Abstractype 2 diabetes (T2D) is a multifactorial disease with a range of genetic and environmental risk factors. Increasing evidence shows that DNA methylation is likely to play a role in the aetiology of T2D. In this thesis, I investigated the association of DNA methylation with T2D focusing on middle-age adults of European background, and ascertained causality of this association by using Mendelian randomization.
I have used genome-wide DNA methylation data to identify markers associated with prevalent T2D in four European cohorts, combining results across studies via meta-analysis to increase power. This approach allowed the confirmation of well-known markers mapping to the genes TXNIP and ABCG1, and the identification of novel markers which are good candidates for validation. I also investigated DNA methylation in relation to different glycaemic traits using disease-free participants. Results suggested that variation in methylation preceding disease occurrence can indicate early stage of disease. Various in silico explorations allowed me to explore the potential mediating role of DNA methylation in the genetics of T2D and gene expression, biological pathways enriched for differentially-methylated sites, and the relevance of blood as a source of methylation markers for T2D. Genetic variants associated with T2D were used to determine causality of signals detected observationally using a bidirectional Mendelian randomization. Results revealed that methylation can be both, a consequence and a cause of T2D.
The findings of this thesis suggest that DNA methylation might be an important factor in the aetiology of T2D. However, further studies are needed to confirm novel signals, and to assess their generalizability in other populations. Results of the causal analysis should be reinforced by including larger datasets, especially when DNA methylation is regarded as the outcome.
|Date of Award||7 May 2019|
|Supervisor||Hannah R Elliott (Supervisor), Caroline L Relton (Supervisor) & Gemma C Sharp (Supervisor)|