Premium
Transcriptome‐based analyses of EADB reveal new insights into Alzheimer's disease
Author(s) -
Küçükali Fahri
Publication year - 2020
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 - 10.1002/alz.037974
Subject(s) - rna splicing , genome wide association study , biology , transcriptome , alternative splicing , genetics , expression quantitative trait loci , gene , phenotype , computational biology , single nucleotide polymorphism , gene expression , exon , genotype , rna
Background For the majority of Alzheimer’s disease (AD) genetic risk loci, the mechanism of action of variants or genes in AD pathophysiology remains unidentified. One way to approach this is to investigate the downstream effects of AD‐associated variants on molecular phenotypes such as expression and splicing. By integrating EADB and AD‐relevant brain expression and splicing data we performed the largest transcriptomic investigation of AD GWAS to date. Method We generated both eQTL and sQTL catalogues on six AD‐relevant brain regions by integrating whole‐genome sequencing (WGS) genotypes of a total of 1067 individuals with 1552 brain RNA‐Seq samples after quality control, available at AMP‐AD Knowledge Portal. We quantified expression and splicing using GTEx and Leafcutter pipelines, and used enhanced FastQTL to perform linear regression of WGS genotypes against normalized molecular phenotypes with adjustment for sex, genetic and transcriptomic principal components. We investigated the overlap between AD‐associated EADB GWAS variants and e/sQTL variants. We then generated expression and splicing weight panels using these catalogues and conducted expression and splicing TWAS for AD on summary statistics of the EADB GWAS including 100,000 Europeans. Long‐read nanopore sequencing was performed using 60 lymphoblastoid cell lines, 18 frontal cortex BA10 and 16 hippocampal cDNA samples on a MinION platform to further study complex splicing events. Result We identified 1225 genome‐wide significant variants that overlap with a significant eQTL (associated with 175 genes) and 525 genome‐wide significant variants that overlap with a significant sQTL (associated with 260 splice junctions) in at least one of the six catalogues created. Expression and splicing TWAS confirmed known AD‐associated genes and highlighted new genes in novel loci, including a gene on chromosome 10 located within one of the most significant novel loci of EADB GWAS, whose TWAS‐associated splice junctions ( P = 7.24 × 10 −10 and 7.33 × 10 −10 in MayoRNASeq TCX and ROSMAP DLPFC TWAS) are hinting at novel cryptic exons, which we confirmed by Nanopore sequencing, within its functional domain. Conclusion Our results indicate that in addition to eQTLs, part of AD GWAS loci can be explained by sQTLs that are understudied in the field; and together these catalogues provide a valuable resource for understanding the genes involved in AD.