Integrative polygenic statistical approaches for genetic and epigenetic data to elucidate the origins of Alzheimer’s disease
Final Report Abstract
Alzheimer’s disease (AD) is a devastating condition, currently without effective cure, treatment or prevention that robs people of their memories and abilities. It is well known that, beside age and sex, genetic and epigenetic factors, along with environmental factors, lifestyle choices and psychological factors, have an impact on AD risk. However, how all these risk factors interact and thereby cause AD remains unclear. The aim of this research project was to extend and combine existing statistical approaches to improve the polygenic integrative analysis of genetic and epigenetic risk factors across multiple correlated phenotypes. These approaches were applied to data generated from post-mortem human dorsolateral prefrontal cortex from the Religious Order Study and Rush Memory and Aging Project (ROS/MAP) to improve the understanding of the interplay between genotypes, DNA methylation and depression leading to AD pathology. We extended a kernel-based approach, which was originally developed to test associations with rare genetic variants, to allow for DNA methylation data. This approach was then applied to brain tissue-based DNA methylation and cognitive trajectory in 636 participants from the ROS/MAP cohorts, which resulted in the identification of a novel gene (CLDN5) that influences cognitive trajectory beyond traditional neuropathologies. Findings were validated in independent postmortem frontal cortex methylation samples, and linked to gene expression and genotype data. In addition, we found epigenome-wide significant associations between brain-tissue-based DNA methylation and late-life major depressive disorder (MDD) in the same cohort. The most significant and robust association was found with altered methylation levels in the YOD1 / PFKFB2 loci. This association was not confounded by dementia or genetic risk factors and significant in both the single site and region-based analysis. Overall, our results highlight the phenotypic complexity of late-life depression, cognitive decline and AD pathology, which has not been considered by most previous studies. Particularly, we showed that methylation patterns in CLDN5 influence cognitive trajectory beyond traditional neuropathologies with varying effects for different cognitive domains, and identified altered methylation in the YOD1, UGT8, FNDC3B and SLIT2 loci as new epigenetic factors associated with late-life depression, an association which was not confounded by dementia. In addition, we adjusted our models for genetic risk factors (single genotypes and polygenic risk scores) to show that our epigenetic associations were not confounded by genetic risk factors. These findings have implications for future (epi)genetic association studies as well as other association studies aiming to identify risk factors for late-life depression, cognitive decline and AD pathology. Detailed phenotyping and the application of advanced statistical models that account for the complexity of these phenotypes are crucial to identify risk factors for neuropsychological outcomes
Publications
- Brain DNA Methylation Patterns in CLDN5 Associated with Cognitive Decline. bioRxiv 2019
Hüls A, Robins C, Conneely KN, Edgar R, De Jager PL, Bennett DA, Wingo AP, Epstein MP, Wingo TS
(See online at https://doi.org/10.1101/857953) - Association between DNA Methylation Levels in Brain Tissue and Late-Life Depression in Community-Based Participants. medRxiv 2020
Hüls A, Robins C, Conneely KN, De Jager PL, Bennett DA, Epstein MP, Wingo TS, Wingo AP
(See online at https://doi.org/10.1101/2020.04.21.20074021) - Methodological challenges in constructing DNA methylation risk scores. Epigenetics. 2020 Jan - Feb;15(1-2):1-11
Hüls A, Czamara D
(See online at https://doi.org/10.1080/15592294.2019.1644879)
