Developing a novel 3D model of the intestinal epithelium as a tool to study the pharmacokinetics of new chemical entities

Abstract

New chemical entities (NCE) are in continuous development in pharmaceutical companies across the globe, in a never-ending arms race of human ingenuity against human disease. Methods for the testing of NCEs in the past have relied on the widespread use of basic cellular epithelial equivalents, generally made by culturing a single epithelial cell line on a 2D permeable plastic membrane, and animal models for the validation of NCEs. These techniques are utilised before progression onto animal and human trials and eventual commercial availability after thorough efficacy and safety testing.
The use of cell lines to create two-dimensional (2D) models of the intestinal epithelium have been the gold standard since the 1980’s. These models benefit as both a cost saving exercise due to their simplicity and in reducing the need to use animals in research. which then, as today, is both ethically important and physiologically justified because animal models are often unreliable models of human anatomy, tissue and function. However, 2D models are unable to recreate complex structural variations present in vivo, usually incorporating a single cell phenotype in a non-physiologically based system. Whilst these simple models are cheap and mass producible, their use for NCEs analysis leads to progression of poor clinical candidates to later phases of development which fail due to in vivo functional irrelevance, toxicity or poor pharmacokinetics. Newer methods have been developed which improve the in vivo characteristics of 2D models, often through inclusion of additional cells lineages, such as goblet cells to make use of their distinct functions. Likewise, paracrine effects of fibroblasts or immune cells are increasingly shown to have critical functions in directing epithelial homeostasis and development. Three-dimensional (3D) tissue equivalents, are an emerging technology able to model a number of systems in vitro, bridging the gap between 2D models and human tissues. Ultimately however, conventional 2D monoculture models such as Caco-2 remain the gold standard for pharmacokinetic and toxicity analysis of NCE.
It was hypothesised that Caco-2 model phenotypes can be improved through the use of fibroblast conditioned medias and application of cell lines into a 3D model. The aim was to develop a more developed understanding of the effect of the paracrine microenvironment and 3D culture on epithelial, specifically Caco-2, phenotype.
4
Through utilisation of the Caco-2 cell line along with fibroblast cells of varying origin (Colon carcinoma, normal small intestine and skin) this study was able to create both 2D Transwell paracrine cultures and 3D models of both the intestinal epithelium and mucosae respectively. Colon derived fibroblast cells were shown to secrete significant concentrations of Keratinocyte Growth Factor (KGF) into media under normal 2D culture conditions. Moreover, the addition of paracrine factors released by fibroblast cells into culture and direct 3D co-culture allows for the creation of models with enhanced structural characteristics over Caco-2 monolayers with distinct epithelial and sub-epithelial compartmentalisation and similar structural morphologies as seen in in vivo tissues. Functionally, models were tested for their pharmacokinetic capability to a number of model compounds. Comparison of model functional between models and reported literature values for tissues and Caco-2 controls suggests that paracrine fibroblast secretome and 3D cells culture has a significant effect on Caco-2 function. Significant variation between models was observed in this study suggesting further research into the mechanistic actions behind the morphological changes seen is required.