SEATTLE HEADLINES in the first week of April were dominated by Judge Thomas Penfield Jackson's monopoly judgment against Microsoft. But the news in a brief AP dispatch elsewhere in the papers may turn out to be a lot more significant in the long run. At a scientific conference in Beijing April 3, biotech superstar Leroy Hood announced that a group under his supervision at the University of Washington Department of Molecular Biotechnology had completed a draft map of the genetic sequence of the rice plant, the largest genome so far sequenced and the one with by far the greatest potential economic impact.
A few weeks earlier, an announcement just as significant went completely unregarded in the Seattle papers when the genomics-for-profit behemoth PE Biosystems of Framingham, Massachusetts, declared its intention to license a procedure developed by another UW biotech group associated with Hood, this one led by Ruedi Aebersold.
Hardly a week goes by anymore without some such news, or the publication of an important paper by UW scientists in a scientific journal, or the announcement of a startup company to capitalize on their discoveries. The biotech business has been a growth area hereabouts ever since Robert Nowinski broke away from Fred Hutchinson Cancer Center back in 1981 to set up his own company, Genetic Systems Corp., to exploit the fruits of his publicly funded researches there. In time, biotech could well come to have a greater impact on the local economy—and local society—than the software business has.
The two businesses have a lot more in common than appears on the surface. When biomedical research is mentioned, we still tend to think in terms of people in white lab coats squinting into microscopes. But biotechnology today, particularly in the exploding genomics field, has at least as much to do with constructing algorithms, designing databases, and running search engines as it does with wet work in the lab.
Ever since Jacques Monod and Fran篩s Jacob figured out back in 1960 that genetic information is stored in cells as a linear, "digital" code, cutting-edge cell biology and information science have been on a collision course. Today the subjects are inseparable, as the current buzzword "bioinformatics" suggests, and progress is largely dependent on state-of-the-art computer programming, not to mention state-of-the-art machinery and massive infusions of cash. Sequencing the rice genome took 80-odd machines, a staff of 200, and heavy investment from Monsanto Corporation, and Gregory Mahairas' rice group is just one of the local teams hacking its way through the genetic jungle.
AND SEQUENCING a genome is only the beginning. Raw sequence data from the chromosomes of mice or men is nothing but an endless stream of four code-bits: billions upon billions of bits, repeating in apparently random sequence. In some species, including humans, a lot of the sequence is random, or close to it: "junk DNA," produced and discarded through millions of generations but still cluttering up the chromosomes like the trash in Granny's attic. A genetic sequence isn't much use until someone's gone over it looking for the occasional meaningful sentence lost among the babble of nonsense syllables.
And once you've identified such passages (by sophisticated computer comparisons with known meaningful sequences), you still aren't much closer to putting your knowledge to work. The same set of genes (the meaningful bits among the junk) is present in every single cell of a plant or animal, but only a few genes are active in any particular cell, and a different few in every cell type. Trying to figure out what's wrong with an organism with nothing but the sequence of its genes to
work with is like trying to look up something in an encyclopedia that hasn't been alphabetized.
So, even before the subscience of genomics has produced extensive results (despite all the claims and counterclaims in recent months, we're still a long, long way from possessing the key to the human genome), a new science is emerging, known as "proteomics." Proteins are the primary functional units of every cell. Every gene codes for just one protein. But not all genes are "switched on" at any given time.
So if we could somehow catalog all the proteins present in a cell at any one time, we would also know which genes were currently switched on and producing them. And if we could repeat the cataloging process at various stages in a cell's life cycle often enough, we would have a kind of biochemical freeze-frame motion picture of the particular vital process we're interested in.
The UW's Aebersold has come up with such a technique for sorting and counting proteins—and automating the process, so that machines can reliably perform the hundreds of thousands of measurements required for each "frame." No wonder PE Biosystems wanted a piece of Aebersold's action. PE is the hardware arm of the biotech empire founded by Craig Venter, the scientist-entrepreneur who last month claimed to have sequenced the entire human genome.
Celera, the software-and-sequencing half of Venter's business empire, has attracted close to a billion dollars in investment from venture capitalists and pharmaceutical firms, thanks to Venter's amazing gifts as a self-promoter and his promise to complete the sequencing of the entire human genome years ahead of the nonprofit Human Genome Project. Last month, just two years after starting the job, Venter claimed victory in the race.
Leaders of the official HGP bitterly dispute his claim. But whether Venter can make good on his boast or not, sequencing a genome is a long way from identifying individual genes, and it's an even bigger step from identifying a gene to exploiting the knowledge for medical applications, let alone commercializing them.
So far those who have invested in Celera and other firms like it have essentially been doing so on blind faith; how could anything that sounds so promising not pay off? But a payoff is expected sooner or later, and with the market for high-tech stocks as jittery as it's been recently, the biotech business badly needs some evidence that the payoff will come in the foreseeable future.
PROTEOMICS ARE the kind of boost Celera and the like need. Techniques like those developed by Aebersold's UW group offer the promise of rapidly bringing abstract genomic discoveries out of the database and into the lab—and the marketplace. Cells exhibiting a particular disease syndrome can be analyzed: Once their active proteins and the links between them are identified, it's a relatively short step to figuring out how to intervene in some crucial link in the process.
Best of all, such interference can prove very profitable: Genes themselves may belong only to Mother Nature, but innovative ways of blocking, enhancing, or otherwise interfering with their operation are eminently patentable. And promotable: It doesn't take an advanced degree in biostatistics or database design to understand how proteomic studies can be made to turn a profit.
The promise of proteomics is great enough that there's every reason to expect another major round of wealth creation centered on the people developing the methods and machines that drive its progress. And for better or worse—do we really need more new billionaires?—Seattle and the Eastside look to be the West Coast capital of the action. So long, Silicon Valley; all aboard for Enzyme Junction.