Frequency of familial Alzheimer’s disease gene mutations within the Alzheimer Disease Sequencing Project (ADSP)

Background: Mutations within the amyloid precursor protein (APP), presenilin-1 (PSEN1), and presenilin-2 (PSEN2) genes are known to cause familial Alzheimer’s disease (AD). Our goal is to examine the distribution of clinical variants in these genes within the ethnically diverse participants in the Alzheimer Disease Sequencing Project (ADSP) whole genome sequence (N = 4789) and whole exome sequencing (N = 20,158) data. Method: We aggregated mutations reported in APP, PSEN1, and PSEN2 from the Alzheimer’s disease and Frontotemporal Dementia Mutation Database, ClinVar, and the Online Mendelian Inheritance in Man, and selected rare variants (minor allele frequency < 1%) described as pathogenic and related to AD or dementia yielding 304 variants (53 in APP, 233 in PSEN1, 18 in PSEN2). We counted the presence of these clinically implicated variants in the currently available ADSP data. Result: Among 22,558 ADSP subjects passing quality control filters we observed 16 PSEN1, 5 PSEN2 and 6 APP variants. We identified 48 individuals (13 controls, 30 AD cases, 5 missing AD diagnosis) that were heterozygous for one of these variants. An additional AD case was a compound heterozygote for two APP variants. Out of the total sample, 0.06% of African Americans (AA), 0.21% of whites (non-Hispanic white or white with ethnicity unknown), and 0.33% of Hispanics (HI) were carriers. In contrast, AA, white, and HI represented 21%, 63%, and 15% of the sample, respectively. The carriers included two pairs of second-degree relatives: one Hispanic pair and one pair consisting of one Hispanic and one white/ethnicity unknown individual. Conclusion: A reported mutation in APP, PSEN1 or PSEN2 was observed in 0.22% of ADSP participants, with the lowest carrier rate in AA. Studies in other diseases indicate that disparities in inclusion in genomic sequencing may result in misclassification or higher rate of uncertain significance for patients in underrepresented populations (e.g. Manrai et.al. 2016, Spratt et.al. 2016). It is unknown if the current difference in observed carrier rate by ancestry reflects study design, a true difference in frequency of AD causal APP, PSEN1, and PSEN2 mutations across populations, or limitations in identified pathogenic mutations due to populations historically included in genetic studies of AD.