Refining the genetic landscape of Alzheimer's disease : insights from a novel neuropathological cohort

Abstract

Alzheimer’s disease (AD) is the leading cause of dementia in aging populations, affecting approximately 10% of individuals over the age of 65 and accounting for roughly 70% of dementia cases. While AD is classically characterized by extracellular amyloid-β (Aβ) plaques and intracellular neurofibrillary tangles (NFTs) composed of hyperphosphorylated tau, the neuropathological landscape is markedly heterogeneous. Most patients exhibit additional co-morbid lesions, including cerebral amyloid angiopathy (CAA), phosphorylated TDP-43 (TDP-43) inclusions, Lewy bodies, granulovacuolar degeneration (GVD), and Hirano bodies (HBs), which influence disease progression and complicate diagnosis. Although genome-wide association studies (GWAS) have identified numerous genetic risk loci for AD, these are typically derived from clinically defined cohorts, potentially introudcing phenotypic heterogeneity. This PhD project aimed to elucidate the genetic architecture of both hallmark and co-existing neuropathological features using a deeply phenotyped autopsy cohort of European ancestry. In the first phase, 85 known AD risk variants were evaluated for association with 12 neuropathological traits in 325 individuals. A significant association between a variant in APH1B and NFT pathology was identified and validated through meta-analysis in an independent cohort, with network modelling suggesting a direct link to tau pathology. Additional subthreshold and suggestive associations, including TPCN1 with TDP-43 and UMAD1 with Lewy bodies, were observed. In the second phase, novel GWAS was performed for a total cohort of 414 individuals to allow the discovery of novel risk genes. GWAS for major neuropathological features identified a robust association between KCNQ4 and NFTs. Largely understudied but frequently observed lesions were studied as well. GWAS of HB pathology revealed signals near NRF1 and within GDPD3, with replication in an independent cohort confirming the findings. GWAS of GVD stage identified six genome-wide significant loci (FANCD2OS, XAF1, HIPK1, SMIM21, SPDYE3, and PREX2), with implicated genes predominantly involved in apoptotic and inflammatory pathways. Collectively, this work expands current understanding of the genetic determinants underlying both hallmark and co-morbid AD neuropathologies. By leveraging a uniquely characterized autopsy cohort, it provides novel insights into disease mechanisms and identifies potential targets for therapeutic intervention.