By Kevin Davies
October 21, 2011 | SAN FRANCISCO -- The genesis of Genia, a promising Silicon Valley nanopore sequencing start-up, took place – not for the first time – with a serendipitous meeting at a popular branch of Starbucks in Menlo Park, California.
Roger Chen was reading a book on the origins of life. Stefan Roever was buying a cappuccino. The two men struck up a conversation. Chen said he was developing a DNA sequencing machine in his garage, to which Roever replied, “What’s that?!”
But it wasn’t long before Roever, a serial entrepreneur, was offering to help assemble a team and write a business plan. He recruited a couple of additional angel investors, helping Genia raise a few hundred thousand dollars over two years, and signed on as CEO.
“If Ion Torrent -- electrical detection but requiring amplification -- and Pacific Biosciences – single-molecule but optical -- are 3rd generation [sequencing technologies], then we're 4th generation -- single molecule, electrical detection,” says Roever. “That's the holy grail, because it combines low-cost instruments with simple sample prep. So we'd like to think of it as last-gen!”
With claims like that, no surprise that Genia has caught the attention of the big guns. In April 2011, Genia closed a strategic investment with Life Technologies that Dietrich Stephan, who sits on Genia’s scientific advisory board, calls “a double digit [millions of dollars] investment.” The investment was not formally announced last spring, but nor is it any great secret, says Stephan. “Life is putting significant additional in-house resources behind the product,” he says.
A Life Technologies spokesperson reserved comment on the investment.
“Genia is an amazing technology,” Stephan told Bio-IT World. “They have developed key proprietary innovations around the pore that allow single molecules of single-stranded DNA to move through the pore slowly, so the sequence can be measured accurately. Key innovations around the array-based sensor allow the pores to be electronically self-assembled into lipid films, and the DNA molecules to be moved bidirectionally under tight control.”
Stephan continued: “The sensor itself is truly transformative and allows very small signals to be seen high above the noise floor, which is one of the issues all the other nanopore companies are struggling with, as well as allowing massively parallel measurements to be made directly on an integrated circuit.”
Roever moved to California in 2000 after successfully building an Internet banking company in his native Germany in the late ‘90s, which he took public and grew to 2,000 people. He has since been involved with various technology start-ups, including a silicon wafer reclamation company, although he readily admits he’s neither a biology guy nor an electrical engineer.
“If we take this to market, at some point, we may be looking for someone with a stronger industry background. But right now, we’re more focused on building a product. Technology development is what I’ve done all my life,” he says.
The name “Genia” has an interesting origin: it was actually one of dozens of unused candidates that Roever received in an earlier naming study for a previous company that cost him $50,000. “I dug out that list and one of them stuck. Sounds like a good name for a sequencing company!”
What Roever lacks in engineering experience is more than compensated for by Chen and the other two co-founders. Indeed, Stephan believes that one of Genia’s key differentiators is that most of the firm’s senior management has roots in the semiconductor industry.
Chen (chief technology officer) and the other co-founders, Pratima Rao (VP marketing) and Randy Davis (VP R&D), are all alums of Maxim Integrated Products, one of the leading analog-to-digital chip companies, who gravitated independently to biochemistry. Chen worked with David Deamer in nanopore sequencing at the University of California Santa Cruz. Davis managed the UCSF cancer center core lab, and Rao previously headed product marketing for Affymetrix.
In addition to Stephan, the scientific advisory board includes George Church (Harvard Medical School), and two of Stephan’s former Navigenics colleagues, Sean George (CEO Locus Development) and computational biologist Eran Halperin. They are joined by Bob Dobkin, founder of Linear Technology, who is also an angel investor.
Following the Life Technologies investment, the company has graduated from Chen’s garage. It is now housed in a Mountain View incubator facility, employing some 15 people.
A number of companies are vying to commercialize nanopore sequencing technology, including Oxford Nanopore in the UK, NABsys in Providence, RI, and NobleGen, which is commercializing a fluorescence detection method in conjunction with nanopores. (Oxford Nanopore has an existing deal with NGS heavyweight Illumina.)
So what does Roever consider is Genia’s advantage?
“At our core, we’re an electronics platform,” says Roever. “We’ve got some developments on biochemistry, but we’re developing a massively parallel analog sensor to do nanopore-based sequencing, supporting a variety of chemistries on the [chip] surface. But at the core, we’re a chip company.”
“Most chips today are purely digital – digital input, digital output, with some processing in the middle,” Roever continues. “A few chips take analog inputs, e.g. temperature sensors or a mobile phone sensor. It’s a bit of a black art.”
Chen previously showed that by using custom electronics -- state-of-the-art analog-to-digital processing with very small currents (picoamp range) and signal (femtoamp range) – it is possible to count “literally hundreds of electrons,” potentially a much finer resolution than competing technologies.
A popular model in a nanopore sequencing scheme is to measure a single molecule of DNA traversing the pore. “The single base difference you’re looking for is effectively about 10 atoms,” says Roever. Those 10 atoms can be detected by measuring the electronic stream. “The signal is in the hundreds of electrons. You’re looking at the very edge of what’s electrically detectable.”
“The published data in nanopore sequencing using off-the-shelf electronics shows a lot of noise, so the resolution isn’t as good as you want,” says Roever. While other companies are engineering genetically modified pores, Genia, by contrast, is building an integrated circuit. “It’s not a chip in the Oxford Nanopore sense or a typical biological sense, which is a typical passive chip with a bunch of leads and some signal processing,” says Roever.
The Genia integrated circuit is essentially a checkerboard array of analog-to-digital sensors or electrodes exposed on the chip surface. Each electrode is a few microns in diameter, potentially enabling as many as 1 million electrodes to be packed onto a chip. The analog-to-digital signal conversion occurs on the chip.
The process sounds deceptively simple: put the DNA in solution above the array of nanopores and sensors. “We can make it all electronically – setting up the bilayer and the pore, plug the chip in, power it on to make the bilayers and pores,” says Roever. Next, apply the potential such that current flows through the pores. “We detect the operating cells and start sucking in the DNA and measuring it. It’s a very scalable platform for molecular analysis – sequencing is just one application. For example, you could put transporter proteins in the bilayer for drug discovery applications.”
Roever draws parallels with a digital camera. Just as each sensor in a digital camera has its own integrated circuit underneath, the Genia platform measures minuscule electronic currents right on the chip. “That gives us two advantages,” says Roever. “First, we can resolve signal much better than if you have to take signal off and process it outside. It’s easier for us to see single base differences. We can react to any changes by moving the DNA back and forth across the pore.”
“Second, we can dynamically build these lipid bilayer/nanopore complexes on the chip surface. Instead of doing that mechanically, we’ve got software to make the bilayer and start the nanopore, so we can easily create this array of sensors with their own bilayers. That allows us to do massively parallel DNA analysis. Instead of a single sensor or a few hundred or a thousand, we have hundreds of thousands or 1 million.”
The knock on nanopores has been the need to hinder the passage of single-stranded DNA through the pore. “We’ve got a way to slow it down, by a combination of electronics and some biochemistry,” says Roever. “We’re not sequencing yet but showing we can do single-base resolution, looking at individual molecules and we can control the movement,” says Roever. The current platform requires less than 2 seconds to read a base, but Roever expects to push that down to less than 1 second/base. “There could be 1 million sensors at 1 base/second,” says Roever.
Roever says Genia can actively control the DNA template, moving it back-and-forth through the nanopore multiple times if required, in a kind of flossing action. “We can oversample, rewind and read again. You change the applied voltage and the DNA goes backward. If you capture the DNA in the pore, you can ‘dental floss’ it – you can read it 10-20 times.” Roever would not detail the read-out mechanism, other than to say, “Our approach relies on some IT to reassemble those sequences.”
By contrast to other nanopore companies that are focusing on alpha-hemolysin or other bacterial nanopore complexes, Roever says Genia is focusing on building “the underlying platform to run in a massively parallel array.” The chips themselves, he says, will be extremely affordable and have a cost structure similar to a standard semiconductor chip today.
Roever says Genia does not have read-length estimates yet, but expects them “to compare favorably to what's out there today.”
The precise choice of nanopore has not yet been settled. “The one thing we know is we have the better control and resolution… We don’t have to get all fancy on the chemistry.” Patents have been filed, says Roever, who adds: “What we have to license will depend on what chemistry we decide to use, what pore we use, what technique we wind up doing... But for the core approach, we think we have freedom to operate.”
Goals and Milestones
One of the next goals for Genia is to build an instrument, which Roever says shouldn’t be any more elaborate than Ion Torrent’s Personal Genome Machine. “Ultimately, it could be a chip in a handheld reader connected to a PC,” he says. A milestone early next year will be receipt of the first working chip, containing hundreds of sensors, from a Taiwanese foundry.
One of the advantages of the chip is that it is amenable to the screening of genetically modified nanopores. “You don’t have to design, you can randomly mutate and throw them at the chip and see what works. You don’t need the best protein engineer, you can just randomly mutate and screen.” That makes it amenable to a variety of potential applications, including drug screening.
“Nanopores are the best way to look at DNA, because the one thing that’s proven to read an individual DNA strand is a protein… All we are is a platform to let nature do its thing,” says Roever. As for other firms commercializing nanopore sequencing, Roever says: “I don’t see Oxford Nanopore as a competitor – there’s no reason their chemistry couldn’t run on our platform, right? If they develop a working chemistry, get a better mutated pore, why couldn’t they run on our platform?”
It is tempting for Roever and his colleagues to dream of the future of medicine when cheap, rapid sequencing is ubiquitous. The Genia website even dares to proclaim “the $100 genome.” Roever checks himself, but will say this: “Within the next decade, you’ll go to the doctor, the nurse will take a saliva sample and sequence everything in there. By the time you see the doctor, he’ll say it’s just a rhinovirus or Epstein-Barr virus or what have you. Instead of the guessing game today, people will look back to this time as medieval medicine!”