Study of the role of the epithelium in the initiation and progression of intestinal fibrosis in Ulcerative Colitis

Introduction
Intestinal fibrosis in Ulcerative colitis (UC) is poorly studied whereas pathological fibrosis involves extracellular matrix (ECM) accumulation and/or muscularis mucosae thickening (in ~100% of UC surgical specimen with chronic injury and not with active inflammation). The mucosa of UC may reach a level of activation leading to proinflammatory cytokines and profibrotic factors (TGF-β) release, the induction of Fibroblast to Myofibroblast transition (FMT) (α-smooth muscle actin +) leading to excessive ECM deposition. Indeed, myofibroblasts derive from epithelial cells trough epithelial to mesenchymal transition (EMT), proliferation and FMT. The mechanism switch between repair and fibrosis remain unknown while its understanding could offer anti-fibrotic therapeutic targets. The Endoplasmic reticulum (ER) stress and chronic inflammation have been shown increased in CD stricture’s epithelium. The response to ER stress involves production of chaperones, transcription factors, the unfolded protein response (UPR), apoptosis and EMT. Anterior gradient protein homolog 2 (AGR2) is such an UPR ER chaperone protecting from EMT and presenting paracrine effects promoting fibroblast migration and FMT. Blockade of AGR2 or other factors involved in ER stress response produced by the epithelium in IBD might be of interest as new therapeutic target.

Aim
Our objective is to better understand the role played by the epithelium in UC intestinal fibrosis initiation and progression. We aim to develop a model of apical out organoid showing a profibrotic phenotype and to characterise the response of such inverted organoids derived from UC samples.

Methods
Apical-out organoids, derived from stem cells of patients undergoing surgeries or mucosectomies were characterized by proliferation (Ki67, OLMF4, LGR5) and differentiation markers (AGR2, MUC2, E-Cadherin, Villin1, Chromogranin A) by RT-q-PCR and by histology for polarity inversion. Differentiated organoids were treated transiently by Tunicamycin (Tm) at increasing doses (0-10 µg/mL). The profibrotic phenotype was obtained after a recovery phase taking place after ER stress induction and media change. This involves EMT markers detection on organoids materials (Villin1 vs E-Cadherin, ACTA2), as well as ER stress (BiP, AGR2, CHOP, sXBP1) by WB, RT-q-PCR or immunofluorescence (IF) imaging. The profibrotic capacities of organoids to produced paracrine drivers under such induction include extracellular AGR2 and functional assays on primary intestinal fibroblasts. These undergo FMT (increase in α-SMA by IF) after exposure to preconditioned organoids media collected during the recovery phase for selection of the best timing. Investigations of 2,5 Di-Tert-Butyl-P-Benzoquinone (TuBHQ), another ER stressor, is ongoing.

Results
Three cellular architecture models were tested: basal-out, apical-out organoids. The apical-out organoid showed a self-differentiating propensity upon complete inversion with KI67 (2 fold decreased in proliferation apical-out compared to basal-out organoids). MUC2, Vill1, E-Cadherin, Vimentin, Chromogranin A and AGR2 displayed heightened expression levels in inverted organoids (2.26 / 1.87 / 1.89 / 3.34 / 1.55 respectively). Differentiation was confirmed by higherfold changes for Muc2 compared to proliferation phase: 3.67x for Apical-out , 7.07x for 2D Mono-layer and 93.67x for basal-out oragnoids respectively. BiP, CHOP, sXBP1 showed no significant variation across the three models under basal condition. Histological analysis confirmed morphological disparities: Inverted organoids, outwardly oriented apical pole facilitating mucus drainage in contrast to, the basal-out model with mucus retention within goblet cells and the organoid lumen. Dose-response investigations done on inverted organoids utilizing Tm (1, 2, 4, 6, 10µg/mL) or TuBHQ (10, 20, 40, 60, 80µM) showed the release of eAGR2 into culture media (WB) without detectable cell lysis (GAPDH and B-actin negative). Analyses of cell extracts revealed the presence of AGR2 along with BiP itself with a positive Tm dose correlation. Tm increasing concentrations (1 to 6µg/mL) induced ER stress markers increases (fold change mean of 5 for BIP, sXBP1, CHOP and AGR2). Functional assessments on fibroblasts using recovery media, showed increases in α-SMA and morphological FMT changes for the Tm doses of 2/4/6µg/mL with an optimum at 2µg/mL. The recovery time after Tm stress showing highest eAGR2 release in the absence of cell mortality was 32 hours.

Conclusions
The characterization of apical-out organoids showed all the expected characteristics: architecture, morphology, as well as self-differentiations. The profibrotic phenotype using Tm induction and apical-out polarity organoids have been optimized and comparison with TuBHQ optimum is pending. Our biobank includes IBD and healthy controls enabling further multi-Omic comparisons of factors produced by organoids induced to a profibrotic phenotype.