A single-molecule DNA “navigator” that can successfully find its way out of a maze constructed on a 2D DNA origami platform might be used in artificial intelligence applications as well as in biomolecular assembly, sensing, DNA-driven computation and molecular information and storage. The device works thanks to a domino-like process dubbed “proximal strand exchange cascade”.
DNA origami exploits the base pairing of DNA’s four nucleotides, A, T, C and G, to produce an infinite variety of self-assembled engineered shapes. The resulting nanostructures can be used as scaffolding or as miniature circuit boards. A team of researchers led by Friedrich Simmel of the Technische Universität München and Chunhai Fan of the Chinese Academy of Sciences in Shanghai has now used the technique to make a maze whose structure resembles a mathematical “tree graph”. The edges of the tree are defined via anchoring sites constructed using DNA staple strands on the origami and vacant areas without staples correspond to ‘walls’ in the maze.
The structure is equivalent to a ten-vertex rooted tree with three junctions and contains one entrance and one exit defined and denoted as vertices ENT and EXIT, respectively (see figure).
Path paving
The researchers then placed a navigator made from a “DNA walker” in the maze. This nanobot can be directed to take a specific path along the network of tracks laid down on the DNA origami platform thanks to the progression of hybridization chain reactions on it, explains Fan.
“When we activate this system (with an initiator DNA molecule), a DNA hairpin structure immobilized on the origami substrate triggers a sequence of conformational changes (also known as path paving) on other DNA hairpins, from one neighbour to another, at well-defined positions. This mechanism, which we have dubbed proximal strand exchange cascade (PSEC), is like what happens in a domino show.”
OAS_AD('Middle');
Parallel depth-first search
The progressing PSEC stochastically turns at the junctions and corner points of the maze structure, he says. Each single-molecular navigator therefore autonomously explores one of the possible paths through the maze in a process known as a parallel depth-first search (PDFS).
“Our system is composed of a huge number of such single-molecule DNA navigators that collectively explore all possible paths through the maze and a vast number of PSEC events thus occur simultaneously.”
To filter out the correct exit path from all the other possible solutions, the researchers chemically modified the vertex of the exit of the maze. They then imaged this path using atomic force or super-resolution microscopy. The first technique measures the height difference between formed paths and unreacted areas, explains Fan, while the second “DNA-PAINT” technique allows for florescence imaging of the path with nanoscale resolution.
Towards autonomous and intelligent nanoscale robotic systems
“Our work is another important step towards realizing autonomous and intelligent nanoscale robotic systems,” he tells Physics World. “Such systems behave more like the biomolecular machines found in Nature that couple molecular action to simple decision-making processes.
“Our research should help advance the fields of DNA nanotechnology and biomolecular self-assembly as well as embodied artificial intelligence. Our navigator system could also find use in applications such as single-molecule sensing for intelligent disease diagnosis and treatment, as well as in molecular information storage and transfer.”
Fan and colleagues report their work in Nature Materials 10.1038/s41563-018-0205-3.
Want to read more?
Register to unlock all the content on the site
E-mail Address
Register
