Interaction Analysis Reveals Complex Genetic Associations with Alzheimer's Disease in the CLU and ABCA7 Gene Regions

Abstract

Sporadic Alzheimer's disease (AD) is a polygenic neurodegenerative disorder. Single-nucleotide polymorphisms (SNPs) in multiple genes (e.g., CLU and ABCA7) have been associated with AD. However, none of them were characterized as causal variants that indicate the complex genetic architecture of AD, which is likely affected by individual variants and their interactions. We performed a meta-analysis of four independent cohorts to examine associations of 32 CLU and 50 ABCA7 polymorphisms as well as their 496 and 1225 pair-wise interactions with AD. The single SNP analyses revealed that six CLU and five ABCA7 SNPs were associated with AD. Ten of them were previously not reported. The interaction analyses identified AD-associated compound genotypes for 25 CLU and 24 ABCA7 SNP pairs, whose comprising SNPs were not associated with AD individually. Three and one additional CLU and ABCA7 pairs composed of the AD-associated SNPs showed partial interactions as the minor allele effect of one SNP in each pair was intensified in the absence of the minor allele of the other SNP. The interactions identified here may modulate associations of the CLU and ABCA7 variants with AD. Our analyses highlight the importance of the roles of combinations of genetic variants in AD risk assessment.

Keywords: Aging; SNPs interaction; compound genotype; dementia; genetic heterogeneity.

Figure 1

Volcano plot for the associations between Alzheimer’s disease (AD) and the 32 CLU and 50 ABCA7 single-nucleotide polymorphisms (SNPs). The x-axis displays effect sizes for SNPs, while the y-axis shows minus-logarithm-base-10-transformed q-values, representing false-discovery rate (FDR) adjusted significance. The dashed line indicates the significance threshold, set at −log10 (q-value = 0.05) = 1.3. Above this cutoff line, blue and red dots represent AD-associated SNPs in the CLU and ABCA7 genes, respectively. Non-significant SNPs in these two genes are depicted in light and dark gray, respectively. Numerical estimates are provided in Table 2, Tables S3–S6.

Figure 2

Volcano plot for compound genotype (CompG) analyses of the 496 single-nucleotide polymorphism (SNP) pairs selected within the CLU gene region. The x-axis displays effect sizes of CompGs, while the y-axis shows minus-logarithm-base-10-transformed q-values, representing false-discovery rate (FDR) adjusted significance. The dashed line indicates the significance threshold, set at −log10 (q-value = 0.05) = 1.3. Dark-gray dots below the cutoff line represent non-significant effects for Mm, mM, and mm CompGs. Light-gray dots above the cutoff line indicate significant SNP pairs whose comprising SNPs were associated with Alzheimer’s disease (AD) individually. Red and green dots above the cutoff line denote SNP pairs whose comprising SNPs were not associated with AD individually and were presented in Table 3 and Table S3.

Figure 3

Volcano plot for compound genotype (CompG) analyses of the 1225 single-nucleotide polymorphism (SNP) pairs selected within the ABCA7 gene region. The x-axis displays effect sizes of CompGs, while the y-axis shows minus-logarithm-base-10-transformed q-values, representing false-discovery rate (FDR) adjusted significance. The dashed line indicates the significance threshold, set at −log10 (q-value = 0.05) = 1.3. Dark-gray dots below the cutoff line represent non-significant effects for Mm, mM, and mm CompGs. Light-gray dots above the cutoff line indicate significant SNP pairs whose comprising SNPs were associated with Alzheimer’s disease (AD) individually. Red and green dots above the cutoff line denote SNP pairs whose comprising SNPs were not associated with AD individually and were presented in Table 4 and Table S5.