CAR T-cell therapy, a type of immunotherapy, requires a specific mix of CD8+ (killer) and CD4+ (helper) T-cells. Killer T-cells recognize and kill cancer cells, while helper T-cells help coordinate the immune response. Though current treatments extract these cells directly from the patient, T-cells derived from human induced pluripotent stem cells (hiPSCs) can be precisely engineered to have more reliable and reproducible function. However, existing differentiation protocols are not yet robust, with several protocols producing a mix of cells that are double-positive. The goal of this project is to develop a closed-loop controller that will be able to produce the optimal mix of helper/killer T-cells in a single differentiation protocol. Diverging from the traditional control problem in synthetic biology, where an integral controller is used to set the concentration of a protein to some desired level, here we are instead designing a controller that senses the current ratio of cells, regulates a bistable system through noise, and autonomously achieves the desired programmable ratio of cell type in a population of cells. The scope of this project is currently focused on immunotherapy applications, though closed-loop multicellular control is an important problem with applications across systems and synthetic biology.

I am participating in the SuperUROP program to conduct rigorous research and to strengthen my scientific communication skills. I am excited to gain experience in the fields of systems/synthetic biology and control theory while applying the theory and tools I have learned from some of my favorite courses, 6.310 (Dynamic System Modeling and Control Design), 18.03 (Differential Equations), and 6.370 (Probability).