When Stevan Jovanovich talks about his company’s development of a sequencing system designed to achieve seven-fold coverage of the mammalian genome for $100,000, it’s clear that even with this familiar destination, the road he’s describing is certainly less traveled.
Jovanovich is president and co-founder of Microchip Biotechnologies, a company formed in 2003 to address small-volume sample preparation on microchips. At the heart of Microchip’s sequencing device is a microfluidic sample prep system based on Richard Mathies’ work at the University of California, Berkeley. Mathies, a chemistry professor and co-founder of the company, realized early on that optimizing the sample prep process is key to cutting costs. “The problem is sample preparation: it’s hard, it’s unreliable, it takes lots of space and people,” he says.
Enter the NanoBioProcessor, Microchip’s sample prep platform scheduled to hit the market later this year. It hinges on microbeads that capture and concentrate specific targets, which are then shuttled around a multi-layered chip via microfluidic valves, pumps, and routers. “Our approach is to take chemistries and processes that already exist and miniaturize them,” Jovanovich says.
Microchip uses its sample prep technology to power a sequencing platform called the microbead-based integrated DNA sequencer system, which is expected to be commercialized as the NanoBioSequencer in 2008. Jovanovich says the goal is to scale up to 400 channels and achieve seven-fold throughput of 7 million bases per day on a single machine.
The patent- and publication-pending sequencer works by coupling library construction using micro-emulsion PCR on microbeads, each of which contain a single DNA fragment, with thermal cycling. Cycle sequencing is run at nanoliter volumes, as are the sample prep and purification steps. Paired reads are accomplished in the same cycling chamber at the same time, and these are separated on a pair of capture matrices. Finally, extension fragments are injected into a micro-capillary array electrophoresis sequencer for separation of DNA sequencing fragments.
“The niche where this will fit in the sequencing world is getting long paired reads. We have shown 600 base reads and we will push that towards 800 to 1000 base reads,” Jovanovich says. For de novo sequencing, the benefits are obvious. In terms of resequencing, both Jovanovich and Mathies cite the importance of long reads when sorting through the repetitive sequences in cancer genomes.
Microchip has received an NHGRI sequencing grant, and the company has collaborations with Annelise Barron at Northwestern and Jingyue Ju at Columbia. While Mathies is developing new techniques for amplifying and selecting clones, as well as integrating those processes, Barron is contributing novel separation matrices that can be loaded onto the chips for high resolution separation. Ju’s lab tests the resulting prototypes.