Epigenome-wide association study of cerebrospinal fluid-based biomarkers of Alzheimer's disease in cognitively normal individuals

Abstract

Introduction: Cerebrospinal fluid (CSF) biomarkers of Alzheimer's disease (AD) are reliable predictors of future AD risk. We investigated whether pre-clinical changes in AD CSF biomarkers are reflected in blood DNA methylation (DNAm) levels in cognitively normal participants.

Methods: We profiled blood-based DNAm with the EPIC array in participants without a diagnosis of cognitive impairment in the Emory Healthy Brain Study (EHBS; N = 495), Alzheimer's Disease Neuroimaging Initiative (N = 122), and Parkinson's Progression Markers Initiative (N = 118) cohorts. Their CSF amyloid beta 42, total tau (t-tau), and phosphorylated tau181 levels were quantified using Elecsys immunoassays. We conducted epigenome-wide association studies to assess associations between DNAm and CSF biomarkers of AD.

Results: In EHBS, no loci were Bonferroni significant after adjusting for confounding factors. In the meta-analysis of all three cohorts, DNAm in cg22976567 (LMNA) was significantly associated with higher CSF t-tau levels.

Discussion: Our study showed little evidence of an association between differential blood-based DNAm and pre-clinical AD CSF biomarkers.

Highlights: We conducted one of the largest (n = 735) blood DNA methylation (DNAm) studies of Alzheimer's disease cerebrospinal fluid (AD CSF) biomarkers. This is the first epigenome-wide association study in cognitively normal participants examining AD CSF biomarkers. Limited associations between blood DNAm and AD CSF biomarkers were identified.

Keywords: Alzheimer's disease; DNA methylation; amyloid; biomarker; blood; cerebrospinal fluid; epigenetics; epigenome‐wide association study.

FIGURE 1

EWAS of AD CSF biomarkers in 450 cognitively normal EHBS participants. Manhattan plots for the association between DNAm beta values and (A) tau, (B) p‐tau, (C) Aβ/tau ratio, (D) Aβ± , (E) p‐tau± . The dotted line represents the Bonferroni threshold ($p=7.55\times {10}^{-8}$). F, UpSet plot showing overlapping associations across the five CSF biomarkers (t‐tau, p‐tau, Aβ42/t‐tau, Aβ42± , p‐tau ±). A blue dot represents an association between DNAm beta values and the corresponding CSF biomarker with a p value < $1\times {10}^{-5}$ for at least one CpG site assigned to the corresponding gene. All associations were adjusted for age at baseline, sex, smoking (with or without smoking history), and estimated cell‐type proportions (B lymphocytes, natural killer cells, CD4 + T lymphocytes, CD8+ T lymphocytes, monocytes, neutrophils). Aβ, amyloid beta; AD, Alzheimer's disease; CSF, cerebrospinal fluid; DNAm, DNA methylation; EHBS, Emory Healthy Brain Study; EWAS, epigenome‐wide association study; p‐tau, phosphorylated tau; t‐tau, total tau.

FIGURE 2

Meta‐analysis of CSF biomarkers EWAS of cognitively normal individuals from EHBS (N = 450), ADNI (N = 122), and PPMI (N = 118). Manhattan plots for the association between DNAm beta values and (A) tau, (B) p‐tau, (C) Aβ/tau ratio, (D) Aβ ± , (E) p‐tau ± . The dotted line represents the Bonferroni threshold ($p=7.61\times {10}^{-8}$). F, UpSet plot showing overlapping associations across the five CSF biomarkers (t‐tau, p‐tau, Aβ42/t‐tau, Aβ42 ± , p‐tau ±). A blue dot represents an association between DNAm beta values and the corresponding CSF biomarker with a p value < $1\times {10}^{-5}$ for at least one CpG site assigned to the corresponding gene. Aβ, amyloid beta; AD, Alzheimer's disease; ADNI, Alzheimer's Disease Neuroimaging Initiative; CSF, cerebrospinal fluid; DNAm, DNA methylation; EHBS, Emory Healthy Brain Study; EWAS, epigenome‐wide association studies; PPMI, Parkinson's Progression Markers Initiative; p‐tau, phosphorylated tau; t‐tau, total tau.