Background/Purpose: Juvenile-onset systemic lupus erythematosus (jSLE) is an autoimmune disease characterised by chronic inflammation, multi-organ damage, and/or death. Despite their young age, jSLE patients have a 100-300-fold increased cardiovascular disease (CVD) risk, potentially associated with type I interferons (IFNs). We used multi-omics to explore the heterogeneity, cell specificity, and CVD risk associated with type I IFN signatures in jSLE.

Methods: RNA sequencing was used to assess differentially expressed genes (DEGs) in peripheral blood mononuclear cells (PBMCs) between jSLE patients with inactive disease (n=29, mean age/disease duration=19/6.5 years) and healthy controls (HCs, n=8, mean age=18). Data was analysed by weighted gene co-expression network analysis (WGCNA), pathway enrichment, hierarchical clustering, and receiver operating characteristic (ROC) analysis. Tetherin (established IFN-regulated protein) expression was quantified by spectral flow cytometry on PBMCs over a 30-parameter immune landscape (UMAP/FlowSOM). Proteomics (Olink) and Metabolomics (Nightingale) assessed serum proteins (validated by ELISA) and metabolites associated with CVD. Clinical data was analysed by trajectory analysis (mean 4.9 years/17 encounters).

Results: Independent of disease activity, type I IFN signalling was the most upregulated network module in jSLE patients compared to HCs (280 DEGs, p< 0.0001) from unbiased WGCNA. Despite this, jSLE patients clustered heterogeneously into high (H-IFN, 66%, ROC: p< 0.0001/AUC=1.00 vs HCs) and low (L-IFN, 34%, ROC: p=0.53/AUC=0.59 vs HCs) type I IFN signature groups, validated by established type I IFN gene scores (p< 0.0001).

Transcriptomic type I IFN scores in jSLE patients correlated positively with serum LAMP3 protein levels (Olink, p=0.0018/r=0.67). Furthermore, LAMP3 levels differentiated between H-IFN and L-IFN jSLE patients with strong accuracy (ROC: p=0.0069/AUC=0.95); this finding was validated by ELISA (ROC: p=0.027/AUC=0.89), and outperformed the biomarker accuracy of PBMC tetherin expression, supporting a more translational biomarker of H-IFN.

Importantly, no difference in clinical serology or disease activity trajectory was observed between type I IFN groups (well-controlled patients). However, transcriptomic type I IFN scores correlated positively with proteomic (galectin-9/ICAM1/VCAM1) and metabolomic (glycoprotein acetyls/ApoB:ApoA1) biomarkers, all heavily associated with CVD risk mechanisms (p< 0.01).

Finally, unbiased PBMC clustering identified relatively higher sensitivity to type I IFN (quantified by tetherin expression) in jSLE-specific sub-populations of B cells, plasmacytoid dendritic cells, CD8+ T cells, and NK cells.

Conclusion: Multi-omics allows us to better understand type I IFNs in jSLE: 1) patients have heterogeneous type I IFN signatures associated with CVD risk, independent from disease activity; 2) LAMP3 is an accurate serum biomarker that could stratify H-IFN patients in a clinical setting for tailored anti-IFN treatment to improve disease/CVD outcomes; and 3) specific immune populations may respond more favourably to anti-IFN therapy, providing mechanistic insight into treatment.

« Back to ACR Convergence 2024

ACR Meeting Abstracts - https://acrabstracts.org/abstract/multi-omic-analysis-defines-heterogeneous-and-cell-specific-type-i-interferon-signalling-in-juvenile-onset-sle-patients-associated-with-biomarkers-of-cardiovascular-risk/