Crisscross cooperativity for robust synthesis of molecular hierarchies

Abstract

To what extent is it possible to program molecules to self-assemble into increasingly complicated arrangements? While structural DNA nanotechnology has provided tools for scientists and engineers to self-assemble approximately any single nanostructure, it has proven remarkably difficult to assemble many of these subunits together. Existing methods are subject to yield-limiting failures resulting from spontaneous nucleation of subunits and are not suitable to create higher-order assemblies with sophistication rivaling the hierarchical complexity of natural systems. As a solution, this thesis presents crisscross cooperative assembly as a generalizable design principle with elongated slat monomers that engage other slats beyond their nearest neighbors. The length of the slats can be increased as desired to achieve any coordination number which enables the simultaneous realization of two prize characteristics in seeded self-assembly: Large kinetic barriers to spontaneous assembly are maintained using reaction conditions which promote fast growth of the desired structure if initiated from a seed. We proceed to show the crisscross architecture can be physically realized with two distinct DNA designs. Both elongated DNA origami slats and single-stranded DNA (ssDNA) slats which were ~200 times smaller in mass were resilient to spontaneous assembly across a breath of temperatures, ionic concentrations, and slat concentrations. By leveraging this nucleation control, we create a multitude of finite and periodic megastructures from DNA origami slats. The most intricate structure is a fully addressable canvas with lateral dimensions exceeding 2 μm and is composed of 1022 unique DNA origami slats, which is over an order of magnitude more distinct origami components compared to the literature precedent. With the ssDNA slats, we attained fast seed-dependent growth using high concentrations of monomers at up to ~10°C below the where slat binding was reversible, which has not been previously attained with other DNA tile-based monomers. Taken together, these specific implementations of crisscross with DNA are enabling for all-or-nothing formation of one- and two-dimensional periodic and finite structures with addressable nanoscale features. With further adaptation, these could be adapted for three-dimensional structures, algorithmic self-assembly, and zero-background signal amplification in diagnostic applications requiring extreme sensitivity. Longer term, we foresee crisscross assembly as a generalizable design principle of significance to the broader world or molecular self-assembly, where it could be adapted to other slat designs using other DNA and non-DNA monomers.