Dynamic Optimal Experimental Design Yields Marginal Improvement over Steady-state Results for Computational Maximization of Regulatory T Cell Induction in ex vivo Culture

The isolation of T cells, followed by differentiation into Regulatory T cells (Tregs), and re-transplantation into the body has been proposed as a therapeutic option for inflammatory bowel disease. A key requirement for making this a viable therapeutic option is the generation of a large population of Tregs. However, several factors, including cytokines in the local microenvironment produced by the T cells can impact the yield of Tregs during differentiation. As such, experimental design is an essential part of deriving and evaluating the importance of different cytokines and their concentrations in Treg differentiation. However, currently only single, constant concentrations of the cytokines have been investigated, which forms the main motivation of this work: perform experimental design in silico based upon time varying profiles of the concentrations of TGF-β, IL-2, IL-6, and IL-23. An optimization problem was formulated and solved that seeks to maximize the predicted induction of Tregs relative to Th17 cells, by selecting an optimal input function for each of the four cytokines. While this approach sounds promising, the results show that only marginal improvements in the concentration of Tregs can be achieved for dynamic cytokine profiles as compared to optimal constant concentrations. Since constant concentrations are easier to implement in experiments, it is recommended for this particular system to keep the concentrations constant and, furthermore, keep IL-6 as low as possible while continuously stimulating the system with high concentrations of TGF-β, IL-2, and IL-23.

Reference

A. Sinkoe, A. Jayaraman, and J. Hahn. "Dynamic Optimal Experimental Design Yields Marginal Improvement over Steady-state Results for Computational Maximization of Regulatory T Cell Induction in ex vivo Culture"

IET Systems Biology 12, No. 6, pp. 241-246 (2018)