Shaping up nucleic acid computation

Abstract

Nucleic acid-based nanotechnology has always been perceived as novel, but has begun to move from theoretical demonstrations to practical applications. In particular, the large address spaces available to nucleic acids can be exploited to encode algorithms and/or act as circuits and thereby process molecular information. In this review we not only revisit several milestones in the field of nucleic acid-based computation, but also highlight how the prospects for nucleic acid computation go beyond just a large address space. Functional nucleic acid elements (aptamers, ribozymes, and deoxyribozymes) can serve as inputs and outputs to the environment, and can act as logical elements. Into the future, the chemical dynamics of nucleic acids may prove as useful as hybridization for computation.

Figure 1

Adleman’s solution of a Hamiltonian path problem (adapted from [2]). (a) The schematic figure of a 7-city Hamiltonian path problem. Each city is named by a number and color coded. (b) Examples of city strands and path strands, with color-coding as in (a). (c) Ligation of two path strands is facilitated by a city strand as a splint.

Figure 2

Toehold-mediated strand displacement. (a) Schematic figure of toehold-mediated strand displacement. Top panel: slow displacement without toehold. Middle panel: largely irreversible strand displacement. Bottom panel: reversible strand displacement, also called toehold exchange. (b) Reaction pathway for toehold-mediated irreversible strand displacement. Round and triangular ends of the lines represent 5′ and 3′ termini, respectively. Dashed line with arrow indicates toehold binding.

Figure 3

Examples of DNA circuits based on toehold exchange. (a) An AND gate, adapted from [8**]. (b) An entropy driven amplifier, adapted from [12]. Names of strands are shown in blue and names of gate complexes are shown in black. Solid gray lines with arrows indicate the reactants and products of individual reactions. Dashed gray lines with arrows indicate toehold binding.

Figure 4

Deoxyribozyme-based logic gates. (a) Sequence and secondary structure of deoxyribozyme E6. The substrate is shown in red. The substrate-binding and catalytic domains of E6 are shown in orange and black, respectively. The bolded “A” in the substrate denotes the ribonucleotide adjacent to the cleavage site (dashed red line). A schematic representation of this structure is shown in the inset. These schematics represents (b) a single-input YES gate, (c) an AND gate, and (d) a NOT gate. In (b) to (d) the input strands are shown in blue and dark green. The corresponding receptors (input-binding loops) are shown in cyan and light green, respectively.