Nanoscale programming of celluar and physiological phenotypes

University Of Virginia, Charlottesville VA

Investigators

Linked publications & trials

Abstract

ABSTRACT The foundation for our funded MIRA grant 1R35 GM134864 was the concept of a nanocomputing agent (NCA), a protein-based device whose function is responsive to specific stimuli. Specifically, we proposed designing computing devices based on a single protein, the function of which was “conditioned” on a set of extracellular or intracellular instructions. Our grand intent is to design subcellular molecular controllers of cellular phenotype that can be exploited in a variety of biomedical and pathological settings (like cancer or Alzheimer’s disease). The function of this protein follows the logic imposed by the communicated or sensed instructions. Hence, the manipulation of the input signals allows a computation to be performed by the engineered protein. Conceptually, the NCA agent is controlled by engineered response units (RUs) into target proteins, which sense diverse signals, such as light, small molecules (e.g., drugs), RNAs, proteins, and pH. Some RUs are natural, such as the LOV2 protein that responds to blue light by undergoing order-disorder transition, while some RUs are engineered, such as uniRapR which undergoes disorder-to-order transition upon binding a small ligand. The modularity of the NCA design allows for programmable control of arbitrary biological systems using desired external and internal queues. Over the past four years, we have made significant progress in the proposed directions to create functional NCAs. Not only were we able to create the first-ever single protein- based logic gate, but we also built the first-ever circuit featuring non-commutative logic displayed by a single protein. While we did not anticipate such rapid success with our tools, these proof-of-concept NCAs demonstrate the feasibility of the methodology. Furthermore, the biological findings based on the developed NCAs allowed us to develop several translational biomedical applications, including a new cancer immunotherapy approach that we are testing in melanoma mice models. Here, we propose to continue developing computational and experimental tools for rapidly designing and deploying NCAs for a wide range of applications. We will use molecular modeling, simulations, machine learning, and macromolecular design in combination with experimental studies of engineered biomolecules. Specifically, we plan to (1) develop a computational platform for designing allosteric wiring in proteins; (2) develop a repertoire of robust input triggers (RUs), specifically focusing on designing a small pH-sensing domain; and (3) develop a computational platform for RNA design. In addition, we plan to continue applying our technology to various biomedical systems with numerous collaborators. These directions highlight our shorter-term (5-year) goals towards our long-term vision of developing robust Lego-like NCA systems that enable precise control of cellular phenotype.