Abstract
DNA methylation age (DNAm age) estimation is a powerful biomarker of human ageing. To date, epigenetic clocks have not been evaluated in age-related macular degeneration (AMD). Here, we perform genome-wide DNA methylation analyses in blood of AMD patients with a documented smoking history (14 AMD, 16 Normal), identifying loci of differential methylation (DML) with a relaxed p-value criterion (p ≤ 10−4). We conduct DNAm age analyses using the Horvath-multi tissue, Hannum and Skin & Blood epigenetic clocks in both blood and retinal pigment epithelium (RPE). We perform Ingenuity Pathway Analysis Causal Network Analysis (IPA CNA) on the topmost significantly differentially methylated CpG probes in blood and RPE. Results show poor performance of epigenetic clocks in RPE. Epigenetic age acceleration (EAA) was not observed in AMD. However, we observe positive EAA in blood of smokers, and in smokers with AMD. DML analysis revealed hypomethylation at cg04953735 within RPTOR (p = 6.51 × 10−5; Δβ = −11.95%). IPA CNA in the RPE also identified RPTOR as the putative master regulator, predicted to be inhibited in AMD. In conclusion, this is the first study evaluating an association of epigenetic ageing in AMD. We posit a role for RPTOR as a common master regulator of methylation changes in the RPE in AMD.
Highlights
Ageing is a complex physiological process characterised by progressive loss of tissue functionality and an increased risk of death [1]
We might expect any non-zero epigenetic age acceleration” (EAA) in the retinal pigment epithelium (RPE) to be associated with weak correlation, our findings are key in illustrating the potential biological and technical deviations present when assaying RPE samples as opposed to whole of the epigenetic clocks in RPE [33]
The consistent poor correlation of predicted DNA methylation (DNAm) age with chronological age observed in the RPE markedly improved when analysing whole blood-derived genomic DNA (gDNA) data, explained by the datasets used to train each respective epigenetic clock [3,4,8]
Summary
Ageing is a complex physiological process characterised by progressive loss of tissue functionality and an increased risk of death [1]. Horvath’s multi-tissue [3], Hannum’s [4], and the Skin & Blood epigenetic clocks [8] were constructed using penalised multivariate elastic net regression models that select CpG probes for each respective clock by minimising the residual error of predicting age while removing redundant coefficients from the model. Significant negative EAA has been identified in the cerebellum of extremely elderly humans (>100 years old) and in certain cancers [12,13,14,15], acting as a predictor of poor prognostic outcomes in the latter [15] Together these studies indicate that both positive and negative accelerations of epigenetic age represent cellular adaptive responses to genomic/epigenomic instability, contributors to age-related disease [7]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
