Functional rare genetic variation in Alzheimers disease: an exome-wide association study in the CHARGE consortium. J. Jakobsdottir1, S. J. van der Lee2, J. C. Bis3, V. Chouraki4,5, A. V. Smith1,6, A. L. DeStefano7,5, J. Brody3, N. Amin2, L. J. Launer8, C. A. Ibrahim-Verbaas2,9, S. Choi7,5, A. Beiser7,5, R. Au4,5, P. A. Wolf4,5, O. L. Lopez10, M. A. Ikram2, A. Hofman2, A. G. Uitterlinden2,11, D. Levy5,12,4, C. J. O'Donnell5,12, M. L. Grove13, E. Boerwinkle13,14, A. L. Fitzpatrick15, B. Psaty16,15, S. Seshadri4,5, V. Gudnason1,6, C. M. van Duijn2 1) Icelandic Heart Association, Kopavogur, Iceland; 2) Department of Epidemiology, Erasmus Medical Center, Rotterdam, The Netherlands; 3) Cardiovascular Health Research Unit, University of Washington, Seattle, WA, USA; 4) Boston University School of Medicine, Boston, MA, USA; 5) Framingham Heart Study, Framingham, MA, USA; 6) Faculty of Medicine, University of Iceland, Reykjavik, Iceland; 7) Boston University School of Public Health, Boston, MA, USA; 8) Laboratory of Epidemiology and Population Sciences, National Institute on Aging, Bethesda, MD, USA; 9) Department of Neurology, Erasmus Medical Center, Rotterdam, The Netherlands; 10) Department of Neurology, University of Pittsburgh Medical Center, Pittsburgh, PA, USA; 11) Department of Internal Medicine, Erasmus Medical Center, Rotterdam, The Netherlands; 12) Heart, Lung, and Blood Institute, Framingham, MA, USA; 13) School of Public Health, University of Texas Health Science Center at Houston, Houston, TX, USA; 14) Human Genome Sequencing Center, Baylor College of Medicine, Houston, TX, USA; 15) Department of Epidemiology, University of Washington, Seattle, WA, USA; 16) Department of General Internal Medicine, University of Washington, Seattle, WA, USA.

   Alzheimers disease (AD) is a complex disease with a heritability of ~60%. Relatively common SNPs with modest effects on AD risk have been identified as well as high-risk variants that lead to early onset forms of AD. In CHARGE, a large consortium of population-based studies, we tested the hypothesis that functional coding and splice-site variants modulate the risk of AD. Four cohorts (AGES-Reykjavik Study, Cardiovascular Health Study, Framingham Heart Study, and Rotterdam Study) with 1327 AD cases and 8912 controls of European ancestry were genotyped for the ExomeChip (EC). The EC is a comprehensive genotyping array of functional, mostly rare, variation in the human exome. Of ~240,000 variants on the EC 90% are predicted to alter the protein structure and 45% are predicted to be damaging. Over 62,000 samples were jointly called within CHARGE, which allowed for high accuracy in the rare variant genotype calls. In each study we tested for associations of AD to single and multiple variants, grouped by genes, using score tests and then meta-analyzed the scores. The strongest associations to single SNPs were found to be at common SNPs near the APOE locus, previously established to affect AD risk (rs769449 in APOE, MAF 12%, p=4x10-34). The Sequence-Kernel Association Test (SKAT) was used for the gene-based analysis. We tested ~14,000 genes, containing non-synonymous or splice-site variants of MAF5%. 14,000 tests correspond to p=4x10-6 for a Bonferroni multiple testing correction. The SKAT analysis identified the ERVFRD-1 gene at a suggestive level of exome-wide significance (p=5x10-5, cumulative MAF 2.5%). The most significantly associated SNP in ERVFRD-1 was rs55714642 (MAF 2%, p=7x10-6), which we observed with the same direction of protective effect in all four studies. Associations were also found with the SKAP2 and TM2D3 genes. Due to low cumulative MAF, the association signal in the SKAP2 gene (cumulative MAF 0.06%, p=3x10-8) was driven by three studies while the signal at TM2D3 (cumulative MAF 0.23%, p=5x10-6) was driven by one study enriched for the putative high-risk SNP. -amyloid plaques in the brain are an important characteristic of AD and dementia and all three genes are directly connected to -amyloid precursor or binding proteins, supporting the role of the -amyloid pathway in AD. Further studies, such as replication, sequencing, and functional studies are warranted to firmly establish these newly found associations.

You may contact the first author (during and after the meeting) at