Section of Rheumatology and Science for Life Laboratory, Department of Medical Sciences, Uppsala University, Uppsala, Sweden.
Molecular Medicine and Science for Life Laboratory, Department of Medical Sciences, Uppsala University, Uppsala, Sweden.
Front Immunol. 2019 Jul 30;10:1686. doi: 10.3389/fimmu.2019.01686. eCollection 2019.
To perform a cross-comparative analysis of DNA methylation in patients with systemic lupus erythematosus (SLE), patients with primary Sjögren's syndrome (pSS), and healthy controls addressing the question of epigenetic sharing and aiming to detect disease-specific alterations. DNA extracted from peripheral blood from 347 cases with SLE, 100 cases with pSS, and 400 healthy controls were analyzed on the Human Methylation 450k array, targeting 485,000 CpG sites across the genome. A linear regression model including age, sex, and blood cell type distribution as covariates was fitted, and association -values were Bonferroni corrected. A random forest machine learning classifier was designed for prediction of disease status based on DNA methylation data. We established a combined set of 4,945 shared differentially methylated CpG sites (DMCs) in SLE and pSS compared to controls. In pSS, hypomethylation at type I interferon induced genes was mainly driven by patients who were positive for Ro/SSA and/or La/SSB autoantibodies. Analysis of differential methylation between SLE and pSS identified 2,244 DMCs with a majority of sites showing decreased methylation in SLE compared to pSS. The random forest classifier demonstrated good performance in discerning between disease status with an area under the curve (AUC) between 0.83 and 0.96. The majority of differential DNA methylation is shared between SLE and pSS, however, important quantitative differences exist. Our data highlight neutrophil dysregulation as a shared mechanism, emphasizing the role of neutrophils in the pathogenesis of systemic autoimmune diseases. The current study provides evidence for genes and molecular pathways driving common and disease-specific pathogenic mechanisms.
为了对系统性红斑狼疮 (SLE)、原发性干燥综合征 (pSS) 患者的 DNA 甲基化进行交叉比较分析,解决表观遗传共享问题,并旨在检测疾病特异性改变,我们对来自 347 例 SLE 患者、100 例 pSS 患者和 400 例健康对照者的外周血 DNA 进行了分析,这些 DNA 采用人类甲基化 450k 阵列进行分析,靶向基因组上 485000 个 CpG 位点。建立了一个包含年龄、性别和血细胞类型分布作为协变量的线性回归模型,并对关联值进行了 Bonferroni 校正。基于 DNA 甲基化数据,我们设计了一个随机森林机器学习分类器来预测疾病状态。我们在 SLE 和 pSS 患者与对照组之间建立了一个由 4945 个共享差异甲基化 CpG 位点 (DMCs) 组成的综合集。在 pSS 中,I 型干扰素诱导基因的低甲基化主要是由 Ro/SSA 和/或 La/SSB 自身抗体阳性的患者驱动的。SLE 和 pSS 之间的差异甲基化分析确定了 2244 个 DMCs,这些 DMCs中的大多数位点在 SLE 中与 pSS 相比显示出甲基化减少。随机森林分类器在辨别疾病状态方面表现出良好的性能,曲线下面积 (AUC) 在 0.83 到 0.96 之间。SLE 和 pSS 之间的大多数差异 DNA 甲基化是共享的,但存在重要的定量差异。我们的数据突出了中性粒细胞失调作为一种共享机制的重要性,强调了中性粒细胞在系统性自身免疫性疾病发病机制中的作用。本研究为驱动共同和疾病特异性致病机制的基因和分子途径提供了证据。