Late-onset Alzheimer's Disease Risk Loci Research Articles

Single-nucleus multiomics reveals the disrupted regulatory programs in three brain regions of sporadic early-onset Alzheimer's disease.

Sporadic early-onset Alzheimer's disease (sEOAD) represents a significant but less-studied subtype of Alzheimer's disease (AD). Here, we generated a single-nucleus multiome atlas derived from the postmortem prefrontal cortex, entorhinal cortex, and hippocampus of nine individuals with or without sEOAD. Comprehensive analyses were conducted to delineate cell type-specific transcriptomic changes and linked candidate cis- regulatory elements (cCREs) across brain regions. We prioritized seven conservative transcription factors in glial cells in multiple brain regions, including RFX4 in astrocytes and IKZF1 in microglia, which are implicated in regulating sEOAD-associated genes. Moreover, we identified the top 25 altered intercellular signaling between glial cells and neurons, highlighting their regulatory potential on gene expression in receiver cells. We reported 38 cCREs linked to sEOAD-associated genes overlapped with late-onset AD risk loci, and sEOAD cCREs enriched in neuropsychiatric disorder risk loci. This atlas helps dissect transcriptional and chromatin dynamics in sEOAD, providing a key resource for AD research.

Alzheimer's disease risk allele of PICALM causes detrimental lipid droplets in microglia.

Despite genome-wide association studies of late-onset Alzheimer's disease (LOAD) having identified many genetic risk loci1-6, the underlying disease mechanisms remain largely unknown. Determining causal disease variants and their LOAD-relevant cellular phenotypes has been a challenge. Leveraging our approach for identifying functional GWAS risk variants showing allele-specific open chromatin (ASoC)7, we systematically identified putative causal LOAD risk variants in human induced pluripotent stem cells (iPSC)-derived neurons, astrocytes, and microglia (MG) and linked PICALM risk allele to a previously unappreciated MG-specific role of PICALM in lipid droplet (LD) accumulation. ASoC mapping uncovered functional risk variants for 26 LOAD risk loci, mostly MG-specific. At the MG-specific PICALM locus, the LOAD risk allele of rs10792832 reduced transcription factor (PU.1) binding and PICALM expression, impairing the uptake of amyloid beta (Aβ) and myelin debris. Interestingly, MG with PICALM risk allele showed transcriptional enrichment of pathways for cholesterol synthesis and LD formation. Genetic and pharmacological perturbations of MG further established a causal link between the reduced PICALM expression, LD accumulation, and phagocytosis deficits. Our work elucidates the selective LOAD vulnerability in microglia for the PICALM locus through detrimental LD accumulation, providing a neurobiological basis that can be exploited for developing novel clinical interventions.

Cell type-specific histone acetylation profiling of Alzheimer's disease subjects and integration with genetics.

We profile genome-wide histone 3 lysine 27 acetylation (H3K27ac) of 3 major brain cell types from hippocampus and dorsolateral prefrontal cortex (dlPFC) of subjects with and without Alzheimer's Disease (AD). We confirm that single nucleotide polymorphisms (SNPs) associated with late onset AD (LOAD) show a strong tendency to reside in microglia-specific gene regulatory elements. Despite this significant colocalization, we find that microglia harbor more acetylation changes associated with age than with amyloid-β (Aβ) load. In contrast, we detect that an oligodendrocyte-enriched glial (OEG) population contains the majority of differentially acetylated peaks associated with Aβ load. These differential peaks reside near both early onset risk genes (APP, PSEN1, PSEN2) and late onset AD risk loci (including BIN1, PICALM, CLU, ADAM10, ADAMTS4, SORL1, FERMT2), Aβ processing genes (BACE1), as well as genes involved in myelinating and oligodendrocyte development processes. Interestingly, a number of LOAD risk loci associated with differentially acetylated risk genes contain H3K27ac peaks that are specifically enriched in OEG. These findings implicate oligodendrocyte gene regulation as a potential mechanism by which early onset and late onset risk genes mediate their effects, and highlight the deregulation of myelinating processes in AD. More broadly, our dataset serves as a resource for the study of functional effects of genetic variants and cell type specific gene regulation in AD.

An integrated genome and phenome-wide association study approach to understanding Alzheimer's disease predisposition

Genome-wide association studies (GWAS) have identified common single nucleotide polymorphisms (SNPs) that increase late-onset Alzheimer's disease (LOAD) risk. To identify additional LOAD-associated variants and provide insight into underlying disease biology, we performed a phenome-wide association study on 23 known LOAD-associated SNPs and 4:1 matched control SNPs using UK Biobank data. LOAD-associated SNPs were significantly enriched for associations with 8/778 queried traits, including 3 platelet traits. The strongest enrichment was for platelet distribution width (PDW) (p = 1.2 × 10−5), but increased PDW was not associated with LOAD susceptibility in Mendelian randomization analysis. Of 384 PDW-associated SNPs identified by prior GWAS, 36 were nominally associated with LOAD risk (17,008 cases; 37,154 controls) and 5 survived false-discovery rate correction. Associations confirmed known LOAD risk loci near PICALM, CD2AP, SPI1, and NDUFAF6, and identified a novel risk locus in epidermal growth factor receptor. Integrating GWAS and phenome-wide association study data reveals substantial pleiotropy between genetic determinants of LOAD and of platelet morphology, and for the first time implicates epidermal growth factor receptor – a mediator of β-amyloid toxicity – in Alzheimer's disease susceptibility.

Genetic Background for Alzheimer's Disease: Knowledge Accumulated from AD GWAS

Dementia is a leading cause of death in many countries. Late-onset Alzheimer's disease (LOAD) is the most common form of dementia, with approximately 35 million affected people. LOAD shows a high heritability (h2) of 58-79%. Clarifying the genetic architecture of LOAD could contribute to precision medicine. In recent years, large-scale genome-wide association studies (GWASs) and their meta-analyses with a large sample size have elucidated the disease-susceptible genes and disease-causing pathways. To date, meta-analyses of GWASs in the Caucasian population have successfully identified approximately 40 LOAD risk loci. The gene set and disease pathway analysis obtained from the results of GWASs suggested biological mechanisms involving brain immune function, lipid-related processes, tau binding proteins, and degradation of amyloid precursor proteins in the pathogenesis of LOAD. Furthermore, the exome sequencing analysis in Japanese individuals with LOAD also revealed a rare variant with a large effect of SHARPIN in LOAD susceptibility, and the variant protein possibly affects the immune response through aberrant cellular localization, which may result in attenuated NF-κB activity in the brain. These findings could provide biological and pharmaceutical approaches in precision medicine for LOAD.

HIGH-THROUGHPUT IDENTIFICATION OF NONCODING FUNCTIONAL SNPS

Genome wide association studies (GWAS) have provided significant insight into complex diseases by identifying numerous disease-associated non-coding single nucleotide polymorphisms (SNPs). However, GWAS cannot distinguish functional SNPs (fSNPs) from other disease-associated SNPs that are in linkage disequilibrium (LD). To address this challenge, we developed a novel approach to bridge the gap between GWAS and biological mechanisms with a set of high throughput methodologies. We use SNP-seq (Single Nucleotide Polymorphism-Sequencing) and Reel-seq (Regulatory Element-Sequencing) to identify disease-associated fSNPs and FREP-MS (Flanking Restriction Enhanced DNA Pulldown-Mass Spectrometry) to characterize these fSNPs by identifying regulatory proteins that specifically bind to the fSNPs. To prove the feasibility, we applied this approach to a couple of disease-associated loci. For this study, we applied the same approach to a library that contains SNPs in LD on 27 loci that are associated with late onset Alzheimer's disease (LOAD). We confirmed 27 fSNPs on 12 LOAD risk loci by allele-imbalanced gel shifting including 5 on the HLA-DQA/ HLA DRB1 locus. We identified multiple proteins that specifically bind to each of these fSNPs. The regulation of the risk gene expression on these 12 loci will be explored.

Association of Alzheimer's Disease Genetic Risk Loci with Cognitive Performance and Decline: A Systematic Review.

The association of Apolipoprotein E (APOE) with late-onset Alzheimer's disease (LOAD) and cognitive endophenotypes of aging has been widely investigated. There is increasing interest in evaluating the association of other LOAD risk loci with cognitive performance and decline. The results of these studies have been inconsistent and inconclusive. We conducted a systematic review of studies investigating the association of non-APOE LOAD risk loci with cognitive performance in older adults. Studies published from January 2009 to April 2018 were identified through a PubMed database search using keywords and by scanning reference lists. Studies were included if they were either cross-sectional or longitudinal in design, included at least one genome-wide significant LOAD risk loci or a genetic risk score, and had one objective measure of cognition. Quality assessment of the studies was conducted using the quality of genetic studies (Q-Genie) tool. Of 2,466 studies reviewed, 49 met inclusion criteria. Fifteen percent of the associations between non-APOE LOAD risk loci and cognition were significant. However, these associations were not replicated across studies, and the majority were rendered non-significant when adjusting for multiple testing. One-third of the studies included genetic risk scores, and these were typically significant only when APOE was included. The findings of this systematic review do not support a consistent association between individual non-APOE LOAD risk and cognitive performance or decline. However, evidence suggests that aggregate LOAD genetic risk exerts deleterious effects on decline in episodic memory and global cognition.

Genetic meta-analysis of diagnosed Alzheimer's disease identifies new risk loci and implicates Aβ, tau, immunity and lipid processing.

Risk for late-onset Alzheimer's disease (LOAD), the most prevalent dementia, is partially driven by genetics. To identify LOAD risk loci, we performed a large genome-wide association meta-analysis of clinically diagnosed LOAD (94,437 individuals). We confirm 20 previous LOAD risk loci and identify five new genome-wide loci (IQCK, ACE, ADAM10, ADAMTS1, and WWOX), two of which (ADAM10, ACE) were identified in a recent genome-wide association (GWAS)-by-familial-proxy of Alzheimer's or dementia. Fine-mapping of the human leukocyte antigen (HLA) region confirms the neurological and immune-mediated disease haplotype HLA-DR15 as a risk factor for LOAD. Pathway analysis implicates immunity, lipid metabolism, tau binding proteins, and amyloid precursor protein (APP) metabolism, showing that genetic variants affecting APP and Aβ processing are associated not only with early-onset autosomal dominant Alzheimer's disease but also with LOAD. Analyses of risk genes and pathways show enrichment for rare variants (P = 1.32 × 10-7), indicating that additional rare variants remain to be identified. We also identify important genetic correlations between LOAD and traits such as family history of dementia and education.

Late Onset Alzheimer’s Disease Risk Variants in Cognitive Decline: The PATH Through Life Study

Recent genome wide association studies have identified a number of single nucleotide polymorphisms associated with late onset Alzheimer's disease (LOAD). We examined the associations of 24 LOAD risk loci, individually and collectively as a genetic risk score, with cognitive function. We used data from 1,626 non-demented older Australians of European ancestry who were examined up to four times over 12 years on tests assessing episodic memory, working memory, vocabulary, and information processing speed. Linear mixed models were generated to examine associations between genetic factors and cognitive performance. Twelve SNPs were significantly associated with baseline cognitive performance (ABCA7, MS4A4E, SORL1), linear rate of change (APOE, ABCA7, INPP5D, ZCWPW1, CELF1), or quadratic rate of change (APOE, CLU, EPHA1, HLA-DRB5, INPP5D, FERMT2). In addition, a weighted genetic risk score was associated with linear rate of change in episodic memory and information processing speed. Our results suggest that a minority of AD related SNPs may be associated with non-clinical cognitive decline. Further research is required to verify these results and to examine the effect of preclinical AD in genetic association studies of cognitive decline. The identification of LOAD risk loci associated with non-clinical cognitive performance may help in screening for individuals at greater risk of cognitive decline.

Effects of multiple genetic loci on age at onset in late-onset Alzheimer disease: a genome-wide association study.

Because APOE locus variants contribute to risk of late-onset Alzheimer disease (LOAD) and to differences in age at onset (AAO), it is important to know whether other established LOAD risk loci also affect AAO in affected participants. To investigate the effects of known Alzheimer disease risk loci in modifying AAO and to estimate their cumulative effect on AAO variation using data from genome-wide association studies in the Alzheimer Disease Genetics Consortium. The Alzheimer Disease Genetics Consortium comprises 14 case-control, prospective, and family-based data sets with data on 9162 participants of white race/ethnicity with Alzheimer disease occurring after age 60 years who also had complete AAO information, gathered between 1989 and 2011 at multiple sites by participating studies. Data on genotyped or imputed single-nucleotide polymorphisms most significantly associated with risk at 10 confirmed LOAD loci were examined in linear modeling of AAO, and individual data set results were combined using a random-effects, inverse variance-weighted meta-analysis approach to determine whether they contribute to variation in AAO. Aggregate effects of all risk loci on AAO were examined in a burden analysis using genotype scores weighted by risk effect sizes. Age at disease onset abstracted from medical records among participants with LOAD diagnosed per standard criteria. Analysis confirmed the association of APOE with earlier AAO (P = 3.3 × 10(-96)), with associations in CR1 (rs6701713, P = 7.2 × 10(-4)), BIN1 (rs7561528, P = 4.8 × 10(-4)), and PICALM (rs561655, P = 2.2 × 10(-3)) reaching statistical significance (P < .005). Risk alleles individually reduced AAO by 3 to 6 months. Burden analyses demonstrated that APOE contributes to 3.7% of the variation in AAO (R(2) = 0.256) over baseline (R(2) = 0.221), whereas the other 9 loci together contribute to 2.2% of the variation (R(2) = 0.242). We confirmed an association of APOE (OMIM 107741) variants with AAO among affected participants with LOAD and observed novel associations of CR1 (OMIM 120620), BIN1 (OMIM 601248), and PICALM (OMIM 603025) with AAO. In contrast to earlier hypothetical modeling, we show that the combined effects of Alzheimer disease risk variants on AAO are on the scale of, but do not exceed, the APOE effect. While the aggregate effects of risk loci on AAO may be significant, additional genetic contributions to AAO are individually likely to be small.

Population-based Analysis of Alzheimer’s Disease Risk Alleles Implicates Genetic Interactions

Reported odds ratios and population attributable fractions (PAF) for late-onset Alzheimer's disease (LOAD) risk loci (BIN1, ABCA7, CR1, MS4A4E, CD2AP, PICALM, MS4A6A, CD33, and CLU) come from clinically ascertained samples. Little is known about the combined PAF for these LOAD risk alleles and the utility of these combined markers for case-control prediction. Here we evaluate these loci in a large population-based sample to estimate PAF and explore the effects of additive and nonadditive interactions on LOAD status prediction performance. 2419 samples from the Cache County Memory Study were genotyped for APOE and nine LOAD risk loci from AlzGene.org. We used logistic regression and receiver operator characteristic analysis to assess the LOAD status prediction performance of these loci using additive and nonadditive models and compared odds ratios and PAFs between AlzGene.org and Cache County. Odds ratios were comparable between Cache County and AlzGene.org when identical single nucleotide polymorphisms were genotyped. PAFs from AlzGene.org ranged from 2.25% to 37%; those from Cache County ranged from .05% to 20%. Including non-APOE alleles significantly improved LOAD status prediction performance (area under the curve = .80) over APOE alone (area under the curve = .78) when not constrained to an additive relationship (p < .03). We identified potential allelic interactions (p values uncorrected): CD33-MS4A4E (synergy factor = 5.31; p < .003) and CLU-MS4A4E (synergy factor = 3.81; p < .016). Although nonadditive interactions between loci significantly improve diagnostic ability, the improvement does not reach the desired sensitivity or specificity for clinical use. Nevertheless, these results suggest that understanding gene-gene interactions may be important in resolving Alzheimer's disease etiology.

Genome-wide association analyses of onset age in late-onset Alzheimer disease (LOAD) demonstrate no strong effect outside of the APOE region

LOAD risk loci may also contribute to variation in age of onset (AAO) of LOAD, as do the allelic variants in APOE, however roles in AAO for the other newly identified risk loci (CLU, BIN1, and others) have not been explored. We examined variants at ten confirmed LOAD risk loci (APOE, CLU, PICALM, CR1, BIN1, CD2AP, EPHA1, the MS4A region, ABCA7, and CD33) to determine if they contribute to variation in AAO among 9,160 LOAD cases in the Alzheimer's disease Genetics Consortium (ADGC). We tested association with AAO for each locus using linear modeling with covariate adjustment for population substructure and performed a random-effects meta-analysis across datasets. We also examined genetic burden using genotype scores weighted by risk effect sizes to examine the aggregate contribution of these loci to variation in AAO. Analyses confirmed association of APOE regional variation with AAO (rs6857, P =3.30×10 -96), with statistically significant associations with AAO (P<0.005) demonstrated at several other LOAD risk loci, including rs6701713 in CR1 (P =0.00717), rs7561528 in BIN1 (P =0.00478), rs561655 in PICALM (P =0.00223). Associations remained largely unchanged after additional adjustment for dosage of APOE ε4 alleles. Burden analyses showed APOE contributes to 3.1% of variation in AAO (R 2 =0.078) whereas the other nine genes contribute to 1.1% of variation (R 2 =0.058) over baseline (R 2 =0.047). Secondary analyses of genome-wide association with AAO among non-risk loci identified several regions with multiple SNPs demonstrating suggestive associations (P<10 -6), including one nearing genome-wide statistical significance: MYO16 (47 SNPs; most significant: rs9521011, P =7.62×10 -8), CDH20 (4 SNPs; most significant: rs12956834, P =6.17×10 -6), and SGCZ (10 SNPs; most significant: rs7016159, P =7.70×10 -6). We confirmed the association of APOE variants with AAO among LOAD cases, and observe associations with AAO in CR1, BIN1, and PICALM. In contrast to earlier hypothetical modeling, we show that the combined effects of other loci do not exceed the effect of APOE on AAO, and if additional genetic contributions to AAO exist, they are likely very small individually or are hidden in gene-gene interactions.

Evaluation of memory endophenotypes for association with CLU, CR1, and PICALM variants in black and white subjects

BackgroundGenetic variants at the CLU, CR1, and PICALM loci associate with risk for late-onset Alzheimer's disease (LOAD) in genomewide association studies. In this study, our aim was to determine whether the LOAD risk variants at these three loci influence memory endophenotypes in black and white subjects. MethodsWe pursued an association study between single nucleotide polymorphism genotypes at the CLU, CR1, and PICALM loci and memory endophenotypes. We assessed black subjects (AA series: 44 with LOAD and 224 control subjects) recruited at Mayo Clinic Florida and whites recruited at Mayo Clinic Minnesota (RS series: 372 with LOAD and 1690 control subjects) and Florida (JS series: 60 with LOAD and 529 control subjects). Single nucleotide polymorphisms at the LOAD risk loci CLU (rs11136000), CR1 (rs6656401, rs3818361), and PICALM (rs3851179) were genotyped and tested for association with Logical Memory immediate recall, Logical Memory delayed recall, Logical Memory percent retention, Visual Reproduction immediate recall, Visual Reproduction delayed recall, and Visual Reproduction percent retention scores from the Wechsler Memory Scale–Revised using multivariable linear regression analysis, adjusting for age at exam, sex, education, and apolipoprotein E ε4 dosage. ResultsWe identified nominally significant or suggestive associations between the LOAD-risky CR1 variants and worse Logical Memory immediate recall scores in blacks (P = .068–.046, β = −2.7 to −1.2). The LOAD-protective CLU variant is associated with better logical memory endophenotypes in white subjects (P = .099–.027, β = 0.31–0.93). The CR1 associations persisted when the control subjects from the AA series were assessed separately. The CLU associations appeared to be driven by one of the white series (RS) and were also observed when the control subset from RS was analyzed. ConclusionThese results suggest for the first time that LOAD risk variants at CR1 may influence memory endophenotypes in blacks. In addition, the CLU LOAD-protective variant may confer enhanced memory in whites. Although these results would not remain significant after stringent corrections for multiple testing, they need to be considered in the context of the LOAD associations with which they have biological consistency. They also provide estimates for effect sizes on memory endophenotypes that could guide future studies. The detection of memory effects for these variants in clinically normal subjects, implies that these LOAD risk loci might modify memory prior to clinical diagnosis of AD.

Novel late-onset Alzheimer disease loci variants associate with brain gene expression

Recent genome-wide association studies (GWAS) of late-onset Alzheimer disease (LOAD) identified 9 novel risk loci. Discovery of functional variants within genes at these loci is required to confirm their role in Alzheimer disease (AD). Single nucleotide polymorphisms that influence gene expression (eSNPs) constitute an important class of functional variants. We therefore investigated the influence of the novel LOAD risk loci on human brain gene expression. We measured gene expression levels in the cerebellum and temporal cortex of autopsied AD subjects and those with other brain pathologies (∼400 total subjects). To determine whether any of the novel LOAD risk variants are eSNPs, we tested their cis-association with expression of 6 nearby LOAD candidate genes detectable in human brain (ABCA7, BIN1, CLU, MS4A4A, MS4A6A, PICALM) and an additional 13 genes ±100 kb of these SNPs. To identify additional eSNPs that influence brain gene expression levels of the novel candidate LOAD genes, we identified SNPs ±100 kb of their location and tested for cis-associations. CLU rs11136000 (p = 7.81 × 10(-4)) and MS4A4A rs2304933/rs2304935 (p = 1.48 × 10(-4)-1.86 × 10(-4)) significantly influence temporal cortex expression levels of these genes. The LOAD-protective CLU and risky MS4A4A locus alleles associate with higher brain levels of these genes. There are other cis-variants that significantly influence brain expression of CLU and ABCA7 (p = 4.01 × 10(-5)-9.09 × 10(-9)), some of which also associate with AD risk (p = 2.64 × 10(-2)-6.25 × 10(-5)). CLU and MS4A4A eSNPs may at least partly explain the LOAD risk association at these loci. CLU and ABCA7 may harbor additional strong eSNPs. These results have implications in the search for functional variants at the novel LOAD risk loci.

Alzheimer's disease genetics consortium: new Alzheimer's disease genes and gene interactions

Late-onset Alzheimer Disease (LOAD) is a highly heritable neurological disorder, with susceptibility influenced by known genetic risk factors at CLU, PICALM, and BIN1. The Alzheimer's Disease Genetics Consortium (ADGC) examined gnome-wide genotyping data from 11,840 LOAD cases and 10,931 cognitive controls to identify novel genetic variants and gene-gene interactions associated with LOAD. The ADGC dataset consists of fifteen datasets genotyped using genome-wide genotyping platforms (Illumina or Affymetrix). All cases met standard clinical LOAD definitions (or were confirmed through autopsy); controls were age-matched to cases and were cognitively intact at exam. To create a common set of SNPs across all datasets, genotype imputation was performed using IMPUTE software and the October 2010 release of the 1,000 Genomes Project data as a reference. Association analysis was performed within each dataset using logistic regression for case-control datasets and generalized estimating equations (GEE) for datasets that include family members. For both family and case-control datasets genotype was modeled using an additive model on the dosage of the minor allele. Age at onset/exam, gender, and variables for population substructure were included as covariates. Data were then analyzed across the datasets using a meta-analysis approach. For the gene-gene interaction analysis, SNPs will be pruned based on LD to reduce redundancy and then filtered using Biofilter to identify potential interactions/effects using known biological connections. Association analysis of the interactions will be performed using logistic regression and an additive model. Ten distinct genomic regions showed SNP associations in the combined meta-analysis, including previously associated regions CLU, CR1, PICALM, BIN1, and APOE10−8. Additional associated regions include MS4A, EPHA1, CD33, and ABCA7. Analysis of the gene-gene interaction is ongoing and will be presented. We have confirmed the association of several putative LOAD risk loci (APOE, CR1, CLU, PICALM and BIN1) and identified new risk loci as well (MS4A, EPHA1, CD33 and ABCA7). These associations confirm that common variations influence the etiology of LOAD. More importantly, they provide novel targets for functional assessment and further association studies.

Genome-wide association of familial late-onset Alzheimer's disease replicates BIN1 and CLU and nominates CUGBP2 in interaction with APOE.

Late-onset Alzheimer's disease (LOAD) is the most common form of dementia in the elderly. The National Institute of Aging-Late Onset Alzheimer's Disease Family Study and the National Cell Repository for Alzheimer's Disease conducted a joint genome-wide association study (GWAS) of multiplex LOAD families (3,839 affected and unaffected individuals from 992 families plus additional unrelated neurologically evaluated normal subjects) using the 610 IlluminaQuad panel. This cohort represents the largest family-based GWAS of LOAD to date, with analyses limited here to the European-American subjects. SNPs near APOE gave highly significant results (e.g., rs2075650, p = 3.2×10−81), but no other genome-wide significant evidence for association was obtained in the full sample. Analyses that stratified on APOE genotypes identified SNPs on chromosome 10p14 in CUGBP2 with genome-wide significant evidence for association within APOE ε4 homozygotes (e.g., rs201119, p = 1.5×10−8). Association in this gene was replicated in an independent sample consisting of three cohorts. There was evidence of association for recently-reported LOAD risk loci, including BIN1 (rs7561528, p = 0.009 with, and p = 0.03 without, APOE adjustment) and CLU (rs11136000, p = 0.023 with, and p = 0.008 without, APOE adjustment), with weaker support for CR1. However, our results provide strong evidence that association with PICALM (rs3851179, p = 0.69 with, and p = 0.039 without, APOE adjustment) and EXOC3L2 is affected by correlation with APOE, and thus may represent spurious association. Our results indicate that genetic structure coupled with ascertainment bias resulting from the strong APOE association affect genome-wide results and interpretation of some recently reported associations. We show that a locus such as APOE, with large effects and strong association with disease, can lead to samples that require appropriate adjustment for this locus to avoid both false positive and false negative evidence of association. We suggest that similar adjustments may also be needed for many other large multi-site studies.

Genome-wide Association Study Implicates a Chromosome 12 Risk Locus for Late-Onset Alzheimer Disease

Only Apolipoprotein E polymorphisms have been consistently associated with the risk of late-onset Alzheimer disease (LOAD), but they represent only a minority of the underlying genetic effect. To identify additional LOAD risk loci, we performed a genome-wide association study (GWAS) on 492 LOAD cases and 498 cognitive controls using Illumina's HumanHap550 beadchip. An additional 238 cases and 220 controls were used as a validation data set for single-nucleotide polymorphisms (SNPs) that met genome-wide significance. To validate additional associated SNPs (p < 0.0001) and nominally associated candidate genes, we imputed SNPs from our GWAS using a previously published LOAD GWAS(1) and the IMPUTE program. Association testing was performed with the Cochran-Armitage trend test and logistic regression, and genome-wide significance was determined with the False Discovery Rate-Beta Uniform Mixture method. Extensive quality-control methods were performed at both the sample and the SNP level. The GWAS confirmed the known APOE association and identified association with a 12q13 locus at genome-wide significance; the 12q13 locus was confirmed in our validation data set. Four additional highly associated signals (1q42, 4q28, 6q14, 19q13) were replicated with the use of the imputed data set, and six candidate genes had SNPs with nominal association in both the GWAS and the joint imputated data set. These results help to further define the genetic architecture of LOAD.