P018 Myeloid diversity in inflammatory bowel disease revealed by single-cell transcriptomics

Garrido Trigo, A.(1);Corraliza, A.M.(1);Melón-Ardanaz, E.(1);Veny, M.(1);Ordás, I.(2);Ricart, E.(2);Masamunt, M.C.(2);Panès, J.(2);Hemberg, M.(3);Mereu, E.(4);Heyn, H.(5);Azucena, S.(1);

(1)IDIBAPS, Inflammatory bowel disease group, Barcelona, Spain;(2)Hospital Clinic de Barcelona, Inflammatory Bowel Disease Unit, Barcelona, Spain;(3)Harvard Medical School, The Evergrande Center for Immunologic Diseases- Brigham and Women's Hospital, Boston- MA, United States;(4)Josep Carreras Research Institute, Cellular Systems Genomics Group, Barcelona, Spain;(5)National Centre for Genomic Analysis- Centre for Genomic Regulation, Single Cell Genomics Group, Barcelona, Spain


Background: Recent studies have applied single-cell RNA sequencing (sc-RNAseq) of the human intestine to resolve the cellular complexity and heterogeneity characteristic of inflammatory bowel disease (IBD). Nonetheless, an in-depth analysis of the myeloid mucosal compartment, a crucial player in gut immune responses, has not yet been conducted.


Methods: Colonic biopsies from healthy controls (HC, n=6), endoscopically active Crohn’s disease (CD, n=5) or ulcerative colitis (UC, n=6) patients were collected and immediately disaggregated. Single-cell suspensions were processed using the 10x Genomics Next-GEM Single Cell kit assays (Chromium 3’ v3.0) and the resulting libraries sequenced (NovaSeq 6000 S1, Illumina). Data from 3452 myeloid cells was separated in silico and analyzed using Cell Ranger v3.1.0 and the R package Seurat.


Results: In HC, mast cells, CD1c+ dendritic cells (DCs) and resident macrophages (RMΦ) in different transcriptional states were identified. All RMΦ clusters were HLA-IIhi and expressed C1Q genes, SELENOP, CD74 and LYZ. Within RMΦ, a distinct cluster of M2-like cells characterized by high expression of CD209, CD163L1, LILRB5, AXL and MRC1 formed an independent cluster. IBD showed a marked increase in both the number of cells and heterogeneity within the myeloid compartment compared to HC. While RMΦ and M2 macrophages were decreased, we observed an infiltration of inflammatory monocytes and M1-like macrophages expressing the monocyte markers VCAN and CD300E.  A cluster of TNIP3-expressing M1 macrophages that co-expressed ACOD1, KYNU, TRAF1, MMP19, IRAK2, TNF and IL1B was present in most UC and CD samples. Additional M1-like macrophage subsets, including a cluster characterized by high expression of CXCL5, HTRA1, SPP1, INHBA, MMP9, MMP14 and MMP10, or a cluster of NRG1, RETN, EREG, HBEGF, TGFB1 and VEGFA-expressing macrophages were identified in UC. Furthermore, IBD samples presented an additional cluster of mature DCs expressing CCL22, CCL17, CCL19 and CCR7. Importantly, our study is the first to capture infiltrating granulocytes in IBD patients by scRNA-seq; specifically, eosinophils expressing CLC, IL4 and IL13, and 3 neutrophil states, expressing FCGR2B, CMTM2, S100A8 and CXCL8.


Conclusion: We describe an unbiased approach for the identification of intestinal myeloid cells that revealed previously undescribed macrophage subpopulations. In addition, we discovered a marked heterogeneity within intestinal inflammatory macrophages that showed high patient-to-patient variability. We suggest that the diverse macrophage states in mucosal lesions could explain important differences in disease pathophysiology and contribute to the observed diversity in patient behavior.