Premium
P4‐042: HIGH‐DIMENSIONAL ANALYSIS OF RNA EXPRESSION WITH CORTICAL THICKNESS
Author(s) -
Knol Maria J.,
Roshchupkin Gennady V.,
Muetzel Ryan L.,
Duijn Cornelia M.,
Vernooij Meike W.,
Ikram M. Arfan,
Adams Hieab H.H.
Publication year - 2018
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2018.06.2444
Subject(s) - neurodegeneration , transcriptome , biology , atf4 , rna , population , neuroscience , gene expression , human brain , transcription factor , gene , pathology , medicine , genetics , disease , environmental health
Background: The majority of Late Onset Alzheimer’s Disease (LOAD) GWAS associated SNPs are in noncoding regions of the genome, suggesting regulatory function. In addition, changes in gene expression in LOAD vs. healthy control brains have been described, and several groups reported expression quantitative trait loci (eQTLs) within LOAD associated regions. However, these disease and expression associations represent indirect links that may be attributed to other variants in high linkage disequilibrium with the associated tagging variants. Our goal was to define regulatory elements in the vicinity of the LOAD associated regions that are likely to influence the expression of genes important in LOAD etiology. Towards this goal we applied a bioinformatics approach using public databases for functional annotations. These analyses represent the first step in a global strategy to identify causal variants involved in LOAD. Methods:Genomic regions 60.5Mb surrounding LOAD GWAS-associated SNPs from the were integrated with data from The Roadmap Epigenomics Mapping Consortium for chromatin state segmentation (25-state model) to identify potential active enhancers for specific brain tissue vulnerable in LOAD. This data was aligned with CTCF transcription factor (TF) binding sites determined by ChIP-seq data from ENCODE to identify 3D chromatin structure, looping, between distal enhancer elements and promoter regions. Results:A total of 494 genes map within the defined LOAD GWAS regions. As an example, across the 1Mb region tagged by rs3865444 (CD33) we found 4 enhancer segments (brain hippocampus middle) that include CTCF ChIP-seq peaks. The CTCF signals mapped to the promoters of 4 genes (KLK6, KLK10, IGLON5 and SPACA6), suggesting that these enhancers are likely to regulate these target genes. Conclusions:We describe a valuable resource for testing the hypothesis that causal variants are positioned in regulatory elements of critical LOAD genes, and contribute directly to LOAD susceptibility by affecting gene regulation. These results will inform experimental work for direct validation of regulatory function using model systems such as isogenic iPSC-derived models generated by CRISPR/Cas9 genome editing. Our study is a foundational step in a larger strategy for progressing from GWAS association signals to target genes and specific causal variants for LOAD.