Computational Design of Self-Assembling Protein Nanomachines

Silicon-based technologies are fast approaching physical limits of miniaturization. Achieving ever-higher densities of useful work under specified quantity of time, material, space, energy and cost, is a critical challenge in the 21st century. Capabilities of natural biomolecular nanomachines suggest that alternative systems could be designed de novo for extreme density and energy-efficiency. However, engineering functional and scalable biomolecular architectures remains elusive since tools have been lacking to achieve accurate assembly at nanoscales. Interestingly, proteins are versatile and modular components capable of self-assembling with sequence-dependent programmability, thus offering a vast engineering playground. Although proteins rely on complex folding, allosteric mechanism and interfaces of non-covalent interactions, recent advances in the development of the Rosetta software now allows the computational design of accurate de novo protein nanomaterials. For the first time, it appears possible to harness the thermodynamics and symmetry of protein-protein interactions to realize protein nanomachines and information processing systems. Our research proposes to investigate design rules and computational methods for the design of multidimensional symmetric self-assemblies with built-in dynamic behavior, to allow the implementation of protein-based information processors and actuators. Specifically, we currently focus on the systematic design of de novo protein homo-oligomers and 3D nanocage lattices of various symmetry groups. Expanding our toolbox of synthetic building blocks available should be useful not only to build new sorts of information processing systems, but also to engineer a vast range of nanodevices for medicine, material sciences and industrial bioprocesses.