October 6, 2011 | DNA Electronics certainly sounds like the quintessential bio-IT company, but this London firm is quietly making waves in the field of next-generation sequencing (NGS) and molecular diagnostics. Some of the firm’s key intellectual property (IP) is providing the foundation for present and future semiconductor sequencing platforms. And some of that same technology lies at the core of the firm’s newly developed handheld device, dubbed the SNP-DR, which can detect dozens of DNA variants in a saliva sample within 20-30 minutes.
The founder and CEO of DNA Electronics is Chris Toumazou, who also holds the chair in Biomedical Circuits in the Department of Electrical and Electronic Engineering at Imperial College in London.
As a postdoc at Imperial, he was interested in consumer electronics, particularly multiband digital mobile phones at a time when GSM was becoming a mobile phone standard. Toumazou viewed “low power technology” as the only way to integrate those standards onto a silicon chip, and started patenting innovative semiconductor techniques to integrate CMOS (complementary metal oxide semiconductor) chips.
Toumazou’s honeymoon in the medical area began in the late 1990s, when he was approached by a Canadian company, Epic Biosonics, which manufactured electrodes that made contact with the 8th nerve in the ear for cochlear implants. “It used hardly any power because it had an excellent low power electrode, but cosmetically, it looked awful, a whacking great processor hanging out of the kid’s ear,” recalls Toumazou.
“I looked at the billions we’d invested in making low power chips for communication and thought, if we can just apply a fraction of that technology to the health care problem, we could make a difference.”
Within a year, Toumazou had helped turn that huge chip into something using nanowatts of power, integrated onto an electrode array. It was his first “bio-inspired platform,” a case where the biology inspired the technology. (The resulting product is now the core of a number of cochlea prosthetics.)
But while Toumazou was excited about the application of semiconductors in health care, he sensed an even bigger opportunity. “It’s not just making sick people better, but healthy people healthier. I wanted to establish a more personal approach to medicine.”
Toumazou didn’t have to look far to realize his ambitions. Imperial College has the largest medical school in Europe, providing a perfect setting to explore health care applications for innovative semiconductor design. He found an ally in the rector, Sir Richard Sykes, the former head of Glaxo, who also saw the synergies between silicon devices and the pharma industry. “He was passionate about the next generation of pharma products that could be medical devices,” says Toumazou. After raising more than £30 million, Imperial created the Institute for Biomedical Engineering.
But Toumazou was also exploring commercialization opportunities. His entrepreneurial skills were honed from experiences developing the world’s first combined analog/digital mobile phone in Thailand with now exiled premier, Thaksin Shinawatra, which produced lots of low power technology IP.
Toumazou founded Toumaz Technology (now publicly traded), which produced a digital patch—a band-aid device that sticks on a patient’s chest to monitor vital signs. It sends information wirelessly to a mobile device, and was recently approved by the FDA. “If it detects something, it becomes preventative medicine,” says Toumazou. Several large hospitals are testing the device, and they have recently established a joint venture with Patrick Soon-Shiong, founder of Abraxis BioScience.
Toumazou’s interest in using semiconductors for biochemical monitoring of substances such as urea or creatinine crystallized during a family emergency. About seven years ago, his teenage son Marcus collapsed with renal failure, due to a previously-undiagnosed genetic disorder. Following a kidney transplant, Marcus was on dialysis for more than three years, teaching Toumazou a lesson in the pains of managing chronic disease.
As he explored technologies that could help a patient like his son monitor and manage his lifestyle, Toumazou learned about ISFETs (ion sensitive field-effect transistors)—discrete off-the-shelf devices which can be made in CMOS. Of particular interest, chemists were using ISFETs as pH probes. “I got hold of one and thought, this is more than a pH electrode; it’s got the guts of a semiconductor. But instead of an electrical gate on top where you inject current or apply voltage, it had chemistry—an insulator surface exposed to an electrolyte.”
Toumazou decided to build some very low power biosensors for measuring glucose, urea and other metabolites, deploying different chemistries on the surface of the device. “We were trying to create a biosensor, while integrating that into the electrical circuits we’d been looking at in our traditional mobile phone technology,” says Toumazou.
Despite his excitement, his students soon encountered a problem: over a few days, the recorded levels of glucose and other metabolites in blood samples were drifting, presumably because of surface chemistry issues.
Around this time, Toumazou was introduced to Craig Venter’s work (they later met through their mutual investor, Malaysian businessman, Tan Sri Kok Thay Lim). Talk turned to DNA, which wouldn’t drift. Toumazou thought, why not take an ISFET and instead of applying glucose, try something discrete like DNA?
In a 2001 experiment, one of Toumazou’s clinical students put DNA polymerase on the ISFET, and added the nucleotide adenine [to a probe-target pair]. “We found that when you match the bases, protons get released. Let’s see if we get incorporation and the change in pH. That was our Eureka moment! Whoosh, there was a whacking change in pH. The transistor turned on in real time. I went back to my office, wrote three paragraphs and sent it to the patent office.”
Toumazou’s team dabbled with various forms of DNA sequencing, but coming from the mobile phone sector, he was much more interested in point-of-care—find an application and get an on-the-spot result. “It lent itself easily to the digital logic of DNA sequencing, but it wasn’t in the remit of what I wanted to do,” says Toumazou (see, “Sequencing Roadmap”). “I started thinking, rather than use it as a discoverer of mutations, let’s use it as a detector of mutations. Then it could be done with a very simple system.”
In late 2003, Toumazou spun out a company called DNA Electronics. At that stage, most detection technologies were optical rather than label-free, and would be difficult to integrate with hydrogen ion-based amplification, especially in a handheld device. But in a semiconductor detector, Toumazou could seamlessly combine sample prep + amplification + detection.
One of his first hires was Leila Shepherd, a Ph.D. student who was studying the synergy of ISFETs and biochemistry. The core idea was to place a known single nucleotide polymorphism (SNP) on a probe such that the chip could achieve a fast ‘digital’ result—a 1 if there’s a match, a 0 if there isn’t. Such SNPs could provide information on a patient’s drug metabolism or an infectious disease agent.
The Genalysis device, or ‘SNP-DR’ as it is nicknamed, works directly from a saliva sample. The first step is to amplify the DNA. “It was blatantly obvious to us that the output of the extension reaction (after thermal cycling) was pH,” he says. If he could simultaneously use the pH shift that occurs during DNA amplification as the pH detection input, then he might be able to simultaneously amplify and detect DNA on a single chip.
DNA Electronics’ first point-of-care chip, which was successfully tested by Pfizer last year, is “probably the most sophisticated complex chip combining DNA detection and mixed CMOS technology,” says Toumazou. “Pfizer ran some genotypes on real patient samples and sent them to us blindly,” says Shepherd. “We ran on the device, and got 100% call rate success.” Trials are ongoing at St Mary’s Hospital in London.
The CMOS chip that runs the SNP-DR features 40 sensors, thermal devices (for PCR), and all the necessary analog/digital circuitry. The PCR has been customized to maximize the pH change while keeping the enzymes happy. “There’s lots of on-chip intelligence,” adds Shepherd. “It has 40 sensors but only 9 wires on the chip. And there’s no heat block, no dye, no optics.”
“You need all this intelligence,” explains Toumazou, in a strategy that harkens back to the cochlea implant. “We want to avoid too much bioinformatics on the back end—we want to do things on the chip so the intelligence is local and you spit out the results.”
“This is where our heart is—a totally lab-free instrument,” says Shepherd. “Take saliva, mix it with chemicals, dispense onto a test cartridge housing the chip. The silicon chip has all the magic inside—sensors, heaters and temperature control. Once we apply the saliva sample, it’ll perform real-time DNA amplification and detection. If you’ve got a matching SNP in the saliva sample, it’ll give you the answer in 15-30 minutes.”
One Chip One Drug
The immediate target is for low-multiplex applications, such as testing warfarin and other drug sensitivities, but eventually there will be a library of different application-specific cartridges.
“Our vision is, one chip one drug, one chip one bug!” says Toumazou. “We’re not in a dark room—we’re out in the real world with instantaneous answers. It’s because we can integrate the PCR and detection. That’s fundamentally the key.”
The key is to compress the data, minimize extraneous information, and end up with the essential results. “Spending all the power on the analysis, it really does take days to get the results and you lose sight of what you’re trying to do,” says Toumazou. “Medics don’t want to be piled with lots of information, they just want results. We’re not getting rid of the bioinformatics, but the more intelligence you can put on the chip or the process, the faster you get the results.”
DNA Electronics employs about 35 people, with expertise in semiconductors, biochemistry, chemistry, and fluidics. Toumazou intends to translate any form of nucleic acid analysis into a pH signal. As he says, “There are plenty more strands to the bow.” •
A few years ago, while DNA Electronics was working on their point-of-care chip, Chris Toumazou hosted a visit from Jonathan Rothberg, the founder of 454 Life Sciences, in London. Not surprisingly, the two entrepreneurs hit it off, approaching NGS from complementary angles: Rothberg from the chemistry side, Toumazou from the semiconductor field.
After a good couple of years of negotiations, Rothberg licensed Toumazou’s technology. Rothberg wanted an exclusive deal, but Toumazou held firm. “I wanted to create the fabless semiconductor industry for the genetics wave for life sciences,” says Toumazou. “Jonathan really wanted exclusivity, but we weren’t going to give exclusivity.”
Toumazou favored the model utilized by ARM Holdings, maker of fabless digital processors found inside iPhones, Androids and tablets that are “agnostic to the silicon,” says Toumazou. The platform is licensed to everybody in a non-exclusive way. “We’re an ARM,” says DNAe’s Leila Shepherd. “We develop the IP cores.”
“Rothberg was the right person to take sequencing forward,” says Toumazou. “He had the array/chemistry experience from 454 and we were happy to work with him. But I wanted to make it non-exclusive so we could talk to the 454’s, the Illumina’s and the Life Tech’s if we wanted to.” Toumazou has also licensed technology and struck a partnership with Rothberg’s former colleagues at Roche/454, and is collaborating on the development of a new semiconductor product. “We have a level of sophistication of the semiconductor chemistry/interface layer that we think will give us a performance edge,” says Toumazou.
“We’ve got a roadmap for semiconductor sequencing that can match 4th-generation sequencing. Nanotech to me is top-down. We can use processors now that go down to 25-40 nm. This is the natural progression. It’s not the semiconductor that will be the limiting factor, it’ll be the interface. Semiconductor technology offers a really good platform to do those things.” K.D.
This article also appeared in the 2011 September-October issue of Bio-IT World magazine.