Fujio et al. Inflammation and Regeneration (2020) 40:11 Inflammation and Regeneration https://doi.org/10.1186/s41232-020-00123-w REVIEW Open Access Review: transcriptome and trans-omics analysis of systemic lupus erythematosus Keishi Fujio*, Yusuke Takeshima, Masahiro Nakano and Yukiko Iwasaki Abstract Systemic lupus erythematosus (SLE), which was recognized as a defined clinical entity more than 100 years ago, is an archetype for systemic autoimmune diseases. The 10-year survival of SLE patients has shown dramatic improvement during the last half-century. However, SLE patients receiving long-term prednisone therapy are at high risk of morbidity due to organ damage. Identification of key immune pathways is mandatory to develop a suitable therapy and to stratify patients based on their responses to therapy. Recently developed transcriptome and omic analyses have revealed a number of immune pathways associated with systemic autoimmunity. In addition to type I interferon, plasmablast and neutrophil signatures demonstrate associations with the SLE phenotype. Systematic investigations of these findings enable us to understand and stratify SLE according to the clinical and immunological features. Keywords: Systemic lupus erythematosus, Transcriptome, Omics, eQTL Background previous assessment; SLEDAI Physician Global Assess- Systemic lupus erythematosus (SLE) is an archetype of ment 1 ≦ 4; a current prednisolone (or equivalent) dose autoimmune disease. It is characterized by its predomin- ≦ 7.5 mg daily; and well-tolerated standard maintenance ance in woman of childbearing age [1, 2]. Clinical het- doses of immunosuppressive drugs. Patients who spent erogeneity, together with the potential severity of these greater than 50% of their observed time in LLDAS had manifestations, make the management of SLE a distinct significantly reduced accrual of organ damage compared challenge for rheumatologists. Despite recent progress in with patients who spent less than 50% of their time in treatment protocols, several studies have pointed out LLDAS and were significantly less likely to have in- that SLE patients still have an overall 2- to 3-fold in- creased organ damage. However, some patients devel- creased risk of death [3, 4]. In this context, an initiative oped organ damage in spite of the maintenance of to evaluate possible therapeutic targets and develop LLDAS. While some patients remain serologically active treat-to-target guidance would be highly appropriate in clinically quiescent (SACQ) indefinitely or become sero- the management of SLE patients. logically quiescent clinically quiescent (SQCQ), others’ The Asia Pacific Lupus Collaboration (APLC) pro- SACQ periods are terminated by disease flares for which posed the existence of a Lupus Low Disease Activity reliable predictors have yet to be identified [6]. There- State (LLDAS). This was subsequently defined oper- fore, novel biomarkers are necessary to improve prog- ationally via a nominal consensus approach [5]. LLDAS nostic stratification, monitoring of treatment responses, was defined as follows: SLE Disease Activity Index (SLE- and detection of early renal flares. Transcriptome and DAI) 2K ≦ 4, with no activity in major organ systems; omics analyses, which evaluate the genome, transcrip- no new lupus disease activity compared with the tome, and epigenome, are powerful approaches to the identification of clinically useful biomarkers. * Correspondence: [email protected] Department of Allergy and Rheumatology, Graduate School of Medicine, The University of Tokyo, 113-8655 7-3-1 Hongo, Bunkyo-ku, Tokyo, Japan © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Fujio et al. Inflammation and Regeneration (2020) 40:11 Page 2 of 6 Key molecular signatures in SLE Banchereau et al. reported a longitudinal modular SLE patients demonstrate transcriptional signatures re- study of 158 SLE patients in the Immunological Genome lated to type I interferon (IFN) and granulocytes [7, 8]. Project Consortium [13]. They found that positive corre- Although a number of studies have focused on the link- lates of disease activity were enriched for the IFN re- age between IFN-induced transcripts or proteins and sponse as well as plasmablast, cell cycle, neutrophil, SLE disease activity, most of them suffer from sample- histone, and B cell modules. Notably, the plasmablast size limitations and essential clinical and therapeutic signature was the most reproducible, and plasmablast re- heterogeneity in SLE. Chaussabel et al. applied modular sponses showed increases in African-American patients analysis in PBMC microarray data from 239 individuals who presented with more severe disease than did Cauca- [9]. Differentially expressed gene (DEG) analysis toler- sians. In addition, the neutrophil, myeloid lineage, and in- ates noise due to the large number of comparisons per- flammation modules were enriched in patients with lupus formed (usually > 10,000). In order to reduce noise, they nephritis. These observations support a model of sequential focused the analysis on small sets of coordinately disease progression, whereby an initial increase in IFN re- expressed transcripts that constitute coherent and bio- sponse and differentiation of B cells into short-lived plas- logically meaningful transcriptional units. Among 28 mablasts, which may occur before clinical onset, is followed transcriptional modules linked with pathways or cell by lupus nephritis and full-blown systemic inflammation types involved in immune processes, two modules, IFN- enhanced by myeloid cells, including neutrophils. inducible and neutrophil genes, exhibited positive corre- One limitation of transcriptome analysis of peripheral lations with SLE disease activity, SLEDAI. In contrast, blood cells or PBMC is the distorting effects of varia- gene modules corresponding to ribosomal proteins and tions in the size of each immune subset. That is, tran- surface molecule of T cells including CD5, CD6, CD7, scriptome analyses of a mixture of cell subsets are CD26, CD28, and CD96 negatively correlated with SLE- obscured by the variations in the frequencies of different DAI. Intriguingly, in one patient who showed disease cell types, as shown in comprehensive analyses in the flaring and subsequent recovery, an increase in the tran- human brain of bulk/single-cell transcriptomes, active scriptional score was detected 2 months before deterior- enhancers, genotypes, and Hi-C data sets, named Psy- ation of the patient’s clinical condition. Their following chENCODE [14]. In PsychENCODE, weighted combina- report also revealed complex IFN signatures in SLE that tions of single-cell signatures could account for most of were composed of 3 modules and involved both an IFN- the population-level expression variation, with an accur- α signature and those for IFN-β and IFN-γ [10]. acy of > 88%. Therefore, more detailed analyses of im- The possibility of clinical application of IFN gene ex- mune cell subsets are required to advance modular pression data has been demonstrated by a two-score sys- findings obtained from peripheral blood or PBMC. Mc- tem for IFN status. Factor analysis by El-Sherbiny et al. in Kinney et al. conducted a subset-specific transcriptomic 328 autoimmune disease patients reduced the 30 ISGs analysis of 58 patients with ANCA-associated vasculitis (interferon signature genes) to 2 factors, IFN-Score-A and (AAV) and 23 patients with SLE. They found that while IFN-Score-B [11]. Although IFN-Score-A and IFN-Score- CD4+ T cell co-stimulation is pronounced, genes specif- B shared several key ISGs, specific genes for IFN-Score-A ically downregulated in exhausted CD8+ T cells during were mostly typical ISGs, i.e., IFIT1, IFI27, RSAD2, IFI44L, chronic murine lymphocytic choriomeningitis virus IRF7, ISG15,andXAF1. On the other hand, specific genes (LCMV) infection were downregulated in CD8+ T cells for IFN-Score-B were LAMP3, UNC93B1, SP100, IFI16, from patients at low risk of relapse of AAV and SLE. IFIH1, PHF11,andBST2. IFN-Score-A differentiated SLE This result suggested that whereas CD8+ exhaustion is from both rheumatoid arthritis (RA) and healthy controls associated with poor outcome in viral infection, it pre- (HC), while IFN-Score-B differentiated SLE and RA from dicts favorable outcome in autoimmune diseases. HC. In SLE, both scores were associated with cutaneous Shi et al. studied monocytes from 9 SLE patients using and hematological but not musculoskeletal disease activ- RNA-seq of both coding and noncoding RNAs [15]. ity. Yusof et al. examined these scores in 118 individuals They found that SLE patients had diminished expression at risk of autoimmune connective tissue diseases [12]. At of most endogenous retroviruses and small nucleolar