By Julia Karow
With intellectual property from Harvard and Boston University and a proof of concept in hand, startup NobleGen Biosciences is setting out to develop and commercialize a single-molecule sequencing technology that uses arrays of solid-state nanopores and optical detection.
The method comes from the lab of Amit Meller at BU and colleagues at the University of Massachusetts Medical School, who published a feasibility study in Nano Letters earlier this month. In that paper, they demonstrated that they can convert DNA biochemically by replacing each nucleotide with oligonucleotides according to a binary code, hybridize fluorescently labeled probes to the converted DNA, strip these probes off by passing the DNA through a solid-state nanopore, and detect the resulting photon bursts from several nanopores simultaneously. Meller presented the results previously at a meeting in March (IS 3/16/2010).
"This is the first paper in which solid-state nanopores were used to identify the converted DNA code for all four bases," Meller told In Sequence. "Additionally, this is the first paper which demonstrates simultaneous readout from multiple pores fabricated on the same micrometer-scale membrane."
Meller filed his first patents on the technology while working at Harvard, and in 2007, Sequenom exclusively licensed the IP from Harvard (IS 10/2/2007). But in mid-January of this year, the firm, which has recently focused on commercializing diagnostic tests based on other technology platforms, agreed to return the IP to the university, clearing the way for the startup, according to NobleGen CEO Frank Feist.
NobleGen was founded in February by Meller; Ory Zik, an Israeli serial entrepreneur; and Zeev Pearl, an IP lawyer; and recently licensed a patent portfolio from Harvard and BU that includes additional IP from Meller's research at BU.
The company currently has a couple of full-time employees — Feist just hired Phil Buzby, former senior director of intellectual property and of nucleic acid chemistry at Helicos BioSciences, as head of biochemistry, and is close to hiring a head of engineering. Feist joined the company earlier this year after serving as executive director of molecular diagnostic firm Advalytix, which was previously part of Olympus before being acquired by Beckman Coulter last year.
NobleGen is currently in the process of securing funding, primarily looking for strategic investors. Those investors could be sequencing vendors searching for a third-generation platform, microarray companies seeking an entry into the sequencing market, or in vitro diagnostic companies who want to add a sequencing platform to their portfolio, Feist said.
NobleGen has no funding at the moment, but Meller has a four-year grant application for further developing his technology under review at the National Institutes of Health, according to BU. In 2004, he won a two-year, $600,000 grant from the National Human Genome Research Institute's Advanced Sequencing Technology program to develop his method, and in 2006, he was awarded a three-year, $2.2 million grant under the same program (IS 8/8/2006). At that time, Meller was collaborating with Norwegian company LingVitae on the DNA conversion, but Feist said that firm is no longer involved and NobleGen is using a different conversion process now.
Feist did not disclose how much funding the company is looking to raise — he said a larger and a smaller scenario are possible — but the firm's immediate goals are to put together a team, set up a lab in the Boston area, and start development work.
The timeline for developing a sequencing instrument depends on NobleGen's access to funding as well as its business model, but "conceivably, we could be in the market in the 2014 timeframe," Feist said. The aim is to develop a "general-purpose" sequencing instrument, "but we see the ultimate application in IVD."
Modular and Optical
NobleGen's sequencing platform will differ from others in that its biochemistry and readout will be modular and physically separate, similar to capillary electrophoresis sequencing, where the sequencing reactions precede the electrophoretic separation. This, Feist said, offers several advantages. "It makes for much simpler engineering if you can engineer the chemistry separate from the instrument, and also from a business side, it gives us more options," he said. "We believe that to be very attractive, in particular to IVD vendors."
Only the upfront reactions — DNA conversion and probe hybridization — but not the nanopore readout step will require reagents, making the platform "extremely low cost from a reagent consumption perspective," Feist said. Reagent costs will also be proportional to the desired read length.
Both the instrument and reagent costs will be able to "meet 2014+ requirements for overall cost per genome," he said. He declined to speculate on what the cost per human genome will likely be at that time but said that "our cost will be very low."
In terms of time to result, he said that the DNA conversion process will be "fully automated and should be completed in less than 24 hours." The nanopore readout would be "very fast," enabling a throughput of "significantly above" 500 gigabases per hour.
Since several DNA samples can be prepared in parallel, the instrument can be kept busy continuously, he said.
It is too early to tell what the platform's read length will be, which is solely determined by the upfront conversion process that is still at the proof-of-principle stage, Feist said. But the company targets a read length of 1,000 base pairs.
Likewise, it is too soon to tell what the system's sequencing error rate will be, but the firm has reason to believe that it will "very good," Feist said. For example, the DNA conversion process uses ligase, which he said is more accurate than polymerase. Also, an error made during the DNA conversion will not be propagated during subsequent cycles of the process, so there will not be systematic errors from the conversion.
"We are very encouraged by the accuracy and efficiency of the conversion we achieved" so far, Meller said, though he noted that more studies will be needed. "The good part is that the Circular DNA Conversion is a relatively simple biochemical process, and that there are many ways to improve and optimize it. This is currently a major focus for our research."
As Meller recently showed (IS 1/5/2010), the nanopore chip can use small DNA concentrations and still efficiently capture single DNA molecules, so any background noise should be low while throughput can be high, Feist said.
By using an optical rather than an electronic readout, the system can take advantage of high-speed digital cameras that have "already been proven" for other single-molecule applications that involve imaging of several thousand densely fabricated nanopores, Meller said. Building electrodes to individually read out large numbers of nanopores in parallel, on the other hand — something that companies like Oxford Nanopore Technologies will have to do — "is not easy," according to Feist.
Also, Feist claimed, distinguishing all four bases from fluorescent signals is easier than distinguishing them via small differences in small currents as bases pass through nanopores.
And because NobleGen's readout step is enzyme-free, its speed is only limited by how fast the DNA probes can be removed, a process that can "be easily tuned for faster and faster readout as digital sensor technologies improve," Meller said. At this stage, he added, the readout speed is already "substantially faster than third-generation single-molecule DNA sequencing platforms," which use enzymes during readout.
But, according to John Milton, senior vice president of research at Oxford Nanopore Technologies, the throughput of any optical system "is ultimately limited by camera speeds." His company is developing an exonuclease-nanopore sequencing technology with electronic readout.
He also pointed out that sequencing errors could arise during NobleGen's readout, because the fluorescently labeled oligo probes may cross-hybridize with non-complementary sequences or detach before they are being read, and because not all probes may have intact fluorescent labels. In addition, labeled oligo probes are expensive reagents, he said.
Meller is actually a member of Oxford Nanopore's technical advisory board, and the company licenses IP he developed, but it exclusively focuses on electronic rather than optical nanopore detection.
There are several technical challenges NobleGen still needs to master, among them optimizing the DNA conversion process, improving and optimizing the fabrication of nanopore arrays, and improving the DNA reader for speed and throughput, Meller said.
He and his team have already improved the DNA conversion strategy and nanopore materials and system, but he declined to provide details prior to publishing a paper.
NobleGen is aware that it is entering an already crowded field, with other companies developing nanopore-based sequencing technologies, such as NabSys (IS 3/16/2010) and Electronic Biosciences (IS 12/1/2009).
"It's a very competitive space," Feist said. "Sequencing is going to become the molecular analysis platform of the future, taking a significant share of microarrays and even PCR in the foreseeable future. It's a big opportunity, and with all big opportunities, there are going to be a lot of people who would like to take advantage of it."