Annie Law1, Chee Jian Pua2, Dianyang Guo3, Chin Teck Ng4, Julian Thumboo1, Andrea Hsiu Ling Low5 and Xiubo Fan1, 1Singapore General Hospital; Duke-NUS Medical School, Singapore, Singapore, 2National Heart Centre, Singapore, Singapore, 3Singapore General Hospital, Singapore, Singapore, 4Singapore General Hospital; Duke-NUS Medical School, Singapore, Malaysia, 5Department of Rheumatology and Immunology, Singapore General Hospital, Singapore, Singapore
Background/Purpose: IFN response, plasmablasts and neutrophils are three hallmarks of systemic lupus erythematosus (SLE). Studies showed that the netting neutrophils stimulated plasmacytoid dendritic cells to secret type I interferon directly and further promoted the expansion of plasmablasts indirectly, suggesting the upstream position of neutrophils in the pathogenesis of SLE. The blood transcriptomic profiling of human SLE, with a focus on lymphoid (T cells, B cells and NK cells) and partial myeloid (monocytes and dendritic cells) lineages, has been uncovered. However, due to the sample processing challenge – a rapid process of fresh blood sample for optimal cell quality, the transcriptomic profiling of neutrophils remains incomplete.
Methods: To determine cell-type specific signatures, we planned to decipher neutrophil transcriptional changes in human SLE using single cell RNA sequencing (scRNA-seq). Briefly, freshly isolated white blood cells from healthy controls (n = 3) and SLE patients (SLEDAI ≤ 4, n = 3) were collected for scRNA-sq sample preparation and sequencing with BD Rhapsody microwell-based single-cell partitioning technology. For data analysis, Scrublet was used for multiplet removal, BBKNN for batch correction, Python-based Scanpy pipeline for data pre-processing, visualization, clustering and differential expression testing and UMAP for data plotting.
Results: Neutrophils were clustered to 5 subpopulations (pre-neutrophils (PreNeu), early neutrophils (EarlyNeu), middle neutrophils (MidNeu), late neutrophils (LateNeu) and late IFN-expressing neutrophils (LateIFNNeu), according to Gustaf et al. In SLE patients (cases), the frequency of EarlyNeu in myeloid cells was markedly increased (48.7% vs 26.3%), whereas the frequencies of MidNeu (18.8% vs 25.4%), LateNeu (7.3% vs 13.3%), LateIFNNeu (10.4% vs 14.8%) were reduced compared to healthy controls. Gene set enrichment analyses were performed using hallmark gene sets to identify the possible pathophysiology of the disease. In line with our speculation, interferon (IFN-αand IFN-γ) response, inflammatory response and TNF-α/NF-κB signals were found to be enriched in cases for PreNeu, EarlyNeu and LateNeu. In addition, relative to controls, IL-6/JAK/STAT3 and apoptosis signals were enriched in cases for EarlyNeu and LateNeu; complement signal was enriched in cases for LateNeu; IL-2/STAT5 signal was enriched in cases for MidNeu and LateNeu; cell cycle (e.g., G2M checkpoints, E2F targets and mitotic spindles) and DNA repair signals were under-represented in cases for PreNeu. Conversely, IFN-γ signal was under-represented in cases for MidNeu and LateIFNNeu.
Conclusion: In SLE patients, pronounced inflammatory signals were observed in neutrophils across different stages.
A. Law: None; C. Pua: None; D. Guo: None; C. Ng: None; J. Thumboo: None; A. Low: Boehringer-Ingelheim, 6, Janssen, 6; X. Fan: None.
