Immunophenotyping Reveals No Significant Perturbation to PBMC Subsets When Co-cultured With Colorectal Adenocarcinoma Caco-2 Cells Exposed to X-Rays

Abstract

In vitro co-culture models between tumor cells and peripheral blood mononuclear cells (PBMCs) allow studying the interplay between these cell populations, potentially gaining insight into the in vivo response of the immune system to the presence of the tumor, as well as to possible other agents as radiation used for therapeutic purposes. However, great care is needed in the experimental optimization of models and choice of conditions, as some setups might offer a limited possibility to capture subtle immune perturbations. A co-culture model of PBMCs from healthy donors and colorectal adenocarcinoma Caco-2 cells was successfully adopted in a previous work to measure effects on Caco-2 and modulation of signaling when these latter are irradiated. We here tested if the same experimental setting allows to measure perturbations to the main PBMC subsets: we performed immunophenotyping by means of flow cytometry and quantified helper and cytotoxic T cells, NK cells, and B cells, when PBMCs are cultured alone (control), in presence of non-irradiated Caco-2 cells or when these latter are exposed to a 10 Gy X-ray dose from a conventional radiotherapy accelerator. To measure a baseline response in all experimental conditions, PBMCs were not further stimulated, but only followed in their time-evolution up to 72 h post-irradiation of Caco-2 and assembly of the co-culture. In this time interval PBMCs maintain a high viability (measured via the MTT assay). Caco-2 viability (MTT) is slightly affected by the presence of PBMCs and by the high radiation dose, confirming their radioresistance. Immunophenotyping results indicate a large inter-individual variability for different population subsets already at the control level. We analyzed relative population changes and we detected only a small but significant perturbation to cytotoxic T cells. We conclude that this model, as it is, is not adequate for the measurements of subtler immune perturbations (if any, not washed-out by inter-individual differences). For this purpose, the model needs to be modified and further optimized e.g., including a pre-treatment strategy for PBMCs. We also performed a pooled analysis of all experimental observations with principal component analysis, suggesting the potential of this tool to identify subpopulations of similarly-responding donors.