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

Abstract

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.

Figure 1. Principal components analysis of the complete sample, based on all ethnicities.

Red: European-American subjects. PC1 and PC2: first and second principal component.

Figure 2. First four principal components (PCs) in the European-American sample alone.

Colors represent inferred ancestry. Black: northwest (NW) Europe; green: southeast (SE) Europe; cyan: Ashkenazi Jewish (AJ); magenta: indeterminate (omitted from subpopulation analyses).

Figure 3. Cumulative distribution of absolute value of allele frequency differences between subpopulations and APOE genotypes.

Panel A: subjects from NW and SE (dotted line), AJ and SE (dashed line), and NW and AJ (solid line) groups. Panel B: cumulative distribution of European-American PC4 values as a function of APOE genotype for ε4 homozygotes (dotted line), ε4 heterozygotes (dashed line); genotypes with no ε4 (solid line). In panel A, the horizontal axis is truncated at 0.25 despite a few rare allele frequency differences that extend to 0.59; in panel B the vertical axis is only presented for the upper quartile of the distributions, where the curves are differentiated.

Figure 4. Genome scan of European-American subjects.

Panel A: CC_all sample analyzed as a single population; panel B: stratified analysis of CC_all sample that accounts for three subpopulations (NW, SE, AJ); panel C: stratified analysis of CC_all sample across four APOE genotypes; panel D: CC_un sample, with covariate adjustment for the number of ε2 and ε4 alleles. Plots have been truncated at −log₁₀p = 10 on the vertical axis to more easily visualize results for most of the genome. Multiple SNPs near APOE on chromosome 19 yielded −log₁₀p≫10 in the analyses that did not control for APOE (Panels A and B, see text for details), and are represented by a single triangle at the top of each such panel. Horizontal line shows genome-wide significance level.

Figure 5. Stratified analysis of APOE-defined subgroups of all European-American subjects.

Panels A: ε4/ε4 genotype, B: ε3/ε4 genotype; C: ε3/ε3 genotype; and D: ε2/ε2+ε2/ε3 combined genotype. Horizontal line shows genome-wide significance level.

Figure 6. Quality control evaluation of association tests in the CC_un and CC_all samples.

Panels A, B: Quantile difference plots for association tests excluding SNPs in the APOE region; and panel C: −log₁₀(p) for the same analyses for the 95 SNPs in the APOE region. For panels A and B, results are shown, for N tests, as the difference of the i^th of N ordered observed (−log₁₀(p_i)) and expected (−log₁₀(i/N)) quantiles plotted against the expected quantiles. A: results for the CC_un sample, with grey: PCA adjusted; magenta: unadjusted analysis; cyan: ε4 adjustment; black: full adjustment. B: results for the sample containing related individuals; grey: unadjusted analysis of NW subgroup; magenta: unadjusted analysis of CC_all; cyan: ε4-stratified analysis of CC_all; black: full adjustment. C: UN depicts results for analysis of CC_un; REL depicts results for analysis of the larger sample, in both cases for the same four conditions and colors as in panels A and B.

Figure 7. Quantile difference plot of tests of allele frequency differences in APOE ε4-carrier versus non-carrier cases.

SNPs in the APOE region are not included.