Interrogating the protective effect of African APOE alleles on Alzheimer's disease risk through pleiotropy

Project Summary/Abstract Alzheimer’s disease (AD) is a highly heritable trait, where much of the phenotypic variation is explained by genetic variation. A significant proportion of AD risk is explained by ancestry and APOE genotype. We and others have shown that ancestry differences across the genome (global ancestry) and at the APOE locus (local

ancestry) are significantly associated with risk of AD in admixed populations. African-derived APOE alleles are associated with reduced risk of AD after controlling for global ancestry and APOE ε2/ε3/ε4 genotype in African American, Caribbean Hispanic, and Puerto Rican samples. This local ancestry effect may be driven by

variants on African-derived haplotypes, European-derived haplotypes, or both. The focus of this study is to determine which variants on these haplotypes may explain the association between African-derived APOE alleles and reduced risk of AD and to characterize their potential consequences at the cellular level.

APOE is involved in many pathways relevant to AD, including lipid biosynthesis and two ends of the efferocytosis pathway: recognition/engulfment and adaptation. This may explain why the ε2 and ε4 alleles associated with AD risk are associated with many phenotypes, including peripheral lipid traits (lipids) and

inflammatory markers like C-reactive protein (CRP). Both ε2 and ε4 are associated with unfavorable lipid and CRP profiles that are themselves associated with AD risk, AD biomarkers, and other causes of morbidity and mortality. As with AD, the strength of association between ε2 and ε4 and both lipids and CRP vary with

ancestry. Joint consideration of AD with lipid and inflammation-related traits is expected to reveal additional pathophysiological information not identifiable in separate analyses. Therefore, we propose to investigate local ancestry effects at APOE and AD risk within the Alzheimer’s Disease Sequencing Project (ADSP) whole genome sequence (WGS) data, along with their effects on

harmonized lipid traits and CRP levels paired with WGS data from the Trans-Omics for Precision Medicine (TOPMed) project among admixed participants who were not ascertained for AD. We will estimate local ancestry probabilities, then test for association between global (across the genome) and local (at APOE)

ancestry using two alternative approaches. Contrasting ancestry-specific APOE haplotypes within and between these studies will allow us to nominate variants with pleiotropic vs. trait-specific effects. Our bioinformatics pipeline will link variants to their potential consequences using gene expression and regulatory

data tailored to AD, prioritizing the appropriate model systems for future studies. This work will facilitate therapeutic target selection that avoids negative effects on lipid and inflammatory phenotypes associated with significant morbidity and mortality.