OP06 Multiomic interrogation of the epithelium in active Ulcerative Colitis reveals dysregulated myeloid cells associated with non-response to anti-Tumour Necrosis Factor therapy. (30th January 2023)

• Record Type: Journal Article
• Authors: Jha, D; Al-Taie, Z; Krek, A; Eshghi, S; Tankelevich, M; Meringer, H; Cossarini, F; Canales-Herrerias, P; Livanos, A E; Polydorides, A D; Martin, J; Ko, H M; McBride, J; Hackney, J; Colombel, J F; Argmann, C; Suarez-Farinas, M; Petralia, F; Mehandru, S
• Is Part Of: Journal of Crohn's and colitis. Volume 17(2023)Supplement 1
• Page Start: i11
• Page End: i12
• Publication Date: 2023-01-30
• Subjects: Inflammatory bowel diseases -- Periodicals; 616.344005
• DOI: 10.1093/ecco-jcc/jjac190.0006
• Languages: English
• ISSNs: 1873-9946

Abstract

Abstract: Background: While the majority of Ulcerative Colitis (UC)-related mucosal studies have focused on whole intestinal tissues or the lamina propria (LP), epithelial compartment (EC)-specific studies are largely lacking. Here, we have defined EC-associated molecular and cellular dynamics during inflammation and studied their response to anti-tumor necrosis factor inhibitor (TNFi) therapy. Methods: EC-focused analyses that included total RNA sequencing (RNAseq), single-cell (sc) RNAseq, spatial transcriptomics (ST), microscopy and flow cytometry (FC) were performed in a cohort of UC patients (n=103) and healthy controls (HC, n=116); Primary Cohort (PC). Inflammation-associated signatures were validated in an internal validation cohort (VC-1; UC, n= 401; HC, n = 243). Additionally, 3 distinct validation cohorts (VC-2 a, n=23; VC-3 b, n=48; VC-4 c, n=214) were used to determine cellular and molecular phenotypes that were associated with TNFi-treatment response. a. Gut, 2018; PMID-27802155 b. Am J Gastroenterol., 2011; PMID- 21448149 c. Lancet Gastroenterol Hepatol., 2022; PMID: 34798036 Results: Total RNAseq and scRNA seq analyses and FC revealed distinct immune perturbations in the EC of patients with UC, including a major increase in neutrophils, monocyte-macrophages (MoMac) and inflammatory macrophages, while EC-resident, homeostatic gd T cells were significantly reduced (Fig 1A, B). ST identified significantly reduced frequencies of mature epithelial subtypes in UC and significantly increased co-localization between multiple cell types, including epithelial cells and myeloid cells (Fig 1C), that was confirmed by microscopy (Fig 1D). A signature of 255 EC inflammation-associated genes was derived that reversed with TNFi. This included treatment associated reduced expression of genes such as CSF3R, FCGR3B, MZB1, PDPN, TREM, FPR2 with a concomitant increased expression of genes like BEST4, CA2, SLC16A1, UGT1A10 (Fig 1E). Interrogation of UC-associated inflammatory cell types within VC-2, VC-3, and VC-4 demonstrated that reduction in neutrophils, MoMac, macrophages, DCs and plasma cells and an increase in epithelial cells was associated with TNFi response. Furthermore, early (W8-W10) reductions in myeloid cell- and plasma cell-associated genes was associated with TNFi response (Fig 1F, G). Conclusion: Detailed multiomic characterization of the EC in UC reveals myeloid cell-, plasma cell- and epithelial cell-associated modules that are associated with non-response to TNFi and charts a course to define rational drug sequencing and combinations in UC.