Determining the genetic basis of Keratoconus and implications for treatment

Keratoconus (KC) is a non-inflammatory, progressive thinning of the cornea resulting in a conically shaped protrusion. Its incidence in the population is 1/2000. There are known environmental causes, but familial aggregation, bilateralism and twin studies point to a genetic component, which is poorly understood. This project aims to identify these genetic causes by studying a cohort of familial cases and KC and wild type (WT) corneal tissue. The family cohort consisted of 15 multiplex recessive KC families from consanguineous and endogamous ethnic backgrounds. DNA from at least one family member was subjected to autozygosity mapping and whole exome sequencing (WES). Tissue from 6 sporadic KC and 5 WT corneas was analysed by RNA-seq. Genomic DNA from the 6 KC patients who donated their corneas was also subjected to WES. Autozygosity mapping and WES did not identify clear pathogenic alleles, though WES alone yielded long lists of variants segregating in families. An underpowered enrichment test on the WES results also yielded lists of variants potentially implicated in KC. RNA-seq generated a transcriptome profile of the normal anterior human cornea, and highlighted putative new cornea-specific transcripts. The differential expression (DE) data between KC/WT corneas generated lists of significantly DE genes. The combination of all these analyses identified 5 genes, FLNB, ITGB4, KIAA0100, LAMA5 and PCDH1, harbouring coding variants that cosegregated with KC in the families, enriched in variation in the cohort and significantly down-regulated in KC corneas. These are now strong candidates to harbour variants increasing susceptibility to KC. The work describe shows that recessive alleles of large effect are either rare or non-existent in the cohort studied, suggesting that KC is indeed a genetically complex disease. Combining the datasets highlights variants in 5 large structural proteins requiring further analysis to confirm their involvement in KC pathogenesis.