DNA sequences called telomeres, located at the ends of chromosomes, fold into compact structures called G-quadruplexes. These structures are named for the four repeated sequences of the nucleotide guanine that characterize telomeric DNA and shape the folding of the quadruplex. Cynthia J. Burrows, a chemist at the University of Utah and a recently elected member of the National Academy of Sciences, studies chemical modifications to damage-prone DNA structures, such as G-quadruplexes, and the mechanisms of folding and unfolding in these important DNA regions. By attaching DNA tails to G-quadruplexes and threading the tails through a hollow, mushroom-shaped α-hemolysin nanopore, Burrows and her colleagues examined how different folding structures affected quadruplex unraveling in a confined space. Burrows recently spoke to PNAS about her findings.

• Download figure
• Open in new tab
• Download powerpoint

Cynthia J. Burrows. Image courtesy of Alyssa Geisler.

PNAS: What does the work described in your Inaugural Article (1) reveal about G-quadruplexes?

Burrows: We’re working with the human telomere sequence GGGTTA, repeated at least four times. You need 12 guanines all together to fold up into this G-quadruplex. If we put this sequence into physiological salt conditions, mostly potassium, the sequence will fold into a hybrid fold. If you only have sodium present as the cation, then things wrap around a little bit differently and you end up with a basket fold. If you fold the same sequence in the presence of potassium but with low water activity, then you get the propeller-folded structure. With these three different folds, just coming from one sequence but different ion and solvent conditions, you get three different structures. They have very different dynamics for unraveling. If we just studied the thermodynamics of how stable a folded structure is around sodium or around potassium, then the thermodynamic answer would be about the same for all three. They’re all pretty stable folded structures.

But we found that the kinetics for unraveling were different by orders of magnitude. That, to us, was really interesting. It’s telling us something about how these macromolecules behave in a confined space. Imagine you’ve got a ball of string and you thread some of it through a hole that you’ve drilled in a piece of wood. As you pull on the other side of the string, the ball of string is going to bounce around and unravel really quickly. But if instead of a flat surface you had a cup that was about the same size and shape as the ball of string, and now to unravel it has to roll around in a confined space, then this is going to be much harder to do. You’ve got all of the interactions between the surface of this ball and the confined space it’s in, so it’s really hard to unfold within the cavity of the protein. If things are in confined space, then they’re going to behave very differently than if they are just out in a vast solution.

PNAS: What other applications do you see for α-hemolysin as a DNA sieve?

Burrows: α-Hemolysin and other protein nanopores are of considerable interest for DNA sequencing. These are open pores that form ion channels across a lipid bilayer and, using two electrodes on either side of the membrane, DNA is driven electrophoretically through this pore. You can watch the current flux as DNA interrupts the flow of ions; the goal is to sequence DNA very rapidly, very cheaply, on a single-molecule level.

We can capture molecules in α-hemolysin, and as things rattle around in this cavity we get an electrical signal and know something about what we capture and how it is folded. If there is an equilibrating mixture of species in solution, we can capture and release various molecules and sample the population distribution of these differently folded strands in the solution. So, for example, a previous paper (2) homed in just on the hybrid structure, which can fold in two analogous ways: hybrid 1 and hybrid 2. We could distinguish those using this technique. There’s a folding intermediate called a triplex instead of a quadruplex, because it’s partially unfolded; we could determine that triplexes were present in solution. It’s as though the nanopore acts as a sieve where we capture molecules one at a time, we can put them in bins, and we say what’s present in solution under these specific conditions.

This particular study focused on the sequence called the human telomere sequence at the ends of chromosomes, but if you look at all of the data from human genome sequencing, there are around 375,000 potential G-quadruplex–forming sequences in the human genome. That’s a huge number. They are likely formed as regulatory elements in gene promoter sequences, potentially at intron/exon boundaries. The structures of these could be very dynamic and could play a huge role in how we read out DNA and how DNA responds to different conditions, cellular stress of different sorts. So there are all kinds of structural information we can go after now, having this tool available to sample structures out of different solution conditions.

PNAS: What do your results reveal about the behavior of G-quadruplexes in a cellular environment?

Burrows: We start by building molecules and we want to be physiologically correct. We pick the right pH, the right metal ions, high potassium, low sodium, little bit of magnesium, etc. But then we really have to go much further than that to be anywhere close to what the cell looks like, because DNA is mostly wrapped up around proteins in the cell. It’s part of a very confined space and has a supercoil to it as well. It’s only when you reconstitute a huge number of components that you start to learn about what really might be going on in the cell. I think what we begin to learn is the effect of confinement on unfolding and refolding of structures as we study these coiled DNA polymers in small spaces.

So, would this happen in a cell where we have α-hemolysin in a cell membrane with a big chunk of DNA trying to go through? Probably not. I would say that this specific example is more interesting to how you thread DNA through the nanopore for sequencing applications or for other analytical applications of DNA. But still I think we learn something equally profound about how biopolymers that are all folded up can unravel and refold in confined spaces.