Premium
[O2–13–01]: GENETICALLY REGULATED TRANSCRIPTOMIC STUDY OF ALZHEIMER's DISEASE YIELDS MECHANISTIC INSIGHTS
Author(s) -
Raj Towfique
Publication year - 2017
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.2017.07.216
Subject(s) - transcriptome , biology , gene , genome wide association study , genetics , gene expression , genetic association , genome , gene expression profiling , computational biology , genotype , single nucleotide polymorphism
Background: Genome-wide association studies have identified many genomic regions associated with late-onset Alzheimer’s disease (AD) but elucidating the underlying genes and their regulatory mechanisms remains a challenge. Large-scale transcriptomic datasets have enabled identification of variants regulating gene expression in specific tissues. These data can be used to impute tissue-specific genetic expression levels from genotypes in larger samples, which can be tested to identify potentially novel associated genes. Methods: We performed a transcriptome-wide association study for AD by jointly analyzing gene expression data from 7,088 individuals across 61 tissues including 1,028 brains (of which 461 are AD) from the AMPAD and CommonMind Consortium, peripheral blood, primary immune cells (monocytes, T-cells, neutrophils, and B-cells), and 44 tissues from GTEx including 10 brain regions. We performed: 1) heritability for each expression unit; 2) cross-validation of five predictive modeling techniques. Finally, using the heritable expression weights, we tested tissue-specific total gene expression and alternative splicing association to AD. Results: We identified 31 transcriptome-wide significant genedisease associations (p<10), of which 13 did not overlap a genome-wide significant variant. Majority of the associations (14 genes) stemmed from monocytes or macrophages, highlighting the importance of these cells in AD etiology. The genes in myeloid cells include SPI1, CD33, MS4A/6A, PILRA and other genes not in AD loci (PCDHA5/11, C1RL, UBE2D2, OAS1, TNFRSF21, GPR141, ITGB6 and PFKFB2). In addition, we identified 9 associations from differentially spliced mRNA in brains that were independent of total gene expression. These include genes in AD GWAS loci (CLU, PTK2B, PICALM, PILRB and MTCH2) and four novel associations genes (TBC1D7, AP2A1/2, and MAP1B). Conditional analysis of the MHC identifies a novel AD gene (complement component 4A). We replicated these associations via internal cross-validation, as well as in an independent cohort of 531 brains. Conclusions: Our study highlights several AD susceptibility genes in myeloid cells as drivers of at least part of the genetic component of AD. Our results also suggest that alternative mRNA splicing in the brain is an important source of diseaserelevant variation. Overall, this study provides a foundation for further mechanistic studies that will elucidate the molecular drivers of AD. O2-13-02 DECODING LOAD-GWAS DISCOVERIES: GENE EXPRESSION CHANGES IN SINGLE NEURONS THROUGHOUT LOAD PATHOLOGICAL PROGRESSION