"A challenge in nanotechnology is creating highly complex 3D nanoscale patterns," says Luvena L Ong, a researcher at Harvard University and MIT in the US, and first author of a Nature letter describing how to self-assemble as many as 33,000 DNA components into a pre-programmed 3D structure. Alongside a team of researchers spanning the US, France and China, Ong built structures from bricks of DNA strands. She describes these bricks as "molecular lego" since each brick only fits a predefined partner. The team chemically synthesized tens of thousands of these unique DNA bricks with specific chemical sequences to pair up in such a way that when thrown in together and annealed, they would self-assemble into a given structure. Using this approach, the researchers were able to assemble discrete structures up to 1 GDa and uniquely addressable structures up to 0.5GDa.

The approach is not new but the scale is so far unprecedented, exceeding the size of previous benchmark structures by two orders of magnitude. "The major challenges for scale-up of the DNA bricks approach include low yields, assembly times, and strand concentration," Ong tells nanotechweb.org. This is because the number of unique strands increases for larger structures, so that the overall concentration of each strand species decreases. Since the self-assembly process is stochastic, it takes longer for strands to find the partners they fit at lower concentrations, resulting in lower yields and longer assembly times. The solution Ong and her collaborators came up with was to increase the number of base pair binding domains in each brick from 8 to 13.

"This longer binding domain showed higher yields and faster assembly," says Ong. "Additionally, we found that our structures also assembled under a narrow range of temperatures, so assembling isothermally at a fixed temperature allowed us to achieve high yields." The researchers demonstrate the capabilities of the technique by constructing cuboids with complex shaped voids inside, including bunnies, teddy bears, interlinked rings, as well as sculpting the word "LOVE".

DNA origami

Alongside the success of the DNA lego approach there has been advance in "DNA origami", where "staple" strands of DNA fold a long "scaffold" strand of DNA into a bespoke structure. "DNA origami is particularly versatile in this context because each DNA strand in the origami nanostructure occupies a unique position and can serve as a uniquely addressable pixel," explain Grigory Tikhomirov, Philip Petersen and Lulu Qian at Caltech, US, in their letter in the same issue of Nature. "However, the scale of such structures has been limited to about 0.05 square micrometres, hindering applications that demand a larger layout and integration with more conventional patterning methods." To get around this they produce 2D arrays of DNA origami tiles using a hierarchical, multistage assembly process, which they demonstrate in a 700 nm × 700 nm replica of the Mona Lisa.

Also in the same issue Klaus F Wagenbauer, Christian Sigl and Hendrik Dietz at the Technical University of Munich scale up DNA origami in three dimensions by combining the approach with some natural selection rules. Based on the principle that to build closed rings or containers in 3D, you need to be able to define angles between components, their building blocks are V shaped DNA structures.

According Dietz the main limiting factors for scaling up DNA origami are assembly defects, deviations between the actual and believed geometry, interaction motifs that are too strong, and too much flexibility. "It is like building a big IKEA cabinet: if the pieces are not quite right, you won’t be able to assemble the whole thing," says Dietz.

Whereas the building blocks used by Ong and her team were defined by chemical sequence, Wagenbauer, Sigl and Dietz use geometry to encode higher-order assemblies in an hierarchical fashion. "The objects produced by Luvena Ong et al. could potentially also be used as a starting point for such an approach," says Dietz.

Mass production

In a second letter in the same issue Dietz’s team report on DNA self-assembly in mass quantities. They use bacteriophages – viruses that infect bacteria and replicates inside them – to generate single-stranded precursor DNA interleaved with self-excising ‘cassettes’. The precursor DNA contains the target strand sequence and the cassettes comprise two DNA-cleaving enzymes so that they can produce single strands of DNA of virtually arbitrary length and with virtually arbitrary sequences in a scalable and cost-efficient manner.

"In both papers we have set new records," says Dietz. "We report the biggest bottom-up assembled defined-size nanostructures, with 1.2 Gigadalton or 1.8 million basepairs. And we also succeeded for the first time to produce macroscopic quantities for DNA origami – personally I think that this is the advance with the immediately highest impact."

Further details on all four papers can be found in:
• Nature doi:10.1038/nature24648
• Nature doi:10.1038/nature24655
• Nature doi:10.1038/nature24651
• Nature doi:10.1038/nature24650