Abstract 16448: Expression Quantitative Trait Loci in Human Myocardium
Background: Inter-individual and inter-tissue variations in gene transcript levels are important in mediating disease susceptibility. We performed an expression quantitative trait loci (eQTL) analysis in normal and ischemic human left ventricular myocardium by treating the expression levels of genes as quantitative traits and examining the association with nearby (cis) genetic variants.
Methods: Apical punch biopsies from the left ventricle were taken from 24 patients undergoing cardiac surgery at two time points; immediately after aortic cross clamping (pre-ischemia), and immediately before aortic cross clamp removal (post-ischemia). The Illumina HiSeq 2000 was used to quantify genome-wide mRNA expression. Variants from RNA-Seq data were called using GATK Unified Genotyper. Associations between called variants and gene expression levels were examined with Matrix eQTL using linear regression. Only cis- (local) eQTL with a distance of ±1Mbp from the nearest gene end were considered. False Discovery Rate (FDR) was applied to account for the large number of testing between variants and expressed genes. We examined Response-eQTLs, defined as the genetic variation associated with the delta between baseline and post-ischemia expression levels (quantified as log2 fold-change of post- versus pre- ischemia).
Results: In all, 32,686 variants (25,530 single nucleotide polymorphisms and 7,156 indels) and 16,995 genes were analyzed with a total of 1,594,514 variant-gene pairs tested. Of those, 46 unique genes of the discovered Response-eQTLs had a FDR below 0.1.
Discussion: We demonstrated the presence of genetic variants contributing to inter-individual variation for gene expression change in cardiac surgical patients in left ventricular myocardium in response to ischemia, hereby providing novel evidence of a genetic basis for variable transcriptional response.
- © 2012 by American Heart Association, Inc.