DNA origami gets curves
A few years ago a friend of mine became mildly obsessed with the idea of building a “DNA cube”. At the time it seemed like a random and useless concept to me, but it turns out I was wrong (not the first time). In the hands of a skilled chemist, DNA represents a tiny and versatile building block in the realm of nanotechnology.
So why was my initial gut reaction to the DNA cube “that’s weird…why would you want to do that”? Well, the role of DNA in a living cell is as a database of genetic information. Contained within an organism’s DNA is (almost) all the information it needs to grow and reproduce. Nowhere in the central dogma does it say “DNA also makes a rather lovely scaffolding material”.
But the idea of using DNA as a structural nanomaterial rather than as a genetic element is not a new one. Almost 30 years ago Nadrian Seeman recognized that DNA possessed certain crucial characteristics that could be exploited by a nano-engineer. And down the rabbit hole we go, my friends…
1. Base-pairing between nucleotides is highly predictable and specific.
In the 1950s James Watson and Francis Crick (and Rosalind Franklin…but that’s a topic for another day) characterized DNA (deoxyribonucleic acid) as the genetic information responsible for heritable traits in all cellular organisms on this planet. At the molecular level, single-stranded DNA (ssDNA) polymers are strings of deoxyribonucleotides. There are four of these nucleotides, adenine, cytosine, guanine, and thymine, and the order in which they appear in the polymer defines the identity of that piece of DNA. But DNA doesn’t characteristically spend a lot of time in a single-stranded conformation. Two strings of ssDNA can come together and form double-stranded DNA if they are “complementary”. This complementarity is based on the fact that the four DNA nucleotides have specific affinities for one another. Adenines like to “base-pair” with thymines, and cytosines with guanines. Therefore, if two pieces of ssDNA share perfect complementarity they can come together and form a piece of dsDNA.
2. The structure of the DNA double-helix is well characterized and stereotypical.
dsDNA is not a 2-D entity. It forms a characteristic double-helix in the vast majority of situations (there are exceptions to this, but for the sake of simplicitylet’s ignore them for now). We know a lot about this double-helix at the atomic level. For example, we know that it has a 2nm diameter and the height of a single turn is 3.4nm. We also know that each turn, or helical repeat, comprises 10.5 nucleotide base-pairs. Knowing exactly how DNA behaves at its most fundamental therefore allows scientists to precisely exploit and manipulate its structure.
I’m sure a small alarm is starting to bleep in your head. Why did life evolve this double-helical dsDNA molecule? Wouldn’t a single strand do just as good of a job at holding genetic information? Well yes, but the double-stranded nature of DNA allows for very efficient and accurate replication (and repair). All cells at some point in their life need to replicate themselves, including (rather critically) replicating their genetic information. During the replication process each strand acts as a template for a new complementary strand, resulting in two identical copies of the double-stranded genome. Gen(ome)ius right?
3. Double-stranded DNA is relatively rigid, whereas single stranded pieces are more floppy.
Though DNA likes to be in a dsDNA situation, ssDNA has the useful property of being far more flexible. dsDNA is also restricted in terms of how it can move and what shapes it can make. This rigidity is of course beneficial if you want to make a cube, but what if you want to make a sphere? It turns out that a combination of dsDNA and ssDNA is the perfect malleable material for modeling curved shapes, as you’ll see later.
4. DNA molecules spontaneously self-assemble into predictable higher-order structures in a test tube.
As I mentioned, two complementary pieces of ssDNA can come together and form dsDNA, rather like a zipper. Importantly for scientists, DNA doesn’t need much help: It can recognize its complementary partner and form a stable double helix.
5. As a bio-compatible molecule, DNA could be used to engineer more complicated structures.
Seeman’s initial vision was to use DNA to form a regular lattice made up of repeating units of DNA molecules. By using certain nucleotide sequences, he hypothesized that the lattice could be exploited to organize proteins. Such organization could then aid in x-ray crystallography, in which a large number of proteins must be lined up in a repeating array in order to generate an image of them at atomic resolution. This model could then be extended, with proteins being used as cement or reinforcement in a larger DNA sculpture.
The first example of a 2-D DNA array was published by Seeman’s group in 1999. It was constructed in a piecemeal way from individual DNA blocks. But more recently the field has been addressing the challenge of folding one large piece of DNA into a particular shape. In 2006 the first example of DNA “origami” was published in Nature. In this biochemical version of the Japanese art of paper folding, Paul Rothemund figured out how to fold a dsDNA molecule (made up of two strings of 7000 nucleotides, a.k.a. 7 kilobase-pairs). In order to fold the long molecule he employed hundreds of scaffolding ssDNA molecules. These short “helper” strands were designed to base-pair specifically with regions of the longer ssDNA in such a way that the resulting dsDNA folded into a pattern, or became a smiley face.
A square peg in a round hole
In Rothemund’s system defined units of DNA, kind of like pixels, made up the finished shape. You could think of the DNA array as being just like your computer screen. You are reading these words because a bunch of tiny, square, black, pixels are organized into the shapes of the individual letters. The same is true for the DNA units.
But what if you wanted to make curved structures? We’ve all seen a horribly low-resolution picture in which the edges of people’s faces are “pixelated”. The same flaw is true in the DNA origami pictures above. Well last week a report appeared in Science that showed the construction DNA shapes that were truly curved. As I said, dsDNA has to conform to certain rules due to an inherent rigidity. It can however bend a little bit, and this bending is enhanced if the traditional double-helical geometry is disrupted slightly. In the simplest shape in this paper, a series of concentric circles, Han et al designed a model in which the short helper strands orchestrated “cross-over” events within the long 7-kilobase origami strand. They then tested their model by synthesizing the scaffolding molecules and mixing them in a test tube with the origami strand. Using a very sensitive form of microscopy they imaged their DNA, and sure enough they saw the shapes they had predicted.
Perhaps the biggest breakthrough in this paper is that the authors were able to construct 3-D curved shapes. Shown are half and whole spheres, a hollow ovoid, and a “nanoflask”.
There are a surprising number of applications for this technology. In 2000 the first DNA tweezers were synthesized and used to build molecular-sized electrical circuits. But arguably the most interesting use of DNA nanostructures is in medicine. Potential applications include DNA-mediated drug delivery systems for unstable medicines, and diagnostic tests in which DNA structure can be used as a read out for circulating biomarkers.
This is still a technology in its infancy, an exciting time for a scientific idea. How do you think these ideas could be exploited? Let me know in the comments, I’ll give you a cut of the profits…
Han D, Pal S, Nangreave J, Deng Z, Liu Y, & Yan H (2011). DNA origami with complex curvatures in three-dimensional space. Science (New York, N.Y.), 332 (6027), 342-6 PMID: 21493857
Rothemund, P. (2006). Folding DNA to create nanoscale shapes and patterns Nature, 440 (7082), 297-302 DOI: 10.1038/nature04586
Mao, C., Sun, W., & Seeman, N. (1999). Designed Two-Dimensional DNA Holliday Junction Arrays Visualized by Atomic Force Microscopy Journal of the American Chemical Society, 121 (23), 5437-5443 DOI: 10.1021/ja9900398Mao, C., Sun, W., & Seeman, N. (1999). Designed Two-Dimensional DNA Holliday Junction Arrays Visualized by Atomic Force Microscopy Journal of the American Chemical Society, 121 (23), 5437-5443 DOI: 10.1021/ja9900398