Ex Vivo Mimicking of Inflammation in Organoids Derived From Patients With Ulcerative Colitis

Despite extensive research, the exact mechanisms causing ulcerative colitis (UC) and Crohn’s disease, the main entities of inflammatory bowel diseases (IBD), remain unclear. Due to the complex nature of IBD, experimental models mimicking the in vivo situation are hard to find.1de Souza H.S. et al.Nat Rev Gastroenterol Hepatol. 2016; 13: 13-27Crossref PubMed Scopus (835) Google Scholar Intestinal organoids are 3-dimensional structures developing from multipotent epithelial stem cells residing at the base of intestinal crypts (crypt-based columnar cells). Crypts can be isolated from intestinal biopsies and, when cultured in the presence of specific growth factors, the epithelium stem cells containing crypts expand toward 3-dimensional organoids with a crypts and villi morphology resembling the epithelium of the gut in vivo. The conservation of self-renewal, self-organization, and tissue-specific differentiation provide major advantages compared with epithelial tumor cell lines, primary biopsies, or animal models.2Sato T. et al.Gastroenterology. 2011; 141: 1762-1772Abstract Full Text Full Text PDF PubMed Scopus (2109) Google Scholar Although studies showed that organoids maintain preserved genomic stability and disease characteristics,2Sato T. et al.Gastroenterology. 2011; 141: 1762-1772Abstract Full Text Full Text PDF PubMed Scopus (2109) Google Scholar,3Dotti I. et al.Gut. 2017; 66: 2069-2079Crossref PubMed Scopus (118) Google Scholar it remains unclear whether cultured organoids from patients with UC maintain the inflammatory character present in freshly isolated biopsies. In this study, we compared the transcriptional signature of biopsies, crypts, and colonic organoids derived from biopsies harvested from both inflamed and noninflamed regions of patients with UC as well as non-IBD controls, to evaluate the inflammatory characteristics over time during ex vivo culture. Next, we tried to (re-)induce inflammation in organoids of patients with UC and non-IBD controls. Eight non-IBD controls and 8 patients diagnosed with active UC, classified as an endoscopic Mayo sub-score of ≥2 and with an accessible border of macroscopic inflammation, were included at the University Hospitals Leuven (Leuven, Belgium). Fresh mucosal biopsies were obtained in both noninflamed and macroscopically inflamed (patients with UC only) regions to culture organoids.4Vancamelbeke M. et al.J Crohns Colitis. 2019; 13: 1351-1361Crossref PubMed Scopus (44) Google Scholar Biopsies, isolated crypts, and organoids after 1 and 4 weeks ex vivo culture were collected from all previously mentioned regions. Organoids (>4 weeks) were exposed during 24 hours to 4 inflammatory mixtures, to select the optimal mix to re-induce inflammation in organoids (>4 weeks) of inflamed, noninflamed, and non-IBD origin. Samples were analyzed by RNA sequencing by Lexogen QuantSeq 3ʹ and analysis was performed by DESeq2,5Love M.I. et al.Genome Biol. 2014; 15: 550Crossref PubMed Scopus (32893) Google Scholar xCell,6Aran D. et al.Genome Biol. 2017; 18: 220Crossref PubMed Scopus (1464) Google Scholar and Ingenuity Pathway Analysis.7Krämer A. et al.Bioinformatics. 2014; 30: 523-530Crossref PubMed Scopus (2799) Google Scholar Additional methods can be found as supplementary data. Baseline analysis of colonic biopsies showed an increased inflammatory character in biopsies of inflamed UC regions in comparison with noninflamed UC regions and non-IBD controls by up-regulation of inflammatory genes and pathways (Figure 1A, Supplementary Figure 1A and B). Also in crypts of inflamed UC regions, this inflammatory gene expression was maintained (Figure 1B, Supplementary Figure 1A and B), whereas xCell analysis and immune cell marker expression levels showed that this signature was not the result of remaining immune cells (Supplementary Figure 1C and D). After 1 week ex vivo culture, organoids of inflamed origin already lost most of their inflammatory gene expression (Supplementary Figure 1A and B). After 4 weeks ex vivo culture, organoids of inflamed and noninflamed UC origin were indistinguishable and clustered per patient instead, whereas organoids of non-IBD origin still maintained a different phenotype (Figure 1C, Supplementary Figure 1A and B, 2A). To enable the study of UC in the ex vivo organoid model, inflammation had to be re-induced. Organoids (>4 weeks) of inflamed, noninflamed, and non-IBD origin were exposed to different mixtures to select mix 4 (100 ng/mL tumor necrosis factor-α, 20 ng/mL interleukin-1β, 1 μg/mL Flagellin) as the optimal mix to re-induce inflammation (Supplementary Figure 2B). Inflammation could efficiently be re-induced in organoids of all conditions (Figure 1D and E) and showed activation of inflammatory pathways (Figure 1F, noninflamed origin). Remarkably, also after inflammatory stimulation, organoids of inflamed and noninflamed UC origin were indistinguishable and clustered per patient, whereas organoids of non-IBD controls still maintained a different phenotype (Figure 1D, Supplementary Figure 2C). Finally, we demonstrated that the re-induced inflammation in organoids of noninflamed UC origin showed similar activated pathways and genes compared with the inflammatory profile observed in crypts of inflamed origin (Supplementary Figure 2D and E). We demonstrated that organoids, derived from inflamed regions in the colon of patients with UC, lose their transcriptional inflammatory phenotype over time in the absence of inflammatory stimuli. xCell enrichment and immune cell markers demonstrated that the inflammatory signature in isolated crypts was not the result of remaining immune cells. Ideally, the decreased proportion of immune cells suggested by xCell should be validated by flow cytometry, but the limited yield did not allow this approach. After 7 days of culture of the crypts as organoids, the original inflammation was nearly completely abolished. One could hypothesize that too damaged cells may die off in the first days of ex vivo culture and cannot give rise to daughter cells and lead to the loss and transmission of some inflammatory signals. To address this, future single-cell RNA sequencing studies might be of interest. Dotti et al3Dotti I. et al.Gut. 2017; 66: 2069-2079Crossref PubMed Scopus (118) Google Scholar previously described that organoids of patients with UC maintain permanently dysregulated genes, compared with non-IBD organoids while Kraiczy et al8Kraiczy J. et al.Gut. 2019; 68: 49-61Crossref PubMed Scopus (86) Google Scholar demonstrated disease-associated DNA methylation in the gut epithelium. We confirmed differences between non-IBD and UC organoids but cannot exclude that organoids from active regions of patients with UC maintained any epigenomic signature. Interestingly, instead of clustering per inflammatory status, 4-week-old organoids clustered per patient, illustrating again conservation of patient-specific characteristics and supporting the feasibility of individualized analysis. Inflammation is one of the main features of UC and therefore essential to create a representative model. Remarkably, organoids that originated from inflamed or noninflamed regions in patients with UC, did not behave different on inflammatory stimulation, in contrast to non-IBD controls, indicating that intestinal epithelial stem cells seem to be intrinsically affected by the disease, although we cannot exclude previous inflammation in macroscopically unaffected areas. The biological relevance of the observed difference in inflammatory response should still be clarified in further experiments. Pathway and gene expression analysis showed reactivation of pathways in organoids on inflammatory stimulation that were up-regulated in crypts of inflamed origin. We conclude that for organoid cultures, it is not essential to obtain biopsies in inflamed regions of patients with UC, but to re-induce inflammation before evaluation of other mechanisms underlying IBD. The R-spondin-1 (originally from Calvin Kuo laboratory) and WNT3A producing cells were kindly provided by Hans Clevers (Hubrecht Institute, The Netherlands). The Noggin-Fc cell line was kindly provided by Gijs van den Brink (AMC, The Netherlands). Transcript Profiling: EMBL-EBI ArrayExpress database (www.ebi.ac.uk/arrayexpress), accession number E-MTAB-7860 Part of these data have been presented at the plenary session European Crohn’s and Colitis Organisation, March 2019, Copenhagen, Denmark; plenary session Belgian Week of Gastroenterology, February 2019, Antwerp, Belgium; poster presentation Digestive Disease Week, May 2019, San Diego, CA; and plenary session European Crohn’s and Colitis Organisation, February 2020, Vienna, Austria. Kaline Arnauts, Master (Conceptualization: Equal; Data curation: Lead; Formal analysis: Lead; Methodology: Lead; Writing – original draft: Lead). Bram Verstockt, MD, PhD (Conceptualization: Supporting; Data curation: Supporting; Formal analysis: Supporting; Methodology: Supporting; Supervision: Supporting). Anabela Santo Ramalho, PhD (Data curation: Supporting; Methodology: Supporting). Séverine Vermeire, MD, PhD (Conceptualization: Supporting; Supervision: Supporting; Writing – review & editing: Supporting). Catherine Verfaillie, MD, PhD (Conceptualization: Supporting; Supervision: Supporting; Writing – review & editing: Supporting). Marc Ferrante, MD, PhD (Conceptualization: Lead; Funding acquisition: Lead; Supervision: Lead; Writing – original draft: Supporting; Writing – review & editing: Lead). Download .pdf (1.13 MB) Help with pdf files Supplementary Data