Genome Sequencing of a Familial Keratoconus Cohort

Abstract

Purpose : Keratoconus (KC) is a complex, progressive corneal ectasia with a large proportion of inheritance undetermined. Our recent genome wide association study (GWAS) identified risk factors which explain 12.5% of heritability in Europeans, with a moderately predictive value (AUC = 0.737) when tested in a small, ethnically matched group (Hardcastle et al (2021)). Here we aim to further understand disease aetiology and complement our GWAS by performing genome sequencing on a familial KC cohort, to identify rarer, deleterious variants of large effect to increase the predictive value of genetic screening.

Methods : Seventeen families comprising two or more affected individuals were recruited to this study, based on a diagnosis determined though clinical signs of corneal distortion and corneal thinning, later confirmed by corneal imaging. Whole genome sequencing (WGS) was preformed using Illumina NovaSeq 6000. Data was aligned using Burrows-Wheeler Aligner and annotated using ANNOVAR (GRCh37, Novogene). Variants were filtered to identify rare (MAF<0.01; gnomAD v1.3.2) single nucleotide variants and indels shared between affected individuals. Rare variants were assessed for potential pathogenicity (CADD, PolyPhen, SIFT, SpliceAI), biological relevance, effect on gene expression or function, and expression in relevant corneal tissue and cells. Rare and potentially disease-associated variants were then compared to existing associated loci. Variant validation was performed by Sanger sequencing.

Results : Family 1 is a large family comprising nine affected individuals, with KC segregating in an apparent autosomal dominant fashion. Three heterozygous, exonic, rare variants were identified on 15q22.2-q22.31 that co-segregate with KC in Family 1. Our GWAS has identified common variants within this region that are significantly associated with KC (p-value = 1.75E−08, rs76194223). Further analysis is underway, to determine if we have identified the causative variant(s) that underlie disease in Family 1.

Conclusions : Common variants associated with KC could reflect the effect of one or multiple rarer variants in linkage disequilibrium with the genotyped or inferred SNP(s). Here we are harnessing familial WGS data to identify rarer variants with potentially large effects that confer substantial risk of disease. Integrating GWAS and familial data, has the potential to identify causative variants underlying existing risk loci that are responsible for KC.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.