Background: Alzheimer’s disease (AD) is the leading cause of dementia in the elderly. While evidence for an underlying genetic etiology of AD is strong, it has proven to be quite challenging to elucidate. The APOE gene accounts for less than half of the susceptibility and thus other genetic factors are likely to be involved. Genetic heterogeneity is a major compli- cating factor hindering further gene identification. To overcome this prob- lem and maximize our power to identify AD risk genes, we are studying the genetically isolated and well-defined Amish populations of middle Ohio and northern Indiana. To date we have enrolled over 1550 Amish individ- uals with over 100 of these having either probable or possible dementia. Methods: We have undertaken a whole-genome SNP linkage screen (Il- lumina Linkage Panel IVb) using 672 Amish individuals (103 with AD) within 21 sub-pedigrees ranging in total size from 18 to 157 individuals. We performed both 2-pt linkage analysis (using dominant and recessive models within the Superlink program) and a novel test of association using the MQLS test (by Thornton & McPeek), on 5,645 SNPs. Results: Pre- liminary linkage analysis found 162 SNPs with 2-pt lod scores 1.0. Suggestive linkage to AD was found for 18 SNPs across 16 independent loci (2-pt lod scores 2.0: 1q, 2p, 2q, 3q22, 3q26, 4q, 5q, 7q, 8p, 13q, 14q23, 14q24, 16p, 18q, 20q, 21q). The MQLS test also identified 284 SNPs with nominally significant p-values (p0.05). Subsequently, we utilized the Simwalk program to perform multipoint linkage analysis for multiple 3-SNP sliding windows across each of the regions identified by the 2-pt analysis. Five of these regions continued to demonstrate strong evidence for linkage (1q, 3q22, 5q, 18q, and 21q), with three of these also demonstrating MQLS p-values 0.05 (1q, 3q22, 18q). Additional fol- low-up of these more narrowed regions is currently ongoing. Conclusions: Our genome scan suggests that a number of regions within our Amish population may contain AD susceptibility genes, with the strongest current evidence suggesting involvement of chr 1q, 3q, and 18q. P3-226 DISSECTING GENETIC VARIATION AT THE NICASTRIN LOCUS USING FUNCTIONAL GENOMICS Gillian Hamilton 1 , John Powell 2 , Richard Wade-Martins 1 , 1 University of Oxford, Oxford, United Kingdom; 2 Institute of Psychiatry, London, United Kingdom. Contact e-mail: gillian.hamilton@dpag.ox.ac.uk Background: Given the many problems encountered using traditional genetic methods to identify risk genes for Alzheimer’s disease (AD) a new approach to this challenge is timely. Using a novel viral-based approach our laboratory is investigating the functional effect of variation at complete genomic loci. Nicastrin (Nct) is a component of the -secretase complex and represents an excellent functional candidate gene for late-onset AD. Five major haplotypes (A-E) are located at this locus and association results indicate that one, HapB, is a potential risk factor for AD. However, like the majority of studies, results have been inconsistent. We are inves- tigating the functional role of haplotype variation at the Nct locus using this novel viral-based approach. Methods: The infectious BAC (iBAC) system is a newly developed method permitting in vitro functional analysis of whole genomic loci. This system has a high genomic capacity (up to 150kb), a broad range of tropism and is helper-free. A series of BACs expressing the five common Nct haplotypes were created using established recombination methods. These were further modified by the insertion of sequences from Herpes Simplex Virus-1 (HSV-1). Nct knock-out mouse embryonic fibroblasts were infected with each Nct-haplotype and clonal cell lines generated. A Notch functional assay has been established to determine the level of functional -secretase activity. Results: All avail- able Nct-BACs were found to carry the A haplotype, and the initial step to create a series of BACs expressing the remaining four haplotypes (B-E) was achieved by modification of the DNA sequence using homologous recombination. Each BAC was successfully retrofitted with pEHHG, a vector carrying both HSV-1 elements required for amplicon production and Epstein Barr Virus elements needed for episomal retention. Clonal cells lines successfully maintaining each BAC episomally were identified. These cells lines are currently being assessed for differential functional abilities. Conclusions: Clonal cell lines expressing the five Nct haplotypes have been successfully generated and functional analysis is ongoing. The results obtained from these experiments will inform as to whether variation at the Nct locus affects -secretase activity, and consequently whether Nct is a valid genetic risk factor for AD. P3-227 POLYMORPHISMS IN THE DNA REPAIR GENES, HOGG1 AND APE1 AND THE RISK OF ALZHEIMER’S DISEASE Hasmet A. Hanagasi 1 , Hande Parildar-Karpuzoglu 2 , Semra Dogru-Abbasoglu 2 , Berrin Karadag 3 , Hakan Gurvit 1 , Murat Emre 1 , Mujdat Uysal 2 , 1 Istanbul Medical Faculty, Department of Neurology, Istanbul, Turkey; 2 Istanbul Medical Faculty, Department of Biochemistry, Istanbul, Turkey; 3 Cerrahpasa School of Medicine, Department of Internal Medicine, Division of Geriatrics, Istanbul, Turkey. Contact e-mail: hasmet@yahoo.com Background: Oxidative stress has been claimed to play a central role in neurodegenerative diseases, including Alzheimer’s disease (AD). One abundant type of DNA damage resulting from reactive oxygen species (ROS) exposure produces 8-oxoguanine (8-oxoG), which has been shown to be highly mutagenic, yielding G:C to T:A transversions. hOGG1 is the major human 8-oxoguanine DNA glycosylase which removes 8-oxoG. Following the glycosylase reaction, the apurinic/apyrimidinic (AP) site is subsequently recognized by AP-endonuclease. The major mammalian AP- endonuclease, APE1 incises the phosphodiester backbone of DNA imme- diately 5’ to the baseless lesion, leaving behind a strand break with a normal 3’-hydroxyl group and a non-conventional 5’-abasic terminus. Methods: In our study, the polymorphisms of hOGG1 (exon 7, codon 326) and APE1 (exon 5, codon 148) were investigated in the DNA samples of 61 patients with AD and age-matched 63 control subjects. The diagnosis of probable or possible AD was made according to the NINCDS-ADRDA criteria. Results: The mean age was 76.126.32 for AD cases. No statis- tically significant differences were found between cases and controls for genotypes and alleles for these polymorphisms. Conclusions: Our findings show that the polymorphic variants of hOGG1 (exon 7, codon 326) and APE1 (exon 5, codon 148) may not be considered as risk factors in AD. However, it is necessary to investigate these polymorphisms in larger study populations. P3-228 GENETIC ASSOCIATION OF VITAMIN D RECEPTOR GENE ON CHROMOSOME 12q WITH LATE-ONSET ALZHEIMER’S DISEASE Kenju Hara 1 , John Gilbert 1 , Paul Gallins 1 , Patrice Whitehead 1 , Gary Beecham 1 , Michael Slifer 1 , Eden Martin 1 , Jonathan Haines 2 , Margaret Pericak-Vance 1 , 1 Miami Institute for Human Genomics, University of Miami Miller School of Medicine, Miami, FL, USA; 2 2) Center for Human Genetics Research, Vanderbilt University, Nashville, TN, USA. Contact e-mail: khara@med.miami.edu Background: Late-onset Alzheimer’s disease (LOAD) has a complex etiology and a strong genetic component. Several linkage studies have shown evidence of linkage to chromosome 12q for LOAD. We have previously conducted a genome-wide association study for LOAD and identified a strongly associated SNP (rs11610206, P=1.4E-6) on chromo- some 12q13.11. Vitamin D is known to be an antioxidant which up- regulates  glutamyl transpeptidase activity and controls brain detoxifica- tion processes. It also regulates the synthesis of neurotrophic factors and alter calcium uptake in some excitable cells by modulating the L-type voltage sensitive calcium channels in hippocampus. Methods: Using genomic convergence, combining the information from linkage and asso- ciation studies and expression data (SAGE; serial analysis of gene expres- sion), we focused on the vitamin D receptor (VDR) gene that had a previously published or inferred functional role in LOAD. We screened for potential risk-associated sequence variants in all exons and 3’-UTR in VDR by sequencing 28 cases and 20 controls. We further investigated the T587 Poster Presentations P3: