Divergent spatial microdomains drive inflammation and repair in Ulcerative and Immune Checkpoint Therapy Colitis - whole tissue scRNA-Seq

Adult inflammatory bowel disease is incompletely understood. We combine unbiased single-cell RNA sequencing (gene expression profiling, CITE-seq derived cell surface protein data, TCR and BCR sequence data) with unbiased spatial transcriptomics to interrogate changes across immune and non-immune populations in colitis and health, across tissue and blood. We compare idiopathic ulcerative colitis with hitherto under-studied immune checkpoint therapy induced colitis, utilizing non-inflamed disease states as additional controls. We identify patterns of inflammation and response unique and common to both diseases, and infer changes in cell trafficking with potential therapeutic implications. We go on to localize disease-specific changes in tissue using spatial transcriptomics. We leverage this data to interrogate cellular interactions in an unbiased manner, allowing us to describe novel microdomains of inflammation and repair. Cells from tissue (colon) and peripheral blood (PBMC), in patients with active ulcerative colitis (UC_I), paired non-inflamed areas at the same endoscopy (UC_NI), active immune checkpoint inhibitor-induced colitis (CC_I), patients given checkpoint inhibitors without colitis (CC_NI) all compared with healthy controls (HC). Patients with CC_I and CC_NI were either treated with anti-PD-1 immunotherapy (Monotherapy) or a combination of simultaneous anti-PD-1 and anti-CTLA4 immunotherapy (Dual therapy). Cell pools were either comprised of peripheral blood cells (PBMC) or enriched for Epithelium (Epi) or the remaining Stromal/Immune fraction (Stromal) - see methods for full details. Samples were subjected to 5' scRNA sequencing, yielding Gene expression (GEX), CITE-seq (ADT), T-cell receptor (TCR) and B-cell receptor (BCR) information.