Russell J. Deaton

A DNA based implementation of an evolutionary search for good encodings for DNA computation

Computation based on manipulation of DNA molecules has the potential to solve problems with massive parallelism. DNA computation, however, is implemented with chemical reactions between the nucleotide bases, and therefore, the results can be error-prone. Application of DNA based computation to traditional computing paradigms requires error-free computation, which the DNA chemistry is unable to support. Careful encoding of the nucleotide sequences can alleviate the production of errors, but these good encodings are difficult to find. In this paper, an algorithm for evolutionary computation with DNA is sketched. Evolutionary computation does not require error-free DNA chemistry, and in fact, takes advantage of errors to produce change and variation in the population. An application of the DNA based evolution program to a search for good DNA encodings is sketched.

Biomolecular computing and programming

Molecular computing is a discipline that aims at harnessing individual molecules at nanoscales for computational purposes. The best-studied molecules for this purpose to date have been DNA and bacteriorhodopsin. Biomolecular computing allows one to realistically entertain, for the first time in history, the possibility of exploiting the massive parallelism at nanoscales inherent in natural phenomena to solve computational problems. The implementation of evolutionary algorithms in biomolecules would bring full circle the biological analogy and present an attractive alternative to meet large demands for computational power. The paper presents a review of the most important advances in biomolecular computing in the last few years. Major achievements to date are outlined, both experimental and theoretical, and major potential advances and challenges for practitioners in the foreseeable future are identified. A list of sources and major events in the field has been compiled in the Appendix, although no exhaustive survey of the expanding literature is intended.

DNA‐Linked Nanoparticle Building Blocks for Programmable Matter

Programmable matter is a distributed system of agents that act cooperatively to configure themselves into arbitrary shapes with arbitrary functions. Molecular self-assembled structures containing many nanoparticles are candidates for programmable matter. Programmability implies that system designers are able to control the properties of assembly products. The system should be able to assemble into arbitrary, anisotropic shapes, like an electronic circuit, with the capability of incorporating different materials at specific locations within the structure. Defects or errors should be minimized, and 3D assembly should be possible. In self-assembly, component parts, or building blocks, interact locally to produce a coherent and organized whole. At the molecular level, the interactions are determined by “patches” that react between building blocks. Frequently, the assemblies exhibit collective properties that are distinct from those of their constituent components. These properties often depend upon the shape of the structure. Thus, the difficulty of programmability is really the difficulty of controlling the shape of resulting nanostructures. The ability to program the shape of a final assembly is computationally difficult and subject to frequent errors. Nevertheless, through careful design and implementation of building blocks, desired shapes and properties might be achieved. To maximize programmability (i.e., control), there should be a large number of types of patches available. Otherwise, there is no variety of interactions to assemble complicated shapes. The placement and relative orientation of patches on the surface of the building block should be controlled. Different types of patches should be able to be placed on the same building block to diversify the shapes available. Finally, the chemistry for patch conjugation to the building block should be relatively simple and sustainable, and it should be able to be used with a variety of materials. Because of its unique molecular recognition properties, structural features, and ease of manipulation, beginning with seminal work by Seeman and Chen, DNA has been considered as a promising material to achieve programmable assembly of nanostructures. Nanoparticle (NP) building blocks with different surface functionalities for DNA linkers have been reported. DNA computing verified the programmability of DNA-based nanotechnology and, in fact, demonstrated that DNA self-assembly was computationuniversal. DNA programmability has demonstrated the ability to assemble a variety of shapes 9] and, when NPs are incorporated, to control the position of NPs in linear, 2D, and 3D assemblies, including those based upon origami techniques. 10] Nevertheless, the rational self-assembly of functional structures with arbitrary shapes in all dimensions and at all scales that can incorporate many different NPs into a variety of final geometries remains difficult to attain. Herein we present a strategy to control the number, placement, and relative orientation of DNA linkers on the surface of a colloidal NP building block to maximize its programmability and realize enhanced control over the shape and function of final self-assembled structures. Figure 1 shows a schematic illustration of the assembly sequence to produce the DNA-linked colloidal gold NP building blocks (termed nBLOCKs). A measure of control was achieved by the sequential ligand replacement approach and stiff DNA linkers, which were shorter than the persistence length of double-stranded DNA. In the assembly reaction, the electrostatic repulsions and steric hindrances of DNA molecules influence the layout of DNA on a NP. The net charge of DNA at a pH value above its isoelectric point (i.e., pH 5) is negative, so it would tend to position on a NP to minimize its mutual electrostatic repulsions, in analogy to the valence shell electron pair repulsion model, thus contributing to the molecular geometry. Also, mutual steric hindrance could be another factor to further constrain the overall geometry, particularly using small NPs. In our strategy, a Au NP is functionalized by DNA strand by strand: for example, a NP with one DNA strand is the starting material for the second DNA attachment, a NP with two DNA strands is the starting material for the third DNA attachment, etc. (Figure 1; see Figure S1 in the Supporting Information). With this constraint, the position of DNA attachment would be chosen to minimize the electrostatic and steric interactions with existing DNA on the NP, yielding the optimal arrangement of DNA on a NP with up to sixfold symmetry, that is, linear (one and two DNA strands), T-shaped (three DNA strands), square planar (four DNA strands), square pyramidal (five DNA [*] Prof. J.-W. Kim, Dr. J.-H. Kim Bio/Nano Technology Laboratory Department of Biological and Agricultural Engineering and Institute for Nanoscience and Engineering University of Arkansas, Fayetteville, AR 72701 (USA) E-mail: jwkim@uark.edu

A PCR-based Protocol for In Vitro Selection of Non-crosshybridizing Oligonucleotides

DNA computing often requires oligonucleotides that do not produce erroneous cross-hybridizations. By using in vitro evolution, huge libraries of non-crosshybridizing oligonucleotides might be evolved in the test tube. As a first step, a fitness function that corresponds to noncrosshybridization has to be implemented in an experimental protocol. Therefore, a modified version of PCR that selects non-crosshybridizing oligonucleotides was designed and tested. Experiments confirmed that the PCR-based protocol did amplify maximally mismatched oligonucleotides selectively over those that were more closely matched. In addition, a reaction temperature window was identified in which discrimination between matched and mismatched might be obtained. These results are a first step toward practical manufacture of very large libraries of non-crosshybridizing oligonucleotides in the test tube.

A Software Tool for Generating Non-crosshybridizing Libraries of DNA Oligonucleotides

Under an all or nothing hybridization model, the problem of finding a library of non-crosshybridizing DNA oligonucleotides is shown to be equivalent to finding an independent set of vertices in a graph. Individual oligonucleotides or Watson-Crick pairs are represented as vertices. Indicating a hybridization, an edge is placed between vertices (oligonu-cleotides or pairs) if the minimum free energy of hybridization, according to the nearest-neighbor model of duplex thermal stability, is less than some threshold value. Using this equivalence, an algorithm is implemented to find maximal libraries. Sequence designs were generated for a test of a modified PCR protocol. The results indicated that the designed structures formed as planned, and that there was little to no secondary structure present in the single-strands. In addition, simulations to find libraries of 10-mers and 20-mers were done, and the base composition of the non-crosshybridizing libraries was found to be 2/3 A-T and 1/3 G-C under high salt conditions, and closer to uniform for lower salt concentrations.

Codeword design and information encoding in DNA ensembles

Encoding of information in DNA-, RNA- and other biomolecules is animportant area of research in fields such as DNA computing,bioinformatics, and, conceivably, microbiology and genetics. This surveyfocuses on two fundamental problems, the codeword design problemand the representation problem of abiotic information, formassively parallel processing with DNA molecules. The first problemrequires libraries of DNA sequences to be designed so that specificduplexes are formed during annealing while simultaneously preventingother undesirable hybridizations from occurring in the course of acomputation in the tube. The second involves a search for efficient andcost-effective methods of representing non-biological information in DNAsequences for storage and retrieval of large amouns of data (tera- andpeta-byte scales). Two approaches are treated, namely thermodynamic andcombinatoric-computational. Both experimental and theoretical resultsare described. A reference list of major works in the area is given.Finally, some open problems deemed important for their possible impacton encoding of abiotic information representation and processing arediscussed.

Design and test of noncrosshybridizing oligonucleotide building blocks for DNA computers and nanostructures

DNA oligonucleotides that anneal to form duplexes in specific, planned configurations are a basic construction material for DNA-based computers and nanotechnology. Unplanned duplex configurations introduce errors in computations and defects in structures, and thus, the sequences must be designed to minimize these effects. A software design tool has been developed that uses thermodynamic models of DNA duplex thermal stability and algorithms from graph theory to select good sets of oligonucleotides. An example set was tested in the laboratory, and the designed sequences formed no unplanned duplexes and had no detectable secondary structure.

In situ fluorescence microscopy visualization and characterization of nanometer-scale carbon nanotubes labeled with 1-pyrenebutanoic acid, succinimidyl ester

In order to characterize hybrid bio/abio technology utilizing carbon nanotubes (CNTs), in situ, real-time, yet noninvasive methods of accurate and reliable imaging are needed for observing CNTs’ interactions with biological materials, i.e., DNA, in biologically relevant aqueous environments. Optical visualization and characterization of individual CNTs in aqueous solutions were explored in this study using 1-pyrenebutanoic acid, succinimidyl ester (PSE) and a conventional fluorescence microscope. The results demonstrate the potential of fluorescence microscopy based on PSE-based staining methodology monitoring with nanometer resolution of individual CNTs and their manipulation with biological materials in bio/abio hybrid systems.

Characterization of Non-crosshybridizing DNA Oligonucleotides Manufactured in vitro

Libraries of DNA oligonucleotides manufactured by an in vitro selection protocol were characterized for their non-crosshybridizing properties. Cloning and sequencing after several iterations of the protocol showed that the sequences, in general, became more non-crosshybridizing. Gel electrophoresis of protocol product, also, indicated non-crosshybridization, and showed evolution in the population of molecules under the non-crosshybridization selection pressure. Melting curves of protocol product also indicated non-crosshybridization when compared to control samples. Thus, it appears that the protocol does select populations of non-crosshybridizing sequences.

Virtual test tubes: a new methodology for computing

Biomolecular computing (BMC) aims to capture the innumerable advantages that biological molecules have gained in the course of millions of years of evolution to perform computation unfeasible on conventional electronic computers. While biomolecules have resolved fundamental problems as a parallel computer system that we are just beginning to decipher, BMC still suffers from our inability to harness these properties to bring biomolecular computations to levels of reliability, efficiency and scalability that are now taken for granted with conventional solid-state based computers. The authors explore an alternative approach to exploiting these properties by building virtual test tubes in software that would capture the fundamental advantages of biomolecules, in the same way that evolutionary algorithms capture in silico the key properties of Darwinian evolution. We use a previously built tool, Edna, to explore the capabilities of the new paradigm.