Inteins in the Loop: A Framework for Engineering Advanced Biomolecular Controllers for Robust Perfect Adaptation

Homeostasis is one of the cornerstones of life shaped by billions of years of evolution. A notion that is similar to homeostasis, but yet more stringent, is Robust Perfect Adaptation (RPA). A system is endowed with RPA if it is capable of driving a variable of interest to a prescribed level despite the presence of disturbances and uncertainties in the environment. Designing and building biomolecular controllers capable of achieving RPA have been identified as an important task which has immediate implications for various disciplines. Here, we develop systematic theoretical and experimental frameworks for custom-built proteins that exploit split inteins -- short amino acid sequences capable of performing protein-splicing reactions -- to design, genetically build and analyze a wide class of RPA-achieving integral feedback controllers. We first lay down a theoretical foundation that facilitates the screening of intein-based controller networks for RPA, and then usher an easy-to-use recipe to simplify their, otherwise complex, underlying mathematical models. Furthermore, we genetically engineer and test various controller circuits based on commonly used transcription factors in mammalian cells. We experimentally and theoretically demonstrate their ability of robustly rejecting external disturbances (that is achieving RPA) over an exquisitely broad dynamic range. Due to their small size, flexibility, modularity, lack of side effects and applicability across various forms of life, inteins serve as promising genetic parts to implement RPA-achieving controllers. To this end, we believe "inteins in the control loop" will leave a significant impact on various disciplines spanning synthetic biology, biofuel production, metabolic engineering and cell therapy among others.