Researchers in Germany have found that a new combination of near-field optical techniques and vibrational spectroscopy could enable label-free single-molecule DNA sequencing.
In their proof-of-principle study, which appeared online in the international edition of Angewandte Chemie last month, the scientists used a combination of atomic force microscopy and Raman spectroscopy to analyze a single molecule of homopolymeric RNA.
Although they did not read the molecule base by base, they say that the technology provides enough sensitivity, contrast, and resolution to do so.
Atomic force microscopy on its own, which scans a sample with a silver-coated tip less than 20 nanometers in diameter, offers enough resolution to sequence single strands of nucleic acid, but it cannot distinguish between different bases, according to Volker Deckert, one of the study’s authors and director of proteomics at the German government-funded Institute for Analytical Sciences in Dortmund.
Raman spectroscopy, on the other hand, can distinguish DNA bases by their characteristic spectra, but its resolution is not good enough for detecting single bases. In order to overcome each method’s shortcomings, Deckert and his colleagues combined them into what they call tip-enhanced Raman scattering, or TERS.
They started out using TERS to study larger structures, such as virus particles, but “we wanted to get the ultimate lateral resolution and still get the spectra that we are used to,” Deckert told In Sequence last week.
Initially, he and his colleagues had planned to use DNA as their test case because “everyone knows what DNA is,” he said. However, it turned out to be “much easier” to obtain long RNA oligonucleotides with defined homopolymeric stretches than DNA oligos of that type, so the researchers chose RNA for their initial study.
In their study, the scientists were able to measure cytosine-specific Raman spectra at seven positions along a single RNA strand, each spectrum coming from a region about 30 to 60 bases long underneath the AFM tip. This showed that the signal-to-noise ratio of the method is good enough to obtain base-specific spectra from a single nucleic acid strand.
“If it was not only one [type of] base — like in our case, poly-C — we would be able to see differences between different bases,” Deckert said.
Secondly, the researchers were able to position the tip precisely over the RNA molecule several times. “This is not straightforward, because you have to have sub-nanometer positioning,” which required that the system was isolated from even small vibrations, Deckert explained.
The reason why they did not scan the RNA base by base is that they could not move the tip even more precisely, in one-base intervals. But Deckert, who is a physical chemist by training, is confident that a €15,000 ($22,000) add-on that will allow them to move in 100-picometer steps will solve that problem.
The total setup, consisting of an atomic-force microscope and a Raman spectrometer with a “very good” detector, costs approximately €250,000, he estimated.
Because the AFM tip covers several RNA bases, the lateral resolution is not good enough to interrogate each base directly. “I don’t think that can be achieved at all,” Deckert said.
But if the tip can be shifted by a single base, the difference in the spectrum would be enough to determine the next base. “If we have, let’s say, 10 to 20 bases underneath our tip, but we can move by one-base spacing, … for instance, A goes out [underneath the tip and] T goes in,” he explained.
In terms of analyzing DNA, “this AFM approach is very interesting for some specific occasions, for example looking at [DNA from] single virus particles that are very difficult to do PCR on,” Deckert said.
The main advantages of sequencing by TERS are that the method is label-free and does not require sample amplification, he said.
However, to develop the approach to enable routine, high-throughput DNA sequencing, Deckert and his colleagues plan to use a different approach that “will work on the same principle,” he said. “It is about bringing the sample to the tip, instead of bringing the tip to the sample.” Deckert declined to elaborate on how that technology will work.
“We wanted to get the ultimate lateral resolution and still get the spectra that we are used to.”
To increase the throughput, hundreds of probe tips could run in parallel. “At the moment, you can bring the active tips together in a size that’s limited by your laser focus spot,” which is 10 to 20 micrometers in size and allows 20 tips to be mounted per millimeter, Deckert said.
The speed of the system could also be increased. Right now, a speed of a second per base seems feasible, he said, but it “probably could go down by a factor of 10.”
The scientists will have to overcome a number of challenges on their way to developing a DNA sequencing technology.
In their study, they write that “entangling of the strands was a major issue and an extensive search for the linear single RNA strands needed for the measurements is required.”
Part of the reason was that the researchers used as little buffer solution and other chemicals as possible in order to avoid background signal, which increased strand entangling. “That’s a challenge for us, and we don’t have the background of molecular biologists,” Deckert said.
Also, the DNA needs to be single-stranded to be sequenced by their approach, which would require a heating or chemical step to melt double-stranded DNA, but Deckert does not believe that will be a major obstacle.
Finally, the AFM probe tips the researchers use are currently difficult to obtain, but Deckert said he and his colleagues are “in close collaboration” with an undisclosed company that they hope will make the tips commercially available.
The scientists have not patented their method, and are currently not working with any commercial entities to develop it for DNA sequencing, said Deckert.
“It’s definitely good work,” Kumar Wickramasinghe, a professor of electrical engineering and computer science at the University of California, Irvine, told In Sequence this week. “It’s good progress in tip-enhanced Raman spectrometry.”
However, Wickramasinghe, who developed the apertureless near-field optical microscope and the vibrating mode atomic force microscope, cautioned that some practical issues may be difficult to solve. “There is a long step from what [Deckert] has got right now to [determining] a DNA sequence,” he said.
For example, he pointed out that the researchers still need to show that individual bases can be distinguished by their Raman spectrum, and that the change in overall signal when the tip moves can be attributed to a specific base.
Also, the method will likely encounter problems in homopolymeric DNA regions where the incoming and the outgoing base under the probe tip are the same, and the overall signal does not change when the tip moves.
Wickramasinghe won a grant under the National Human Genome Research Institute’s advanced sequencing technology program last year to accelerate and miniaturize Sanger sequencing using nanoscale electrophoretic DNA separation along the surface of an AFM probe tip. He and his team plan to develop a “massively parallel sequencing platform” containing hundreds of probe tips, according to the grant abstract.
Wickramasinghe said he believes that Deckert’s method will probably have applications other than DNA sequencing, such as determining the amino acid sequence of proteins.
The German team is indeed using TERS to analyze the surface of bacterial cells and would like to apply it to protein sequencing as well, Deckert said.