Engineering in vitro models of the intestinal mucosa to study the role of the stroma on the epithelium

Abstract

[eng] The standard in vitro model of the small intestine in the field of drug development still consists in a monolayer of cancer-derived cells seeded on a hard substrate. Despite the huge advances that this model has provided, it does not faithfully recapitulate the complexity of the native tissue. Recently, the long-term culture of primary intestinal epithelial cells has been achieved through intestinal organoids, self-organizing 3D structures retrieving many of the features of the native tissue. Yet, they still have some limitations: (i) they are closed structures whose apical side is difficult to access, (ii) to grow, they require to be embedded in a highly heterogeneous matrix where well-controlled gradients of biochemical factors cannot be established, (iii) despite recapitulating the native cell compartmentalization, they do not retrieve the dimensionality of the tissue and (iv) they do not include the stromal compartment (mainly fibroblasts), proven to be key to maintain the intestinal stem cell niche. Therefore, the general objective of this thesis is to contribute to the development of complex in vitro models of intestinal epithelium focusing on the incorporation of the input of the stroma.

In the first section of this thesis, we successfully adapted an in-house developed photolithography-based technique to fabricate poly(ethylene glycol) diacrylate (PEGDA)-based hydrogels with crypt-like cavities. Although the nature of the chemical reaction used yielded partially-crosslinked polymer trapped in the cavities, the fabricated hydrogels could be functionalized with collagen I and organoid-derived epithelial monolayers could be formed on top.

In the second section of this thesis, we have developed an in vitro model that allows the culture of organoid-derived cells on hydrogels with villus-like architecture under gradients of the ISC niche. The fabricated hydrogels showed native tissue-like elasticity and microstructures resembling those of the native villi. The mesh size of the hydrogels allowed the diffusion of the factors of the ISC niche through the hydrogel, which formed stable gradients along the villus vertical axis. An in silico model of protein diffusion through the hydrogel was developed to predict the profiles of the gradients of the factors of the ISC niche. Organoid-derived single cells seeded on collagen I-functionalized hydrogels formed complete monolayers. Only under gradients of the ISC niche factors, as opposed to uniform concentrations, the monolayers displayed native-like cell compartmentalization. Finally, organoid-derived cells cultured under gradients that promote more the proliferation showed faster growth dynamics but also a depletion of the proliferative pool. Altogether, these findings demonstrate that our villus-like 3D intestinal epithelial model retrieves native-like cell compartmentalization under appropriate gradients of ISC niche factors. The in vitro model presented here could be used to systematically screen gradients of factors, of particular interest those secreted by fibroblasts, and study the cellular response.

In the third section of this thesis, we aimed to study if intestinal fibroblasts have a physical role in epithelial cell migration under non-pathological conditions. Through the coculture of fibroblasts with intestinal organoids, we found that the paracrine signalling from the fibroblasts induced a cystic morphology in the organoids, whereas their physical presence triggered their expansion into 2D monolayers. This physical effect of fibroblasts leading the expansion of epithelial cells was also observed in a 2D engineered coculture in vitro model. Organoid-derived monolayers exposed to a cell-free space (gap) migrated directionally towards the gap and fully closed it. During this process, fibroblasts were found to also migrate but towards the monolayers, align perpendicularly to the migration front and produce equally oriented ECM proteins that might be used for epithelial cells to migrate directionally. Under only the paracrine signalling from fibroblasts, epithelial monolayers did not migrate that efficiently nor fully closed

the gaps. Under no fibroblast input, epithelial cells migrated randomly and could not recover their integrity. All in all, we could demonstrate that fibroblasts are indispensable for epithelial integrity restoration under non-pathological conditions in a 2D coculture in vitro model.

Given the physical role of fibroblasts on the epithelial tissue, in the fourth section of this thesis we aimed to develop an in vitro model that incorporated a lamina propria mimicking compartment, together with the epithelial tissue, all the while recapturing the native architecture. To do so, we employed an in-house developed 3D printing technology to fabricate flat and villus-like gelatine methacryloyl (GelMA)-PEGDA hydrogels. First, using flat hydrogels, embedded fibroblasts were shown to be viable, have proliferation capacity and spread over the hydrogel surface. Organoid-derived epithelial cells seeded on top could only grow and form monolayers when fibroblasts were embedded within the GelMA-PEGDA hydrogels. In villus-like hydrogels, fibroblasts not only migrated to and spread on the surface, but they preferentially located at the tips of the villus-like microstructures. Organoid-derived single cells seeded on top of grew mainly at the tips and bases of the hydrogels. In these cocultures, fibroblasts migrated to the regions where epithelial cells were growing, mainly the tips and the bases of the hydrogels. Finally, fibroblasts cocultured with epithelial cells had their nuclei more elongated than when cultured alone, indicating a physical communication between the two cell types. Overall, we developed an in vitro model that mimics the native tissue architecture and encompasses not only the epithelium but also the mesenchymal compartment. Through this platform, we have unveiled a bidirectional communication between epithelial cells and fibroblasts.

[cat] Avui dia, el model in vitro estàndard de l'intestí prim al camp del desenvolupament de fàrmacs encara consisteix en una monocapa de cèl·lules cancerígenes sembrades sobre un substrat dur. Tot i els grans avenços que aquest model ha suposat, no recapitula de manera fidedigna la complexitat del teixit. Recentment s’ha posat apunt el cultiu d’organoids intestinals, estructures tridimensionals autoorganitzades que recapitulen moltes de les característiques del teixit nadiu. Tot i això, presenten diverses limitacions: (i) en ser estructures tancades, l'accés a la part apical és complicat; (ii) per créixer, necessiten estar dins una matriu altament heterogènia que dificulta poder establir gradients de factors bioquímics ben controlats; (ii) malgrat recapitular l'organització cel·lular del teixit, no mimetitzen la seva arquitectura i (iv) no inclouen el compartiment “estromal” (principalment fibroblasts), tot i que s'hagi demostrat ser essencial en el manteniment del nínxol de cèl·lules mare.

Al llarg d’aquesta tesi s’han desenvolupat diferents models que recapitulen moltes de les característiques mencionades més amunt. La primera secció d'aquesta tesi ha tingut per objectiu reproduir l’arquitectura del teixit intestinal. Per això, hem adaptat una tècnica desenvolupada al nostre laboratori que es basa en fotolitografia per fabricar hidrogels de diacrilat de polietilè glicol (PEGDA) amb cavitats similar a les criptes intestinals. En la segona part, ens hem centrat en crear un model que permetés establir gradients bioquímics de factors del nínxol de cèl·lules mare. En concret, hem desenvolupat un model in vitro que permet el cultiu de cèl·lules epitelials derivades d'organoids en hidrogels amb unes microestructures que reprodueixen les vellositats intestinals sota l’efecte de gradients bioquímics fisiològicament rellevants. En la tercera part, hem desenvolupat un model que permet el co-cultiu de cèl·lules epitelials derivades d’organoids amb fibroblasts primaris aïllats de l’intestí. A més, a través d’aquest model, hem pogut veure com els fibroblasts son essencials per a la migració de les cèl·lules epitelials en un context no patològic. En la quarta i última secció d’aquesta tesi, hem incorporat el compartiment estromal en un model que també reprodueix l’arquitectura del teixit, establint doncs un co-cultiu de cèl·lules epitelials i fibroblasts espacio-fisiològicament rellevant.