Genome-wide association study identifies CBFA2T3 affecting the rate of CSF Aβ42 decline in non-demented elders

Abstract

Brain amyloid deposition is an early pathological event in Alzheimer's disease (AD), and abnormally low levels amyloid-β₄₂ peptide (Aβ₄₂) in cerebrospinal fluid (CSF) can be detected in preclinical AD. To identify the genetic determinants that regulate the rate of CSF Aβ₄₂ decline among non-demented elders, we conducted a genome-wide association study involved 321 non-demented elders from Alzheimer's Disease Neuroimaging Initiative (ADNI) 1/GO/2 cohorts restricted to non-Hispanic Caucasians. A novel genome-wide significant association of higher annualized percent decline of CSF Aβ₄₂ in the gene CBFA2T3 (CBFA2/RUNX1 translocation partner 3; rs13333659-T; p = 2.24 × 10^-9) was identified. Besides displaying abnormal CSF Aβ₄₂ levels, rs13333659-T carriers were more likely to exhibit a greater longitudinal cognitive decline (p = 0.029, β = 0.097) and hippocampal atrophy (p = 0.029, β = -0.160) in the non-demented elders, especially for the participants who were amyloid-positive at baseline. These findings suggest rs13333659 in CBFA2T3 as a risk locus to modulate the decline rate of CSF Aβ₄₂ preceding the onset of clinical symptoms.

Keywords: Alzheimer’s disease; GWAS; amyloid; cerebrospinal fluid; genetics.

Figure 1

Manhattan plot (A), quantile-quantile plot (B) and regional plots (C and D) for the GWAS of longitudinal changes of CSF Aβ_42. (A) A genome-wide significant association (P < 5×10⁻⁸; red line) with longitudinal change of CSF Aβ₄₂ was identified on chromosome 16 within CBFA2T3. Suggestive associations are at the threshold of P < 1×10⁻⁵ (blue line). (B) Quantile-quantile plot. (C) Regional association results for the 88.9 Mb to 89.2 Mb region of chromosome 16. (D) Association results for the 88.9 Mb to 89.2 Mb region of chromosome 16 controlling for rs13333659.

Figure 2

Mean changes of CSF Aβ₄₂ over time in rs13333659 minor allele carriers vs. non-carriers. (A) Mean concentrations of CSF Aβ₄₂ ± SE (standard error) change for rs13333659 minor allele carriers vs. non-carriers in total GWAS cohort during follow-up period. (B) Mean concentrations of CSF Aβ₄₂ ± SE change for rs13333659 minor allele carriers vs. non-carriers in baseline amyloid-positive group during follow-up period.

Figure 3

Effects of CBFA2T3 rs13333659-T on cognitive performance and hippocampal volume over time. Data from linear mixed-effects models adjusted for age, gender, educational level, APOE ε4 genotype, disease status, follow-up duration, as well as intracranial volume for hippocampal volume. ADAS-cog 11 indicates Alzheimer Disease Assessment Scale-cognitive subscale.

Figure 4

Cryptic relatedness and population stratification checked with genomic identity-by-descent (IBD) and multidimensional scaling (MDS) components. (A) MDS plot of ADNI non-Hispanic White samples. Two samples were outliers based on the second MDS component (at above of plot; 024_S_2239 and 024_S_4084). (B) MDS plot of ADNI samples overlaid on HapMap samples. The ancestry of the HapMap participants is shown by the point color. Abbreviations: ADNI: Alzheimer’s Disease Neuroimaging Initiative; ASW: African ancestry in Southwest USA; CEU: Utah residents with Northern and Western European ancestry from the CEPH collection; CHB: Han Chinese individuals from Beijing, China; CHD: Chinese in Metropolitan Denver, Colorado; GIH: Gujarati Indians in Houston, Texas; JPT: Tokyo, Japan; LWK: Luhya in Webuye, Kenya; MEX: Mexican ancestry in Los Angeles, California; MKK: Maasai in Kinyawa, Kenya; TSI: Tuscans in Italy; YRI: Yoruba in Ibadan, Nigeria.