Hippocampal transcriptome-wide association study and neurobiological pathway analysis for Alzheimer’s disease

Abstract

Genome-wide association studies (GWASs) have identified multiple susceptibility loci for Alzheimer’s disease (AD), which is characterized by early and progressive damage to the hippocampus. However, the association of hippocampal gene expression with AD and the underlying neurobiological pathways remain largely unknown. Based on the genomic and transcriptomic data of 111 hippocampal samples and the summary data of two large-scale meta-analyses of GWASs, a transcriptome-wide association study (TWAS) was performed to identify genes with significant associations between hippocampal expression and AD. We identified 54 significantly associated genes using an AD-GWAS meta-analysis of 455,258 individuals; 36 of the genes were confirmed in another AD-GWAS meta-analysis of 63,926 individuals. Fine-mapping models further prioritized 24 AD-related genes whose effects on AD were mediated by hippocampal expression, including APOE and two novel genes (PTPN9 and PCDHA4). These genes are functionally related to amyloid-beta formation, phosphorylation/dephosphorylation, neuronal apoptosis, neurogenesis and telomerase-related processes. By integrating the predicted hippocampal expression and neuroimaging data, we found that the hippocampal expression of QPCTL and ERCC2 showed significant difference between AD patients and cognitively normal elderly individuals as well as correlated with hippocampal volume. Mediation analysis further demonstrated that hippocampal volume mediated the effect of hippocampal gene expression (QPCTL and ERCC2) on AD. This study identifies two novel genes associated with AD by integrating hippocampal gene expression and genome-wide association data and reveals candidate hippocampus-mediated neurobiological pathways from gene expression to AD.

RNA Dynamics in Alzheimer’s Disease

“…GWAS and TWAS analyses have identified around 50 disease-associated loci in AD, both with familial early onset as well as late onset AD [ 45 , 116 , 117 , 118 , 119 , 120 ]. Many of these genes present disease-associated changes in their isoforms due to changes in AS and/or APA of their transcripts, including presenilin 1 and 2 ( PSEN1 and PSEN2 ), microtubule-associated protein Tau ( MAPT ), amyloid beta precursor protein ( APP ), phosphofructokinase ( PFKP ), phosphatidylinositol binding clathrin assembly protein ( PICALM), clusterin ( CLU ), protein tyrosine kinase 2 beta (PTK2B), and NDRG family member 2 (NDRG2) [ 35 , 45 , 121 , 122 ].…”

Multivariate Analysis and Modelling of multiple Brain endOphenotypes: Let’s MAMBO!

“…They often rely on the assumption that, under the null hypothesis, summary Z-statistics from univariate tests performed on individual phenotypes have an asymptotic multivariate normal distribution with correlation equal to the phenotype correlation matrix. In recent years, this strategy has become very popular, being employed for the analysis of multiple phenotypes in a wide variety of scenarios, including IG studies [82] , [83] , [84] . The main advantage of these approaches is their ability to carry out fast GWAS analyses across many phenotypes and with very large sample sizes, without requiring complex and time-consuming access to individual-level data.…”

Role of primary aging hallmarks in Alzheimer´s disease

“…Remarkably, the telomeres of hippocampal neurons become shorter in AD, and much evidence supports its correlation with elevated levels of oxidative stress [102,108]. Interestingly, telomeres are not only favoured by oxidative attack due to their unique sequence but also less efficiently repaired when compared to non-telomeric damage [109][110][111][112].…”