Abstract
Purpose :
Fifty-two nuclear-encoded single nucleotide polymorphisms (nSNPs) in 34 loci have been associated with AMD. However, main effects at these loci do not explain all phenotypic variation. Interactions with the mt genome may explain some of the remaining effect. A nonsynonymous mtSNP, A4917G, in MT-ND2 has been associated with AMD (Canter et. al, 2008). Thus, mtSNP A4917G-nSNP interactions were explored in a genome-wide analysis to identify loci not previously associated with AMD.
Methods :
In the International AMD Genomics Consortium (IAMDGC) exome chip dataset, 17,810 controls and 16,122 advanced AMD cases of European descent were genotyped on the Illumina HumanCoreExome array. Imputation using MINIMAC and the 1000 Genomes Project Phase I reference panel produced genotypes at 12 million nSNPs for analysis. This was reduced to 3.9 million nSNPs by pairwise linkage disequilibrium (LD) pruning (one SNP of a pair with r2 > 0.8 was kept). A two degree of freedom test (Kraft et al., 2007) was used to test the joint effect of each nSNP and mtSNP A4917G. It compares two logistic regression models: a full model containing mtSNP, nSNP, and mtSNP-nSNP interaction terms and a partial model lacking nSNP and mtSNP-nSNP interaction terms. The models adjusted for covariates, including the first two principal components from analysis of population stratification.
Results :
All known AMD risk loci generated genome-wide significant joint test results. One novel locus, TRPM1 (transient receptor potential cation channel, subfamily M, member 1), had genome-wide significant joint effects at two intronic nSNPs in strong LD: rs6493454 and rs7182946 (minimum p=1.7x10-8). Tests of interaction for these nSNPs and mtSNP A4917G were nominally statistically significant (p<0.01). Stratified analysis by mt allele showed that significant association was limited to mtSNP 4917A (major allele) carriers (minimum p=4.4x10-9, OR=1.11, 95% CI=1.07-1.15). No association was detected in mtSNP 4917G (risk allele) carriers (minimum p=0.25, OR=0.94, 95% CI=0.85-1.05).
Conclusions :
The genome-wide analysis of mtSNP A4917G-nSNP interactions on AMD has detected a locus, TRPM1, not previously identified by genome-wide association studies in this dataset. This underscores the need to consider gene-gene interactions, particularly with mtSNPs, to better understand the genetic architecture of AMD.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.