Dna: The Secret of Life

by Watson, James

  In principle the patenting of methods could stifle innovation by restricting the application of important technologies, but Stanford handled the matter wisely, and there were no such negative consequences. Cohen and Boyer (and their institutions) were rewarded for their commercially significant contribution, but not at the expense of academic progress. In the first place, the patent ensured that only corporate entities would be charged for use of the technology; academic researchers could use it free of charge. Second, Stanford resisted the temptation to impose a very high licensing fee, which would have prevented all but the wealthiest companies and institutions from using recombinant DNA. For a relatively modest $10,000 a year with a maximum 3 percent royalty on the sales of products based on the technology, the Cohen-Boyer method was available to anyone who wanted to use it. This strategy, good for science, proved to be good for business as well: the patent has contributed some quarter of a billion dollars to the coffers of UCSF and Stanford. And both Boyer and Cohen generously donated part of their shares of the proceeds to their universities.

  It was only a matter of time before organisms genetically altered by technology would themselves be patented. The test case had in fact originated in 1972; it involved a bacterium that had been modified using not recombinant DNA technology but traditional genetic methods. The implications for the biotech business were clear nevertheless: if bacteria modified with conventional techniques were patentable, then those modified by the new recombinant methods would be too.

  In 1972, Ananda Chakrabarty, a research scientist at General Electric, applied for a patent on a Pseudomonas bacteria strain he had developed as an all-in-one oil-slick degrader. Before this, the most efficient way to break down an oil spill was to use a number of different bacteria, each of which degraded a different component of the oil. By combining different plasmids, each coding for a different degradation pathway, he managed to produce a superdegrader strain of Pseudomonas. Chakrabarty's initial patent application was turned down, but after wending its way through the legal system for eight years it was finally granted in 1980, when the Supreme Court ruled five to four in his favor, concluding that "a live, human-made microorganism is patentable subject matter" if, as in this case, it "is the result of human ingenuity and research."

  Despite the clarification supplied by the Chakrabarty case, the early encounters between biotechnology and the law were inevitably messy. The stakes were high and – as we shall see in the case of DNA fingerprinting in chapter 10 – lawyers, juries, and scientists too often speak different languages. By 1983, both Genentech and Genetics Institute had successfully cloned the gene for tissue plasminogen activator (t-PA), which is an important weapon against the blood clots that cause strokes and heart attacks. Genetics Institute did not, however, apply for a patent, deeming the science underlying the cloning of t-PA "obvious" – in other words, unpatentable. Genentech, however, applied for and was granted a patent, on which, by definition, Genetics Institute had infringed.

  The case first came to court in England. The presiding judge, Mr. Justice Whitford, sat behind a large stack of books for much of the trial, appearing to be asleep. The basic question was whether the first party to clone a gene should be granted all subsequent rights over the production and use of the protein. In finding for Genetics Institute and its backers, the drug company Wellcome, Justice Whitford concluded that Genentech could justify a narrow claim for the limited process used by them to clone t-PA but could not justify broad claims for the protein product. Genentech appealed. In England when such esoteric technical cases are appealed they are heard by three specialist judges, who are led through the issues by an independent expert – in this instance, Sydney Brenner. The judges turned down Genentech's appeal, agreeing with Genetics Institute that the "discovery" was indeed obvious, and therefore the Genentech patent was invalid.

  In the United States, such cases are argued in front of a jury. Genentech's lawyers ensured that no member of the jury had a college education. Thus what might be obvious to a scientist or to legal experts trained in science was not obvious to members of that jury. The jury found against Genetics Institute, deeming the broad-based Genentech patent valid. Not, perhaps, American justice's finest hour, but the case did nevertheless establish a precedent: from then on, people applied for patents on their products regardless of whether or not the science was "obvious." In future disputes, all that would matter was who cloned the gene first.

  Good patents, I would suggest, strike a balance: they recognize and reward innovative work and protect it from being ripped off, but they also make new technology available to do the most good. Unfortunately, Stanford's wise example has not been followed in every case of important new DNA methodology. The polymerase chain reaction (PCR), for instance, is an invaluable technique for amplifying small quantities of DNA. Invented in 1983 at the Cetus Corporation, PCR – about which we shall hear more in chapter 7, in connection with the Human Genome Project – quickly became one of the workhorses of academic molecular biology. Its commercial applications, however, have been much more limited. After granting one commercial license to Kodak, Cetus sold PCR for $300 million to the Swiss giant Hoffmann-LaRoche, makers of chemical, pharmaceutical, and medical diagnostic products. Hoffmann-LaRoche in turn decided that, rather than granting further licenses, the way to maximize the return on their investment was to establish a monopoly on PCR-based diagnostic testing. As part of this strategy, it cornered the AIDS testing business. And only as the patent expiration date drew near did the firm grant any licenses for the technology; those granted have generally been to other major diagnostic companies that can afford the commensurably large fees. To create a subsidiary revenue stream from the same patent, Hoffmann-LaRoche has also levied hefty charges on producers of machines that carry out PCR. And so, to market a simple device for schoolchildren to use, the Cold Spring Harbor Dolan DNA Learning Center must pay the company a 15 percent royalty.

  An even more pernicious effect on the productive availability of new technologies has been exerted by lawyers moving aggressively to patent not only new inventions but also the general ideas underpinning them. The patent on a genetically altered mouse created by Phil Leder is a case in point. In the course of their cancer research, Leder's group at Harvard produced a strain of mouse that was particularly prone to developing breast cancer. They did this using established techniques for inserting a genetically engineered cancer gene into a fertilized mouse egg cell. Because the factors inducing cancer in mice may be similar to those at work in humans, this "onco-mouse" was expected to help us understand human cancer. But instead of applying for a patent limited to the specific mouse Leder's team had produced, Harvard's lawyers sought one that covered all cancer-prone transgenic animals – they didn't even draw the line at mice. This umbrella patent was granted in 1988, and so was born the cancerous little rodent dubbed the "Harvard mouse." In fact, because the work in Leder's laboratory was underwritten by Du Pont, the commercial rights resided not with the university but with the chemical giant. The "Harvard mouse" might have been more aptly called the "Du Pont mouse." But whatever its name, the impact of the patent on cancer research has been profound and counterproductive (see Plate 35).

  Companies interested in developing new forms of cancer-prone mice have been put off by the fees demanded by Du Pont, and those keen to use existing cancer mouse strains to screen experimental drugs have likewise curtailed their programs. Du Pont has begun demanding that academic institutions disclose what experiments are being performed using the company's patented onco-mice. This represents an unprecedented, and unacceptable, intrusion of big business into academic laboratories. UCSF, MIT's Whitehead Institute, and Cold Spring Harbor Laboratory, among other research institutions, have refused to cooperate.

  When patents involve "enabling technologies" that are fundamental to carrying out the necessary molecular manipulations, the patent holders can literally hold an entire area of research for ransom. And while every patent application should be trea
ted on its particular merits, there are nevertheless some general rules that should be observed. Patents on methods clearly vital to scientific progress should follow the precedent set by the Cohen-Boyer case: the technology should be generally available (not controlled by a single licensee) and should be reasonably priced. These limitations by no means go against the ethic of free enterprise. If a new method is a genuine step forward, then it will be extensively used and even a modest loyalty will result in substantial revenue. Patents on products, however – drugs, transgenic organisms – should be limited to the specific product created, not the entire range of additional products the new one might suggest.

  Genentech's insulin triumph put biotechnology on the map. A quarter of a century later, genetic engineering with recombinant DNA technology is a routine part of the drug-discovery industry. These procedures permit the production in large quantities of human proteins, which are otherwise difficult to acquire. In many cases, the genetically engineered proteins are safer for therapeutic and diagnostic uses than their predecessors. Extreme short stature, dwarfism, often stems from a lack of human growth hormone (HGH). In 1959, doctors first started treating dwarfism with HGH, which then could be obtained only from the brains of cadavers. The treatment worked fine, but it was later recognized to carry the risk of a terrible infection: patients sometimes developed Creutzfeldt-Jakob disease, a ghastly brain-wasting affliction, similar to so-called mad cow disease. In 1985, the FDA banned the use of HGH derived from cadavers. By happy coincidence, Genentech's recombinant HGH – which carries no risk of infection – was approved for use that same year.

  During the biotech industry's first phase, most companies focused on proteins of known function. Cloned human insulin was bound to succeed; after all, people had already been injecting themselves with some form of insulin for more than fifty years when Genentech introduced its product. Another example was epoetin alpha (EPO), a protein that stimulates the body to produce red blood cells. The target population for EPO is patients undergoing kidney dialysis who suffer from anemia caused by loss of red blood cells. To meet the need for this product, Amgen, based in Southern California, and Genetics Institute both developed a recombinant form of EPO. That EPO was a useful and commercially viable product was a given; the only unknown was which company would come to dominate the market. Despite being trained in the arcane subtleties of physical chemistry, Amgen CEO George Rathmann has adapted well to the rough and tumble of the business world. Competition brings out a decidedly unsubtle side in him: negotiating with him is like wrestling with a large bear whose twinkling eye assures you that it is only mauling you because it is obliged to. Amgen and its backer, Johnson & Johnson, duly won the court battle with Genetics Institute, and EPO is now worth $2 billion a year to Amgen alone. Amgen is accordingly today the biggest player in the biotech stakes, worth some $64 billion.

  After biotech's pioneers had rounded up the "obvious" products, proteins with known physiological function like insulin, t-PA, HGH, and EPO, a second, more speculative phase in the industry got under way. Having run out of surefire winners, companies hungry for further bonanzas began to back possible contenders, even long shots. From knowing that something worked, they went to merely hoping that a potential product would work. Unfortunately, the combination of longer odds, technical challenges, and regulatory hurdles to be cleared before a drug is approved by the FDA has taken its toll on many a bright-eyed biotech start-up.

  The discovery of growth factors – proteins that promote cell proliferation and survival – provoked a proliferation of new biotech companies. Among them, both New York-based Regeneron and Synergen, located in Colorado, hoped to find a treatment for ALS (amyotrophic lateral sclerosis or Lou Gehrig's disease), the awful degenerative affliction of nerve cells. Their idea was fine in principle, but in practice there was simply too little known at the time about how nerve growth factors act for these efforts to be anything more than shots in the dark. Trials on two groups of ALS patients failed, and the disease remains untreatable today. The experiments did, however, reveal an interesting side effect: those taking the drugs lost weight. In a twist that illustrates just how serendipitous the biotech business can be, Regeneron is today developing a modified version of its drug as a weight-loss therapy.

  Another initially speculative enterprise that has seen more than its fair share of dashed commercial hopes is monoclonal antibody (MAb) technology. When they were invented in the mid-1970s at the MRC Laboratory of Molecular Biology at Cambridge University by Cesar Milstein and Georges Köhler, MAbs were hailed as the silver bullets that would quickly change the face of medicine. Nevertheless, in an oversight that would today be unthinkable, the MRC failed to patent them. Silver bullets they proved not to be, but, after decades of disappointment, they are just now coming into their own.

  Antibodies are molecules produced by the immune system to bind to and identify invading organisms. Derived from a single line of antibody-producing cells, MAbs are antibodies programmed to bind to a unique target. They can be readily produced in mice by injecting animals with the target material, inducing an immune response, and culturing the blood cells from the mouse that produced the MAb. Because MAbs can recognize and bind to specific molecules, it was hoped that they could be used with pinpoint accuracy against any number of pernicious intruders – tumor cells, for instance. Such optimism prompted the founding of a slew of MAb-based companies, but they quickly ran into obstacles. Ironically, the most significant of these was the human body's own immune system, which identified the mouse MAbs as foreign and duly destroyed them before they could act on their targets. A variety of methods have since been devised to "humanize" MAbs – to replace as much as possible of the mouse antibodies with human components. And the latest generation of MAbs represents the biggest growth area in biotech today.

  Centocor, based near Philadelphia, now owned by Johnson & Johnson, has developed ReoPro, an MAb specific to a protein on the surface of platelets, which promote the formation of blood clots. By preventing platelets from sticking together, ReoPro reduces the chance of lethal clot formation in patients undergoing angioplasty, for instance. Genentech, never one to lag in the biotech stakes, now markets Herceptin, an MAb that targets certain forms of breast cancer. Immunex in Seattle produces an MAb-based drug called Enbrel, which fights rheumatoid arthritis, a condition associated with the presence of excessive amounts of a particular protein, tumor necrosis factor (TNF), involved in regulating the immune system. Enbrel works by capturing the excess TNF molecules, preventing them from provoking an immune reaction against the tissue in our joints.

  Still other biotech companies are interested in cloning genes whose protein products are potential targets for new pharmaceuticals. Among the most eagerly sought are the genes for proteins usually found on cell surfaces that serve as receptors for neurotransmitters, hormones, and growth factors. It is through such chemical messengers that the human body coordinates the actions of any individual cell with the actions of trillions of others. Drugs developed blindly in the past through trial and error have recently been found to operate by affecting these receptors. And that same new molecular understanding has also explained why so many of these drugs have side effects. Receptors often belong to large families of similar proteins. A drug may indeed effectively target a receptor relevant to the disease in question, but it also may wind up inadvertently targeting similar receptors, thus producing side effects. Intelligent drug design should permit more specific targeting of the receptors so that only the relevant one is blocked. However, as with MAbs, what seems a great idea on paper is too often hard to apply in practice, and even harder to make big bucks from.

  This depressing lesson was learned by SIBIA, a San Diego start-up associated with the Salk Institute. The discovery of membrane receptors for the neurotransmitter nicotinic acid promised a breakthrough treatment for Parkinson disease, but as so often in biotech a good idea was only the beginning of a long scientific process. Ultimately, after giving promising res
ults in monkeys, SIBIA's drug candidate failed in humans.

  Like the unexpected weight loss associated with Regeneron's nerve growth factor, breakthroughs in this area too are often born of pure luck rather than the scientific calculus of rational drug design. In 1991, for instance, a Seattle-based company, ICOS, led by George Rathmann of Amgen fame, was working with a class of enzymes called "phosphodiesterases," which degrade cell-signaling molecules. Their quarry was new drugs to lower blood pressure, but one of their test drugs had a surprising side effect. They had stumbled onto a Viagra-like therapy for erectile dysfunction, which may well yield a bigger jackpot than any they previously dreamed of.*

  *Viagra itself has a similar history. Also originally developed to combat high blood pressure, trials on male medical students convinced researchers that it had other properties.

  The market for easier erections notwithstanding, the search for cancer therapies has, not surprisingly, become the single greatest driving force for the biotech industry. The classic "cell-killing" approach to attacking cancer, using radiation or chemotherapy, invariably also kills healthy normal cells, typically with dreadful side effects. With developing DNA methodologies researchers are finally closing in on drugs that can target only those key proteins – many of them growth factors and their receptors on the cell surface – that promote cancer cell growth and division. Developing a drug that inhibits a desired target without disabling other vital proteins is a formidable challenge even for the best of medicinal chemists. And the uncertain journey from a successfully cloned drug target gene to the widespread availability of an FDA-approved pharmaceutical is a veritable odyssey that seldom takes less than ten years.