Development of an improved 3D in vitro intestinal model to perform permeability studies of paracellular compounds

Abstract

The small intestine is the primary site of drug absorption following oral administration, making paramount the proper monitoring of the absorption process. In vitro tools to predict intestinal absorption are particularly important in preclinical drug development since they are less laborious and cost-intensive and raise less ethical considerations compared to in vivo studies. The Caco-2 model is considered the gold standard of in vitro intestinal models regarding the prediction of absorption of orally delivered compounds. However, this model presents several drawbacks, such as the expression of tighter tight junctions, not being suitable to perform permeability of paracellular compounds. Besides, cells are representative of only one intestinal cell type, without considering the role of non-absorptive cells on the absorption pathway of drugs. In the present study, we developed a new three-dimensional (3D) intestinal model that aims to bridge the gap between in vitro tools and animal studies. Our 3D model comprises a collagen layer with human intestinal fibroblasts (HIFs) embedded, mimicking the intestinal lamina propria and providing 3D support for the epithelium, composed of Caco-2 cells and mucus-producing HT29-MTX cells, creating a model that can better resemble, both in terms of composition and regarding the outcomes of drug permeability when testing paracellular compounds, the human small intestine. The optimization of the collagen layer with HIFs was performed, testing different collagen concentrations and HIF seeding densities in order to avoid collagen contraction before day 14, maintaining HIF metabolically active inside the collagen disks during time in culture. HIF morphology and extracellular matrix (ECM) deposition were assessed, confirming that fibroblasts presented a normal and healthy elongated shape and secreted fibronectin and laminin, remodeling the collagen matrix. Regarding the epithelial layer, transepithelial electrical resistance (TEER) values decreased when cells were in the 3D configuration, comparing with the 2D analogs (Caco-2 and coculture of Caco-2+HT29-MTX models), becoming more similar with in vivo values. The permeability assay with fluorescein isothiocyanate (FITC)-Dextran 4 kDa showed that absorption in the 3D models is significantly higher than that in the 2D models, confirming the importance of using a more biorelevant model when testing the paracellular permeability of compounds.