Programmable DNA Lattices: Design Synthesis and Applications

Recent Contract Accomplishments: July, 2002

Duke and NYU are characterizing various novel DNA tiles, and investigate the use of these and related tiles to form DNA lattices. We are working to characterize 1 D tilings of CTX tiles; we expect these form long cylinders which may be able to capture linear structures (e.g., carbon nanotubes). We are investigating of modifications of the TX molecules with additional Holiday junctions between the top and bottom dsDNA, so the resulting tile, a “Cylindrical TX tile” (CTX), has a cylindrical conformation and we will investigate the use of these and related families of DNA tiles to form 3D DNA lattices based on hybrids of CTX plus more conventional DX and TX tiles. We are investigating design of DNA molecular building blocks (MBBs) that can be rigidly attached to DNA lattices with a known orientation, and of unbounded binary counter (a component of demultiplexing RAM lattice). Caltech will rely on tried-and-true DNA tile structures. A new highly symmetric 2-tile DAE lattice has produced unexpected results, which we are investigating.

Duke and Caltech continued development of a mathematical/algorithmic framework for design of multi-strand DNA structures. We have begun to formulate the DNA design problem in terms of partition functions for multi-stranded DNA complexes, to examine tractable models of DNA pseudo-knots, and to develop software for specifying and creating 3D molecular models of DNA structures. Began design of nucleating structures for de-multiplexing RAM lattice. We also improved existing software for design of DNA nanostructures and their DNA sequences and tested that software for the design of improved triple-crossoverand single-strand DNA tiles.

A research objective of Caltech is development of a mathematical/algorithmic framework for design of multi-strand DNA structures. Caltech developed the first known algorithm for computing partition functions for possibly pseudo-knotted, single-stranded DNA structures. This allows the calculation of the probability that a desired target structure will result for a given DNA sequence; this provides an avenue for establishing both positive and negative design of DNA structures. The main student working on this project, Robert Dirks, passed candidacy with his proposal of a concrete example -- de novo design of an allosteric switchable ribozyme -- on which to test the design system experimentally. We have also established a framework for energy calculations on multi-stranded DNA complexes, and we are initiating work coding algorithms for minimum free-energy, partition functions, and stochastic kinetic simulations. Program design is beginning.

Duke and Caltech continued investigation of various assembly techniques for patterned 1D and 2D DNA lattices of moderate length, using techniques of unmediated algorithmic self-assembly, step-wise assembly, and directed nucleation assembly. We have identified strategies for patterning surfaces at the nanometer scale, including patterns required for nanoelectronic circuits, such as a RAM memory array and addressing circuits. We will soon move this project toward experimental realization using algorithmic DNA self-assembly. A research objective is design of nucleating structures for de-multiplexing RAM lattice.

Caltech has designed and synthesized strands for a preliminary two-dimensional algorithmic self-assembly system, and we have shown by atomic force microscopy that the linear border for the lattice forms, and we are working to dope this assembly with a seed to grow an X-shaped border. We have initiated a series of experiments to identify conditions where lattice grows only in the presence of border tiles.

Duke is targeted gold nanospheres into desired patterns by templating with DNA lattices. Our long-term goal has been to use self-assembling DNA templates to fabricate nanostructures with novel and technologically significant electronic transport properties. Materials displaying useful electrical characteristics are organized into desired patterns via the self-ordering properties of DNA complementarity. We made use of well-known gold-sulfur chemistry for binding of gold nanospheres to our DNA lattice. First, thiolated oligonucleotide was incorporated into the DNA lattice such that the -SH group is displayed on the free end of the stem helix protruding from the lattice at fixed sites. Gold nanoparticles have been added to the annealed lattice and should bind the immobile sulfur (see figure above). Alternatively, thiolated oligonucleotide can be reacted directly with gold nanoparticles to yield single-strand DNA labeled gold which can be subsequently annealed to its complementary strand displayed on the lattice on protruding stem helices. We employ DNA lattice to impose patterns of interest on the assembling gold, thereby allowing increasingly complex constructs with only small changes in the overall scheme. As shown in the figure above, the final step in the production of long continuous wires involves fusion of the immobilized spheres in the presence of dissolved gold salt and hydroxylamine. We incorporated thiol containing oligonucleotides into TAO AB lattice with surprising results(see below subsection on tubular DNA nanostructures). The thiols appear to be forming disulfide bridges under the conditions tested which distorts the flat lattice sheets into regular sized filaments. We are adapting the conditions to reestablish flat sheets and also working with metallization of the filament structures. We have begun our studies using gold-sulfur chemistry and are also developing other methods for targeting and immobilization of gold, other metals and single-wall carbon nanotubes.

Duke is improving visualization of self-assembled nanostructures by platinum rotary shadowing and electron microscopy. Duke executed further work on visualization of self-assembled DNA nanostructures by platinum rotary shadowing on transmission electron microscopy (TEM), yielding higher resolution images of DNA lattice than any previously available from atomic force microscopy (AFM), including the ability to visualize individual tiles. We have adapted a sample preparation technique typically used for examination of large, fibrous proteins to visualize our self-assembling DNA lattices. In collaboration with Harold Erickson’s lab in Cell and Molecular Biology at Duke, we have successfully applied the following protocol. Annealed DNA lattice was mixed 1:1 with 80% glycerol, sprayed onto freshly cleaved mica with compressed air and a micro-pipette, dried under high vacuum, rotary shadowed with platinum at a 4° angle, layered with carbon film from directly above, floated on water,captured on a 400 mesh copper grid, and examined by transmission electron microscopy (TEM). This new technique has yielded higher resolution images of DNA lattice than any previously available including the ability to visualize individual tiles. We have successfully examined TAO AB lattice and TAE computational complexes of various lengths. These results will be reporting in the literature in a paper which is currently in preparation.

The work by NYU for the period were largely devoted to defining DNA motifs that are compatible with 3D self-assembly of periodic DX DNA lattices. We have designed at two different motifs that produce some extent of X-ray diffraction, one, a TX motif diffracting to ~7.5 Å and a DX motif diffracting to ~8 Å. We suspect that the problems with these crystals result from DNA helicities incommensurate the design. We are screening molecules with varied helicities to solve this problem. We find that some helicities in the trigonal TX motif lead to crystals and some do not, supporting this hypothesis, although diffraction has not been obtained yet from the new species. We are performing a similar scan of the TX system.

Selected Recent Publications

J. H. Reif, DNA Lattices: A Programmable Method for Molecular Scale Patterning and Computation, special issue on Bio-Computation, Computer and Scientific Engineering Magazine, IEEE Computer Society. February 2002, pp 32-41.

J. H. Reif, The Emergence of the Discipline of Biomolecular Computation in the US (invited paper to the special issue on Biomolecular Computing), New Generation Computing, edited by  Masami Hagiya, Masayuki Yamamura, and Tom Head, Volume 20, No. 3, p 217-236, 2002.

H. Yan, X. Zhang, Z. Shen and N.C. Seeman, A Robust DNA Mechanical Device Controlled by Hybridization Topology, Nature 415, 62-65 (2002).

N.C. Seeman & A.M. Belcher, Emulating Biology:  Nanotechnology from the Bottom Up, Proceedings of the National Academy of Sciences (USA), in press (2002).

N.C. Seeman, Key Experimental Approaches in DNA Nanotechnology, Current Protocols in Nucleic Acid Chemistry, in press (2002).

N.C. Seeman, It Started with Watson and Crick, But it Sure Didn't End There:  Pitfalls and Possibilities beyond the Classic Double Helix, Natural Computing, in press (2002), p 53-84.

N.C. Seeman, DNA Nanotechnology:  Life's Central Performer in a New Role, Biological Physics Newsletter, in press (2002).

R. Sha, F. Liu and N.C. Seeman, Atomic Force Measurement of the Interdomain Angle in Symmetric Holliday Junctions, Biochemistry, in press (2002).