Genome-wide Interaction Analysis Of Exogenous Estrogen In Age-related Macular Degeneration (AMD) Implicates CFD

Abstract

Purpose: : AMD’s multifactorial etiology includes both genetic and environmental risk factors. Previous studies found that women who take exogenous estrogen in the form of hormone replacement therapy (HRT) have reduced risk of AMD. The purpose of this study was to detect novel genes for AMD by accounting for interactions with estrogen in a genome-wide interaction study.

Methods: : Affymetrix 6.0 chipsets were used to genotype 668,238 SNPs in 540 Caucasian women with AMD and 239 unaffected women controls. 2.5 million SNPs were imputed with quality > 0.3 using the HapMap CEU reference panel. History of ever taking HRT was collected with a self-administered questionnaire. Kraft’s two degree of freedom (2df) joint test of main effect and interaction was tested using a likelihood ratio test for each imputed SNP. The likelihood was computed using a logistic regression model comparing a full model with the SNP and HRT main effects, a pair wise interaction term, age, and smoking status, and a restricted model excluding the SNP and interaction terms. Stratified analyses were conducted on interesting SNPs (p<10^-4) to see if genetic effects were restricted to particular exposure strata.

Results: : Although no SNPs met Bonferroni corrected genome wide significance, there were 313 suggestively associated variants (p<10^-4) in 37 genes. These genes were then filtered based on established biological interactions with estrogens and previous associations with ocular development or disease. One gene was previously implicated in AMD. For women who have never taken HRT, a suggestive association with AMD was found for Complement Factor D (CFD, rs3826945: OR=0.30, p=4.39x10^-5).

Conclusions: : CFD is a member of the alternative complement pathway, but its association with AMD has been equivocal. It was recently reported that associations for CFD SNPs and AMD or CFD plasma concentrations were mostly restricted to women (Stanton et al, IOVS 2011). Our results support this finding and suggest exogenous estrogen as a mediator of CFD’s effect. This also demonstrates how genome-wide environmental interaction studies can implicate additional genes for complex traits that have been missed in studies of genetic main effects.

Keywords: age-related macular degeneration • genetics • clinical (human) or epidemiologic studies: risk factor assessment