- Email: [email protected]
P1-311
Poster Presentations P1 gene binds and regulates the transcription of the only known risk factor for late onset Alzheimer’s disease, APOE. In Alzheimer’s disease, the temporal cortex is particularly susceptible to developing pathology; thus, it is tentatively possible that Zic1 might underlie this effect. P1-308
ASSOCIATION OF VARIANTS IN A PROCESSING GENES WITH LATE ONSET ALZHEIMER’S DISEASE (LOAD)
Linda H. Younkin1, Toros Dincman1, Johann Lok1, Manu Raam1, Fanggeng Zou1, Minerva Carrasquillo1, Li Ma1, Ron Budjak1, Samuel Younkin1, Ronald Petersen2, Neill Graff-Radford1, Steven Younkin1, 1Mayo Clinic School of Medicine, Jacksonville, FL, USA; 2Mayo Clinic School of Medicine, Rochester, MN, USA. Contact e-mail: [email protected] Background: The mutations in APP, PSEN1, and PSEN2 that cause early onset familial AD alter A processing in a way that fosters A aggregation. Risk for late-onset AD (LOAD) is, therefore, likely to be associated with common variants in these and other genes involved in A processing. Objective(s): To evaluate this possibility, we analyzed putative functional variants in three A processing genes (APP, PSEN1, BACE1). Methods: Putative functional variants in conserved regions (100 bp window, ⬎70% match human vs. mouse) of APP, PS1, and BACE1 were mined from publicly available databases, genotyped, and tested for association in large case/control series. Results: Previous meta-analysis of PSEN1 SNP rs16592 (Alzgene at Alzforum.org) in many Caucasian case/control series (3248 AD, 2767 Control) showed significant association with LOAD (p⫽0.03). This SNP, which is located in a conserved region of intron 8, had the same OR when it was genotyped in our combined Caucasian case/ control series (1231 AD, 1326 CON). This association was not significant in our series alone (p⫽0.17), but inclusion of our series in the metaanalysis improved significance (p⫽0.008) supporting the validity of the previously reported association. None of six additional SNPs in conserved regions of PSEN1 was significant at the 0.05 level, but two showed more significant association than rs16592. Analysis of 7 SNPs in conserved regions of BACE1 showed that four associated at p⬍0.25. In many genes all common variants are on a single haplotype block where they form a small number of haplotypes that pair to form relatively few genotypes that can be analyzed individually. We analyzed 39 SNPs in conserved regions of APP, and analysis by HaploView (solid spline of LD) showed them to be distributed over 7 haplotype blocks creating a large number of complex APP genotypes. Our preliminary analysis suggests that these complex APP genotypes may substantially influence risk for AD. Conclusions: Collectively our findings in these three LOAD candidate genes suggest that many LOAD genes may have modest effects that will be difficult to demonstrate convincingly until large case/control series with sufficient power to show replicable association are assembled as we are now endeavoring to do. P1-309
EVIDENCE FOR INTERACTION BETWEEN THE CHOLESTERYL ESTER TRANSFER PROTEIN (CETP) GENE AND APOE IN LATE ONSET ALZHEIMER’S DISEASE
Alejandro Arias1, Aaron Isaacs1, Yurii Aulchenko1, Albert Hofman1, Monique Breteler1, Ben Oostra2, Cornelia van Duijn1, 1Department of Epidemiology & Biostatistics, Erasmus Medical Centre, Rotterdam, The Netherlands; 2Department of Clinical Genetics, Erasmus Medical Centre, Rotterdam, The Netherlands. Contact e-mail: [email protected] Background: Late onset Alzheimer’s disease (LOAD) is a complex neurodegenerative disorder. Of the many genes studied thus far with respect to the disease, the only consistently replicated genetic risk factor has been the APOE gene. The APOE*4 allele of this gene may promote amyloid beta (A) plaque fibrillation, which is one of the hallmarks of LOAD. APOE may relate to LOAD through cholesterol by regulating the production of A, with high cellular cholesterol reducing A levels. Cholesteryl ester
S187
transfer protein (CETP) is involved in the same pathway through regulation of the HDL levels. Objective(s): We studied both the APOE gene and the I405V polymorphism of the CETP gene in relation to LOAD and investigated whether this polymorphism is independently associated with LOAD risk, or acts in concert with the APOE gene. Methods: We genotyped 525 LOAD cases and 5404 controls from the Rotterdam Study using a TaqMan allelic discrimination assay. Our results show that in the APOE*4 noncarriers group, those homozygous for the V allele of the I405V polymorphism have a 1.6 fold (95% CI 1.1-2.5, p ⫽ 0.02) increase in risk of LOAD compared to non-carriers. This risk increases to 2.5 (95% CI 1.4-4.5, p ⫽ 0.001) in the APOE*4 carriers group (p ⫽ 0.04 for gene-gene interaction). Conclusions: Our results suggest that the interaction of CETP with APOE increases the risk of LOAD, probably through cholesterol metabolism in the brain. P1-310
LATE ONSET ALZHEIMER’S DISEASE (LOAD) SUSCEPTIBILITY ALLELES IN THE INSULINDEGRADING-ENZYME GENE (IDE) AND OTHER A〉-DEGRADING ENZYMES
Minerva M. Carrasquillo1, Samuel G. Younkin1, Mariah R. Kashino1, Samantha L. Wilcox1, Toros A. Dincman1, Melissa E. Howard1, Ma Li1, Fanggeng Zou1, Nilu¨fer Ertekin-Taner2, Mariet Allen1, Chloe¨ M. Stanton1, Christopher M. Watson1, Linda H. Younkin1, Ronald C. Petersen2, Neill Graff-Radford1, Steven G. Younkin1, 1Mayo Clinic, Jacksonville, FL, USA; 2Mayo Clinic, Rochester, MN, USA. Contact e-mail: [email protected] Background: Genes that degrade A (IDE, PLAU, MME, ECE1) are strong LOAD candidate genes. Objective: We analyzed variants in conserved regions of these 4 genes, most notably IDE, for association with LOAD. Methods: Conserved sequences encompassing all IDE exons, 3’ and 5’ flanks (10 kb each), and several highly conserved IDE intronic regions were screened for variants likely to be functional. Variants in IDE conserved regions (⬎70% human vs. mouse) with a minor allele frequency ⬎1% were tested for association with AD in exploratory (445 AD:422 control) and follow-up series (857 AD:957 control) using age, ApoE, and gender as covariates. PLAU, MME, and ECE1 were analyzed similarly. Results: Only 15 of the 63 variants found in conserved IDE DNA had minor allele frequencies ⬎1%. Remarkably, 7 showed association at p ⬍ 0.25, and formed 5 haplotypes with frequencies over 1% that comprised 97% of all haplotypes. Of the five haplotypes, three were protective and two were risky, and they showed the same significant association with AD in the exploratory (nominal global p ⫽ 0.007) and follow-up series (global p ⫽ 0.004). The three protective haplotypes pair to form 8 genotypes likely to be protective. These were used as a referent group and compared to the five remaining common genotypes. Four of the five genotypes were risky compared to the set of 8 protective pairs with a global significance of 0.007 in the exploratory and 0.004 in the follow-up series. Combined, the set of risky IDE genotypes had an OR of 2.0 (1.5-2.6) with a PAR of 46% (30% to 58%). Importantly, we observed a significant (p⫽0.005) 1.5-fold elevation of IDE message in the cerebella of 140 subjects with protective as compared to 22 subjects with risky genotypes. Conclusions: These results indicate that IDE is likely to be an important AD gene with variants that influence risk of AD by altering IDE expression. Exploratory analysis of variants in conserved regions of PLAU, MME, ECE1 revealed that PLAU haplotypes also show significant association with LOAD (global p ⫽ 0.02), but ECE1 and MME variants showed no compelling evidence of association. P1-311
APOE LOCUS AND SURVIVAL IN A FRENCH COHORT OF DEMENTED CASES
Florence Richard1, Ame´lie Bruandet1, Violette Sorel2, Audrey Dusart3, Marie-Christine Chartier-Harlin4, Philippe Amouyel1, Florence Lebert2,
S188
Poster Presentations P1
Florence Pasquier2, Jean-Charles Lambert1, 1INSERM U744, Institut Pasteur de Lille, Lille, France; 2Memory Center, Hopital Salengro, Lille, France; 3Memory Center, Lille, France; 4EA 2683 MENRT, Hopital Roger Salengro, Lille, France. Contact e-mail: [email protected] The common apolipoprotein E (APOE) e4 allele is strongly associated with risk of dementia and age at onset, but studies are inconclusive as to whether the e4 allele affects rate of progression or survival in demented patients. Furthermore, previous observations suggest a contribution of two APOE promoter polymorphisms (-491 A/T and -219 G/T) in dementia, but the influence of these two polymorphisms on survival in demented patients have not been evaluated yet. With this background, we investigated whether the APOE locus may affect survival in a French cohort of demented patients. The APOE -491 G/T, -219 G/T and e2/e3/e4 genotyping was performed on 139 consecutive patients with probable AD or mixed dementia enrolled at the memory centre of Lille between January 1995 and march 1996. Data were rightcensured 2000 days after the patient enrollment. During this period, 7 patients (5.0%) were lost to follow-up and 48 patients (34.8%) died. A Cox proportional hazards regression model adjusting for age and MMSE score at baseline, education and gender was performed. As previously described, we observed that survival rates in dementia were better in women than in men (HR⫽0.5, 95% CI [0.3-0.9], p⫽0.02) and were lower in the oldest compared with the youngest (HR⫽1.1, 95% CI [1.0-1.2], p⫽0.0006). The -491 A/T polymorphism was not significantly associated with mortality risk whereas patients bearing at least one -219 T allele had such an increased mortality risk (HR⫽2.9, 95% CI [1.1-7.5], p⫽0.03). Interestingly, no association was found with the ⑀2 or ⑀4 alleles (respectively, HR⫽0.5, 95% CI [0.2-2.0], ns and HR⫽0.7, 95% CI [0.31.3], ns). In conclusion, we observed that the -219 T allele, which has been previously reported to increase the risk of developing AD independently of the ⑀4 allele, may also be associated with an increased mortality risk in demented patients. P1-312
SUCCESSFUL REPLICATION OF ASSOCIATION BETWEEN LATE-ONSET ALZHEIMER’S DISEASE AND THE INSULIN DEGRADING ENZYME (IDE)
Behnosh Fakhri Bjo¨rk1, Hagit Katzov2, Patrick Kehoe3, Laura Fratiglioni4, Bengt Winblad1, Jonathan A. Prince2, Caroline Graff1, 1Neurotec, Stockholm, Huddinge, Sweden; 2Centre for Genomics and Bioinformatics (CGB), Stockholm, Solna, Sweden; 3 Department of Clinical Science at North Bristol, Bristol, BS16 1LE, UK, United Kingdom; 4Neurotec, Division of Geriatric Epidemiology, Stockholm, Sweden. Contact e-mail: [email protected] Background: To date, the Apolipoprotein ⑀4 allele (APOE ⑀4) is the only identified risk factor for the late-onset sporadic form of Alzheimer’s disease (LOAD). Recent linkage studies have highlighted a large area on chromosome 10q, within which the gene for Insulin degrading enzyme (IDE) is located. IDE is a 110-kDa thiolmetalloendopeptidase, known to cleave small proteins such as hormones and bioactive peptides, including those with the ability to form -pleated sheet-rich amyloid fibrils and may thus have an important role in the degradation process and clearance of -Amyloid peptide in brain of Alzheimer’s disease (AD) patients. Previous association studies have produced inconsistent results, which is why we decided to investigate the relationship between IDE and AD development in a large case-control study including 1269 LOAD cases and 980 controls from Sweden and the UK. Objective(s): To investigate the relationship between IDE and AD in a large case-control population in an attempt to test whether the results from previous association studies were replicable, and thereby lead to a greater understanding of the importance of genetic variations in the IDE gene in AD patients. Methods: The genotyping process was performed by using Dynamic Allele Specific Hybridization (DASH). Four previously reported IDE polymorphisms IDE_7, IDE_9,
IDE_14 and HHEX_23 (rs2251101, rs1887922, rs1832196 and rs1544210 respectively) were genotyped. 1269 LOAD cases and 980 controls from Sweden and the UK were included in the study. Results: In this study we show highly significant association to IDE_9, with the strongest association among the heterozygote subjects (P⬍0.0001) in the total material. After stratification based on gender the association in males was P⬍0.0001 and after stratification based on the absence and presence of APOE ⑀4 the association in APOE ⑀4 non-carriers was P⬍0.0006. Conclusions: Our findings strongly indicate a role for IDE in AD, and provide models that may improve chances of further independent replication. P1-313
EXAMINATION OF THE EFFECT OF HETEROGENEITY ON THE CHROMOSOME 10 RISK IN LATE-ONSET ALZHEIMER’S DISEASE
Nathalie Schnetz-Boutaud1, Xueying Liang1, Eden R. Martin2, Brent M. Anderson1, Stephan Zuchner2, John R. Gilbert2, Margaret A. Pericak-Vance2, Jonathan L. Haines1, 1Center for Human Genetics Research, Vanderbilt University, Nashville, TN, USA; 2Center for Human Genetics and Department of Medicine, Duke University, Durham, NC, USA. Contact e-mail: [email protected] Background: Numerous candidate gene studies have been done to identify risk factors for Late-onset Alzheimer’s disease (LOAD), but there isn’t a gene that has been replicated in multiple studies except ApoE. One explanation for this failure is that the candidate gene selection strategy is biased by the known biological function and the function of most genes is unknown. Objective(s): This prompted us to combine linkage and association studies to locate the susceptibility genes underlying LOAD. Methods: Five previously proposed candidate genes, VR22, LRRTM3, PLAU, TNFRSF6 and IDE, on chromosome 10 were selected for study since they also fell within a broad area of linkage. 53 snps were genotyped across these genes in Caucasian American case-control dataset (745 cases, 998 controls) and an independent family-based dataset (1337 affected discordant sib pairs). Simultaneously we attempted to narrow down the peak region of linkage by analyzing family-based dataset using both covariate and subset analyses to increase the homogeneity. 36 SNPs evenly spaced across 80 Mb region on chromosome 10 were genotyped in 567 multiplex families (922 affected sib pairs). Results: 23 out of 53 snps in candidate genes showed positive results in at least one dataset, with at least one snp shown association on each gene. Marker rs2441718 on VR22 and marker rs203162 on TNFRSF6 showed association in both family-based and case-control datasets, with an Odds Ratio (OR) in the case-control dataset of 1.29 (95% CI⫽1.02-1.50, p⫽0.03) and 1.23 (95% CI⫽1.04-1.46, p⫽0.02), respectively. We analyzed the family-based data using Ordered Subset Analysis (OSA), and a two point LOD score of 2.69 was obtained at marker rs1890739 at 45.1 Mb (p⫽0.03 in 21% families) when the families were ordered from low to high by ApoE LOD score. In the autopsy confirmed subset, the peak of LOD⫽1.91 was generated at 49.7 Mb between rs7097397 and rs14327 in OSA multipoint analysis when the covariate ApoE LOD score ranked families from low to high. Conclusions: Our data suggest that there is a locus associated with LOAD in the region between 40 to 60 Mb, near rs7097397 (49.7 Mb) on chromosome 10. These results will facilitate further association studies in this important LOAD region. P1-314
A NOVEL PSEN2 MUTATION IN EARLY-ONSET DEMENTIA WITH PROFOUND SEMANTIC LOSS
Suzanne G. Lindquist1, Jørgen E. Nielsen1, Marianne Schwartz2, Lis Hasholt3, Anne Nørremølle3, Gunhild Waldemar1, 1Memory Disorders Research Group, Department of Neurology, Copenhagen University Hospital, Rigshospitalet, Copenhagen Ø, Denmark; 2 Department of Clinical Genetics, Copenhagen University Hospital, Rigshospitalet, Copenhagen Ø, Denmark; 3Department of Medical
ASSOCIATION OF VARIANTS IN A PROCESSING GENES WITH LATE ONSET ALZHEIMER’S DISEASE (LOAD)
Linda H. Younkin1, Toros Dincman1, Johann Lok1, Manu Raam1, Fanggeng Zou1, Minerva Carrasquillo1, Li Ma1, Ron Budjak1, Samuel Younkin1, Ronald Petersen2, Neill Graff-Radford1, Steven Younkin1, 1Mayo Clinic School of Medicine, Jacksonville, FL, USA; 2Mayo Clinic School of Medicine, Rochester, MN, USA. Contact e-mail: [email protected] Background: The mutations in APP, PSEN1, and PSEN2 that cause early onset familial AD alter A processing in a way that fosters A aggregation. Risk for late-onset AD (LOAD) is, therefore, likely to be associated with common variants in these and other genes involved in A processing. Objective(s): To evaluate this possibility, we analyzed putative functional variants in three A processing genes (APP, PSEN1, BACE1). Methods: Putative functional variants in conserved regions (100 bp window, ⬎70% match human vs. mouse) of APP, PS1, and BACE1 were mined from publicly available databases, genotyped, and tested for association in large case/control series. Results: Previous meta-analysis of PSEN1 SNP rs16592 (Alzgene at Alzforum.org) in many Caucasian case/control series (3248 AD, 2767 Control) showed significant association with LOAD (p⫽0.03). This SNP, which is located in a conserved region of intron 8, had the same OR when it was genotyped in our combined Caucasian case/ control series (1231 AD, 1326 CON). This association was not significant in our series alone (p⫽0.17), but inclusion of our series in the metaanalysis improved significance (p⫽0.008) supporting the validity of the previously reported association. None of six additional SNPs in conserved regions of PSEN1 was significant at the 0.05 level, but two showed more significant association than rs16592. Analysis of 7 SNPs in conserved regions of BACE1 showed that four associated at p⬍0.25. In many genes all common variants are on a single haplotype block where they form a small number of haplotypes that pair to form relatively few genotypes that can be analyzed individually. We analyzed 39 SNPs in conserved regions of APP, and analysis by HaploView (solid spline of LD) showed them to be distributed over 7 haplotype blocks creating a large number of complex APP genotypes. Our preliminary analysis suggests that these complex APP genotypes may substantially influence risk for AD. Conclusions: Collectively our findings in these three LOAD candidate genes suggest that many LOAD genes may have modest effects that will be difficult to demonstrate convincingly until large case/control series with sufficient power to show replicable association are assembled as we are now endeavoring to do. P1-309
EVIDENCE FOR INTERACTION BETWEEN THE CHOLESTERYL ESTER TRANSFER PROTEIN (CETP) GENE AND APOE IN LATE ONSET ALZHEIMER’S DISEASE
Alejandro Arias1, Aaron Isaacs1, Yurii Aulchenko1, Albert Hofman1, Monique Breteler1, Ben Oostra2, Cornelia van Duijn1, 1Department of Epidemiology & Biostatistics, Erasmus Medical Centre, Rotterdam, The Netherlands; 2Department of Clinical Genetics, Erasmus Medical Centre, Rotterdam, The Netherlands. Contact e-mail: [email protected] Background: Late onset Alzheimer’s disease (LOAD) is a complex neurodegenerative disorder. Of the many genes studied thus far with respect to the disease, the only consistently replicated genetic risk factor has been the APOE gene. The APOE*4 allele of this gene may promote amyloid beta (A) plaque fibrillation, which is one of the hallmarks of LOAD. APOE may relate to LOAD through cholesterol by regulating the production of A, with high cellular cholesterol reducing A levels. Cholesteryl ester
S187
transfer protein (CETP) is involved in the same pathway through regulation of the HDL levels. Objective(s): We studied both the APOE gene and the I405V polymorphism of the CETP gene in relation to LOAD and investigated whether this polymorphism is independently associated with LOAD risk, or acts in concert with the APOE gene. Methods: We genotyped 525 LOAD cases and 5404 controls from the Rotterdam Study using a TaqMan allelic discrimination assay. Our results show that in the APOE*4 noncarriers group, those homozygous for the V allele of the I405V polymorphism have a 1.6 fold (95% CI 1.1-2.5, p ⫽ 0.02) increase in risk of LOAD compared to non-carriers. This risk increases to 2.5 (95% CI 1.4-4.5, p ⫽ 0.001) in the APOE*4 carriers group (p ⫽ 0.04 for gene-gene interaction). Conclusions: Our results suggest that the interaction of CETP with APOE increases the risk of LOAD, probably through cholesterol metabolism in the brain. P1-310
LATE ONSET ALZHEIMER’S DISEASE (LOAD) SUSCEPTIBILITY ALLELES IN THE INSULINDEGRADING-ENZYME GENE (IDE) AND OTHER A〉-DEGRADING ENZYMES
Minerva M. Carrasquillo1, Samuel G. Younkin1, Mariah R. Kashino1, Samantha L. Wilcox1, Toros A. Dincman1, Melissa E. Howard1, Ma Li1, Fanggeng Zou1, Nilu¨fer Ertekin-Taner2, Mariet Allen1, Chloe¨ M. Stanton1, Christopher M. Watson1, Linda H. Younkin1, Ronald C. Petersen2, Neill Graff-Radford1, Steven G. Younkin1, 1Mayo Clinic, Jacksonville, FL, USA; 2Mayo Clinic, Rochester, MN, USA. Contact e-mail: [email protected] Background: Genes that degrade A (IDE, PLAU, MME, ECE1) are strong LOAD candidate genes. Objective: We analyzed variants in conserved regions of these 4 genes, most notably IDE, for association with LOAD. Methods: Conserved sequences encompassing all IDE exons, 3’ and 5’ flanks (10 kb each), and several highly conserved IDE intronic regions were screened for variants likely to be functional. Variants in IDE conserved regions (⬎70% human vs. mouse) with a minor allele frequency ⬎1% were tested for association with AD in exploratory (445 AD:422 control) and follow-up series (857 AD:957 control) using age, ApoE, and gender as covariates. PLAU, MME, and ECE1 were analyzed similarly. Results: Only 15 of the 63 variants found in conserved IDE DNA had minor allele frequencies ⬎1%. Remarkably, 7 showed association at p ⬍ 0.25, and formed 5 haplotypes with frequencies over 1% that comprised 97% of all haplotypes. Of the five haplotypes, three were protective and two were risky, and they showed the same significant association with AD in the exploratory (nominal global p ⫽ 0.007) and follow-up series (global p ⫽ 0.004). The three protective haplotypes pair to form 8 genotypes likely to be protective. These were used as a referent group and compared to the five remaining common genotypes. Four of the five genotypes were risky compared to the set of 8 protective pairs with a global significance of 0.007 in the exploratory and 0.004 in the follow-up series. Combined, the set of risky IDE genotypes had an OR of 2.0 (1.5-2.6) with a PAR of 46% (30% to 58%). Importantly, we observed a significant (p⫽0.005) 1.5-fold elevation of IDE message in the cerebella of 140 subjects with protective as compared to 22 subjects with risky genotypes. Conclusions: These results indicate that IDE is likely to be an important AD gene with variants that influence risk of AD by altering IDE expression. Exploratory analysis of variants in conserved regions of PLAU, MME, ECE1 revealed that PLAU haplotypes also show significant association with LOAD (global p ⫽ 0.02), but ECE1 and MME variants showed no compelling evidence of association. P1-311
APOE LOCUS AND SURVIVAL IN A FRENCH COHORT OF DEMENTED CASES
Florence Richard1, Ame´lie Bruandet1, Violette Sorel2, Audrey Dusart3, Marie-Christine Chartier-Harlin4, Philippe Amouyel1, Florence Lebert2,
S188
Poster Presentations P1
Florence Pasquier2, Jean-Charles Lambert1, 1INSERM U744, Institut Pasteur de Lille, Lille, France; 2Memory Center, Hopital Salengro, Lille, France; 3Memory Center, Lille, France; 4EA 2683 MENRT, Hopital Roger Salengro, Lille, France. Contact e-mail: [email protected] The common apolipoprotein E (APOE) e4 allele is strongly associated with risk of dementia and age at onset, but studies are inconclusive as to whether the e4 allele affects rate of progression or survival in demented patients. Furthermore, previous observations suggest a contribution of two APOE promoter polymorphisms (-491 A/T and -219 G/T) in dementia, but the influence of these two polymorphisms on survival in demented patients have not been evaluated yet. With this background, we investigated whether the APOE locus may affect survival in a French cohort of demented patients. The APOE -491 G/T, -219 G/T and e2/e3/e4 genotyping was performed on 139 consecutive patients with probable AD or mixed dementia enrolled at the memory centre of Lille between January 1995 and march 1996. Data were rightcensured 2000 days after the patient enrollment. During this period, 7 patients (5.0%) were lost to follow-up and 48 patients (34.8%) died. A Cox proportional hazards regression model adjusting for age and MMSE score at baseline, education and gender was performed. As previously described, we observed that survival rates in dementia were better in women than in men (HR⫽0.5, 95% CI [0.3-0.9], p⫽0.02) and were lower in the oldest compared with the youngest (HR⫽1.1, 95% CI [1.0-1.2], p⫽0.0006). The -491 A/T polymorphism was not significantly associated with mortality risk whereas patients bearing at least one -219 T allele had such an increased mortality risk (HR⫽2.9, 95% CI [1.1-7.5], p⫽0.03). Interestingly, no association was found with the ⑀2 or ⑀4 alleles (respectively, HR⫽0.5, 95% CI [0.2-2.0], ns and HR⫽0.7, 95% CI [0.31.3], ns). In conclusion, we observed that the -219 T allele, which has been previously reported to increase the risk of developing AD independently of the ⑀4 allele, may also be associated with an increased mortality risk in demented patients. P1-312
SUCCESSFUL REPLICATION OF ASSOCIATION BETWEEN LATE-ONSET ALZHEIMER’S DISEASE AND THE INSULIN DEGRADING ENZYME (IDE)
Behnosh Fakhri Bjo¨rk1, Hagit Katzov2, Patrick Kehoe3, Laura Fratiglioni4, Bengt Winblad1, Jonathan A. Prince2, Caroline Graff1, 1Neurotec, Stockholm, Huddinge, Sweden; 2Centre for Genomics and Bioinformatics (CGB), Stockholm, Solna, Sweden; 3 Department of Clinical Science at North Bristol, Bristol, BS16 1LE, UK, United Kingdom; 4Neurotec, Division of Geriatric Epidemiology, Stockholm, Sweden. Contact e-mail: [email protected] Background: To date, the Apolipoprotein ⑀4 allele (APOE ⑀4) is the only identified risk factor for the late-onset sporadic form of Alzheimer’s disease (LOAD). Recent linkage studies have highlighted a large area on chromosome 10q, within which the gene for Insulin degrading enzyme (IDE) is located. IDE is a 110-kDa thiolmetalloendopeptidase, known to cleave small proteins such as hormones and bioactive peptides, including those with the ability to form -pleated sheet-rich amyloid fibrils and may thus have an important role in the degradation process and clearance of -Amyloid peptide in brain of Alzheimer’s disease (AD) patients. Previous association studies have produced inconsistent results, which is why we decided to investigate the relationship between IDE and AD development in a large case-control study including 1269 LOAD cases and 980 controls from Sweden and the UK. Objective(s): To investigate the relationship between IDE and AD in a large case-control population in an attempt to test whether the results from previous association studies were replicable, and thereby lead to a greater understanding of the importance of genetic variations in the IDE gene in AD patients. Methods: The genotyping process was performed by using Dynamic Allele Specific Hybridization (DASH). Four previously reported IDE polymorphisms IDE_7, IDE_9,
IDE_14 and HHEX_23 (rs2251101, rs1887922, rs1832196 and rs1544210 respectively) were genotyped. 1269 LOAD cases and 980 controls from Sweden and the UK were included in the study. Results: In this study we show highly significant association to IDE_9, with the strongest association among the heterozygote subjects (P⬍0.0001) in the total material. After stratification based on gender the association in males was P⬍0.0001 and after stratification based on the absence and presence of APOE ⑀4 the association in APOE ⑀4 non-carriers was P⬍0.0006. Conclusions: Our findings strongly indicate a role for IDE in AD, and provide models that may improve chances of further independent replication. P1-313
EXAMINATION OF THE EFFECT OF HETEROGENEITY ON THE CHROMOSOME 10 RISK IN LATE-ONSET ALZHEIMER’S DISEASE
Nathalie Schnetz-Boutaud1, Xueying Liang1, Eden R. Martin2, Brent M. Anderson1, Stephan Zuchner2, John R. Gilbert2, Margaret A. Pericak-Vance2, Jonathan L. Haines1, 1Center for Human Genetics Research, Vanderbilt University, Nashville, TN, USA; 2Center for Human Genetics and Department of Medicine, Duke University, Durham, NC, USA. Contact e-mail: [email protected] Background: Numerous candidate gene studies have been done to identify risk factors for Late-onset Alzheimer’s disease (LOAD), but there isn’t a gene that has been replicated in multiple studies except ApoE. One explanation for this failure is that the candidate gene selection strategy is biased by the known biological function and the function of most genes is unknown. Objective(s): This prompted us to combine linkage and association studies to locate the susceptibility genes underlying LOAD. Methods: Five previously proposed candidate genes, VR22, LRRTM3, PLAU, TNFRSF6 and IDE, on chromosome 10 were selected for study since they also fell within a broad area of linkage. 53 snps were genotyped across these genes in Caucasian American case-control dataset (745 cases, 998 controls) and an independent family-based dataset (1337 affected discordant sib pairs). Simultaneously we attempted to narrow down the peak region of linkage by analyzing family-based dataset using both covariate and subset analyses to increase the homogeneity. 36 SNPs evenly spaced across 80 Mb region on chromosome 10 were genotyped in 567 multiplex families (922 affected sib pairs). Results: 23 out of 53 snps in candidate genes showed positive results in at least one dataset, with at least one snp shown association on each gene. Marker rs2441718 on VR22 and marker rs203162 on TNFRSF6 showed association in both family-based and case-control datasets, with an Odds Ratio (OR) in the case-control dataset of 1.29 (95% CI⫽1.02-1.50, p⫽0.03) and 1.23 (95% CI⫽1.04-1.46, p⫽0.02), respectively. We analyzed the family-based data using Ordered Subset Analysis (OSA), and a two point LOD score of 2.69 was obtained at marker rs1890739 at 45.1 Mb (p⫽0.03 in 21% families) when the families were ordered from low to high by ApoE LOD score. In the autopsy confirmed subset, the peak of LOD⫽1.91 was generated at 49.7 Mb between rs7097397 and rs14327 in OSA multipoint analysis when the covariate ApoE LOD score ranked families from low to high. Conclusions: Our data suggest that there is a locus associated with LOAD in the region between 40 to 60 Mb, near rs7097397 (49.7 Mb) on chromosome 10. These results will facilitate further association studies in this important LOAD region. P1-314
A NOVEL PSEN2 MUTATION IN EARLY-ONSET DEMENTIA WITH PROFOUND SEMANTIC LOSS
Suzanne G. Lindquist1, Jørgen E. Nielsen1, Marianne Schwartz2, Lis Hasholt3, Anne Nørremølle3, Gunhild Waldemar1, 1Memory Disorders Research Group, Department of Neurology, Copenhagen University Hospital, Rigshospitalet, Copenhagen Ø, Denmark; 2 Department of Clinical Genetics, Copenhagen University Hospital, Rigshospitalet, Copenhagen Ø, Denmark; 3Department of Medical