Effects of substrate-derived cues in driving the selforganization of organoid-derived intestinal epithelia

Abstract

Intestinal cells self-organize into 3-dimensional (3D) organoids that recapitulate the in vivo structural and functional characteristics when embedded in a 3D cell-derived protein mixture (Matrigel). However, these very same cells self-organize into 2-dimensional (2D) intestinal epithelial monolayers that recapitulate the in vivo-like cell organization when seeded on thin layers of the same cell-derived protein mixture. Moreover, in vivo, the intestine integrates regulation from paracrine signals to establish its characteristic crypt-villus axis self- organization. However, direct experimental manipulation of these paracrine signals, as well as their functional concentrations and effects at the cellular level, has been hampered by limitations of the in vivo and in vitro currently available systems. In general, changes in epithelial cell organization are characterized by a cross-talk between cell-substrate and cell-cell interactions, but the role of the ECM dimensionality, protein composition and spatial distribution in the intestinal epithelial cells’ organization is not fully understood. In this thesis, we show that intestinal epithelial cells self-organize in 2D-monolayers or 3D-tubular networks depending on the Matrigel protein concentration when the dimensionality is fixed. This self-assembles tubular networks have inner apical polarization and are similar to soap foams or de-wetted collagen networks. They have well defined topological and metrical properties and become spontaneously ordered at large length scales. Interestingly, stem cells have a particular dynamic during the formation of each self-organized patterns. On low Matrigel concentration, stem cells present a confined random movement to form a 2D-monolayer. In contrast, on higher Matrigel concentration, stem cells perform a direct motion towards a specific target to form the 3D- tubular networks. By reducing the proportion of stem cells in the culture, the formation of 3D-tubular networks is impaired. Instead, primary cells form aggregates when seeded above the transition protein concentration, similar to two other epithelial cell types (Caco-2 and MDCK cells). On the other hand, the 2D-monolayers formed on low Matrigel concentration contain crypt- and villus-like domains resembling those found in vivo. These compartments are randomly distributed and their shape is not uniform. However, by producing localized micropatterns of immobilized Wnt and ephrin factors on freeze-dried Matrigel-coated substrates by microcontact printing, we can drive the compartmentalization of the intestinal epithelial monolayers by spatially positioning the crypt- and villus-like domains. Finally, by changing the shape and dimension of the patterns we can control the distance between the crypt-like domains as well as their dimensions and shape. Overall, our experiments illustrate how Matrigel concentration regulates intestinal epithelial cell organization as a function of cell-substrate adhesion, and show that primary intestinal epithelial cells self-organize in structures with well-defined sizes and shapes independently of dimensionality or external signaling gradients. Also, we show that the amount of stem cells in the culture regulates the geometry of those self-organized structures. On the other hand, micropatterns of immobilized proteins to the ECM provides accurate control of the crypt-villus domain positioning in our epithelial monolayers. Thus, our work could yield insights about the roles of stem cells and protein concentration in tissue morphogenesis and their influence in the in vivo tissue morphological features such as the dimension of the crypts. In addition, we believe our platform will allow an easy and reliable manner to analyze the effect of relevant proteins on the epithelial cell compartmentalization, as well as the study of important intestinal epithelial processes such as stem cells proliferation, cell migration and differentiation both in homeostasis and pathological processes.

Las células intestinales cultivadas dentro de Matrigel se auto-organizan en organoides tridimensionales (3D) que recapitulan la organización del tejido in vivo. Sin embargo, estas mismas células cultivadas sobre láminas del mismo sustrato se auto-organizan en monocapas bidimensionales (2D) que también recapitulan la organización del tejido in vivo. Además, in vivo, el intestino integra señales paracrinas para establecer su característica auto- organización en criptas y vellosidades. Sin embargo, la manipulación de estás señalizaciones se ha visto obstaculizada por limitaciones en los sistemas in vivo e in vitro actuales. En general, la organización epitelial se caracteriza por interacciones célula-célula y célula-sustrato. Sin embargo, no se acaba de entender el rol de la dimensión, la composición y la distribución de la matriz extracelular (ECM) sobre la organización de las células epiteliales del intestino. En esta tesis, se analiza la organización de células epiteliales en monocapas 2D o redes tubulares 3D en función de la adhesión célula-sustrato. De esta manera, se ilustra la organización en estructuras con tamaños y formas bien definidas independientemente de la dimensionalidad o señalizaciones externas. Además, la proporción de células madre regula la geometría de dichas estructuras. Por otro lado, en contraste con lo que se observa in vivo, los dominios de cripta de las monocapas están desordenados y su forma no es uniforme. Mediante una plataforma que localiza micropatrones de proteínas sobre la ECM, controlamos el posicionamiento de los dominios de cripta-vellosidad. Para concluir, nuestro trabajo proporciona información sobre como influencia la composición y la distribución de la ECM y las células madre en la morfología del tejido in vivo, como la dimensión de las criptas. Además, la plataforma permite analizar el efecto de diferentes proteínas en la compartimentación de las células y en otros procesos epiteliales como proliferación, migración o diferenciación celular, tanto en homeostasis como en proceso patológicos.