THE GENE KINGS
Freud was wrong: The gene, not anatomy, is the closest biology comes to destiny. The 60,000 to 80,000 genes in each of our cells are the blueprint--the operating system, if you will--of humanity. Inscribed in their double helixes of DNA are tales of life and death, sickness and health. By instructing cells to make proteins, they determine whether our eyes are blue or brown, whether or not we can dunk a basketball, and whether we should worry about developing heart disease or breast cancer.
What's turning pharmaceutical executives into gene aficionados is the prospect mf vast commercial treasure in those spiraling strands. A single gene, if it makes a protein that works as a drug, such as Amgen Inc.'s anemia-fighting Epogen, could mean a product worth more than half a billion dollars a year. And that's just the beginning. Tests to spot the faulty pieces of DNA that cause such diseases as cystic fibrosis or colon cancer--a market worth $376 million a year--should balloon into a multibillion-dollar business early next century. Greater knowledge of human genes promises "smarter" drugs and the ability to nip disorders in the bud. Genes "are the raw material for the next wave of therapeutic discoveries," says Lawrence Livermore National Laboratory's Gregory G. Lennon.
Five years ago, science's DNA sleuths--most of them academic researchers--had nabbed less than 5% of all genes. And it appeared that reading the rest would take many years. One man changed all that: J. Craig Venter, an ex-surf bum and obscure National Institutes of Health scientist, perfected a method to rapidly find and sequence genes.
Venter still seeks respect in the snobbish world of science. But his innovations are transforming the pharmaceutical industry. Even skeptics such as Dennis Henner at Genentech Inc. admit Venter's work is "making companies reevaluate all their technology and drug discovery methods." The impact of this genetic information "is going to be enormous," adds Upjohn Co. Distinguished Scientist Jerry Slightom. "It may be the mainstay of drug companies in the future." It is also turning Venter and fellow scientist William A. Haseltine into potential Gene Kings.
The two men have formed one of the oddest alliances in biotech: Venter, 48, heads the nonprofit The Institute for Genomic Research (TIGR) in Gaithersburg, Md. Haseltine, 50, is CEO of Human Genome Sciences Inc. (HGS) in nearby Rockville, Md., which has rights to commercialize TIGR's findings. Together, they have deciphered DNA sequences representing parts of a staggering 85% to 90% of all human genes. The functions of more than half of these remain unknown. Still, they say, their databases contain leads for scores of new drugs. To Haseltine, their discoveries rival that of Balboa, who crested a mountain range in Panama to see--and claim for Spain--a whole new world. "This isn't like oil--there's not more than one gene pool," Haseltine exults. "All people who come later can do is repeat what we've done."
Critics see that as classic hype from Haseltine, who is known as much for his hubris as for his scientific brilliance. But giant drugmakers and startups alike have anted up millions of dollars in a frenzied race to acquire and mine this genetic treasure trove. SmithKline Beecham was the first believer. It committed $125 million to HGS in May, 1993, for a 7% equity stake and first dibs on promising genes. Now, everyone wants to be king. Hoffmann-LaRoche Inc. has invested $70 million in Millennium, a Cambridge (Mass.) startup. Upjohn and Pfizer Inc. have paid millions to look into the data banks of HGS's main rival, InCyte Pharmaceuticals in Palo Alto, Calif. And last September, Merck & Co. funded a major gene-sequencing operation at Washington University.
THE REAL GENETIC JACKPOT. Experts see Merck's approach as a direct attempt to undermine the Gene Kings. The drug Goliath will make all the gene-sequencing data public. To some, it's the biotech equivalent of a computer company giving away its new operating system--and making money on the applications. Merck executives figure that if everyone has the same information, the company's vaunted research and development department can win most of the races to market. "Making drugs from genes is like going from a dictionary to the works of Shakespeare," explains Merck's Alan R. Williamson, vice-president for research strategy worldwide.
The great debate, however, goes beyond HGS vs. Merck, or proprietary vs. "open" gene strategies. The real genetic jackpot may have nothing to do with the sheer number of genes identified and cataloged. That's why, rather than blindly sequencing tens of thousands of unknown pieces of DNA, outfits such as Millennium and Sequana Therapeutics are hunting for genes that cause diseases such as diabetes and obesity.
Even this more targeted approach, critics say, doesn't guarantee success. After all, the world is awash in DNA data, but brilliant new therapies are rare. "Identifying genes is only the beginning of a long, painful, and expensive process of drug development," explains Michael Steinmetz, vice-president for clinical R&D at Hoffman-LaRoche. In the end, says skeptic Stephen G. Pagliuca, managing director of Bain Capital Inc., a Boston consulting firm: "Investing in genomics is like going to Las Vegas."
Venter and Haseltine are unfazed. They've deployed powerful supercomputers to pin down functions of thousands of unknown genes and, with SmithKline, pull out those that could lead to products. HGS is also madly filing patents on everything from gene fragments that help diagnose cancer to proteins with possible therapeutic benefits. Even where HGS and SmithKline opt not to develop a product, the patents may allow them to reap royalties on others' drugs (page 76).
A GOLDEN RECORD. This aggressive patent stance has incensed rivals. "They want to be the gatekeeper of the genome," fumes one pharmaceutical executive. "They think they have everyone else over a barrel." It may also have incited Merck's radical actions. But Venter belittles the threat from the pharmaceutical giant. "They're just validating my approach," he says, "except they're a couple of years late." Even when the Merck project ends in 1996, says HGS's drug-development chief, Michael J. Antonaccio, HGS and TIGR will still have a far thicker "dictionary" of genes. Merck "may use its dictionary to come up with some very good poetry," he says, "but our potential is so much greater."
If HGS lives up to that potential, much of the credit will go to Venter. A champion backstroker in high school, "I was anti-intellectual to an extreme," he recalls. He shunned college to surf at Newport Beach, while working as a night clerk at Sears, Roebuck & Co. As the Vietnam War draft loomed, Venter enlisted in the Navy--with the understanding that he would be on the swim team.
When Venter was in boot camp, President Johnson escalated the war and shut down military sports teams. Fortunately, Venter recounts, he scored highest out of 30,000 on his military intelligence test and got his choice of training. He picked hospital corpsman because it wouldn't extend his three-year enlistment. That left him patching up wounds in a Navy hospital in Da Nang and in a Vietnamese village. "It was a lifetime of education packed into one year," he says.
When he left the Navy, a now academically driven Venter raced through his undergraduate degree and his PhD in biochemistry in six years. He quickly snared a faculty position at the University of California at San Diego and, in 1984, was recruited by the NIH.
If Venter's background seems unusual for a scientist, Haseltine is pure pedigree. A Navy physicist's son, he grew up steeped in science. His life has been a golden record of accomplishment: high honors at the University of California at Berkeley, training with Nobel laureates James D. Watson and Walter Gilbert, a prestigious post at Harvard University, and discovery of several HIV genes.
With that, Haseltine leaped into the limelight. A leading spokesman on AIDS, he lobbied Congress for more funding, warning the disease would spread to heterosexuals. Critics called him a publicity hound. "I believe it's your responsibility to speak out if you see a major health problem, even if that poses a risk to your own career," he replies. He also ruthlessly exposed flaws in others' work while extolling his own. "His personality is a real liability," says friend and Harvard colleague Max Essex. Were his manner more diplomatic, Essex adds, "he could be ruling the world." Haseltine shrugs off the criticism. "Some people interpret enthusiasm as arrogance," he retorts.
But Haseltine also earned respect for his research and his ability to nurture talent in others. "He was a powerful mentor," says former student Alan D'Andrea, an associate professor at Harvard Medical School. "I was always glad I was on the same side of the table."
While Haseltine thrived in high-profile science, it was Venter who took an obscure idea and transformed it into a critical breakthrough. In 1986, he spotted a paper by geneticist Leroy E. Hood that suggested a method for doing DNA-sequencing with robots instead of manually. Having spent months doing it the old-fashioned, tedious way, "I was one of the few people who got excited," Venter recalls. He scrounged $110,000 in funding to get one of the first sequencing machines from Applied Biosystems Inc., developed from Hood's ideas. And Venter was the first scientist to get it to work, he says.
RECAST AS A VILLAIN. It was a time of growing excitement and controversy in biology as scientists were proposing the mammoth 15-year, $3 billion Human Genome Project. The idea was to first map every part of the 23 pairs of human chromosomes, using known genes and other types of markers to place signposts along the chromosomes. Next, scientists planned to sequence all of the DNA in between the signposts. Venter's idea--using the new machine to plunge into the second part of the genome project by reading big chunks of the X chromosome--won support from Watson, the co-discoverer of the structure of DNA, who was heading the NIH's genome project. But review committees repeatedly denied it funding because the technology was deemed too new.
Then Venter had a better idea. Although each human cell harbors some 3 billion individual molecules of DNA, strung together like rungs of an immense ladder, only 3% of them are found in actual genes. These instruct cells to make specific proteins, which in turn, control how the body develops and functions. The rest of the genome is so-called junk DNA, with largely unknown functions.
If he just read the code spelled out by every piece of DNA, Venter figured, the presence of all that junk would make it nearly impossible to find--or understand--the genes themselves. But, as he says, "cells are smarter than scientists." In order to create necessary proteins, cells ignore the junk and copy the DNA of important genes into a related molecule, messenger-RNA (mRNA), which tell the cells' protein what to make. "My insight--or my perception of the obvious--was recognizing that biology works pretty damn well," he says.
To pinpoint the genes, Venter planned to scoop up the easily spotted but fragile mRNA, then make sturdier DNA copies--cDNA--to feed into his sequencing machines. To speed things even more, he read just part of the DNA, since each interesting fragment could later be used to fish out the whole gene. The approach cut the cost of sequencing an unknown gene from an estimated $50,000 using older methods to roughly $20.
It wasn't a new idea, and the gene-hunting elite didn't buy it. They argued that nature was essentially uncooperative: Only a tiny fraction of the genes in a given group of cells are turned on. For example, muscle cells might be making only a couple of key muscle proteins, leaving everything else switched off. They figured the method would snare mRNA from 8% of the genes, at most--many of them uninteresting.
They were wrong. "Venter lucked out," says University of Wisconsin gene-sequencer Frederick R. Blattner. "Nature was not throwing us the terrible curve ball that people feared." Other scientists were spending years to collar one gene; Venter was able to sequence bits of 100 human brain genes in months.
Then, all hell broke loose. Figuring the sequences had value that should be protected, the NIH filed patent applications on 315 gene fragments in June, 1991. It later withdrew all the applications, but not before Venter had been recast as a scientific villain. Part of the flood of criticism was aimed at the idea of patenting gene fragments, but much was a direct assault on Venter and his method. Watson charged at a July, 1991, Senate hearing that Venter's operation "could be run by monkeys." Other luminaries sniffed that Venter hadn't found anything major, such as the Huntington's disease gene, and that what he was doing was uninspired, unsporting, and a threat to the genome project. "I had three strikes against me," Venter recalls. "I had a radical idea, it worked, and I was an outsider."
Even now that cDNA has become mainstream, many scientists belittle Venter's contribution. "He gets credit for inventing this whole technology--that's blatant nonsense," says Hood. "He has never invented anything. The only thing he deserves credit for is scaling up the process." Retorts Venter: "What really pisses people off is once they see it works, they think: `I could have done that."'
But if scientists failed to applaud, the business world took notice. At a time when only a few human genes were known, "Craig offered an approach that might enable you to quickly discover most of the rest," says venture capitalist Wallace H. Steinberg, chairman of HealthCare Investment Corp., which founded TIGR and HGS. "It was not an opportunity that would ever come along again."
By May, 1992, Venter had two $70 million offers: one from Steinberg and one from Amgen. Steinberg was ready to commit only $20 million. But Venter wanted $70 million (later upped to $85 million) for a nonprofit outfit that would let him do the academic work he hoped would boost his standing in the scientific community. Given the tremendous value of the approach, Steinberg says, he had no choice but to ante up. At the same time, he founded HGS to commercialize TIGR's gene sequences and began looking for a CEO to run it.
Enter Haseltine. Even his numerous critics readily admit he is an exceptional scientist. "Bill has a smell for what's important in molecular biology," says NIH virologist Robert C. Gallo, co-discoverer of HIV. In fact, venture capitalist Steinberg had launched several Boston-area companies based on the Harvard virologist's ideas. But he couldn't induce Haseltine to leave the ivory tower until HGS came along. Haseltine took a leave from Harvard to move--with his socialite wife, Giorgio perfume creator Gale Hayman--to Washington.
Meanwhile, Steinberg began approaching drug companies, trying to sell Venter's gene data. Rhone-Poulenc Rorer was one company that declined. "We couldn't handle all that data," explains one insider. SmithKline R&D chief George Poste, on the other hand, recognized "a technology that was absolutely fundamental for our future competitiveness." Despite widespread skepticism in the drug industry, he persuaded his bosses to sign on. "It was a big gamble," says Stanford University biochemist Paul Berg.
ON-SWITCHES. The colossal question now is: how to go from the hundreds of thousands of gene fragments with unknown functions to products. As always, skeptics abound. The DNA code of those fragments offers "no insights into anything," says Hood.
Some scientists, though, see tantalizing clues in the sequences of the cDNA fragments. They have learned to discern key features in snippets of genetic code. For example, one pattern, or motif, of individual DNA molecules usually results in a protein attached to a cell membrane, such as a receptor for outside signals. With an unknown gene, says Berg, "it's possible to make a good guess at the function of the protein it encodes." The guess may sometimes be wrong, so HGS fully sequences scores of its genes, testing the proteins they make.
In addition, academic researchers have been making enormous strides in probing the DNA of simple critters such as yeast and nematodes, whose genes usually have human counterparts. When the machines at HGS spit out a new sequence, scientists comb their databases for telltale motifs or similarities to known genes. SmithKline scientists have used this approach to discover what Poste describes as "a totally novel enzyme" that dissolves bone. He hopes a chemical that inhibits the enzyme may offer a treatment for osteoporosis.
The strategy works in reverse, too. In December, 1993, Johns Hopkins University oncologists Bert Vogelstein and Kenneth W. Kinzler were racing to catch up with rivals at Harvard and the University of Vermont in a hunt for the gene that, when flawed, causes inherited colon cancer. They suspected that the gene normally fixes errors made when cells copy DNA during cell division and had in hand such a gene from yeast. So, in what many see as a vindication of the cDNA approach, they agreed to cede product rights in return for access to HGS's database. In minutes, they found the human version. "It's the way things are going to be done in the future," says Kinzler.
That's especially true for the pharmaceutical industry's lifeblood: drug discovery. In the past, drugmakers tested thousands of chemicals to find one that eased pain or lowered blood pressure. More recently, they have begun to grasp the underlying biology, finding enzymes and biochemical pathways that can be blocked or activated by drugs. But the body typically has substitute pathways, so drugs often don't function as well as expected. With a database of cDNA sequences, "you can start with the full set [of genes and pathways] and choose the best one," says HGS Senior Vice-President for R&D Craig A. Rosen.
cDNA could help solve another major problem: a dearth of novel drug targets. Experts believe a database of all human genes must be laden with clues to previously unknown biochemical pathways that could be manipulated to treat or prevent disease. The trick is finding them. One clever method is comparing cells from different organs, or from normal vs. diseased tissue. When scientists at SmithKline studied stroke in animals, for example, they found more than 20 genes that switched on at the onset of the illness and more than 30 that switched off. Poste expects the key to treating or preventing stroke lies hidden in those 50-plus genes.
Similarly, scientists at HGS and TIGR have found that a whole new batch of genes turn on when normal prostate cells turn cancerous. HGS and SmithKline are now using the genes to make diagnostic tests for spotting signs of malignancy. Moreover, some of the genes could offer new strategies for halting cancer in its tracks. The finding, says Samuel Broder, recently departed director of the National Cancer Institute, "is fabulously interesting."
Venter is also pushing the frontiers of basic science. In late May, he will announce the DNA code for the first two microbes ever fully sequenced, offering crucial insights into the biology of these bacteria. It will also provide a better understanding of human biology and disease, since nature reused most bacterial genes in complex mammals, including humans. The NIH had declined to fund this effort, claiming his cDNA method was no match for the task. But once again, Venter confounded popular wisdom. "No one else thought it could be done," says Nobel laureate and molecular geneticist Hamilton O. Smith of Johns Hopkins.
THE PERILS OF PUBLISHING. Few now doubt that the cDNA approach is transforming science. But who will reap the commercial benefits? Haseltine is determined to stay in the lead--even if it requires a change in strategy. Originally, HGS planned to rely primarily on DNA sequences produced by Venter at TIGR. But by 1993, Venter was turning more to basic science--and HGS faced growing competition from rivals such as InCyte, which were also using the cDNA approach to sequence thousands of genes.
So Haseltine changed HGS's course to downplay TIGR's role. He duplicated Venter's sequencing and computer systems on a grander scale and began churning out 750,000 pieces of DNA code a day, while racing to find and patent the most important new genes. He wants HGS to develop some drugs on its own, help SmithKline make others, and license the rest.
The strategy is risky. For one thing, it has helped create a schism between the two Gene Kings. Venter's unusual deal with HGS gave him the right to publish his findings quickly. In return, HGS got commercial rights to all the genes he discovered. At a cost to his academic standing, Venter initially bowed to pressure from HGS to delay publishing and to focus only on human genes. But now, he's making his data public, with major papers coming out soon. The result is conflict. "Craig feels that HGS basically repeated what he did and stole the glory," says one company insider. "HGS sees Craig's desire to publish as undermining their commercial position. Now, they're handing over $8.5 million a year and not getting much in return."
Internal clashes, however, are the least of Haseltine's worries. HGS's plan puts the bold startup against a host of competitors, especially Merck. With sequences to more than 4,000 gene fragments now flowing each week from the Merck-funded operation at Washington University into public databases, "Merck has really broken the monopoly," says Baylor University gene-sequencer Richard A. Gibbs. What's more, other genome companies, such as Myriad Genetics, Sequana, Mercator, and Millennium, are trying a different tack. They see no point in HGS's approach of blindly sequencing thousands of unknown genes. Instead, they are studying families in which the incidence of diseases such as diabetes or cancer is unusually high. That way, they can snare the underlying genetic mechanisms, then develop drugs for those diseases. HGS may win the race to nail many genes, argues Millennium CEO Mark J. Levin, but his own, more focused strategy will be a faster way of getting drugs to market.
And so, the race is on: the Gene Kings and their partners against the rest of a drug industry, which is straining to catch up. A few skeptics notwithstanding, "nearly everyone is into genomics," says Bear, Stearns & Co. analyst David T. Molowa. "Everyone is a believer." The first wave of products--new diagnostics--may be on the market in a year or two, with potentially important drugs emerging by the end of the decade. By then, it may be clear which strategy will carry the day--and whether Haseltine and Venter have managed to stay on their thrones. Whatever the outcome, their position as pioneers in a medical revolution is already secure.
The Tales of the Genes
1 Each human cell contains some 60,000 to 80,000 genes, which are made of DNA. But only 3% of all DNA is actually genes. The rest is regulatory elements and "junk." So reading the code of each DNA strand, or "sequencing" it, is difficult.
2 Craig Venter takes a shortcut. Cells know which parts of DNA are genes. They copy the genes into messenger-RNA (mRNA). So Venter grabs the mRNA and makes a DNA copy (cDNA), thus snaring only genes, not the junk.
3 Venter then takes thousands of such cDNA pieces, and uses automated machines to read parts of their genetic code. He and HGS now have sequenced parts of
an estimated 85% to 90% of all human genes.
4 New gene sequences go into a giant database, where they can be compared with the known genes of everything from humans to rats and fruit flies. The method also reveals crucial information about when and where a particular gene is active.
Probes for spotting everything from pathogenic bacteria to the various stages of cancer. Left: A probe spots genes turned on only in a tumor.
Newly discovered genes code for proteins that could act as drugs, boosting the immune system, say, or dulling the appetite. Left: Protein binds to cells in the brain's appetite center, turning off hunger.
Many genes code for receptors on cells. Finding drugs that bind to these receptors offers potential therapies for everything from heart disease to Alzheimer's. Left: Drug attaches to a receptor, turning a pathway on or off.
Newly discovered genes could be put into cells to help fight cancer, brain disorders, or other diseases. Regulatory DNA segments ensure that genes work properly. Left: Cystic fibrosis genes go into lung and pancreas cells.
This treasure trove of genes may hold a cornucopia of potential products: By John Carey in Rockville, Md., with Joan O'C. Hamilton in San Francisco, Julia Flynn in London, and Geoffrey Smith in Boston