Fine Mapping of Genetic Variants in BIN1, CLU, CR1 and PICALM for Association with Cerebrospinal Fluid Biomarkers for Alzheimer's Disease

Abstract

Recent genome-wide association studies of Alzheimer's disease (AD) have identified variants in BIN1, CLU, CR1 and PICALM that show replicable association with risk for disease. We have thoroughly sampled common variation in these genes, genotyping 355 variants in over 600 individuals for whom measurements of two AD biomarkers, cerebrospinal fluid (CSF) 42 amino acid amyloid beta fragments (Aβ42) and tau phosphorylated at threonine 181 (ptau181), have been obtained. Association analyses were performed to determine whether variants in BIN1, CLU, CR1 or PICALM are associated with changes in the CSF levels of these biomarkers. Despite adequate power to detect effects as small as a 1.05 fold difference, we have failed to detect evidence for association between SNPs in these genes and CSF Aβ42 or ptau181 levels in our sample. Our results suggest that these variants do not affect risk via a mechanism that results in a strong additive effect on CSF levels of Aβ42 or ptau181.

About the Authors:

John S. K. Kauwe

Contributed equally to this work with: John S. K. Kauwe, Carlos Cruchaga

* E-mail: [email protected]

Affiliation: Department of Biology, Brigham Young University, Provo, Utah, United States of America

Carlos Cruchaga

Contributed equally to this work with: John S. K. Kauwe, Carlos Cruchaga

Affiliations Department of Psychiatry, Washington University School of Medicine, St. Louis, Missouri, United States of America, Hope Center for Neurological Disorders, Washington University School of Medicine, St. Louis, Missouri, United States of America, Charles F. and Joanne Knight Alzheimer's Disease Research Center, Washington University School of Medicine, St. Louis, Missouri, United States of America

Celeste M. Karch

Affiliations Department of Psychiatry, Washington University School of Medicine, St. Louis, Missouri, United States of America, Hope Center for Neurological Disorders, Washington University School of Medicine, St. Louis, Missouri, United States of America

Brooke Sadler

Affiliation: Department of Psychiatry, Washington University School of Medicine, St. Louis, Missouri, United States of America

Mo Lee

Affiliation: Department of Biology, Brigham Young University, Provo, Utah, United States of America

Kevin Mayo

Affiliations Department of Psychiatry, Washington University School of Medicine, St. Louis, Missouri, United States of America, Hope Center for Neurological Disorders, Washington University School of Medicine, St. Louis, Missouri, United States of America

Wayne Latu

Affiliation: Department of Biology, Brigham Young University, Provo, Utah, United States of America

Manti Su'a

Affiliation: Department of Biology, Brigham Young University, Provo, Utah, United States of America

Anne M. Fagan

Affiliations Department of Neurology, Washington University School of Medicine, St. Louis, Missouri, United States of America, Hope Center for Neurological Disorders, Washington University School of Medicine, St. Louis, Missouri, United States of America, Charles F. and Joanne Knight Alzheimer's Disease Research Center, Washington University School of Medicine, St. Louis, Missouri, United States of America

David M. Holtzman

Affiliations Department of Neurology, Washington University School of Medicine, St. Louis, Missouri, United States of America, Hope Center for Neurological Disorders, Washington University School of Medicine, St. Louis, Missouri, United States of America, Charles F. and Joanne Knight Alzheimer's Disease Research Center, Washington University School of Medicine, St. Louis, Missouri, United States of America

John C. Morris

Affiliations Department of Neurology, Washington University School of Medicine, St. Louis, Missouri, United States of America, Charles F. and Joanne Knight Alzheimer's Disease Research Center, Washington University School of Medicine, St. Louis, Missouri, United States of America

Alzheimer's Disease Neuroimaging Initiative

Alison M. Goate

Affiliations Department of Psychiatry, Washington University School of Medicine, St. Louis, Missouri, United States of America, Department of Neurology, Washington University School of Medicine, St. Louis, Missouri, United States of America, Hope Center for Neurological Disorders, Washington University School of Medicine, St. Louis, Missouri, United States of America, Charles F. and Joanne Knight Alzheimer's Disease Research Center, Washington University School of Medicine, St. Louis, Missouri, United States of America

Introduction

Alzheimer's disease (AD) is the most common form of dementia and is neuropathologically characterized by extracellular senile plaques containing amyloid beta (Aβ) and intracellular neurofibrillary tangles containing hyperphosphorylated tau protein. Mendelian forms of the disease are caused by mutations in the amyloid precursor protein (APP) gene and the presenilin 1 and 2 genes (PSEN1 and PSEN2 respectively). While only apolipoprotein E (APOE) has been clearly identified as a susceptibility gene in the more common form of AD, data from recent genome-wide association studies has implicated several other common risk variants [1], [2], [3], [4], [5], [6], [7], [8]. Variants in bridging integrator 1 (BIN1), clusterin (CLU; also referred to as APOJ), complement component receptor 1 (CR1) and phosphatidylinositol binding clathrin assembly protein (PICALM) have already been reported to show replicable association with risk for AD [5], [6], [7], [8].

Identifying associated variants is an important first step toward understanding novel aspects of the etiology of disease. Characterization of the mechanisms by which these variants, or other functional variants in strong linkage disequilibrium, influence risk for disease will provide a better understanding of the biology of disease. Initial publications for these novel, AD associated variants provided some hypotheses for each of the reported genes. Previously reported work suggests that CLU and APOE may have additive effects on Aβ deposition [9]. CR1 may contribute to Aβ clearance [10]. Convincing evidence for an Aβ-related mechanism for risk exists for both of these genes. Less is known about the effects of BIN1 and PICALM on Aβ or tau metabolism: BIN1 function may affect risk for AD by altering neuronal membranes and clathrin mediated synaptic vessel formation [8], [11] and changes in PICALM function result in perturbation at the synapse, possibly altering synaptic vesicle cycling and leading to altered risk for AD [12], [13].

In our previous work we have shown the utility of using two well-established cerebrospinal fluid (CSF) biomarkers for AD, 42 amino acid fragments of amyloid beta (Aβ42; decreased in AD) and tau phosphorylated at threonine 181 (a proxy for hyperphosphorylated tau; ptau181; increased in AD), as endophenotypes for genetic studies of AD [14], [15], [16], [17]. In this approach we test variants for genetic association with CSF levels of Aβ42 and/or ptau181 levels. In cases where risk variants have already been identified this approach allows us to validate or generate hypotheses regarding the biological mechanism of risk. We can also take advantage of the increased statistical power and decreased heterogeneity of the biomarker phenotype relative to qualitative clinical diagnosis to identify novel variants that affect biomarker levels and aspects of disease [18]. In our previous studies using this approach we have successfully validated hypothesized effects of rs2986019 in CALHM1 on CSF Aβ42 levels [19], generated testable biological hypotheses for AD implicated variants [16], and identified novel variants in MAPT and PPP3R1 that are associated with both biomarker levels and rate of progression of AD [14], [17]. In this study we use an endophenotype-based approach to test predictions of biological effects on Aβ42 levels for variants in CLU and CR1 and to attempt to generate biological hypotheses of risk mechanism for BIN1, CLU, CR1 and PICALM.

Methods

Samples

CSF for the Washington University in St. Louis (WU) series was collected from 407 individuals after overnight fasting. CSF collection and processing as well as CSF biomarker measurements were performed as described previously [20]. Characteristics of the sample, including a breakdown of demographic information in demented and non-demented individuals can be found in table 1.

[Figure omitted. See PDF.]

Table 1. Sample characteristics.

https://doi.org/10.1371/journal.pone.0015918.t001

Data from 257 samples with biomarker data and either AD or cognitively normal diagnoses from the Alzheimer's Disease Neuroimaging Initiative (ADNI) were also used. Data used in the preparation of this article were obtained from the ADNI database (http://adni.loni.ucla.edu). The Principal Investigator of this initiative is Michael W. Weiner, M.D., VA Medical Center and University of California – San Francisco. ADNI is the result of efforts of many co-investigators from a broad range of academic institutions and private corporations, and subjects have been recruited from over 50 sites across the U.S. and Canada. The initial goal of ADNI was to recruit 800 adults, ages 55 to 90, to participate in the research — approximately 200 cognitively normal older individuals to be followed for 3 years, 400 people with mild cognitive impairment (MCI) to be followed for 3 years, and 200 people with early AD to be followed for 2 years.” For up-to-date information see www.adni-info.org. Sample characteristics, including age, clinical dementia rating, gender, APOE ε4 status and mean and standard deviation of the CSF biomarkers can be found in table 1. ADNI phenotype and GWAS data are publically available (http://adni.loni.ucla.edu/). The genotypes from this study will be provided upon request to the authors.

Biomarker values in both samples were measured using internal standards and controls that ensure consistent and reliable measurements. Differences between the measured values in the WU and ADNI samples are likely to be due to differences in the antibodies and measurement technologies used for each series (e.g. standard ELISA with Innotest in the WU samples, Luminex with AlzBia3 in the ADNI samples). It is also possible that the inclusion of more AD cases and older individuals in the ADNI data or differences in the number of freeze thaw cycles prior to analysis (1 cycle for WU samples and 2 cycles for ADNI samples) accounts for some of the variation in the biomarker measurements. CSF biomarkers in the two samples show association with similar covariates [17], [19].

SNP selection and genotyping

For each gene we downloaded the list of SNPs in the gene region (and approximately 500 kb of flanking sequence) from HapMap. These SNPs were then evaluated for putative functional effects using SNPseek and Pupasuite. SNPs with putative function and SNPs that showed association in the original published reports were designated as forced tags in the tagging algorithm in Haploview when an r2 cutoff of 0.8 was applied. A total of 283 SNPs were selected (see table S1 for a list of all SNPs in the study).

Genotyping was performed using Applied Biosystems OpenArray technology (http://www3.appliedbiosystems.com/cms/groups/mcb_support/documents/generaldocuments/cms_058198.pdf), a means of running multiple TaqMan assays together on one chip. 125 ng per sample was added to the reaction mix and spread over 64 assays, the chips were thermocycled as described in the linked protocol and imaged. Results were analyzed using the Applied Biosystems Genotyper software (https://products.appliedbiosystems.com/ab/en/US/adirect/ab?cmd=catNavigate2&catID=607267&tab=Literature). Samples were analyzed on a plate-by-plate basis in the context of all the samples to eliminate variation in calls between plates. SNPs that deviated from HW equilibrium (p-value threshold 0.001), had a genotyping rate lower than 95% or a minor allele frequency of less than 5% were removed. Samples with a genotyping rate lower than 95% were also removed. After application of these quality control criteria there were 664 samples and 233 SNPs.

Analysis

Ptau181 levels were normally distributed after log-log transformation. Using stepwise regression analysis we identified age, APOE ε4 genotype and Clinical Dementia Rating (CDR) as significant covariates to be included in the model. Gender was not significantly associated with CSF ptau181, and was not included in the model. Association with genotype was tested using ANCOVA after adjustment for these covariates. For the combined analysis we also included site as a covariate. All analyses were also performed without CDR as a covariate but there were no qualitative differences in the results.

Aβ42 levels were not normally distributed even after a variety of transformations were applied. For this reason Aβ42 analyses were performed using permutation based testing in PLINK (1 million permutations) [21]. For the WU data the statistical analyses included age, CDR and APOE ε4 genotype as covariates; the ADNI model included CDR and APOE ε4 genotype. Age was not significantly associated with biomarker levels in the ADNI sample (due to lack of variation in age in the ADNI sample) and was therefore not included as a covariate for analyses of that sample alone. Site was included in the combined analysis in addition to age, CDR and APOE ε4 genotype. There were no qualitative differences in the results when run without CDR as a covariate.

The alpha level for this study using Bonferroni correction for 233 tests is 0.00021. A less conservative correction using the Eigen values of the SNP correlation matrix to estimate the effective number of tests yielded an adjusted alpha of 0.00022 [22], [23]. Using either adjusted alpha yields the same qualitative conclusions from these data.

Haplotype and set-based analyses were performed using PLINK with default settings [21]. The SNPs selected for fine mapping around each GWAS hit were defined as a set and 10,000 permutations were run using the same models as described previously for each phenotype.

Power

Power for the overall F test in a one-way, three-group analysis of variance was calculated using proc power in SAS. The effect size, measured in “fold-difference” between the means at which power was estimated at 0.80 was calculated for minor allele frequencies from 0.10 to 0.50 and alpha levels of 0.05 and 0.00021 (the Bonferroni correction for 233 tests) assuming markers do not deviate from Hardy-Weinberg Equilibrium (table 2).

[Figure omitted. See PDF.]

Table 2. Power analyses.

https://doi.org/10.1371/journal.pone.0015918.t002

Results

We failed to detect significant association between the CSF biomarker levels and SNPs in BIN1. rs3820757 (p = 0.31), rs744373 (p = 0.44) and rs2276582 (p = 0.48) had the smallest p-value for association with CSF Aβ42 levels but did not show statistically significant association in the combined sample (WU+ADNI CSF samples, table 3). The three top hits in BIN1 for CSF ptau181 did not show statistically significant association (table 4). The SNP identified in previous GWAS, rs744373, did not show an association with CSF ptau181 levels in the combined sample (p = 0.79; table 4). Set-based analyses of the 14 BIN1 fine mapping SNPs were not significant for either biomarker phenotype (Aβ42 p = 0.42; ptau181 p = 0.37). Haplotype analyses also failed to identify significant association with Aβ42 and ptau181.

[Figure omitted. See PDF.]

Table 3. Top hits and GWAS SNPs for CSF Aβ42.

https://doi.org/10.1371/journal.pone.0015918.t003

[Figure omitted. See PDF.]

Table 4. Top hits and GWAS SNPs for ptau181.

https://doi.org/10.1371/journal.pone.0015918.t004

We failed to detect evidence for association between rs11136000 in CLU, which has been implicated in risk for AD, and CSF Aβ42 (p = 0.79) or ptau181 (p = 0.78) levels (Tables 3 and 4) in the combined sample. The top hits for CSF Aβ42 levels in CLU were rs10216623 (p = 0.011), rs2640734 (p = 0.036) and rs17057419 (p = 0.056); but these p-values do not pass multiple test correction. The top hits in CLU for association with CSF ptau181 were rs2439497 (p = 0.0010), rs2640734 (p = 0.004) and rs576256 (p = 0.0081) in the combined sample. The p-value threshold for Bonferroni correction for the entire study is 0.00021; therefore none of these p-values pass the multiple test correction. Set-based analyses of 57 SNPs from the CLU fine mapping set showed that there was evidence for association with ptau181 levels (p = 0.034). However this p-value is not significant after correction for the 4 SNP sets that were tested. There was no evidence for association in the set-based analyses for Aβ42 levels (p = 1). Haplotype analyses failed to identify significant association with either CSF phenotype.

The SNP in CR1 that is implicated in risk for disease from recent GWAS is rs6656401. We failed to detect association between this SNP and either CSF Aβ42 (p = 0.63) or ptau181 (p = 0.52) levels in the combined sample. In CR1 no SNPs were significant with either phenotype (table 3 and 4) and set-based analyses of the 24 SNPs within the CR1 fine-mapping region provided no evidence for association (Aβ42 p = 1; ptau181 p = 1). Haplotype analyses failed to detect significant association with Aβ42 and ptau181.

Rs3851179, the PICALM SNP identified in the recent GWAS studies showed no evidence of association with either CSF Aβ42 or ptau181 levels (Tables 3 and 4). The top hits for CSF Aβ42 levels were rs7113656 (p = 0.0090), rs11234454 (p = 0.010) and rs10792828 (p = 0.011). The top hits for CSF ptau181 levels were rs638509 (p = 0.00098), rs694353 (p = 0.0010) and rs10898433 (p = 0.0012). Set-based analyses of 138 SNPs in the PICALM fine-mapping region failed to detect evidence for association with either Aβ42 (p = 0.56) or ptau181 (p = 0.47). Haplotype analyses also failed to identify significant association with these CSF phenotypes.

There is evidence of an interaction between SNPs in PICALM and APOE ε4, in at least one study the effects of risk associated SNPs in PICALM were found to be much stronger in the presence of the APOE ε4 allele [6]. To investigate this interaction we included an interaction term for PICALM SNPs and the presence or absence of APOE ε4 and performed association analyses between PICALM SNP genotypes and CSF Aβ42 and ptau181 in APOE ε4 positive and APOE ε4 negative substrata and using an APOE ε4 by SNP interaction term in the combined sample. We failed to detect statistically significant associations in the APOE ε4 negative and APOE ε4 positive substrata and in the interaction analysis (table S2). The most significant p-value from these three analyses is for association of rs11234542 with CSF ptau181 levels in the APOE ε4 negative substratum (p = 5.31×10-5; table S2). In this case the minor allele of rs11234542 was associated with higher CSF ptau181 levels.

Power to detect additive effects of more than an approximately 1.02 fold difference between the means was greater than 0.80 when alpha is 0.05 for all SNPs in this study. Even with an extremely conservative alpha of 0.00021 (Bonferroni correction for 233 tests) all SNPs in this study had power estimated at greater than 0.80 for at least a 1.05 fold difference (for reference significant association detected between rs2986019 in CALHM1 on CSF Aβ42 levels by Kauwe et al was a 1.05 fold difference [19]).

Discussion

While there were some suggestive associations of CSF ptau181 levels with PICALM SNPs, we failed to detect association that was significant after multiple test correction between SNPs in BIN1, CLU, CR1 or PICALM and CSF Aβ42 or ptau181 levels in our analyses. The power calculations suggest that our single snp tests had a very high probability of detecting a strong, additive effect (1.05 fold difference) on CSF biomarker levels if it were present. The lack of significant associations suggests that there is not likely to be a strong additive genetic effect between the SNPs in this study and CSF levels of Aβ42 or ptau181. A recently published GWAS of 17 plasma lipoproteins in a sample of over 17,000 individuals identified 43 associated loci [24]. Close review of the results of that study shows that approximately one third of the significant associations show less than a 1.05-fold difference and about one sixth show less than a 1.03-fold difference. These findings suggest that small additive effects on protein levels are common and that much larger numbers of CSF samples will be required to precisely determine associations between Alzheimer's disease risk variants and biomarker levels. Greater sample sizes, while not immediately available, will be possible as we and other groups continue to collect additional specimens. Our set-based analyses suggest that there may be a signal for association with CSF ptau181 in the CLU gene region. This result, and the lack of signal with Aβ levels, are unexpected given data suggesting additive effects of CLU and APOE on Aβ deposition in mice [9]. The association is not significant after correction for the four sets that were tested but suggests that with increased power significant biomarker association may be detected.

An alternative interpretation of our results is that, given the lack of association with Aβ42 and ptau181, variants in these genes may modulate risk for AD through mechanisms that do not directly alter CSF levels of Aβ42 or ptau181. CLU, PICALM and CR1 participate in other processes not related to Aβ or tau aggregation, processing or clearance, and therefore studies of the role of these proteins in the brain may reveal evidence for additional disease mechanisms, which go beyond Aβ or tau accumulation. In fact there are several studies that link these genes with lipid metabolism and inflammatory pathways. Two of the identified AD susceptibility genes (CLU, CR1) have known functions in the immune system, which suggests a possible role for the immune system in the risk for AD. [25], [26]. Possible links between the genes in this study and lipid metabolism have also been identified and are reviewed by Jones et al [27].

Our study was designed specifically to detect additive genetic effects of common SNPs. Failure to detect significant association in this study design does not rule out, or even directly address, the possibility that these genes harbor rare variation that influence these biomarkers or that common variants in these genes have very small effects on these biomarkers. Finally, this approach may not detect complex, non-additive genetic mechanisms, such as complex gene-gene or gene-environment interactions that may modulate biomarker levels.

Supporting Information

[Figure omitted. See PDF.]

Table S1.

A complete list of SNPs in the study, position, minor allele frequencies (MAF), and p-values for association with CSF ptau181 and Aβ42 levels in the Washington University (WU), Alzheimer's Disease Neuroimaging Initiative (ADNI) and combined sample sets. SNPs with values of #N/A failed to meet QC criteria.

https://doi.org/10.1371/journal.pone.0015918.s001

(DOCX)

Table S2.

Association of SNPs in interaction with APOE ε4 alleles. For each SNP in the PICALM gene region p-values for association with CSF ptau181 and Aβ42 levels for the SNP by presence/absence of the APOE ε4 allele interaction term, association in individuals without an APOE ε4 allele and association in individuals with an APOE ε4 allele are shown.

https://doi.org/10.1371/journal.pone.0015918.s002

(DOCX)

Acknowledgments

We thank contributors, including the Alzheimer's Disease Centers who collected samples used in this study, as well as patients and their families, whose help and participation made this work possible. The authors also acknowledge the contributions of the Genetics, Clinical, Psychometric, Biomarker and Biostatistics Cores of the Washington University Alzheimer's Disease Research Center.

Some of the data used in the preparation of this article were obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database (http://adni.loni.ucla.edu/). As such, the investigators within the ADNI contributed to the design and implementation of ADNI and/or provided data but did not participate in analysis or writing of this report. ADNI investigators include (complete listing available at http://www.loni.ucla.edu/ADNI/About/About_Investigators.shtml).

Author Contributions

Conceived and designed the experiments: AMG JSKK CC. Performed the experiments: CC CMK BS KM. Analyzed the data: JSKK CC CMK BS ML WL MS. Contributed reagents/materials/analysis tools: DMH AMF JCM ADNI. Wrote the paper: JSKK CC AMG ML KM WL CMK.

Citation: Kauwe JSK, Cruchaga C, Karch CM, Sadler B, Lee M, Mayo K, et al. (2011) Fine Mapping of Genetic Variants in BIN1, CLU, CR1 and PICALM for Association with Cerebrospinal Fluid Biomarkers for Alzheimer's Disease. PLoS ONE6(2): e15918. https://doi.org/10.1371/journal.pone.0015918

References

1. Beecham GW, Martin ER, Li YJ, Slifer MA, Gilbert JR, et al. (2009) Genome-wide association study implicates a chromosome 12 risk locus for late-onset Alzheimer disease. Am J Hum Genet 84: 35–43.GW BeechamER MartinYJ LiMA SliferJR Gilbert2009Genome-wide association study implicates a chromosome 12 risk locus for late-onset Alzheimer disease.Am J Hum Genet843543

2. Bertram L, Lange C, Mullin K, Parkinson M, Hsiao M, et al. (2008) Genome-wide association analysis reveals putative Alzheimer's disease susceptibility loci in addition to APOE. Am J Hum Genet 83: 623–632.L. BertramC. LangeK. MullinM. ParkinsonM. Hsiao2008Genome-wide association analysis reveals putative Alzheimer's disease susceptibility loci in addition to APOE.Am J Hum Genet83623632

3. Carrasquillo MM, Zou F, Pankratz VS, Wilcox SL, Ma L, et al. (2009) Genetic variation in PCDH11X is associated with susceptibility to late-onset Alzheimer's disease. Nat Genet 41: 192–198.MM CarrasquilloF. ZouVS PankratzSL WilcoxL. Ma2009Genetic variation in PCDH11X is associated with susceptibility to late-onset Alzheimer's disease.Nat Genet41192198

4. Coon KD, Myers AJ, Craig DW, Webster JA, Pearson JV, et al. (2007) A high-density whole-genome association study reveals that APOE is the major susceptibility gene for sporadic late-onset Alzheimer's disease. J Clin Psychiatry 68: 613–618.KD CoonAJ MyersDW CraigJA WebsterJV Pearson2007A high-density whole-genome association study reveals that APOE is the major susceptibility gene for sporadic late-onset Alzheimer's disease.J Clin Psychiatry68613618

5. Harold D, Abraham R, Hollingworth P, Sims R, Gerrish A, et al. (2009) Genome-wide association study identifies variants at CLU and PICALM associated with Alzheimer's disease. Nat Genet 41: 1088–1093.D. HaroldR. AbrahamP. HollingworthR. SimsA. Gerrish2009Genome-wide association study identifies variants at CLU and PICALM associated with Alzheimer's disease.Nat Genet4110881093

6. Jun G, Naj AC, Beecham GW, Wang LS, Buros J, et al. (2010) Meta-analysis Confirms CR1, CLU, and PICALM as Alzheimer Disease Risk Loci and Reveals Interactions With APOE Genotypes. Arch Neurol. G. JunAC NajGW BeechamLS WangJ. Buros2010Meta-analysis Confirms CR1, CLU, and PICALM as Alzheimer Disease Risk Loci and Reveals Interactions With APOE Genotypes.Arch Neurol

7. Lambert JC, Heath S, Even G, Campion D, Sleegers K, et al. (2009) Genome-wide association study identifies variants at CLU and CR1 associated with Alzheimer's disease. Nat Genet 41: 1094–1099.JC LambertS. HeathG. EvenD. CampionK. Sleegers2009Genome-wide association study identifies variants at CLU and CR1 associated with Alzheimer's disease.Nat Genet4110941099

8. Seshadri S, Fitzpatrick AL, Ikram MA, DeStefano AL, Gudnason V, et al. (2010) Genome-wide analysis of genetic loci associated with Alzheimer disease. Jama 303: 1832–1840.S. SeshadriAL FitzpatrickMA IkramAL DeStefanoV. Gudnason2010Genome-wide analysis of genetic loci associated with Alzheimer disease.Jama30318321840

9. DeMattos RB, Cirrito JR, Parsadanian M, May PC, O'Dell MA, et al. (2004) ApoE and clusterin cooperatively suppress Abeta levels and deposition: evidence that ApoE regulates extracellular Abeta metabolism in vivo. Neuron 41: 193–202.RB DeMattosJR CirritoM. ParsadanianPC MayMA O'Dell2004ApoE and clusterin cooperatively suppress Abeta levels and deposition: evidence that ApoE regulates extracellular Abeta metabolism in vivo.Neuron41193202

10. Wyss-Coray T, Yan F, Lin AH, Lambris JD, Alexander JJ, et al. (2002) Prominent neurodegeneration and increased plaque formation in complement-inhibited Alzheimer's mice. Proc Natl Acad Sci U S A 99: 10837–10842.T. Wyss-CorayF. YanAH LinJD LambrisJJ Alexander2002Prominent neurodegeneration and increased plaque formation in complement-inhibited Alzheimer's mice.Proc Natl Acad Sci U S A991083710842

11. Wigge P, Kohler K, Vallis Y, Doyle CA, Owen D, et al. (1997) Amphiphysin heterodimers: potential role in clathrin-mediated endocytosis. Mol Biol Cell 8: 2003–2015.P. WiggeK. KohlerY. VallisCA DoyleD. Owen1997Amphiphysin heterodimers: potential role in clathrin-mediated endocytosis.Mol Biol Cell820032015

12. Harel A, Wu F, Mattson MP, Morris CM, Yao PJ (2008) Evidence for CALM in directing VAMP2 trafficking. Traffic 9: 417–429.A. HarelF. WuMP MattsonCM MorrisPJ Yao2008Evidence for CALM in directing VAMP2 trafficking.Traffic9417429

13. Masliah E, Mallory M, Alford M, DeTeresa R, Hansen LA, et al. (2001) Altered expression of synaptic proteins occurs early during progression of Alzheimer's disease. Neurology 56: 127–129.E. MasliahM. MalloryM. AlfordR. DeTeresaLA Hansen2001Altered expression of synaptic proteins occurs early during progression of Alzheimer's disease.Neurology56127129

14. Kauwe JS, Cruchaga C, Mayo K, Fenoglio C, Bertelsen S, et al. (2008) Variation in MAPT is associated with cerebrospinal fluid tau levels in the presence of amyloid-beta deposition. Proc Natl Acad Sci U S A 105: 8050–8054.JS KauweC. CruchagaK. MayoC. FenoglioS. Bertelsen2008Variation in MAPT is associated with cerebrospinal fluid tau levels in the presence of amyloid-beta deposition.Proc Natl Acad Sci U S A10580508054

15. Kauwe JS, Jacquart S, Chakraverty S, Wang J, Mayo K, et al. (2007) Extreme cerebrospinal fluid amyloid beta levels identify family with late-onset Alzheimer's disease presenilin 1 mutation. Ann Neurol 61: 446–453.JS KauweS. JacquartS. ChakravertyJ. WangK. Mayo2007Extreme cerebrospinal fluid amyloid beta levels identify family with late-onset Alzheimer's disease presenilin 1 mutation.Ann Neurol61446453

16. Kauwe JS, Wang J, Mayo K, Morris JC, Fagan AM, et al. (2009) Alzheimer's disease risk variants show association with cerebrospinal fluid amyloid beta. Neurogenetics 10: 13–17.JS KauweJ. WangK. MayoJC MorrisAM Fagan2009Alzheimer's disease risk variants show association with cerebrospinal fluid amyloid beta.Neurogenetics101317

17. Cruchaga C, Kauwe JSK, Mayo K, Spiegel N, Bertelsen S, et al. (2010) SNPs Associated with Cerebrospinal Fluid Phospho-Tau Levels Influence Rate of Decline in Alzheimer's Disease. PLoS Genet 6: e1001101.C. CruchagaJSK KauweK. MayoN. SpiegelS. Bertelsen2010SNPs Associated with Cerebrospinal Fluid Phospho-Tau Levels Influence Rate of Decline in Alzheimer's Disease.PLoS Genet6e1001101

18. Price JL, Morris JC (1999) Tangles and plaques in nondemented aging and “preclinical” Alzheimer's disease. Ann Neurol 45: 358–368.JL PriceJC Morris1999Tangles and plaques in nondemented aging and “preclinical” Alzheimer's disease.Ann Neurol45358368

19. Kauwe JS, Cruchaga C, Bertelsen S, Mayo K, Latu W, et al. (2010) Validating Predicted Biological Effects of Alzheimer's Disease Associated SNPs Using CSF Biomarker Levels. J Alzheimers Dis. JS KauweC. CruchagaS. BertelsenK. MayoW. Latu2010Validating Predicted Biological Effects of Alzheimer's Disease Associated SNPs Using CSF Biomarker Levels.J Alzheimers Dis

20. Fagan AM, Mintun MA, Mach RH, Lee SY, Dence CS, et al. (2006) Inverse relation between in vivo amyloid imaging load and cerebrospinal fluid Abeta42 in humans. Ann Neurol 59: 512–519.AM FaganMA MintunRH MachSY LeeCS Dence2006Inverse relation between in vivo amyloid imaging load and cerebrospinal fluid Abeta42 in humans.Ann Neurol59512519

21. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, et al. (2007) PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 81: 559–575.S. PurcellB. NealeK. Todd-BrownL. ThomasMA Ferreira2007PLINK: a tool set for whole-genome association and population-based linkage analyses.Am J Hum Genet81559575

22. Cheverud JM (2001) A simple correction for multiple comparisons in interval mapping genome scans. Heredity 87: 52–58.JM Cheverud2001A simple correction for multiple comparisons in interval mapping genome scans.Heredity875258

23. Li J, Ji L (2005) Adjusting multiple testing in multilocus analyses using the eigenvalues of a correlation matrix. Heredity 95: 221–227.J. LiL. Ji2005Adjusting multiple testing in multilocus analyses using the eigenvalues of a correlation matrix.Heredity95221227

24. Chasman DI, Pare G, Mora S, Hopewell JC, Peloso G, et al. (2009) Forty-three loci associated with plasma lipoprotein size, concentration, and cholesterol content in genome-wide analysis. PLoS Genet 5: e1000730.DI ChasmanG. PareS. MoraJC HopewellG. Peloso2009Forty-three loci associated with plasma lipoprotein size, concentration, and cholesterol content in genome-wide analysis.PLoS Genet5e1000730

25. McLaughlin L, Zhu G, Mistry M, Ley-Ebert C, Stuart WD, et al. (2000) Apolipoprotein J/clusterin limits the severity of murine autoimmune myocarditis. J Clin Invest 106: 1105–1113.L. McLaughlinG. ZhuM. MistryC. Ley-EbertWD Stuart2000Apolipoprotein J/clusterin limits the severity of murine autoimmune myocarditis.J Clin Invest10611051113

26. Jozsi M, Prechl J, Bajtay Z, Erdei A (2002) Complement receptor type 1 (CD35) mediates inhibitory signals in human B lymphocytes. J Immunol 168: 2782–2788.M. JozsiJ. PrechlZ. BajtayA. Erdei2002Complement receptor type 1 (CD35) mediates inhibitory signals in human B lymphocytes.J Immunol16827822788

27. Jones L, Harold D, Williams J (2010) Genetic evidence for the involvement of lipid metabolism in Alzheimer's disease. Biochim Biophys Acta 1801: 754–761.L. JonesD. HaroldJ. Williams2010Genetic evidence for the involvement of lipid metabolism in Alzheimer's disease.Biochim Biophys Acta1801754761