Epigenetic Clocks: Sequencing the Epigenetic DNA Modifications Involved in Biological Ageing

Abstract

Mammalian ageing encompasses a broad set of phenotypic changes in cells and tissues, coinciding with their functional decline. This process is driven by factors including epigenetic alterations, which can regulate changes in gene expression without altering the DNA sequence itself. This includes changes in the epigenetic modification of the cytosine base in the form of 5-methylcytosine (5mC). Age-related changes in the distribution of this base modification are widespread, reflecting both the accumulation of stochastic molecular insults, as well as consistent changes at specific loci in age-relevant biological pathways. Consistent sites of age related methylation change are of special interest to ageing research, with efforts to use such changes to inform predictive models of ageing, known as “epigenetic clocks”, which can predict the age or produce a biological measure of ageing in a sample tissue.

Most of the research attention in this field has focused on age-related changes in 5mC; however, its less common oxidation derivative, 5-hydroxymethylcytosine (5hmC), also undergoes distinct and functional changes during ageing. Efforts to explore these changes have been hindered by the difficulty sequencing this modification using current methods.

In this research, long read nanopore sequencing is used to characterise age-related changes in 5mC and 5hmC. The robustness of this method for sequencing these modifications is first validated through comparison to orthologous epigenetic sequencing techniques. Nanopore sequencing is then used to collect DNA sequence and epigenetic modification data for a murine model of ageing based on the cerebellum, unearthing widespread age-related change in 5hmC. Using paired transcriptomic data, gene body changes in 5hmC are found to significantly associate with differentially expressed genes involved in known hallmarks of ageing. Finally, epigenetic data are used to train machine learning models for an age classification problem, demonstrating first-of-its-kind use of base-resolution 5hmC data as predictive of age.

Date of Award	10 Dec 2025
Original language	English
Awarding Institution	University of Bath
Supervisor	Adele Murrell (Supervisor), Sandipan Roy (Supervisor), Stefan Bagby (Supervisor) & Adrien Leger (Supervisor)