Sequencing Stakes: Celera Genomics Carves Its Niche

J. Craig Venter is no stranger to contradiction and controversy. He seems to thrive on it. In 1991, when the National Institutes of Health was haggling over patenting expressed sequence tags (ESTs)--a shortcut to identifying protein-encoding genes--Venter the inventor accepted a private offer to found The Institute for Genomic Research (TIGR) in Rockville, Md. TIGR would discover ESTs and give most of them to a commercial sibling, Human Genome Sciences (HGS), to market. ESTs are now a standard way to tackle a cumbersome genome.1 J. Craig Venter In 1994 TIGR asked NIH to fund sequencing of the bacterium Haemophilus influenzae. "NIH trashed the proposal. They thought it couldn't be done, that it was full of holes," recalls Claire Fraser, current president of TIGR and Venter's spouse. Even when Venter announced at a meeting that they'd assembled 90 percent of the genome and requested that their grant be reconsidered, NIH declined. Too bad--H. influenzae was the first organism to have its genome sequenced,2 with support from HGS. Eventually, TIGR and HGS split, largely over HGS's profit motive. Yet in May 1998 Venter boldly announced that he and Perkin-Elmer Corp.'s new company, Celera Genomics, would sequence the human genome faster and more economically than the publicly funded international consortium. Scientists may have been upset, but they weren't surprised. The mainstream media ate it up but couldn't avoid contradiction, branding Venter a maverick while treating him like a deity. USA Weekend magazine's Jan. 29, 1999 cover depicted Venter in the company of Copernicus, Newton, Einstein, and Galileo. Most scientists would blanch at the comparison, but the hallway leading to Venter's office at Celera is lined with that and other magazine covers, making a visitor feel a little like Dorothy at the door of the great and powerful Oz. Now the biotech community is watching carefully as Celera carves out a large niche for itself in the genome stakes. Claire Fraser Sequencing the human genome by Dec. 31, 2001, for a paltry $300 million is only the tip of the iceberg. A look at Celera's production lines of high-throughput sequencers and the massive data center that will house the world's second-largest supercomputer indicates loftier goals. The facility will sequence 115 million bases per day and process 2.5 billion pairwise comparisons of gene pieces per hour, says Paul Gilman, director of policy planning at Celera. "This is not a research lab. ... We crank out information, so there is a lot of redundancy in equipment." And that information is a hot commodity. "We provide the data, and our clients can make the discoveries and claim intellectual property. From 100 to 300 genes per genome will stay here, and we will license them for commercial discovery." Celera's first product, due any time now, will be the fruit fly genome, sequenced with Gerald Rubin's group at the University of California, Berkeley. "If we can do Drosophila melanogaster in a few months for a few million dollars, versus years and several million, that will open up a universe to do other genomes. The ones we're doing now are the 'foundation genomes,'" explains Venter. Following D. melanogaster will be Homo sapiens, then mouse, each one easier to assemble than the last as newly sequenced genomes are overlaid on existing ones. Biotech analysts say that Celera's business strategy to sell sequences is consistent with the increasingly information-based nature of life science. "Selling data is a common practice in the biotechnology arena," says Sue Markland Day, president of the Bay Area Bioscience Center, a nonprofit organization that provides public information on biotechnology. "Early access investors" already signed up for Celera's human sequence data include Novartis Pharma of Basel, Switzerland; Amgen Inc. of Thousand Oaks, Calif., and Pharmacia & Upjohn, in Bridgewater, N.J.--each taking out five-year subscriptions to the human database for $50 million. Venter sees even bigger customers, suggesting that the government subscribe to Celera's rice genome, then make it available to its scientists. And he insists that the company's strategy will lessen competition. "The Celera model is that instead of everyone competing, we [will] be the central clearinghouse and make genome data available to everyone for a tiny fraction of what it would cost them to sequence it." But the company is certainly not the only group generating genome sequences. Venter's vision for Celera is based on the success of New England Biolabs, a Beverly, Mass., company founded in 1976 to supply the restriction enzymes that underlie recombinant DNA technology. "Otherwise, you'd have to purify the enzymes in the lab. To have a business to make them and sell them was considered ludicrous at the time. Now we're all dependent on them to get our restriction enzymes," says Venter. The analogy to New England Biolabs is apt, says Henry Miller, former director of the Office of Biotechnology at the Food and Drug Administration and currently a research fellow at the Hoover Institution of Stanford University. "Celera is OK. They are doing enabling research that the private sector should do. The government should fund early work that the private sector wouldn't otherwise do, not the mindless technique of genome sequencing," he says. David Parker, managing partner at the law firm of Arnold, White, and Durkee in Austin, Texas, and vice president at Introgen Therapeutics, offers a different analogy: a hardware store. "If you want to build a better mousetrap, you go to a hardware store to purchase wood, nuts, bolts, glue, fasteners, the tools that you use to invent. In the Amgen/Celera collaboration, Amgen is in the business of making new pharmaceutical products, and they need the hardware store to buy the tools--in this case, knowledge of the sequences of different genes." Redundancy is a recurring theme in genomics, from shattering multiple genomes to generate sets of overlapping fragments, to installing the equipment to scale up the process. Corporate redundancy exists in the almost incestuous relationship between Celera and TIGR. Venter founded Celera, yet he remains as TIGR's chief scientific officer, chairman of the board of trustees, principal investigator on the project to sequence the plant Arabidopsis, and an author on various papers.3 Both companies will independently sequence the rice genome--Celera will do it through private funding, while TIGR is applying for federal dollars. And once again Venter ruffled feathers, this time with a widely publicized offhand comment to a Japanese official that he could do the job in just six weeks. Japan is a big player in rice. Neither Venter nor Fraser sees a conflict between the two organizations, located about four miles apart in Rockville. "The companies have an agreement to share technological information and planning for now. But we must maintain a separation. If Celera has a database that we consider worth paying for, then we will," says Fraser. Adds Venter, "The work at TIGR and Celera is complementary, but yes, there will be some overlaps." Fraser says she and hubby don't talk business over the dinner table. "We both get home so late, and seldom at the same time, that it just isn't an issue." Still, since some people think "there must be one," Fraser is concerned that a perceived conflict of interest could threaten TIGR's public funding. The conflict that has garnered the most headlines is between Celera and the international public consortium to sequence the human genome. In March, Francis Collins, director of the U.S. National Human Genome Research Institute, and Michael Morgan, executive for genome sequencing of the London-based Wellcome Trust, announced a tentative shift in goals from a 99.99 percent "highly accurate" map with 10-fold redundancy ready by 2003, to a 90 percent accurate, fivefold redundancy "working draft" to be unveiled by spring 2000.4,5 The plan was confirmed at a closed meeting held at Cold Spring Harbor in May. Although the consortium claims its changes are to release more sequence sooner, the aura of competition lurks. "During the past year, NIH has had several revised visions. First was to wait and see. Now NIH is hustling to do a rough draft to compete with Celera's finished project. When the NIH [officials] shifted and said they would do a rough draft, that made it into a race," says Fraser. But it's a race that may prove to be synergistic. "The public effort will provide sequences that Celera, and also companies such as InCyte Pharmaceuticals in Palo Alto, [Calif.], will clean up, repackage, and successfully sell to clients, because the public sector can't afford to do as good a job. They don't have the organization, the manpower, or the motivation to provide a commercial-quality package that large pharmaceutical companies want," maintains Charles Cantor, chief scientific officer of Sequenom, in San Diego. Venter thinks Uncle Sam should spend his money on projects less amenable to commercial muscle, such as basic research on cancer. But he's not averse to taking federal dollars if offered, for example, for sequencing rice. What do the sagas of the ESTs, H. influenzae, and now Celera's promise of a human genome sequence by New Year's Eve 2001 share? "What all these accomplishments reflect is a willingness to take risks. When we learned molecular biology, we took risks. But those risks led Craig to explore automated sequencing and ESTs. And when we left NIH for TIGR, we were leaving permanent government positions," recalls Fraser. Venter adds another secret to his success: fortifying his claims with evidence. "The way that science moves forward is by convincing the critics by coming up with data. And the great beauty of genomics is just that--data!" Ricki Lewis and Barry A. Palevitz are contributing editors for The Scientist. M.D. Adams et al., "Complementary DNA sequencing: expressed sequence tags and the Human Genome Project," Science, 252:1651-6, 1991. R. Fleischmann et al., "Whole genome sequencing and assembly of Haemophilus influenzae," Science, 269:496-512, 1995. K.E. Nelson et al. "Evidence for lateral gene transfer between Archaea and bacteria from genome sequence of Thermotoga maritima," Nature, 399:323-8, May 27, 1999. E. Marshall, "Sequencers endorse plan for a draft in one year," Science, 284:1439-40, May 28, 1999. E. Marshall, "A high-stakes gamble on genome sequencing," Science, 284:1906-9, June 18, 1999. Whereas Celera is still being set up, and Venter speaks of results to come, TIGR's lobby boasts several large posters depicting already sequenced and analyzed genomes, with genes color-coded by function. Upstairs, Fraser shares a roomy office with Cricket and Marley, two black standard poodles. Like many scientists, she diverged at the last minute from the premed trajectory thanks to wonderful mentoring in an undergraduate research project at Rensselaer Polytechnic Institute that led her to fall in love with science. "That was 23 years ago. There was beginning to be a problem with pollutants from pesticides and herbicides in runoff. We took soil and water samples from various sites in Lake George and enriched for bacteria that could grow in the presence of an herbicide. If something we isolated survived, it had bioremediation potential. That is exactly what the Department of Energy (DOE) is interested in today. But now, rather than approaching and then isolating bacteria (in the field) and identifying the species, we look at it through whole-genome analysis," she says. Fraser went on to graduate school at State University of New York, Buffalo, where she worked on neurotransmitter receptors and met her husband-to-be. They were married in 1981. "We went to a meeting for our honeymoon and wrote a grant proposal there," Fraser recalls. The two moved to NIH in 1984, and set about teaching themselves the art and science of gene cloning. "The first year, we hired a postdoc who was straight out of a molecular biology program, and we trained ourselves. We got the Maniatis cloning manual and started on page one," says Fraser. Tiring of manual DNA sequencing, Venter signed on as the first test site for Applied Biosystems' first generation of automated DNA sequencers. And from 1987 until 1990, he accumulated the cDNAs (pieces of DNA reverse-transcribed from messenger RNAs) that would form the basis of EST technology. Fraser continued in her own lab investigating G-protein coupled receptors, then moved to TIGR to search among the stockpiled sequences for receptor signatures. The shift from amassing ESTs to sequencing genomes came from exploiting the overlaps among the cDNAs. TIGR software engineers developed a program called "Assembler" to align the overlaps and derive a sequence. Hamilton Smith, a professor of molecular biology and genetics at Johns Hopkins University who helped develop Assembler, realized that the approach would work for bacterial genomes and suggested they use Haemophilus influenzae, on whose restriction enzymes he had won the Nobel Prize for physiology or medicine. By 1994 TIGR was deep into the sequencing, and Fraser fell in love again. "I became so excited by microbial genomes that I haven't worked on receptors since!" she says. TIGR and Celera are the yin and yang of genomics. Celera is a powerhouse, churning out high-demand "foundation" genomes. TIGR, in contrast, dissects genomes simply because the organisms are interesting. Fraser hopes that this can continue. "Whole-genome analysis so far has been a fairly skewed sample. I worry that it will be difficult to study biodiversity because even though the cost of sequencing has gone down, it is still expensive. NIH focuses on pathogenic organisms, and DOE on those of industrial relevance. So much of the microbial world may remain unexplored." --Ricki Lewis and Barry A. Palevitz

Sequencing Stakes: Celera Genomics Carves Its Niche

Interested in reading more?

Become a Member of