In recent years, DNA nanotechnology has emerged as one of the most powerful strategies for bottom-up construction of nanomaterials. Due to the high programmability of DNA molecules, their self-assembly can be rationally designed. Engineered 3D DNA crystals, as critical products from the design of DNA self-assembly, have been proposed as the structural scaffolds for organizing nano-objects into three-dimensional, macroscopic devices. However, for such applications, many obstacles need to be overcome, including the crystal stability, the characterization methodology, the revision of crystal designs as well as the modulation of crystallization kinetics. My PhD research focuses on solving these problems for engineered 3D DNA crystals to pave the way for their downstream applications.<br>In this thesis, I started by enhancing the stability of engineered 3D DNA crystals. I developed a highly efficient post-assembly modification approach to stabilize DNA crystals. Enzymatic ligation was performed inside the crystal lattice, which was designed to covalently link the sticky ends at the crystal contacts. After ligation, the crystal became a covalently bonded 3D network of DNA motifs. I investigated the stability of ligated DNA crystals under a wide range of solution conditions. Experimental data revealed that ligated DNA crystals had significantly increased stability. With these highly stabilized DNA crystals, we then demonstrated their applications in biocatalysis and protein encapsulation as examples.<br>I also established electron microscope imaging characterization methods for engineered 3D DNA crystals. For crystals from large-size DNA motifs, they are difficult to study by X-ray crystallography because of their limited diffraction resolutions to no better than 10 Å. Therefore, a direct imaging method by TEM was set up. DNA crystals were either crushed or controlled to grow into microcrystals for TEM imaging. To validate the imaging results, we compared the TEM images with predicted models of the crystal lattice. With the advance in crystal characterization, DNA crystals of varying pore size between 5~20 nm were designed, assembled, and validated by TEM imaging.<br>The post-assembly ligation was further developed to prepare a series of new materials derived from engineered 3D DNA crystals, which were inaccessible otherwise. With the directional and spatial control of ligation in DNA crystal, I prepared new DNA-based materials including DNA microtubes, complex-architecture crystals, and an unprecedented reversibly expandable, self-healing DNA crystal. The integration of weak and strong interactions in crystals enabled a lot of new opportunities for DNA crystal engineering.<br>In the final chapter, I investigated the effect of 5’-phosphorylation on DNA crystallization kinetics. I found that phosphorylation significantly enhanced the crystallization kinetics, possibly by strengthening the sticky-ended cohesion. Therefore, DNA crystals can be obtained at much lower ionic strength after phosphorylation. I also applied the result to controling the morphology of DNA crystals by tuning the crystallization kinetics along different crystallographic axes. Together with previously methods to slow down DNA crystallization, the ability to tune DNA crystallization kinetics in both ways is essential for DNA crystal engineering.

<p> DNA origami is a powerful method to construct DNA nanostructures. It requires long, single-stranded DNAs. The preparation of such long DNA strands is often quite tedious and has a limited production yield. In contrast, duplex DNAs can be easily prepared via enzymatic reactions in large quantities. Thus, we ask a question: can we design DNA nanostructures in such a way that the two complementary strands can simultaneously fold into the designed structures in the same solution instead of hybridizing with each other to form a DNA duplex? By engineering DNA interaction kinetics, herein, we are able to provide multiple examples to concretely demonstrate a positive answer to this question. The resulting DNA nanostructures have been thoroughly characterized by electrophoresis and atomic force microscopy imaging. The reported strategy is compatible with the DNA cloning method; thus, would provide a convenient way for large-scale production of the designed DNA nanostructures. </p>