This week, a team of academic scientists working with Microchip Biotechnologies, a Dublin, Calif.-based startup founded by three Amersham veterans and a UC Berkeley researcher, published details of an integrated Sanger-based micro DNA sequencer they developed that fits into the palm of your hand. The company is one of at least three groups trying to lower the cost of and miniaturize Sanger sequencing.
To be sure, while the actual sequencing performance of the prototype device is comparable to a typical Sanger-based platform, their developers promise it will cut reagent and personnel costs while maintaining the high accuracy and long sequence reads for which the Sanger method is known. By using tiny amounts of starting material, they also plan to eventually abolish clonal libraries, a cumbersome step in sample preparation, and replace them with bead-based amplification.
While the device, which integrates the three steps of Sanger sequencing, is "superb engineering," according to Richard Gibbs, director of the Human Genome Sequencing Center at Baylor College of Medicine in Houston, "when they are at the point of deploying this thing, they have to measure themselves against the current standards. That might be their big challenge," he said. He added that "the [Applied Biosystems] technology is also evolving."
At the moment, the technology has enabled its inventors to create a 556-base sequence read with 99-percent accuracy. So far, none of the rivaling emerging sequencing technologies, like those developed by 454 Life Sciences and Solexa, have published similar read lengths. Such read lengths are required to detect structural genome variations, including repetitive sequences, gene inversions, deletions, and duplications, that play an important role in diseases like cancer, according to Richard Mathies, a professor of biophysical and bioanalytical chemistry at the
"This is one of the reasons why we thought it was important pushing Sanger to its ultimate limit," said Mathies, who is the senior author of the paper, which appears in this week's online early edition of Proceedings of the National Academy of Sciences.
Solexa's chief scientist David Bentley agreed that long reads will have a role to complement short read, low-cost methods such as those developed by his company. He also thinks, though, that researchers can use short read data for de novo assemblies of genomes and to detect copy number polymorphisms.
William Spencer, director of worldwide systems sales of 454 Life Sciences, pointed to whole genome physical mapping, such as that provided OpGen, as an easier alternative to Sanger sequencing for studying large-scale structural changes of a genome.
Novel and Sanger-based technologies will likely compete with and complement each other in core sequencing labs of the future, according to Gibbs. "Most of us are predicting that there will be heterogeneous platforms in the future, depending on which particular applications [are used]," Gibbs told GenomeWeb News. "Some methods will be better suited for some things and some for others, and there will be a blend for yet other problems. Given that, it's very hard without seeing the precise performance of the technology to know how it's going to fit into that complicated picture."
Microchip Biotechnologies CEO Stevan Jovanovich said he wants to commercialize an integrated microfabricated DNA sequencer based on Mathies' developments that includes bead-based template amplification, sample prep, and separation, in 2008 as the last part of three developmental stages. MBI wants to target core labs that could benefit from the hundredfold cost savings the machine could eventually generate Jovanovich said.
First is a sample-prep station that will include thermal cycling and sample clean up, scheduled to enter beta-testing later this year. This product might reduce core facilities' costs on the order of tenfold, according to Jovanovich.
The second product, to enter beta-testing at the end of next year, will add microscale capillary electrophoresis separation to the first device, lowering costs another estimated 10 times. The company has just delivered a standalone prototype of this separation technology to Jingyue Ju, a collaborator who will test the instrument in his
Finally, MBI wants to add bead-based template amplification by the end of 2008. "That's when it is all scheduled to come together," Jovanovich said.
The microdevice described in the PNAS article integrates the three steps of Sanger sequencing, but does so on a microfabricated disk 10 cm wide: It performs thermal cycling to create the dye-labeled fragments, purifies the samples, and separates the fragments by capillary electrophoresis. Using just a femtomole of DNA template in 250 nanoliters of reaction volume -- about tenfold less than what is typically used at sequencing centers, according to Mathies, who is a co-founder and director of MBI -- the scientists created a 556-base-long sequence with 99-percent accuracy.
By comparison, a study by 454 Life Sciences last year showed read lengths of 110 bases; an article published by George Church's group on polony sequencing last year reported 26-base reads; and Solexa said it aims for 35- to 50-base reads. 454 has said that a prototype instrument achieves up to 400-base reads.
With the microdevice, the thermocycling reaction efficiency is comparable to conventional Sanger sequencing, and close to 100 percent of the product is purified on the chip, Mathies said. Still, the researchers said they could improve how much of the sample is injected into the capillary. "That's the most inefficient stage," Mathies said.
The performance of the microdevice "basically sets a benchmark [and] tells us what is possible in that format," he said. "From that we are able to extrapolate what the ultimate limits are."
Providing what he called a conservative estimate, Mathies said he might be able to lower the template and reaction volume another tenfold, which would translate to an 800-fold reduction in template and a 400-fold reduction in reagents, compared to what is used at Sanger-based sequencing centers today.
Eventually, with improved scanner sensitivity, researchers might get away with using as little as 10 attomoles in their DNA template, he said. "The background of this, of course, is that the cost of sequencing consists heavily on the cost of the reagents," he said. Also, since the instrument integrates all the steps of sequencing, personnel could be reduced, he added, another factor in cost reduction that is not easy to quantify at this stage.
Another improvement in conventional Sanger sequencing will become possible once the amount of template enters the attomole range: being able to get rid of clonal libraries and instead using PCR colonies, or polonies, the product of single molecule PCR reactions.
"That's revolutionary," according to Mathies. It's "the part [of the Sanger process] that has never been changed," because only cloning was able to provide the amounts of DNA needed for sequencing. "Everybody is still doing it the same old way."
Mathies' lab is currently working on a method for sample preparation that involves PCR-colony beads. The eventual goal is to build a device that integrates the entire Sanger sequencing process, including library production. The PNAS study explored the back end of this, namely the feasibility to sequence from a small amount of material. "My strategy is always a bottom-up strategy: If you can't sequence it, there is no point in making [the template]," he said.
MBI, was founded in 2003 by Mathies and Amersham alumni Jovanovich, Roger McIntosh, and Dennis Harris. Harris is currently chief scientific officer of Serologicals. MBI has been funded by several government contracts: In October 2004, the National Human Genome Research Institute gave it a three-year, $6.1-million grant to developing its DNA sequencer with Mathies, Ju, and Anneliese Barron at Northwestern University. That month, the company also obtained a $70,000 Phase I SBIR grant from the department of the army to develop a pathogen bioprocessor.
Last June, MBI won a $100,000 SBIR grant form the Homeland Security Advanced Research Projects Agency to explore a fluidic sample bioprocessor, and earlier this month the Department of Defense gave MBI a one-year, $365,000 Phase II SBIR grant that is renewable for another year to continue developing its biothreat detection device.
The company is also currently raising private capital but would not disclose how much or when it is hoping to achieve this. "We are close," Jovanovich said.
MBI is not the only group trying to improve and miniaturize Sanger sequencing: Network Biosystems of Woburn, Mass., received a $4.5-million NHGRI last year to develop a microfabricated Sanger sequencer. Also, Vera Gorfinkel, a researcher at SUNY Stony Brook, last year received a $1.5-million grant from NHGRI to develop two-dimensional monolith multi-capillary arrays using nanoliter reaction volumes.
Julia Karow covers the next-generation genome-sequencing market for GenomeWeb News. E-mail her at [email protected].