By Kevin Davies
July 1, 2010 | Roche, the manufacturer of the first next-generation sequencing technology to reach the market, originally commercialized by 454 Life Sciences, has backed a promising third-generation sequencing technology under development by IBM.
The two firms will jointly develop a nanopore-based sequencing system that IBM has been developing in house. The collaboration marries IBM’s leadership in IT, microelectronics, and computational biology (see, There is Only One IBM) and Roche’s expertise in medical diagnostics and genome sequencing. Although the technology is unlikely to see commercial release for several years, the combination of an electronic read out and easy sample prep could deliver very cheap sequences.
“That’s the promise: you can do a whole human genome for hundreds of dollars in hours,” says Christopher McLeod, CEO of 454 Life Sciences.
“We’re constantly evaluating new technologies, but this is the first one that really caught our attention,” McLeod told Bio-IT World. “We feel, even though it’s very early and risky, this has the potential to take sequencing to a whole new level of performance and dominate the current technologies. We’re not looking for an incremental improvement but revolutionizing what we and Illumina have already done to Sanger sequencing.”
IBM has been developing a synthetic nanopore-based single-molecule DNA sequencing technology for the past few years. The idea is to ratchet a single-stranded DNA molecule through a nanometer-sized pore in a custom-designed silicon chip. IBM believes it has the capability to ease the molecule through the pore one base at a time.
The project began in 2006 with a serendipitous meeting between systems biologist Gustavo Stolovitzky and Stanislav Polonsky, “a hardware guy” as Stolovitzky refers to him, in one of the lengthy hallways at IBM’s T.J. Watson research center just outside New York City.
Such accidental meetings are not uncommon within IBM, Stolovitzky told Bio-IT World in a recent interview. “In order to go from one place to another, we have very long corridors!” Because of the design of an ‘eccentric Finish architect,” Stolovitzky said, which features vast glass exteriors surrounding windowless interior offices, “we’re forced to go towards the light.”
Polonsky wanted to do something to advance bionanotechnology, and Stolovitzky suggested sequencing. What began as a “skunk works project” soon led to a patent application, a paper, some hires, then some funding from NIH, and further internal support Stolovitzky’s boss, Ajay Royyuru. The project has even been mentioned by IBM CEO Sam Palmisano in a recent keynote speech.
The platform will make use of synthetic nanopores. “One thing we can do at IBM is to create a layered structure of metal and dielectric,” says Stolovitzky, allowing the nanopore to interact with a strand of DNA in very precise ways. “We have electrodes inside the pore. That opens a lot of possibilities to create an electrical field that traps the DNA, similar to a tug of war. You pull with an electrical field to one side, and you pull with another field to the other side, so you reach a balance and the DNA stops. That place is stable.”
Stolovitzky continues: “If you can stop the DNA from going through, like pausing while eating a piece of spaghetti, then you can interrogate a set of bases that are parked in front of a sensor.”
Manipulating the DNA is in some sense the easy part. Stolovitzky acknowledges that he doesn’t yet have a preferred method to sense or read the actual DNA sequence. Two likely contending methods are measuring the capacitance, which is likely to shift due to the changing bases, or by measuring the ionic current, similar to the idea promoted by Oxford Nanopore, except that IBM would not break the DNA.
Speaking to Bio-IT World earlier this year, Stolovitzky said that IBM’s usual strategy was to partner with a major player to the marketing and co-development. Roche says it will fund further development of the technology at IBM, and develop and market all products based on the technology.
454’s McLeod did not recall exactly how the idea of the collaboration began, other than his colleagues and IBM met at several meetings. “It’s been a long process,” he said. “The [IBM] technology has been well publicized. We’ve done diligence on a number of potential systems. We wanted to fully understand what IBM had.”
McLeod says the advantages of the real-time, single-molecule sequencing system is that it will read the sequence of native DNA as it passes through a nanopore. “The attraction is its very fast, inexpensive, that’s why it will take it to a different level.”
The challenge, however, is that “unless there’s a mechanism to slow it down, DNA will go through the pore much too quickly to identify individual bases. IBM has a novel approach on how to modulate the translocation speed with which the DNA passes through the nanopore. It really is fundamentally different than anyone else. The challenge to get there is not only do you need to couple the translocation control with accurate sensing -- there are a number of technologies for how you do that – but what hasn’t happened is to put all the pieces together to show you can actually sequence by measuring electrical variation as DNA transits a nanopore.”
McLeod said that 454 would share some of the developmental work with IBM, and expressed optimism that 454’s expertise in sequencing chemistry will complement IBM’s strengths in semiconductors, computational biology, and to manufacture prototypes. 454 will also be responsible for commercialization and biochemical experts. “We will have an exclusive license for distributing that product, and we’ll be sharing the success of any products in a royalty stream.” McLeod was reluctant to give a timeline, but suggested that a “five-year horizon” was a reasonable target.
Roche currently sells the flagship GS FLX system and earlier this year introduced a benchtop model, the GS Junior system. Nevertheless, 2010 has the makings of a transitional year for Roche/454. The firm’s pyrosequencing platform has been overshadowed to an extent by the release of ultra-high-throughput instruments from Illumina and Life Technologies that widen the gap in throughput between the respective platforms. Meanwhile, 454 founder Jonathan Rothberg unveiled a new benchtop personal genome sequencing machine from Ion Torrent Systems that, while not available until the end of 2010, could impact 454’s market share (see, Electronic DNA Sensing and Ion Torrent Systems).
Roche/454 has also lost two of its key executives in recent months. Tim Harkins, who drove the marketing efforts for 454, left to take a position with Life Technologies. And Michael Egholm, 454’s chief scientist who took over the evangelical role at 454 following Rothberg’s departure, recently left to join the Pall Corporation.
McLeod said that in spite of the high-profile departures, 454 had built a strong organization that is not dependent on one individual. “We’ve built a very strong bench,” he said. Todd Arnold was assuming development activities for the 454 line, while Vinod Makhijani will head the IBM collaboration.
McLeod said it would be wrong to assume that this is the only third-generation sequencing bet that Roche would make. “We continue to invest in 454 sequencing, and we continue to look at what’s out there and we might have other conversations.”
As for the launch of the GS Junior, McLeod said the roll out was going great. “We’re distributing worldwide and we’re right on target. There’s a lot of hype in the market, but we’re quite content to just write the orders.”
McLeod believes his organization has as good a chance as any in developing “a totally revolutionary sequencing technology and getting down to a $100 human genome. Whether or not it happens, time will tell. The question is: Can you take sequencing, which has already declined by four orders of magnitude (price per base), and drive it another couple of orders of magnitude? Certainly this combined team is one that must be considered one of the leading contenders.”