A team of collaborating researchers from Columbia University, Harvard Medical School, Genia, and the National Institute of Standards and Technology has shown proof of principle for a tag-based single-molecule nanopore sequencing approach they have been developing.
The data, which has only emerged over the last few weeks, "shows the system works, but it's not at the level where we would be comfortable calling it sequence data" because metrics like quality scores or read length are not available yet, according to Stefan Roever, Genia's CEO. "Hopefully, over the next few weeks it will be at a point where we can really start calling it sequence data and be comfortable about it."
As previously reported, Genia plans to start a beta testing program with outside researchers by the end of this year and to launch its first commercial sequencing instrument in late 2014 (IS 10/2/2012).
About a year ago, Genia, which has developed a chip-based nanopore detection platform, started to collaborate with the teams at Columbia, led by Jingyue Ju, and NIST, headed by John Kasianowicz, and licensed their so-called NanoTag technology exclusively. Also involved in the project is George Church's group at Harvard, which has been working on coupling polymerase enzymes with protein nanopores.
Earlier this month, the collaborators won $5.25 million in funding over three years under the National Human Genome Research Institute's Advanced DNA Sequencing Technology program to develop their technology (GWDN 9/9/2013).
The Columbia and NIST teams outlined the basic principle of NanoTag sequencing a year ago in a paper in Scientific Reports: A polymerase coupled to a protein nanopore incorporates tagged nucleotides into a DNA template, each type of nucleotide carrying a tag of a different size. In the paper, the researchers used polyethylene glycol molecules of different lengths for that. When a tag is released at the end of the polymerase reaction, it enters the nanopore, where it creates a unique ionic current blocking signal.
In the paper, the researchers showed that a polymerase can incorporate nucleotides labeled with four different tags efficiently and accurately into DNA. Separately, they demonstrated that each tag generates a distinct ion current signal when traveling through an alpha-hemolysin nanopore.
Now, the scientists have put the two steps together: they have attached a single polymerase to a nanopore and shown that it can catch template DNA and start synthesizing DNA using tagged nucleotides, which can be identified by the nanopore.
"We clearly see four distinct nanopore current blockade signatures specific to the four bases," said Ju, a professor of engineering and pharmacology at Columbia. "That's a significant advance, but that's still the first step, to distinguish the four bases."
"When you perform sequencing, you need to do this continuously. You need to further refine the tag, the polymerase, and other elements of the system," Ju said. "I think within a year, we probably should be able to do that."
According to Roever, there are a few key differences between the current approach and the one described in last year's paper: The original idea, he said, was that the polymerase cleaves the tag first, which is then captured and measured in the nanopore. Instead, the tag is now captured in the pore while it's still part of the nucleotide that the polymerase incorporates, and is immobilized there for about 100 milliseconds, enough time to detect it. After that, the polymerase releases the tag, which is then pulled through the pore.
"So we are not trying to detect the moving tag, we are detecting an immobilized static tag, and we have a lot more time to do that," Roever explained. "We also don't have to deal with the issue that the tag may diffuse away."
The modified approach also allows the researchers to operate at salt concentrations low enough to be compatible with the polymerase, which was a point of criticism with the original paper, where the pore measurements were conducted at high salt concentrations.
NanoTag sequencing, although it requires polymerase enzymes and nucleotide analogs, has advantages over strand sequencing, which is the nanopore approach pursued by Oxford Nanopore Technologies, Roever said. "It's much easier to just detect these tags than to try to detect the DNA in the pore, where you have many bases competing with each other," he said. "All of that goes away."
The researchers are working on improving the NanoTag chemistry now to make it more robust. "There are a lot of levers you have to play with by using these synthetic tags, you have a lot of degrees of freedom that you don't have in the traditional strand sequencing approach," Roever said.
"Work is going on to refine the tagged nucleotides, the length of the tags, and the chemical properties such as the charge, and the bulkiness of the tags," said Ju. "The other effort is to engineer the polymerase to produce the optimized parameters for this real-time single-molecule electronic sequencing platform."
While the Columbia team is primarily in charge of designing and manufacturing the tagged nucleotides, as well as some protein engineering, the Harvard group has been working on attaching the polymerase to the pore, and Genia has been changing the kinetic characteristics of the polymerase through mutagenesis, so it operates "at the desired speed," and has been integrating the assay on its nanopore chips, Roever said.
The company and its collaborators have not committed to publicly present data on the technology by a certain time, he said. The proof-of-principle data demonstrates single nucleotide resolution, "but we have not had it long enough to really be able to analyze it and say, 'it comes with this quality and this [read] length.'"
In principle, the platform should be able to measure the kinetics of nucleotide incorporation and use that information to identify DNA modifications, such as methylated cytosines, similar to how Pacific Biosciences' platform detects methylation from kinetic data. However, the team has not started working on that yet, Roever said.
By the end of this year, Genia plans to start a beta testing program with "friendly researchers," and plans to launch a first commercial instrument, which will be for research use only, in late 2014.
The price of that platform, which has not been finalized yet, will be "highly competitive with existing instruments" and "below where anything else is priced," Roever said.
In addition, it will compete with existing sequencing platforms on ease of use, sample preparation, time to answer, and read length, he said.
A second version of the platform for diagnostics, to be launched at a later point, will have an expected instrument price in the range of $1,000, and a cost per run of $100.
Roever said he is not sure whether Genia's nanopore system will beat Oxford Nanopore's to the market. "We are actually hoping that they do come out," he said. "We do have certain features that clearly distinguish us from them, but at the same time, any way that can prove that a nanopore system works is actually a good thing, in our view."