It’s no secret that the first big period of revolutionary activity in the sphere of DNA origami produced some encouraging results. In fact, one of the creations that materialised during that time still smiles back at us. An immortal relic, perhaps, of a brighter, happier age. Heralded as a ‘positive breakthrough’, Paul WK Rothemund’s infamous 2006 paper – widely credited with introducing DNA origami to the world – featured 20 or so 100nm-sized smiley faces all photographed grinning under microscopic inspection.

For those less interested in the subtle mechanics of how the core building blocks of life really operate, Rothemund’s work functioned as an impressive, if relatively simplistic, art installation. But, for more learned admirers, it was a successful example of how the natural self-assembly process inherent to DNA could be utilised to create new 2D and 3D nanostructures. The origami method, as Rothemund called it, takes advantage of the predilection for DNA bases A, G, C and T to pair up with one another. By changing their sequence on DNA strands, researchers can bind them together to create different shapes in much the same way that the traditional Japanese art of paper folding can turn a square sheet into a paper crane. Along with cheery visages, Rothemund’s example included squares, five-pointed stars and discs.

Barriers ahead

Since Rothemund’s cover story was published in Nature, DNA origami has produced an assortment of interesting structures – from basic ‘maps’ of China and the Americas, to 3D cubes – but it has largely failed to overcome a problem first outlined by Lloyd M Smith, a chemistry professor at the University of Wisconsin, back in that same year. The barrier the industry will “have to surmount next is to deploy our knowledge to develop structures and devices that are really useful”, said Smith. It’s a problem that Khalid Salaita, senior professor of chemistry at Atlanta’s Emory University, has spent a significant portion of his career striving to overcome. “The vast majority of DNA structures, like the famous smiley face, for instance, tend to be static. They’re not dynamic,” Salaita says. “My angle has always been this coupling between mechanics and instructions, so, the question is how do we create a machine that can consume chemical energy – fuel if you will – and translate that into mechanical work.”

As Salaita clarifies, his use of the term ‘we’ is not merely directed at himself and his colleagues at the Salaita lab at Emory University – which focuses on developing cellular mechanobiology for a range of functions spanning fundamental developmental biology to cancer diagnostics – but it also applies to the nanoscience community more broadly. “From the inception of the entire [field of] nanoscience, there’s always been this dream of moving motors, machines and rotors,” he says.

Fundamentally, this means that Salaita spends a great deal of time trying to reconcile mechanical design and chemistry, unifying these disciplines to create natural ‘devices’ built from DNA that can fulfil very specific functions. Rather than viewing mechanical structures through the prism of their differences from biological counterparts, however, Salaita stresses the importance of seeing complex bodily processes as intrinsically machine-like in their own way. Grasping this concept, he explains, is the key to understanding the full potential of DNA origami.

“Even muscle contraction is driven by biological proteins, so actomyosin can, if coordinated with countless other motors, drive muscle contraction and generate huge amounts of mechanical work,” Salaita says. “The building block of a motor in biology is an enzyme that takes fuel and converts it into mechanical work. So, our angle is that we can create a completely synthetic version.”

Walk before running

As well as building on the earlier work of pioneering nanotechnologists such as the New York-based crystallographer Nadrian Seeman – who won the Feynman Prize in 1995 for synthesising the first ever 3D nanoscale object – the field has been buoyed by advancements in the genomics industry, specifically the decrease in the cost of synthesising DNA. Equally, the development of more sophisticated software tools means that nanotechnologists can now envisage more creative DNA structures.

From a medical standpoint, the hope is that one day, in the not too distant future, these machine-like vessels can travel deep within bodily cavities, acting as sensors or perhaps even drug delivery pods. Unlike artificial devices, which are often rejected by the body, DNA motors can function as Trojan horses, smuggling drugs deep into cellular cavities and directly targeting tumours, or so the thinking goes. Researchers at Harvard University’s Wyss Institute, for instance, have created a tube-shaped DNA nanorobot to target leukaemia and lymphoma. In theory, such a tool would hold anti-cancer drugs in place with aptamers – tiny DNA molecules – able to break apart and release the medication within the infected cells to target tumours. For now, however, these devices need to walk before they can run, let alone save lives. DNA motors need to first travel at sufficient speeds along a desired trajectory outside the body.

“The building block of a motor in biology is an enzyme that takes fuel and converts it into mechanical work. So, our angle is that we can create a completely synthetic version.”

This is the area in which Salaita and his contemporaries have had their most recent success. By collaborating with Yonggang Ke, assistant professor at Emory’s School of Medicine and the Wallace H Coulter Department of Biomedical Engineering, the Salaita lab was able to create the fastest, most persistent DNA nanomotor yet. The motor – created from 16 strands of DNA stacked four-by-four, one on top of the another – is fuelled by tiny portions of Ribonucleic acid (RNA) that attach to the main body. Bits of DNA protrude from the chassis like tiny little feet. The RNA binds with the DNA feet on the bottom face of the chassis, and an enzyme targeting bound RNA then destroys these RNA molecules. The process repeats, as more RNA pulls the DNA feet, tipping the chassis forward and causing it to roll. By gripping and detaching in this way the rolling device continues to move in a straight line, as opposed to the more random motion of other DNA motors. Due to its quirky rod design, Emory’s motor can travel the length of a human stem cell within two to three hours, unlike previous models that took an entire day to do so – if they actually made it that far. Moreover, as Salaita explains, the motor’s rod design means it is sturdier than the vast majority of prior DNA motors, which typically ‘walked’ on two legs.

“Originally, when we designed it, we thought they would slide or walk on the surface. We imagined these motors functioning like a lawnmower, sliding along the surface in a linear fashion,” Salaita says. “But we discovered that the motion could be through a rolling mechanism, making it fundamentally isotropic.”

The sturdiness of the motor is a crucial feat in itself, given how impacted these machines are by the powerful forces that govern interactions at the nanoscale level; a mysterious realm where ordinary rules and logic are turned on their head. The main problem with travelling in the microscale is the Brownian motion of molecules, which batter and buffet DNA motors like a kayak in a sea storm. Not only is this barrage unrelenting, but it is also fundamentally random in nature, hitting the motor from different angles. Liquids – specifically water – pose a similar threat, turning as viscous and dense as honey, making it hard for tiny motors to move in. “The viscous drag becomes enormous because your mass becomes inconsequential,” Salaita explains. “So, you’re really subject to the thermal energy, the bombardment of molecules and the water and the solvent around you – that’s what really dominates your motions.”

Due to its sturdy rolling pin-like structure and the RNA fuel that powers it, the motor managed to successfully overcome these obstacles, albeit within the safe and controlled confines of a laboratory. For now, the model will serve as a blueprint for other designers seeking to build more complex DNA structures.


Minutes it takes for Salaita’s motor to travel the width of a hair.

Emory University

A long way to go

Despite these advancements, Salaita maintains that there is still a long way to go before DNA ‘machines’ can operate within the body. Despite its rapidity compared with its predecessors, the motor still takes 1,000 minutes to travel the width of a hair. To become a viable option for drug delivery it will need to be able to cross centimetre distances at a quicker pace.

“We still have a way to go to get direct delivery, where we deliver the drug at one point in the body and it moves to its destination,” Salaita says. “It would still take a long time… but [we’re] not too far out. We’re quite close to the speed of biological motor proteins.”

For DNA motors to really work, two crucial obstacles must be overcome. First, researchers need to work out how to coordinate multiple devices in tandem – greatly increasing their combined power. Not only do these motors need to communicate with one another effectively, but also operate simultaneously as a unit without sacrificing speed or stability. Then there’s the tricky task of transitioning from artificial lab experiments to the body by designing structures that can weave through the bloodstream.

“We need to think of a new generation of motors that use, for example, the sugar coating on cells as the fuel,” Salaita explains. “Then the next challenge is how do you harness the biomolecules on the surface of cells or the endothelial lining of vessels to fuel the motor translocation? That’s very interesting.”

Despite these obstacles, the future looks bright for DNA nanobots, and not just because of the happy visages that first brought the domain into the realm of public attention. In a tech-obsessed world that puts robotics and sophisticated technologies on a pedestal, it might be that for all the impressive feats achieved by artificial machines, the most invaluable breakthroughs in drug delivery and diagnostics are achieved by harnessing the natural codes by which humans are made.