DNA methylation landscape of ocular tissue relative to matched peripheral blood

Alex W Hewitt,Jihoon E Joo,Maria Franchina,Vania Januar,Helena Liang,Jie Jin Wang,Alexandra Sexton-Oates,Richard Saffery,Jamie E Craig

doi:10.1038/srep46330

Abstract

Epigenetic variation is implicated in a range of non-communicable diseases, including those of the eye. However, investigating the role of epigenetic variation in central tissues, such as eye or brain, remains problematic and peripheral tissues are often used as surrogates. In this study, matched human blood and eye tissue (n = 8) were obtained post-mortem and DNA methylation profiling performed on blood, neurosensory retina, retinal pigment epithelium (RPE)/choroid and optic nerve tissue using the Illumina Infinium HumanMethylation450 platform. Unsupervised clustering and principal components analysis revealed tissue of origin as the main driver of methylation variation. Despite this, there was a strong correlation of methylation profiles between tissues with >255,000 CpG sites found to have similar methylation levels. An additional ~16,000 show similarity across ocular tissues only. A small proportion of probes showing inter-individual variation in blood co-varied with eye tissues within individuals, however much of this variation may be genetically driven. An improved understanding of the epigenetic landscape of the eye will have important implications for understanding eye disease. Despite a generally high correlation irrespective of origin, tissue type is the major driver of methylation variation, with only limited covariation between blood and any specific ocular tissue.

Highlights

Eye, but these have generally been undertaken in non-ocular tissues and as such, the relevance to the pathology of interest remains unclear[8,9,10]
Multidimensional Scaling plot visualisation of the 1,000 most variable probes highlighted the similarity of retinal pigmented epithelium (RPE)/choroid and optic nerve methylation profiles relative to retina, with blood the most dissimilar in methylation profile (Supplementary Figure 1)
Most of the variation within the dataset (55.84%) was due to tissue type, with the first four principal components (PCs) all associated with tissue of origin at p < 0.05 (Fig. 2)

Summary

Introduction

Eye, but these have generally been undertaken in non-ocular tissues and as such, the relevance to the pathology of interest remains unclear[8,9,10]. A clearer understanding of the methylation landscape of tissues relevant to specific conditions is required in order to gain insights into the relevance of measuring epigenetic profile in a proxy tissue such as blood or saliva[11]. Such knowledge gaps are common when dealing with living individuals and have only been directly assessed in limited postmortem studies. The overriding hypothesis was that there would be both tissue and individual specific methylation signatures, and that a blood DNA methylation profile could be identified for potential use as a proxy for the direct study of eye tissue, generally not possible in living humans. We determine the relative importance of ocular tissue methylation specificity at the genome-wide level, and quantify the number of sites that co-vary within an individual across neurosensory retinal, RPE/choroidal and optic nerve tissue

Objectives

Results

Conclusion