News this Week

Science  29 Jul 2005:
Vol. 309, Issue 5735, pp. 678

You are currently viewing the .

View Full Text

Log in to view the full text

Log in through your institution

Log in through your institution


    Controversial Study Finds an Unexpected Source of Oocytes

    1. Gretchen Vogel

    Scientists have made some surprising claims about bone marrow and blood cells in the last few years, but this week brings perhaps the most surprising of all: that cells in the bone marrow and blood are a source of developing oocytes found in the ovaries. If true, this work in mice would rewrite the current understanding of the female reproductive system. It could also open new discussions about the ethics and potential consequences of bone marrow and even blood donation.

    Although the study's authors do not have evidence that such blood-derived oocytes could be fertilized and develop into babies, they suggest that human donors might be sharing germ cells along with their lifesaving immune cells and clotting factors. They also say they hope this work will lead to new treatments for infertility, especially for women who must undergo chemotherapy.

    For decades, scientists have thought that female mammals are born with a lifetime supply of potential oocytes in the ovary. That view was challenged last year by Jonathan Tilly, Joshua Johnson, and their colleagues at Massachusetts General Hospital in Boston, who reported in a controversial paper in Nature that new oocytes could form throughout an adult mouse's lifetime (Science, 12 March 2004, p. 1593). That finding has not been replicated in another lab.

    Tilly, Johnson, and colleagues have now dropped another bombshell at a meeting* and in the 29 July issue of Cell: They report that they have found ovary-replenishing germ cells in the bone marrow and circulating blood of adult mice. They build their case on several lines of evidence. First, looking for the source of oocyte stem cells that might explain their previous results, the team found signs that genes typical of germ cells were expressed in samples of bone marrow from mice and from humans. They also found that the level of at least one of these genes, called Mvh, varies during the animals' estrus cycle. That made them wonder if cells in the bone marrow might be a source of new oocytes.

    To check that idea, the team treated mice with two chemotherapy drugs that cause infertility, cyclophosphamide and busulfan. Mice that received the drugs, as expected, suffered extensive ovary damage and stopped producing new oocytes. But in the ovaries of treated mice that later received bone marrow transplants from female donors, the scientists found “several hundred” oocyte-containing follicles at various stages of maturity.

    The effect of treatment was rapid: New oocytes appeared 28 to 30 hours after a transplant. Some oocyte development experts are dubious, noting that fruit fly oocytes take a week to mature from stem cells. “You just can't do it in a day,” says Allan Spradling of the Carnegie Institute of Washington in Baltimore, Maryland. But Tilly says the oocytes might begin to mature in the bone marrow and continue developing as they travel through the bloodstream.

    Blood borne?

    Jonathan Tilly of Massachusetts General Hospital and colleagues claim that bone marrow transplants and blood transfusions can prompt the ovaries of genetically infertile mice to begin producing oocytes (inset).


    The team also reports using bone marrow and blood transplants to prompt the growth of oocytes in mice that are genetically infertile. Mice with a mutation in a gene called ataxia-telangiectasia mutated can't produce mature germ cells, and their ovaries usually lack follicles and developing oocytes. But after receiving either bone marrow or blood from healthy donors, the team reports, the animals' ovaries started producing follicles containing healthy-looking oocytes. The team concludes that bone marrow provides a continuous source of germ cell stem cells to the ovaries throughout adult life.

    So far, however, they have not been able to prove that these cells can trigger ovulation or give rise to new offspring. “Until the authors have shown that the putative oocytes are functional, we should be cautious,” says Margaret Goodell of Baylor College of Medicine in Houston, Texas, who studies bone marrow stem cells. She and others say the markers the team used to identify oocytes can be misleading. For instance, similar techniques have led others to conclude mistakenly that bone marrow cells had become neurons or lung cells. “It will be important to transplant [green fluorescent protein] positive bone marrow cells into GFP-negative adult mice to test whether those mice go on to give birth to GFP-positive pups,” says Sean Morrison of the University of Michigan, Ann Arbor. “This experiment should be straightforward.”

    Tilly says the team is working on such experiments but has had to find a new approach because the drugs they were using can damage the uterus and fallopian tubes, possibly preventing mice from becoming pregnant.

    Turning to the clinic, Tilly suggests that the mouse results could explain a number of surprising reports of cancer patients and others who were expected to be infertile but who gave birth to children after receiving bone marrow transplants. One patient with Fanconi's anemia, for example, had a single menstrual period and then entered menopause at age 12. After receiving a bone marrow transplant from a sibling, Tilly says, her periods resumed, and she later gave birth to two children.

    Although genetic tests of patients and their children might answer the question, Tilly says, they would be ethically problematic. And such cases wouldn't necessarily be easy to detect, he says, because bone marrow donors are often siblings.

    Even if the new oocytes can't be fertilized, Tilly says, they may nevertheless enhance a woman's fertility. He speculates that they may function as “drone oocytes” that keep the ovary functioning to support the original “queen” oocytes set aside for procreation. If so, he says, the results open new possibilities for preserving or restoring the fertility of young cancer patients and might even provide a way to postpone menopause.

    But until the team produces mice that can be traced without a doubt to a bone marrow donor, scientists are likely to remain wary. “The experiments will have a stimulating effect on the field,” says Hans Schöler of the Max Planck Institute for Molecular Biomedicine in Münster, Germany, “even if they stir quite some controversy.”

    • * Society for the Study of Reproduction, Quebec City, Canada, 24-27 July.


    Dinosaur Embryos Hint at Evolution of Giants

    1. Erik Stokstad

    Paleontologists have long assumed that giant dinosaurs called sauropods, like all other dinosaurs, evolved from smallish bipedal ancestors and dropped down on all fours only as their bodies grew too large to be carried on two feet. But when they examined a pair of embryos dug up about 30 years ago—the oldest fossilized dinosaur embryos so far discovered—they got a surprise. As described on page 761 by Robert Reisz of the University of Toronto's Mississauga campus in Canada and colleagues, the embryos suggest that sauropods were already quadrupedal even as smaller creatures. “This would be significant because it means we might have to re-evaluate the origin of many features in sauropod skeletons we assumed had to do with weight support,” says Matthew Bonnan of Western Illinois University in Macomb.

    The clues are indirect, because the embryos are not sauropods but members of their closest kin, a group of much smaller herbivores called the prosauropods. Paleontologists found them inside remarkably well-preserved eggs of a 5-meter-long animal called Massosponodylus, which 190 million years ago roamed the floodplains of what is now South Africa. “It's a really cool discovery,” says Kristi Curry Rogers of the Science Museum of Minnesota. The eggs clearly contained embryonic bones, but only recently did paleontologists dare to prepare them. It took Reisz's lab technician Diane Scott more than a year of full-time work to expose the delicate bones of the 6-centimeter-long eggs. As Reisz studied the specimens with colleagues from the Smithsonian Institution and the University of the Witwatersrand in Johannesburg, South Africa, he identified the largish skull as that of Massospondylus.

    What was unusual was the rest of the body. “The proportions are just ridiculous,” Reisz says. The neck was long, the tail short, and the hind and forelimbs were all roughly the same length. “It was an awkward little animal,” he concludes. Because of the lack of developed teeth, huge head, and tiny pelvis (where leg muscles attach), the group proposes that Massospondylus hatchlings would have required parental care. “This is certainly suggestive but very difficult to test,” says Martin Sander of the University of Bonn, Germany.


    Embryos suggest that prosauropod dinosaurs grew up from four-legged hatchlings.


    To Reisz, the horizontal neck, heavy head, and limb proportions all suggest that the embryo would have walked quadrupedally after hatching. That's strange, because it means that as the Massospondylus hatchlings developed, they had to become bipedal—a pattern of development almost unheard of among vertebrates. To figure out how the hatchlings changed as they matured, the researchers measured nine other Massospondylus fossils of various sizes. They found that the neck grew much more rapidly, relative to the femur, than the rest of the body did, while the forelimbs and skull grew more slowly.

    If the earliest sauropods also developed from embryos with quadrupedal proportions, Reisz and his colleagues propose, sauropods may have become quadrupedal adults by retaining their juvenile state into adulthood, a phenomenon called pedomorphosis. “It sheds some light in the evolutionary pathways through which the peculiar adaptations of giant dinosaurs were attained,” says Eric Buffetaut of France's major basic research agency, CNRS, in Paris.

    Bonnan notes that other traits of adult sauropods seem to fit the same pattern. For example, the rough ends of sauropod limb bones indicate that the animals sported lots of cartilage in their joints. Paleontologists had assumed that the joints evolved because they helped sauropods support their weight. But cartilage-rich joints are more typical of young vertebrates, so adult sauropods might have acquired them by retaining a youthful trait.

    Some paleontologists, however, are wary of trying to read too much of the history of sauropod evolution from two embryos. So little is known about dinosaur embryology, they say, that it's dicey to reconstruct the loco-motion of hatchlings and extrapolate to other taxonomic groups. “It's a stunning find,” says Anusuya Chinsamy-Turan of the University of Cape Town, South Africa, but “I have all these questions.”


    Rogue Fruit Fly DNA Offers Protection From Insecticides

    1. Elizabeth Pennisi

    Genomes are full of DNA that doesn't belong there. Called transposons, these small bits of sequence jump between chromosomes, often disrupting genes in the process. But sometimes, these interlopers do some good. Dmitri Petrov, a population geneticist at Stanford University in California, and his colleagues have discovered a transposon that, by changing a gene, seems to help fruit flies evolve resistance to certain insecticides. The work, reported on page 764 of this issue of Science, is one of a growing number of examples of natural selection preserving transposons, indicating that “they may play a much larger role in evolutionary novelty than is currently appreciated,” says Todd Schlenke, an evolutionary geneticist at Cornell University.

    Typically, researchers have stumbled on such beneficial transposons while searching for mutations involved in disease or traits such as resistance to toxins. The general assumption has been that these movable DNA elements have long been intertwined with the gene in question. But Petrov and his colleagues demonstrated that transposon-mediated evolution can happen in real time to create novel solutions to changing conditions.

    Working with Petrov, Stanford graduate student Yael Aminetzach had determined which of the 16 members of the Doc family of transposable elements were common in populations of the fruit fly Drosophila melanogaster. One stood out, Doc1420. Unlike other Doc transposons, which proved to be quite rare, this one appeared in 80% of fruit flies tested from eight different countries, suggesting that it plays some useful role. “The paper is a tour de force of population genetics,” says David Heckel, a geneticist at the Max Planck Institute for Chemical Ecology in Jena, Germany.

    A little help from …

    Although transposable elements tend to be harmful, one has helped make Drosophila melanogaster tougher to kill.


    When the Stanford researchers then looked more closely at this transposon, they found that it had landed in a gene that, to date, has defied characterization. The gene exists intact in distantly related fruit flies, suggesting that it has a key function—one that was disrupted as Doc elements jumped around the D. melanogaster genome. By comparing Doc1420 to the other Doc sequences, Aminetzach and graduate student Michael MacPherson estimate that Doc1420 buried itself in this gene 90,000 years ago but did not become widespread until between 25 and 240 years ago, when human activities began to alter the environment dramatically. This recent expansion suggested that, rather than rendering the gene nonfunctional, the transposon altered it, possibly resulting in a different protein product—one that became important to the species' survival.

    The sequence of the unaltered gene provided a clue to this new gene's role. That sequence resembles that of genes for choline metabolism, which operate in nerves affected by organophosphate pesticides. To test whether the new protein was involved in this pathway, the researchers bred fruit flies to create strains that differed only in whether they carried the Doc1420 insertion. The Doc1420 strain fared much better when Aminetzach and her colleagues treated the insects with an organophosphate insecticide: 19% died, compared to 68% of the fruit flies lacking Doc1420.

    Researchers have already identified a few other examples of transposon-induced insecticide resistance, but this is the first to disrupt a gene whose protein is not a target of the pesticide, Petrov says. But Schlenke, Heckel, and others say that more work is needed to verify the transposon's role in resistance. “The data showing pesticide resistance [are] very weak,” notes Richard ffrench-Constant, a molecular entomologist at the University of Bath, U.K.

    Nonetheless, Martin Feder of the University of Chicago is quite enthusiastic. “The paper is the latest in a series of recent discoveries that transposons can play a role in 'real time' microevolution in natural populations,” he says. “The phenomenon is [now] difficult to ignore.”


    Two Mines in Running for Underground Lab

    1. Adrian Cho

    The U.S. National Science Foundation (NSF) has decided that it's in the business of experimentation, not excavation. On 21 July, the $5.5 billion research agency chose two established mines—the Homestake Mine in Lead, South Dakota, and the Henderson Mine in Empire, Colorado—as possible sites for a multipurpose underground laboratory. In doing so, NSF passed over four “green field” sites that would have required builders to excavate thousands of feet of rock and existing sites in Nevada and Ontario, Canada.

    The proposed Deep Underground Science and Engineering Laboratory would house experiments in particle physics, geoscience, and microbiology. The original idea was for federal lawmakers to salvage Homestake for scientific ends before it was abandoned and flooded (Science, 6 June 2003, p. 1486). But that initiative was derailed by political and environmental considerations, leaving NSF free to pursue a more deliberate process that engaged a larger section of the scientific community. Last October, the agency solicited proposals for other sites.

    The two preliminary winners in that competition “stood out significantly above the rest” because they are deep, have desirable geologic characteristics, and come with some infrastructure already in place, says John Lightbody, executive officer of NSF's division of physics. Each team will receive $500,000 to work up a full conceptual design for the laboratory, which backers hope could win funding as early as 2009.

    Rocky Mountain high.

    The Henderson molybdenum mine west of Denver, Colorado, has made the first cut to become an underground laboratory.


    Both mines present challenges. Henderson is an active molybdenum mine, meaning that researchers would have to coordinate their activities with the mining operations. But a working mine also provides functioning lifts, vents, and other infrastructure that researchers can take advantage of, says Chang Kee Jung, a particle physicist at Stony Brook University in New York and spokes-person for the Henderson Mine collaboration.

    In contrast, the abandoned Homestake gold mine was sealed in 2003 and is currently filling with groundwater. Once it reaches 1480 meters below the surface, possibly by 2007 or 2008, the mine's infrastructure could be ruined. However, South Dakota officials plan to open the upper levels of the mine for experiments and begin pumping out water as early as 2006, says Dave Snyder, executive director of the South Dakota Science and Technology Authority. Barrick Gold Corp. has agreed to transfer the mine to the state if the state legislature approves funds to open the site or if NSF builds the lab at Homestake, Snyder says.

    Last weekend, the University of Minnesota, Twin Cities (UMTC), hosted a workshop to discuss the scientific mission of an underground lab. Some scientists feel that NSF short-circuited its own process by narrowing the choices to just two alternatives and excluding green-field sites. “If what they wanted was cheap and deep, they could have told us that right away, and we wouldn't have had to do all this work,” says Priscilla Cushman, a UMTC physicist who worked on a losing proposal to dig the laboratory at the Soudan Mine in Minnesota.

    Despite their disappointment, most scientists are expected to rally behind one of the two remaining collaborations, says Bernard Sadoulet, a cosmologist at the University of California, Berkeley. “I'm convinced that the science is so compelling that the community will pull together,” says Sadoulet, who is leading a study to define the scientific mission of the lab. That teamwork, however, is only the first step in a long process.


    U.S. University Backs Out of Biolab Bid

    1. Andrew Lawler

    The University of Washington (UW), Seattle, last week abruptly abandoned its attempt to build a biosafety level 3 (BSL-3) facility to study infectious diseases and bioterrorism agents. University officials say they were unable to come up with the $35 million required by the National Institutes of Health (NIH) to keep the proposal alive. But there was also intense opposition to the proposed $60 million facility from community activists, who saw it as a public health and safety hazard.

    The university was one of several institutions that applied last December for a Regional Biocontainment Laboratory grant, part of a post-9/11 push to increase the nation's ability to study infectious agents. NIH has set aside approximately $125 million for a second national competition to complement an earlier round of nine labs funded in 2003 (Science, 10 October 2003, p. 206). It expects to make from five to eight awards for the BSL-2 and -3 labs, which handle materials such as plague.

    Three public forums in Seattle this spring drew hundreds opposed to the 5200-square-meter facility, which would have employed 100 scientists and staff. In May, university officials noted that community trust “has been dramatically undermined” and that building the lab despite opposition could prove “devastating” to community relations. An NIH grant to Boston University to build a lab to study even more dangerous biological agents is moving ahead despite citizen protests (Science, 28 January, p. 501).

    Despite that opposition, chief UW spokesperson Norm Arkans says that the real deal breaker for Washington was money: “We knew it would be difficult to raise the $35 million, since the university has a number of capital needs.” A letter from NIH asking for details of its cost-sharing plans triggered the university's pullout, according to Arkans. NIH officials declined comment on the competition, the winners of which are expected to be announced in September.

    Community activists were delighted, but they don't take credit for preventing construction. “I think it came down to money,” says Kent Wills, head of the University Park Community Club. And some scientists are unhappy with the university's withdrawal. “We desperately need better facilities in the Pacific Northwest,” says Samuel Miller, a UW infectious disease specialist. The decision won't keep BSL-3 work away from Seattle: Two dozen university labs already provide that level of containment.


    Technique Uses Body as 'Bioreactor' to Grow New Bone

    1. Robert F. Service

    Tissue engineers have long dreamed of starting with a small clutch of cells in a petri dish and growing new organs that can then be transplanted into patients. The strategy has worked for relatively simple, thin tissues such as skin and cartilage that don't depend on a well-formed network of blood vessels to deliver food and oxygen. But it hasn't panned out for more complex tissues shot through with vessels, such as bone and liver. Now a novel approach to tissue engineering that grows bone inside a patient's own body could change all that.

    In a paper published online this week by the Proceedings of the National Academy of Sciences, researchers from the United States, the United Kingdom, and Switzerland report that they grew large amounts of new bone alongside the long leg bones of rabbits. When they harvested and transplanted the new bone into bone defects in the same animal, the defects healed and were indistinguishable from the original.

    “This is a fresh, new strategy for tissue engineering that relies on the body's own capacity to regenerate itself,” says Antonios Mikos, a tissue engineering specialist at Rice University in Houston, Texas. “I think it will have an enormous impact on the field.”

    The field of tissue engineering could use some help. Attempts to grow complex tissues outside the body have progressed in fits and starts. Italian researchers, for example, have coaxed bone marrow cells injected into a ceramic matrix to create new bone. But organisms have been unable to resorb and remodel the tissue, as occurs with normal bone. To avoid such problems, researchers led by tissue engineers Prasad Shastri at Vanderbilt University in Nashville, Tennessee, and Molly Stevens and Robert Langer at the Massachusetts Institute of Technology in Cambridge decided to see if they could let the body handle it itself.

    Good as new.

    A surgically formed cavity acts as a “bioreactor” to grow new bone between the periosteum (Ps) and mature bone in the tibia of a rabbit (right), producing a slight bulge of new bone (left).


    Bones are sheathed in a thin membrane of cells called the periosteum. If a small wound or fracture occurs, cells in the periosteum can divide and differentiate into replacement tissue, including new bone, cartilage, and ligaments. Shastri wanted to see if he and his colleagues could use this same wound-healing response to generate new tissue.

    The researchers injected a surgical saline solution between the tibia—the long, lower leg bone—and the periosteum of white rabbits, a standard small animal model for studying bone. This created a small, fluid-filled cavity into which they hoped new bone would grow. To prevent the cavity from collapsing as the saline is absorbed by the body, the researchers injected a gel containing a calcium-rich compound called alginate. Previous studies have suggested that calcium helps trigger cells in the periosteum to differentiate into new bone, and that is exactly what happened, the researchers report. Within a few weeks, the alginate cavities were filled with new bone. And when that bone was removed and transplanted to damaged bone sites within the same animals, the new bone integrated seamlessly.

    “I think the strength of this approach is its simplicity,” Mikos says. “It doesn't rely on the delivery of exogenous growth factors or cells.” That could make it a boon to ortho-pedic surgeons, who often need to harvest large amounts of bone from patients to fuse vertebrae in spinal fusions. That harvested bone usually comes from a patient's hip, a procedure that often produces pain for years. But if this approach works in people, it could enable physicians to generate new bone alongside a patient's shin, for example, which could then be transplanted to other sites.

    The technique could also prove useful for other tissues. With a few tweaks, says Shastri, it works to generate healthy new cartilage. Now the team is looking to see if it can be used to generate liver tissue as well. If so, it may turn tissue engineers' dreams into reality.


    Gene Bank Proposal Draws Support--and a Competitor

    1. Jennifer Couzin

    The U.S. Department of Veterans Affairs (VA) is quietly moving forward with plans for a national gene bank that would link DNA donated by up to 7 million veterans and their family members with anonymous medical records. The bank, which is widely supported inside and outside the VA, would represent the first massive U.S. gene banking effort. But it is causing a furor among scientists, some VA employees, and politicians from New York state. They charge that top VA officials accepted a gene bank proposal from a cancer biologist at Stratton VA Medical Center and the State University of New York (SUNY), Albany, but are now privately circulating another gene bank plan that may leave Albany out. Most senior officials and scientists involved in both plans declined to comment for this story.

    Although some smaller gene banks are sprouting in the United States, none can match those gearing up in Iceland, Estonia, the United Kingdom, and Japan (Science, 8 November 2002, p. 1158). In these cases, DNA samples from hundreds of thousands of people are linked with health information stripped of identifiers, making the banks powerful tools for sorting out "the complex interactions between gene and environment that lead to disease," says Alan Guttmacher, deputy director of the National Human Genome Research Institute (NHGRI) in Bethesda, Maryland.

    The VA, say outside scientists, is a natural home for such a project because health records for the 7 million people it serves are computerized and standardized. The VA "has not only samples but histories," says Karen Hitchcock, president of SUNY Albany until early 2004 and now the principal and vice chancellor at Queens University in Kingston, Canada. Although there are potential dis-advantages to a VA bank—namely low numbers of females, if veterans but not family members are included—the population includes minorities underrepresented in gene banks overseas, says Guttmacher.

    According to documents obtained by Science, in July 2002, Paulette McCormick, who held joint appointments at the Stratton VA Medical Center and as head of SUNY Albany's Center for Functional Genomics, sent a gene bank proposal to Mindy Aisen, then the VA's deputy chief of research and development and now chief of the VA's rehabilitation research division. McCormick's plan was to collect blood samples from at least 2 million volunteers. The data bank would be open "to VA scientists and other academic and industry scientists" after their projects were approved by the VA and the bank's scientific and ethics committees, one version of her proposal states. The samples would be owned by the VA; they and computers containing the data were to be stored in locked rooms at SUNY Albany. McCormick also proposed having companies pay to access gene bank data as a means of funding the bank. Strict privacy controls would protect DNA donors.

    SUNY Albany officials and New York politicians saw the plan as a flagship project that could raise the profile of the university and the state. "We all kind of whooped. It was an absolutely fantastic idea," says Hitchcock.

    On 11 December 2003, the VA signed an agreement with Albany suggesting that it would move forward with McCormick's plan and base the bank in New York state. SUNY Albany modified plans for a cancer research center then under construction, making "add-ons" to accommodate space for a gene bank at a cost of "multiple millions," says Hitchcock. In an e-mail sent on 19 March 2004, Jonathan Perlin, now VA undersecretary for health, wrote to three colleagues in VA headquarters that the gene bank "is a VA resource, first and foremost, and Albany would be a lead partner."

    Not in the bank.

    Paulette McCormick's proposal seemed to have won approval, but the VA is now circulating a similar plan of its own.


    That May, a small VA delegation, including Perlin, traveled to Albany and met with New York State Senator and majority leader Joseph Bruno (R) and New York Governor George Pataki (R), say sources familiar with the meetings. At the time, it was generally understood that New York would supply most of the project's pilot funding—estimated at $10 million—while the VA would offer nominal support, such as staff to collect blood samples.

    But behind the scenes, the project was unraveling. An e-mail from Perlin sent in February 2004 noted that McCormick's proposal "has raised significant ethical, privacy and operational issues." An e-mail from Nora Egan, then VA Secretary Anthony Principi's chief of staff, reported that the secretary felt that "issues related to medical ethics, privacy, … and benefit to be derived by VA" needed to be addressed. Precise concerns were not specified. A fall 2003 review of McCormick's proposal by the director of the VA's National Center for Ethics in Health Care had concluded: "On the whole, the … Gene Bank proposes ethically appropriate measures to protect subjects' privacy and the confidentiality of their personal health and genetic information."

    Earlier this year, VA officials at the agency's headquarters began circulating memos of a separate gene bank proposal, reportedly crafted by Perlin, Timothy O'Leary, who heads VA's Biomedical Laboratory Research and Development Service, and Stephan Fihn, acting head of VA research and development until 31 May 2005. A recent confidential draft, obtained by Science, is dated 13 July 2005.

    Conceptually, the proposal is similar to McCormick's: It recommends gathering blood samples from "all enrollees" in the VA system over 5 years and linking them "to data in other clinical and administrative databases" within the VA. Clinical information would be stored in "highly secure" areas. A scientific advisory committee would offer advice on specimen collection, storage, and other matters; the proposal notes that NHGRI Director Francis Collins has agreed to serve on this committee. (Collins declined to comment.) Biotechnology firms seeking access to the gene bank for specific projects could provide "commercial support." Initial costs are pegged at $40 million to $60 million, and the proposal notes that given tight federal budgets, Congress is unlikely to supply the funds. The proposal diverges from McCormick's in its suggestion that the bank's infrastructure be based in Texas or in Colorado, the home of VA Secretary James Nicholson, to "capitalize on VA support" in those states.

    "In my view, there's an evolution in thinking rather than a competition," says Fihn, who explains that on this project of unprecedented scope, VA headquarters realized it had to be in control. Furthermore, Fihn says, it's ludicrous to argue that Albany owned the concept. "Anybody who takes credit for the idea of creating a gene bank in this day and age—it's like saying you invented the Internet," he notes. He can't say what role, if any, Albany will play in the bank and anticipates a competition for participation.

    "We were rather upset" by how VA has handled the project, says Richard Roberts, a board member at SUNY Albany's Center for Functional Genomics and the chief scientific officer of New England BioLabs in Ipswich, Massachusetts. Roberts, a Nobel laureate, says it appears that McCormick's idea is being "seized" by "people in Washington."

    Last year, as concerns from New York politicians intensified that the VA was backing out of the December 2003 agreement it had signed with SUNY Albany, VA officials asked the agency's general counsel, Tim McClain, for advice. He prepared a memorandum arguing that the agreement isn't binding. "Execution of the subject Agreement by VA did not constitute acceptance of the gene bank research proposal," it reads.

    McCormick, meanwhile, has returned full-time to SUNY Albany after being released from the VA last year. Late last month, McCormick's successor on the gene bank, her SUNY Albany colleague Richard Cunningham, was also released from his part-time appointment at the VA, although he continues to work there without pay. "Employee privacy" rules preclude elaborating on those releases, says Linda Blumenstock, a spokesperson for the Stratton VA Medical Center.


    WHO Faults China for Lax Outbreak Response

    1. Dennis Normile

    Worried that Asia's bird flu outbreak could be on the verge of spreading worldwide, increasing the risk of a human pandemic, international health organizations are warning that China is not rigorously following up on a recent outbreak of the deadly H5N1 strain among wild birds in the western Qinghai region. In particular, the World Health Organization (WHO) is pressing Chinese officials to study migratory birds to see whether they may be able to spread the virus to previously unaffected areas. Chinese scientists point out that they have already sequenced virus from migratory birds and made the results publicly available through GenBank.

    Concerns are focused on the H5N1 outbreak at China's Lake Qinghai. The unprecedented 6000 death toll among wild birds, previously only slightly affected by infections, has experts worried that the virus has become more lethal and that surviving migratory birds could carry it to wintering grounds in India, which has not yet reported any H5N1 outbreaks.

    To assess this risk, WHO and the United Nations Food and Agriculture Organization (FAO) have urged Chinese authorities to sample surviving birds to see whether any are carrying the virus without obvious symptoms, as well as to tag birds for tracking. China's Ministry of Agriculture could not be reached for comment. But in an interview with the Wall Street Journal that appeared on 19 July, Jia Youling, director general of the ministry's Veterinary Bureau, was quoted as saying they haven't tested live migratory birds “because in catching them, it is easy to harm them.” FAO animal epidemiologist Juan Lubroth in Rome says that there are humane ways of testing of live birds. Such data, he adds, “would allow for preventive actions on the ground, such as vaccinating domestic poultry flocks near known rest areas” along migratory routes.

    Roy Wadia, a spokesperson for WHO in Beijing, says China has also not yet responded to requests for isolates of the virus circulating in Qinghai. Time is of the essence, he says, because authorities want to determine whether the virus has changed before the return migration. Wadia was unaware that DNA sequence information from samples from Lake Qinghai had been deposited in GenBank by a group at China's Institute of Microbiology; they reported online in Science that the virus appears to have changed in ways that could make it more lethal (Science, 8 July, p. 231).

    Meanwhile, Indonesia confirmed its first human deaths from bird flu, among a family that apparently had no contact with infected poultry—the usual route of transmission—raising questions about possible human-to-human transmission. And as Science went to press, Russian officials were trying to determine the H5 subtype responsible for an outbreak of avian influenza among poultry in Novosibirsk.


    New Array Takes Measure of Energy Dispute

    1. Adrian Cho

    Amid the incessant hail of cosmic rays striking Earth's atmosphere from outer space, every now and then one comes screaming in with the energy of a walnut-sized hailstone (Science, 21 June 2002, p. 2134). Such ultrahigh-energy cosmic rays could herald bizarre astronomical phenomena or new fundamental particles, so physicists are eager to know how often they come along. In recent years, Japanese experiments have indicated that the particles are unexpectedly common; American experiments say they're rare. Now the first results from the Pierre Auger Observatory, a gargantuan cosmic ray detector under construction on an ancient lakebed near Malargüe, Argentina, may have pinpointed the crux of the dispute: The apparent energy of the cosmic rays depends on which method is used to measure it.

    Auger's preliminary findings “go a long way to resolving the difference between the two [previous] data sets,” says Floyd Stecker, a theoretical astrophysicist at NASA's Goddard Space Flight Center in Greenbelt, Maryland. Auger researchers will present their results next week at a conference in Pune, India.*

    When a high-energy cosmic ray crashes into the atmosphere, it triggers an avalanche of billions of lower energy particles known as an “air shower.” Between 1990 and 2004, researchers working with the now-defunct Akeno Giant Air Shower Array (AGASA) about 120 kilometers west of Toyko, Japan, caught some of the particles with detectors on the ground. They compared their readings with the results of a computer simulation to deduce the energy of the original cosmic ray. The researchers spotted about a dozen cosmic rays with energies exceeding 100 exa-electron volts (100 EeV, or 1020 eV).

    Cosmic dissonance.

    Auger's particle detectors (foreground) and telescopes measure different energies for particles from space.


    As they stream earthward, the particles in a shower excite nitrogen molecules in the air and cause them to fluoresce. Researchers working with the High-Resolution Fly's Eye (HiRes) detector at the U.S. Army's Dugway Proving Grounds in Utah use specialized telescopes to detect that light and estimate the energy of the original cosmic ray somewhat more directly. They observed only a few cosmic rays with energies above 100 EeV.

    The Auger Observatory possesses both types of detectors. Auger researchers observed dozens of cosmic rays with both the telescopes and the ground detectors and used the “hybrid events” to calibrate the ground detectors without resorting to the computer simulations. The results suggested that the computer simulations overestimate the energies of the cosmic rays by about 25%, says James Cronin, a physicist at the University of Chicago and co-founder of the Auger collaboration.

    Some physicists, however, question whether the energy estimates from the fluorescence detectors are really more accurate than those from the simulations. “The Auger measurement clearly explains the difference between the AGASA and HiRes results,” says Masahiro Teshima, a cosmic ray physicist at the Max Planck Institute for Physics in Munich, Germany, and former spokesperson for AGASA. “But at the moment, I don't know which is right.”

    All agree that as it gobbles up data, the massive Auger Observatory should settle the issue once and for all. “In a year and a half with a quarter of the array, we've matched the data set of the existing experiments,” Cronin says. “It's looking good.” The complete array will comprise 24 light telescopes and 1600 surface detectors covering 300 square kilometers. Within 2 years, Auger researchers expect to have collected seven times more data.

    • *29th International Cosmic Ray Conference, 3-10 August.


    El Niño or La Niña? The Past Hints at the Future

    1. Richard A. Kerr

    Two teams of researchers, studying the same evidence with the same techniques, have painted diametrically opposite pictures of a key period in the history of Earth's climate, which climatologists are probing for hints of what's to come. “It's a tough issue to sort out,” says climate modeler Raymond Pierrehumbert of the University of Chicago in Illinois. “What's at stake is the regional distribution of climate,” both past and future. But he's going to have to wait for more data from the past.

    The two groups, one British and one American, are studying what temperatures in the equatorial Pacific Ocean were like during the early Pliocene epoch, about 4.5 million to 3.0 million years ago. The world was about 3°C warmer then than it is today—much as it may be a century or two from now. Today, the tropical Pacific is the “engine” that drives much of the global climate system. Computer climate models disagree about how future global warming will affect it: whether the region will get stuck in the warmth of a permanent El Niño, slip into the relative cool of an endless La Niña, or keep swinging from one to the other as it does today. By showing how the tropical Pacific worked the last time the world got hot, climatologists hope the Pliocene will help them forecast what to expect next time.

    To find out ancient ocean temperatures, each group studied a pair of deep-sea sediment cores from either end of the pivotal equatorial Pacific, one taken from near the Galápagos Islands and one from 13,000 kilometers to the west. From the mud, they extracted the fossils of microscopic creatures called foraminifera, or forams, that lived in Pliocene surface waters and sank to the bottom after they died. By studying the ratio of the elements magnesium and calcium preserved in forams' carbonate shells, scientists can estimate the temperature of the water the creatures once floated in.

    No match.

    Ancient sea surface temperatures determined by two groups (blue and gold dots) from forams (right) starkly disagree during the early Pliocene (3.0 million to 4.5 million years ago).


    The British group weighed in first (Science, 25 March, p. 1948). Rosalind Rickaby and Paul Halloran of the University of Oxford, U.K., published six eastern Pacific temperatures spanning the past 5 million years, including one from the Pliocene warm period. It showed that the eastern Pacific was dramatically cooler than the west—the hallmark of a dominant La Niña.

    Now, on page 758, the American group—Michael Wara, Christina Ravelo, and Margaret Delaney of the University of California, Santa Cruz—reaches a different conclusion. They produced more than 200 temperatures over 5 million years, including more than 50 from the time of Pliocene warmth. Wara and colleagues conclude that at that time the eastern Pacific was only slightly cooler than the west. The implication: El Niño, not La Niña, ruled the early Pliocene.

    It's a big difference. A dominant La Niña would have made the world slightly cooler on average than the alternative. More important, La Niña's regional climate effects—such as a wetter western Pacific and a cooler northwestern North America—would have been felt around the globe. If El Niño prevailed, on the other hand, that would have meant a warmer climate overall and much warmer and drier conditions in southern Africa, for example.

    So who is right? Outside experts say the Californians' hundreds of temperature readings give El Niño a tentative edge. “You need really dense data sets to do this work well, in my opinion,” says paleoceanographer David Lea of the University of California, Santa Barbara. “This is difficult work, and it's easy to be misled.” Paleoceanographer Gary Dwyer of Duke University in Durham, North Carolina, agrees, noting that sampling as sparse as the Oxford group's could make it easy to mistake a few rare cold-water interludes for a long-term La Niña regime. But Rickaby stands by her team's results and hints that superior British sample cleaning more than closes the numerical gap in data points.

    Researchers say only more research can settle what really happened during the Pliocene. “There may be missteps before it's done,” says Pierrehumbert, but “I can't overemphasize the importance of such data” to testing climate models


    Learning to Adapt

    1. Erik Stokstad

    The ambitious Northwest Forest Plan tried to balance desires for timber and biodiversity, but preservation trumped logging—and research. Can the plan be made as adaptable and science-friendly as intended?

    For decades, a steady stream of logging trucks rolled out of forests in the Pacific Northwest, piled high with ancient Douglas firs, valued for their huge trunks. Old-growth forests on private lands were the first casualties, and as they disappeared, the loggers turned to national forests. Despite outcries from environmentalists, the pace of clear-cutting intensified in the 1980s—reaching a peak of more than 5 billion board feet a year, enough to build 350,000 three-bedroom houses, much of it from old growth. Then in the early 1990s, environmentalists finally found a weapon powerful enough to fight destruction of these venerable forests: the northern spotted owl, which needs large tracts of old trees to survive.

    Not long after the owl was added to the endangered species list in 1990, environmental groups sued on its behalf, and a federal judge ordered a moratorium on logging in owl habitat. The rumble of trucks from the national forests silenced, but the volume of the debate only got louder. As it played on national media, the bitter battle pitted birds against jobs. Activists spiked trees to damage mills, while loggers held protests and cut down old-growth trees at night. The tension ratcheted up.

    Out of this political crisis came the largest, most ambitious forest conservation plan ever. Called the Northwest Forest Plan (NWFP), it covers 9.8 million hectares of federal land in California, Oregon, and Washington. Striving for compromise, the plan tried to balance the needs of loggers and endangered species. To meet that tall order, the architects set up special research areas to devise new ways of cutting timber that would be benign or even beneficial to wildlife. Economic and ecological progress would be monitored, and the plan would be altered decade by decade as needed—a process called adaptive management.

    Now, more than 10 years and $50 million in monitoring costs later, researchers and forest managers have taken the first major stab at assessing how well the plan is working. This fall, they will publish a series of extensive reports, with a synthesis slated for release this month. The bottom line, they say, is that the plan is basically on track: Old-growth forest has been preserved, and watersheds are improving. But several key goals have not been met. Some forests face the risk of catastrophic fires; the spotted owl population is still declining; and timber sales never came near projections, meaning lost jobs and dollars for both the timber industry and the U.S. Forest Service (USFS).

    Another shortcoming is the relative dearth of new approaches for improving the plan. Despite good intentions, the goal of devising and studying alternative management strategies essentially fizzled. Officials say that fixing this is a top priority, as is reducing fire risk.

    Flash point.

    Cutting of old-growth trees, like this Douglas fir, created bitter conflict and led to the Northwest Forest Plan.


    But keeping the plan on track—let alone boosting its activities—faces serious challenges, as funding for the USFS in the Pacific Northwest has fallen dramatically. Forest service officials say that changes in regulations governing the plan, implemented by the Bush Administration, will give them needed flexibility, but environmentalists worry that the changes provide license for irresponsible logging that could threaten remaining old-growth forests.

    Legal logjam

    Several broad environmental laws passed in the 1970s made the conflict between logging and old-growth conservation all but inevitable. The Endangered Species Act (ESA) of 1973 requires the conservation of habitat that listed species depend on, and sections of the National Forest Management Act mandate that populations of species be kept viable. Forest service officials knew in the 1980s that the spotted owl was likely to be listed but, under pressure from politicians in the northwest, continued to allow cutting of old-growth forests—until the Seattle Audubon Society and other groups sued.

    In March 1989, a federal circuit judge blocked sales of timber within the range of the owl, an area encompassing the remaining old growth. Congress intervened, allowing a few timber sales to go through, enraging environmentalists. The issue rose to prominence in the 1992 presidential campaign.

    A few months after the election, President Clinton asked a large group of scientists from USFS, the Bureau of Land Management (BLM), and universities to provide a range of options that could end the judicial moratorium. The Forest Ecosystem Management Assessment Team (FEMAT) was charged with finding ways to protect the long-term health of the forest across the range of the spotted owl while providing “a predictable and sustainable level of timber sales and nontimber resources that will not degrade the environment.”

    A core team of several dozen researchers, led by wildlife biologist Jack Ward Thomas of USFS, holed up for 3 months in a Portland office building, working around the clock and calling on more than 100 outside scientists when needed. “The mood was one of great intensity and focus,” says FEMAT participant Norman Johnson of Oregon State University in Corvallis. From this came a 1366-page document that laid out 10 distinct management options. All of them took a broad view, focusing on managing the entire ecosystem rather than just the spotted owl. But to survive court challenges, any plan had to comply with laws aimed at species protection.

    Clinton picked Option 9, which set up a patchwork of old-growth areas—45 so-called Late Successional Reserves, totaling 2.8 million hectares or almost 30% of federal land in the plan area. The primary objective in these reserves was to ensure the survival of old-growth forest habitat that the owl requires. Some 1.9 million hectares outside the reserves, called the matrix, would be available for logging, except near owl nests.

    To figure out what type of management would be most compatible with conservation and timber goals, the plan set aside 10 areas (see map, below), totaling 603,000 hectares, for experimentation with restoration and harvesting approaches. It also called for different management strategies in various reserves, depending on local conditions. For instance, the pine forests east of the Cascade Range are drier and more prone to fire than those to the west, and decades of fire suppression had led to a buildup of brush and deadwood. They would need aggressive management, including thinning and prescribed burns, to prevent catastrophic fires. To the west of the mountains, by contrast, the idea was to accelerate the development of old-growth habitat by thinning second-growth plantations.

    Mixed success.

    Old growth was preserved on federal land, but not many experiments took place.


    Because officials expected salmon to be listed under ESA, the plan also includes a substantial Aquatic Conservation Strategy. To prevent erosion, which adds sediment and can destroy fish habitat, the plan creates a system of riparian reserves: 100-meter-wide no-logging strips on either side of streams, totaling 903,000 hectares. As more was learned about watershed ecology, the buffers were to be adjusted to the minimum size necessary to conserve fish, thus allowing more logging.

    Before it was implemented, Option 9 went to the departments of Interior and Agriculture, where it was modified—presumably to make it legally more airtight—without scientific advice from FEMAT. The biggest change was to expand the scope of protection beyond species listed under the ESA to include several hundred largely unstudied species whose status was unknown. “The precautionary principle went berserk at that point,” Thomas says.

    Under this additional “survey and manage” program, before any ground-disturbing activity could take place, the agency had to check for the presence of any of these organisms, including lichens and invertebrates, and devise a plan to minimize impact on them. Although this provision has helped the overall plan hold up to court challenges, it had unintended and wide-ranging consequences. In particular, because it made the plan substantially trickier to implement, much logging and many adaptive-management experiments never got off the ground. “It almost made it impossible to pursue the actions in Option 9,” says Thomas, who was chief of USFS from 1993 to 1996.

    Charting progress

    This spring, USFS and BLM began previewing the first monitoring results. In some cases, the data are too sparse to yield a useful assessment, because it took several years to design and implement the monitoring programs. Researchers also note that a decade isn't much time compared to the pace of forest succession and the century-long horizon of the plan.

    For old-growth forests, however, the trend appears positive. Older forest increased by 245,000 hectares between 1994 and 2003, about the amount originally expected. “Perhaps we can conclude for the short term that the policies are working,” says USFS's Melinda Moeur, who led the old-growth monitoring team. But environmentalists counter that the net increase—tabulated when an average tree diameter crosses a certain threshold—means only marginal improvement in habitat, while the 6800 hectares of older forest that were clear-cut represent real setbacks. “The losses are catastrophic, while the gains are incremental,” says Doug Heiken of the Oregon Natural Resources Council in Eugene.

    The plan fell far short of its goal in terms of timber production. About 0.8 billion board feet per year were expected to be put up for sale each year; in most years less than half of that was. A major factor was the stringent requirements of the “survey and manage” program. Environmental groups also slowed things down with lawsuits to prevent any harvesting they thought detrimental.

    This decline in timber harvesting had both economic and ecological effects. Although it cost roughly 23,000 timber-related jobs, that was less than some had feared. Jobs with USFS also disappeared and were not replaced. Yet over the decade, some 800,000 other jobs were created in the region. As former timber workers and USFS employees moved out, they were replaced by retirees and telecommuters. Overall, the Pacific Northwest did not suffer economically because of the plan, says forest economist Richard Haynes of USFS, but some rural communities were hit quite hard. The shortfall of cutting also has ecological implications. The paucity of clear-cutting in former plantations, which would mimic the effects of a severe windstorm or major fire, means that the northwest could end up many decades from now with a lack of early successional forests, which are prized for their biological diversity. And because there was little thinning, which both provides timber and helps accelerate forest succession to old growth, the fire hazard continued to increase in eastern old-growth forests.

    Another disappointment is that despite the progress in habitat preservation, the population of spotted owls is estimated to be declining at 3.4% per year. The culprit is a surprise: invasive species. Barred owls, which are native to the central and eastern United States, have moved west over the past few decades. The newcomers seem to dissuade spotted owls from hooting, and spotted owls are apparently more likely to leave their territory if barred owls appear. Moreover, their diets overlap 75%, so they may be competing for food as well. “Barred owls may ultimately be as big or bigger a threat than habitat loss,” says Eric Forsman, a wildlife biologist with USFS in Corvallis.

    Thin is in.

    Selective logging can speed forest maturation, reduce fire risk, and produce timber.


    Trying to adapt

    A cornerstone of the original plan was adaptive management—essentially, learning by doing and monitoring—which had never been tried on this scale before. The plan called for setting aside 10 adaptive- management areas (AMAs), where scientists would test ideas about how to create or restore forest or riparian habitat and protect threatened species while integrating timber harvest. Most never got off the ground, which leaves the Forest Service with few new ideas to guide efforts to improve the plan. “It's been an extremely frustrating decade,” says forest ecologist Bernard Bormann of USFS. “The progress has been very slow.”

    Several factors scuttled the projects. Tension and lack of trust between forest managers and environmental groups figured large. When environmental groups felt that foresters were using AMAs primarily to extract timber rather than to improve the ecosystems, they sued. However, Dave Werntz of the Northwest Ecosystem Alliance in Bellingham, Washington, says that trust has been building, thanks to better communication and good-faith efforts: “We're doing a better job today at implementing the Northwest Forest Plan than any time in the past.”

    Other problems remain: When national forest budgets got tight, these experiments were axed or fell lower on priority lists. In addition, rather than being encouraged to try novel approaches, local managers had to offer evidence to the U.S. Fish and Wildlife Service (FWS) that experiments wouldn't harm listed species. In many cases, managers simply gave up trying to make projects work or walked on eggshells to avoid legal trouble. “Caution seems to have trumped creativity,” says Elaine Brong, BLM's director for Oregon and Washington.

    There were a few exceptions. The Blue River Adaptive Management Area, for instance, was set up to recreate the effects of historical patterns of forest fires across 23,000 hectares in the Cascades near Eugene, Oregon. Cutting, combined with prescribed burns, has yielded timber at a low but constant rate. The project began only 5 years ago, so no results have emerged yet. But modeling indicates that the experiment will create more old forest than the standard design of the NWFP will and much more intermediate-age forests. “We'll end up with what we believe is a more natural system,” says geomorphologist Fred Swanson of USFS. And thinning experiments in the Siuslaw National Forest near Waldport, Oregon, are probing the best way to accelerate the maturation of younger forests, says Bormann, the lead scientist. Thanks to the thinning, the Siuslaw now produces more timber than any other national forest in the NWFP.

    Overall, scientists say the plan is succeeding at its goal of conserving old-growth ecosystems. “So far so good,” sums up Thomas Spies, a forest ecologist with USFS. Conservation wasn't the exclusive goal at the outset, of course, but the agency seems resigned that it won't meet its timber harvests. “If we can keep them flat, then we'll be doing pretty good,” says USFS spokesperson Rex Holloway.


    Spotted owls face competition from an invasive species.


    That state of affairs—if it holds—distresses the timber lobby but pleases environmentalists. The Bush Administration has, however, implemented several changes that could swing the balance, such as eliminating the “survey and manage” requirements last year to boost timber production. Other major changes, which affect all national forests, include removing the concept of retaining viable populations from the National Forest Management Act and lessening mandatory monitoring and requirements for environmental-impact statements. The changes “give total discretion to the local forest manager on how to manage the forest,” says Michael Leahy of Defenders of Wildlife in Washington, D.C., which has filed suit.

    How these changes specifically affect the operation of the plan will be determined by the Regional Interagency Executive Committee (REIC), made up of officials from USFS, BLM, and other agencies. This group will also decide how to modify the plan based on what's been learned over the past decade. A key priority is “getting the AMAs to work,” says Linda Goodman, regional forester of USFS's Pacific Northwest Region and a REIC member. One strategy is increased involvement of FWS and the National Oceanic and Atmospheric Administration's National Marine Fisheries Service, which are responsible for endangered species, in research design so that scientists and managers have more latitude to take risks.

    Yet as they hope to ramp up research and management activities for the next decade, Forest Service managers face a declining budget and downsizing. The agency's budget dropped 35% in the NWFP area during the first decade, which forced it to cut 36% of positions and close about 23% of its field offices in the plan area. “I'm very concerned,” says Jerry Franklin of the University of Washington, Seattle. “What's happening is a real threat to carrying forward the plan successfully.” To a large extent, the question of funding will determine how much monitoring and experimentation will continue—and what researchers will have learned about managing the forests 10 years from now.


    New National Academy Head Is No Stranger to Spotlight

    1. Eli Kintisch

    Ralph Cicerone came to Washington, D.C., this month to lead the National Academy of Sciences—and walked smack into a hot climate debate

    Last week, Ralph J. Cicerone showed the U.S. Senate what he might be like as the new president of the National Academy of Sciences (NAS): a politically savvy administrator who intends to make the voices of scientists heard in Washington, D.C., and beyond.

    On consecutive days, the 62-year-old atmospheric scientist testified before separate panels examining the science of climate change. To the first panel, he explained firmly why the National Academies had waded into a fight brewing between an influential House committee chair and scientists whose research has linked rising temperatures with human causes by volunteering to look into the questions that Representative Joe Barton (R-TX) had raised about Michael Mann's work (Science, 22 July, p. 545). In the second, he addressed a legislator's concerns about the economic costs of capping greenhouse gas emissions by ticking off seven ways in which efficient energy use would help average Americans.

    Colleagues say his performance, scarcely 2 weeks into his 6-year term as NAS president, was typical of someone who knows how to talk to politicians, peers, and the public. “He's very good at putting all the pieces together from different disciplines to provide a simple answer for societal questions,” says atmospheric chemist Guy P. Brasseur of the Max Planck Institute for Meteorology in Hamburg, Germany.

    Policy-oriented answers to complex problems are the academies' stock in trade. More than 200 times a year, it delivers measured judgments on issues from teaching evolution to energy policy. In 2001, while still chancellor of the University of California (UC), Irvine, Cicerone himself chaired a White House-requested academies' review of climate science that said human activities could result in higher temperatures, drought, and increased rainfall while noting uncertainties. “We were all on the hot seat,” says botanist Peter Raven, who led the academy committee that nominated Cicerone to succeed Bruce Alberts. “But he really came through, with rigor and accuracy.”

    Although Cicerone called his back-to-back Senate appearances “probably more than I'd like to do,” a busy, high-profile schedule is hardly a novelty for him. He maintained a productive research lab at Irvine during his 7-year stint at the helm, avoiding serious cuts to programs and personnel despite a tough budget environment. Raven says that Cicerone's public relations and fundraising skills helped him nab the NAS job.

    Cicerone began his career as an electrical engineer studying atmospheric plasmas. At the University of Michigan, Ann Arbor, in 1973, he and Richard Stolarski showed that free chlorine atoms could decompose ozone catalytically, earning the pair a citation when UC Irvine colleague Sherwood Rowland won the Nobel Prize in 1995. His interests steadily broadened, from methane's role in greenhouse warming to climate change, and he reported his findings in regular testimony on Capitol Hill.

    Cicerone spoke last week with Science about his new job. Here are excerpts from that conversation.

    Hot seat.

    New NAS President Ralph Cicerone prepares to testify at a Senate hearing on climate change.


    On his goals for NAS:

    “In my lifetime, I think I've seen a pretty pronounced slippage of the public's enthusiasm for and understanding for science. And I'm going to try to get a number of academy members together and some of our staff to look at our past efforts on communicating and see what we can do better. …

    “I'm [also] really worried about the U.S. science and technology base. … We have a couple of groups working right now to assemble some measures of how we track our progress and our relative standing around the world. … We'll be working this one with the National Academy of Engineering and with scientific and engineering society leaders, too.”

    On the timeliness of reports:

    “That's always been a criticism, but I think things have sped up a little bit. … There have been some fast ones lately, like what to do with the Hubble [Space Telescope]. … You couldn't take on the number of studies we're doing now if all of them were, let's say, 2-month turnaround. And I think by nature, many of the questions we're asked to look at are longer term, anyway.”

    On the number of women members:

    “Last year's [entering class] was the all-time record, with 19 out of 72. … We're doing better, but there are still a lot of ways in which women are not being involved enough, like in our choice of award winners and officers of the academy. We've got a long way to go.”

    On his career progression:

    “I think there's a real difference between leadership and management and administration. … [In 1994] we had a fantastic dean of physical sciences who had to step aside for personal reasons, and they asked me to take over the job. I was out of town when the faculty met. … [But] I've always enjoyed trying to do several things at once. Then when the opportunity came to be chancellor of the campus, … someone said to me, 'You've complained a lot at the way other people do these jobs. Maybe it's time for you to try it.'”

    On a funding gap between the life and physical sciences:

    “In the physical sciences, I think there are many discoveries out there waiting to happen, largely because of our new capabilities in measurement. … I think it was necessary to increase the portfolio for biological and health sciences, and I'm really glad we've done it. But the physical sciences have fallen too far behind.”


    Tackling the Cancer Genome

    1. Jocelyn Kaiser

    Genome sequencers and cancer experts hope a pilot NIH project to find genetic glitches in tumors will build support for a complete catalog of human cancer genes

    The scientists who brought you the human genome project are teaming up with cancer researchers for another big-biology moon shot. They want to compile a catalog of all common mutations found in human cancers, with the goal of jump-starting molecular approaches to treating cancer. Last week, at a Washington, D.C., workshop to explore the idea, scientists sketched out a game plan for a 3-year pilot project. And two institutes of the National Institutes of Health (NIH) in Bethesda, Maryland, announced a $100 million down payment on a project expected to cost $1.5 billion over a decade.

    The human cancer genome project was hatched by a group of advisers to the National Cancer Institute (NCI) led by Eric Lander of the Broad Institute in Cambridge, Massachusetts, who unveiled the initial plan in February (Science, 25 February, p. 1182). It would identify the genetic glitches that lead to uncontrolled cell growth in most cancers. Lander's group proposes systematically searching for the common mutations in 12,500 tumor samples from 50 major cancer types. “If everybody were to pull together, we could at least know the enemy in a decade,” Lander says.

    Planners say the project dovetails with the push by NCI Director Andrew von Eschenbach to translate genomic discoveries into the clinic. Von Eschenbach says the idea “will embed in the entire strategy of the NCI,” which has agreed to share the cost of the pilot project with the National Human Genome Research Institute. The full project, however, would require additional funding from Congress.

    Cancer genome project backers hope to repeat the success of the human genome effort. But determining the sequence of the 3 billion bases in human DNA, although controversial at first, was a well-defined task compared to what cancerresearchers are proposing. The new project would collect samples from thousands of patients, analyze those samples for mutations found in at least 5% of cases, and measure gene-activity patterns in the tumors.

    Because a full sequencing of each sample to find mutations would cost too much, Lander's group has proposed at first sequencing only the coding regions of 2000 or so genes implicated in cancer. Even then, much of the data may be meaningless, notes Michael Stratton of the Sanger Institute in Hinxton, U.K., which has a smaller cancer genome project under way. Stratton presented data at the workshop on protein kinases, enzymes involved in cell signaling, for several cancers. Only a small fraction of mutations in kinase genes cause abnormal cell growth, he reported, and although some tumor samples carried several mutations, others had none.

    Know the enemy.

    A new cancer project will look for genetic changes, such as this chromosomal translocation in a salivary gland tumor.

    CREDIT: NATURE GENETICS 30, 208-213 (2003)

    Given the uncertain results from sequencing, cancer researcher Ronald DePinho of Harvard University pushed for analyzing gross genetic changes that are relatively easy to detect and known to lead to cancer, such as extra copies of genes and chromosomal translocations. “That might be the quickest way to get the most bang for the buck,” DePinho says. Other workshop participants suggested that mutations in regulatory regions—which determine how much of a protein is produced—could prove even more important than coding regions. And some urged using emerging technologies to take a closer look at epigenetics, such as changes in DNA methylation patterns that affect whether genes are turned on or off.

    Scientists also debated exactly where to begin the pilot. Some argued that an in-depth analysis of one type of cancer would be more likely to hit a home run—for example, find a mutation that flags which patients would benefit most from a particular treatment. But others argued that studying several cancers would boost the odds of a treatment breakthrough and keep more patient advocacy groups on board. “We need deliverables,” said Bruce Stillman, president of Cold Spring Harbor Laboratory in New York.

    All this testing will require large amounts of tumor tissue with reliable clinical information attached and proper consent from patients. Because genes expressed in tumors change over time, scientists may need to test tumors at different stages. Attendees pondered whether to collect new samples over several years or to rely on existing tissue banks, assuming the researchers who collected them are willing to share. They also worry about community resistance to making the data freely available quickly. That provision may give some cancer researchers “the heebie-jeebies,” said one speaker.

    Despite the challenges, workshop participants agreed to start by focusing on a few tumor types drawn from existing samples. Tissue banks will be invited to participate later this year, and the best proposals will determine which cancers to study. Meanwhile, the pilot will also begin developing methods for collecting new samples for later—and presumably cheaper—analysis. Other requests for applications will seek proposals for technologies, both high-throughput sequencing at genome centers and small-lab techniques such as microarrays for expression analysis.

    To succeed, proponents will need lots of friends from a research and advocacy community that may have doubts not only about the project's eventual price tag but also about the value of fishing for data rather than investigating a hypothesis. “We have a lot of questions,” says Fran Visco, president of the National Breast Cancer Coalition in Washington, D.C., which is still studying the idea. “How are we going to prioritize so it's not creating data to keep scientists busy and not really helping patients?” That's one of many concerns scientists must address to make the cancer genome project a success.


    Strong Personalities Can Pose Problems in the Mating Game

    1. Elizabeth Pennisi

    A closer look at confrontational behavior in various animals shows that aggression may help individuals survive, but it can impair reproductive success

    For male fishing spiders, courtship is dangerous business. Females of the species are notoriously aggressive, and the male—which signals his arrival by gently tapping the surface of the water—often ends up as a meal rather than a mate. Yet each time the female eats her would-be partner, she lessens her chance of reproducing, leaving evolutionary biologists wondering just why this behavior persists. Aggressive female spiders just can't stop themselves, says J. Chadwick Johnson, a behavioral ecologist at the University of Toronto, Scarborough.

    Johnson is among a small group of researchers investigating the “personalities” of animals from spiders and fish to insects and birds. Although many biologists once strongly protested attributing human qualities such as personalities to animals, more and more investigators are adopting such descriptive language. Individual animals, even simple invertebrates, do have consistent behavioral quirks that endow them with discernible dispositions, says Andrew Sih, a behavioral ecologist at the University of California, Davis.

    Although he and his colleagues think of these dispositions as personalities, they have tried to steer clear of being criticized as anthropomorphic by instead coining the term “behavior syndromes.” In addition to identifying such syndromes in animals, Sih, Johnson, and several other investigators are finding that animal personality traits, such as being bold toward potential predators or aggressive toward cohorts, can have drawbacks, despite the traits' apparent value, say in hunting or defending territories. For example, Renee Duckworth of Duke University in Durham, North Carolina, has shown how one bluebird species' aggressiveness allows it to steal habitat from another—yet that same trait impairs the bird's reproductive fitness in certain conditions. Looking at animal personalities, and the good and bad they bring, represents “an important paradigm shift in our approach to the evolution of behaviors,” says Duckworth.

    Eight-legged dominatrix.

    A female fishing spider devours her suitor.


    Dangerous liaisons

    Many researchers credit Sih for bringing to prominence the idea that animal personalities carry survival risks. The notion plays off a proposal made 25 years ago by the late paleontologist Steven J. Gould and geneticist Richard Lewontin, both from Harvard. At that time, the two stirred up the evolutionary biology community by arguing that maladaptive traits could persist if they were linked with beneficial ones in an often-precarious balancing act. For example, guppies living around predators reproduce as early as possible so as to pass on their genes before being eaten. But the eggs slow gravid females down, making them easier prey earlier in life, a finding that lent credibility to Gould and Lewontin's idea.

    Now, by showing that a personality trait that is counterproductive in one context perseveres because of its utility in another, Sih is moving Gould and Lewontin's ideas “into a new arena,” says evolutionary ecologist Andrew Hendry of McGill University in Montreal, Canada. Sih argues that because some animals are very limited in their ability to moderate their personalities according to particular situations, they are stuck with the consequences throughout their daily lives.

    Take the North American fishing spider, the subject of Johnson's studies. In 1997, Göran Arnqvist of Uppsala University in Sweden and a colleague suggested that aggressive females who eat males who come courting were simply following their strong instincts to catch prey. The drive to hunt would serve juvenile females quite well, enhancing their growth, particularly when competition for food was intense. But those instincts, if unfettered, may backfire when the females become adults and need mates.

    No love lost.

    A female water strider struggles to get a male off her back.


    Johnson has recently followed up on this proposal, verifying key elements. He found that even as young spiders, certain females were aggressive hunters, spending more time than their cohorts searching for the next meal and, as a result, bulking up more. This aggressiveness was also reflected as boldness in encounters with predators, Johnson discovered when he mimicked a bird's approach by tapping the water near these spiders. Although all fishing spiders dove into the water when they detected such tapping, the female superpredators surfaced more quickly.

    These daredevils also were more likely than less aggressive females to try to snack on males, Johnson reported last month at Evolution 2005 in Fairbanks, Alaska. “Boldness to a simulated predator is proportional to the tendency to attack males,” he said. Overall, he concluded, the bold, aggressive female spiders ate more food, but they compromised their survival and productivity by treating males as food and taking predation risk lightly.

    Daniel Promislow of the University of Georgia, Athens, is surprised that aggression can pervade all aspects of a female spider's life. If the fishing spiders could modulate their personality, he explains, then the females should be as aggressive as possible in hunting, less aggressive in the face of danger, and mild-mannered when approached by males—but that's not what the experiments indicate. “We often think of behaviors as relatively plastic traits compared to morphology, physiology, or life history,” he says, but Johnson's results challenge that premise.

    Counterproductive aggression is not limited to female arachnids. Sih has found that militant males are the troublemakers among insects known as water striders. Sih graded aggressive tendencies in males by observing, for example, how much they fight, how long they were active, and how often they chased after potential female mates. He then put together 12 groups of water striders, each consisting of males with similar personalities from least aggressive to most aggressive, in separate artificial ponds. The researchers then put females into the ponds and monitored each group's mating successes and failures, keeping track of each individual's partners within their group. The investigators also tracked each water strider's feeding and tallied how often an individual retreated to riffles, supposedly a more dangerous habitat but also a refuge from aggressive peers.

    Females tended to avoid the most aggressive males, the researchers found. Indeed, females often refused to put up with any “Rambo” male in their midst and moved as far away from him as they could, diminishing both his and his peers' mating opportunities. Aggressive individuals couldn't turn down their swagger. They ultimately “hurt not only themselves but, by being too aggressive, the entire group,” Sih reported at the evolution meeting.

    Buzz off.

    A test of aggression shows that western bluebirds are quite fierce against swallows.


    Group dynamics

    Working with small fishes called three-spined sticklebacks, Alison Bell of the University of Glasgow, Scotland, has found that living conditions may narrow the range of personalities within a group of animals. Whereas researchers such as Sih and Johnson typically focus on the behavior of individuals in a population, she is assessing variation in “in your face” behavior—the combination of boldness and aggression—between and within whole populations of the fish. Because stickleback populations have diverged genetically, so might their behavior in different places, she hypothesized.

    To examine this possibility, Bell collected groups of 20 juveniles from 13 different populations of freshwater and marine sticklebacks in various lochs and harbors around Scotland. Some of these populations regularly faced predators—pike, trout, and the like—and others lived in relatively predator-free environments. To measure boldness of the fish from each population, she set up a tank with a pike behind a glass divider, then counted how often individual fish approached the pike to inspect it. For a gauge of aggressiveness, she counted the number times a fish isolated in one tank tried to nip at other sticklebacks in an adjoining tank separated by glass.

    The fish within each of the 13 populations seemed to share similar mindsets. Bell found that when one individual from a population fearlessly approached the pike, so did most of the others from those groups. In general, most of the fish within a particular group acted the same way, she reported. And fish from the boldest populations, as measured by the pike test, were also the most confrontational toward other sticklebacks. In the wild, says Bell, this bullying could translate into bigger territories, better food, and even increased mating for the biggest bully. But the fearlessness toward predators may also cost fish in these aggressive groups their lives, suggesting that whole groups of animals, not just individual ones, can have personality traits that threaten reproductive success at times.

    Bell also observed that the bold, aggressive stickleback populations had higher breathing rates, more spines, and heavier body armor than more wimpy populations. Those correlations suggest that “behavioral syndromes might be part of a larger package of evolutionary [traits],” says Sih.

    Duckworth's studies indicate that sometimes the bold personality of one species can help it beat out similar, but shyer, species, at least in a particular environment. Observations over the past 40 years show that western bluebirds have greatly expanded their range in Montana, displacing mountain bluebirds. By tallying the number of each bluebird in places where both species are present, Duckworth documented that western bluebirds in just a few years supplanted mountain bluebirds at valley study sites. Much of the western bluebird's success sprang from its fierceness, suggests Duckworth.

    She placed tree swallows, a bluebird competitor, in nest boxes, and then watched as either of the bluebird species approached the box. She found that western bluebirds were more aggressive, an indication that they are better able to acquire and defend their territories against the swallows. The male western bluebirds also were fiercer than mountain bluebirds when competing for mates, another sign of pushy temperaments.

    In this case, aggressiveness seems to go hand in hand with reproductive success. But a closer looks suggests that, as with fishing spiders and water striders, the western bluebird's obnoxiousness can come with a cost. Duckworth points out that western bluebirds spend so much time defending their nests and courting that they neglect their offspring. This poor parental behavior is especially problematic in tough environments, such as mountains. In contrast, mountain bluebirds are loyal parents and have an edge where weather can be rough, says Duckworth. As a result, they have maintained their foothold in Montana's mountains. “Behavioral syndromes can have profound ecological and evolutionary consequences by mediating species coexistence,” Duckworth says. Thus, in animals, as in people, personality can make or break one's success in life.

  15. The Hunt for a New Drug: Five Views From the Inside

    1. Jeffrey Mervis

    The world of drug discovery in big pharma can seem pretty mysterious to outsiders. But some patterns are visible from the inside


    Waiting for his lunch to arrive, Graeme Bilbe wants to make sure that the reporter on the other end of his cell phone understands how hard it is to discover a new drug. The U.K.—born, Basel, Switzerland—based head of global neuroscience research at Novartis is dining in southern California with a former pharma colleague, Tamas Bartfai, now chair of the department of neuropharmacology at the Scripps Research Institute in La Jolla, California. The hors d'oeuvre is a lecture on the industry's staggering attrition rates.

    “How many ideas do you think you need [to develop a drug]?” demanded Bilbe, who's been with Novartis since 1989. “Take a guess. One thousand? Ten thousand? You need at least that many, if not more. The chances that any of those ideas will ever become a drug are vanishingly small.”

    Those mind-boggling numbers color everything about research in big pharma and make this research sector distinct from any other area of industrial research. Very few pharma scientists actually work on products. Instead, the vast majority toil at a much more basic level, looking for potential targets, synthesizing compounds that might act on those targets in a way that would be therapeutic, and then making the compound “druggable.” “I have never worked on a successful drug,” confesses Derek Lowe, a medicinal chemist with 16 years in the industry who writes what may be the only Web log (blog) dedicated to pharmaceutical research ( “Heck, I haven't worked on anything that anybody with a disease has ever put in their mouths.”

    The research environment has also been reshaped dramatically in the past decade or so by mergers, which can abruptly shift a researcher's focus onto a whole new area of study. And unlike research on a new computer chip or a more efficient engine, the output from pharma research labs is not so easy to measure (see p. 726).

    The world of big pharma research is shrouded in a culture of secrecy that goes well beyond the specific compounds and targets a company is working on. Here's how an otherwise candid Lex Van der Ploeg, head of Merck's new research lab in Boston (see sidebar, p. 723), puts it when asked about his productivity goals. “If I told you that this lab was going to generate, say, eight lead candidates this year, then our competitors could look at the number of people we employ and figure out how many people it takes us to develop a candidate compound,” he says. “Then they would compare it to how many it takes them. And if we're lower, they'd try to figure out why, and what they can do to become more efficient. That would give them a competitive advantage.”

    Knowing where they stand is an all-consuming interest for pharma executives. As a result, investment analysts and corporate consultants churn out reams of reports each year on industry trends, from early-stage alliances with biotech companies possessing intriguing compounds to the latest technology “platforms” that can improve efficiency. The documents are sprinkled liberally with breathless predictions about how these trends “will change everything.”

    Most of the time they don't, of course. In the meantime, however, these big-picture studies provide little idea of what the view is like from inside the industry's labs. From the scores of scientists and research managers we interviewed for this special section, we have chosen five individuals whose stories provide glimpses of how those big trends trickle down to the labs and computer workstations around the globe that represent ground zero in the hunt for new drugs.

    A view from the bench: Change as a constant

    Eric Gulve joined big pharma in 1993 in hopes of ameliorating the ravages of diabetes. A research assistant professor at Washington University in suburban St. Louis, Missouri, Gulve became part of a team at G. D. Searle (the pharmaceutical arm of Monsanto) that was just beginning to tackle insulin resistance in type 2 diabetes. Since then, he's worked on cholesterol metabolism for Monsanto, cardiovascular diseases for Pharmacia, and then diabetes again for Pharmacia. Today he's with Pfizer, seeking potential targets to treat two forms of cardiovascular disease, thrombosis and hypertension.

    Although he's worked for three companies in 12 years, Gulve is no hired gun. He hasn't even changed his commute to work. Rather, the 46-year-old physiologist has spent his entire pharma career in the same four-story industrial lab in the St. Louis suburb of Creve Coeur. The job changes were the result of three corporate mergers, culminating in Pfizer's $53 billion acquisition of Pharmacia in April 2002. Those mergers triggered top-down reviews of existing research, followed by projects or entire areas of therapeutic research being cancelled or transferred to another site. During one gutwrenching transition, Gulve spent weeks interviewing scientists for a revamped department—without knowing whether he would be their boss or even if he would still have a job with the new company.

    Growth industry.

    Today's Pfizer is built upon a decade and more of dealmaking.


    There's no way to know if Gulve's career path is typical. Some scientists remain at one company their entire lives, and others switch jobs often and voluntarily. But mergers have clearly changed the landscape of big pharma in the past decade. Gulve's current employer, with $52 billion in sales last year, has become the industry's leader thanks to its ingestion of Warner-Lambert in 2000 and Pharmacia, each of which in recent years had swallowed smaller fish such as Parke-Davis, Upjohn, Monsanto, and G. D. Searle.

    “I'm not complaining about any of the decisions that were made,” he says. “But it is frustrating when you've worked so long and hard on a project and still haven't gotten far enough along to know if your hypothesis is right or wrong. I know that mergers are part of the business. But I hope that I never have to go through another one.”

    A view from a loyal critic: The art of drugmaking

    Derek Lowe may be unique in the pharmaceutical industry: He's a medicinal chemist for a big pharma who writes a blog on drug discovery. His column ( is an irreverent look at the industry. It's filled with pinpricking commentaries on the latest clinical results, corporate reshufflings, and overhyped trends in the business. He's not embarrassed to describe his own failures, either, including an on-again, off-again attempt to test a hypothesis that stubbornly resists verification.

    His daily musings generate 25,000 hits a month. That traffic feeds Lowe's need for an audience, a hunger that offsets the lack of payment for his labors. “It hasn't helped my research,” he confesses about the blog, which he started in 2002. “But it's given me a much broader perspective on the business.” His readers are both colleagues—“insiders write me about how they've tried the same things in their labs that I write about”—and outsiders with a voyeuristic streak. “Where else would I get to hear from people saying, 'When I took that drug you wrote about …' I've also done some historical reading about the fashions that sweep through the industry and the fact that most of them don't pan out.”

    A 1988 chemistry Ph.D. from Duke University, Lowe wanted to teach at a small liberal arts college but couldn't find the right job. Answering a job ad has led to a career in industry that he says “has worked out pretty well.” He currently works for Bayer but goes to great lengths to separate his dual identities as a researcher and blogger.

    Lowe doesn't hesitate to point out the foibles of the pharmaceutical industry. “We're not angels. And when we mess up, I say so. If I was rah-rah all the time, nobody would read me.”

    Even so, he's as dedicated to improving human health through modern drug discovery as any pharma bigwig. Taking umbrage at a recent story in Business Week entitled “Biotech, At Last” that paints academic research as nimble and pharma science as hidebound, Lowe writes: “It's true that many of the basic discoveries that have led to the current biotechnology industry came from academic research. That's just as it should be. But none of it would have been turned into human therapies without corporate research and development.” And he's personally offended at the article's characterization of pharma's relationship with biotech in the 1980s and 1990s. “Shied away from biotech for years? We pumped uncountable billions into it, much of which we never saw again.”

    A career-eye view: A taste of industry

    What's a postdoc doing in pharma? Scottishborn David Dornan has spent nearly 3 years at Genentech, which has a 30-year-old policy of seeking out promising young scientists to pursue basic research. And although the company has a rule that its postdocs don't move into permanent positions, Dornan sounds like someone whose career aspirations may have been altered by working at the South San Francisco, California, biotech giant.

    “My future? I think of it every day,” says the 27-year-old Dornan, who earned his Ph.D. in molecular oncology at the University of Dundee, U.K. “And the longer I'm here, the more difficult it is to envision becoming an academic.”

    A member of a team led by Genentech's head of oncology V. M. Dixit, Dornan was a co-author of papers in Science and Nature last year that describe the group's work on how cancer-related proteins are degraded by the ubiquitin system. And although the work is fundamental science, Dornan has also been bitten by the drug discovery bug. “We found something that could be a therapeutic, and we have a unique chance to put it into development. It depends on the next phase. And if it works, we'll be handing it off to the chemists. The point is that it's possible.”

    Dornan is realistic about his chances of staying on the West Coast. “California is great, but you have to be willing to go where there's a job.”

    A view from a distance: Landing on her feet

    When Myrlene Staten saw the job ad in the New England Journal of Medicine in 1989, she thought it could have been written just for her. “Roche wanted a junior faculty member with clinical experience in metabolic diseases,” she recalls. Her work as an endocrinologist at Washington University in St. Louis made her a perfect fit, she realized, and before long she had moved from Missouri to New Jersey to help the company develop drugs for obesity and diabetes.

    It was the start of a 15-year odyssey through big pharma that she recalls with mixed emotions from her current post at the National Institute of Diabetes and Digestive and Kidney Diseases in Bethesda, Maryland, where she runs a program to encourage academics and small companies to develop new therapies for type 1 diabetes. After 4 years at Roche, she moved to Lederle, which was soon bought by Wyeth. In 1995, she headed out west to Amgen, where she was part of the team doing the company's first clinical trials on the protein leptin, once highly regarded as a potential diet drug. Then it was back to the East Coast for a 2-year stint with Bristol-Myers Squibb before joining Upjohn/Pharmacia, where she was head of metabolic diseases.

    A 2000 merger with Searle resulted in a spinoff of the new company's metabolic diseases portfolio to a new biotech based in Stockholm, Sweden. But Staten found a way to stay in New Jersey. “Searle had a cardiovascular group, and it had an opening. So when the boss called me in and asked me how I felt about working on cardiovascular diseases, I said, 'Real good.'”

    Two years later, Staten had to handle another merger, this one with Pfizer. After helping the new company analyze its combined portfolio—“you present your project to senior managers, who then go behind closed doors and come out months later with a list of what stays and what goes”—she was faced with finding a new position. Choosing a very different direction, Staten jumped back into the nonprofit world, landing at the National Institutes of Health.

    She's lost none of her zeal for finding new medicines that can help people. But she's 15 years wiser about how difficult that is, and how failure is a much more likely option. “My goal, then and now, is to develop a drug that can achieve a 20% permanent weight loss with no unusual side effects. But I think maybe I'll have to leave that to the next generation.”

    A view from the executive office: Finding his niche

    At 54, Bob Stein is a drug industry veteran who has done it all, including two stints with biotech. After several coast-to-coast moves, he says he's exactly where he wants to be: president of Roche's research lab in Palo Alto, California.

    A graduate of a joint M.D./Ph.D. program at Duke University in Durham, North Carolina, Stein came to Merck in 1981 along with Edward Scolnick, who later became the company's legendary research chief. (Scolnick had originally offered him a job at the National Cancer Institute in Bethesda, Maryland, but then signed on at Merck before Stein said yes.) Stein says he found the company's scientific environment “much more exciting” than academic jobs he had been offered. Several years later, as head of pharmacology at Merck, “where I worked on some pretty good drugs,” he was dispatched one day to San Diego, California, to assess a possible collaboration with a company, Ligand Pharmaceuticals, that “had great science but no infrastructure for drug discovery.” After recommending that Merck walk away from the deal, Stein was headhunted to become Ligand's chief scientific officer.

    All for one.

    Drug discovery puts a premium on teamwork, says Roche's Bob Stein.


    Within a year, Ligand had raised $250 million in a public offering, and Stein had negotiated eight collaborations with big pharma. But the grueling schedule—including talks at 13 investment meetings and 250 business presentations—and the amount of work it took to “get other people to do what needed to be done” at the small company led him to embrace an offer from a mentor to return to East Coast pharma. The job was as head of research and preclinical development at DuPont Merck Pharmaceuticals, a joint venture of the two companies.

    “I'd have been happy to stay there, too,” Stein says about his 6 years there. But DuPont decided that the joint venture was chewing up too much of its research budget, he says, and after Bristol-Myers Squibb bought the company for $7.8 billion, “I didn't like what I saw.” So he jumped to Incyte Corp., where he spent 2 years as president and chief scientific officer before joining Roche in 2003.

    Stein clearly loves the horsepower of a big pharma and enjoys the chance to apply what he's learned over a quarter-century about drug discovery. But greater capacity means a greater chance to fail, too. “The goal is to develop superior medicine,” he says. “But the process includes a million handoffs, and any dropping of the ball could be potentially devastating.”

  16. Boston Means Business for Drug Companies

    1. Jeffrey Mervis

    BOSTON, MASSACHUSETTS—Asked why he robbed banks, Willie Sutton is said to have responded: because that's where the money is. After more than a century, big pharma is following that logic by setting up shop here amid what may be the world's largest concentration of biological brainpower.

    Two of the world's biggest drug companies, New Jersey-based Merck and Connecticut-based Pfizer, have opened small outposts to supplement their global R&D networks and put company turnaround artists in charge of them. A third pharma, Novartis, has gone even further by relocating its main research facility, the Novartis Institutes for Biomedical Research (NIBR), in a spectacularly remodeled former candy factory and two other buildings adjacent to the Massachusetts Institute of Technology (MIT) and picking an industry novice to run it. The 1000-strong scientific work force assembled in the past 2 years represents a serious bid by the Swiss-based company to find the sweet spot in drug discovery.

    “Our kickoff career fair attracted more than 2000 people, and it was a fabulous opportunity to meet and greet leading scientists and business leaders,” says Lynne Cannon, vice president for human resources at NIBR. “That would have been difficult to do in Groton [Connecticut, the site of Pfizer's largest lab] or Princeton, New Jersey.”

    Boston may be the cradle of American independence, and Cambridge the home of the country's oldest and most prestigious university, but until the past few years the region wasn't even on the map of big pharma. Area academics with backgrounds in molecular biology had formed many biotechnology companies, some of which aspired to become the next big pharma. However, the nation's chemical-based drug industry was confined to the mid-Atlantic region and the Midwest.

    Pfizer made the first move in 1999, opening up a Discovery Technology Center in Cambridge that offered the latest technology to drug discovery scientists throughout the company. Last year, officials expanded the center's mission to the entire pipeline of drug development and plucked Phil Vickers from the company's ranks to run it. A 45-year-old biochemist who enjoys a challenge and a change in scenery, Vickers was born in England, received his Ph.D. at the University of Toronto, and did a postdoc at the National Institutes of Health in Maryland before joining Merck's Frosst laboratory in Montreal in 1988. He came to Pfizer in 1994 and earned his stripes in a series of management posts on both sides of the Atlantic.

    Perched on the edge of the MIT campus, the renamed Research Technology Center aims to satisfy Pfizer's need for technological support by mixing in-house expertise with the skills of local academics and start-up companies. Vickers says his youthful but growing shop—he plans to add 25 scientists to the current 110-person roster by the end of the year—“offers the attributes of a biotech with the resources of a big pharma.”

    Despite running an operation almost 10 times the size of Pfizer's, Mark Fishman describes NIBR in similar terms. A molecular cardiologist who had pioneered the use of zebrafish for gene discovery at Harvard Medical School (HMS) and Massachusetts General Hospital, Fishman is hoping to “functionalize the genome” by applying it to diseases where the biological mechanism is already understood. The lab's location—the region has supplied more than half the institute's talent, not to mention an ever-widening network of academic collaborations—provides an added boost, he says.

    Already, Fishman has raided HMS to find global heads in cardiovascular research and modeling disease. He's also tapped biotech and pharma for chiefs in oncology, molecular pathways, and discovery chemistry, luring them with the prospect of painting on a fresh canvas. “We're getting who we want, and almost nobody has left,” he crows.


    Novartis converted a candy factory for its main research institute.


    Across the Charles River and adjacent to Boston's medical complex sits Merck's Edward M. Scolnick Research Laboratory. Named in honor of its former research chief, the new 12-story, glass-faced lab opened last fall, and its site head, Lex Van der Ploeg, is busily recruiting talent. Van der Ploeg, 50, a specialist in infectious diseases who joined Merck in 1991, took on the challenge after a year spent shifting the focus of Merck's San Diego facility from neuroscience to stroke. Soon after he left, however, corporate officials decided to shut the lab and shift some resources to other sites.

    His mission is to rev up the company's efforts in developing treatments for cancer, obesity, and Alzheimer's disease. He expects to double the size of the basic research team, now 140, by 2007, beginning with oncology and then moving into the neurosciences. “The proximity to talent is terrific, and our success rate is about 90%,” he says about current recruiting efforts. About a quarter of the scientists have migrated from other Merck labs.

  17. It's Still a Man's World at the Top of Big Pharma Research

    1. Jeffrey Mervis

    For a few years after their company was acquired by Wyeth in 1995, molecular biologist Abbie Celniker and several female colleagues at Genetics Institute in Cambridge, Massachusetts, hoped that the new management might boost their careers. But eventually they came to the opposite conclusion. “There was an established culture [at Wyeth] that said it would be harder to influence our peers…. Simply put, we didn't see a career progression unless we learned to play golf and use the men's room.”

    What Celniker, now senior vice president for strategic research at Millennium Pharmaceuticals in Cambridge, had sensed becomes obvious by looking at the leadership rosters of the research divisions of big pharma: Drug discovery is a man's world. Not one of the chief scientists or heads of research at these companies is a woman. The precious few senior women executives with science Ph.D.s or M.D.s are most often found on the development/business side of the company or holding corporate posts without line responsibilities.

    Why that's the case, however, is much less clear. Ask a man and you're likely to hear that the industry is no different from the rest of society. Then he'll note that his company is very concerned. “It's a tough issue that I think about a lot,” says Jonathan Knowles, head of global research for Roche. “I'd like to understand it better.” He'll also say that things are getting better.

    Ask a woman—who by definition has not made it to the top—and her answer will be quite different, although equally nuanced. “The forces keeping women scientists down are more psychological and cultural than legal,” says Joanne Kamens, a project team leader at Abbott Bioresearch Center in Worcester, Massachusetts, and president of the state chapter of the Association for Women in Science. “People still have a problem seeing women as leaders rather than as caretakers and mothers. Men who decide to spend more time with their families also tend to be seen as weaker. But at least they have the option. If the father can't help out at home, it falls on the women.”


    Novartis's Lijun Wu makes room for both career and family.


    Lijun Wu, a 41-year-old unit head within the cardiovascular group at the flagship Novartis Institute for Basic Research (NIBR) in Cambridge, Massachusetts, remembers being asked as a graduate student if her decision to get married meant that she planned to drop out of the program to have a family. Several years later, after becoming pregnant with the first of her two children, colleagues told her that her bosses at Millennium were wondering if she'd return after giving birth. “My career was going well, and they didn't ask me directly. But I think it's unfair; they wouldn't have wondered that about a man.”

    Wu doesn't understand why any employer would care whether she even has a family. But most pharma executives acknowledge that family responsibilities do matter. “One possible reason [for the dearth of women] is that any senior position requires a huge commitment,” says Knowles. “It would be difficult for someone to do that type of job while also looking after a home and small children.”

    Amgen's research chief Richard Perlmutter offers similar thoughts. “I'm reluctant to generalize about gender differences,” he says. “At the same time, you can't get around the fact that the burden of early child rearing may be a career breaker [for some women]”

    That burden can show up in subtle ways, notes Lynne Cannon, vice president for human resources at NIBR. “It's not just a question of having the door open to women,” she says. “Sometimes it's about how the door gets opened. If I can't stay until 8 p.m.—when a lot of decisions get made—because I have to pick up my kid at 6 from daycare, then I may miss out on something important.”

    Many pharma companies have recently begun to identify and assist women scientists who want to move up the corporate ladder. Novartis has a “women to watch within the lab” program, Cannon says, to provide ongoing career guidance and support for outstanding women. “Mentoring is great,” says Cannon, “but there's a danger if you attach yourself to one person and that person leaves.” Although that's true for men, too, the dearth of women makes any loss of support costly.

    Wyeth has a similar program for top-performing women, says Robert Ruffolo, president of research and development, that's modeled on a gender-blind program for the top 1% of its researchers. Gail Cassell, vice president for strategic planning for Eli Lilly, says that the Indianapolis, Indiana-based company offers a variety of programs for women scientists, from tips on how to ask for a promotion to networking with colleagues in other fields.

    None of the programs has run long enough to accumulate meaningful data, however. And it's not clear that company executives have thought in much detail about what they want to achieve. “We don't know what enough is,” Ruffolo admits. “But we consider it a win as long as we're attracting more women and minorities each year than are leaving the company.”

  18. Productivity Counts--But the Definition Is Key

    1. Jeffrey Mervis

    With costs soaring, every company says it's becoming more efficient. But what exactly does that mean?

    For all but a tiny fraction of big pharma scientists, their work isn't really about discovering new drugs to cure disease and improve human health. It's about looking for druggable compounds: molecules that might bind to targets that could block or enhance a biochemical process that leads to a particular pathological state or impairment. And success isn't measured by how much they have contributed to a drug or therapeutic medicine on the market. Rather, it means “hitting your numbers,” that is, achieving a preset goal of “deliverables”— be they compounds, animal data, or patients—that argue for moving along to the next step in the process.

    Trouble is, that approach is hugely inefficient. The current cost of discovering and developing a new drug may be as high as $1.9 billion, according to an extrapolation by Joseph DiMasi of the Tufts University Center for the Study of Drug Development in Boston, Massachusetts, whose 2001 report pegging the number at $802 million was based on medicines that entered clinical trials as long as 20 years ago. Lowering that number is the current Holy Grail of the industry. “Productivity is our biggest challenge and the number one topic of conversation among my colleagues,” says Steven Paul, president of Lilly Research Laboratories and the top scientist at the Indianapolis, Indiana-based drug giant.

    But consensus on the goal doesn't mean agreement on how to get there. Big pharma management features a multiplicity of organizational models, all aimed at achieving greater efficiency. Some companies such as Pfizer are highly centralized, whereas others pride themselves on having small, semiautonomous units. “Pfizer is probably at one end of the spectrum. Everything related to drug discovery has to go through either New London [Connecticut] or Sandwich, U.K.,” says industry analyst Roger Longman, co-managing partner of Windhover Information Inc. in Norwalk, Connecticut.

    The wrong direction.

    The industry's overall record of success in testing drugs in humans has declined in recent years across each step of the process.


    At the other end, he notes, is U.K.-based GlaxoSmithKline (GSK), second in global pharmaceutical sales to Pfizer. Under the leadership of research chief Tachi Yamada, GSK has created Centers of Excellence in Drug Discovery around the world in six therapeutic areas, plus one center for biologics. Each has its own budget and hiring authority. “I wanted them to be small, and studies show that you can know the names of 300 people but no more,” says Yamada. The centers “have total control of their budgets and hiring. But they still have targets.”

    Falling somewhere in the middle is a “hub-and-spokes” system that Roche follows that allows its corporate headquarters in Basel, Switzerland, to keep tabs on research sites in the United States, Europe, and China. And although that arrangement can mean 2 a.m. teleconferences for Bob Stein, who oversees 1100 people at Roche Palo Alto, California, he says it's vastly preferable to having “one big R&D operation that, like a 10-foot spider, has outgrown its body plan.”

    There are also many views on which metrics are the most meaningful, and if metrics can even take you where you want to go. One popular view, espoused by Pfizer CEO Hank McKinnell and others, embraces “shots on goal.” That's the belief that more compounds going into clinical trials translates into more successful outcomes and, ultimately, more marketable drugs.

    But what kind of shots are most important? For Yamada, the key metric “is not the number of targets validated, or the number of chemicals selected. It's proof-of-concept in patients.” His counterpart at Novartis, Mark Fishman, puts it even more bluntly. “[A drug candidate] is not a success until we've treated a patient with it.”

    At New Jersey-based Wyeth Pharmaceuticals, which sits on the centralized end of the management spectrum, R&D president Robert Ruffolo has done a scientific analysis of the science of drug development. A 55-yearold pharmacologist and 28-year industry veteran, Ruffolo likes to say that “we've got numbers on everything.” And since coming to Wyeth in 2000, Ruffolo has probably gone further than any other pharma honcho in trying to quantify what his researchers should accomplish at each stage of the process.

    “Some people say that they can pick winners,” Ruffolo told a meeting of pharma scientists gathered this spring in Washington, D.C. “But I believe that it's still a crapshoot. I can't pick winners, and after 30 years in this business, I haven't met anybody who could.”

    What Ruffolo can do, he says, is ride herd on the factors that he can control. Hence his insistence on production targets that take attrition into account and, if met, would allow for a sufficient flow of new compounds through the pipeline. Raises are based on achieving the goals, and it's all computerized.

    The magic numbers for Ruffolo are 12, 8, and 2. That's a three-link chain of the annual number of compounds entering development, the number of investigational new drugs entering clinical trials each year, and the annual number of new drug applications submitted to the U.S. Food and Drug Administration. He says that his approach has helped turn around what he calls the company's “pathetic” track record of submitting new drug applications in the years before he arrived. And best of all, it's proven to be sustainable: Wyeth has met the targets every year since 2001, he says. “That's the most important point. It's a steady-state model.”

    Ruffolo admits that approach didn't win him any popularity contests at Wyeth. “Scientists hate this approach,” he says. “When I was a scientist, we used to say that you can't manage science. But it needs to be.” Those who didn't buy into the approach left the company, he says—and those who have remained appreciate knowing where they stand.

    Richard Scheller takes a very different approach as executive vice president of research at Genentech, which has eschewed large acquisitions and does all research at its ever-expanding South San Francisco, California, campus. A neuroscientist and former Howard Hughes Medical Institute investigator at Stanford University, Scheller came to Genentech in 2001 after deciding that its culture meshed with his own philosophy of doing science. Genentech's corporate strategy, labeled Horizon 2010, does include research goals for its more than 600 scientists over the next 5 years. But although they specify the number of new products to be moved forward for each of the company's three major therapeutic areas, some goals omit key steps in the process. And they aren't linked together in a formal manner.

    Sitting in a top-floor office overlooking San Francisco Bay—and the pier that was allegedly the favorite fishing hole of cofounder Herbert Boyer—Scheller describes an ongoing study of Genentech's attrition rate and the nature of its pipeline in a way that suggests he doesn't view it as quite the priority that Ruffolo does. “It turns out that different types of projects fail for different reasons,” notes Scheller, who says that he “doesn't know very much about big pharma” despite the fact that, based on the value of its stock, Genentech is the fifth-largest drug company in the world.

    “For example,” Scheller says, “I'm expecting small-molecule throughput rates to be lower than for protein therapeutics. I'm also leading a project to understand the bottlenecks. And I think that they will turn out to be what you'd expect: Some projects will be underresourced, some will suffer from poor internal communications. When we're finished, we'll react appropriately. But I suspect that when we f ix one problem, some other bottleneck will appear.”

    Don't be fooled by that dispassionate tone, however. Scheller isn't afraid to be just as hard-nosed as Ruffolo in assessing the performance of his troops. But he doesn't plan to do it from a spreadsheet. Knowing how to maintain a healthy pipeline, he says, “is more or less a matter of intuition.” And the most important thing about dealing with scientists, he says, “is to be clear about the reasons for your decision [for killing a project or shifting resources]. I'm not always going to be right. But I've earned a lot of respect from my credentials at Stanford and my achievements as a scientist.”

  19. I See You've Worked at Merck …

    1. Jeffrey Mervis

    Senior hires at Amgen demonstrate how one company's loss can be another's gain

    In 2000, Merck CEO Raymond Gilmartin came down from New Jersey to Washington, D.C., to extol the value of research partnerships involving the government, academia, and industry. In a talk marking the centennial of the Association of American Universities, Gilmartin mentioned two executives he hoped would help the company set up a research lab in Boston (see sidebar, p. 723) to tap its rich talent pool: Roger Perlmutter, head of basic research, and Ben Shapiro, his predecessor and current head of external research. He also noted that Merck's success in the development of the first protease inhibitor to treat AIDS, Crixivan, by a team led by chemist Paul Reider, rested on the government's long commitment to basic biomedical research.

    Fast-forward 5 years—a generation in big pharma—and none of the four team members still works at Merck. In May, Gilmartin suddenly stepped down earlier than expected, a casualty of the company's voluntary withdrawal last fall of Vioxx, its arthritis painkilling COX-2 inhibitor pill. Shapiro had retired in 2003, in keeping with company policy for executives who reach age 65.

    Perlmutter and Reider remain very active in the drug business. But they now work for Amgen, the southern California biotech giant that industry wags have dubbed “Merck West.” Their migration is illustrative of the company's role over the years as both a magnet for top academic talent and a fertile hunting ground for competitors. And even as market analysts wonder if Merck can recover from the blow to its reputation from Vioxx and the financial burden of stock shares trading at one-third their 2000 level, several successful alumni say that they retain warm feelings for “Mother Merck.” Several senior research officials at Merck declined comment for this story.

    Coast to coast.

    Amgen's senior research team is led by former Merck scientist Roger Perlmutter (left) and includes several former—and current—colleagues such as Paul Reider (right).


    Shapiro, for one, thinks “it's logical that Merck would help seed the leadership ranks of other companies.” In 1990, when he was a department chair at the University of Washington (UW), Shapiro says he jumped at an invitation from Ed Scolnick, the former head of Merck research, to become head of basic research because “Merck had a reputation for caring about science.” In 1997, Shapiro, in turn, recruited UW immunologist Perlmutter.

    While Shapiro was grooming Perlmutter for the top research job, saying: “He was special. There aren't that many academics who would be good at senior management in big pharma,” Scolnick had other plans. In December 2000, he brought on Peter Kim, a biochemist at the Massachusetts Institute of Technology. Although it would be another 2 years before Scolnick retired and Kim succeeded him, the line of succession was clear. So it was no surprise that in January 2001, Perlmutter was named executive vice president of research and development at Amgen. He says that he had been weighing several career options and chose Amgen because of “the magnitude of its commitment to building up its R&D operation.” He also was attracted to its relative youth—it was founded in 1979—compared with its century-plus-old pharma competitors, and its strength in biologics.

    But Perlmutter wasn't turning his back on Merck. He says he had declined an offer to join Amgen in 1996 because “I wasn't ready. I didn't understand the totality of drug discovery and development enough to have the impact that I wanted to have.” Merck gave him that knowledge; he says: “My experience there informs everything that I do here.”

    One lesson was to tap into his Merck connections. One call went to a former medical and graduate school colleague, pathologist Joseph Miletich, who had spent most of his career at Washington University in St. Louis, Missouri, before Shapiro convinced him that Merck offered “a bigger canvas.” Four years and several promotions later, Miletich heard a similar recruiting pitch from Amgen, which he joined in 2002 as senior vice president for research and preclinical development.

    Not long after, Perlmutter reached out to Reider, who had come to Merck in 1980 right out of graduate school on a mission to conquer dread diseases. Reider found Perlmutter's description of Amgen as an eager teenager full of promise appealing, as well as his pledge that the company would only work on treatments for important and unmet medical needs.

    “I can't get very excited working on the ninth molecule to correct male impotence, or to treat male pattern baldness,” says Reider, Amgen's vice president for chemistry. “I'm somebody who needs to go home every day with a sense that I've accomplished something. I'm also 53, and it would be nice if we could find a treatment for Alzheimer's by the time I need it.”

    Reider says that his “dream job” would be to work with both Perlmutter and Kim, whom he says he “cherishes.” But he's worried that Merck could stumble and lose its way. “Amgen today is where Merck was 15 or 20 years ago. The ability to pounce on an idea and take it into development quickly is so important. Merck still has tons of good people. It would take 30 years to lose that edge. But when you get so big that your chief concern is what products to bring to market in what time frame, that's a warning sign.”

  20. The Brains Behind Blockbusters

    1. Jennifer Couzin

    The inventors of top-selling drugs talk about their unlikely paths to success, and whether today's scientists can pull off similar feats

    How does a scientist hit a home run in the drug business? For Kenneth Koe and Willard Welch, it took curiosity, determination, and a series of lucky breaks. The payoff for their employer, Pfizer, was huge: the antidepressant Zoloft, one of 19 drugs that last year generated more than $2 billion in revenues in the U.S., according to IMS Health, a company that collects and markets health care data.

    Koe, a biochemist, and Welch, an organic chemist, are members of a small, exclusive club of drug discoverers whose labors have helped catapult their companies into the ranks of the world's most profitable. But while aggressive advertising has made household names of drugs such as Lipitor, Nexium, and Celebrex, their inventors remain relatively unknown. Science interviewed nearly a dozen of them to learn the stories behind their discoveries. Although these superinventors identified very different drugs, across oceans and decades, their experiences are more similar than one might expect.


    Zoloft, a drug with sales of $3 billion a year, likely wouldn't exist without the work of Kenneth Koe and Willard Welch.


    For one, few grasped the value of their discovery at the time or anticipated the hurdles standing in their way. Even fewer profited from their accomplishment, a fact that many quietly resent. Some doubt that they would be allowed to pursue the same lines of research today that they chased 15—or 40—years ago. “It was nice and unsophisticated”—and more fun—“in those days,” says Bruce Roth, the Pfizer chemist who, at the tender age of 31, helped invent Lipitor in 1985.

    For many scientists near or at retirement age, the advantages of today's powerful drug-hunting technologies are offset by what they see as a loss of freedom to stretch one's mind around novel ideas. “Too much computer and not enough brain,” grumbles former Merck biochemist Alfred Alberts, who helped invent Mevacor, the first successful statin, as well as its $6-billion-a-year successor Zocor. Strategies to unearth blockbusters today are “not working,” says Alberts, who retired in 1995 after 20 years at Merck. “I think that's fairly clear.”

    Graced by luck

    What distinguishes past generations of drugmakers from the present? One difference is their starting point. Scientists then were often running blind, chasing new therapies without the benefits of modern biochemistry and the clues it can provide. Serendipity often planted the seeds of a new drug. Celebrex, Pfizer's antiarthritis drug, was chosen for further testing over another chemical in what was “more or less a coin flip,” says John Talley, one of its inventors. Talley did the work while at Searle, which later became part of Pfizer. He now heads drug discovery at Microbia, a biotech in Cambridge, Massachusetts.

    But luck was only part of the equation. “Chance favors the prepared mind—is that how the saying goes?” asks Welch, who joined Pfizer's Groton, Connecticut, lab in 1970 and retired 3 years ago. Koe came to Pfizer in 1955 and studied penicillin offshoots before he was transferred to the company's tiny team of central nervous system researchers. After dabbling in potential antianxiety compounds, Koe turned to a then-new concept: the effects of serotonin in depression. He soon roped Welch into working on candidate antidepressants. Within a year he and Koe had hit on Zoloft.

    Protracted search.

    Ludo Kennis, shown working at Janssen Pharmaceuticals in the 1970s, chased several false leads before he hit on the antipsychotic Risperdal.


    One thing that hasn't changed for drug inventors, says Welch, is the need to stay abreast of developments in a given field. This is particularly crucial because, like drugmakers today, researchers are frequently transferred from their area of expertise into a new therapeutic area—say, from endocrinology to cardiology—and are expected to bring themselves up to speed quickly. Inventors recall their obsessive tracking of published chemical structures, patent filings by rival companies, and clinical trials of comparable drugs—even those that fail—as signposts helping point the way to their blockbuster.

    Tenacity also helps. Without it, medicinal chemist Ludo Kennis would never have succeeded in creating the antipsychotic drug Risperdal, which grossed $3 billion last year for its parent Johnson & Johnson.

    The late Paul Janssen, a pharmacologist who founded the company in 1953, wanted an alternative to the antipsychotic Haldol, whose patent was expiring. But competition in the field was fierce. Kennis, who joined Janssen's research unit near Antwerp, Belgium, in the mid-1960s, and his colleagues began with little more than a hunch that blocking the neurotransmitters dopamine and serotonin simultaneously might do the trick. But their knowledge of the underlying neuroscience was thin. It was a lucky goal set “without really knowing that this was a real improvement,” says Kennis.

    Every drug they made was tested in rats to see if it blunted dopamine and serotonin. Three compounds that did so progressed to human trials. “At that time, it was very easy to do a clinical experiment,” says Kennis. Regulations governing human experimentation were far more lax than today's, he notes.

    All three drugs failed. The most promising displayed wildly different metabolic properties in humans from those in rodents. But Kennis took another look at the compound, setoperone, and refashioned it. More than 20 years after he started, the U.S. Food and Drug Administration approved Risperdal.

    Such delays were not uncommon. Merck's multibillion-dollar asthma and allergy drug Singulair entered clinical trials in 1994 after 13 years of sometimes tortuous development, including clinical trials of two related compounds that failed due to animal and human liver complications. There was a “willingness of the corporation to take these risks,” says Robert Zamboni, one of Singulair's discoverers and now the vice president of medicinal chemistry at the company's Merck Frosst research center just outside Montreal in Canada. “These days, it wouldn't happen.”

    Kennis, now partly retired from Johnson & Johnson, says that nowadays “management people give us a certain period of time to find a new compound. … You cannot spend 10 years on something [for] which you don't know the outcome.”

    Facing doubters

    The current skittishness about a drug's prospects might have sunk today's top-selling class of drugs, the statins. “There was no consensus then, as there is today, that lowering cholesterol [is] good,” says Jonathan Tobert, who oversaw statin trials of Mevacor and Zocor at Merck and is now a consultant. Pursuing anticholesterol drugs, he says, “was a high-risk proposition.”

    In the 1960s, an anticholesterol drug from Marion Merrill Dow was withdrawn from the market after it was found to cause cataracts and other problems. Despite this troubled history, Merck stuck with a cholesterol program it had begun around that time, and Alberts was assigned to it after joining the company in 1975 as an expert on lipid synthesis. His chief competitor was Japan's Akira Endo, who worked at Sankyo; Endo had made breakthrough discoveries in cholesterol metabolism.

    Alberts had little idea where to find anticholesterol compounds. Based in part on what he knew of Sankyo's work, he turned to extracts from fungus, which had been rejected by a Merck lab in Spain seeking new antimicrobials and contained a jumble of products made by the organism. In the first week, 17 samples arrived in the lab and failed to blunt a preselected cholesterol target; the 18th, which reached Alberts's group a week later, worked so well that Julie Chen, running tests alongside Alberts, was convinced she'd erred. She hadn't. The extract that became Mevacor surfaced after just 2 weeks in 1978.

    That was the easy part. Despite the company's research in the area, many at Merck were leery of cholesterol drugs. A Merck medication had recently failed in clinical trials, says Alberts, and people from that project “had become very negative.”

    But back then, “a few people who really felt strongly about something could have quite an impact” on company decisions, says Tobert. And so, despite differences of opinion over its fate, the compound moved forward. It suffered a near-fatal setback in the fall of 1980, when a Sankyo cholesterol trial was abruptly halted due to safety issues in animals. Sankyo was quiet about what those were, but Merck's compound was similar enough to Sankyo's that its trials, too, stopped. Three years of analysis on Mevacor, and much “agitating” by Tobert, Alberts, and others, sent the drug back to the clinic. After a cataract scare in dogs, it was approved in 1987.

    Mevacor and its successor Zocor, a slightly modified version of its parent in which Alberts played a lesser role, have brought in tens of billions of dollars for Merck. But Tobert thinks that Mevacor might have been “shot down” had it emerged more recently. In the 1980s, he says, “somebody [who felt] strongly about something” and was “nowhere near a vice president… [was] heard.” The company was also willing, if sometimes reluctantly, to buck traditional wisdom. Part of the change, Tobert thinks, is a result of steadily growing research and corporate staffs and the accompanying bureaucracy.

    When the statin Lipitor was discovered in the mid-1980s, its backers faced a different challenge. Lipitor would be the fifth statin on the market, and in animals “it looked a little better [than existing statins], but not a lot, not enough to make a couple-million-dollar bet on,” says Pfizer's Roth, then at Parke-Davis. Parke-Davis's marketing department predicted that Lipitor would earn no more than $350 million a year.

    The company's chair of atherosclerosis and its head of research thought the drug deserved a chance, however, and coaxed senior management into funding a short-term clinical trial of healthy employees. The drug appeared to be more effective than any other statin. Today, by a $6 billion margin, it's the world's best-selling drug—and Roth, now a vice president of chemistry, swallows his own invention daily.

    Reflecting on how drug discovery has changed since the days of Lipitor, Roth mourns the luxury of time that's been lost. But he also sees important enhancements. Regulatory hurdles mean that drugs developed today are safer, more selective, and more potent, he believes. Back then, “we didn't understand the science”—evinced by Parke-Davis's surprise at how well Lipitor performed in people. A drug's biology is still often baffling, but its behavior in humans is better understood, scientists agree.

    Changing times.

    In 40 years, Pfizer's revenues have grown 100-fold, and the company now relies far more heavily on the popularity of individual drugs—in particular, its $12-billion-a-year Lipitor.


    Tobert, however, perceives a more fundamental breakdown in the drug discovery process. “A company's got to be prepared to take some risks,” he says, risks that must be balanced against possible harm to patients. Public fears about drug safety are widespread these days, driven in part by recent revelations that blockbusters are not free of serious hazards: best-selling selective serotonin reuptake inhibitors such as Zoloft can trigger suicidality, and COX-2 inhibitors such as Celebrex and Vioxx increase the risk of heart attacks.

    No financial windfall

    Even if their drug is later roiled in controversy—an experience that Celebrex inventor Talley says he's found deeply distressing—scientists take enormous pride in their creations. Nearly all those interviewed recalled being told, in their early days as industry scientists, that few drug hunters ever find one that makes it to market. “Virtually everything we do fails,” says Roth. Success, if it comes, tastes sweet.

    But it doesn't fatten the discoverer's wallet. At best, scientists receive some stock options as a reward for their work. Although companies can informally thank researchers, such as with pay raises or promotions or internal awards, they do not normally offer drug discoverers a revenue share or a substantial cash reward once the drug reaches the market.

    It's a policy many would like to see changed. Talley, who in addition to Celebrex discovered the related drug Bextra (which was recently removed from the U.S. and European markets due to safety concerns), struggled to pay college tuition for his two daughters while the therapies he created brought in billions for the company. “The ideal … would be to get royalties,” says Alberts, who discovered Mevacor. “I don't know how to do it, but I think there should be a better way.”

    Not everyone agrees. Roth and some others say the current setup—in which companies suffer the risks and reap the benefits, and scientists enjoy a steady salary whether they hit on a blockbuster or not—may be the fairest. Monetary rewards could “create an enormous competition [between] people internally,” says Kennis, a competition he believes would be unhealthy. Pharmaceutical giants such as Pfizer, AstraZeneca, and GlaxoSmithKline say they do not pay discoverers of new drugs for their finds. It's not something his pharmaceutical industry group has been asked to consider, says Jeffrey Trewhitt, spokesperson for the Pharmaceutical Research and Manufacturers of America in Washington, D.C.

    In the end, it's not lack of financial reward that drives the occasional inventor to leave the place where his discovery was made. Rather, what most bothers these scientists is how drug discovery has evolved. Company management increasingly favors “the short-term developmental route” rather than investing in projects more speculative in nature, says Craig Smith, the co-inventor, with Raymond Goodwin, of the rheumatoid arthritis drug Enbrel. The pair discovered Enbrel while at the Seattle, Washington, biotech Immunex, initially working on the side when their bosses “weren't looking,” says Smith. Both left after Immunex was bought for $16 billion by biotech giant Amgen in 2001.

    Smith predicts that drugmakers who avoid untrodden territory, based on inquiries that may not lead anywhere, will hit a wall down the road. “In the longer run,” he says, “you're shooting yourself in the foot.”

  21. Saving the Mind Faces High Hurdles

    1. John Travis

    Fierce competition to find a drug that could delay onset of or prevent Alzheimer's disease is a relatively recent phenomenon. Why was this potential blockbuster shunned for so long?

    Cancer has been arguably the most feared disease in the United States for the past several decades. Now, as the baby boom generation starts to inch past middle age, a new contender has emerged for that unappealing label: Alzheimer's disease (AD).

    An estimated 4.5 million Americans already have the neurodegenerative condition, and that number could more than triple by 2050. Devastating to both those afflicted and their caregivers, the illness exerts a $100-billion-a-year drain on the U.S. economy, according to the Biotechnology Industry Organization. “Alzheimer's disease probably has a larger impact on society than any other disease, in terms of economic and emotional costs,” says Dale Schenk, chief scientific officer at Elan, a biotechnology company based in Dublin, Ireland.

    So when Science looked for a condition that illustrates the challenges confronting the pharmaceutical industry—and the opportunities that beckon—AD was an obvious candidate. A drug that slows the disease could be especially lucrative because it presumably would need to be taken well before the first symptoms are likely to appear, and then for life. “Everyone recognizes that this is a great, unmet medical need. The drug company that succeeds here will be a very successful company,” says Peter Boxer, associate director of central nervous system (CNS) pharmacology at Pfizer Global Research and Development in Ann Arbor, Michigan.

    That recognition is fairly new, however. Academic and federal scientists had to lobby hard in the late 1980s to get Parke-Davis to conduct the first major clinical trial of an Alzheimer's drug. Although that drug, tacrine, and related compounds known as acetycholinesterase inhibitors rang up about $3 billion in sales for AD therapy in 2003, they are less than ideal medicines. They don't halt the underlying progression of the disease, and their slowing of cognitive decline is temporary.

    But even a less-than-perfect AD drug could still be a blockbuster for companies. It would also be a boon for society: Because the prevalence of Alzheimer's disease increases exponentially with age, drugs that provide a modest 5-year delay in the onset of symptoms would reduce the number of affected people by as much as 50%.

    Close look.

    By studying brain tissue from people who had Alzheimer's disease, drug companies are racing to develop drugs that slow or treat the disorder.


    Industrial nihilism

    Although drug development for AD is a relatively young endeavor, the condition was identified nearly a century ago by German neuropathologist and psychiatrist Alois Alzheimer. In 1906, he gave a lecture on a 51-year-old woman who had died with dementia. An autopsy found that her brain was littered with extracellular masses (plaques) and intracellular clumps (neurofibrillary tangles) that have since become the diagnostic hallmark of the disease that now bears his name. But for decades, because it wasn't diagnosable until after death, AD remained an obscure condition, and study of the illness was a scientific backwater. “No one wanted to get into [AD research] because it was seen as an unpromising career path leading to a scientific dead end,” recalls Zaven Khachaturian, former director of the Office of Alzheimer's Disease Research at the National Institute on Aging (NIA).

    The same pessimism about AD held true in industry. “There was very little interest because the disease could not be diagnosed, and the prevailing wisdom considered it an untreatable normal consequence of aging,” says Khachaturian. The lack of a cause was equally stifling to drug development. “There was a nihilism around [AD],” says neuroscientist Geoff Dunbar, who has worked on CNS drugs at several major companies and is now at a small biotech firm, Targacept, in Winston-Salem, North Carolina. “No one knew what to do with the plaques and tangles.”

    In the absence of hard evidence, a few vague theories took root. Some researchers argued that the dementia in general stemmed from inadequate blood flow within the brain, giving a slight boost to a class of drugs called cerebral vasodilators. Similarly, compounds that promoted learning and memory in animals—drugs known as nootropics, which means “growing the mind”—were also suggested as dementia treatments. “The assumption was that that would be sufficient to help the deficits in Alzheimer's disease,” recalls Boxer.

    The scientific hook

    Drug development for AD didn't truly get started until the cholinergic hypothesis emerged in the late 1970s, largely through the efforts of British neuroscientists such as Peter Davies, now at Albert Einstein College of Medicine in New York City. In 1976, for example, he and a colleague reported that compared to normal brains, those from several people who had had the brain disorder had decreased levels of an enzyme that helps make the neuro-transmitter acetylcholine. Those data, combined with earlier evidence that drugs blocking the cholinergic system produced memory problems in people, led Davies and others to argue that the core defect in AD was a lack of acetylcholine.

    Double trouble.

    The pathological hallmarks of Alzheimer's disease are extracellular brain deposits of β amyloid called plaques (large blue oval in corner) and intracellular clumps of tau known as tangles (smaller blue masses).


    “Until that time, dementia was primarily looked at as an amorphous mental disorder,” says Khachaturian. “The cholinergic hypothesis was the first scientific hook that could provide a clear path to understand the underlying neurochemistry of AD. It also gave us a plausible scientific rationale for developing treatments because so much was known about the cholinergic system.” That knowledge, says Dunbar, “meant we were in neuropharmacology that the industry understood.”

    There was also an obvious therapeutic road map to follow. It drew from work a decade earlier showing that the symptoms of Parkinson's disease stemmed from the death of dopamineproducing neurons and that L-dopa, a dopamine precursor, could bring about miraculous recoveries in patients. Could curing AD, researchers asked, be as simple as replacing acetylcholine?

    Not quite. Efforts to deliver acetylcholine precursors to the brain met with little success. In 1986, however, a different strategy grabbed the spotlight. A research team reported remarkable benefits for a few AD patients taking the well-studied compound oral tetrahydroaminoacridine, also called tacrine, which blocks the activity of an enzyme that breaks down acetylcholine.

    Quickly deciding to push for a validation study on the efficacy of tacrine, Khachaturian and the directors of the recently created, NIA-funded network of Alzheimer's Disease Research Centers sought a company to formulate the compound, which was off-patent, into various doses and quantities needed for a full-scale trial. They found an advocate in Elkan Gamzu at Parke-Davis. “Having a person inside that company lobbying for an efficacy study was very important to getting that first drug to go,” says Khachaturian.

    Parke-Davis, a division of Warner-Lambert Co. that later became part of Pfizer, started its tacrine study in 1987. But the drug, marketed as Cognex, failed to pass muster with a Food and Drug Administration (FDA) advisory board in 1991. After further trials with higher doses, the drug won FDA approval in 1993, albeit not without controversy. “The scientific community was not very enthusiastic about it because the benefits were marginal and it had a lot of side effects,” says Khachaturian.

    Still, its approval validated the cognitive tests that had recently been developed to gauge drug efficacy for the disease and provided clear guidelines on how to conduct clinical trials for AD. “If the FDA had set the bar very high and not approved it, then that would have been the kiss of death. No other company would have gotten into developing [AD] therapies,” says Khachaturian. “Once tacrine was approved, a lot of other companies jumped on the bandwagon” to develop safer and more potent acetylcholinesterase inhibitors, notes Boxer. (Six of the seven drugs currently approved in the U.S. for AD are in this class.)

    Keep apart?

    Now in phase III study, the drug Alzhemed blocks proteoglycan molecules from helping β amyloid form fibrils.


    The quick follow-up to tacrine by other drugs targeting the same enzyme illustrates an important principle of drug development. Even before a company with a head start on a target proves the value of a class of drugs, other firms will generally have similarly acting “me, too” drugs with improved properties in their pipeline. For competition's sake, says Boxer, “you can't wait for other companies' clinical data.”

    The acetylcholinesterase inhibitors spurred research into other ways of tweaking the cholinergic system. Acetylcholine operates through two classes of receptors, muscarinic and nicotinic, and major pharmaceutical companies vigorously pursued muscarinic agonists until troublesome side effects slowed their development. “Big pharma is still plugging away at the muscarinic hypothesis,” says Dunbar. That has left room for his current firm, Targacept, to develop AD drugs that target nicotinic receptors.

    The amyloid hypothesis

    Still, halting the decline of the cholinergic system in AD is not the same as curing, preventing, or even slowing the actual pathology of the illness. In fact, the benefits of acetycholinesterase inhibitors are so questionable that a government panel evaluating drugs for the U.K. health care system recently issued a preliminary opinion that the drugs aren't worth buying, a viewpoint the makers of the drugs have strongly challenged.

    Most companies seeking more fundamental treatments for AD are focusing on a protein fragment called β amyloid, which in 1984 was shown to be the primary component of the brain's plaques. That discovery spawned the amyloid hypothesis, which holds that the buildup of β amyloid causes AD by harming or killing brain cells. In 1991, scientists found that several families plagued by an early-onset form of AD had mutations in the gene encoding β amyloid precursor protein (APP), from which β amyloid is derived. A few years later, similar disease-causing mutations were found in genes encoding proteins called presenilins that were subsequently shown to affect APP processing into β amyloid.

    The amyloid hypothesis provided a bounty of new targets and potential strategies. Some companies tried to prevent β-amyloid molecules from clumping together, for example, while others began testing whether known drugs, such as statins and nonsteroidal anti-inflammatories, alter β-amyloid production.

    The novel hypothesis opened the door for small biotech companies, too. Neurochem, which was founded in 1993 in Laval, Canada, drew upon research licensed from Queen's University in Kingston regarding proteoglycan molecules in the brain that bind to β amyloid and promote the formation of the amyloid fibrils that make up plaques. The company has developed small organic molecules that mimic these proteoglycans, occupying their binding sites on β amyloid and preventing fibrils. Earlier this year, Neurochem launched a phase III trial of its lead Alzheimer's treatment, Alzhemed, seeking to become the first to bring an amyloid-modifying drug to market.

    Few firms are trying to directly block β-amyloid molecules from aggregating, notes Dennis Garceau, senior vice president of drug development at Neurochem. “Big companies like to target enzymes; it's a more conventional target,” he says. Indeed, the fiercest competition has been to develop secretase inhibitors, compounds that block the enzymes that cut APP into the smaller β-amyloid fragment.

    The race began in 1999 when a β secretase that acts upon APP was identified. (After the initial published report by Amgen, several other firms quickly revealed that they too had identified the same potential β secretase, perhaps setting the stage for a patent fight.) “Everyone went after that target right away. It was such a rational target,” says Boxer, who recalls hearing that another company had launched a major effort to inhibit the enzyme within a week of the announcement of its discovery.

    The identified β secretase was a particularly inviting target because it belonged to the same family of enzymes as HIV's protease. Several protease inhibitors had already been approved as AIDS drugs, allowing companies to draw on those experiences.

    It takes two cuts to make β amyloid out of APP, however. Drug companies weren't ignoring the other key enzyme, γ secretase, but they just weren't clear what it was. A theory that presenilins were γ secretases took several years to be accepted after its 1999 proposal. Still, even without a clear identification of the enzyme, several firms had developed in vitro systems displaying γ-secretase activity upon which they could test potential inhibitors.

    PET project.

    Using a compound developed at the University of Pittsburgh, researchers can now use PET scans to image the amount of β amyloid in brains of people with (left) or without (right) Alzheimer's disease. Such an ability could help drug companies monitor whether a drug is helping a person with Alzheimer's disease.


    Current efforts to develop secretase inhibitors remain shrouded in corporate secrecy. Bristol-Myers Squibb reportedly began clinical testing a γ-secretase inhibitor in 2001 and stopped because of side effects, but it has never publicly reported those results. Eli Lilly has also just begun clinical testing of a γ-secretase inhibitor. The challenge in developing such drugs seems to be blocking their action on enzymes needed for activities other than cutting up APP. γ secretases also cleave a protein called Notch, for example, that's important in development and the immune system. As a result, companies must find compounds that more specifically affect APP processing.

    A surprise vaccine

    While the amyloid hypothesis has offered drug researchers a number of obvious targets and strategies, it also led to the most surprising attempt to thwart AD. In the late 1990s, long after his colleagues at Elan had tested their most promising compounds, Schenk suggested injecting a few mice with β amyloid itself. His goal was to raise an antibody or other immune response against plaques. “No one thought it would work. Even after the experiment was done, the results weren't analyzed for a while,” recalls Schenk.

    The results were stunning. The immunization slowed or prevented the development of β-amyloid plaques in young mice and even wiped away preexisting ones in older mice. The episode illustrates how one person's idea can change the direction of a company or a field. “Dale was really brave,” says John Trojanowski of the University of Pennsylvania School of Medicine in Philadelphia.

    How does big pharma react when a disease-treating strategy such as the Elan vaccine comes out of the blue? Most large companies working on CNS drugs have experience with small-molecule drugs, not antibodies, says Boxer. And although firms can always tweak an enzyme inhibitor to make a better drug and carve out some market share, vaccines tend to either work or not. “We look at this stuff and go, 'Huh?'” says Boxer. “Where is your unique drug?” As result, he says, most companies have conceded the vaccine approach to Elan.

    The unexpected emergence of the Elan vaccine illustrates the importance and limitations of animal models. For several years, companies pursuing the amyloid hypothesis were largely stuck in vitro. Attempts to genetically engineer mice that overproduced APP seemed fruitless; there was even a notable fraud case in which a researcher published a picture of a human plaque as evidence that his mice had developed β-amyloid clumps. “The entire field was trying to make a mouse model,” says Schenk.

    Then a failing biotech company trying to sell off its assets approached Elan and saved the day, ultimately setting the stage for the vaccine's proof of principle. The struggling company's transgenic rodents were greatly overexpressing APP, and when Elan scientists checked out the mice, they found numerous brain plaques. Elan acquired the rights to the mice and quickly began testing its compounds. The company eventually allowed the β-amyloid vaccine strategy to be tested. Without that animal model, the idea might have faded away.

    Having animal models reduces the risk, and thus the cost, of developing drugs. For small companies such as Neurochem, they can also be a lifeline to continued funding from venture capitalists and other sources. “Until we got proof of concept in vivo, people were a little bit skeptical,” says Garceau.

    Going up.

    A number of factors influence predictions of how many Americans will have Alzheimer's disease in the coming years, but all such estimates suggest a rapid increase.

    CREDIT: HERBERT ET AL., ARCH. NEUROL. 60, 1119-1122 (2003).

    Yet animal models also reveal the risks of drug development. Elan's vaccine approach seemed to work well in mice, but brain inflammation in a few patients triggered an abrupt halt to the clinical trial. Elan, together with its partner Wyeth, is now conducting clinical trials with plaque-targeting immunotherapy strategies such as passive administration of antibodies to β amyloid.

    But how can a company pursuing β amyloid-based therapies for AD know if its drug or treatment is working? Showing that people maintain the same cognitive and memory skills, or improve such skills, can be difficult and time-consuming. Unfortunately, there are no well-accepted AD biomarkers, like cholesterol levels for heart disease or viral load for AIDS. A lack of animal models and biomarkers are “two difficult issues for developing a drug,” says Boxer.

    The biomarker obstacle has led companies such as Pfizer, Merck, Eli Lilly, and Elan to partner with the Alzheimer's Association, NIA, the National Institute of Biomedical Imaging and Bioengineering, and FDA to identify ways of measuring progression of mild cognitive impairment and AD in people. Industry will pick up one-third of the cost of the $60 million, 5-year effort, known as the Alzheimer's Disease Neuroimaging Initiative, that will test various ways of imaging brain plaques and tangles as well as measuring levels of proteins in blood, urine, and cerebrospinal fluid. “It's so difficult [to develop an Alzheimer's treatment without biomarkers] that the drug companies are collaborating,” says Boxer.

    What about tau?

    It's sometimes forgotten that the effort to develop β amyloid-based treatments represents a huge and costly gamble on a single, unverified theory of AD. There are many other hypotheses being explored by small numbers of scientists or a handful of tiny biotech firms. One is the second major theory of AD, which involves tangles, the intracellular brain lesions identified by Alois Alzheimer.

    In the early days, Alzheimer's researchers were divided over whether plaques or tangles were more important. The identification of β amyloid in plaques and disease-causing mutations in the APP gene relegated tangles and their primary constituent, a hyperphosphorylated form of a protein called tau, to a sideshow. “We were the token other pathway at every meeting,” recalls Trojanowski; he and his wife Virginia Lee have been the most vocal proponents of tangles and tau research.

    Surprise shot.

    Mice genetically engineered to overproduceβ amyloid develop brain deposits (a, b) similar to the plaques in Alzheimer's disease, but injecting such rodents with β amyloid stirs an immune response that can clear such deposits (c, d).

    CREDIT: NATURE 400, 173-177 (1999) ARCH. NEUROL. 60, 1119-1122 (2003).

    For companies, that lack of interest was partly a matter of simple economics. “Even big pharma can only pick a certain number of targets,” says Dunbar, noting that Bristol-Myers Squibb, where he used to direct clinical development of CNS drugs, has never had a tau program to his knowledge.

    Tau is now drawing more attention, in part because of a 1998 paper in which researchers showed that mutations in a gene encoding one of the human versions of tau lead to a rare form of dementia that bears some similarities, such as tau tangles, to AD. “It launched studies that should have been done in the early 1990s,” says Trojanowski.

    Trojanowski contends that tau, when it becomes overloaded with phosphate groups, can no longer bind to and stabilize cellular filaments called microtubules. That change disrupts the ability of neurons to transport molecules down the long extensions known as axons. Back in 1994, his team proposed that microtubule-stabilizing compounds, such as the cancer drug Taxol, might treat AD. And earlier this year, in the 4 January Proceedings of the National Academy of Sciences, they offered a proof of concept in mice genetically engineered to overproduce a human version of tau.

    These rodents suffer from a neurodegenerative disorder that includes tanglelike masses of hyperphosphorylated tau and impaired axon function. As hypothesized, the administration of Taxol sped up the animals' axonal transport and ameliorated their motor problems. Trojanowksi and his colleagues are now working with Angiotech Pharmaceuticals in Vancouver, British Columbia, and other firms are sniffing around. “I know pharma is interested,” he says. “My phone rings more often.”

    Partnerships and future

    Will the next significant drug for AD come from a small biotech company or big pharma? Given the economics of drug development, it's likely that the Davids and Goliaths will end up working together.

    “It's very difficult for a small company to take a drug all the way to market,” notes Targacept's Dunbar. His company's strategy, for example, is to push a drug only through phase II trials and then “outsource it to big pharma.” And Neurochem says it would be open to partnerships with bigger companies given the right deal.

    Big pharma is certainly happy to let smaller companies take the initial plunge before it swoops in and buys up a promising drug. “They have such big wallets they can wait until almost all the risk is taken out,” says Dunbar.

    Still, the search for Alzheimer's drugs should leave room for many companies, small and large, to prosper. “This disease will need a cocktail of treatments,” predicts Neurochem's Garceau.

  22. Pharma Moves Ahead Cautiously in China

    1. Yidong Gong*
    1. Gong Yidong writes for China Features in Beijing.

    Companies can't resist the lure of China. But full-service research labs remain on the horizon

    SHANGHAI—As recently as 5 years ago, China was terra incognito for big pharma research organizations. To be sure, the global drug giants have been selling their products in China since the 1980s, and quite a few have built manufacturing plants there. Yet concerns over enforcement of the country's fledgling laws governing intellectual property rights (IPR) had prevented companies from taking the logical next step: opening a lab to do drug discovery.

    Those qualms remain. But they are balanced by the industry's growing desire to dip into China's intellectual talent pool. In 2002, Novo Nordisk broke from the pack and set up a small research facility in Beijing, the company's only research site outside its home in Denmark. Later that year, U.K.-based AstraZeneca set up the first Western-owned clinical research organization in China to collaborate on multisite trials. The next year, U.S.-based Eli Lilly inked a deal with the Chinese company ChemExplorer to purify, synthesize, and analyze compounds supplied by its researchers. And last fall, when the Swiss-based Roche dedicated its new research and development lab in Shanghai, Roche Chair and CEO Franz Humer predicted that China will “someday [be] one of Roche's important R&D centers rather than a mere market and production base.”

    Humer may well be right. And yet, his open-ended time reference sent the subtle but unmistakable message that big pharma still harbors doubts about China's ability to protect any valuable intellectual property that a company might create within its borders. Indeed, the research director of the Novo Nordisk site, Wang Baoping, concedes that his company is taking a risk. “China has in place a series of laws related to IPR protection. But their enforcement, particularly the amount that would be paid to the damaged side, remains a problem that needs to be addressed,” says Wang, a U.S.-trained geneticist. “I am not sure when the IPR environment in China will be truly favorable.”

    China's growing appetite for Western drugs—the current $15 billion market is expected to quadruple by 2010, and then double again by 2020—has certainly caught the attention of every drug company. So has its cheap but skilled scientific labor force. Not only do Ph.D.s receive annual salaries of $10,000 or less, but the most expensive aspect of drug development—clinical trials—costs an estimated 30% less in China than in the United States or Europe. And then there is its growing prowess in science. “I'd say that setting up our own research lab there is only a matter of time,” Novartis CEO Dan Vasella remarked this spring. “It's not so much a need as it is a hunger to take advantage of the opportunities.”

    Despite those inducements, the research centers being set up are shadows of pharma's existing full-service shops in the West. The Novo Nordisk and Roche labs are much smaller—some 40 to 50 scientists—and narrower in focus, typically medicinal chemists and biologists. But company officials still hope that the labs can make big contributions. Roche's Chen Li, chief scientific officer for the Shanghai site, says the key to its success in medicinal chemistry will be “giving full play to the initiatives of the scientists here, including access to information” throughout Roche's global research network. At Novo Nordisk's Beijing lab, scientists focus on protein expression to supplement the company's portfolio of diabetes drugs.

    First steps.

    Chinese scientists at work in the Novo Nordisk R&D Center in Beijing.


    Lorenz K. Ng, vice president of research alliance and business development for Lilly Asia, says, “We looked at China because of its good supply of chemists.” Its partner Chem-Explorer has a team of 175 chemists. Still, as one ChemExplorer scientist notes, “What we do is only one piece of the core technology of drug development.” AstraZeneca's clinical research unit focuses on another piece of drug development by taking advantage of lower costs and access to a different population. The unit has been involved in six multicenter trials, totaling 765 patients. And Pfizer China is recruiting biometricians to staff a clinical trials data management center that it hopes to open early next year to help the company crunch the numbers from trials already under way.

    Most Chinese scientists believe the arrival of Western pharmaceutical companies will be a long-term benefit for the country. “The biggest contribution of foreign research centers is the opportunity to learn drug development. That's a huge gap that needs to be filled in China,” says Hu Zhuohan, a professor of pharmacology at Fudan University in Shanghai.

    But a few observers worry that the trend will stifle China's own efforts. “Research and development is one of the least profitable links of the pharmaceutical chain,” says Zhang Hua, a financial analyst in east China's Shandong Province, who predicts that small, young local drug companies “will inevitably fall prey to foreign companies.” Even so, Zhang thinks the process is irreversible, and he holds out hope that Chinese companies will learn drug innovation more quickly by watching it firsthand