P2-077: Alzheimer's Disease Sequencing Project: Search for Alzheimer's Disease Resilience Genes That May Modify Disease Susceptibility in Specific Apoe Genotype Backgrounds

Common variation at the APOE locus is the strongest known genetic risk factor for Alzheimer’s disease (AD). Specifically, the epsilon 4 (ε4) allele increases risk ∼3-fold when heterozygous and ∼14-fold when homozygous, as compared to homozygous epsilon 3 (ε3) carriers. Moreover, ε3 homozygosity itself confers ∼1.6-fold risk versus epsilon 2 (ε2) homozygotes and ε2/ε3 individuals. Despite this strong genetic liability, some APOE-ε4 carriers never develop symptomatic AD or have a very late age of onset, suggesting these resilient individuals may harbor protective alleles. Understanding the pathogenic mechanism behind such alleles holds great promise for novel therapeutic approaches. Two subsets of non-Hispanic Caucasians were selected from the Alzheimer’s Disease Sequencing Project Case-Control exome sequencing dataset: [1] APOE4 extremes (ε4 allele-carrying cases [n = 492] with age of onset ≤ 65 years, and non-demented controls [n = 625] with age at last assessment ≥ 75 years for ε4/ε4 individuals and ≥ 80 years for ε3/ε4 individuals) and [2] APOE3 extremes (ε3/ε3 cases [n = 496] with age of onset ≤ 70 years and ε3/ε3 non-demented controls [n = 2126] with age at last assessment ≥ 85 years). Sequencing data were subjected to strict quality control. Only low frequency or rare variants annotated as non-synonymous were retained. Primary association analyses were conducted using RvTest, which implements the Score test for single variant and SKAT-O for gene aggregation, adjusting for gender, principal components, sequencing center, and ε4 carrier status. Single variant analyses uncovered several preliminary associations, including TREM2 Arg47His that increased AD susceptibility. Aggregation analyses identified a handful of genes demonstrating preliminary study-wide significance, including TREM2 and KRTAP19-4 in the APOE3 extremes. Our data suggest complex relationships connect variants in these genes with AD outcomes: combinations of risk, protective, and neutral variants likely drive the observed associations with altered disease susceptibility. Similar analyses in an independent cohort are ongoing. If replicated, our findings have pinpointed several novel genes that appear to influence AD risk in specific APOE genotype backgrounds, and suggest they could be prioritized for mechanistic studies and evaluation as novel therapeutic targets.