POS0847 IDENTIFICATION OF MOLECULAR PHENOTYPES IN SYSTEMIC SCLEROSIS BY INTEGRATIVE SYSTEMS ANALYSIS

M. J. Chang,S. X. Zhang,Q. Wang,J. Qiao,R. Zhao,S. Song,Y. Zhang,Q. Yu,P. F. He,X. Li

DOI: https://doi.org/10.1136/annrheumdis-2021-eular.1846

IF: 27.973

2021-01-01

Annals of the Rheumatic Diseases

Abstract:Background:Systemic sclerosis (scleroderma, SSc) is a systemic autoimmune disease characterized by inflammation, fibrosis and vasculopathy and associated with high mortality and high morbidity1. Stratification based on whole-genome gene expression data could provide a new basis for clinical diagnosis from a micro perspective2.Objectives:The objective of this study is to stratify patients with SSc, combine with clinical skin scores and clinical features, and provide a preliminary assessment and novel insights for assessing disease severity, and treatment design.Methods:The original data mRNA expression profiles of GSE95065 (including 18 SSc patients and 4 healthy controls) and GSE130955 (including 58 SSc patients and 33 healthy controls) were downloaded from the public Gene Expression Omnibus (GEO) database. After batch correction, background adjustment, and other pre-processing, a large gene matrix was obtained to identify the differently expressed genes (DEGs) of SSc compared with healthy controls. Then the gene expression matrix decomposition was used to identify SSc subtypes by NMF algorithm. The cluster-based signature genes were applied to pathway enrichment analysis by Metascape3. Immune infiltrating cells and clinical skin scores were evaluated in all SSc subtypes.Results:Total 325 DEGs were imputed to NMF unsupervised machine learning algorithm. Patients were divided into 2 subtypes (Figure 1A), one of which (sub1) was mostly enriched in the defense response to bacterium and cellular response to lipopolysaccharide pathway and another subtype (sub2) was enriched in the PPAR signaling and alcohol metabolic process pathway (Figure 1B-C). According to immune infiltration, sub1 had higher level of immune cells such as B cells, CD4+T cells, DC cells, Th2 cells and Tregs compared with sub2 (P < 0.01). Sub2 had more skin-related cells, including Epithelial cells, Fibroblasts and Sebocytes (P < 0.05). Interestingly, combined with clinical information, sub1 showed a severe clinical skin score over those of Sub2 patients (P < 0.05)(Figure 1D-E).Conclusion:Our findings indicated that SSc patients could be stratified into 2 subtypes which had different molecular profiles of disease progression and clinical disease activities. This result could serve as a template for future studies to design stratified approaches for SSc patients.References:[1]Xu X, Ramanujam M, Visvanathan S, et al. Transcriptional insights into pathogenesis of cutaneous systemic sclerosis using pathway driven meta-analysis assisted by machine learning methods. PLoS One 2020;15(11):e0242863. doi: 10.1371/journal.pone.0242863 [published Online First: 2020/12/01].[2]Xu C, Meng LB, Duan YC, et al. Screening and identification of biomarkers for systemic sclerosis via microarray technology. Int J Mol Med 2019;44(5):1753-70. doi: 10.3892/ijmm.2019.4332 [published Online First: 2019/09/24].[3]Zhou Y, Zhou B, Pache L, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun 2019;10(1):1523. doi: 10.1038/s41467-019-09234-6 [published Online First: 2019/04/05].Acknowledgements:This project was supported by National Science Foundation of China (82001740), Open Fund from the Key Laboratory of Cellular Physiology (Shanxi Medical University) (KLCP2019) and Innovation Plan for Postgraduate Education in Shanxi Province (2020BY078).Disclosure of Interests:None declared