Cross-ancestry genome-wide association analysis of corneal thickness strengthens link between complex and Mendelian eye diseases

Central corneal thickness (CCT) is a highly heritable trait associated with complex eye diseases such as keratoconus and glaucoma. We perform a genome-wide association meta-analysis of CCT and identify 19 novel regions. In addition to adding support for known connective tissue-related pathways, pathway analyses uncover previously unreported gene sets. Remarkably, >20% of the CCT-loci are near or within Mendelian disorder genes. These included FBN1, ADAMTS2 and TGFB2 which associate with connective tissue disorders (Marfan, Ehlers-Danlos and Loeys-Dietz syndromes), and the LUM-DCN-KERA gene complex involved in myopia, corneal dystrophies and cornea plana. Using index CCT-increasing variants, we find a significant inverse correlation in effect sizes between CCT and keratoconus (r = −0.62, P = 5.30 × 10−5) but not between CCT and primary open-angle glaucoma (r = −0.17, P = 0.2). Our findings provide evidence for shared genetic influences between CCT and keratoconus, and implicate candidate genes acting in collagen and extracellular matrix regulation.


NEIGHBORHOOD (National Eye Institute (NEI) Glaucoma Human Genetics Collaboration Heritable Overall Operational Database)
The NEIGHBORHOOD dataset is the meta-analysis of imputed GWAS summary data for 8 independent studies (total sample = 3,853 cases and 33,480 controls) 5 . The study was approved by the Massachusetts Eye and Ear Infirmary institutional review board and all subjects signed consent forms approved by the local IRB prior to enrolling in the study. For all enrollees, primary open angle glaucoma (POAG) cases were defined as individuals for whom reliable visual field (VF) tests show characteristic VF defects consistent with glaucomatous optic neuropathy. Individuals were classified as affected if the VF defects were reproduced on a subsequent test or if a single qualifying VF was accompanied by a cup-disc ratio (CDR) of 0.7 or more in at least one eye. For some cases VFs were not reliable and these cases had either CDR> 0.7 in both eyes are a difference in CDR of at least 0.2 between the eyes. In the OHTS study (one of the 8 NEIGHBORHOOD datasets) an alternative case definition based on progression of optic nerve degeneration was also used 6 . Patients with clinical features of secondary glaucoma based on the examination of the ocular anterior segment were excluded from this study. Elevation of IOP was not a criterion for inclusion; however, 67% of cases did have a history of elevated IOP (≥22 mm Hg) measured in a clinical setting (typically between the hours of 8AM and 5PM) and were classified as high-pressure glaucoma (HPG).
For all datasets genome-wide genotypes were obtained from either Illumina or Affymetrix platforms details have been described in detail elsewhere 5  *Nearest gene (reference NCBI build37) is given as locus label, but this should not be interpreted as providing support that the nearest gene is the best candidate, a list including all the genes +/-200kb of the lead SNP is presented in Supplementary Table 12 NA= SNPs that did not meet our filtering criteria in the Asian-specific meta-analysis (i.e., minor allele frequency (MAF) > 0.01, Info > 0.3 and available in at least two cohorts). *Nearest gene (reference NCBI build37) is given as locus label, but this should not be interpreted as providing support that the nearest gene is the best candidate, a list including all the genes +/-200kb of the lead SNP is presented in Supplementary Table 12 Supplementary Results from the look-up in the keratoconus meta-analysis (933 cases and 5,946 controls of European ancestry). MarkerID, rsID; Chr:bp_hg19, chromosome: base pair NCBI build37, Cyto loci, cytogenic location; Sentinel_CCT_variant, CCT lead SNP; Nearest gene (reference NCBI build37) is given as locus label, but this should not be interpreted as providing support that the nearest gene is the best candidate, a list including all the genes +/-200kb of the lead SNP is presented in Supplementary Table 12;EA; Effect Allele, OR; Odd ratio, SE, standard error of the odds ratio; P, P-value.

Supplementary Figures
Supplementary Figure 1

. Study design and main findings
We conducted a three-stage GWAS meta-analysis of CCT. Stage 1 included 14 studies of European ancestry. In stage 2, we examined the 28 lead SNPs from stage 1 in the Asian-specific meta-analysis. In stage 3, we conducted a cross-ancestry meta-analysis in more than 25,000 individuals to detect additional CCT loci. We then investigated whether the CCT-associated SNPs influence susceptibility to keratoconus and POAG. In total, we identified 44 loci  56

Supplementary Figure 6. Histone modification annotation enrichment amongst CCT associated variants, using European-specific GWAS results
The radial axis represents the fold enrichment in regulatory histone modifications features amongst the CCT European meta-analysis association signals. Eleven histone modification marks based on ENCODE ChIP-Seq data from the human cell-lines legended in the outer circle were available: H2AFZ (marking nucleosomal mobility), H3K27ac (marking active enhancer elements), H3K27me3 (inactive transcription site, poised enhancers), H3K36me3 (facultative and constitutive heterochromatin), H3K4me1(active, primed or poised enhancers), H3K4me2 (Transcription factor binding regions), H3K4me3 (active promoters), H3K79me2 (5' end of gene