(713d) Structure Modeling and Design of Coiled-Coil Protein Origami

Nanotechnology requires methodologies for high-throughput and scalable nanofabrication. A promising “bottom-up” approach is biomolecular self-assembly because of the programmability of biomolecules in their length, conformation, and intermolecular interactions. For example, DNA molecules have been used to fabricate strikingly complex and well-controlled nanostructures, using the method of DNA origami. The use of proteins as building blocks for origami, instead of DNA, can offer several advantages that include low cost and scalability for manufacturing as well as programmable biological functionality. Coiled-coils protein motifs, which have been widely used as versatile toolkits to create protein assemblies in various systems, can serve as DNA-like building blocks for origami. They form oligomers through the protein-protein interaction with controllable binding affinity, specificity, orientation, and oligomeric states. The sequence-to-structure relationships of coiled-coil motifs are well established, which have enabled the de novo design of synthetic coiled-coil sequences that orthogonally interact. Indeed, orthogonally interacting coiled-coil motifs have been modularly arranged to assemble well-defined protein nanostructures that consist of connected coiled-coil helices. This method is termed coiled-coil protein origami (CCPO), which is based on the modular recombination of coiled-coil motifs into fusion proteins. Mediated by specific and orthogonal intra- or inter-strand dimerization of the coiled-coil modules, the coiled-coil fusion proteins are expressed, folded, and form a nano-assembly. In CCPO, modeling has been used as a powerful tool for the characterization of origami structures in atomistic resolution. However, there are no reliable tools to assess the model quality, which has limited structure prediction without experimental data. In this study, we demonstrate a method of comparative modeling that we use to build CCPO structure models, predict the radius of gyration, and evaluate the models. The comparative CCPO models were built by templating with coiled-coil crystal structures or models and refining through MD optimization with simulated annealing. We investigated the effective MD optimization length to improve the modeling quality and derived a new method for model valuation. The modeling results were validated by small-angle X-ray scattering data, and our model evaluation method showed significantly improved predictive power than the methods implemented in comparative structure modeling tools. The modeling method demonstrated in this study is fast and does not require a significant amount of computational resources. Thus, it can enable high-throughput computational screening of a library of CCPO designs with enhanced structure characterization. We extended our approaches to various CCPO designs, which showed the potential as a useful design tool for computational CCPO design.