Abstract
Purpose: :
Keratoconus is a bilateral corneal ectasia which commonly affects working-age individuals. The prevalence is approximately 1 in 2000. Although many cases are sporadic, the aetiology is likely multifactorial with a genetic predisposition. Linkage studies have identifed several loci involved, however no genes have been associated with common keratoconus. We have conducted a genome-wide association study utilising equimolar DNA pools to identify novel keratoconus genes.
Methods: :
DNA from 100 patients with severe keratoconus and 216 elderly controls was accurately quantitated. Equimolar amounts of DNA from each patient were pooled into single tube for the cases and a single tube for controls. Both pools were typed on the Illumina Human 1M Duo array on multiple arrays each. Following normalisation and quality control checks the allele frequencies for cases and controls were estimated and tested for association. Associated SNPs were then genotyped in each sample individually as well as in an independent cohort of 100 cases and controls.
Results: :
After rigorous quality control, 23 SNPs survive Bonferroni correction (ie. p<5x10-8) for the 1 million polymorphisms typed. Nine of these SNPs cluster in a region on chromosome 11 not previously reported to be linked to keratoconus. Seventy-seven SNPs were followed up in individual samples. The chromosome 11 region did not retain the same level of association observed in the pooled study, but was found to be associated in both the original Australian sample (p=0.0002) and the independent but smaller sample from Northern Ireland (p=0.02). There are no annotated genes in this region. In total, 25 individually typed SNPs were associated with a p-value < 9x10-3. Genes of interest (p<9x10-5 in the discovery sample) include CDH11, NUB1 and COL27A1. SNPs in HGF were found to be associated in both the discovery sample (p=0.005) and the independent replication samples from Australia and Northern Ireland (p=0.008).
Conclusions: :
Several SNPs appear to be associated with keratoconus and are currently undergoing further testing to validate and replicate the findings. These genes are plausible functional candidates for this common disease. Understanding the genetic causes of keratoconus may lead to both predictive testing and novel therapies.
Keywords: keratoconus • genetics