MIT software could bring 'DNA origami' to the masses
New program helps predict complex 3D structures based on a given DNA template, opening the door to developing more targeted drug delivery systems, synthetic photocells, and more.
DNA molecules are not merely carriers of information. They are also highly stable and programmable, which is why researchers have been working so feverishly on a design strategy called DNA origami.
And now a team at MIT is developing a program that makes the game playable by more than just a select few.
DNA origami--constructing specific 2D and 3D shapes out of DNA strands--could prove to be a highly effective means of developing nanoscale tools, such as synthetic photocells that perform artificial photosynthesis and highly targeted drugs (think of sending a cancer drug to hunt down a specific tumor).
But it's still young. Paul Rothemund of CalTech first introduced DNA origami in 2006 (thereby making the cover of Nature and delivering a TED Talk showing tiny DNA smiley faces), and William Shih's lab at Harvard Medical School was able to up the game from 2D to 3D a few years later.
The result is that today a small number of brilliant and highly specialized minds are bent over a nanoscale game of origami, playing with various sequences to try to build specific shapes for specific tasks. Imagine a room of highly sophisticated gamers playing with building blocks in a world without Tetris; if they had the game, they'd be able to work faster.
This is where the team at MIT, led by biological engineer Mark Bathe, comes in. They've developed software that makes it far easier, with a given DNA template, to predict the three-dimensional shape that will result.
"They're sort of building blocks, but it's even more crude because DNA is just a sequence," Bathe says. "It's taking the places you would connect the DNA together and predicting with a computer what it would look like in the final shape. The goal is to really have this be in the inverse, so the designer wants to make a box or a basket or a gear and then the program tries different folding combinations to give you the shape you want."
DNA comprises a string of four nucleotide bases called A, T, G, and C, with A binding only with T and G only with C. Rothemund found that he was able to get a long strand of DNA to fold using a viral genome that consisted of 8,000 of these nucleotides to create 2D stars, triangles, and yes, those smiley faces. That one strand served as a scaffold for the rest of the structure, with literally hundreds of shorter strands (only 20 to 40 bases in length) combining with the long strand to hold its desired shape.
Bathe says his software presented a mathematical and computation challenge, but that because DNA is governed by physics in terms of how it bends and twists and folds, DNA origami is very clean and obedient. Proteins, he says, are much messier, making protein-folding far more complex, which is why the game Foldit exists. (Researchers opened the process up to the masses in the hopes that a greater volume of people working on the problem might speed up progress.)
Bathe and his team, who haven't resorted to a game just yet, provide a primer of their software in the Feb. 25 issue of Nature Methods, and they're already working on making the program more automated and "unsupervised," because at this point it's still largely manual.
"Designers still have to guess the rules and then based on the shape modify the rules to get closer to the shape," Bathe says. "It's the Holy Grail to say, 'I want this,' and then it happens. We've made quite some progress already, so I think in the next half year to a year that should be coming out."
Ever the optimist, Bathe was quoted in the MIT news release saying, "Once nonspecialists can design arbitrary 3D nanostructures using DNA origami, their imaginations can run free." Ever the realist, I had to ask whether such an achievement might also be risky in the wrong hands. For Bathe, this is the conundrum we face in light of most advancements; the potential for progress, he hopes, far outweighs the risks.