A sticky blob could be the next bottleneck-breaker in genome sequencing
By Aaron J. Sender
Would you like to play with some goo? John Nelson, clad in tie-dyed lab coat and protective goggles, is two parts Bill Nye the Science Guy, one part Dr. Seuss character as he shows off his labs newest enzymatic creation. He proudly presents a 20-microliter drop atop his lab bench. The fluid, chock-full of DNA, is so thick and gummy that he has difficulty drawing it into his pipette.
The gunk is the product of Amersham Pharmacia Biotechs new DNA amplification kit, TempliPhi. Like its predecessor, polymerase chain reaction, TempliPhi relies on primers to initiate replication and enzymes to extend the primers by dragging in free As, Cs, Gs, and Ts to copy the target sequence.
But whereas PCR requires a several-thousand-dollar thermal cycler to take the sample through a series of temperature changes, TempliPhi doesnt. Instead, it employs a technology developed by a New Haven, Conn., startup that imitates the way some viruses copy themselves in infected host cells.
Molecular Staging, a Yale spinout, calls its patented process Rolling Circle Amplification Technology or RCAT, known by insiders as RCA. The technology is poised to take a bite out of the amplification market, which, according to Strategic Directions International, will be generating nearly $1 billion annually by 2003.
Nelson and his mates at Amersham, which holds the exclusive license to develop RCA for nucleic acid research products, say that, especially for genome sequencing labs, their goo could be gold. TempliPhi cuts the 16-hour trip from plasmid to sequencing reaction to as little as four hours. Plasmid prep takes people, it takes machines, it takes time, it has variability, says Nelson. Well eliminate the requirement for you to grow cell cultures, well eliminate boiling minipreps. Well eliminate alkaline lyses. Whatever prep you do to make your plasmid, well eliminate all that work.
True to its name, the RCA polymerase actually rolls in a circle around its target, attaching to it a spool of DNA. When the reaction starts, several sets of primers and polymerases begin copying at different points of the circle. As each polymerase whirls around the template it bumps into the product of another polymerase and forces it to peel away. These displaced strands get primed as well, and in this way a circular strand of genetic material quickly becomes a many-branched tree of double-stranded DNA, each spawning its own copies, resulting in exponentially exponential amplification, Nelson explains. Just a drop of the clear liquid transforms a single DNA plasmid colony into more than enough genetic material for sequencing within a few hours.
In its first incarnation as Templi- Phi, RCA already promises success where PCR has failed. Genome sequencing operations rarely use PCR for sample prep: It tends to introduce errors and can only copy short, well-defined sequence segments. Says Joint Genome Institute R&D manager Paul Richardson, There are a lot of problems with getting a good, clean template out of PCR. You get stuttering and just inconsistent PCR, or just some stuff that wont PCR at all.
Instead, many genomics labs (including JGI) have resorted to a tedious dozen-step labor- and equipment-intensive plasmid prep or miniprep process, which grows enough cells and then isolates and cleans the DNA. As sequencing technologies become faster and better, sample prep remains a bottleneck. TempliPhi could reduce it to a simple two-step process.
In addition, while PCR hits its limit at about 20,000 bases, Nelson believes TempliPhi can generate at least a million bases without falling off the template. It hops on a piece of DNA and were not able to measure how far it can go before it falls off, he says. Thats because it makes DNA so large that it wont enter an agarose gel.
And the reaction runs to completion, exhausting all the free nucleotides so the amount of DNA at the end is predictable, clean, and ready for sequencing.
To be sure, industry observers advise you not to ditch your thermal cyclers and PCR protocol guides just yet. After all, PCR has nearly 20 years on TempliPhi in terms of product development and user mind-share. Every low-level technician knows how to use PCR and what results to expect.
Furthermore, if you want to copy a specific short piece of DNA, PCR is still your best bet. TempliPhi, which is as yet the only commercial use of RCA, copies everything in the tube. And while TempliPhis beta-testers give positive feedback, Brian Taylor, Qiagens genomics marketing manager, says its just too early to tell whether it poses a real threat to the $331 million prep market.
The Amersham goo is new, but RCAs development began nearly a decade ago in Mexico. It was there at the National Autonomous University that biologist Paul Lizardi was working on nucleic acid detection technologies as a Howard Hughes Medical Institute international scholar. In 1994 he received notice from HHMI that it was time to work on something new. He recalls, I started to scratch my head and think, What can I do to get this money again?
When Lizardi saw a Science paper by long-time Lee Hood lab member Ulf Landegren on circular DNA probes, he knew what his next project would be. The minute I saw this paper, I thought, Gee, could one use rolling circle reactions for surface-anchored amplification? Lizardi says. I thought this would be a very nice idea to base my new HHMI application on.
But the funding agency had other plans. It decided to disqualify American citizens living abroad from its award. So Lizardi took his idea to a friend at Yale, David Ward. After a few experiments demonstrated its promise, the university filed a series of patent applications and spun out Molecular Staging to build a business around the technology. Today, Lizardi continues to work on RCA at Yale with funding from Molecular Staging.
For a while, RCA was a technology in search of an application. Because Molecular Stagings business plan does not include creating and marketing its own products, it licensed rolling circle to Amersham to develop and market it as an R&D tool.
After several brainstorming sessions, scientists at the two companies decided that the perfect first application would be templates for DNA sequencing.
Enter 38-year-old polymerase pioneer John Nelson. Nelson did a PhD with David Hinkle, a developer of Sequenase, one of the first mass-marketed DNA sequencing polymerases. And as a postdoc he discovered two of his own enzymes: DNA polymerase zeta and REV1, which both enable cells damaged by events such as ultraviolet radiation to continue to divide. When you sit outside and get a sunburn, my enzymes start kicking in, Nelson says.
The Amersham/Molecular Staging team needed a polymerase that would stay on larger circles long enough, displacing strands and generating multiple tandem repeats. Very early on in this project, I found that all of the DNA polymerases that we were looking at didnt do this very well, says Nelson. But then Phi29, an enzyme that had been sitting around the Amersham lab for several years, became the superhero, he says. Nelson leaps to his feet and draws rapid circles in the air with his finger to illustrate his point: It keeps going and going.
Earlier this year Amersham shipped TempliPhi off to the Sanger Centre, Washington University, the Whitehead Institute, and the Joint Genome Institute for beta testing. Richardson, who evaluates new technologies for adoption at JGI, jumped at the opportunity. That you could essentially turn a nanogram of DNA into a big thick bowl of jelly that you cant pipette was pretty amazing, he says.
JGI has since signed on as a paying customer and converted all its DNA template prep to TempliPhi. Richardson credits TempliPhi with improving his rate of accurate reactions per run from 82 to 93 percent and increasing read length by as much as 90 bases. Thats a huge improvement right there and reason enough to switch, he says.
Even so, labor and real-estate savings were what sold JGI. Before TempliPhi, 20 technicians labored over plasmid prep. Now only seven do. The 3,500 square feet devoted to the process has been cut to 1,100.
Richardson says: I see [RCA] as a very versatile tool. The amount of amplification youre able to achieve in a relatively short time without PCR offers tremendous opportunities for all areas of molecular biology.
Rolling in new directions
Whats been rolled out by Amersham Pharmacia Biotech is product number one, says Stephen Kingsmore, chief operating officer at Molecular Staging. But Molecular Staging is at work modifying RCA and multiple displacement amplification, another Lizardi invention that copies linear DNA, for a variety of applications.
Nelson says the TempliPhi kit was designed as a starting application that we wanted to test, stick our toe in the water, and see what we can do in the field of DNA amplification.
Amersham will soon release an RCA-based SNP-scoring assay called SNiPer and is developing a quantitative version of RCA that will directly compete against quantitative PCR kits such as TaqMan.
And while Amersham holds the exclusive license for R&D uses, Molecular Staging has some RCA-based diagnostics products up its sleeve. (The recent lapse of an option agreement for RCA diagnostics development with Motorola, though, means such products could be further off than was expected.)
Molecular Staging says it will go after areas where PCR is not reliable or not possible. For example, theres never been a powerful amplification method for proteins before, says Kingsmore. By combining RCA with conventional immuno- assays, Molecular Staging claims a thousand-fold increase in sensitivity. On chips, we can measure 50 to 400 proteins at once. The chip can test which of hundreds of antigens a patient is allergic to. And when performed on beads, immunoRCA can detect viral antigens in blood at earlier stages of infection. If you amplify a nucleic acid by PCR, the amplified material comes out into the solution, Kingsmore says. RCA keeps the product localized on a surface.
There are so many, many applications of the technology, Kingsmore says. But for the immediate future, the most likely adaptations to RCA will be in TempliPhi. In the works are kits to amplify larger circles, such as BACs, cosmids, prokaryotic genomes, and eventually even linear genomes. JGI has already successfully used TempliPhi to amplify mitochondrial DNA directly from whole cells and microbial genomes up to 6.5 megabases long.
Whether polymerase chain reaction will give way to rolling circle amplification is still uncertain. If you want to detect a single copy of DNA, just one copy, today PCR is superior to hyperbranched RCA, says Lizardi. But I dont know if that would be the case five years from now. Sure, he notes, PCR could not detect one molecule when it was discovered by Kary Mullis, [and] after 15 years of industrial development it can.
But his previous experience with would-be PCR alternatives has taught Lizardi to be cautious. Even if RCA proves to be better than PCR in many applications, it may be too late, he says. Like most things in life, theres a certain advantage to being first. PCR will prevail because its on the market today. [RCA] would have to be 10 times better, he says. I dont know if it will be.