The epigenomics of human ageing

Abstract

The global population is ageing and age is a primary risk factor for many life threatening diseases including cancer, cardiovascular and neurodegenerative conditions. This also encompasses conditions such as osteoporosis which increase the risk of death indirectly through elevated risk of the fracture of major bones and subsequent complications. The aim of this thesis was to extend the understanding of the epigenetic processes involved in ageing and ageing-related disease. Three specific areas were investigated. Firstly, the relationship between early life epigenetic state and long term bone health was analysed. This investigation took the form of Epigenome-wide association studies (EWAS) conducted on umbilical cord blood DNA methylation. This identified candidate CpGs whose, DNA methylation state is associated with bone mineral content at 6 years of age (p<2.52×10−8, n = 402) and periosteal circumference at 6 years of age (p<4.24×10−8, n = 141) respectively. Secondly, the changes in human tRNA gene DNA methylation with age were interogated. tRNAs permit the look-up of amino acids matching a given codon and as such are an essential core component of the translation of mRNA into protein by the ribosome. Human tRNA genes were found to be enriched for age-related DNA hypermethylation and three specific tRNA loci show genome-wide significant (p<4.34×10−9) hypermethylation with age. Two of which, tRNA-iMet-CAT-1-4 and tRNA-Ser-AGA-2-6, were validated using the 450k Illumina DNA methylation array and replicated in an independent cohort, using targeted bisulfite sequencing. Thirdly, a DNA methylation based age predictor based on the Alu family of SINE repeat elements was constructed. These repeats comprise a region of the genome not previously broadly accessible to DNA methylation assays used in the construction of age predictors. Age predictors using the DNA methylation state of Alu repeat elements were able to predict human age with an R of 0.65 and a median absolute error of 8.1 years. This predicition was possible using MeDIP-seq training data from 774 individuals and validated on an unrelated set of 664 individuals. An attempt was made to use ageing-related changes in Alu DNA methylation as a potential measure of ‘biological’ ageing. A Genome-wide Association study (GWAS) was performed for age accleration, the difference between predicted and chronological age. However, this age acceleration calculation was observed to be still strongly driven by actual age. This work brings together epigenomic changes related to ageing from the beginning of the lifecourse through to later life, uniquely examining areas of the DNA methylome not previously studied in depth.

More information

Identifiers

ORCID for Richard, James Acton: orcid.org/0000-0002-2574-9611

ORCID for Cyrus Cooper: orcid.org/0000-0003-3510-0709

ORCID for Karen Lillycrop: orcid.org/0000-0001-7350-5489

ORCID for Christopher G Bell: orcid.org/0000-0003-4601-1242

Catalogue record

Export record