OP0212 EXAMINING CLINICAL HETEROGENEITY IN PRIMARY SJOGREN’S SYNDROME: PROTEOMIC AND NETWORK ANALYSIS

Background A symptom-based stratification tool using data from three European cohorts identified four clinical subtypes of pSS: Low symptom burden (LSB), high (HSB), dryness dominant with fatigue (DDF), and pain (PDF).[1] These have distinct molecular profiles, manifestations, responses to immunomodulatory therapies. Detailed knowledge the biological pathways underlying is vital understanding heterogeneity pSS has relevance for trial design therapeutic development. Objectives Clinical manifestations an immune-mediated inflammatory disease are likely underpinned by networks dysregulated pathways. Proteomic network analysis was performed identify that differentially expressed between subtypes. Methods We profiled serum proteins O-link technology 180 subjects (45 each subtype). used 5 proteomics panels which included 454 proteins. As we anticipate these may act directly or indirectly through interactions other proteins, a approach. Using LSB group as reference Network reconstruction ARACNE algorithm calculates mutual information variables multidimensional dataset construct protein network.[2] Differential expression estimates were overlaid on reveal key regions phenotypes. Statistical in R v4.1.3 JMP Pro v15 visualisation cytoscape v3.8.2. Results The DDF characterised clinically glandular dysfunction. proteome demonstrated higher chemokines orchestrate lymphocyte migration ectopic lymphoid structure formation salivary glands cytokines promoting B-cell hyperactivity autoantibody production. In contrast DDF, HSB PDF groups report extra-glandular such pain, fatigue, mood disturbances. proteomes show increased linked response, oxidative stress response cellular metabolism (IL-1α, IL-6, APE/Ref-1, FOXO1, TIGAR BACH1). revealed clusters upregulated downregulated different subtypes, providing added support concept contribute Figure 1. Selected showing differences associated A), B) C) subgroups. Derived data, nodes represent individual edges nodes. Colour scaling shows estimated difference comparator group, red indicating blue lower expression. Estimates derived generalised linear model post adjustment baseline age, sex batch Conclusion Protein subgroups cytokines, master transcription factors redox imbalance altered energy metabolism. Furthermore, seem be interrelated our analysis. Our provide clues contributing systemic potential targets References [1]Tarn JR et al . Symptom-based patients primary Sjogren’s syndrome: multi-dimensional characterisation international observational reanalyses randomised trials. Lancet Rheumatol. 2019 Oct 1;1(2):e85–94. [2]Margolin AA ARACNE: An gene regulatory mammalian context. BMC Bioinformatics. 2006 Mar 20;7(Sp1). Acknowledgements behalf UKPSSR. Work funded FOREUM, EU MRC. Disclosure Interests None Declared.