London-based DNA Electronics said that it is building a semiconductor sequencing device that will focus on small gene panels, rather than whole-genome or exome sequencing, for point-of-care applications. Key to the platform is the ability to perform all the necessary steps on the chip, a process the company demonstrated in a proof-of-principle study published this week in Nature Methods.
The company said that it would launch its first product, a genotyping device for tailor-made cosmetics targeting the consumer market, this year through the company geneOnyx, to which it licensed nucleic acid testing technology last year.
Eventually, the amplification technique will be applied to targeted sequencing panels for such applications as point-of-care diagnosis of diseases and hospital-acquired infections, as well as companion diagnostic tests, CEO Christofer Toumazou told In Sequence.
DNA Electronics initially licensed its semiconductor sequencing technology non-exclusively to Life Technologies' Ion Torrent, which Ion incorporated into its PGM. Following that deal, it entered into a collaboration with Roche to develop a semiconductor sequencing platform that would compete with Ion Torrent. But that collaboration ended earlier this year when Roche shut down those research and development efforts (IS 4/23/2013). Roche had previously said that the collaboration was a major area of focus and that the two companies had demonstrated that DNA could be sequenced on their semiconductor chips (IS 1/22/2013).
After ending the collaboration, DNAe said it would continue to develop a semiconductor sequencer on its own. But, rather than continuing to focus on a higher-throughput sequencing machine, Greg McGuinness, the company's vice president of business development, told IS that the company is now focusing on developing a point-of-care sequencing platform for gene panels of around 20 megabases to 100 megabases.
"We believe we have a niche there," Toumazou added. "We can sequence gigabases if we want to, but there's no need for that information if you're going for something rapid in a clinical setting."
The company said it anticipates having prototypes available to early-access customers in around one year.
The Nature Methods publication is a first look into the on-chip amplification technology it will incorporate into such a platform.
The goal of the semiconductor sequencing platform will be to perform all steps, including amplification, library preparation, and sequencing on the chip.
The DNAe researchers tested two different methods of amplification — PCR and isothermal amplification. For both, they relied on monitoring pH signatures to confirm that amplification took place. The integrated circuit contains 40 ion-sensitive field effect transistor, or ISFET, sensors, 10 temperature sensors, resistive heating tracks, signal-processing circuitry and control circuitry all embedded in the chip to allow continuous monitoring of pH. It is fabricated using standard CMOS production, similar to what is used for making cell phone and computer chips. The chip includes 120,000 transistors and is mounted on a disposable cartridge the size of a memory card. An analyzer board provides power and an interface to read data, and also performs the genotyping and sends the results to a personal computer for analysis.
The team demonstrated that both the pH-PCR and pH-loop-mediated isothermal amplification, or pH-LAMP, techniques could amplify and detect nucleic acid, testing the platform's ability to distinguish SNP variants of the CYP2C19 gene, a pharmacogenetic gene associated with the ability to metabolize anti-coagulant drugs such as Pfizer's Plavix (clopidogrel).
For the initial demonstration, the team included three flow cell chambers onto the chip — one that served as a reference control, one that included mutant primers targeting the *1 allele, and one that included wild-type primers.
Amplification was detected by changes in voltage, and the researchers found that the changes were comparable to the flourescence signals detected by conventional real-time PCR. Additionally, the 40-cycle pH-PCR process from purified DNA was completed in 35 minutes.
Next, they tested the pH-LAMP technique on the NAT2 gene, also involved in the metabolism of drugs and xenobiotics, and observed amplification after 20 minutes.
They also tested the technique on saliva samples to demonstrate its clinical utility. For this, they designed primer sets for amplification of the wild-type CYP2C19 gene and two allelic variants, CYP2C19*2 and CYP2C19*17. The *2 allele is associated with poor metabolism of Plavix and the *17 allele is associated with ultrametabolism of the drug.
They tested 30 samples, both saliva and purified DNA, of the *2 genotype and 32 saliva samples of the *17 genotype. Results were verified with TaqMan. DNAe's method was able to successfully detect 29 of 30 samples for the CYP2C19*2 analysis, and 30 of 32 samples for the CYP2C19*17 analysis.
Finally, the authors demonstrated that the system can scale by connecting four chips, which allowed them to expand the capacity without redesigning the chip. Additionally, they tested multiplexing by looking at two known biomarkers simultaneously using a 12-chamber flow cell on a quad chip.
Moving forward, the company plans to continue to develop both the pH-PCR and pH-LAMP techniques. There are strengths and weaknesses of both, DNAe COO Sam Reed told IS. With LAMP, there is a "steady temperature and it goes continuously, but the chemistry is much more complex," he said. Whereas with PCR, the chemistry is simpler, but there are thermal cycles.
Toumazou said that the company is agnostic about which amplification method it uses, and said that the "first version will be the one that enables us to meet the application area," he said. For instance, if the application is for a panel that requires a different type of thermal cycling or amplification, then the system would be based on pH-PCR. But, for smaller panels where speed is the main concern, then the company might consider the pH-LAMP approach, he said.
"In terms of the design of the chip, there's not much difference," he said. And, "both of them will work in exactly the same way on the chip."
The goal is "to become the first to market with a point-of-care sequencer," said McGuinness, "combining analysis with sequencing." The company is focused on targeting the Sanger sequencing space. Aside from clinical applications, the company will also focus on the AgBio and forensics markets.
Initial chips will be between 20 megabases to up to 100 megabases, but the company has already built chips internally that go up to 10 gigabases to 20 gigabases, McGuinness said. "We have the capacity for a range of any targeted sequencing application," such as "oncology, pre- and postnatal genetic screening, and transplantation," he said. "But at the moment what we want to focus on is really highly-targeted sequencing."
DNAe declined to provide performance characteristics of its sequencing platform but said that it would be competitive with offerings available today.
Toumazou said that in about one year the company would have early-access customers for its sequencing system with the first commercial sales in around two years. In the meantime, it is continuing to push forward with genotyping applications. Aside from its deal with geneOnyx, DNAe is also collaborating with Eric Topol's group at the Scripps Institute to incorporate the technology into a genotyping test to determine a patient's compatibility with drugs such as Plavix, metformin, and interferon.
The company sees its main competition as the Sanger sequencing market, but may also overlap with gene panels that are run on the PGM or Illumina’s MiSeq. Additionally, a number of other companies are aiming for the lower-throughput targeted sequencing market.
Qiagen is developing a next-gen sequencing workflow and gene panels, initially in the area of oncology, and has said its early-access customers will have systems sometime this year (IS 2/26/2013). Qiagen's system will still initially be higher-throughput than DNAe's system, with its 20 flow cells able to generate up to 2 gigabases each of data.
Nabsys, too, aims to launch a semiconductor-based sequencing system this year, initially for genome mapping with future versions having applications in assembly and validation of larger genomes, structural variation detection, targeted sequencing, and eventually whole-genome sequencing (IS 1/15/2013).