Patent maps on human genes

DRUG FIRMS DISCUSS LINKING UP TO PURSUE DISEASE-CAUSING GENES

By Robert Langreth, Michael Waldholz and Stephen D. Moore Staff Reporters, The Wall Street Journal, 4 March 1999

Major drug companies are scrambling to form an unusual collaboration to hasten the hunt for genes involved in intractable diseases.

The aim is a pathbreaking joint venture that would have the highly competitive companies, which normally share nothing involved in basic science, working together on the biological blueprint for all human life. Its ultimate goal: to spur the creation of a new era of "personalized medicine," in which treatment could be tailored to an individual's unique genetic profile.

Though no agreement has been reached yet, the companies are racing the clock to find one. The reason for their urgency: They are worried that small biotechnology companies, rushing on their own to decode the human genome, will patent their discoveries and monopolize crucial gene information -- and then be able to either keep it private or charge drug companies huge fees for access to it.

If the discussions succeed, the drug makers will form an unusual nonprofit entity whose goal would be to identify tens of thousands of chemical landmarks found throughout human DNA. These landmarks, in turn, would be used to create a catalog of the ever-so-slight genetic variations from one person to another that make some individuals susceptible to disease, such as lung cancer or asthma.

In recent years, scientists have discovered that these tiny genetic differences that reside inside or near genes are sprinkled at regular intervals along the vast DNA molecule, like road signs and mile markers on a long stretch of highway. Many drug-company scientists now believe a jointly funded venture would have a high likelihood of generating a map of all the landmarks, in perhaps as short a time as two years.

Scientists would use the map to sift rapidly through the genetic material from thousands of patients to uncover which of the 100,000 or so genes that make up human DNA (the majority of which remain unknown) predispose people to such common but hard-to-treat ills as diabetes, depression, cancer, arthritis, memory loss and cardiovascular problems. These all are maladies whose underlying biological causes remain largely mysterious, but if uncovered could lead to a treasure trove of new medicines.

A complete record of the minute differences would form "the most powerful map in human genetics to date," says Alan Williamson, a retired Merck & Co. research executive who has helped organize the talks. "Suddenly, there will be a genetic map powerful enough to define which patients respond to a given drug, vs. which don't respond to a given drug," he says. "It will allow

doctors to tailor treatments to patients more exactly than ever before." Like others involved in the talks, Dr. Williamson declines to give any details of their nature or status.

Executives and scientists have worked hard to keep the yearlong talks private for fear that premature disclosure could jeopardize agreement, rattling the companies that are still trying to decide whether to participate. Those familiar with the negotiations say about four or five major drug makers, including Merck, still haven't agreed to join the proposed venture. A Merck official says the company dropped out of the talks.

Even so, scientists and research executives at most of the large pharmaceutical companies are enthusiastic about the map and want to join in a venture if possible. That is because they agree that open access to a map of the genetic landmarks, called SNPs, is crucial to the new way drugs will be tested and developed in the 21st century.

The proposed not-for-profit consortium would be funded with about $75 million to $100 million from participants. It would compete directly with the biotech companies that are making proprietary catalogs of genetic landmarks, firms that include Incyte Pharmaceuticals Inc., Millennium Pharmaceuticals Inc. and France's Genset SA. Through the consortium, major drug companies hope to make an end run around the proprietary databases and prevent the biotech companies from possibly locking up all the key genetic data for themselves.

The map of SNPs (or "snips," as they are called) would complement the federal government's human-genome project, whose biggest goal is to come up with a complete and accurate DNA sequence. The federal project recently began identifying snips as well, but it doesn't have the money to identify enough snips for applications such as making individually tailored therapies.

"SNPs serve as a blinking light on DNA sequences showing there is something very interesting here -- for example, something that is contributing to diabetes. Finding these SNP variations will thus provide new diagnostic tools and incredibly important clues to new therapies," says Francis Collins, who directs the National Human Genome Research Institute. Dr. Collins, who is concerned about gene data being patented by corporations, says, "This is the kind of public-private partnership that I have hoped for. I am waiting anxiously to see it happen." He says he doesn't know details of the proposed venture.

Many big drug makers, among them Merck, Glaxo Wellcome PLC, SmithKline Beecham PLC, Pfizer Inc. and Bristol-Myers Squibb Co., have been rushing to build their own snip maps. But several drug-company executives say they now believe a jointly developed and owned mapping project is essential.

It is crucial "that we have something whose accuracy we all agree upon," says Allen D. Roses, vice president and world-wide director of genetics research at Glaxo Wellcome. Under the direction of Dr. Roses, a renowned gene-hunter who joined Glaxo from Duke University in 1997, Glaxo has been constructing its own snip map.

"If each of us produced our own map, it would, for one thing, take much longer to create," Dr. Roses says, "and it would be very unlikely that the companies would accept one another's map as being valid." Like most drug-company executives, he declines to discuss details of the proposed joint venture.

But he and a number of other scientists agree that patient profiles generated from a complete SNP map could revolutionize the practice of medicine within the next decade. Thanks to rapid advances in gene-sleuthing technology, many pharmaceutical executives believe the key to future blockbuster therapies is identifying which newly invented drugs work best for which patients. Scientists are increasingly convinced that inborn but minute genetic differences largely account for different health traits in people and also explain why a given drug works for one person but is useless or produces bad side effects in another.

An example of how a complete snip map could speed discovery of a drug for Alzheimer's disease was outlined by Dr. Roses and his colleagues at Glaxo Wellcome last week at a scientific meeting in Paris. Several years ago, a Duke laboratory led by Dr. Roses showed that people with a particular version of a gene called APOe are at high risk of getting Alzheimer's. Glaxo researchers are testing drugs that act on the protein that this gene variant produces in cells.

Dr. Roses's discovery at Duke took several years. But a team he leads at Glaxo has sought to determine whether a SNP map might have found APOe much faster. The scientists gathered genetic material from 800 Alzheimer's patients from around the world and, using the snip markers, were able to identify where, in the human DNA, people with the disease consistently had a genetic variation not seen in a similar group of people without Alzheimer's.

The company had expected the effort to take a year or two. But within five months, the scientists found APOe's location. "We were surprised by the speed and accuracy" the snips provided in locating the gene, Dr. Roses says. "It was then, inside Glaxo, we realized that a full SNP map would help us get at genes that predispose people to a wide range of other diseases, too."

In early 1998, based on the Alzheimer's-gene success, Glaxo undertook to build its own map, in a project code-named Atlas. It hoped Atlas would help identify genetic differences that make people susceptible to about 20 major disorders. Glaxo already has used the technique to identify the location of a gene that makes some susceptible to psoriasis, an autoimmune disease.

Dr. Roses says the Atlas project also helps those running clinical trials of new drugs determine which test subjects best respond to the new medicines.

Best of all is the potential for personalized medicine, growing out of gene-based diagnosis. "In the future, before a doctor prescribes a medicine, the doctor will take some blood, have it analyzed at a nearby lab and identify which of, let's say, 12 drugs are most likely to treat the patient effectively with the minimal side effects," Dr. Roses predicts.

Such treatment wouldn't be possible without a complete snip map readily available to doctors.

Glaxo and others building their own maps realized that a common map was crucial for the success of personalized medicine. Among other things, the Food and Drug Administration would need to know that the map was accurate, reliable and agreed to by the scientific community. So beginning last April, scientists from a dozen or more companies gathered under the leadership of Dr. Williamson, the retired Merck executive, to explore the possibility of a joint venture. Some of the companies had at first considered creating a proprietary snip-mapping company, but they soon agreed that a nonprofit entity that dumped all the snip data into a public database would avoid many problems, such as antitrust.

Wary as they were of sharing their valuable data with rivals, they worried even more about biotech companies' projects. For instance, Genset in Paris is racing to complete its own proprietary snip map. Its lead scientist, Daniel Cohen, says Genset has already discovered two genes that predispose men to prostate cancer. He hasn't revealed their identity nor those of several other disease-related genes Genset claims to have found.

In recent conversations, Dr. Cohen has said Genset's plan was to patent the genes and sell them to the highest bidder, perhaps for $50 million to $100 million each. Dr. Cohen says he believes Genset's snip map will enable it to uncover 10 or so disease-related genes for each of about 20 major disorders. "If all goes well," he says, "we could uncover about 200 genes within the next few years."

Genset Chief Executive Pascal Brandys says he is aware of the proposed consortium, but it won't affect his plans because Genset has already been finding snips for two years. A Millennium spokeswoman says that any public snip database won't affect its plans, either, as it uses many gene-hunting tools in addition to snips. Randy Scott, president of Incyte, says he has heard of the proposed collaboration, but it won't affect Incyte's efforts.

Those familiar with the consortium plan say a goal would be to level the playing field for gene-hunting biotech firms, large drug companies and academic scientists whose work is backed with federal funds from the genome project. The consortium would release its map into a public database. All the drug companies, biotech firms and academics would have free and equal access. The consortium would file intellectual-property claims on the public snips so the biotech firms couldn't patent them and sell them back to the corporations.

However, any gene identified by a drug maker, biotech firm or academic by using the snip as a gene-hunting tool could still be patented by the discoverer. "The idea here isn't to restrict the ability of biotech firms or anyone else to patent genes," says one executive. "The idea is to make sure the underlying map we all need to find genes is available to anyone who wants to use it."

Executives familiar with the venture say lawyers have told them that by making the map public, the companies should be able to avoid antitrust

problems.

Though the membership of the proposed consortium remains uncertain, insiders say Britain's Wellcome Trust, a foundation that funds roughly $500 million a year of medical research and has backed many gene-hunting projects, is a major driver of the plan. Other companies that have expressed interest include Novartis AG and Roche Holding Ltd. of Switzerland and Zeneca Group PLC of Britain. None would comment.

People familiar with the talks say that between 12 and 15 drug makers may become charter members -- while emphasizing that a contract hasn't yet been signed. In another effort to avoid antitrust problems, all pharmaceutical and biotech companies would have an open invitation to join even after the snip project was unveiled. The project's director is expected to be Arthur Holden, an American who previously was chief executive of biotech firm Celsis International PLC of Britain. He declines to discuss the venture.

Estimates for how much the effort would cost range from $50 million to $100 million. That would top the estimated $40 million Genset and its partner Abbott Laboratories are spending on development of their own commercial genetic map, expected to be completed by the end of this year.

Drug-company executives contend that the consortium's snip map would be more powerful than Genset-Abbott's effort, boasting well over 100,000 SNPs, or markers, compared with 60,000 markers in the Genset-Abbott map. The consortium is negotiating to contract out the actual work of identifying the snips to academic and commercial laboratories, a potential funding boon expected to quicken the world-wide public and private efforts to map the entire human genome.

Some industry executives caution that for all the excitement about snips, the value of such maps in the discovery of genes and development of medicines remains unproved. The potential applications are "all on paper. Until we start gathering the data, all bets are off," says Lee Babiss, a top researcher at Roche's U.S. drug unit. Dr. Babiss, who won't comment on the talks, nevertheless predicts that snips will eventually prove essential to drug development.

Drug researchers generally hope negotiations over the next few weeks proceed without too many snags. That's because these days, most all of them agree with George Poste, a top science executive at SmithKline, who several months ago said that gene data such as snip maps "will start having an impact" in finding new medicines and tailoring drug treatments "within the next three to five years."

© The Wall Street Journal