OA10 Using single-cell technology to understand the cellular landscape of colonic disease in systemic sclerosis

Abstract Background/Aims Gastrointestinal (GI) disease is one of the most frequent and disabling manifestations of systemic sclerosis (SSc). Unlike the skin, GI disease does not respond to immunosuppression and generally worsens over time. Our previous analysis of gastric tissue biopsies from SSc patients and healthy controls (HC) showed significant differences in fibroblast subsets, expansion of the CD8+ T cell compartment, and significant reduction in plasma cells. Here we utilise single-cell technologies to elucidate pathogenesis in the colon and compare it with gastric tissue. Methods Colonic and gastric biopsies were obtained from four SSc patients and four HC. Single-cell RNA sequencing (scRNAseq) and spatial transcriptomics were performed on whole biopsies using 10X Genomics Chromium workflow and 10X Xenium, with a 5K gene panel. Data were analysed using scanpy, scvi-tools, and squidpy. Results After filtering, 27,274 cells were analysed using scRNAseq. Within the immune cell compartment of the colon tissue, there was an increase in IgA plasma cells and memory B cells in SSc compared to HC. The most abundant fibroblast population in the colon was a proinflammatory fibroblast (PDGFRA+). This was not significantly different in abundance between SSc and HC in this cluster, but gene-set enrichment revealed downregulation of extracellular matrix organisation, and collagen fibril organisation biological process pathways. Comparison across disease and tissue types highlighted that even in health, there were fewer fibroblasts in colon compared to gastric tissue. We found that IgA plasma cells were more abundant in the colon of SSc patients compared with HC, but this was reversed in the stomach. In turn clonally expanded CD8+ T cells were more frequent in the stomach than the colon in SSc. These results suggest that there are location-specific differences in pathology across GI tissues. Spatial transcriptomic analysis of 16525 cells identified significant interaction between macrophages and fibroblasts in niches present in SSc but not HC. There was a profound increase in PDGFRA+ fibroblasts and proinflammatory macrophages (CD206+/MERTK+) within the lamina propria surrounding the crypts of SSc tissue, whereas both these fibroblasts and macrophages were barely present in corresponding sites in HC tissue. There was also a significant increased expression of MMP12, MFGE8 and NR4A1, genes involved in fibrosis and macrophage polarisation. Conclusion Using single-cell and spatial analysis, we show significant differences in tissue composition, which differ between locations and disease state. Thus, cellular pathology is not uniform throughout the GI tract. Notably, although scRNAseq confirms similar abundance of proinflammatory fibroblasts between SSc and HC, spatial transcriptomics suggests location-specific differences in interaction between fibroblasts and macrophages that may drive GI fibrosis. Disclosure K.E.N. Clark: None. M. Attar: None. M. Friedrich: None. C. Murray: None. A. Clarke: None. C.P. Denton: None.