ABSTRACT The APOE ɛ2 and ɛ4 alleles are by far the strongest and most well-established common genetic risk factors for Alzheimer disease (AD). The deleterious effect of ɛ4 on AD risk is dependent on dose, age, sex, and ethnicity, but the protective effect of ɛ2 is is less well studied. We will build upon our previous work aimed at identifying mechanisms underlying the associations of AD with the ɛ2 and ɛ4 alleles in multiple population groups. Recently, we identified a highly significant association of AD with PPP2CB among ɛ2 carriers. PPP2CB encodes the catalytic subunit of protein phosphatase 2A, which is known to dephosphorylate tau protein. Complement C4A and C4B were the most significantly differentially expressed genes in brain tissue from AD cases and controls with the ɛ2/ɛ3 genotype, and we showed that levels of C4b and PPP2CB proteins are significantly correlated. We also identified a novel genome-wide significant association of AD with MGMT that is specific to women lacking ɛ4 (ε4-). Omics analyses of brain-derived data showed significant association of (1) MGMT AD risk alleles with increased MGMT methylation (2) MGMT methylation with MGMT expression and worse AD-related pathology in ε4- women. This proposed project will identify new ε2- and ε4-mediated AD associations using a greatly enlarged multi-ethnic Alzheimer Disease Genetics Consortium GWAS sample and GWAS datasets from European AD genetics consortia which include more than 240,000 subjects. Next, we will identify functional variants in the identified loci in large whole genome whole exome sequencing datasets from diverse populations assembled by the Alzheimer Disease Sequencing Project. We will calculate ethnicity-specific AD polygenic risk scores (PRS) for APOE genotype subgroups and evaluate the moderating effect of each PRS score on AD risk associated with ɛ2 and ɛ4. In a second aim, we will perform proteomic studies in plasma from 2,813 Framingham Heart Study participants to identify proteins whose expression is associated with AD status and measures of cognitive performance in the total sample and APOE genotype groups. We will then investigate methylation and transcriptome profiles in the APOE related loci in 1,391 blood and 952 brain samples from European and African ancestry subjects. APOE genotype-dependent omics signatures derived from these analyses will be validated by testing their association with AD-related neuropathological traits and other proteins measured in brain by multiplex immunofluorescence. Finally, in a third aim we will determine mechanisms through which genes identified by GWAS confer protection against AD pathophysiology using a novel human iPSC-derived human 3D spheroid model of AD that allows testing of multiple brain cell types. We will generate the model by growing induced neurons, astrocytes and microglia separately in 2D cultures, and then combining them in 3D cultures where the integrated neurodegenerative process ensues. We will selectively t...