What needed to happen was to scale not up but down. “We needed to miniaturize sequencing.” Over the next two weeks—the time allotted for his paternity leave—Rothberg hashed out a new design based on cloning by limiting dilution and a solid-phase version of pyrosequencing, a technique that at time was used only for calling SNPs. The result was the massively parallel sequencing-by-synthesis strategy that was the foundation for 454 Life Sciences.
Church also realized the key to cheaper sequencing was miniaturization, and had spent much of the previous two decades testing and developing strategies to get there. One method called multiplex sequencing, which was developed in 1988, featured several characteristics of modern next-generation sequencing. DNA sequences were immobilized on a flat surface and then subjected to many alternating cycles of probing and capturing the data using a digital camera. Church had also developed a strategy in which tiny pools of identical molecules are created in situ by a polymerase—so-called polymerase-generated colonies, or “polonies.”
“I try to familiarize myself with many different technologies,” Church says, describing himself as “constantly on the prowl for combinations that I could match up with sequencing human genomes, because clearly that cost was off by a million at least.”
In 2005, Church and his team rolled his various ideas into a new sequencing-by-ligation strategy that they published in the journal Science (1). That same week, Rothberg's team at 454 described their strategy in Nature (2).
According to Rothberg, it takes more than good sequencing technology to produce a successful sequencing strategy. There's also the ability to eliminate the DNA cloning steps that remain so critical in Sanger-based approaches. “You really have to crack both problems to make it into a successful technology,” he says. “Sequencing was half the problem, sample prep was the other.”
Both groups solved the problem the same way: by making libraries on-the-fly on the surface of micron-sized beads, arraying and immobilizing them on a planar surface, and applying cycles of sequencing chemistry in situ. Then, instead of running gels or capillaries, they imaged the array over and over again, using computers to sort out the data. It's a strategy that is now used in broad strokes by every major sequencing vendor on the market.
Since Church and Rothberg's 2005 publications, next-generation DNA sequencing methods have largely eclipsed Sanger sequencing, and genome-sequencing costs have dropped to nearly $1,000. Church's ligation-based approach has been used or adapted by at least four commercial sequencing firms, including Life Technologies and Complete Genomics, arguably the biggest sequencer of human genomes to date. Roche Diagnostics bought 454, and Rothberg went on to found Ion Torrent, whose semiconductor-based method the Broad Institute's Chad Nusbaum has called “the first post-light” sequencing technology.
For those who have been in the thick of sequencing technology for its 30-plus year history, there is today perhaps a sense of déjà vu. Back in the 1970s, it was Sanger versus Maxam-Gilbert. Today, fundamentally different approaches are competing in a crowded market— optical versus non-optical, sequencing-by-synthesis vs sequencing-by-ligation, with nanopores and other strategies on the horizon. Like VHS versus Betamax, there's a period in technology development when competing technologies can co-exist, but inevitably, the balance shifts. Only time will tell where it will land.
Making PCR Quantitative
1983 was also a banner year for Kent Vrana.
Vrana, the Elliot S. Vesell Professor and Chair of Pharmacology at Penn State College of Medicine in Hershey, Pennsylvania (and a member of the BioTechniques editorial board), earned his PhD that year. His thesis was biochemical, but in 1983, he says, “molecular biology was just breaking,” so he took a postdoc to learn the new technology.
Moving on to West Virginia University, he started his own lab studying transcript abundance. At the time, Vrana says, that meant Northern blotting. But the technique wasn't easy. “Very labor intensive, very difficult to quantify, [and] we were using a lot of radioactivity to do this.”
PCR, Vrana says, had an immediate impact. “It cut the time to analyze RNA down by probably a factor of five.”