Epigenetic aging is associated with atypical neurodegenerative patterns in Alzheimer’s disease
Poster No:
318
Submission Type:
Abstract Submission
Authors:
Lasya Sreepada1, Sandhitsu Das1, Paul Yushkevich1, Wanding Zhou1, David Wolk1, Corey McMillan1
Institutions:
1University of Pennsylvania, Philadelphia, PA
First Author:
Co-Author(s):
Introduction:
While Alzheimer's disease (AD) is typically considered an amnestic, multi-domain disorder, at least 15% of individuals are considered atypical presentations. Atypical presentations are associated with younger age of onset, whereas late-onset AD cases tend to present typically. Although atypical presentations tend to have younger age of onset, age is often defined chronologically, and we hypothesize that epigenetic clock measures of biological age may capture variance contributing to atypical neurodegenerative patterns. We operationalize atypicality in AD as relative neurodegeneration in cortex versus medial temporal lobe (MTL) to investigate whether epigenetic age acceleration (EAA), a robust measure of biological aging, is associated with this neurodegenerative pattern.
Methods:
Subjects: 875 (55.7% female, 75.5 +/- 7.4 years) subjects from Alzheimer's Disease Neuroimaging Initiative (ADNI) with whole blood DNA methylation (DNAm) samples or T1 MRI (646 subjects with both). Using clinical diagnoses and established PET or cerebrospinal fluid (CSF) cut-offs we defined two groups: amyloid-negative controls (N=267) and amyloid-positive MCI/AD (N=608). Subjects were further classified into chronological age groups, such that those below 65 years were labeled "early" and those older than 80 years were labeled "late".
Methylation: DNAm was assayed on Illumina EPIC arrays covering 800K+ CpG sites and beta matrices were generated using the 'SeSAMe' R package. Epigenetic age was computed by applying the Shireby cortical clock to the beta matrices using the 'dnaMethyAge' R package. EAA is defined by regressing the epigenetic clock age against chronological age and extracting the residual. Subjects were classified as biologically accelerated, neutral, or decelerated based on whether EAA was above, within, or below 1 standard deviation of the regression, respectively.
Imaging: Regional cortical thickness measures were generated using FreeSurfer 5.1 and downloaded from ADNI. We applied longComBat to remove batch effects due to scanner and variation in field strength (1.5 or 3T) and adjusted for age and sex relative to healthy controls. We then computed two composite thickness z-scores in previously defined regions of interest (ROI) reflecting age-related and AD signatures of neurodegeneration (Figure 1). We define a composite thickness score in MTL as the bilateral thickness average of entorhinal cortex and parahippocampal cortex. Finally, we defined two "mismatch" scores reflecting relative MTL to age-related and AD signature thickness, respectively.
Methylation: DNAm was assayed on Illumina EPIC arrays covering 800K+ CpG sites and beta matrices were generated using the 'SeSAMe' R package. Epigenetic age was computed by applying the Shireby cortical clock to the beta matrices using the 'dnaMethyAge' R package. EAA is defined by regressing the epigenetic clock age against chronological age and extracting the residual. Subjects were classified as biologically accelerated, neutral, or decelerated based on whether EAA was above, within, or below 1 standard deviation of the regression, respectively.
Imaging: Regional cortical thickness measures were generated using FreeSurfer 5.1 and downloaded from ADNI. We applied longComBat to remove batch effects due to scanner and variation in field strength (1.5 or 3T) and adjusted for age and sex relative to healthy controls. We then computed two composite thickness z-scores in previously defined regions of interest (ROI) reflecting age-related and AD signatures of neurodegeneration (Figure 1). We define a composite thickness score in MTL as the bilateral thickness average of entorhinal cortex and parahippocampal cortex. Finally, we defined two "mismatch" scores reflecting relative MTL to age-related and AD signature thickness, respectively.
·Figure 1
Results:
Overall MCI/AD had reduced cortical thickness in both age-related and AD signatures relative to controls. As expected, mismatch scores reflecting greater cortical neurodegeneration relative to MTL, consistent with greater atypicality, were more pronounced in younger onset relative to older onset MCI/AD. Critically, the degree of cortical to MTL mismatch was greater in decelerated cases (0.13 ageSig, 0.33 ADSig) relative to accelerated cases (1.45 ageSig, 1.07 ADSig). The mismatch difference was statistically significant when using either the ageSig (p=0.003) or ADSig (p=0.01; Figure 2).
·Figure 2
Conclusions:
Our results demonstrate epigenetic age acceleration is significantly associated with variation in neurodegenerative patterns beyond that explained by chronological age: specifically cortical relative to MTL atrophy, a metric of atypicality. Notably, there were significant differences between decelerated and accelerated cases for both the aging and AD signatures, with decelerated cases demonstrating relatively greater cortical involvement akin to effects of younger age. This study motivates future investigation to evaluate the role of biological age specifically in heterogeneity of clinical and pathological outcomes and the enrichment of specific epigenetic markers involved in atypical disease mechanisms.
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s) 1
Genetics:
Genetics Other 2
Lifespan Development:
Aging
Modeling and Analysis Methods:
Multivariate Approaches
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
Cortical Anatomy and Brain Mapping
Keywords:
Aging
Cortex
Degenerative Disease
Informatics
Modeling
MRI
Neurological
STRUCTURAL MRI
Sub-Cortical
Other - epigenetics
1|2Indicates the priority used for review
Provide references using author date format
Beer JC (2020), Alzheimer’s Disease Neuroimaging Initiative. Longitudinal ComBat: A method for harmonizing longitudinal multi-scanner imaging data. Neuroimage
Dickerson BC (2017), Approach to atypical Alzheimer's disease and case studies of the major subtypes. CNS Spectr, 22(6):439-449.
Fischl B (2012), FreeSurfer. Neuroimage, 62(2):774-81
Petersen RC (2010), Alzheimer's Disease Neuroimaging Initiative (ADNI): clinical characterization. Neurology, 74(3):201-9.
Shaw LM (2018), Derivation of cutoffs for the Elecsys® amyloid β (1-42) assay in Alzheimer's disease. Alzheimers & Dementia, 10:698-705.
Shireby GL (2020), Recalibrating the epigenetic clock: implications for assessing biological age in the human cortex. Brain, 143(12):3763-3775.
Wang HF (2016), Application of the IWG-2 Diagnostic Criteria for Alzheimer's Disease to the ADNI. J Alzheimers Dis, 51(1):227-36.
Wang Y (2023), Insights into ageing rates comparison across tissues from recalibrating cerebellum DNA methylation clock. Geroscience
Zhou W (2018), SeSAMe: reducing artifactual detection of DNA methylation by Infinium BeadChips in genomic deletions. Nucleic Acids Res, 46(20):e123