Algorithmic self-assembly of DNA: Theoretical motivations and 2D assembly experiments

Biology makes things far smaller and more complex than anything produced by human engineering. The biotechnology revolution has for the first time giv en us the tools necessary to consider engineering on the molecular level. R esearch in DNA computation, launched by Len Adleman, has opened the door fo r experimental study of programmable biochemical reactions. Here we focus o n a single biochemical mechanism, the self-assembly of DNA structures, that is theoretically sufficient for Turing-universal computation. The theory c ombines Hao Wang's purely mathematical Tiling Problem with the branched DNA constructions of Ned Seeman. In the context of mathematical logic, Wang sh owed how jigsaw-shaped tiles can be designed to simulate the operation of a ny Turing Machine. For a biochemical implementation, we will need molecular Wang tiles. DNA molecular structures and intermolecular interactions are p articularly amenable to design and are sufficient for the creation of compl ex molecular objects. The structure of individual molecules can be designed by maximizing desired and minimizing undesired Watson-Crick complementarit y. Intermolecular interactions are programmed by the design of sticky ends that determine which molecules associate, and how. The theory has been demo nstrated experimentally using a system of synthetic: DNA double-crossover m olecules that self-assemble into two-dimensional crystals that have been vi sualized by atomic force microscopy. This experimental system provides an e xcellent platform for exploring the relationship between computation and mo lecular self-assembly, and thus represents a first step toward the ability to program molecular reactions and molecular structures.