University of Utah chemists have come up with an approach for chemically marking sites in DNA that are missing a purine or pyrimidine base so that this form of DNA damage can be detected using a protein nanopore system.
The method is somewhat akin to nanopore-based sequencing approaches being explored by other groups. But rather than measuring changes in current flow through the pore to determine nucleotide sequence, the researchers slap a chemical called a crown ether onto the DNA sugar exposed at so-called 'abasic' DNA sites so that they can specifically see these lesions.
As they demonstrated in a proof-of-principle study appearing in the online edition of the Proceedings of the National Academy of Sciences last week, the crown ether then reacts with cation-containing electrolytes to slow the movement of abasic DNA sites through the pore, while altering the electrical signal produced in this system in a way that allows for the detection of the abasic sites.
"We create these, sort of, big loops that generate a signal as it goes through the nanopore," University of Utah chemist Cynthia Burrows, the study's senior author, told In Sequence.
Burrows and her colleagues would ultimately like to develop a system that can detect DNA damage while simultaneously discerning the identity of bases moving through the pore. For the time being, though, they are less focused on the sequencing applications and more interested in finding ways to systematically identify DNA damage so that they can look at its distribution and relationship to disease risk, if any.
"Rather than going after the ATCG [nucleotide sequence] problem, I'm interested in finding the location of damage sites within the strand of DNA," Burrows explained.
"Let's assume we more or less know what the sequence is," she added. "We'll be able to now see a pattern of damage sites on top of that that will be quite interesting."
Though they are interested in learning about other kinds of DNA damage as well, such as oxidative damage or thymine dimers, Burrows and her co-authors decided to focus on abasic sites for the current study, citing the potential biological importance of these lesions and the dearth of single-molecule methods available to study these events.
"Failure to repair [abasic] sites poses a severe challenge to polymerases and also leads to strand breaks, DNA cross-links, and transcription mutations," they wrote, "resulting in cellular dysfunction."
"There is urgent need for methods that provide surrounding sequence information and have the ability to detect multiple damage sites per strand, a phenomenon of significant biological importance," they added.
In particular, Burrows explained, researchers suspect that clusters of DNA damage at sites in the genome that are particularly prone to such events would be more deleterious than random, sporadically distributed DNA damage.
But so far it has been difficult to determine if and where such clusters occur in the genome — studies that are needed before researchers can fully understand the DNA damage profiles associated with exposure to different environmental conditions or the clinical relevance of various DNA damage patterns.
Methods have been developed for labeling abasic sites with tags that can be detected by fluorescence microscopy, enzymatic assays, mass spectrometry, or other visualization techniques. But these approaches typically provide a population average across many DNA strands rather than offering information about where DNA damage occurs on each strand, Burrows noted.
And because abasic sites can't be amplified from a DNA template, she added, the ability to look at this type of damage in detail hinges on the availability of single-molecule methods.
Among the existing DNA sequencing methods, Burrows noted, there are hints that the Pacific Biosciences platform can identify some forms of DNA damage. Still, because unmodified abasic sites tend to stall polymerase enzymes indefinitely, it is expected to be tricky to see multiple abasic events along the same DNA strand in any system that relies on DNA polymerization.
"The delay [in polymerization] for an abasic site would be almost infinite," Burrows noted. "You really do not insert well at all opposite an abasic site."
In their own attempt to detect the "potholes" that occur when DNA loses a base spontaneously or as a consequence of enzyme-mediated processes, Burrows and her colleagues pursued a chemical modification method for flagging these lesions as DNA strands move through a nanopore.
The initial chemical modification takes advantage of the chemical properties at abasic sites — namely the aldehyde group that's exposed when the ribose sugar in DNA is missing a purine or pyrimidine base.
"When you have a missing base, the sugar and phosphate remain behind," Burrows said, explaining that is it possible to specifically attach a primary amine group to the ribose aldehyde.
They selected a crown ether amine called 2-aminomethyl-18-crown-6, or 18c6, that not only interacts with the aldehyde group exposed at abasic DNA sites, but also binds cations such as sodium, lithium, and potassium to form a more rigid ring shape.
As a result, the inclusion of such cations in the electrolyte leads to a distinct signal that can be detected when the crown ether-tagged DNA moves through a pore.
Based on their experiments, the researchers found that sodium was the ideal electrolyte candidate in the current nanopore system given its size and affinity for the 18c6 crown ether.
"Potassium binds too strongly, it's too rigid, it completely clogs up the pore so things don't go through, and lithium doesn't bind enough, so nothing happens," Burrows said. In contrast, sodium binds the crown ether in a manner that allows it to move through the nanopore while also producing a discernable change in current.
On the nanopore side, meanwhile, researchers took advantage of the very small holes that collaborator Henry White, an electrochemist at the University of Utah, and his lab members have learned to fabricate in glass through prior nanopore studies. These exceptionally small holes support a very stable lipid bilayer in which nanopore proteins are then embedded, Burrows explained.
"With that hole being just a couple hundred nanometers across, the lipid bilayer is very stable," she said. "So we get a very well behaved lipid bilayer that we can then insert just one protein into."
That protein is alpha hemolysin — a bacterial protein similar to the one that Oxford Nanopore Technologies is using in its proposed nanopore sequencing system, anticipated later this year (IS 2/7/2012).
Even so, Burrows noted that the version of alpha hemolysin that her group is using lacks the proprietary modifications that Oxford Nanopore researchers have introduced to the protein to enhance its performance (IS 10/12/2010).
"We've just worked with wild-type alpha hemolysin but instead [of modifying the protein] have tried to perfect our treatment of the DNA so we get an interesting signal coming out of it," Burrows explained.
Moreover, she noted, the same strategy outlined in the paper is expected to be compatible with other kinds of nanopores as well — for instance, the Mycobacterium smegmatis porin A, or MspA, protein that University of Washington biophysicist Jens Gundlach and his team are testing on the nanopore sequencing front (IS 8/24/2010) — though it may be necessary to tweak the crown ether-electrolyte combination used depending on the specific properties of the pore.
As such, Burrows does not believe her team's approach currently conflicts with any of Oxford Nanopore's alpha hemolysin patents. "Of course," she added, "if they got excited about this and wanted to collaborate, that's OK, too."
So far, Burrows and her colleagues have tested out their system using pieces of DNA that they synthesized in the lab. By introducing abasic DNA damage at specific sites, they were able to determine what type of electrical signal is produced as strands of DNA with crown ether and sodium-tagged abasic sites sit in or slip through the alpha hemolysin nanopore.
"We knew exactly where the damage was," Burrows said. "Then we can see, for something we absolutely know, 'How does it behave in the nanopore?'"
In a series of static experiments using streptavidin and biotin, for example, the researchers were able to hold specific bases within the pore for around a second — a lifetime in nanopore terms — to see what kind of signal they produced and determine the average signal associated with abasic sites under different conditions.
"By doing that kind of immobilization study," Burrows explained, "we were able to study a whole bunch of different adducts — crown ethers and things that we had added on — and that's where we homed in on the 18-crown-6 [ether] and the sodium-potassium balancing act that we were working with."
From there, they went on to do experiments on pieces of DNA with one or two known abasic sites as they moved all the way through the alpha hemolysin nanopore.
Because this transit occurs very quickly, Burrows noted, rapid data acquisition strategies and good signal-to-noise ratios are needed to catch DNA damage signals in those sorts of translocation experiments. Nevertheless, she said, the researchers saw a clear spike — a roughly 10 percent change in current signal — when 18c6 ether-tagged damage sites moved through the nanopore.
Now that it has demonstrated the feasibility of finding abasic sites in synthesized DNA, the team plans to test a similar system using DNA isolated from cells.
"The next step — and I think it will be straightforward — will be to go to DNA that we obtain from cellular samples and then take a look at the kinds of signals we get from that DNA," Burrows said.
There are details that need to be ironed out before such studies can go forward — for instance, to ensure that DNA isolation methods don't introduce artifactual DNA damage. But, she said, "I don't see it as a conceptual leap."
In addition, she and her co-authors noted that it might be possible to specifically convert other types of DNA damage to abasic sites so that these alterations, too, can be detected in the nanopore system.
Even further down the road, the team hopes to develop a nanopore system capable of both DNA damage detection and DNA sequencing, though Burrows cautioned that there are still questions about whether the error rate of nanopore sequencing will be low enough to reliably identify all or most of the single nucleotide changes in a given sequence.
A patent on the chemical approach used to modify abasic sites for detection using nanopores is pending. As for potential commercialization of this DNA damage detection system, Burrows said such efforts will likely depend on the results that come out of upcoming studies aimed at characterizing DNA damage and potential correlations with disease state or disease risk.
If such research does suggest that DNA damage detection is applicable in personalized medicine or other clinical arenas, the system could have commercial potential, she explained. But that is a possibility that still needs to be explored in a basic research setting.
"If we decide that this could be used to monitor damage that is really indicative of a disease state, that really gives us some predictive power, then there's a market for it," Burrows said. "Otherwise, and in the short run, the market is simply researchers interested in correlating DNA damage to disease outcome."