We aim to generate macroscopic crystals of nanoscale objects based on selective recognition and three-dimensional (3D) self-assembly. Such structures can be used e.g. for molecular storage. Initially, we have pursued the overall goal by using DNA hybrids. 3D crystals require branching in all three Cartesian directions during growth. While DNA is normally an unbranched molecule, nature provides some exceptions, e.g., in form of replication forks or Holliday junctions. Unfortunately, such naturally branched DNA structures are too flexible for defect-free “crystalline” recognition. Therefore, the specific approach of project C5 has been to make use of “designer” DNA hybrids made of symmetrical (and rigid) organic cores with covalently attached oligonucleotide arms.
First we have explored hybrid assembly and stability of supra-molecular DNA structures using atomistic simulations. In this fashion the growth of crystallites comprising up to 1,000 building blocks could be characterized and corresponding structural features predicted as a function of temperature, concentration, interaction strengths, coordination number of the cores and core stiffnesses (see subproject C5.1: Modelling Assembly and Growth of Crystals Based on DNA-Hybrids).
DNA hybrids were generated in a three-step fashion. First new rigid tetrahedral and pseudo-octahedral organic core structures were designed and synthesized. Different linker systems were then developed that allowed for attachment of the DNA strands to the corresponding cores. Using these preformed cores and linkers, complete DNA hybrids were finally prepared. Subsequently, the utility of these hybrids for self-assembly of macroscopic lattices was explored experimentally. Overall, well over 20 different hybrids, with varying lengths of the DNA chains, base compositions, and core molecules were generated. In a number of cases it has recently proved possible to isolate insoluble crystalline material from specific self-complementary hybrids (see C5.2: Synthesis of Functionalized Organic Nanostructures).
Enhancing DNA Hybrid Self-Assembly using a Biological Approach
Additionally, a biological approach has been pursued towards further enhancing DNA hybrid self-assembly processes. This exploits the properties of specific proteins, which are known to enzymatically catalyze relevant reactions, e.g., elongation of oligonucleotides. The possibility of assembling novel DNA structures via “kissing-loop” interactions (as first described for RNA-RNA recognition) has also been explored (see subproject C5.4: Protein-Assisted Assembly of Superlattices).