Purpose. Systemic lupus erythematosus (SLE) is a systemic and multifactorial autoimmune disease, and its diverse clinical manifestations affect molecular diagnosis and drug benefits. Our study was aimed at defining the SLE subtypes based on blood transcriptome data, analyzing functional patterns, and elucidating drug benefits. Methods. Three data sets were used in this paper that were collected from the Gene Expression Omnibus (GEO) database, which contained two published data sets of pediatric and adult SLE patients (GSE65391, GSE49454) and public longitudinal data (GSE72754) from a cohort of SLE patients treated with IFN-α Kinoid (IFN-K). Based on disease activity scores and gene expression data, we defined a global SLE signature and merged three clustering algorithms to develop a single-sample subtype classifier (SSC). Systematic analysis of coexpression networks based on modules revealed the molecular mechanism for each subtype. Results. We identified 92 genes as a signature of the SLE subtypes and three intrinsic subsets (“IFN-high,” “NE-high,” and “mixed”), which varied in disease severity. We speculated that IFN-high might be due to the overproduction of interferons (IFNs) caused by viral infection, leading to the formation of autoantibodies. NE-high might primarily result from bacterial and fungal infections that stimulated neutrophils (NE) to produce neutrophil extracellular traps (NETs) and induced individual autoimmune responses. The mixed type contained both of these molecular mechanisms and showed an intrinsic connection. Conclusions. Our research results indicated that identifying the molecular mechanism associated with different SLE subtypes would benefit the molecular diagnosis and stratified therapy. Moreover, repositioning of IFN-K based on subtypes also revealed an improved therapeutic effect, providing a new direction for disease treatment and drug development.

中文翻译：

---

血液基因组学鉴定出系统性红斑狼疮的三种亚型：“IFN-高”、“NE-高”和“混合”

目的。系统性红斑狼疮（SLE）是一种全身性、多因素的自身免疫性疾病，其多样化的临床表现影响分子诊断和药物疗效。我们的研究旨在根据血液转录组数据定义 SLE 亚型、分析功能模式并阐明药物益处。方法。本文使用了从基因表达综合 (GEO) 数据库收集的三个数据集，其中包含儿童和成人 SLE 患者的两个已发表的数据集 (GSE65391、GSE49454) 以及来自 SLE 队列的公共纵向数据 (GSE72754)接受 IFN- α Kinoid (IFN-K)治疗的患者。基于疾病活动评分和基因表达数据，我们定义了全局 SLE 特征，并合并了三种聚类算法来开发单样本亚型分类器 (SSC)。基于模块的共表达网络的系统分析揭示了每个亚型的分子机制。结果。我们确定了 92 个基因作为 SLE 亚型和三个内在亚型（“IFN 高”、“NE 高”和“混合”）的特征，这些亚型的疾病严重程度各不相同。我们推测，IFN高可能是由于病毒感染导致干扰素（IFN）产生过量，导致自身抗体的形成。NE-high 可能主要是由于细菌和真菌感染刺激中性粒细胞 (NE) 产生中性粒细胞胞外陷阱 (NET) 并诱导个体自身免疫反应。混合型包含这两种分子机制并显示出内在的联系。结论。我们的研究结果表明，识别不同SLE亚型相关的分子机制将有利于分子诊断和分层治疗。此外，基于亚型的IFN-K重新定位也显示出改善的治疗效果，为疾病治疗和药物开发提供了新的方向。