Bioengineering Novel in vitro Co-culture Models That Represent the Human Intestinal Mucosa With Improved Caco-2 Structure and Barrier Function

The Caco-2 monolayer is the most widely used <i>in vitro</i> model of the human intestinal mucosa to study absorption. However, models lack communication from other cells present in the native intestine, such as signals from fibroblasts in the lamina propria. In this study, we have investigated the effects of fibroblasts upon the Caco-2 epithelium through two mechanisms: indirect signaling from fibroblasts and direct contact with fibroblasts. Culture of Caco-2 cells with paracrine signals from fibroblasts, through the use of conditioned media, did not induce a significant change in epithelial cell morphology or function. To examine the effects of direct contact between the epithelium and fibroblasts, we developed novel, humanized three-dimensional (3D) co-culture models whereby Caco-2 cells are grown on the surface of a subepithelial-like tissue construct containing intestinal or dermal fibroblasts. In our models, we observed endogenous extracellular matrix production from the fibroblasts that provides support to the above epithelium. The Caco-2 epithelium displayed morphological changes in 3D co-culture including enhanced polarization and the formation of a basement membrane-like attachment to the underlying stromal compartment. An important structural alteration was the significantly straightened lateral membrane that closely mimics the structure of the <i>in vivo</i> intestinal mucosa. This enhanced lateral membrane phenotype, in correlation with an reduction in TEER to levels more similar to the human intestine, is thought to be responsible for the increased paracellular permeability observed in 3D co-cultures. Our results demonstrate that direct contact between epithelial and mesenchymal cells results in an enhanced epithelial barrier. The <i>in vitro</i> models described herein have the potential to be used for studying intestinal epithelial-fibroblast interactions and could provide more accurate tools for drug permeability studies.