In-depth analysis of differential DNA methylation for systemic lupus erythematosus

Abstract

Systemic lupus erythematosus (SLE), a prototypic autoimmune disease, has been extensively studied by genome-wide genetic and epigenetic approaches. However, genetic risk loci revealed by genetic association studies so far only explained less than 30% of disease heritability. Epigenetics, including DNA methylation, histone modification, microRNA regulation, and transcription factors (TF) binding, plays a vital role in maintaining genome stability, regulating gene expression and translation, and determining cell fate and tissue development. DNA methylation was the most studied epigenetic mechanism in autoimmune diseases such as SLE.

In this study, integrative analysis of differentially methylated regions (DMRs) for SLE across three immune cell types (T cells, B cells, and monocytes) was performed by using publicly available DNA methylation datasets. In addition to Type I interferon signalling which is known, hyper-methylated genes were found enriched in T cell receptor (TCR) signalling, as demonstrated by gene clusters from optimal subnetwork analysis based on protein-protein interaction (PPI) of the STRING database. By integrating in-house genetic data from genome-wide association studies (GWAS) on SLE, significant co-localization of DMRs with SLE susceptibility loci was observed, exemplified by well-studied risk genes such as ETS1. Furthermore, TFs, upon treatment by different interferon stimuli, presented distinct binding enrichment among DMRs, where significantly more differentially methylated genes contained TF binding sites only after longer interferon treatment. Last, a machine learning algorithm, LASSO, was used to select a panel of CpGs that seems to reveal both known and novel disease mechanisms, such as calcium signalling, immunoproteasome processing, and ubiquitinase functioning. They accurately distinguished SLE patients from healthy individuals and were significantly correlated with disease activities indicated by SLEDAI scores.

In summary, this study presented an in-depth investigation of aberrant epigenetic alterations in SLE by integrating publicly available omics datasets and may have useful implications for epigenetics research of other diseases, especially other autoimmune diseases.