is the leading candidate. In family CSA93, variants were identified in PRX and IPO11, genes that both result in an altered lens morphology in murine knockout models. Investigation into the role of PRX and IPO11 as putative cataract susceptibility genes is recommended. Investigation of structural variants are also recommended as a future step before further conclusions can be made in families CRCH65 and CSA93. The third study utilised in vitro and in vivo techniques to investigate candidate variants identified in a previously studied family (CRVEEH66) with significant linkage at Xq24-25 (LOD=2.53), containing a truncating 127kb deletion of PGRMC1, and suggestive linkage at 1q42.2-43 (LOD=2.44), containing an ERO1B c.662C>T (p.Ala221Val) variant. The ERO1B variant is not highly conserved and a minigene splicing assay was unable to detect altered splicing in lens epithelial cells despite in silico predictions. Morpholino (MO) antisense oligo knockdown of PGRMC1 or ERO1B was performed in zebrafish (Danio rerio). Loss of either gene in early development lead to cataract formation. Cataract formation from ero1b gene knockdown (p=