Multiple novel loci for corneal curvature identified through genome-wide association studies: The CREAM Consortium

Abstract

Purpose: Corneal curvature is an important clinical endophenotype for corneal astigmatism and refractive error. We aimed to identify genetic variants influencing corneal curvature through a large-scale meta-analysis of genome-wide association studies (GWAS).

Methods: The Consortium for Refractive Error and Myopia (CREAM) conducted the largest international GWAS meta-analysis of corneal curvature using a two-stage approach. The discovery stage included 18 European-ancestry cohorts consisting of 28,661 individuals, followed by a replication stage of 14,464 individuals from 12 Asian cohorts. The radius of corneal curvature was measured with an auto-keratometer or IOL master. Study individuals were genotyped using either Illumina or Affymetrix platforms. After stringent quality controls, genotype data in each cohort were imputed to the 1000 Genomes reference panel. The genotype and phenotype association was assessed using linear regression analysis under an additive effect model, adjusting for age, gender and principal components.

Results: We identified a total of 23 loci influencing corneal curvature at the genome-wide significance level (P < 5 x 10^-8) in Europeans; of these, 17 were novel and were also replicated in Asians. The combined analysis of all discovery and replication cohorts revealed an additional 19 novel loci for corneal curvature with low cross-cohort heterogeneity. The most significant locus harbors RSPO1 (P = 2.29 x 10^-29), a gene previously identified by CREAM as influencing axial length. We also confirmed association with four previously identified corneal curvature loci, including PDGFRA (rs7682912, P = 8.24 x 10^-24), FRAP1 (rs3737611, P = 8.16 x 10^-23), CMPK1 (rs60078183, P = 4.61 x10^-14) and RBP3 (rs11204213, P = 7.94 x10^-12). The novel gene loci implicate a number of different pathways but reinforce the likely involvement of cell differentiation and proliferation playing an important role in corneal curvature.

Conclusions: Our study represents the largest multi-ancestry GWAS on corneal curvature, which identified a total of 36 novel loci influencing corneal curvature. The results also provide direct evidence of shared genes, such as RSPO1, that appears to influence both corneal curvature as well as axial length.