Authors: Logan Dumitrescu, Emily R. Mahoney, Shubhabrata Mukherjee, Michael L. Lee, William S. Bush, Corinne D. Engelman, Qiongshi Lu, David W. Fardo, Emily H. Trittschuh, Jesse Mez, Catherine C. Kaczorowski, Keith F. Widaman, Rachel F. Buckley, Michael J. Properzi, Elizabeth C. Mormino, Hyun‐Sik Yang, Shea J. Andrews, R. Elizabeth Sanders, Neha S. Raghavan, Badri N. Vardarajan, Gerard D. Schellenberg, Carlos Cruchaga, Jonathan L. Haines, C. Dirk Keene, Andrew J. Saykin, Eric B. Larson, Reisa A. Sperling, Richard Mayeux, David A. Bennett, Julie A. Schneider, Paul K. Crane, Angela L. Jefferson, Timothy J. Hohman,
Published: 2020-12-07
DOI: 10.1002/alz.043259
Source: Full article
AbstractBackgroundIdentifying genetic factors that provide resilience against the clinical consequences of Alzheimer’s disease (AD) pathology is likely to accelerate the development of novel therapeutics. Sex differences in AD prevalence, neuropathological presentation, and clinical progression suggest that exploring the sex‐specific genetic architecture of resilience to AD may be a critical first step towards the characterization and development of precision interventions.MethodWe completed sex‐stratified and sex‐interaction genome‐wide association studies (GWAS) of resilience across 5,109 non‐Hispanic white individuals from two autopsy cohorts (ACT and ROS/MAP) and two positron emission tomography (PET) cohorts (ADNI and A4). A continuous measure of resilience was quantified using latent variable modeling and represented better‐than‐expected cognition for the given level of amyloid in the brain. The final dataset included 2,130 males (77 ± 9 years) and 2,979 females (77 ± 10 years), the majority of whom were cognitively normal (73% of males; 76% of females). GWAS of resilience were performed in the combined autopsy dataset and combined PET dataset individually and then meta‐analyzed. Covariates included age and three principal components. Analyses were run in all individuals as well as in the subset of cognitively normal individuals.ResultAn intergenic variant on chromosome 10 (rs827389, MAF = 0.46) showed a female‐specific genome‐wide significant association with resilience in cognitively normal females (β = 0.08, p = 7.6 × 10−9) but not in males (β = ‐5.3 × 10‐3, p = 0.77; sex interaction p = 1.4 × 10−4). This variant is a modest brain eQTL for the KIN gene that is implicated in DNA repair. We also observed a strong sex‐interaction effect among all individuals on chromosome 3 (rs113968105, MAF = 0.10, p = 7.5 × 10−8), in which the minor allele was associated with higher resilience in males (β = 0.10, p = 4.2 × 10−7) but lower resilience in females (β = ‐0.04, p = 0.03). This variant is an eQTL for three genes, including the acetylcholinesterase binding gene COLQ.