AD is the sixth leading cause of death in the United States and national direct and indirect annual costs of caring for individuals with AD are at least $100 billion. The majority of cases (90-95%) are late-onset AD (LOAD) that can show familial clustering without a clear Mendelian mode of inheritance. Several chromosomal regions have been consistently identified that may harbor LOAD loci in genome-wide linkage scans on several cohorts. In our original genome scan of the NIMH-ADGI sibling cohort of 1439 individuals from 437 families we identified six candidate regions of interest (CRI) with multipoint LOD scores (MLS) >2.0 {Blacker, 2003 #1067}. Next to the 19q13 peak, which probably represents APOE, the 9q22 signal was the next most suggestive, with a peak MLS = 2.9 at 101 cM and located between 78 and 126 cM. We have followed-up the 9q22 linkage signal by genotyping additional microsatellites within this region and have confirmed the linkage signal with an increase of the MLS from 2.9 at 101 cM to 3.8 at 95 cM and the 1 LOD interval narrowed from 21.5 cM to 11 cM (~92-103 cM) {Perry, 2007 #2444}. Evidence for linkage in the 9q21-31 region has also been confirmed in data sets containing NIMH and other LOAD families. Furthermore, single nucleotide polymorphisms (SNPs) located in three genes adjacent to this 1 LOD region of 11 cM, Ubiquilin 1 (UBQLN1, 84.6 cM, Build 36) Death-Associated Protein Kinase 1 (DAPK1, 90.9 cM), and Golgi Phosphoprotein 2 (GOLPH2, 87.1 cM) have been reported to be significantly associated with AD in the NIMH sample and other data sets. They and we provide evidence that additional variants in this or other AD susceptibility genes remain to be identified in this region. We recently reported a significant association (P=0.009) of a three SNP haplotype in the middle of the neurotrophic tyrosine receptor kinase 2 (NTRK2) gene, located at 84.6 cM, to the NIMH cohort {Chen, 2008 #2458}. Thus, there is consistent and compelling evidence suggesting the 9q22 CRI harbors one or more AD susceptibility genes and a dense SNP mapping array is the next logical step to identify additional susceptibility genes in this 1 LOD region of 11 cM. In order to obtain at least 80% power to detect an OR of 1.5 or more, we will need to obtain another larger dataset of AD samples, such as available at the National Cell Repository for Alzheimer's Disease (NCRAD) consisting of ~850 families with affected siblings (~2,070 samples). Although it would be preferable to genotype all htSNPs representing all LD blocks in the entire 11 cM CRI, budgetary constraints limit us for this approach. We believe genes are of first priority and therefore propose to identify and genotype htSNPs in LD blocks covering all the genes, SNPs located between these blocks, and SNPs located in intergenic areas of the 9q22 CRI where copy number variations (CNVs) are known to be located. Any preliminary signals from this approach are then followed by sequencing to identify possible new variants in LD with these signals. This should sufficiently cover the region for identification of possible functional variants of genes contributing to LOAD risk in this region on 9q22.
Alzheimer's disease (AD) is the sixth leading cause of death in the United States and national direct and indirect annual costs of caring for individuals with AD are at least $100 billion. Identification of genes involved in AD would lead to better understanding of the cause and pathology of the disease and thus lead to improved treatment and eventual prevention of this devastating disease in the future.
