Fig. 1: (a) Schematic representing the photochemical transitions required to perform a two-bit logic operation on DNA; (b) A circuit diagram of a two-bit transition; (c) Fluorescence image of binary addition 10+10 (taken from Reference 1).

The fabrication of the current generation of computers relies on lithography to generate the tiny transistors which process information. But the transistor size and hence processing power for conventional computing is restricted by the limits of lithography. So researchers are investigating new concepts for producing future computing devices.

Molecular computing offers an alternative approach to information processing. DNA is a particularly intriguing candidate as it can encode an enormous amount of information through the predictable molecular recognition between its base pairs. DNA computing has the potential to store greater amounts of information and carry out more complex processing tasks than conventional electronic processing.

Now, scientists from Japan have integrated DNA information processing with photochemical molecular manipulation to generate an autonomous computing machine which can carry out 16 binary addition calculations.

In DNA computing single stranded DNA molecules are introduced to a second DNA molecule on which the logic operation is performed. Molecular recognition occurs between the input molecules and the calculator molecule and the resulting supramolecule generates a 1 or 0 (on or off) output which can be read optically using fluorescent markers tagged to the DNA.

Fujimoto and colleagues1 have taken this concept much further. Their two-bit DNA adding machine requires an input molecule, an initial logic gate and components for molecular transition leading to the appropriate molecular set-up for a second input and second addition operation. The result of the one bit operation on the initial gate results in two possible pathways: the transformation of the initial logic gate to a transition gate created through photocleavage of the initial gate, binding of a new DNA sequence via the transition template and a photoligation process, or no transformation of the initial gate. A second input is then introduced and the second addition carried out.

Separate test tubes were used for the one-bit and two-bit operation and the outputs were measured by optical outputs from binding between the biotin-tagged DNA gates and streptavidin-cytochrome C. This photochemical transition scheme has many advantages compared to previously reported enzymatic approaches and could lend itself to other binary digit processing such as subtraction, say the researchers.