T he backlash against gene patenting is heating up, and not a moment too soon. The U.S. Patent and Trademark Office has already granted more than 1,000 patents on human genes or their fragments, with over 20,000 pending. The patent office plans to issue new guidelines by the end of the year: Researchers will now have to indicate a gene's function--its "specific and substantial credible utility"--and its chemical code to get a patent. (The industry successfully lobbied against applicants' having to show they could actually make something with the gene.) Though the patent office's revised guidelines represent a minor improvement over the previous open-door policy, they won't do much to slow what Francis Collins, the head of the government-run Human Genome Project, ruefully compares to a gold rush.
Collins's analogy is apt. There are an estimated 40,000 to 100,000 human genes. Once they're claimed, they're gone.
The prospectors' vision is too narrow. The ultimate medical benefit of genomic research, however expansive or limited it turns out to be, will depend not on patents but on a farsighted government willing to finance basic, long-term life science. Most of this research will take place over the next 50 years in university or government labs funded by the National Institutes of Health (NIH), not by scientists in private firms.
But many public researchers worry that they will have to secure dozens of licenses from private firms before they can conduct their studies. Many independent scientists are sounding alarms about the impact gene patenting will have on research. The red tape will discourage independent biomedical research or at least slow it way down. Besides, it's going to take decades to unravel the complicated biochemical processes encoded by genes, and it is unclear how useful that information will be. Many credible observers of genetic research argue that the medical hoopla surrounding completion of the Human Genome Project--the decade-long, government-funded effort to map the three billion-plus base pairs of human genes--was mostly biohype. "It is very unlikely that a simple and directly causal link between genes and most common diseases will ever be found," Richard Horton, editor of the British medical journal The Lancet, wrote recently in The New York Review of Books. "This message is not one that many scientists want the public to hear; continued political support for funding genetic research depends on persistent public credulity."
There would be little public support for this privatization of our human inheritance if the industry's arguments were widely known. Patenting a naturally occurring substance in the name of innovation offends the sensibilities, not to mention common sense. No one had to patent the silicon molecule before developing the microchip. No one had to patent the cornea to get a better corrective surgical procedure. Why is patenting a gene necessary to develop a drug or diagnostic test?
The justification for these murky patents rests on the claim that only clear ownership of genes will allow private firms to raise enough capital to turn genetic knowledge into useful products. "Without strong and effective patent protection, development, particularly commercial development, of new treatments, diagnostic tools, pharmaceuticals, and other benefits of our nation's investment in science and technology will not be possible," Chuck Ludlam, vice president for government relations at the Biotechnology Industry Organization, told a congressional hearing last July.
The biotechnology wing of the pharmaceutical industry, which has staked most of the claims, grudgingly accepted the patent office's new rules. But even if the biotech companies license their patented genes to nonprofit researchers for free--which they say they're going to do--the firms will demand first rights to all the breakthroughs made by their licensees. This territoriality will undoubtedly influence the direction of research and will eventually set up dozens of tollbooths on the road to a successful therapy, making the end product unnecessarily expensive to the general public.
Gene Machines
The success of the latest generation of gene-sequencing machines offers one powerful argument for rejecting gene patenting. Developed by publicly funded scientists, the technologies behind the machines have revolutionized genomic analysis. In the 1980s and early 1990s, researchers in what was essentially a cottage industry spent years identifying a particular gene and its function. Their search was driven by a desire to help families with diseases that might have a genetic cause. The classic example was the search for the gene associated with about one in seven breast cancer cases, which was found in the genetic material of families with clusters of the disease. Today, because of the machines, the process works in reverse. Machine-aided identification of a gene can take just a few days, in a process largely divorced from understanding its biological function. During a recent tour of the Institute for Genomic Research (TIGR), a nonprofit sequencing group in Rockville, Maryland, I watched robots handle genetic material while automated sequencing machines that could run 24 hours a day downloaded their findings into banks of computers for analysis.
The heart of the process is the gene sequencer. This technological marvel, the size of a small refrigerator, owes its existence to government-funded scientists. The Department of Energy (DOE)--not the two companies that make the machines--developed the technologies in order to speed completion of the Human Genome Project. "Around 1989 there was a great deal of concern about generating the technology to do the genome [mapping] at a rational cost," said Richard Mathies, a chemistry professor at the University of California at Berkeley, who is responsible for inventing several of the machine's key components. The government not only financed research and development but put together a prototype of the next-generation sequencer at the Lawrence Berkeley National Laboratory.
The gene-sequencing machine has allowed private firms to compete with the publicly funded Human Genome Project and file their gene patent claims. Every firm seeking to privatize genes today uses the machines, which have turned gene discovery into a routine laboratory task. One gene privatizer is even owned by a manufacturer of the gene sequencer. It's as if the manufacturer of the lunar lander had laid claim to the surface of the moon.
Who are the companies using the new machine to engage in a genetic gold rush? One major player, Celera Genomics of Rockville, a subsidiary of PE Corporation, makes a version of the latest gene sequencer. Another major gene patenter, Incyte Pharmaceuticals of Palo Alto, California, has close ties with Amersham Pharmacia, which makes a competing version of the machine. Neither machine would exist without public-sector research specifically aimed at developing core technologies.
The problem facing the gene privatizers is coming up with the "specific and substantial credible utility" needed to qualify for a patent. While no one has done a systematic study of patent office decisions, a few that have come to public attention suggest that many utility specifications are not much more than educated guesswork and have only the most tenuous connection to eventual medical use.
But once companies get past that hurdle, there's no stopping them. Human Genome Sciences, Inc.--run by William Haseltine, one of the most outspoken advocates of gene patenting--earlier this year received a patent on a gene that regulates a docking mechanism on cell surfaces. Its utility was defined as "anti-viral." After the patent application had been filed, NIH-funded scientists discovered that the gene played a role in transmission of the AIDS virus. If the mechanism becomes part of a new therapy for fighting the deadly disease, Human Genome Sciences stands ready to collect a windfall.
"Such practices can have detrimental effects on science and its delivery of health benefits," Harold Varmus, president of Memorial Sloan-Kettering Cancer Center and former head of the NIH, warned Congress. But the growing concern about gene patents hasn't curbed privatization. Incyte Pharmaceuticals already holds roughly 500 of these patents, has over 6,000 pending, and spits out dozens of new applications each week. Celera Genomics has more than 6,500 gene patent applications on file.
A Historic Battle
Celera chief Craig Venter, a former NIH-funded researcher, received widespread attention last June when his three-year-old firm completed its edition of the human genome about the same time as the government-funded project. The media billed Celera's accomplishment as the triumph of a nimble private-sector actor over a slow and plodding government team. It earned Venter a spot next to President Bill Clinton and the NIH's Francis Collins during a Rose Garden ceremony announcing the historic achievement. "The announcement, years ahead of the original schedule, itself demonstrated the value of competition," The Wall Street Journal editorial page crowed the next day.
Well, not exactly. A private corporation had harvested for itself the fruits of publicly funded science. Celera's triumph was a repetition of an old story.
Applied Biosystems has been manufacturing gene-sequencing machines since 1987. Back in the mid-1980s, Leroy Hood and Lloyd Smith at the California Institute of Technology developed the key technology behind its original machine, known as a slab-gel sequencer (the genetic fragments traveled down a gel sandwiched between two long plates of glass). Hood and Smith's patents involved attaching fluorescent dyes to the ends of genetic material so lasers could read its code when the fragments reached the end of the plate. The government is currently investigating the financing of Caltech's original research with an eye to reclaiming some of the money paid for the machines over the years. Hood and Caltech claim the inventions preceded federal grants.
But the outcome of that dispute is of less consequence to the current debate than the inadequacy of the equipment. Slab-gel machines, while a vast improvement over previous technology, remained slow and labor-intensive. While the original Applied Biosystems machines (a mock version appeared in the movie Jurassic Park) allowed scientists to dream of mapping the entire human genome, their technical drawbacks dictated the project's snailish pace. The original schedule developed in the late 1980s called for mapping the three billion-plus base pairs of the human genome by 2005, and even that was considered optimistic.
So in 1989, the Department of Energy launched a series of technology initiatives aimed at revolutionizing the process of sequencing. Over the next decade, the department spent over $1 billion on more than 2,000 proposals from academic scientists all over the world. Their goal was to solve the technological riddles that prevented faster and more accurate gene sequencing. For every three dollars spent by the NIH on sequencing, the DOE would spend a dollar on developing better machines.
Meanwhile, Lloyd Smith (who had left Caltech for the University of Wisconsin in 1987) continued his pioneering work in the instrumentation field. In 1990 he published papers outlining what would become the core concept behind the next generation of machines: the use of hair-width capillaries for transporting the gene fragments. This concept would eventually enable scientists to run 96 samples at a time (up from 48 on the slab-gel machines) and to cut the cycle time to three hours from the previous 12 hours. Barry Karger, a chemist at Northeastern University who was also working under DOE funding, developed the polymer that could carry the microscopic genetic material through the capillaries.
Reading the fluorescent tags in this miniaturized environment was another major challenge. That problem was solved by Berkeley's Richard Mathies, who developed a new set of fluorescent dyes readable by a laser that focused its beam inside the capillaries. Norman Dovichi, a chemist at the University of Alberta, came up with an alternative laser optics system that read the fluorescent tags just after the molecules emerged from the capillaries. Both projects were funded by the DOE.
By early 1996, DOE officials realized they had solved most of the technical puzzles. They asked Joseph Jaklevic, a senior staff scientist at the Lawrence Berkeley lab, to put together a prototype, which his team did within a year. Jaklevic's team used the Dovichi laser method for reading the machine's output.
At the same time, scientists from Mathies' lab were lining up behind a private manufacturer who wanted to use their dyes and laser system. Jingyue Ju, a postdoctoral researcher at Berkeley on a DOE fellowship, was lured to Incyte after generating 11 patents on the reagents used in sequencing. The firm, which had teamed up with Molecular Dynamics, wanted to use Ju's expertise to build a better sequencer. Molecular Dynamics--later part of the pharmaceutical giant Amersham Pharmacia--licensed all the Berkeley patents and developed a machine called the MegaBACE.
In September 1997, Ju stunned the gene-sequencing world at a meeting in Hilton Head, South Carolina. The routine meeting brought together the Human Genome Project's leading players and was sponsored by the nonprofit sequencing group TIGR, run by Craig Venter (he would shortly leave to join Applied Biosystems in starting Celera). The first runs from the new Molecular Dynamics MegaBACE machine were in, and they showed large volumes of good results in very short periods of time. The slab-gel machine was suddenly obsolete. "Applied Biosystems was very nervous to see the MegaBACE producing actual data," recalled Ju, today a researcher at Columbia University.
Applied Biosystems scrambled to catch up. CEO Michael Hunkapiller, who'd worked in Leroy Hood's lab in the early 1980s before starting a firm to make slab-gel machines, delivered a hurried presentation to the group. He suggested that Applied Biosystems was well along in the development of its own capillary machine. Eight months later, the company announced a prototype using Dovichi's method of laser readouts (the company licensed both Dovichi's invention and one from Hitachi, which had independently come up with something similar).
Still, the company didn't ship its first machine until January 1999. Celera immediately became one of its biggest customers. Not only would Celera use the machine to create its own version of the human genome, but its business plan called for generating information databases and gene patents that could then be peddled to the public.
Jaklevic, whose DOE-funded team built the first prototype, doesn't minimize the contributions of the private firms in coming up with the new machines. "There's a lot of engineering that goes into making a machine that is reliable, that you can put in the box with an instructional manual and ship across the country," he said. "But once you know someone has done it, it makes it a lot easier."
hile the patent office cooperates in the looting of the human genome, the biochemists and biophysicists whose work made it possible have joined hands with fellow researchers on the biomedical side of the aisle to oppose gene patenting. "Some people argue that nonpatentability would stifle new products, but I don't buy that argument," said Wisconsin's Lloyd Smith, who has devoted most of the past two decades to developing better sequencing machines. "There are so many patents you can write on formulating the proteins or synthesizing variations that you don't have to have a patent on the gene itself."
Warned Berkeley's Mathies: "If Celera and the 20 companies like them have their database as well as the public database, then everyone in the world who works on the genome will have to subscribe to their database and pay fees. That's a fairly monopolistic situation." ¤
