Phenotypically distinguishable eosinophilic cells do not impact epithelial functions in a triple-culture in vitro intestinal model

Abstract

The small intestine is a complex assembly of different cell types, such as enterocytes, secretory, immune, stromal and nervous cells. Due to this complexity, studying human tissue function ex vivo is challenging. As surrogate systems, in vitro co-culture models have been proven to be reliable and affordable. In this study, we used absorptive and secreting epithelial cell lines combined with differentiated eosinophilic cells to establish a triple-culture system to examine the impact of eosinophils on epithelial cell functions. We first differentiated an eosinophilic precursor cell line (EoL-1) using butyrate, forskolin, or dibutyryl-cAMP. In-depth characterization by real-time PCR, flow cytometry, functional assay and electron microscopy showed that only butyrate and dibutyryl-cAMP generated phenotypically distinct eosinophilic cells with different activation statuses, marked by differential expression of surface markers CD11c and CD62L, increased expression of eosinophil specific genes, and development of eosinophilic structural features. Then, a triple-culture system encompassing the enterocytic cell line Caco-2 and the secretory cell line HT29-MTX complemented with eosinophilic differentiated cells was established. Eosinophilic cells altered neither the proliferation nor survival of the culture. In order to get additional insights in possible changes of specific epithelial functions, we assessed the expression profile of different genes that are critical for various functions of the epithelia. The presence of eosinophilic cells did not affect the expression of crucial genes involved in intestinal barrier functions, nor did it modify the epithelial barrier function as demonstrated by electrical resistance and paracellular transport assays. However, mucus staining of the epithelial layer indicated that triple-culture with eosinophilic cells obtained using butyrate showed a tendency to a weaker mucus production. Furthermore, although the eosinophilic cells did not alter the epithelia, we observed the survival of butyrate-differentiated eosinophilic cells over a long period of time. Collectively, our data suggest that different triggers drive EoL-1 cells into phenotypically different eosinophilic cells with possibly distinct functions, mimicking the variability of eosinophils in vivo. Furthermore, this approach could be used as a stable triple-culture assay since differentiated eosinophilic cells showed no detrimental effect on epithelial functions.