A new method for barcoding single DNA molecules may help improve genome sequencing on next-generation sequencers and pave the way for optical sequencing.
The new system tags restriction enzyme sites on genomic DNA and reads these tags, or barcodes, optically after threading the labeled DNA molecules into an array of narrow channels, so-called nanoslits.
It could be used as a front end to any of the existing next-generation sequencing systems, or developed further into its own sequencing device, according to David Schwartz, who invented the mapping system.
The mapping information that the system provides could help place DNA reads, and to assemble reads in de novo genome sequencing. “If you have barcoded, or mapped, molecules, that compensates a good deal for short reads,” said Schwartz, a professor of genetics and chemistry at the University Wisconsin-Madison. “Instantly, you know the location within the genome of any sequence read that you are able to obtain.”
The mapping information would also be useful for genome re-sequencing. “You need to be able to take account of structural variation,” Schwartz said. “With short reads, anything below 500 bases, it’s going to be very hard for you to discern and confidently identify structural variation.”
Schwartz also invented optical mapping, a method that has been commercialized by OpGen Technologies, in which DNA is first stretched out on a surface and then mapped with restriction enzymes and visualized by microscopy. The new system, which Schwartz and his colleagues described in a recent PNAS paper, is an advance over the old method. It promises “significantly increased throughput through facilitated integration with other forms of DNA analysis,” Schwartz said.
In order to map the DNA, the researchers first treat it with a new class of engineered restriction enzymes. These enzymes bind to specific recognition sites, but instead of cleaving both strands of DNA, they only cut one, leaving a nick behind. A polymerase then adds a fluorescent tag to the nicked sites.
Next, the scientists separate the DNA on a disposable nanofluidic device that contains two types of channels. The DNA first enters microchannels, which are about 100 micrometers wide and 3 micrometers high. Single DNA molecules then thread into rows of nanoslits 1 micrometer wide and 100 nanometers high that straddle parallel microchannels.
Usually, DNA molecules would not stretch out in a nanoslit of this dimension. But when the scientists lowered the salt concentration, the DNA molecules stiffened and elongated as if they were sitting in much narrower channels, “which are very hard to make, and very hard to load,” Schwartz said.
Once single DNA molecules are stretched out in the nanoslits, the researchers read the labels by fluorescent resonance energy transfer, or FRET, using a single laser to image both the backbone of DNA and the fluorescent labels. That way, non-incorporated labels are invisible, Schwartz pointed out.
One advantage of the system over optical mapping is that all labeling steps occur in solution, before the DNA is stretched out. “If you have a choice of doing an operation in a test tube or on a surface, it’s better to do it in a test tube,” Schwartz said.
Traditionally, channels with nanoscale dimensions are made using difficult-to-use and expensive lithography. But Schwartz’s nanofluidic device is made out of a rubber-like material, is easy and inexpensive to make and scale up for high-throughput operation, and is disposable, which prevents contamination from re-use, according to Schwartz.
“If you have barcoded, or mapped, molecules, that compensates a good deal for short reads. Instantly, you know the location within the genome of any sequence read that you are able to obtain.”
Right now, he and his colleagues are optimizing the system for mapping, including the molecule presentation, barcoding schemes, and speed. They are also looking for partners to commercialize the technology.
In addition, Schwartz is working on a way to adapt the system for DNA sequencing. Schwartz would not mention any details, saying that the method is not yet IP-protected.
However, he hinted that the new method would have some resemblance to optical sequencing, which he conceived in 1996, and which his lab published in 2004.
In optical sequencing, researchers map single DNA molecules, create regions of single-stranded DNA near the restriction sites, and perform sequencing-by-synthesis reactions using labeled nucleotides and bleaching between each cycle.
Optical sequencing, in principle, “looks very much like the Helicos system,” Schwartz said, except for the fact that the sequenced molecules are mapped. “We know where they are in the genome,” he said.
But the method currently has its limitations. “We didn’t get a lot of bases, [only] one or two,” Schwartz said. “This is why nobody talks about it. But it was proof of principle.” The new system he is working on “is a lot simpler, and a lot more effective.”